Tectonic Evolution of Southeast Asia
Geological Society Special Publications Series Editor
A. J. FLEET
GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 106
Tectonic Evolution of Southeast Asia EDITED BY
ROBERT HALL SE Asia Research Group Department of Geology Royal Holloway London University Surrey, UK AND
DEREK BLUNDELL SE Asia Research Group Department of Geology Royal Holloway London University Surrey, UK
1996
Published by The Geological Society London
THE GEOLOGICAL SOCIETY The Society was founded in 1807 as the Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of 8000. It has countrywide coverage and approximately 1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house, which produces the Society's international journals, books and maps, and which acts as the European distributor for publications of the American Association of Petroleum Geologists, SEPM and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C Geol (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society, Burlington House, Piccadilly, London W1V 0JU, UK. The Society is a Registered Charity, No. 210161
Published by the Geological Society from: The Geological Society Publishing House Unit 7 Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN UK (Orders: Tel 01225 445046 Fax 01225 442836) First published 1996 Reprinted 1997 The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. © The Geological Society 1996. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/95/$07.00.
British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 1-897799-52-7 Typeset by EJS Chemical Composition, Midsomer Norton, Bath BA3 4BQ, UK Printed by The Alden Press, Osney Mead, Oxford, UK
Distributors USA AAPG Bookstore PO Box 979 Tulsa OK 74101-0979 USA (Orders: Tel (918) 584-2555 Fax (918) 548-0469) Australia Australian Mineral Foundation 63 Conyngham Street Glenside South Australia 5075 Australia (Orders: Tel (08) 379-0444 Fax (08) 379-4634) India Affiliated East-West Press PVT Ltd G-l/16 Ansari Road New Delhi 110 002 India (Orders: Tel (11) 327-9113 Fax (11) 326-0538) Japan Kanda Book Trading Co. Tanikawa Building 3-2 Kanda Surugadai Chiyoda-Ku Tokyo 101 Japan (Orders: Tel (03) 3255-3497 Fax (03) 3255-3495)
Contents HALL, R. & BLUNDELL,D. J. Tectonic evolution of SE Asia: introduction
vii
Part 1: Present-day tectonics MCCAFFREY, R. Slip partitioning at convergent plate boundaries of SE Asia
3
MALOD, J. A. & KEMAL, B. M. The Sumatra margin: oblique subduction and lateral displacement of the accretionary prism
19
RANGIN, C., DAHRIN, D., QUEBRAL,R. & THE MODEC SCIENTIFICPARTY. Collision and strike-slip faulting in the northern Molucca Sea (Philippines and Indonesia): preliminary results of a morphotectonic study
29
RICHARDSON,A. N. & BLUNDELL,D. J. Continental collision in the Banda arc
47
SNYDER,D. B., MILSOM,J. & PRASETYO,H. Geophysical evidence for local indentor tectonics in the Banda arc east of Timor
61
HUGHES, B. D., BAXTER, K., CLARK,R. A. & SNYDER,D. B. Detailed processing of seismic reflection data from the frontal part of the Timor trough accretionary wedge, eastern Indonesia
75
MILSOM, J., KAYE,S. & SARDJONO.Extension, collision and curvature in the eastern Banda arc
85
Part 2: Tectonic development of Southeast Asia METCALFE, I. Pre-Cretaceous evolution of SE Asian terranes
97
PACrd-IAM, G. Cenozoic SE Asia: reconstructing its aggregation and reorganization
123
HALL, R. Reconstructing Cenozoic SE Asia
153
SIMANDJUNTAK,T. O. & BARBER,A. J. Contrasting tectonic styles in the Neogene orogenic belts of Indonesia
185
RICHTER, B. & FULLER, M. Palaeomagnetism of the Sibumasu and Indochina blocks: implications for the extrusion tectonic model
203
STOKES, R. B., LOVATTSMITH,P. E & SOUMPHONPHAKDY,K. Timing of the Shan-Thai-Indochina collision: new evidence from the Pak Lay Foldbelt of the Lao PDR
225
LOVATr SMITH,P. F., STOKES,R. B., BRISTOW,C. & CARTER,A. Mid-Cretaceous inversion in the Northern Khorat Plateau of Lao PDR and Thailand
233
HtrrCHISON, C. S. The 'Rajang accretionary prism' and 'Lupar Line' problem of Borneo
247
OMANG, S. A. K. • BARBER,m. J. Origin and tectonic significance of the metamorphic rocks associated with the Darvel Bay Ophiolite, Sabah, Malaysia
263
NGAH, K., MADON,M. & TJIA, H. D. Role of pre-Tertiary fracture in formation and development of the Malay and Penyu basins
281
TJIA, H. D. & LIEW, K. K. Changes in tectonic stress field in northern Sunda Shelf basins
291
CLENYELL, B. Far-field and gravity tectonics in Miocene basins of Sabah, Malaysia
307
McCOURT, W. J., CROW, M. J., COBBING,E. J. & AMIN, T. C. Mesozoic and Cenozoic plutonic evolution of SE Asia: evidence from Sumatra, Indonesia
321
SAMUEL, M. A. & HARBURV,N. A. The Mentawai fault zone and deformation of the Sumatran Forearc in the Nias area
337
WAKITA,K., SOPAHELUWAKAN,J., MIYAZAKI,K., ZULKARNAIN,I., & MUNASRI.Tectonic evolution of the Bantimala Complex, South Sulawesi, Indonesia
353
vi
CONTENTS
WILSON, M. E. J. & BOSENCE, D. W. J. The Tertiary evolution of South Sulawesi: a record in redeposited carbonates of the Tonasa Limestone Formation
365
BERGMAN, S. C., COFFIELD,D. Q., TALBOT, J. P. • GARRARD,R. A. Tertiary tectonic and magmatic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for a Miocene continent-continent collision
391
ALI, J. R., MILSOM, J., FINCH, E. M. & MUBROTO,B. SE Sundaland accretion: palaeomagnetic evidence of large Plio-Pleistocene thin-skin rotations in Buton
431
VROON, P. Z., VAN BERGEN, M. J. & FORDE, E. J. Pb and Nd isotope constraints on the provenance of tectonically dispersed continental fragments in east Indonesia
445
LINTHOUT, K., HELMERS, H., WIJBRANS,J. R. & VAN WEES, J. D. A. M. 4°Ar/a9Ar constraints on obduction of the Seram ultramafic complex: consequences for the evolution of the southern Banda Sea
455
CHARLTON, T. R. Correlation of the Salawati and Tomori Basins, eastern Indonesia: a constraint on left-lateral displacements of the Sorong fault zone
465
MALAIHOLLO,J. E A. & HALL, R. The geology and tectonic evolution of the Bacan region, east Indonesia
483
BAKER, S. & MALAIHOLLO,J. Dating of Neogene igneous rocks in the Halmahera region: arc initiation and development
499
PUBELLIER, M., QUEBRAL,R., AURELIO, M. & RANGIN, C. Docking and post-docking escape tectonics in the southern Philippines
511
CROWHURST, P. V., HILL, K. C., FOSTER, D. A. & BENNETT, A. R Thermochronological and geochemical constraints on the tectonic evolution of northern Papua New Guinea
525
WOPFNER, H. Gondwana origin of the Baoshan and Tengchong terranes of west Yunnan
539
ZI-IOU, Z., LAO, Q., CHEN, H., DING, S. & LIAO, Z. Early Mesozoic orogeny in Fujian, southeast China
549
Index
557
Tectonic evolution of SE Asia: introduction R O B E R T H A L L 1 & D. J. B L U N D E L L
SE Asia Research Group, Department of Geology Royal Holloway, University of London, Egham, Surrey TW20 OEX, UK 1 Fax no. (44) 01784 471780 Email
[email protected]
SE Asia is probably the finest natural geological laboratory in the world yet is still not geologically well known. It is a spectacular region in which the manifestations and processes of plate collision can be observed at present and in which their history is recorded. It is a region that must be understood if we are to understand mountain belts, arc development, marginal basin evolution and, more generally, the behaviour of the lithosphere in collision settings. Furthermore, the region is developing rapidly on the economic front, and a major part of this rapid development is built on natural resources. The geological reasons for the distribution of these resources are therefore of major importance for the inhabitants of the region and for attempts to discover and exploit them. These were some of the thoughts which stimulated the collection of papers (Fig. 1) in this volume. In order to understand the development of this complex region an essential first step is to identify the key features of the active tectonics and determine how plates and sub-plate lithospheric fragments are moving. How successfully can rigid plate tectonics be applied in describing present tectonics? Where are the boundaries between plates? What are the rates at which different parts of the region are moving? The first part of this volume includes a number of papers which deal with these questions, based upon the application of GPS (Global Positioning System) measurements to determining the nature and rates of plate movements and plate boundary zone deformation, results from the BIRPS deep seismic reflection experiment in the Banda arc, the first to cross a modern active margin, and other geophysical and geological studies. MeCaffrey provides a regional overview of recent GPS results and earthquake data bearing on the present plate tectonics. It is ironic that in such an active region the identification of several plates and determination of some important relative plate motions, critical to a full kinematic description, are still very uncertain. McCaffrey discusses the way in which motion is partitioned in obliquely convergent settings, almost the rule in SE Asia. Oblique convergence is commonly inferred in the past for SE Asia (as shown in many of the later papers in the
volume) and is often used as an explanatory tool in orogenic belts elsewhere in the world. He draws attention to the deformation of the upper plate in these convergent settings and emphasizes the importance of a three-dimensional understanding of the process. In this rapidly evolving region, information from slip vectors and geodetic measurements will allow a fourth dimension of time to be included as data accumulate, which will raise the important problem of whether and for how long the present motions can be assumed to extend back into the past. Two well-known areas of oblique convergence, the Sumatra and Philippine Sea plate margins, illustrate the realities of present tectonics, first in identifying small plates or smaller tectonic elements, and second in providing a kinematic description. The increase in obliquity of convergence between Java and north Sumatra has long been considered to result in thrusting normal to the subduction trench and arc-parallel movement on the Sumatran fault. However, as McCaffrey points out, this simple model does not predict some important features, such as subduction west of the Andaman Sea and differences in amounts of extension between north and south Sumatra. Malod & Kemal discuss evidence from marine surveys off Sumatra and propose that this area can be understood as a number of plate slivers between the trench and the Sumatran fault, with variations in the partitioning of movement in different parts of the Sunda forearc. This is explained as the result of differences in coupling between the subducting and overriding plates, possibly reflecting the presence of major structures on the subducting slab, notably an extinct spreading centre on the Indian plate. Further east, the Philippine fault and trench are also considered as the joint expression of partitioning of oblique convergence, but once again applying this simple model presents problems: how does this paired trench-fault system link southwards into the arc-arc collision of the Molucca Sea? Rangin et al. tackle the first important problem of what is actually present in the zone of transition by mapping structures from the south Philippines into Indonesia and determining which features belong
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. vii-xiii.
vii
viii
R. HALL t~z D. J. BLUNDELL
1110°
lo~
.30*
SOUTHcHINA
t/° ,, q
/1;~0°'
0
130°
140°
30~-
.-. !
,"
,,",'," -" 12 ".-,,,.; ,... 200..
" ~ A
China Sea
',.)~.. ',CAMBODIA ~ / 1 8 ~
PHILIPPINE
Pacific Ocean
-10"
N
-~
E
MALAYSIA,,
~MALAYSIA~
.o-
30 MOLUCCAS N ~ SULAWES,
.~
~
32
-
'),SUMATRA~ ~-'
Indian .lo*
~ :
JAVA
~ B ~ U
~
22 ~ , 24 - -
~
i~
Ocean
1O0°
110*
120*
BANDI.~o
oooj~----~
~
: GUINEA
J
.(-fl 4, 5, 61 1
AUSTRALIA
Fig. 1. Areas discussed in the papers in this volume. 1: McCaffrey; 2: Malod & Kemal; 3: Rangin et al.; 4: Richardson & Blundell; 5: Snyder et al.; 6: Hughes; 7: Milsom et al.; 8: Metcalfe; 9: Packham; 10: Hall; 11: Simandjuntak & Barber; 12: Richter & Fuller; 13: Stokes et al.; 14: Lovatt Smith et al.; 15: Hutchison; 16: Omang & Barber; 17: Ngah et al.; 18: Tjia & Liew; 19: Clennell; 20: Mccourt et al.; 21: Samuel & Harbury; 22: Wakita et al.; 23: Wilson & Bosence; 24: Bergman et al.; 25: Ali et al.; 26: Vroon et al.; 27: Linthout et al.; 28: Charlton; 29: Malaihollo & Hall; 30: Baker & Malaihollo; 31: Pubellier et al.; 32: Crowhurst et al.; 33: Wopfner; 34: Zhou et al.
to which plate, a demanding task. Their use of the technique of multibeam mapping, combined with geophysical evidence, is producing highly detailed maps, in many cases of much higher quality than those available for land areas nearby, and revealing unsuspected structures. The arguments surrounding the tectonics of the Banda arc, and some of the complexities of the collision zone, for example its curvature, age of the Banda Sea, and significance of deep basins such as the Weber trough, also illustrate the difficulties of linking simple kinematic models to reality. Very recently two BIRPS deep seismic reflection profiles have crossed the Banda arc and imaged the deep structure of the central part of the AustraliaSE Asia collision zone. Richardson & Blundell discuss how deep reflectors can be linked to recent seismicity, and their connection to structures on land. They identify two sets of divergent structures in the collision zone: a southern set dipping in the same direction as subduction is related to the subducted Australian margin, whereas a northern
set, in the upper plate, is antithetic, and more recent. They argue that these structures changed orientation and migrated north because buoyant continental crust blocked the subduction zone, although Australia continued to move north. Using the BIRPS images to estimate the volume of material in the collision zone they infer that an earlier Neogene collision event, involving a separate microcontinental fragment or outer margin high, must have preceded the present phase of collision which began at about 2.4 Ma. Snyder e t al. draw attention to unusual features of the BIRPS 'Timor' sector of the Banda orogenic zone, including a Bouguer anomaly and thin sediment cover implying a thicker continental crust than normal beneath the accretionary complex, a very narrow forearc, and presence of the volcano Gunung Api in the backarc. Like Richardson & Blundell, they infer the presence of additional Australian margin crust but suggest it formed either a promontory of the Australian margin or a prebreakup extensional basin within the former pas-
INTRODUCTION sive margin, now inverted and thickened. Also like Richardson and Blundell they deduce that the deformation has advanced across the collision zone, causing crustal fracturing in the backarc region and allowing magmatic uprise and underplating beneath Gunung Api. Hughes et al. have reprocessed part of the Timor section to bring out the details of the Timor trough and the accretionary wedge developed on its northern flank. Australian shelf sediments are clearly imaged, relatively undisturbed, and beneath a low-angle decollement above which are highly deformed sediments stacked in a complex set of thrust sheets indicative of the northward underthrusting of the Australian continent. However, a covering drape of sediment indicates that deformation in this area has been inactive for about 1 Ma. One feature that has bedevilled discussions of the Banda arc is the significance of the deep troughs which parallel the arc, which in a superficial way have characteristics of trenches, but when studied more carefully seem to be in the wrong position. Milsom et al. present evidence that the deformation front between Australia and the Banda arc extends between the Kai islands, in the region of maximum curvature of the arc, rather than following the apparent subduction-related feature of the Aru trough. They conclude that the arc must have been a continuous feature before collision with a curvature acquired recently. The present setting of the region is both the start and the end point in attempts to comprehend its development. The start, because an understanding of what is present provides the first clues of what has happened in the past few millions or tens of millions of years, leaving traces in the slabs beneath the arcs, the arcs themselves, and in the stratigraphic record of the arc regions as well as the Sunda continental shelf further north. The end, because the present is the result of what has happened in the past, and many of the present features of the region may be explicable only in terms of what has now disappeared and can only be inferred. Thus, our understanding of the region and its development is an iterative synthesis, each new step forward requiring a reconsideration of present and past. The second section of the volume therefore begins with three overview papers intended to synthesize earlier observations which may provoke new research improving and extending our present regional understanding. Metealfe reviews the pre-Cretaceous development of SE Asia and east Asia and shows that the region has grown by addition of allochthonous terranes which separated from different parts of Gondwana. Their northwards movement was accompanied by the opening and closing of three successive oceans: the Palaeo-, Meso- and
ix
Ceno-Tethys. Assembly began with the formation of Cathaysia in the Late Devonian-Early Carboniferous and its growth continued within the Palaeo-Tethys in the later Palaeozoic. As the Meso-Tethys opened by Late Carboniferous-Early Permian rifting of the northern margin of Gondwanaland so the Palaeo-Tethys began to close, and so subsequently did the opening of the Ceno-Tethys result in elimination of the MesoTethys. The final stage of the reassembly of Gondwana in Asia is not yet complete but the present complexity of the region, as well as the comprehensive grasp of so many disciplines in earth science required to attempt the task of describing its development, give an indication of Metcalfe's achievement in producing a comprehensible synthesis. A description is an essential step in identifying the driving processes and modelling the development of the region. Reconstructing SE Asia in the Cenozoic requires a description of India-Eurasia collision, the motion of the Philippine Sea plate and the present collision of Australia with eastern Indonesia. Eurasian extrusion models have proved popular with many workers in SE Asia, partly because of the striking similarities of plasticine models to tectonic maps, and partly because they offer a means to explain and link different events. As with many major advances, the extrusion model has also been effective in provoking other new ideas and the search for new evidence. Paekham considers the relationships between the observed geology and different models of the region, reviewing the timing of different events, the evidence from crustal volumes for extrusion, and the relationship between predicted and observed palaeomagnetic measurements. In east Asia and western SE Asia the estimates of rates and amounts of movements predicted by the extrusion model, and those that can be determined, are becoming much closer. The inadequacies of the fit cannot just be explained as weaknesses of the model but also emphasize the variable quality of data and their uncertainties. Packham concludes that a regional understanding of SE Asia requires a better understanding of the timing of deformation, especially uplift, in the Himalayas and central Asia, a clearer picture of whether rotations detected using palaeomagnetic data are regional or local, and much new stratigraphic information, particularly in the eastern part of the region. The key role of palaeomagnetic data in developing regional models is a theme discussed by Hall in an attempt to produce a kinematic model for the whole of Cenozoic SE Asia. At the centre of the region is the island of Borneo from which rotations recorded appear to be in conflict with those predicted by an indentation model. Further east, new
X
R. HALL • D. J. BLUNDELL
data from the Philippine Sea plate confirm longterm clockwise rotations of the whole plate suggested by many earlier studies. The attempt to synthesise these results suggests that, even if Cenozoic extrusion of continental fragments from east Asia is accepted, this has not been the most important driving force in the development of the marginal basins of eastern SE Asia. According to this model, development of marginal basins was linked physically and temporally, and opening appears to be mainly subduction-related rather than indentor-driven. This new model does suggest some possible configurations for the region which are different from those previously accepted and may provoke reconsideration of some evidence, especially the habit of looking for local explanations of tectonic phenomena. The animation which accompanies the paper reminds us that regional events may have causes outside the immediate area, and the manifestation of plate movement changes may propagate gradually across the whole region, as plate boundary changes in one area cause changes in others. Several aspects of the model, such as the proposed ages of some basins, for example the Banda Sea, remain to be tested by ocean drilling, palaeomagnetic work, and stratigraphic and structural studies. The remaining papers in the volume are arranged broadly in geographical order. Simandjuntak & Barber illustrate the variation in orogenic styles from different parts of Indonesia. The present diversity and complexity of tectonic processes in SE Asia may provide keys to the interpretation of other orogenic belts. The differences in the history of deformation within the region may leave traces other then geological structures and Richter & Fuller discuss the still thorny question of the implications of palaeomagnetic data from Sundaland and Indochina. They conclude that different parts of Sundaland and Indochina have deformed in different ways; some parts as small blocks with rotations indicating local deformation, some extrusion-driven rotation, principally in Indochina, and some parts recording the results of deformation dominated by the oblique subduction of the Indian plate. Distinguishing between such areas for the whole SE Asia region is an important task for geologists and palaeomagnetists for the future. Even the most carefully constructed regional kinematic models are totally dependent on basic data revealing the timing of tectonic events and the evidence for them. The papers concerned with areas around the South China Sea suggest the need for revisions of models as well as new interpretations of relationships between effects and supposed causes. In Laos, Stokes et al. argue that suturing of the Shan-Thai and Indochina blocks occurred in
the Late Jurassic and that the Indosinian Orogeny, currently assumed to be of Permian or Triassic age is significantly younger than commonly assumed. Based largely on seismic data, Lovatt Smith et al. suggest regional tilting, compressive folding, reverse faulting and basin inversion in Thailand record important phases of structural development which pre-date the currently assumed Tertiary age of structuring. If correct, these interpretations have implications for hydrocarbon exploration and potential. Hutchison draws attention to the contrast between tectonic models of the South China Sea region and the geology of Borneo and proposes that the term Rajang Group should be more carefully applied. An older turbidite sequence, assigned to the Rajang Group proper, represents an accretionary prism compressed and uplifted between the Schwaner Mountains volcano-plutonic arc and a South China Sea microcontinent during an Eocene orogeny. Similar but younger turbiditic rocks deformed by a Miocene orogeny are interpreted not as deposits of a forearc, but as derived from the eroding and uplifted Rajang Group, and should be separated from it. A further record of the late Mesozoic or early Tertiary subduction setting of the NE Borneo margin is to be found in the large Darvel Bay Ophiolite Complex of Sabah. Mineralogical and geochemical studies by Omang & B a r b e r suggest its formation in a suprasubduction zone environment, but with complexities due to high T-low P deformation along a transform fault. High P-T garnet pyroxenites and amphibolites found as clasts in Miocene rocks were derived from a metamorphic sole underlying the complex, formed during subduction and emplacement of the ophiolite. Within the Sunda shelf are sedimentary basins of the South China Sea and adjacent areas which record a link between east Asian tectonics and the plates beyond the subduction zones bounding SE Asia. The importance of pre-existing structures in controlling tectonic development is often forgotten and Ngah et al. suggest that the Malay, Penyu and West Natuna basins originated in the Late Cretaceous as three rift arms that developed during doming of continental crust above a mantle plume. The hydrocarbon potential of these basins was subsequently influenced by changes in stress patterns which Tjia & Liew argue resulted from the interplay of Eurasian extrusion driven by Indian indentation, and changes in directions and rates of motion of the plates in the Pacific and Indian Oceans. Borneo is situated in the middle of this region at the south side of the South China Sea and ought to record the effects of these changes. The geology of Sabah is therefore of considerable regional interest since, if Borneo has rotated, it is an area where the consequences should be most
INTRODUCTION obvious. Clennell discusses the interplay between large-scale regional plate motions (his 'far-field tectonics') and pre-existing structures and local tectonic influences for the development of the unusual circular basins of Sabah. These papers on sedimentary basins show that we are still some way from clearly linking local and regional tectonics. Tjia and colleagues show that there were reversals in the sense of movement on important faults, that the effects of fault movements differ from area to area, and that there is still uncertainty in the timing of fault movements. Clennell infers that basins in Sabah appear not to record some tectonic events because they were decoupled due to the thicknesses of underlying muds and m61anges. Sumatra is an area where a long history of subduction should be recognizable since the island is usually considered to have been situated above the northward-subducting Indian plate from at least the Mesozoic. Despite this there appear to be distinct periods characterized by igneous activity, separated by intervals with little or none. McCourt et al. use isotopic dating and geochemistry to identify plutonic episodes and their character, which they link to plutonism elsewhere in Sunda margin. Understanding the tectonic significance of the igneous episodes needs to be the next step forward. McCourt et al. speculate that variations in the obliquity of convergence, and collision of allochthonous terranes are implicated, although Neogene and younger strike-slip faulting complicates the picture. It would be useful to consider the Cenozoic history of this margin in the light of known plate motions, although the major uncertainty here is not the motion of the Indian plate but the orientation and position of Sumatra; different regional models show very different configurations for the early Tertiary. The Cenozoic history of this margin is clearly complex as indicated by other papers in the volume and Samuel & H a r b u r y show that this complexity is still far from understood. In their paper, based on detailed studies on land in the Sumatran forearc islands, principally Nias, they interpret the Mentawai fault system not as a strike-slip fault, but as an extensional structure with late contractional reactivation. If correct, at least one of the plate slivers of the forearc proposed by Malod & Kemal either does not exist or is of very recent origin. Samuel & Harbury's work also illustrates the importance of field-based studies in providing a firm stratigrapbic basis for the interpretation of other evidence, such as seismic and marine geophysical evidence, and they infer a long extensional history for the Sumatran forearc, with major extensional structures later reactivated, again with possible links to changes in plate motions such as the angle of convergence.
xi
A good stratigraphic base is fundamental to attempts to interpret the tectonic evolution of the region and Wakita et aL provide an example of how this is achieved by detailed radiolarian studies. The Bantimala Complex and Balangbaru Formation of south Sulawesi record critical events in the accretionary growth of the SE Sunda margin and dating based on micropalaeontological evidence is difficult to obtain and interpret since the turbidite sequences yield few fossils, and these may be reworked. Radiolaria, which can often be assigned to narrow zones, provide an additional means of comparing the ages of different lithological units and suggest the Bantimala Complex and Balangbaru Formation are contemporaneous, requiring a modification of previous tectonic interpretations. From the same region Wilson & Bosenee show how redeposited limestones of the Tonasa Limestone Formation can be used as indicators of tectonic activity. Detailed measured sections, well dated by fossils, illustrate how a very clear palaeogeographic picture can be deduced and linked to larger-scale tectonics. Their results provide a basis for regional interpretation and will be of considerable interest to those exploring for hydrocarbon in the frontier regions of east Indonesia. Sulawesi is currently much less wellknown than it deserves to be, especially considering its large size and critical position at the Eurasia-Australia-Pacific junction, and recent results have shown that simple tectonic models for its development need reconsideration. Bergman et al. present data from west Sulawesi which will need to be incorporated in new models and speculate on possible solutions. Of considerable interest is the evidence, based on isotopic studies, for a magmatic contribution from old Australian-type continental crust to the Tertiary plutonic rocks of west Sulawesi. Bergman et al. also focus attention on the structures around the Makassar Strait. This has previously been widely accepted to be an extensional basin but they interpret it as a foreland basin bounded by converging Neogene thrust belts, with the late Miocene western Sulawesi magmatic arc recording continent-continent collision. The collapse of the orogenic belt is seen as the cause of young extension in the region. The rapid changes predicted by the model are certainly consistent with the variety and speed of tectonic processes currently observed in SE Asia. The Neogene collisions of continental fragments in Sulawesi are a principal cause of its geological complexity and Ali et al, provide some insights into how tectonic models can be tested using palaeomagnetic data. In Buton, large rotations are recorded, but are apparently very local, and were very rapid. This work reinforces the value of, and need for, many more palaeomagnetic studies in SE
xii
R. HALL • D. J. BLUNDELL
Asia in order to separate local from regional motions. Buton is one of several continental fragments which are now being reassembled in SE Asia. These include Australian and Sundaland material but their origin can often only be inferred from indirect arguments, commonly controversial. Vroon et al. suggest that isotopic evidence can contribute to solving this problem and show how different types of continental crust can be characterized by analysis of igneous and sedimentary rocks. They suggest that different parts of east Indonesia have provenances in southern New Guinea, north Australia, Pacific New Guinea and Sundaland, leaving the tectonicians with an additional tool but, in this area, some additional problems to solve. In the midst of the continental fragments of east Indonesia are the deep basins of the Banda Sea, as yet unsampled by the ocean drilling programme, and of uncertain age. Linthout et al. report new isotopic ages from Seram implying Neogene spreading in the southern Banda Sea before ophiolite obduction on Seram in the Late Miocene. These ages are broadly consistent with ages of rocks recovered during recent dredging in the Banda Sea and with the tectonic reconstructions of Hall, although the great depths and low heatflow measurements remain apparent inconsistencies. On the north side of the Banda Sea the Sorong fault system separates Australia from the Philippine Sea and Molucca Sea plates and terminates in the continental fragments of east Sulawesi. The timing of movement on the strike-slip faults has never been clear, although recent work suggests this plate boundary zone became a strike-slip system in the early Miocene. The timing of movements and the distribution of continental crust in this region is of major interest, not least in the search for hydrocarbons, since this is an area of established production, recent discoveries, as well as active exploration, all linked to Australian crust. Based on stratigraphic arguments, Charlton argues that two of the basins, the Salawati basin of western New Guinea and the Tomori basin of eastern Sulawesi, were originally a single sedimentary basin, now separated by latest Miocene to Quaternary movements on the fault, implying a left-lateral displacement of about 900 km. Movement on the fault system is one of the latest complications in the development of east Indonesia; on the south side of the fault system is Australian crust while on the north side are the arc-arc collision of the Molucca Sea and the clockwise-rotating Philippine Sea plate. In the fault zone, which includes several major splays, are fragments of both Philippine Sea and Australian origin and Bacan is one of the islands which includes rocks of both provenances. Bacan therefore offers the possibility of elucidating
some of the history of the plate boundary zone, and establishing the timing of tectonic events. Malaihollo & Hall report new stratigraphic data from Bacan which record the early arc history of the Philippine Sea plate and the arrival of continental crust, providing a basis for distinguishing different tectonic models. Baker & Malaihollo discuss the timing of volcanism in the islands of Halmahera immediately north of Bacan, which records the initiation of subduction of the Molucca Sea plate beneath Halmahera and the development of the present-day arc-arc collision. Volcanism began in the middle to late Miocene and migrated northwards implying that the double subduction system was established between about 15-12 Ma. This evidence still remains to be incorporated in models linking east Indonesia to the Philippines. Pubellier et al. provide further evidence from the southern Philippines critical to linking the two areas and understanding the development of the present tectonic setting described earlier in the volume by Rangin et al. Once again, the theme of partitioning of oblique convergence is emphasized. In addition, there are complications reflecting Neogene changes in plate motions and the complexities of intra-arc deformation. Strain has been partitioned between several orientations of faults, reactivated at different stages as thrusts and wrench faults, as well as subduction zones. Of particular interest is the way in which this development has resulted in intra-arc extension and fragmentation within the Philippines. The interplay of subduction and strike-slip faulting is a theme which appears in many of the papers in the volume. Ancient strike-slip motion is often difficult to demonstrate and quantify and is consequently often neglected. However, there is evidence, traditionally linked to the convergent component of collision, which may be differently interpreted. Crowhurst et al. show that fission track data suggest that the Papuan metamorphic rocks may be interpreted as representing early Neogene extension after arc collision, rather than contraction-related metamorphism. These arguments may be applicable in other parts of SE Asia where fission track and isotopic ages are revealing unsuspected events, very young ages, and short time periods for the very complex tectonic evolution of many parts of the region. The volume concludes with two papers from south China accompanied by interpretations of the timing and significance of events. Wopfner suggests that the Baoshan and Tengchong Blocks in western Yunnan have a Gondwana origin, supported by the presence of Upper Palaeozoic glaciomarine deposits, cold-water faunas and Glossopteris. The terranes separated from Gondwana in the Early Permian and docked with
. ° .
INTRODUCTION Cathaysia in the Late Triassic, although Tertiary strike-slip faulting on the Nujiang Line has juxtaposed the two terranes. Z h o u et al. reinterpret part of the history of SE China, diverging in particular from the traditional practice of relating major unconformities to separate orogenies, and suggesting instead that they record different stages in the evolution of a single orogeny following early Mesozoic collision between the South China and the South China Sea blocks. SE Asia is one of the most exciting regions of the globe for any earth scientist. The size, difficulties and practicalities of the region demand a long-term investment of effort but the rewards are illustrated by the papers in the volume. These give an insight into how the history of the region will be uncovered, as well as providing an overview of present regional tectonics and its development. Because of the rapid rates of movements in many parts of the region new geodetic tools and increasingly refined methods of examining earthquake data mean that realistic and rapid tests of regional and global plate models are possible. The challenge for the future is to examine how far interpretations of these data can be pushed back into the past, to improve kinematic descriptions and models, and to identify the processes which have led to the complexity of the region and which can be applied to older orogenic belts. In these tasks, there remains an important role for the field geologist as well as those developing and applying new technologies. An improvement of our understanding will have benefits for knowledge as well
Xlll
as for the many inhabitants of the region, in helping to develop its resources and mitigate its hazards. This Special Publication arose from a conference on the Tectonic Evolution of SE Asia held at the Geological Society in London in December 1994. In addition to the reasons outlined above for the conference, the London University SE Asia Research Group wished to mark the retirement of Dr A. J. Barber in 1994. Tony Barber initiated several of the studies which are presented as publications in this volume and was instrumental in developing the programme of the SE Asia Research Group over many years, particularly by fieldwork throughout the region. We wish him a long and happy retirement. We thank all of the following who provided reviews of manuscripts: J. R. Ali, M. Allen, M. G. Audley-Charles, A. J. Barber, H. Bellon, S. C. Bergman, J. C. Briden, C. S. Bristow, T. R. Charlton, J. Charvet, B. Clennell, D. Q. Coffield, M. C. Daly, J. E. Dixon, C. Elders, R. Ellam, A. Fortuin, M. Fuller, R. J. Garrard, N. S. Haile, N. A. Harbury, K. C. Hill, A. J. Hurford, C. S. Hutchison, S. J. Kelly, J. Malod, S. J. Matthews, R. McCaffrey, M. Menzies, I. Metcalfe, J. S. Milsom, A. H. G. Mitchell, G. Moore, S. J. Moss, R. J. Murphy, G. J. Nichols, G. Packham, C. D. Parkinson, S. Polachan, M. Pubellier, A. J. Racey, C. Rangin, J-P. Rehault, M. A. Samuel, D. Snyder, P. Styles, R. E. Swarbrick, M. E Thirlwall, E Tongkul, J. J. Veevers, R. von Huene, G. K. Westbrook, H. J. Wensink. We are especially grateful to Diane Cameron who carried out many of the major tasks related to organizing the conference, in addition to her assistance in reviewing and editing the manuscripts. Simon Baker, Steve Moss and Moyra Wilson also provided considerable assistance during the preparation of the volume, for which we are extremely grateful.
Slip partitioning at convergent plate boundaries of SE Asia ROBERT McCAFFREY
Department of Earth & Environmental Sciences, Rensselaer Polytechnic Institute, Troy, N e w York 12180, USA Abstract: The active tectonics of SE Asia can be characterized by the interactions of large, rigid plates separated by broad zones of deformation. The relative motion of these plates across their boundaries is often partitioned in the sense that the normal and shear components occur on different structures. Earthquake slip vectors and geological and geodetic measurements are used to infer the degree to which oblique convergence, which is ubiquitous in SE Asia, is partitioned. The active tectonics of Sumatra, the Himalayan thrust, the Philippines, the New Guinea fold-andthrust belt, the Huon-Finisterre collision, and the San Cristobal trench can be understood in terms of upper plate deformation associated with oblique convergence. Western Java may also exhibit partitioning of oblique subduction. Structures accommodating normal and shear components of the motion are often very close. Arc-parallel strain rates are estimated for forearcs of the region. The arc-parallel deformation of forearcs of the SE Asia region demonstrates that plate convergence, whether normal to structure or not, is a three-dimensionalprocess.
When convergence between two plates is not perpendicular to their boundary, shear stress parallel to the plate boundary results in marginparallel shear strain within both plates. In a collision belt or subduction zone, if one of the plates is weak in shear, then the total slip may be partitioned into shear and thrust components (defined by the plate boundary orientation). Often these shear and normal components of slip across the boundary are accommodated on different geological structures. Oblique convergence is globally much more common than trench-normal convergence and often the obliquity varies along the margin. The deformation in many convergence zones may be understood in terms of such slip partitioning. A widely cited example of these parallel structures is along the San Andreas system, which comprises a strike-slip fault and a parallel fold-and-thrust belt (Mount & Suppe 1987) even though the relative motion between the Pacific and North American plates is only a few degrees from the trend of the San Andreas fault. Throughout SE Asia and the SW Pacific, we see several examples of oblique convergence (Fig. 1), both in subduction and collision settings. Studies of Sumatra led Fitch (1972) to first propose that slip was sometimes accommodated by parallel thrust and strike-slip faults. Since that study, it has been found that partitioning of oblique subduction is globally quite common although it is rarely completely partitioned (Jarrard 1986; McCaffrey 1994). Similar geometries, in which a major strike-slip fault accommodates the boundary-parallel shear
strain, exist in the Philippines (Barrier et al. 1991) and possibly in Irian Jaya (Abers & McCaffrey 1988). Other convergence zones that have clear slip partitioning are the Himalayan thrust (Molnar & Lyon-Caen 1989), the Aegean (Gilbert et al. 1994), New Zealand (Anderson et al. 1993) and the Aleutians (Ekstr(Sm & Engdahl 1989; McCaffrey 1992). McKenzie & Jackson (1983) argued that, if plate motions are driven by ductile shear from below, oblique convergence in continental settings should be completely partitioned, that is, slip will occur only on strike-slip and pure thrust faults. This theory provides a test of crustal dynamics in oblique collision zones but requires detailed knowledge of the deformation. SE Asia is a good region to test such ideas because oblique convergence zones are plentiful and fast. In the past few years much has been done to improve our knowledge of the pattems of crustal deformation in SE Asia. In particular, systematic studies of large earthquakes, geological studies of active faults, and geodetic measurements using the Global Positioning System (GPS), that commenced in the SW Pacific region in 1988, are revealing important results, and surprises. The goal of this paper is to review our knowledge of the crustal deformation in SE Asia with attention to the partitioning of slip between thrust and shear faults in the broad deforming boundaries. First, constraints on the motions of the major plates are discussed in order to put bounds on the convergence geometries and then the deformation within these zones is described.
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 3-18.
4
R. McCAFFREY
Himalaya
}
//
30°N
llEUt <,
India
o
_
200N
PSP Marianas
J~INDIAN
10°N <.;3
SEA :/
•
~
!
° ). -/7, I
10 cm_ a -1 ,,,50 SV _ "~
New Guine~
l SV
800E
1200E
100°E
Ca)
olomo°
I.--- -~
?~)/'~~
[~ ~SV
Io
140°E
t
',t 0*
,
t , s--z
o ~
10*S
I -20*S o
.Y o
30°S Pacific
0 o.
Oo
150°E (b)
170°E
i70°W
10°S
SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES
5
Plate motions
The Philippine Sea plate
The broad kinematic framework of SE Asia, determined by the relative motions of the Pacific, Australian, and Eurasian plates, is described to the first order by NUVEL-1A, the global plate motion solution of DeMets et al. (1994a). In addition, the motions of the Philippine Sea, the SE Asian, and Caroline plates control the large-scale tectonics of the region, and these are discussed below.
The motion of the Philippine Sea plate (PSP) is still poorly constrained due to lack of reliable kinematic data; only subduction zone earthquake slip vectors are available (Ranken et al. 1984; S e n t et al. 1993) and these are probably deflected by local deformation of the leading edges of the upper plates (McCaffrey 1994). Moreover, earthquake slip vectors do not contain information about rates of motion. Despite these handicaps, S e n t et al. (1993) revised estimates of PSP motion by constraining it to comply with the NUVEL-1A global solution. However, this does not improve the tie of the PSP to its neighbouring plates; for example the uncertainty in the latitude of the Pacific-PSP pole is nearly 50 ° (Sent et al. 1993). Ongoing programmes by several groups to measure the motion and deformations of the PSP with GPS will provide the required data to constrain the PSP motion relative to the rest of the world. Recent GPS measurements across Taiwan (Yu & C h e n 1994) show that the PSP and Eurasia converge at azimuth 3070_+ 1 and a rate of 86 _+2 m m a -1, which is in a similar direction but 20% faster than the S e n t et al. (1993) PSP-Eurasia pole predicts. A very long baseline interferometry vector (Matsuzaka et al. 1991) between sites on Japan (North American plate) and the Izu-Bonin arc (PSP) also agrees with the azimuth predicted by the combined PSP-North America Euler pole (Sent et al. 1993; DeMets et al. 1994a) but is 15% faster. Both results suggest that Sent et al. may have underestimated the angular velocities for the PSP relative to Eurasia and North America. In the following, the S e n t et al. (1993) poles are used for the PSP but other poles are tried in the kinematic analyses of forearcs around the PSP.
SE Asia as a separate plate
One of the important regional problems is whether or not SE Asia (SEA) moves independently of Eurasia. This question is linked to collision in the Himalayan region and its answer will provide constraints on the mechanism of continental collision. Recently, the first direct measurements of the relative motion between Australia and Indonesia across the Java trench were made. Five GPS measurements over 4 years on a baseline between Christmas Island (Fig. la) and western Java reveal a relative velocity (67 _+7 m m a-t rate at 011 _+4 ° azimuth) that matches the NUVEL-1A vector for Australia relative to Eurasia (71 _+ 2 m m a -1 rate at 020 _+3 ° azimuth) within 1 cm a-1 (Tregoning et al. 1994) (Fig. l a). This single vector can be interpreted in several ways: that SEA moves slowly if at all relative to Eurasia (within the uncertainties of GPS); that SEA moves relative to Eurasia but the pole of rotation is near western Java; or that elastic strains contaminate the geodetic signal due to the proximity of the GPS sites to the plate boundary. Until more results become available, the author favours the first view - that SEA moves very slowly, 1 cm a -1 or less, northward relative to Eurasia. Because the motion of SEA relative to Eurasia is slow and the pole of rotation is poorly constrained, in the following discussions SEA will be treated as part of Eurasia.
The Caroline plate
Weissel & Anderson (1978) argued for the existence of a separate Caroline plate (CAR)
Fig. 1. Tectonic maps. (a) SE Asia. Thrust faults have barbs on the hanging wall. Heavy black arrows show relative plate motions at boundaries (vectors show upper plate moving relative to lower) using poles given in Table 1. Grey arrows with error ellipses show GPS vectors of Cocos Island (CO) and Christmas Island (CI) relative to Java (Tregoning et aL 1994) (black arrows show expected Australia-Eurasia vector). Small arrows pointing landward of trenches show average slip vector (SV) azimuths of interplate thrust earthquakes along 400 km long segments of the forearcs (lengths of arrows are scaled to number of SV in average). Thick grey lines show outline of broad deforming region between the Indian and Australian plates (DeMets et aI. 1994b). SEA, SE Asian plate; PSP, Philippine Sea plate; ODR, Oki Daito ridge; CAR, Caroline plate; JT, Java trench; BS, Banda Sea; MS, Molucca Sea; PF, Philippine fault; PT, Philippine trench; AS, Andaman Sea; SS, Sunda Strait; NGT, New Guinea trench. Bathymetry contours at 2000 m intervals. (b) Southwest Pacific. Format same as (a). GPS vectors are relative to Pacific plate (Bevis et al. 1995). The large dot north of New Zealand is the pole of rotation of the N. Tonga forearc relative to the Pacific determined from GPS vectors. NH, New Hebrides arc; LB, Lau Basin; LR, Louisville Ridge; SV, slip vector.
6
R. McCAFFREY
o ~_~
,/
.j~v@o o
i
~
~°
YapTrench~--~ <'-~-~
,_
t!;~,~/(~lh~'
j ,.3 -
Caroline Plate
//
I0 mma -I
~,
,~~1
Slipvector
~
.
b
r
~
I
137*E
I
IO*N
I
I
1420E
-t5"N
,."
" <'/.-',, ~
132°E
o
~
y..
:
~)
I I
-J
.~ ~ ' ) ~ ,
~ ~k ~J •
.
147*E
10i
152*E
Fig. 2. Plot of bathymetry, plate boundaries, relative plate motion vectors (thick arrows), and earthquake mechanisms and slip vectors (thin black arrows) around the Caroline plate. Grey arrows show the expected motion of the Caroline plate relative to the Pacific and Philippine Sea (PSP) plates and the thicker black arrows show the motion of the Pacific relative to the PSP, all based on the Seno et al. (1993) poles. Earthquake mechanisms are from the Harvard CMT catalogue. PSP, Philippine Sea plate.
between the Pacific, the PSP, and northern New Guinea (Fig. la). This suggestion was based largely on the morphologies of the Mussau trench and Sorol trough (Fig. 2); the former thought to be a subduction zone and the latter predominantly a left-lateral shear zone with some extension. Rates of motion of CAR relative to surrounding plates are based on the observed increase in sediment thickness away from the Ayu Ridge (Weissel & Anderson 1978) and on inferences about the amount of shortening at the Mussau trench (Hegarty et al. 1983). Such motions are estimated at less than 1 cm a-l. Ranken et al. (1984) and Seno et al. (1993) estimated rotation poles for CAR relative to the Pacific and PSP by requiring convergence along the Mussau trench and a combination of extension and strike-slip at the Sorol trough. Recent earthquake mechanisms along the Mussau trench and Sorol trough (Fig. 2) are not consistent with a simple plate boundary between CAR and the Pacific. The slip vector for one event near the intersection of the Sorol trough and Mussau trench is nearly orthogonal to the expected slip based on the Seno et al. Pacific-CAR pole, Events along the Mussau trench itself indicate some thrusting but mostly right-lateral strike-slip on very steep planes that are consistent with the strike and polarity of the trench.
Thrust earthquakes along the Yap trench have slip vectors that are consistent with either Pacific-PSP or CAR-PSP motion. The alleged Caroline-Pacific plate boundary is not a very compelling feature and may simply be a region where stress concentrates due to locally steep, inherited topography. In this paper it is presumed that the 'Caroline plate' is kinematically part of the Pacific plate. Even if the Caroline plate is a separate block, ignoring it in the kinematic analysis will not be critical because it probably moves slowly relative to surrounding plates.
Deforming boundary zones and slip partitioning By far the most elusive aspect of SE Asian tectonics is the description of the deforming zones between the rigid portions of the major plates described above. One way to view the deformation of the plate boundary zones is in terms of slip partitioning. Before discussing slip partitioning at convergent margins of SE Asia, some basic geometrical concepts are reviewed. The mechanics of slip partitioning are discussed by McCaffrey (1992) and Platt (1993) but this paper concentrates on the kinematic consequences. Most of the inferences
SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES drawn here are based on the deflections of slip vectors of interplate earthquakes away from the expected convergence direction between the two plates (Fig. 1 shows orientations of slip vectors and plate vectors along plate boundaries). Earthquake slip vector deflections tell us mostly about the arcparallel component of deformation within the upper plate, because most convergence is at a high angle to the trench normal, arc-normal deformation of the forearc does little to deflect the slip vectors (McCaffrey 1991). If the plate convergence vector and the orientation of the trench are known, then the arc-parallel rate of motion v s of the forearc relative to the upper plate can be inferred from the slip vector based on a simple kinematic relationship (Fig. 3). Moreover, if the obliquity varies along the margin then v s may vary as well, resulting in an arc-parallel strain rate. Note that the arc-parallel strain rate arises from the gradient in obliquity, not from the obliquity itself, and we will see that arc-parallel strain can occur where the obliquity is zero. Hence, even though subduction may be locally perpendicular to the trench and apparently two-dimensional, rapid arc-parallel deformation of the upper plate can occur, producing a threedimensional strain field. Earthquakes used are those that reveal slip between the subducting plate and the overriding forearc. McCaffrey (1994) discusses the selection of data (i.e. trench orientations, plate convergence vectors and slip vectors) needed for this type of analysis. Slip vectors are averaged in 5 0 - 1 5 0 k m long segments of the trenches. This paper uses the Harvard centroid-moment tensor solutions from 1977 through August 1994 and other published solutions for earthquakes prior to 1977. It is assumed that the downgoing plate does not deform rapidly; slip vector deflections require motions of a large percentage of the plate motion rate and it is unlikely that subducting plates are sheared at such high rates. The goal is to infer arc-parallel slip rates v s (Fig. 3) and arc-parallel strain rates (arc-parallel gradients in Vs) in forearcs in the SE Asia region (Table 1). Where available the slip vector derived slip rates are compared to geological or geodetic measurements of slip rates. Sumatra and Java
Fitch (1972) correctly inferred that oblique convergence at Sumatra was partitioned into a component of thrusting at the Java trench that was normal to the trench and a component of arc-parallel shear on the Sumatra fault. In the past few years, detailed investigations utilizing more and better data have refined the kinematics of the Sumatra region. Attempts to treat the Sumatra forearc, between the trench and the Sumatra fault, as a rigid plate (e.g.
7
Upper Plate
/~
A A A VS 1 S l i p vector ~<_....z_.~, / ~"~ - - ~ / ~ Forearc z ' ~ IVn Trench
~"
Vp
r
SubductingPlate
v ; predicted plate convergence vector T = trench-normal azimuth @ - plate motion azimuth 13 - slip vector azimuth g = trench normal - slip vector = T - 13 Obliquity: ~, = T - @ Slip vector residual: A[3 = [3 Forearc slip rate relative to Upper Plate: v s = v ( sin y - cos ~/ tan ~ ) Fig. 3. Geometry of slip partitioning as inferred from earthquake slip vectors. The important items are ~/, the plate convergence obliquity, ~, the slip vector obliquity, and v s, the slip rate of the forearc in the arc-parallel direction relative to the upper plate.
Curray 1989) generally failed to explain subduction west of the Andaman Sea, where the trench takes a turn to the NNE (Fig. l a) or the difference in the amount of extension in the Sunda Strait (100 km) (Huchon & Le Pichon 1984; Harjono et al. 1991) and Andaman Sea (400 km) (Curray et al. 1979). These features can be explained if the forearc stretches in addition to being translated northwestward relative to SEA (Diament et al. 1990; McCaffrey 1991). Using a convergence direction at the Java trench of N3°E and a rate of 75 m m a -t, an arc-parallel extensional strain rate of 3-4 × 10-8 a-1 is required of the Sumatra forearc to match the deflections of the slip vectors (McCaffrey 1991). Tregoning et al. (1994) show that the N U V E L - 1 A AustraliaEurasia vector is approximately correct for the Java trench. The NUVEL-1A Australia-Eurasia convergence direction results in a lower inferred arcparallel strain rate (1.8_+0.5 × 10-8 a-I; Fig. 4a, Table 1) than the N3°E direction does because it matches better the slip vectors west of Sumatra. Such arc-parallel stretching of the forearc implies that the slip rate on the Sumatra fault
8
R. McCAFFREY
Table 1. Arc-parallel strain rates for forearcs Trench name
Pole of rotation*
Number of trench segments
Number of slip vectors
Arc-parallel strain rate (x 10-8 a -1)
Correlation coefficient
RMS (mm a-z)
Himalaya Izu Java Mariana New Hebrides N. Ryukyu N. Tonga Philippine S. Ryukyu S. Tonga Sumatra
N IND-EUR S PAC-PSP N AUS-EUR S PAC-PSP N PAC-AUS S PSP-EUR N PAC-AUS S PSP-EUR S PSP-EUR N PAC-AUS N AUS-EUR
13 7 15 10 11 5 4 23 4 15 18
19 67 40 61 124 27 99 140 14 481 102
2.2 _+0.5 0.7 _-!-0.7 -0.1 _+0.4 1.7 + 0.5 12.3 + 2.5 1.3 _+0.3 16.3 _+5.4 2.6 _+ 1.0 0.6 _ 1.6 -0.1 _+0.4 1.8 _+0.5
0.79 0.52 -0.05 0.81 0.96 0.92 0.93 0.60 0.13 0.15 0.84
12 6 12 6 25 2 14 18 7 11 11
* N, Demets et al. (1994a) NUVEL-1A; S, Seno et al. (1993); IND, India; EUR, Eurasia; PAC, Pacific; PSR Philippine Sea plate; AUS, Australia. Trench segments are between 50 and 150 km long. Positive values of strain rates are arc-parallel extrensional, and negative values are compressional.
increases to the N W (McCaffrey 1991), w h i c h is supported by the few direct m e a s u r e m e n t s of slip rate on the fault (Fig. 4a). At its southern end, near 5°S, the slip rate is less than 10 m m a -z (Bellier et al. 1991) but increases to about 10 m m a -1 near the equator and to about 28 m m a -1 near 2.2°N (Sieh et al. 1994) (Fig. 4a). The average gradient in the slip rate on the fault is close to that predicted f r o m slip vector deflections; an increase in 28 m m a -1 over 1200 k m f r o m the S u n d a Strait to 2.2°N gives a strain rate of 2.3 x 10 -8 a -1. As is seen in Fig. 4a, slip vectors and the N U V E L - 1 A Australia-Eurasia pole give approximately the correct slip rate for the Sumatra fault but also predict left-lateral, arc-parallel shear at a rate of 15-20 m m a -1 at the longitude of W Java. A pole of rotation that fits the earthquake slip vectors south of Java, and hence does not require upper plate d e f o r m a t i o n near Java, predicts h i g h e r than o b s e r v e d slip rates for the S u m a t r a fault
(McCaffrey 1991). Therefore, either the earthquake slip vectors are p o o r indicators o f upper plate deformation, there are faults other than the Sumatra fault that a c c o m m o d a t e forearc deformation off Sumatra (Karig et al. 1980; D i a m e n t et al. 1992), or there is left-lateral shear through Java (Dardji et al. 1991; Malod et al. 1996). The constant v s value at Java (Fig. 4a) indicates a near-zero strain rate. T h e Java forearc m a y be m o v i n g relative to the Sunda shelf but, unlike the Sumatra side, not u n d e r g o i n g internal strain. The m e c h a n i s m for the arc-parallel stretching of the Sumatra forearc is unclear but involves some extension on cross-forearc faults, such as in the strait b e t w e e n Java and S u m a t r a ( H u c h o n & Le Pichon 1984; Harjono et al. 1991; M a l o d et al. 1996), and s o m e strike-slip faults crossing the forearc (Karig et al. 1980). T h e largest shallow, strikeslip earthquake to occur in Sumatra in the past 30 years was beneath the forearc and had nodal planes
Fig. 4. Plot of plate convergence obliquity (shaded curves), slip vector obliquity (open circles), arc-parallel forearc slip rates (v s, closed circles), and arc-parallel slip rate of subducting plate (v., thin solid curve) along convergence zones of SE Asia. The horizontal axis is distance in kilometres along the de(ormation front or trench (also labelled with approximate latitude or longitude). Positive distance is the direction faced if the thrust fault dips to the viewer's right (for example, positive distance is west if the fault dips N). The vertical axis is angles in degrees and slip rates in mm a -z. Plate obliquity is inferred from trench-normals, plate convergence directions (poles used are listed in Table 1), and their uncertainties. Thin dashed curves, when present, show obliquity from poles of rotation based on fitting slip vectors at the trench. Slip vector obliquity is the angle between the slip vector and the direction normal to the trench (Fig. 3). Positive values of v s indicate right-lateral shear and negative values are left-lateral. Heavy dark straight lines show the best-fit to the values of v s, the slope of which (dVs/dX) is the arc-parallel strain rate (Table 1; positive slopes show arc-parallel extension, negative slopes show arc-parallel compression). Triangles show geodetic and geological estimates of arc-parallel slip rates, as discussed in text. NZ, New Zealand.
9
SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES
Himalaya
that were rotated by 45 ° relative to the trend of the Sumatra fault (Zwick & McCaffrey 1991). It is likely that more detailed study of the Sumatra forearc structure will find complex deformation, such as in the Aleutian (Geist et al. 1988) and Cascadia (Goldfinger et al. 1992) forearcs.
/.-.,
"7
i
E
i
Due to the extensive deformation within Asia north of the Himalaya the convergence rate and direction at the Himalayan front is poorly constrained. The convergence rate is estimated to be 15 m m a -1
i
i
a
Sumatra Java
Timor ,
C
¢~
~~" 60 ,
!
90
--
W
~'~' ~'~'~''
.....
30 o
-30 ,0.7E
© 0
' k'C---i122.3E
111.
1104"sE
I 1"5S
(a) Himalaya_ ,... ~ + ~.~~J":~Ji 1~ w 59~ 1000
~E ~ f
2000
3000
5000
o
,
. ~5 4E
0
(b)
90.9E
5.8E
500
1000
80.9E
1500
-60
15"4N
4000
30
76.6E
-75 00
50
2000
0
250
500
750
1000
(e)
-~"
,
,
,
-
1250 ~ 60
25
30
t ...
o
-25 -2.' .::...:.::..~
g"
///
.1~11
•-Q
75~/~~ ..........."........
._.
(c)
-50
New
-5,
~¢~/'~' Hebrides
•~
O 0
'
1171.7E
400
121.1S
17.8SI
800
-30
-75 - 100
1200
0
60
(t)
500
1000
1500
-
2000
-60 60 }0
~D
3 if-,
-30
4 -30
-60 1f'4.6S
O
0
2.4S
1000 (d)
0.9S
2000 Distance, (km)
9.5S
3000
-90
0 (g)
1000 2000 Distance, (kin)
3tRIO
-90
10
R. McCAFFREY vectors tend to remain perpendicular to the thrust front instead of following the convergence direction (Fig. 4b). Kinematically this requires deformation of the overriding plate. The uniform arc-parallel strain rate estimated from slip vectors
(Molnar & Lyon-Caen 1989), or about one-third of the total India-Eurasia motion. Due to the large curvature of the Himalayan front, the obliquity varies by more than 90 ° along it (Fig. 4b). Despite this large variation in obliquity, earthquake slip
/~~-,,,~,
~ i ~
SundaShelf
.
.
.
-k.
m
/\ ~ "~'--~\ ~ -~ " j~ \ ~ - ~ B a n d a S e a _ ~ - - - - - ~
-
.
"
-
......
~
~i
,
l~.~*o)~)~?h
°
"
,o , . ~ . ~ ! ~ j f ~ ' . ~ < ~ 4
~'~ ¥,~ 9.s
.......
-"
--~
-,, 116°E
-
~Z
.4 11°S
, 118°E
120°E
122°E
I,
,
124"E
126°E
,
,
, 128°E
130°E
132°E
(a) I°N
t°S
t°S
i*S
#°S
1330E
135"E
137"E
139"E
141°E
143"E
145"E
147°E
149"E
(b) Fig. 5. Fault plane solutions (1962-1984; M s > 5.5), slip vectors, and plate vectors for (a) eastern Indonesia and (b) New Guinea. Size of earthquake symbol is scaled to logarithm of seismic moment. Asterisks show active volcanoes. Small arrows show slip vectors of interplate thrust earthquakes and large arrows show expected plate convergence directions based on (a) Australia-Eurasia and (b) Australia-Pacific (DeMets et al. 1994a). MT, Manokwari trough; BH, Bird's Head; SF, Sorong fault; CB, Cenderwasih Bay; TAF, Tarera-Aiduna fault; MTB, Mamberambo thrust belt; HFTB, Highlands fold-thrust belt; NGT, New Guinea trench; MV, Markham Valley; HF, Huon-Finisterre Range; BP, Bismarck plate; NBT, New Britain trench; PNG, Papua New Guinea.
SLIP PARTITIONINGAT CONVERGENT PLATE BOUNDARIES is 2.2 + 0.5 × 10-8 a-l (Fig. 4a) if the India-Eurasia convergence rate is used and one-third of that (i.e., 0.7 x 10-8 a-l) if two-thirds of India-Eurasia convergence is taken up in Asia. Molnar & Lyon-Caen (1989) estimate 1.0 _+0.5 x 10-8 a -1 (18 _ 9 mm a-t over a distance of 1800 km) for the rate of E-W extension of Tibet. Extension of the Himalayan frontal arc is thought to be driven by gravitational collapse of the overthickened Tibetan Plateau, instead of shearing due to oblique plate convergence. The evidence for this is the correlation of faulting type with elevation. Normal faulting is confined to elevations above 4000 m, strike-slip occurs at all elevations, and thrust faulting is confined to elevations below 4 5 0 0 m (e.g. Molnar et al. 1993). However, because the earthquakes showing rotated slip vectors occur within 200 km of the thrust front, the Himalayan forearc must also be extending along strike at about 10-8 a-1. Therefore, while normal faulting may be confined to high elevations, E-W extension of the upper plate is not. As is the case for the Sumatra forearc, such extension could be accommodated by strike-slip faulting at low elevations instead of normal faulting. Note that the arc-parallel strain rate applies even to the region near 86°E ( - k m 1000) where subduction is perpendicular to the mountain front. (At several other subduction zones in Fig. 4, similarly, the strain rate is not zero even thought the plate obliquity is near zero.) Even though v s ---0, the margin-parallel gradient in v s is non-zero. In the case of normal convergence, strain will occur in the vertical plane containing the convergence vector and the normal to the margin. Since v s represents motion perpendicular to this plane, a gradient in v s results in strain outside of this plane, producing a three-dimensional strain tensor. The force that stretches the forearc along its length arises from the gradient in the arc-parallel shear stress due to the increase in obliquity on both sides of the point where margin-perpendicular convergence occurs (McCaffrey 1992). Hence, subduction at a curved margin is a three-dimensional process even where it may be locally perpendicular to the trench. Banda arc and Timor
Earthquakes suggest that the present stage of the collision of the Australian continental margin with the Banda arc probably involves more rapid convergence at the backarc thrusts than at the Timor trough (Fig. 5a). The Wetar and Flores backarc thrust zones are considerably more energetic seismically than the subduction thrust fault north of the eastern Java trench and Timor trough (McCaffrey 1988); all known large thrust earthquakes in this decade have occurred at the Flores
||
and Wetar thrusts. The M w = 7.9 December 1992 Flores earthquake is the latest and largest of many to occur along the backarc thrusts. Preliminary GPS results show that the slip rate associated with backarc thrusting increases from Bali eastward to about 50 mm a-1 at the eastern end of the Flores thrust and reaches the full expected rate of Australia-Eurasia relative motion (70 mm a-1) at the Wetar thrust north of Timor (Genrich et al. i994). GPS appears to be confirming the inference based on earthquakes (McCaffrey 1988) and crustal structure (Silver et al. 1983) that the eastern Sunda arc is rotating anticlockwise about a pole in eastern Java. Timor appears to be moving northward relative to the Sunda Shelf (SEA) at about the same rate as Australia, suggesting that the Timor trough no longer accommodates subduction (Johnston & Bowin 1981; McCaffrey & Abers 1991; Genrich et al. 1994). The northward motion of the extinct volcanic islands north of Timor is revealed with GPS to be similar to that of southern Timor (Genrich et al. 1994); hence, little of the northward motion of Australia is now accommodated by internal deformation of the Banda arc structure. The island arc structure, bounded by the Timor trough and the Wetar backarc thrust, is rigidly thrusting over the backarc basin. The present geometry is probably quite young; the Wetar thrust exhibits less than 10 km of shortening (McCaffrey & Nabelek 1986), which would take only 0.15 Ma at the present rate of 60 mm a-1. Probably between 3 Ma, when collision began, and 0.15 Ma convergence was in part accommodated by shortening and thickening of the forearc (Snyder et al. 1993). The Sunda arc now shows a gradual transition from subduction of oceanic lithosphere south of Java to a completed accretion of an island arc terrane to a continental margin at Timor. In the process, convergence has jumped from the forearc trench to the backarc thrust. South of Timor, the Timor trough trends about N70°E while the plate convergence vector is probably close to N10°E (Tregoning et al. 1994), resulting in an obliquity of about 30 ° (Fig. 4a). The geometry predicts that the Timor forearc should be under ENE-trending, left-lateral shear and possibly arc-parallel tension. Interplate earthquake slip vectors (Fig. 5a) are too few to confirm this; values of v s inferred from slip vectors range from zero to the full arc-parallel motion rate (Fig. 4a). The few available slip vectors show motions both parallel to plate convergence and perpendicular to the trench (Fig. 5a), suggesting a complex spatial variation in slip partitioning (below we see that this bimodal distribution of slip vectors is common at forearcs with large curvature). Recent earthquakes in the overriding plate suggest that both shearing and
12
R. M c C A F F R E Y
stretching are occurring. In particular, two shallow earthquakes beneath Timor show left-lateral, strikeslip on NE-trending, vertical planes and shallow, N-striking, normal faulting events are found north of central Timor (Fig. 5a) (McCaffrey 1988). GPS results also suggest that slow E - W stretching of the arc may occur (Genrich e t al. 1994).
H i | l a n d g hs
u i~i:i~i:i
i
~ ~i:!~i!:
.....~:
- 240
•!iii~i
~B
iii~!i!i 180 120
New
Guinea
The island of New Guinea accommodates oblique convergence between the Pacific and Australian plates at a rate of about 12 cm a-1 (Fig. 5b) and azimuth of N70°E. The young, high mountain belt, called the Highlands fold-and-thrust belt, trends at N100°E, so is about 60 ° away from being perpendicular to plate convergence. Large earthquakes suggest that shortening and shear are distributed over much of the island (Abers & McCaffrey 1988; McCaffrey & Abers 1991; Fig. 5b). Earthquakes within the Highlands are predominantly thrust and strike-slip with strikes generally parallel to structure rather than perpendicular to the plate convergence direction (Fig. 6a). Oblique convergence within the Highlands belt of Irian Jaya is therefore largely partitioned. The Mamberambo thrust belt (Fig. 5b) shows many more earthquakes with planes that strike perpendicular to the plate vector, suggesting less partitioning there (Fig. 6b). Recent GPS measurements (Puntodewo et al. 1994) reveal less than 2 cm a-1 of shortening and less than 3 cm a-1 of left-lateral shear between sites at the north and south coasts of E Irian Jaya (at about 140°E) even though nearly 6 cm a -1 of shortening and 10 cm a-1 of left-lateral shear are predicted by Pacific-Australia relative motion. At the longitude of central New Guinea, most of the motion between Australia and the Pacific occurs offshore, probably at the New Guinea trench, contrary to inferences made from geology and seismology. In westernmost New Guinea, plate motion appears to be accommodated in a different manner. GPS sites on the Bird's Head appear to move relative to Australia at nearly the same rate as the Pacific (Puntodewo et al. 1994; Stevens e t al. 1995). In the vicinity of the Bird's Head, the majority of Pacific-Australia motion occurs by large-scale W S W motion of the Bird's Head relative to Australia. The fault (or faults) on which this motion occurs is not clear although in the SW the main boundary is probably the Tarera-Aiduna fault (Hamilton 1979; Abers & McCaffrey 1988). Manuel Pubellier (pers. comm. 1994) suggests that a large, NE-trending shear zone exists landward of the east coast of Cenderwasih Bay (Fig. 5b). If so, it could be among the fastest slipping continental
~!Thrust .............(3 ..............oi!ii!!!ii ' ' ' ~ ......................................!!iiiiiiii.......... .
0
o
n
n
30
60
(a)
60
•
ii!ii!i,o o
o
iiiiiiiii I ~i~i~i~i I
-
PVN
I ili~i i l
90 120 Strike
150
-60 180
Mamberambo I
I
•l
:::i
I
I :!:i |
0
FTB
•
d 240 :~i:J
60 0 •
0 (b)
! 30
! 60
iii!i!i | iiiiiiill | 90 120 Strike
P~N i i!i!ii~!i!- -60 150 180
Fig. 6. Plot of rake angles vs. strike angles for earthquakes in (a) the New Guinea fold-and-thrust belt (FTB) between 135° and 144°E and (b) the Mamberambo thrust belt. For each event, both nodal planes are shown (using open and closed circles). The vertical line labelled FTB shows the trend of the fold-thrust belt, and PVN shows the trend perpendicular to the plate convergence direction. Rake angles near 90 ° indicate thrusting and those near 0 ° and 180° degrees are strikeslip (SS). Events with a strike near 100° and rake near zero reveal left-lateral slip on a plane parallel to the FTB.
shear zones in the world, being at least 7 cm a-l (Stevens e t al. 1995). The Sorong fault, at least where it passes through the Bird's Head, cannot now form the main boundary between the Pacific and Australia because GPS sites on opposite sides
SLIP PARTITIONINGAT CONVERGENT PLATE BOUNDARIES of the fault show little relative motion (Puntodewo et al. 1994). In Papua New Guinea (PNG), the HuonFinisterre (HF) arc terrane is being emplaced onto the Australian continent along a north-dipping thrust fault that crops out in or near the Markham Valley (MV) (Fig. 5b). Slip vectors of thrust earthquakes beneath the HF range north of the valley remain perpendicular to the valley as it swings to a NW trend (Fig. 5b). This rotation of the slip vectors cannot be explained by a single pole of rotation between rigid plates north and south of the valley. Schouten & Benes (1996) argue that the rotation of the slip vectors is caused by rapid deformation south of the New Britain trench and MV. On the bases of geological observations (Abbott et al. 1994), microearthquakes (Kulig et al. 1993), and large earthquakes (Abers & McCaffrey 1994), it is interpreted that the HF range is cut by cross-faults that accommodate relative slip and rotations of blocks of the HF range as it moves southward. In the case of the HF block, slip partitioning is accommodated by faulting at high angles to the trench as in Cascadia and the Aleutians, rather than arc-parallel faults, as in Sumatra. Solomons and N e w Hebrides
The Solomon and New Hebrides trenches form the boundary between the Pacific and Australian plates and, due to their great curvature, are the sites of large variations in convergence obliquity. The deviations of slip vectors from the plate vectors suggest that both forearcs deform significantly and partition the slip between the plates. At the southern end of the New Hebrides trench, west of 173.5°E, slip is partitioned completely so that thrust earthquakes at the trench have their slip vectors perpendicular to the trench and nearly perpendicular to the plate convergence vector (Fig. 7a). Going west around the bend in the trench, the slip vectors remain perpendicular to the trench to 18°S where the trench becomes nearly normal to the plate vector. East of about 173.5°E, the slip vectors rotate so that they form an angle of about 45 ° to the trench and plate vector. The extensional strain rate estimated for the southern New Hebrides forearc from the slip vectors is 12.3 __+2.5 × 10-8 a -1 (Fig. 4c). The Solomon trench becomes nearly parallel to the plate vector at the San Cristobal section (Fig. 7b). Slip vectors for earthquakes between 162 ° and 165°E show two orientations: one nearly perpendicular to the local trend of the trench and the other nearly parallel to the plate vector. If partitioning occurred by a detached forearc sliver moving along the arc, then the events with slip vectors perpendicular to the trench would be closer to the trench
13
(i.e. below the translating sliver) than those showing oblique slip. The uncertainties in the epicentral locations of these events preclude discerning such a pattern. Nevertheless, structures accommodating shearing and thrusting must be within a few tens of kilometres (the uncertainties in epicentres) of each other. Some fault plane solutions (Fig. 7b) show left-lateral, strike-slip faulting on E - W orientated, nearly vertical faults; these faults are consistent with the type needed to partition the slip by a small sliver of the forearc being translated eastward at a high rate relative to the arc.
Tonga trench
According to the NUVEL-1A Pacific-Australia pole of rotation, the Pacific plate subducts beneath the Tonga arc in a direction nearly perpendicular to the trend of the trench throughout most of its length (Fig. lb), although the obliquity gets large at the northern and southern ends (Fig. 4d). Earthquake slip vectors at the Tonga trench from 15°S to 35°S, deviate by 20 ° or less from the NUVEL-1A Pacific-Australia plate vector (Fig. 8a). If this pole applies to the Tonga trench then slip vectors suggest that the southern Tonga forearc translates very slowly (10 mm a-1 on average), left-laterally relative to Australia and does so as a rigid block (Fig. 4d; Table 1). GPS measurements by Bevis et al. (1995) suggest that the northern Tonga forearc rotates rapidly away from Australia as a rigid block (Fig. lb). Slip vectors also suggest that the forearc is rigid as they can be matched throughout most of the length of the island arc with a single pole of rotation at 54°S, 173°E (Fig. 8a). However, the poles of rotation estimated from GPS and from slip vectors are quite distant. The 3 GPS vectors (relative to fixed Pacific plate) from Tonga forearc sites are all matched within 2 mm a-1 in rate and 2 ° in azimuth by a pole at 32.6°S, 178.9°W (Fig. lb), and an angular velocity of-7.1 o Ma (the distance from the pole to the observation points is constrained by the gradient in GPS velocities). This pole falls near the central Tonga trench (Fig. lb) and therefore cannot describe the motion of the entire forearc from north to south because it predicts spreading between the forearc and the Pacific south of 33°S. There is probably a structure within the forearc south of the GPS sites that separates the N and S sections of the forearc. Most probably this boundary is near 24°S where the Louisville Ridge intersects the trench and spreading in the Lau Basin appears to terminate (Fig. lb). The northern Tonga arc goes through a large bend until the trench is nearly parallel to plate
14
R. McCAFFREY
ll*S
15*S
190S
~,3"S 159°E
163*E
167°E
171*E
175*E
(a) ' I ~ ,,'~~
Pacific /_ / ~
;
a-~
'
[
Y 10°S
y t I*S
I
I
I
163*E
164*E
1650E
166°E
(b) Fig 7. (a) Map of earthquake slip vectors (small arrows) and plate convergence vectors (large arrows) based on Pacific-Australia pole (DeMets et al. 1994a) at the Solomon and New Hebrides arcs. SCT, San Cristobal trench. Dashed outline box shows area of b. (b) Map of earthquake mechanisms (Harvard CMT 1977-1994; M s > 5,0) and slip vectors at the San Cristobal trench. The large arrow shows the convergence between Australia and the Pacific.
convergence (Fig. 8b), in the same fashion as the southern New Hebrides, San Cristobal, and southern Marianas arcs. Like the San Cristobal trench, slip vectors at the northern Tonga forearc appear to describe two sets of directions; parallel to the plate vector and perpendicular to the trench (Fig. 8b). The averages of these slip vectors imply
a rapid arc-parallel extensional strain rate (Table 1; Fig. 4d) for the northern Tonga forearc. However, because the trench-normal slip vectors are arcward (SW) of the vectors that parallel plate motion (Fig. 8b), they probably occur either within the subducted slab or in the upper plate and are not indicative of the interplate slip direction. (If this
SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES ! •
!
!
'l
!
i
NTF-PAC pole (GPS)
~
Trench normal i --
s ~
•
150
i
• •
•
~'*~
eee*%e
• "
110
•
-PAC-AUS dl
-45
...... r
(~v)......... ..................".....; ~ a. .
-40
.!'""~!3.~"
.
** l
-35
(a)
" P
" I
_ I
-30 -25 Latitude
-20
"
90
~; i
-15 14°S
150S
16*S
174°Wt
173°W
convergent component taken up by thrusting on faults parallel to the Alpine fault (Anderson et al. 1993).
130
•
'
•
15
172*W
171°W
(b) Fig. 8. (a) Plot of slip vector azimuths (small dots), GPS vector azimuths (large dots), the azimuth normal to the Tonga trench, the Pacific-Australia plate convergence direction (PAC-AUS), and the directions predicted by a pole of rotation between the north Tonga forearc and the Pacific based on GPS vectors (labelled GPS) (Bevis et al. 1995) and by a pole based on slip vectors (labelled SV). (b) Map of earthquake slip vectors (small, black arrows) beneath the northern Tonga forearc, the Pacific-Australia predicted direction (large, black arrow = 84 mm a-]) and the GPS vector (grey arrow --- 240 mm a-l).
rotation of the slip vectors resulted from a detached sliver, these events should be trenchward, NE, of the vectors that parallel plate motion.) Excluding these southerly trending slip vectors results in a near zero arc-parallel strain rate for the northern Tonga forearc, in agreement with GPS results of Bevis et al. (1995). Slip vectors east of northern New Zealand reveal a large degree of slip partitioning and v s values suggest approximately 30 m m a -l of right-lateral shear across the North Island, in agreement with geodetic measurements (Fig. 4d) (31 +__9 m m a-1 from Walcott 1984). In the South Island the plate vector becomes nearly parallel to the island and most of the motion occurs by oblique slip near the Alpine fault, although there is still a small
Philippines
The Philippines are a second region that Fitch (1972) regarded as revealing slip partitioning between the Philippine trench and the left-lateral Philippine fault. Using the Seno et al. (1993) PSP-Eurasia pole, the Philippine fault and trench system appears to be strongly partitioned (Barrier et al. 1991) (Fig. 4e). A difficulty in the kinematic analysis of the Philippines is the lack of a reliable plate convergence vector because the presence of several faults in the Philippines makes it unlikely that the Philippine islands are rigidly attached to Eurasia. A pole of rotation can be found between the forearc and the PSP that fits the Philippine trench slip vectors and removes a large part of the obliquity (Fig. 4e). Hence, we cannot be sure, on the basis of slip vectors alone, that the Philippines are partitioned. From combined trilateration and GPS, Duquesnoy et al. (1994) estimate a left-lateral slip rate on the Philippine Fault near l l°N of 26 + 10 m m a -1, which is consistent with the estimates of v s (Fig. 4e) based on the Philippine Sea-Eurasia pole of rotation (Seno et al. 1993). Hence, this first reliable slip rate for the Philippine fault supports the case for the Philippine system being largely partitioned. The arc-parallel strain rate for the Philippine forearc is poorly determined as the v s values do not vary smoothly with distance along the trench (Table 1; Fig. 4e).
R y u k y u trench
Slip vectors along the Ryukyu trench generally follow the Seno et al. (1993) PSP-Eurasia pole (Fig. 4f) and very little slip partitioning is evident even where the obliquity is high. If there is any arc-parallel slip of the forearc, it is of the order of 10 m m a-t or less. Near the centre of the Ryukyu trench (near 28°N), slip vectors are about 30 ° away from the plate vector even thought the obliquity is near zero (Fig. 4f). In this region the Oki Daito Ridge hits the trench (Fig. 1a) and probably causes the local deflection of slip vectors (these two points were excluded from estimates of strain rates).
I z u - M a r i a n a trench
Although the convergence direction is poorly known, the Marianas forearc shows strong partitioning in that slip Vectors remain nearly perpendicular to the trench despite large changes in
16
R. McCAFFREY
the obliquity (Figs 4g and 9). As observed at other trenches, at the south end of the Marianas trench there appears to be two groups of slip vector directions; one nearly perpendicular to the trench and another, comprising only 2 events, that is much more oblique (Fig. 9). The values of v s at the Mariana forearc are generally equal to v , (Fig. 4g), indicating that the upper plate absorbs" all of the arc-parallel component of plate motion by shear resulting in rapid arc-parallel extension (Table 1; Fig. 4g). At the Izu arc, the slip vectors are similar to the plate vector and no arc-parallel motion is required.
27 °N
23 °N
Conclusions Earthquake slip vectors at convergent margins of SE Asia are used to infer the degree of partitioning of slip on margin-parallel and margin-normal structures. W h e r e independent geological or geodetic estimates are available, slip vector-derived estimates of arc-parallel forearc slip rates appear accurate. Regions where oblique convergence is strongly partitioned are the Himalaya, Marianas, New Hebrides, Philippines, Sumatra, Markham Valley, and New Zealand. At some strongly curved margins, earthquake slip vectors are bimodally distributed between the convergence direction and the trench-normal direction, suggesting that slip partitioning occurs at small spatial scales. Several forearcs reveal arc-parallel deformation even where subduction is normal to the trench, showing that subduction is a three-dimensional process. I thank Tony Barber for many years of good work and cheer, the organizers of the symposium for allowing me to attend, and J. Milsom and D. Snyder for helpful reviews. Supported by NSF grants EAR-9105050 and EAR-9406917.
19°N
15°N
ll°N 139°E
143°E
147°E
Fig. 9. Map of earthquake slip vectors (small, black arrows) beneath the Izu-Mariana forearc, and the PSP-Pacific (Seno et al. 1993) predicted directions (large, black arrows).
References ABBOTT, L. D., SILVER, E. A. & GALEWSKY, J. 1994. Structural evolution of a modern arc-continent collision in Papua New Guinea. Tectonics, 13, 1007-1034. ABERS, G. & MCCAFFREY,R. 1988. Active deformation in the New Guinea Fold-and-Thrust Belt: seismological evidence for strike-slip faulting and basement-involved thrusting. Journal of Geophysical Research, 93, 13 332-13 354. & 1994. Active arc-continent collision: Earthquakes, gravity anomalies and fault kinematics in the Huon-Finisterre collision zone, Papua New Guinea. Tectonics, 13, 227-245. ANDERSON, H., WEBB, T. & JACKSON, J. 1993. Focal mechanisms of large earthquakes in the South Island of New Zealand: implications for the accommodation of Pacific-Australia motion. Geophysical Journal of the Royal Astronomical Society, 115, 1032-1054.
BARRIER,E., HUCHON,P. & AURELIO,M. 1991. Philippine Fault: A key for Philippine kinematics. Geology, 19, 32-35. BELLtER, O., SEBRIER,M. & PRAMUMIJOYO,S. 1991. La grande faille de Sumatra: geometrie, cinematique et quantite de deplacement raises en evidence par l'imagerie satellitaire. Comptes Rendus de l'Acad~mie des Sciences de Paris, 312, 1219-1226. BEVlS, M., TAYLOR, E W., SCHUTZ, B. E., RECY, J., ISACKS,B. L. ETAL. 1995. Geodetic observations of very rapid convergence and back-arc extension at the Tonga island arc. Nature, 374, 249-251. CURRAY,J. R. 1989. The Sunda Arc: A model for oblique plate convergence. In: VAN HINTE, J. E., VAN WEERING, TJ. C. E. & FORTUXN, A. R. (eds) Proceedings. Symposium Snellius-H Expedition, Jakarta, Nov. 23-28, 1989, Vol. 1, Geology and Geophysics of the Banda Arc and Adjacent Areas. Netherlands Journal of Sea Research, 24, 131-140.
SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES --,
HAMILTON, W. 1979. Tectonics of the Indonesian region. USGS Professional Paper, 1078. HARJONO, H., DIAMENT, M., DUBOIS, J. & LARUE, M. 1991. Seismicity of the Sunda Strait: Evidence for crustal extension and volcanological implications. Tectonics, 10, 17-30• HEGARTY, K. A., WEISSEL, J. K. & HAYES, D. E. 1983. Convergence at the Caroline-Pacific plate boundary: Collision and subduction. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Part 2. AGU Geophysical Monograph, 27, 349359. HUCHON, P. & LE P~CHON, X. 1984. Sunda Strait and central Sumatra fault. Geology, 12, 668-672. JARRARD, R. D. 1986. Relations among subduction parameters. Reviews of Geophysics and Space Physics, 24, 217-284. JOHNSTON, C. R. & BOWrN, C. O. 1981. Crustal reactions resulting from the mid Pliocene to Recent continent-island arc collision in the Timor region.
MOORE, D. G., LAWVER,L. A., EMMEL,E J., RAITT, R.W., HENRY,M. & KIECKHEFER,R. 1979. Tectonics of the Andaman Sea and Burma. In: WATKINS, J. S., MONTADERT,L. & DICKERSON, P. W. (eds)
Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 189-198. DARDJI, N., VILLEMIN, T. & RAMPNOUX, J.-E 1991. Cenozoic fault systems and paleostress along the Cimandiri fault zone, west Java Indonesia. In: PRASETYO, H. & SANTOSO(eds) The Silver Jubilee:
Symposium on the dynamics of subduction and its products. Proceedings LIPI Yogjakarta, 233-253. DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S. 1994a. Effects of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophysical Research Letters, 21, 2191-2194. & VOGT, E 1994b. Location of the AfricaAustralia-India triple junction and motion between the Australian and Indian plates: results from an aeromagnetic investigation of the Central Indian and Carlsberg ridges. Geophysical Journal International, 119, 893-930. DIAMENT, M., DEPLUS, C., HARJONO, H., LARUE, M., LASSAL, O., DUBOtS, J., & RENARD, V. 1990. Extension in the Sunda Strait (Indonesia): a review of the Krakatau programme. Oceanologica Acta, 10, 31-42. --, HARJONO, H., KARTA, K., DEPLUS, C., DAHRIN, D. ErAL. 1992. Mentawai fault zone off Sumatra: a new key to the geodynamics of western Indonesia. Geology, 20, 259-262. DUQUESNOY,TH., BARRIER,E., KASSER,M., AURELIO,M., GAULON, R. ET AL. 1994. Detection of creep along the Philippine Fault: First results of geodetic measurements on Leyte island, central Philippine. Geophysical Research Letters, 21, 975-978. EKSTROM, G. & ENGDAHL,E. R. 1989. Earthquake source parameters and stress distribution in the Adak Island region of the central Aleutian Islands, Alaska. Journal of Geophysical Research, 94, 15 499-15 519. FITCH, T. J. 1972. Plate convergence, transcurrent faults and internal deformation adjacent to southeast Asia and the western Pacific. Journal of Geophysical Research, 77, 4432-4460. GEIST, E. L., CHILDS, J. R. & SCHOLL, D. W. 1988. The origin of summit basins of the Aleutian Ridge: Implications for block rotation of an arc massif. Tectonics, 7, 327-341. GENRICH, J., BOCK, Y., MCCAFFREY, R., CALAIS, E., STEVENS, C. E T A L . 1994. Kinematics of the eastern Indonesian island arc estimated by Global positioning system measurements. EOS, 75, 162. GILBERT, L., KASTENS, K., HURST, K., PARADISSIS, D., VEIS, G. ET AL. 1994. Strain results and tectonics from the Aegean GPS experiment. EOS, 75, 116. GOLDFINGER, C., KULM, L. D., YEATS, R. S., APPLEGATE, B., MACKAY, M. E. & MOORE, G. E 1992. Transverse structural trends along the Oregon convergence margin: Implications for Cascadia earthquake potential and crustal rotations. Geology, 20, 141-144.
17
BMR Journal of Australian Geology & Geophysics, 6, 223-243. KARIG, D. E., LAWRENCE,M. B., MOORE, G. E & CURRAY, J. R. 1980. Structural framework of the fore-arc basin, NW Sumatra. Journal of the Geological Society, London, 137, 77-91. KULIG, C., MCCAFFREY, R., ABERS, G. A. & LETZ, H. 1993. Shallow seismicity of arc-continent collision near Lae, Papua New Guinea. Tectonophysics, 227, 81-93. MALOD, J. A., KARTA, K, BESLIER, M. O. & ZEN, M. T. 1996. From normal to oblique subduction: Tectonic relationships between Java and Sumatra. Journal of SE Asian Earth Science, 12, 85-93. MATSUZAKA,S., TOBITA,n . , NAKAHORI,Y., AMAGAI,J. & SUGIMOTO, Y. 1991. Detection of Philippine Sea plate motion by very long baseline interferometry. Geophysical Research Letters, 18, 1417-1419. MCCAFFREY, R. 1988. Active tectonics of the eastern Sunda and Banda arcs. Journal of Geophysical Research, 93, 15 163-15 182. 1991. Slip vectors and stretching of the Sumatran forearc. Geology, 19, 881-884. 1992. Oblique plate convergence, slip vectors, and forearc deformation. Journal of Geophysical Research, 97, 8905-8915. 1994. Global variability in subduction thrust zoneforearc systems. Pure and Applied Geophysics, 141, 173-224. • & ABERS, G. A. 1991. Orogeny in arc-continent collision: the Banda arc and western New Guinea. Geology, 19, 563-566 --. & NABELEK, J. 1986. Seismological evidence for shallow thrusting north of the Timor trough.
-
-
-
-
-
-
Geophysical Journal of the Royal Astronomical Society, 85, 365-381. MCKENZIE, D. & JACKSON, J. 1983. The relationship between strain rates, crustal thickening, paleDmagnetism, finite strain, and fault movements within a deforming zone. Earth and Planetary Science Letters, 65, 182-202. MOLNAR,P. & LYON-CAEN,H. 1989. Fault plane solutions of earthquakes and active tectonics of the Tibetan
18
R. McCAFFREY Plateau and its margins. Geophysical Journal
International, 99, 123-153. --,
ENGLAND, P. & MARTINOD, J. 1993. Mantle dynamics, uplift of the Tibetan Plateau, and the Indian monsoon. Review of Geophysics, 31, 357-396. MOUNT, V. S. & SUPPE, J. 1987. State of stress near the San Andreas fault: Implications for wrench tectonics. Geology, 15, 1143-1146. PLATT, J. P. 1993. Mechanics of oblique convergence. Journal of Geophysical Research, 98, 1623916256. PUNTODEWO, S. S. O., MCCAFFREY, R., CALAIS, E., BOCK, Y., RAIS, J. ET AL. 1994. GPS measurements of crustal deformation within the Pacific-Australia plate boundary zone in Irian Jaya, Indonesia. Tectonophysics, 237, 141-153. RANKZN, B., CAm3WELL, R. K. & KARIG, D. E. 1984. Kinematics of the Philippine Sea Plate. Tectonics, 3, 555-575. SCHOUTEN, H., & BENES, V. 1996. Post-Miocene collision of Australia and the Bismarck arc, western equatorial Pacific. Earth and Planetary Science Letters, in press. SENO, T., STEIN, S. & GRIPe, A. 1993. A model for the motion of the Philippine Sea plate consistent with NUVEL-1 and geological data. Journal of Geophysical Research, 98, 17 941-17 948. SIEH, K., ZACHARIASEN, J., BOCK, Y., EDWARDS, L., TAYLOR, E & GANS, P. 1994. Active tectonics of Sumatra. GSA Abstracts with Programs, 26, A-382. SILVER, E. A., REED, D., MCCAEFREY, R., & JOWOD1WIRYO, Y. 1983. Back-arc thrusting in the eastern Sunda Arc, Indonesia: a consequence of arc-continent collision. Journal of Geophysical Research, 88, 7429-7448.
SNYDEr, D. B., PRASETYO, H., TJOKROSAPOETRO, S., BLUNDELL,D. J., BARBER,A. J., RICHARDSON,A. N. &; MILSOM, J. 1993. A deep seismic reflection transect across the youngest mountain range in the world, the active convergent margin between the Australian craton and the Banda volcanic arc near Timor, Indonesia. EOS, 74, 444. STEVENS, C., PUNTODEWO, S. S. O., MCCAFFREY, R., BOCK, Y. t~ CALAIS, E. 1995. Out of the pot and into the frying pan: A Bird's Head's view of tectonic escape. International Union of
Geodesy and Geophysics, XXI General Assembly, p. A38. TREGONING,P., BRUNNER,E K., BOCK, Y,. PUNTODEWO,S. S. O., McCAFFREY, R. ET AL. 1994. First geodetic measurement of convergence across the Java Trench. Geophysical Research Letters, 21, 21352138. WALCOTT, R. I. 1984. The kinematics of the plate boundary zone through New Zealand: a comparison of short- and long-term deformations. Geophysical Journal of the Royal Astronomical Society, 79, 613-633. WEISSEL, J. K. • ANDERSON, R. N. 1978. Is there a Caroline plate? Earth and Planetary Science Letters, 41, 143-158. Yu, S.-B. & CHEN, H.-Y. 1994. Global positioning system measurements of crustal deformation in the Taiwan arc-continent collision zone. Terrestrial, Atmospheric and Oceanic Sciences 5(4), 477-498. ZWICK, P. & MCCAFFREY, R. 1991. Seismic slip rate and direction of the Great Sumatra Fault based on earthquake fault plane solutions. EOS, 72, 201.
The Sumatra margin: oblique subduction and lateral displacement of the accretionary prism JACQUES
ANDRI~ MALOD 1 & BADRUL
MUSTAFA
KEMAL 2
1 GEMCO, URA 718 du CNRS, UPMC, Case 129, 4 pl. Jussieu, 75252 Paris, France 2 Padang University, 25117 Padang, Indonesia Abstract: The Sumatra active margin extends more than 1600 km from northwest to southeast. The almost north-south direction of subduction is normal in front of Java and oblique in front of Sumatra. The transition between the two regimes of subduction occurs south of the Sunda Strait. A structural and stratigraphical study of the forearc domain, based on data collected during a French-Indonesian programme of cooperation in oceanography, confirms that the convergent motion is partitioned into a convergent motion, more or less perpendicular to the trench, and a strike-slip motion parallel to the trench. This latter motion is taken up along two major faults, the Sumatra and the Mentawai faults. At its northern end, the Mentawai fault is attenuated and seems to terminate within the accretionary prism. It is relayed and connected to the Sumatra fault by the Batee fault. This pattern can be explained by a simple model with two sliver plates: the Mentawai and Aceh sliver plates, on the top of which the forearc basin has developed. The accretionary prism itself is moving northwestwards along the Mentawai fault. Because there is no major evidence of extension within the Mentawai plate, we conclude that the motion along the Sumatra fault may occur at a uniform rate south of latitude 3°N. Strike-slip motion along the Mentawai fault may be explained by relatively better coupling between the subducting slab and the upper plate beneath the accretionary prism compared to that beneath the forearc, perhaps because of the large oceanic structures entering the subduction zone.
Oblique subduction induces the partitioning of the convergent motion into two components, parallel and perpendicular to the subduction zone (Fitch 1972; Jarrard 1986; McCaffrey 1991, 1992). This partitioning results from the coupling effect between the dipping oceanic slab and the upper plate lithosphere. This coupling effect can be identified by study of slip vectors at the subduction zone which indicate the true direction of relative movement of the oceanic slab with respect to the forearc (Jarrard 1986; McCaffrey 1992): In the case where the margin is curved, partitioning increases with obliquity (Fig. 1). If the partitioning is not complete the slip vector will be situated between the convergence vector and the direction perpendicular to the margin. The lateral movement results in the formation of a sliver microplate while the movement perpendicular to the margin, mainly taken up by subduction, is also expressed by compressional deformation within the accretionary prism or the forearc. Marine geophysics cruises, in cooperation with Indonesia, allowed a detailed study of the Sumatra margin to be undertaken (Diament et al. 1992; Malod et al. 1993; Mustafa Kemal 1993; Zen Jr. 1993). The object of this paper is to propose a model for the distribution of the movements in the whole forearc area.
~. plate
k
Oceanic plate
Fig. 1. Partitioning of movements during oblique subduction increases with margin curvature. Full partitioning is shown for sake of clarity.
Structures related to the subduction off Sumatra The subducting oceanic lithosphere of the Wharton basin has a variable age and structure (Fig. 2). Its age ranges from more than 110 Ma in the eastern part of the area, along the Java trench, to approximately 50 Ma south of central Sumatra where the Wharton extinct spreading ridge is being subducted
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 19-28.
Upper plate
~.
19
20
J . A . M A L O D & B . M. KEMAL
1 0 ~ * E
,amai:se; • I:E~ensi°n460
I
1 llO*E
break
: i i i kin;umatra
i .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
;.-C) Sunda
Oblique I Subduction :i'i!i:
~:~~~......... Subduction slab IO°S :~";i;iii~:;;i: : 400-450 km ...;
.
Subduction I
I
/iiiiiiiiiiiiiill i:i\::iiiiiiii:iiii
mccumu'a'n 1
~liiiiiiiiiiiiiii
a-'ill
31;iiiiii;ii;i:ii
--o o ..........
_
14
. .
.
. .
. .
. .
. .
. .
. .
.
.
.
.
. .
. .
. .
. .
. .
. .
. .
Wharton extinct spreading centre . . . . . . . . ~ . . . . . . . . . . . . .
i
-
~5°E
i iiliSundaStrait ~ Extension ~,~ 5 0 - 7 0 km
I Zoneo increasing
+ Ablation
100"E [
il .......
. . . . .
constant obliquity
[
"
....
-- 5os
I
"
' " " iE~'.eO~d.nl . . . . . . . . ) ........... . . 4Fi
I ans'e, I
I
7
I
I
_
t
l
obliquity I
[
t
~?5"E I
I
Fig. 2. The geodynamic framework of Sumatra subduction. Note the Wharton extinct spreading axis and the Investigator fracture zone entering the subduction. The Sumatra and Mentawai faults are the tectonic response to the obliquity of subduction. A schematic distribution of the movements in the forearc is shown. Thick arrows indicate motions with respect to Sumatra mainland. Thin arrows give the relative motion along strike-slip faults. As pointed out in the text, the numerical values are realistic but only indicative, based on published studies. The inset shows the slab geometry indicated by earthquake depths. The stippled areas are high velocity zones reflecting a subducted slab at depths of 400--450 km according to the seismic tomography of Puspito et al. (1993).
THE SUMATRA MARGIN with a velocity of approximately 60-70 mm a-1 (DeMets et al. 1990). The geometry of the dipping oceanic slab can be deduced from the distribution of earthquakes linked to subduction (Fig. 2). The slab curvature follows the island arc of Java and Sumatra and shows some minor rupture zones to the south of Sumatra (Zen Jr. 1993; R. Louat, pers. comm. 1994). Further north in the region of Lake Toba, a major rupture occurs, which is also indicated by seismic tomography (Puspito et al. 1993). This feature corresponds to a marked change in orientation of the continental margin and the subduction trench which follow the shape of the slab and turn to a westward orientation before resuming a NNW direction. The direction of convergence indicated by global kinematic models is 024 ° to 025 ° (DeMets et al. 1990). In fact, the study of slip vectors shows that the subduction is normal beneath Java, where the average of 15 vectors indicates a 003 ° direction (McCaffrey 1991). The Sunda arc therefore allows observation of the change from frontal subduction to oblique subduction (Fig. 2). Subduction becomes increasingly oblique to the margin as the trench turns from E-W to NW-SE west of Java. One can therefore expect that partitioning of the movement will result in an increasing lateral component of motion. The obliquity of subduction is constant along the main linear portion of the Sumatran margin. The lateral component of the movement is taken up here by the great dextral fault of Sumatra, which extends from the south to the north of Sumatra (Katili 1970). McCaffrey (1992) showed that, correlated with the curvature of the margin, slip vectors indicate that the lateral movement increases gradually along the Sumatra fault. Consequently, he predicted a longitudinal extension of the forearc sliver plate between the trench and the Sumatra fault. This model may partly explain the discrepancy between extension at the two ends of the Sumatra fault, from more than 460 km in the Andaman Sea (Curray et al. 1978) compared to only 50-70 km in the Sunda Strait (Huchon & Le Pichon 1984; Lassal et al. 1989; Diament et al. 1992).
Boundaries of forearc microplates The first result of the study of single channel seismic profiles was the discovery that the faulted zone situated between the accretionary prism and the forearc, previously interpreted as a flexure or backthrust (Karig et al. 1978) is linear, has a highly variable tectonic aspect and looks like a flower structure in some places. Therefore, this structure was re-interpreted as the trace of a strikeslip fault zone, the Mentawai fault, which has also
21
accommodated some compression (Diament et al. 1992; Zen Jr. 1993). The Mentawai fault seems to have been active during the Neogene, but it is not possible to identify the relative importance of strike-slip and compressional movements. Furthermore, the amount of lateral motion cannot be measured from seismic profiles.
The Mentawai microplate
The Mentawai fault (Fig. 3) extends from the south of Sumatra up to the south of Nias. Therefore the Sumatra fault and the Mentawai fault bound an elongated sliver microplate, the Mentawai plate, which corresponds to the forearc domain. Evidence for extensional deformation within this plate is weak since no major normal faults are observed to cross the Mentawai plate. There is no significant indication of the extension predicted in previous models (Jarrard 1986; McCaffrey 1992) or that necessary to explain the difference in the amount of extension between the north and south of Sumatra. There remains the possibility of distributed extension by means of small extensional faults. However, recent faults with very small throws are found only in a 50 km long zone southwest of Sumatra where recent seismicity is concentrated (Zen Jr. 1993; Fig. 4). To the north of Nias island, the NNW-SSE striking Batee fault connects the Mentawai fault and the Sumatra fault (Beaudry & Moore 1985; Matson & Moore 1992). On land where it seems not to have been very active in recent time, its dextral strike-slip movement is documented from Spot images (C. Detourbet, pers. comm. 1994). At sea, it appears as a vertical fault with a large normal throw explaining the Nias basin formation (Fig. 5). The transfer of the major part of the movement on the Sumatra fault may in fact occur by means of several faults in the Nias basin (Matson & Moore 1992) or on its southern ~rim (archipelago of Pini island; Fig. 3). Further north, movement would therefore be concentrated on the Sumatra fault, as suggested by the opening of the Andaman Sea north of Sumatra (Fig. 2). Accordingly, to the north of Nias island, the forearc area moves significantly with respect to the main part of Sumatra. The Sumatra fault forms the northern boundary of a pull-apart in the Sunda Strait (Harjono et al. 1988, 1991, 1993; Lassal et al. 1989; Diament et al. 1990). The dextral strike-slip movement is relayed to the south by the Ujung Kulon fracture zone which extends into the accretionary prism (Fig. 6). To the south of the Sunda Strait the Mentawai fault becomes difficult to follow in the accretionary prism. A progressive initiation of the Mentawai fault could explain why it is no longer strongly
22
J . A . MALOD •
Sumatra
/
B. M. KEMAL
I
I
I •
i J~
Et
I
Indo-Australial plate
\
Aceh basin :" ::
\,,,
ION
SUMATRA i ¸¸
Simeulu~
~
!
"~
:
~!i!~i!:
Fault
~iiiiiiiiiii,iiiiiiiiiiiiiii~iiiii!iiiiiiiiiiiiiii lON
I
Thrusts ~1
0ON ~7
Strikeslip !
Subductiontrench
",,,,,,,./ Structures of the oceanic crust 94°E
95OE
Wharton fossil I oceanic |,,.,jr 96OE ridge
| 97°E
\
98°E
Fig. 3. Structural sketch of the northern margin of Sumatra, with the delineation of the main sliver plates.
evident in the structures to the south of the Sunda Strait. The Mentawai microplate is greater than 1000 km in length (Fig. 2). Its existence shows that the accretionary prism has moved relative to the forearc domain. The southern limit of the Mentawai plate is coincident with the zone where subduction changes from oblique along Sumatra to orthogonal along Java. The Mentawai plate ends to the north of Nias, where the subducting slab is interrupted by a large discontinuity.
The Aceh microplate North of Nias, the Sumatra fault, the northern extension of the Mentawai fault, and the southern extension of the West Andaman fault (Curray et al. 1978; Izart et al. 1994), bound another sliver plate, the Aceh microplate (Fig. 3). The Mentawai fault may have played a role in the past, but its actual prolongation does not currently show evidence of active strike-slip movement. Backthrusting of the accretionary prism over the forearc basin, along the
THE SUMATRA MARGIN
23
NW
SE
3.0
3.5
4.0
4.5 s.d.t.t
t
5 km
Fig. 4. Recent vertical normal faults associated with active seismicity southeast of Sumatra (location on Fig. 2).
southern continuation of the West Andaman fault, is the probable effect of flexural bending of the forearc basement under the accretionary prism, which would explain the deepening of the forearc basin toward the west (Malod et al. 1993) (Fig. 7). In the longitudinal direction, the forearc is made up of large blocks tilted southwards. Some part of the movement is absorbed by internal deformation of the Aceh microplate as indicated by the Banyak Island thrust zone, active in latest Pliocene (Fig. 7) or the presently active Tuba anticline ridge.
of sliver plate. However, deformation within the prism is difficult to study. Northwest of Nias island, several NW-SE fault zones, which are ramifications of the Mentawai fault, cross the accretionary prism (Fig. 7). The tectonic pattern of these faults suggests strike-slip motion (although mud diapirism might be possible). To the north of this zone, the accretionary prism may be a part of the Andaman Island and Burma plate. Where the Mentawai plate is present, south of Nias Island, we recognize three different zones from south to north (Fig. 2), as follows.
Transfers in the accretionary prism along Sumatra
1. Extension and material ablation in the south The displacement towards the northwest of the Mentawai microplate as a single block is related to the extension in the Sunda Strait and perhaps further south within the forearc itself. At the same time, oceanic sediments accreted to the deeper part
The definition of Mentawai and Aceh plates leads to a consideration of the accretionary prism between the trench and the Mentawai fault as a kind
W
-o.s~~atee
I
5 krn
fault
I
Fig. 5. The Batee fault. Note the large normal throw along this vertical fault (location on Fig. 4).
E
24
J.A. MALOD • B. M. KEMAL accretionary prism or the Sumatra mainland to the northeast of the Sumatra fault, and mechanisms of movement partitioning in the prism and the forearc. However, the general tectonic disposition of the Sumatra margin can be integrated into a simple model, which actually involves the transfer of material at the edge of Sumatra.
Model The boundaries, the distribution and motion of the plates and microplates are illustrated on a simplified structural diagram (Fig. 2).
Sumatra fault and Mentawai fault
Fig. 6. Simplified structural sketch of the southern prolongation of the Mentawai and Sumatra faults. The Ujung Kulon fault zone relays the Sumatra fault system to the south of the Sunda Strait pull-apart. The Mentawai fault zone prolongation indicated by a dashed line is not observed in this region.
of the margin must have been displaced towards the northwest. These two mechanisms, of extension in the basement and removal of sediments from the accretionary prism, contribute to the concave shape of the margin south of the Sunda Strait.
2. Transfer in the central part In the linear part of the margin west of Sumatra, the Mentawai plate and the accretionary prism are displaced towards the northwest. There is relative movement between the plate and the prism along the Mentawai fault. The amount of relative motion and its timing during the Neogene cannot be established. 3. Accumulation and relay zone in the north
In the northern part of the region, according to our interpretation, movement was concentrated mainly on the Sumatra fault, while materials of the prism accumulated in several slices west of Simeulue island. This process occurred as if the subduction was locally frontal, in agreement with the local E-W orientation of the oceanic slab, and is consistent with the horsetail shape of the Mentawai fault in this region and its reduced role further north. This description of the Sumatra oblique subduction and transfer process leaves some unanswered questions, such as possible deformation within the
These two faults have different natures (Fig. 8). The Sumatra fault is an intra-continental strike-slip fault and its geometry can be followed on land and at sea (Fig. 6). It coincides with the volcanic arc which suggests a connection between volcanism and tectonics (Bellier & Srbrier 1994), the volcanic zone being a zone of weakness. The Mentawai fault appears more like a boundary between the prism constituted of accreted oceanic sediments and the forearc basin established on a crust of intermediate or continental type as shown by refraction (Kieckhefer et al. 1980) or expanding seismic profiles (Liu & Curray 1988; Fig. 8). This fault is not necessarily vertical and the compressional component may explain its complicated surface pattern. This fault separates the relatively rigid area of the Mentawai microplate, as shown by the lack of major deformation, from the complex accretionary prism, possibly more deformable because of the existence of numerous thrust slices.
The geometry o f the margin Down to a depth o f 20 or 25 km, the subduction contact occurs between the oceanic slab and the accretionary prism (Fig. 8). Coupling along this contact, assuming an active Mentawai strike-slip fault, induces partitioning of the movement that drags the prism towards the northwest. At greater depths beneath the margin, the coupling occurs between the oceanic slab and the lithosphere of the forearc. Consequently, the partitioning of the movement induces two decoupling zones: the Mentawai fault and the Sumatra fault. In a horizontal plane, the relative movements depend on the importance of the partitioning of the movement in each of these two zones. The Mentawai fault will have dextral strike-slip movement if coupling is more efficient under the accretionary prism than under the forearc basin (Fig. 9); this seems to happen along the rectilinear part of the fault and the Mentawai plate. If the
THE SUMATRA MARGIN
3.or =W-_~_AAccretionary ~ ~~j~rism
25
Aceh basin
E
3.5 I
4.0 I
4.5 I 5"sOd.t.t.
N
Banyak thrust
~
i i
'_2._ . . . . . . . . . . . .
,
,i 2.0
. . . . . . ~Late Miocene
~
"~''--.'-'
-
:.7"-.~." ..:~/- - 7 : " , - , ;.:- , :;..:.;:2,~. -'-.~-~ ..'.;ZI-... - ., ..;.~ Middle-Late M i o c e n e " : ...... " " " ' " ': " =" "
SW
2.5
~
-
~
~
"-
N
~ ~ . ~'.-
2.5 J
s.d.t.t. _ 3.0
;.~
3.0
3.5 -s.d.t.t.
•
5 krn
I
Fig. 7. Seismic reflection profiles showing deformations around the Aceh microplate (location on Fig. 4): (a) backthrust of the accretionary prism over the Aceh basin; (h) Late Pliocene northward thrust of Banyak Island; (c) possible strike-slip faults in the accretionary prism forming part of a horsetail termination of the Mentawai fault.
partitioning of movement is the same in both regions there will be no relative movement between prism and forearc (Fig. 9). This may be the case to the south of the Sunda Strait and could explain the
poor expression or even the apparent absence of the Mentawai fault in this region. The occurrence of the Mentawai fault and its continuation into the Batee fault indicate that the
26
J.A. MALOD Be; B. M. KEMAL Mentawai fault 0
Sumatra fault
Accretionaryprism l F°rearc coasl i,~ VolcancJarc ~ ~ lOaSl(I ~...... ......................
50
.........
Sumatra
......... :
Fig. 8. Schematic cross-section, without vertical exaggeration, of the Sumatra subduction system.
first case (Fig. 9) is most likely over most pans of the Sumatra forearc. The coupling between the subducting slab and the overriding forearc seems to be more efficient in the prism area than in the forearc basin. This may be the result of several processes including: (1) the accretionary prism may load the downgoing plate and increase the coupling effect; (2) the deeper part of the accretionary prism may have become better consolidated when subruction was more frontal in the EoceneOligocene time; (3) the large structures of the oceanic sea floor
entering the subduction zone, like the Investigator Ridge and the Wharton extinct spreading ridge, may induce a strong coupling between the slab and the accretionary prism. In any case, since the accretionary prism rests directly on the oceanic slab, its lateral transfer is likely to occur provided there is decoupling along the Mentawai fault.
Movement distribution A quantification of motions is possible only for the most recent period of time during which measure-
Sumatra
Sumatra
fault~
Mentaf : i l t ~ fault~
~L ix
T.rench ~ , , % ~
~
~
Prism
Obliquedirection of subduction
Q
Dextralstrike-slip
T~nc.
\ X xv
\
"X.\
,
Fore c
~
Prism
Obliquedirection of subducUon
Q
No strike-slip
Fig. 9. Sketch showing two possibilities of motion distribution by decoupling: (a) the situation in central and north Sumatra with the forearc and prism moving at different rates; (b) the situation south of the Sunda Strait, where it is proposed that the forearc and prism are moving at the same rate.
THE SUMATRA MARGIN ments have been made. Measurements of the recent displacement on the Sumatra fault have been made essentially by means of satellite photographs (Bellier et al. 1991, 1993; Bellier & S6brier 1994). These measurements indicate a velocity of displacement that grows from 6 _ 4 m m a-t in the south close to the Sunda Strait (Bellier et al. 1991 ), to 19 _+ 2 m m a -1 in the Singkarak region at 1°S (Bellier et al. 1993), and to about 23 mm a -i in the Lake Toba region at 2°30'N (Detourbet et al. 1993). The value of 6 _ 4 m m a -1 in the region close to the Sunda Strait is low and might be explained by recent extensional deformation occurring in the forearc (Fig. 4). Along the Batee fault a value of 11_ 2 mm a-l is proposed (C. Detourbet, pets. comm. 1994). Figure 9 shows a possible distribution of the movements that would explain the observed structures and rates of Fig. 2. In order to give some numerical values 20 mm a -1 is chosen for the main Sumatra fault segment. Accordingly movement on the northern part of the Sumatra fault is fixed to the sum of this value plus the strike-slip component on the Batee fault, which is 11 m m a -t, yielding 31 m m a-1. These numerical values serve mainly to illustrate the model, and should not be regarded as absolute values. In central and north Sumatra movement is distributed between the Mentawai and Sumatra faults with the forearc and prism moving at different rates (Fig. 9a). In the region south of the Sunda Strait we propose that the forearc and prism are moving at the same rate (Fig. 9b). At the the northern end of the Sunda forearc the movement is absorbed in three ways: opening of the Andaman
27
Sea, westward migration of the Andaman arc that overlaps the Indian ocean oceanic crust, and finally in backthrusting of the accretionary prism over the forearc basin belonging to the Aceh microplate.
Conclusions The authors observations suggest that the degree of partitioning of movement may be different beneath the accretionary prism and the forearc domain. In the most representative portion of the Sumatra margin with respect to oblique subduction, partitioning is more efficient in the accretionary prism than in the forearc region. This may result from the importance of the oceanic structures entering the subduction zone beneath Sumatra. The Mentawai microplate movements are now well understood and partly quantified for recent periods, but those of the accretionary prism necessitate further studies, particularly to look at possible internal deformation. Quantification of the movements also needs further investigation. We thank all the people who participated in the data collection, specially the Commandant Soepangat and his crew on the Indonesian R.V. Baruna Jaya IlL This research was conducted during an Indonesian-French cooperative programme in oceanography. We acknowledge the help of Mr. M. Pain, attach6 scientifique at the French Embassy in Jakarta. This study has been funded by IFREMER, the French Embassy in Jakarta, CNRS-INSU (grant ATP 933909) and UniversitE Pierre et Marie Curie in France and BPPT in Indonesia. S. J. Matthews and M. A. Samuel reviewed the paper and made many valuable comments.
References BEAUDRY,D. & MOORE, G. E 1985. Seismic stratigraphy and Cenozoic evolution of west~Sumatra forearc basin. AAPG Bulletin, 69, 742-759. BELLmR, O. & SE;BmER,M. 1994. Relationship between tectonism and volcanism along the Great Sumatran Fault Zone deduced by SPOT image analyses. Tectonophysics, 233, 215-231. , - - , D~TOURBET, C. & PRAMUMIJOYO,S. 1993. Long term dextral slip rate along the Sumatran fault system. Terra Abstracts', 1, 253, EUG VII, 4-8 April Strasbourg. , & PRAMUMnOYO,S. 1991. La grande faille de Sumatra: gEomEtrie, cinEmatique et quantitE de dEplacement mises en Evidence par l'imagerie satellitaire. Comptes Rendus de l'Acadgrnie des Sciences de Paris, 312, Sgrie II, 1219-1226. CURRAY, J. R., MOORE, D. G., LAWVER, L. A., EMMEL, E J., RAITT, R. W., HENRY,M. & KIECKHEFFER,R. 1978. Tectonics of the Andaman Sea and Burma. In: WATKINS,J. S., MONTADERt,L. & DICKENSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 189-198.
DEMETS, C., GORDON, R. G., ARGUS, D. E & STEk~, C. 1990. Current plate motions. Geophysical Journal International, 10L 425-478. DETOUm3ET, C., BELL~r, O. & SEBmER, M. 1993. La caldeira volcanic de Toba et le systbme de faille de Sumatra (IndonEsie) vues par SPOT. Comptes Rendus de l'Acad(mie des Sciences de Paris, 316, 1439-1445. DIAMENT, M., DEPLUS, C., HARJONO, H., LARUE, M., LASSAL, O., DUBOIS, J. & RENARD, V. 1990. Extension in the Sunda Strait (Indonesia): Review of the Krakatau programme. Oceanologica Acta, 10, 31-41. ~, HARJONO,H., KARTA,K., DEPLUS,C., DAHRIN,D., ZEN, M. T. J. & MALOD,J. A. 1992. Mentawai fault zone off Sumatra: A new key to the geodynamics of western Indonesia. Geology, 20, 259-262. FITCH, T. J. 1972. Plate convergence, transcurrent faults, and internal deformation adjacent to Southeast Asia and the Western Pacific. Journal of Geophysical Research, 77, 4432.4460. HARJONO, H., DIAMENT, M., DUBOIS, J. & LARUE, M. 1988. SEisrnicitE du detroit de la Sonde (IndonEsie):
28
J . A . MALOD t~Z B. M. KEMAL
presentation des rEsultats d'un rdseau local. Comptes Rendus de l'Acaddmie des Sciences de Paris, 307, SErie II, 565-571. --, --, , -& ZEN, M. T. 1991. Seismicity of the Sunda Strait: evidence for crustal extension and volcanological implications. Tectonics, 10, 17-30. , SI~BRIER, M. & DIAMENT, M. 1993. Correction and addition to "Seismicity of the Sunda Strait: evidence for crustal extension and volcanological implications" by Hery Harjono, Michel Diament, Jacques Dubois, Michel Larue and Mudaham Taufick Zen. Tectonics, 12, 787-790. HUCHOn, E & LE PICHON, X. 1984. Sunda Strait and Central Sumatra Fault. Geology, 12, 668-672. IZART, A., MUSTAFA KEMAL, B. & MALOD, J. A. 1994. Seismic stratigraphy and subsidence evolution of the northwest Sumatra forearc basin. Marine Geology, 122, 109-124. JARRARD, R. D. 1986. Terrane motion by strike-slip faulting of forearc slivers. Geology, 14, 780-783. KARIG, D. E., SUPARKA,S., MOORE, G. E & HEHANUSSA, E E. 1978. Structure and Cenozoic evolution of the Sunda arc in the central Sumatra region In: Watkins, J. S., MONTADERT, L. & DICKENSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 223-237. KATILI, J. A. 1970. Large transcurrent faults in southeast Asia with special reference to Indonesia. Geologische Rundschau, 59, 581-600. KIECKHEFER, R. M., SHOR JR., G. G., CURRAY, J. R., SUGIARTA, W. & HEHUWAT, F. 1980. Seismic refraction studies of the Sunda trench and forearc basin. Journal of Geophysical Research, 85, 863-889. LASSAL, O., HUCHON,P. & HARJONO,H. 1989. Extension
crustale dans le detroit de la Sonde (IndonEsie), donnEes de la sismique rEflexion (Campagne KRAKATAU). Comptes Rendus de l'Acaddmie des Sciences de Paris, 309, SErie II, 205-212. LIU, C. S. & CURRAV, J. R. 1988. Structure and nature of the forearc basin of west central Sumatra. EOS, 69, 1443. MALOD, J. A., MUSTAFA KEMAL, B., BESLIER, M. O., DEPLUS, C., DIAMENT,M. ETAL. 1993. DEformations du bassin d'avant-arc au nord-ouest de Sumatra. Une rEponse ~ la subduction oblique. Comptes Rendus de l'Acaddmie des Sciences de Paris, 316, SErie II, 791-797. MATSON, R. G. & MOORE, G. F. 1992. Structural influences on Neogene subsidence in the central Sumatra forearc basin. In: WATKINS,J. S., ZHIQIANG, F. ET AL. (eds) Geology and Geophysics of Continental Margins. AAPG Memoir, 53, 157-181. MCCAFFREY, R. 1991. Slip vectors and stretching of the Sumatran forearc. Geology, 19, 881-884. 1992. Oblique plate convergence, slip vectors, and forearc deformation. Journal of Geophysical Research, 97, 8905-8915. MUSTAFA KEMAL, B. 1993. La marge active au NordOuest de Sumatra. Mdcanismes gdodynamiques de tran~fert lids glla subduction oblique. Th~se de l'UniversitE E et M. Curie. Paris. PUSPITO, N. T., YAMANAKA,Y., MIYATAKE,Z., SHIMAZAKI, K. & HIRAHARA, K. 1993. Three-dimensional P-wave velocity structure beneath the Indonesian region. Tectonophysics, 220, 175-192. ZEN JR., M. T. 1993. DEformation de l'avant-arc en rdponse ?tune subduction ?1 convergence oblique. Exemple de Sumatra. Th~se de l'UniversitE P. et M. Curie, Paris.
Collision and strike-slip faulting in the northern Molucca Sea (Philippines and Indonesia): preliminary results of a morphotectonic study CLAUDE
R A N G I N 1, D A H A R T A & THE MODEC
D A H R I N 2, R A Y Q U E B R A L 3
SCIENTIFIC
PARTY
ANNE GAELLE BADER, JEAN PAUL CADET, GLEN CAGLARCAN, BENOIT DEFFONTAINES, CHRISTINE DEPLUS, ERNESTO G. CORPUS, ROBERT HALL, YANN HELLO, JACQUES MALOD, SERGE LALLEMAND, DOMINGO B. LAYUGAN, RI~MY LOUAT, REYNALDO MORALES, KEITH PANKOW, MANUEL PUBELLIER, MICHEL POPOFF, REYNALDO T. RODELAS & TRAMANADI YUDHO
1 URA 1759 CNRS, Ddpartement de Gdotectonique, T 26-0 El, Universitd Pierre et Marie Curie, 4 Place Jussieu, 75252 Paris, France e ITB Bandung, Indonesia 3 Mines and Geosciences Bureau, Manila, Philippines Abstract: A swath mapping, gravity and single channel seismic survey was carried out in the northern Molucca Sea with R.V. L'Atalante. Preliminary interpretation of these data reveals the presence of an almost complete Sangihe arc and forearc. The Miangas-Pujada-Talaud ridge in the central part of the Molucca Sea appears to be a backstop within the Sangihe forearc. East of the ridge very contrasting terranes are separated by a major NW-SE crustal discontinuity interpreted as a left-lateral strike-slip fault. North of 6°N the Philippine Trench inner wall is dissected by NW-SE trending left-lateral strike-slip faults, resulting from the dominantly oblique convergence between colliding arcs. South of 6°N the westward subduction of the buoyant Snellius volcanic plateau, a fragment of the Halmahera arc terrane, has induced the formation of a new plate boundary, the Philippine Trench along what is interpreted as a former strike-slip fault zone. East of the Miangas-Pujada-Talaud ridge there is a wide sedimentary wedge separated from the Snellius Ridge to the south by the major NW-SE crustal discontinuity. The thickness of this wedge cannot be explained by subduction along the very young Philippine Trench. It could be either an accretionary wedge developed at the deformed leading edge of the Sangihe forearc or be part of a former intra-arc basin which was part of the colliding Halmahera arc terrane. There is no clear evidence for accreted oceanic crust belonging to the recently subducted Molucca Sea. The Miangas-Pujada-Talaud ridge is part of the Sangihe forearc, and the ophiolites could represent its basement, uplifted along the outer arc ridge. The dog-leg-shaped Philippine Trench is propagating southward across the fragmented Halmahera arc terrane and its southern segment could reactivate a former strike-slip fault zone.
The collision of the Hahnahera arc (a fragment of the Philippine Sea plate) with the Sangihe arc (Fig. 1) has been described by Silver & Moore (1978), Hamilton (1979), and Moore & Silver (1983). The ophiolites and melanges exposed in Talaud, Mayu and Tifore Islands have been interpreted as fragments of the almost completely subducted Molucca Sea floor formerly separating the two colliding island arcs (Silver & Moore 1978; McCaffrey et al. 1980). The creation of a new plate boundary, the Philippine Trench, east of the collision zone therefore marks a step in the incorporation of the intra-oceanic Halmahera island arc into the Eurasian continental margin,
if the Sangihe arc is considered to be built on the thinned margin o f Eurasia (Rangin et al. 1990a). In this paper the geometry of this collision zone is documented by a study of the g e o m o r p h o l o g y of the northern Molucca Sea, based on large coverage m u l t i b e a m mapping, with gravity and single channel seismic data. This dataset was recorded during a cruise of R.V. L'Atalante in April and May 1994 (MODEC cruise), in co-operation with the I n d o n e s i a n Institute of D e v e l o p m e n t and Technological Research, Jakarta (BPPT) and the Philippines Bureau of Mines and Geosciences, Manila (BMG).
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 29-46.
29
i~O L
I~401
I % 1 0 ~"
I~ Qd~,
:)!]ouOow~ ~ImmW~
d!Is-Oq!JtS
qr
'"~/
•
•
•
•
•
="''' o~!i'IOU
•
"
.
,
.
.
•
.
~
+,wvu
ii ~,l un
oAn;~Y
.
.
•
•
•
•
•
.
: • •
", . ' . . ' . s
•
Oluld - Ue!lUJlSni
"
":,
Ills
.;..
r a : m t q
.
•
.~'o~
o
:oTvJ.N"
• ~'.
.
"q,,,,
°
Nisvb
.
+:..,,---,
•
~ ~
~ ' "
uo#low o;nlosqv
~.
~.i-...'
"..'..
.
"
.
•
•
eleld ue.lsLln 3
elml. eros P ' / / / / A 'u!dd.i|!qd V / / / / / ~
•
~
•
•
..
•
.
.
"
" .
• • " "
.
N.O
='-~2
oL;~L
)
•
"
"
•
....,+.~-.+.
.... ~+..~,+.
~ . +/:
. .: ~. . ' "~.;~."..: ".
,.'......'.'........:~.:,..:::... .": ~. . i . . ~:'~,~,8"" ,g~(~)m..
0
O;eld oU!Oed
i ......... :.......~..-...:~',~..
e
o" ",,~ ~(+)si' ~.+~/.+,,+,.... • '
•
A
o ~ I.I.
-p+~.qutu.ts qzutu -smeiu 3!a+qusoql!l U.teLU +ql S~aoqs I+SU! +q& "~9S eoonlOlN oq+ j o ~laO~a+tueaj 3!tUeUApO+D "I "~!~I
ouoz UOllonPlln S V "
"
,<. ~ ' ' • m."X\+"'',,~-~"
. ~ . -
" ~
....
)UOJl uo#s#llOO ~ maonlow
" •
-.
.... • .
]]....
o 5(.,I,
COLLISION AND FAULTING IN THE N MOLUCCA SEA The preliminary interpretations of these data and attempts to answer some basic questions are presented here. How does the collision zone between the Molucca Sea basin margins change from frontal to oblique collision, after oceanic crust of the Molucca Sea has been entirely subducted between the two colliding arcs? What is the morphotectonic signature of a newly created plate boundary, the Philippine Trench? Gravity and seismological data from the Molucca Sea were published by Silver & Moore (1978) and McCaffrey et al. (1980) but the only detailed bathymetric map available before the cruise was that of Krause (1966). Seabed mapping should allow a better understanding of the geometry of this diachronous collision and the rapid structural variations expected in the transition zone between collision and oblique subduction. Here, the focus is on the interpretation of newly acquired seabed surface data (bathymetry and reflectivity, 3.5 kHz echo sounder and single channel seismic reflection with a maximum pene-
123 ° E
~23"3o
124°E
124°3(~
125°E
msO3OE
31
tration of 2 s twt). Gravity, magnetic, seismic reflection and swath mapping surveys were conducted to map geological structures and the thickness of sediments between the Halmahera and Sangihe arcs and the Philippine and Cotabato Trenches. The position, speed and direction of the ship were given by the GPS (Global Positioning System) navigation system. The preliminary interpretation of the structures in the survey area is based on the morphological and geophysical data gathered on board in real time. A total of 3725 nautical miles were surveyed during the cruise providing a detailed bathymetric map (Fig. 2). L'Atalante is equipped with a SIMRAD EM-12 dual multibeam echo sounder which provides bathymetry and reflectivity imagery of the sea floor. Only the bathymetric data are discussed here. The MODEC cruise also involved continuous gravity and magnetic measurements and six channel seismic recording. The few seismic data presented here are single channel records acquired on board and are not yet reprocessed.
126"E
ms*3OE
127°E
127o30E
128°E r
r
i
I
128°3<~ i l
129"E +
7*N
7oh
+°"
-
6ON
+'+'_-
_-
5"30N
5°N
,++,,,., -
]-+/
t
1
[
+.-',
\:
+
-
4o30N
4*N 3°3ON
+' •
Morotai
"
i
~3ON
2=30N --
~
Tagulandang
_
_ 2°N
3°N
123°E
~
--
Halmahera~'i~
2ON 123°30E
124°E
Fig. 2. MODEC cruise tracks.
124°~E
125°E
12S°30E
126°E
1~°3OE
1 27°E
+27*3oE
128°E
t2B*~E
1290E
32
C. RANGIN ET AL.
Geodynamic framework The Molucca Sea is located at the junction of three convergent major lithospheric plates (Fig. 1). The Eurasian plate is bounded by active volcanic arcs (west Mindanao, Sangihe, and north Sulawesi) and includes the marginal basins of the Celebes Sea and South China Sea which opened in Palaeogene times. The northern margin of the Australian plate extends from New Guinea in the east to central Sulawesi in the west. The Philippine Sea plate has a western part formed largely of Palaeogene and older oceanic crust (the West Philippine Basin) which probably formed in a backarc environment. The associated arc currently comprises a major part of the Philippine archipelago, now orientated approximately north-south between Luzon and Mindanao, and extending southwards into the Halmahera arc. Kinematically, the Australian plate is moving northward with respect to Eurasia at a rate of 78 cm a-1. The Philippine Sea plate is moving westwards with respect to Eurasia about a pole of rotation located at approximately 48°N and 157°E (Seno et al. 1993), at a rate which varies from 810 cm a-1 from north to south along the Philippine Trench. The West Philippine Basin therefore subducts along the Philippine Trench more obliquely in the north than in the south. The Philippine fault, parallel to the trench, accommodates part of the lateral component of this movement, assuming a simple model of shear partitioning. As a result the East Philippine crustal fragment is displaced to the north. Th~ recent development (3 to 5 Ma) of subduction at the Philippine Trench is suggested by (1) the presence of a slab extending to less than 200 km depth (Cardwell et al. 1980; McCaffrey 1982); (2) recent arc volcanism in eastern Mindanao (Quebral et al. 1995); and (3) young arc volcanism in Leyte, Samar and Bicol (Aurelio et al. 1991). The formation of this new active margin is interpreted to be the result of a jump in the position of the subduction zone from west to east after the Philippine arc collided with the Eurasian margin. Further south in the Molucca Sea, the Halmahera arc, carried on the Philippine Sea plate, is in frontal collision with the Sangihe arc (McCaffrey et al. 1980). The Philippine Trench terminates at this latitude (Nichols et al. 1990). The collision zone is marked by a thickening of the crust below the basin, mainly deduced from gravity data and interpreted as oceanic crust imbrication in the central part of the basin (Miangas-Talaud-MayuTifore ridge). The oblique movement of the Philippine plate with respect to Eurasia has therefore been interpreted to have resulted in the diachronous closure of the Molucca Sea basin. The
presence of a slab more than 700 km below the Celebes Sea (Cardwell et al. 1980) provides an indication of the extent of the subducted oceanic basin. The Philippine archipelago is the site of the active collision between the Philippine Sea plate carrying an extinct volcanic arc and continental fragments rifted away from the Eurasian margin (Holloway 1982; Rangin et al. 1990a). The present tectonic setting of the Philippines shows an elongated Philippine Mobile Belt, fringed by two recent subduction zones (Lewis & Hayes 1983), and traversed axially by the 2000 km long left-lateral Philippine fault (Willis 1937; Allen 1962; Aurelio et al. 1991). The volcanic arc of Oligocene age, which forms the basement to the eastern part of the mobile belt, is in tectonic contact with the Eurasian margin represented by continental fragments exposed in the western part of this belt, and rifted away from the Eurasian mainland in south China (Holloway 1982; Rangin et al. 1990a). Southeast of Mindanao, the collision between the Halmahera arc and the along-strike equivalent of the western part of Mindanao occurred during the Late Pliocene and Pleistocene (Moore & Silver 1983; Hawkins et al. 1985; Mitchell et al. 1986; Pubellier et al. 1991; Quebral et al. 1995). According to many authors (Roeder 1977; Cardwell et al. 1980; McCaffrey et al. 1980; Moore & Silver 1983; Hawkins et al. 1985), the arc-arc collision of the Molucca Sea should be traceable onshore into Mindanao, and the Philippine fault has been interpreted as the inferred suture. However, geological observations (Mitchell et al. 1986; Pubellier et al. 1991) indicate that the Philippine fault is not a former suture although a possible suture candidate may be traced through the Cotabato basin.
Transect across the Northern Molucca Sea On the basis of the newly acquired bathymetric data (Fig. 2), this paper presents first a summary of the key features and preliminary interpretations of a NE-SW transect across the Molucca Sea south of Mindanao island. This northern transect extends from the Philippine Sea plate to the Celebes Sea immediately south of Mindanao. We then discuss a shorter but complementary transect carried out further south, around the Talaud Islands.
1. Philippine Trench to the Cotabato Trench Various morphostructural provinces are distinguished from east to west along this transect (Fig. 3).
33
COLLISION AND FAULTING IN THE N MOLUCCA SEA
7N
Fig. 3. Shaded image of the sea bottom in the northern most part of the Molucca Sea, south of Mindanao. The main morphological units of the area are shown. DB: Sarangani-Davao Depression; SB: Sangihe forearc Basin; PR: Pujada ridge; CD: Central Depression.
Outer slope of the Philippine Trench Close to 6°N, there is a dog-leg bend of the Philippine Trench. The trench trends north-south in the north and suddenly changes to a mean direction of N N W - S S E further south. North of 6°N, the outer slope shows regularly spaced west-facing fault scarps interpreted as parallel normal faults, down-thrown towards the trench, with vertical offsets which are up to 900 m (Fig. 4). The outer slope is one of the steepest outer slopes observed at a trench anywhere in the world with gradients of up to 11 o to the west. South of 6°N, the outer slope is less steep with an average gradient of 6 ° . Linear sub-parallel scarps trend at 160 ° to 170 ° and are slightly oblique to the north-south trend of the trench. The normal faults dissecting the outer slope vary in orientation from 160 ° to 140 ° and are parallel to the trench axis. In some places the faults are slightly oblique to the trench axis and trend 125 ° (Fig. 4). The very deep (9600+_100 m) V-shaped trench is characteristic of a non-
accretionary convergent margin. Its north-south trend north of 6°N is slightly en echelon. South of 6°N the trench consists of a series of en echelon basins offset along left-lateral faults resembling Riedel fractures, and its depth shallows from north to south by almost 1000 m.
The inner slope of the Philippine Trench The inner slope of the Philippine Trench, which descends eastwards towards the trench, can be divided into three morphological areas in the northern part of the area surveyed: the lower, middle and upper slope regions (Fig. 3). South of 6°N the lower and upper slope areas lose their distinct character and merge with the middle slope. The middle slope is separated from the trench by the Snellius Ridge (Figs 3 and 4).
Lower slope. The lower slope of the upper plate (Figs 3 and 4) is geomorphologically very different
34
C.
RANGIN
ET AL.
~ PHILIPPINE I~i TRENCH
,
\
Outer ~-jSlope
,
5N--
127 E Fig. 4, Bathymetric map of the Philippine Trench Inner wall. Main structures are also indicated. South of 6°N, the complex morphostructural zone between the Snellius Ridge and the Miangas ridge is the middle slope unit.
in the areas north and south of 6°N. North of 6°N it consists of an 8000 m deep terrace. In this area, the scarps at the base of the inner wall that define the front of the margin could be interpreted either as normal or thrust faults. It is very difficult to trace the true plate boundary since no decollement is observed on seismic profiles. South of 6°N a triangular-shaped terrace lies between the Philippine Trench and the middle slope. The major feature observed in this area is a 50 km linear scarp
facing the trench with a 140 ° trend. A vertical offset of more than 2 km is observable along this scarp that could be interpreted as a major normal fault. West of the scarp the lower slope dips gently towards the NE and pinches out along the trench close to 5°30'N. This surface is part of the flank of the Snellius Ridge which broadens southwards.
Middle slope.
The middle slope has an average gradient of 10 ° to the east across the whole of the
COLLISION AND FAULTINGIN THE N MOLUCCA SEA area surveyed. In the northern part of the area, north of 6°N, it is a complex morphostructural zone of interfering structures. The major characteristic of this area is a series of NW-SE trending features, interpreted as a strike-slip faults. One of these fractures appears to be the offshore extension of the Philippine fault (Quebral et al. 1995) exposed in Mindanao (Fig. 5). Discrete 045 ° trending structures are also observable and may represent a fold axial trend. South of 6°N, the orientation of the middle slope break changes from 140 ° to northsouth, and 010 ° close to 5°40'N, and then turns again to 140 ° south of 5°N. This morphology is interpreted as a combination of NW-SE trending left-lateral strike-slip faults and 010 ° trending folds and thrusts. The middle slope break is probably seismically active in the southernmost part of the area since thrusts were observed on seismic profiles. To the west, the middle slope unit is in tectonic contact with the Miangas ridge. At the base of the steep east-facing scarp of this ridge there is a narrow area of broken topography which descends over about 200-300 m to the middle slope (Figs 3 and 4). This large area (Fig. 3) has an average depth of about 3800 m between 5°30'N and 5°50'N and rises to the south and north. A number of lineaments can be seen on the bathymetric and reflectivity maps which mark irregular ridges with principal orientations of 015 °, 140 °, 170 ° and 180 °. The intersection of these ridges outline a number of small rhomb-shaped, flat-bottomed basins.
Upper slope. The upper slope is steeper than the middle slope north of 6°N and absent further south. In its southern part east-verging thrusts were imaged on seismic profiles. Further to the north, lobate bodies with edges which are convex to the east, are interpreted as the manifestation of flat east-verging thrusts crossed by NW-SE trending strike-slip faults (Figs 3 & 4).
Miangas-Pujada ridges This north-south trending bathymetric high is a complex broad feature formed by two ridges which rise from depths of 3000 m on each side (Fig. 5). To the north the western ridge continues into the Pujada peninsula and to the south the eastern ridge widens considerably to form a bank connecting southward with the Talaud Islands (Fig. 2). For most of its length this high is divided by a depression (the Central Depression of Fig. 5). In the north, the Central Depression is little more than an irregular deeper part of a single bathymetrically high area with steep east- and west-facing slopes (Figs 3 and 5). Further south the Central Depression becomes more pronounced and deeper.
35
The Miangas ridge. This eastern ridge remains high along all its length and is emergent on the island of Miangas at 5°36'N. In most of the area surveyed its depth is typically less than 1200 m. The ridge trends broadly towards 160 ° but in detail the eastern face can be seen to be composed of en echelon north- and 140 ° trending segments. It is asymmetrical in cross-section, and the highest points are on its eastern side, with a very steep east slope (average gradient 20% and a much more gently sloping west face (average gradient 9°). This ridge has a consistently high reflectivity, and seismic lines indicate an absence of sediment on most of the upper parts of the ridge and on the very steep east-facing slope. The Miangas ridge is interpreted to be thrust eastwards over the middle slope of the Philippine Trench. This interpretation is supported by preliminary seismicity data obtained in an OBS (Ocean Bottom Seismometer) network installed around Miangas Island during the MODEC Cruise. This temporary network revealed the possible presence of a west-dipping thrust plane below the ridge. This thrust is not observed on seismic profiles due to the steepness of the slope and the absence of sediments. The inferred thrust trace at the western base of the ridge is offset by 140 ° trending lineaments at 5°30'N and 5°40'N, south and north of Miangas Island respectively. These features are interpreted as left-lateral strikeslip faults offsetting the ridge (Fig. 5). The Pujada ridge. The summit of the western ridge deepens southwards, narrows and disappears at 5°05'N (Fig. 3 & 5). South of 5°30'N, the ridge turns east and has an overall orientation parallel to the base of the Sangihe arc (c. 155°). Its western slope becomes somewhat less steep and more irregular although there remains a steep section about half-way up. Its eastern face has a broad, gently east-dipping upper terrace between 2000 m and 2250 m which then descends more steeply, but irregularly, to the well-marked Central Depression with a flat surface at 3700 m. The ridge has a steep west-facing slope. The reflectivity map and the seismic lines show little sediment cover and only a few discontinuous reflectors (Fig. 6); where there are reflections they mainly dip west but no structure can be identified with confidence. It is suggested that these areas are underlain by basement rocks such as ophiolites or arc volcanics. Two principal lineament orientations, 150 ° and 050 ° are parallel to the steeper slopes on the upper part of the ridge. There are no clear offsets on these lineaments and no clear indications that they are faults on seismic lines although the 150 ° direction is close to the orientation of the western margin of the Sangihe trough and inferred axial directions of folds within
P~
i'l
,G.
f ! Z~
,G. ,-) Q.
°/ I:1:1
o > ill i'-'i
,
J
,
.2.-~
J
.
/
,•!.,
.~-:;~
I:
li
<: ,"
J
.. ,
-:'.- u., (-
Z .--=j"
,x- .... ;:'~ <:%d~ -
.... ,
; .:' ; : f .
i'
~.
f
p - , , ~ e ~ . . . . /"
oooP ~,!,
\
•
_/J
,--~
>
>
~
-
7
-*
Ii
~k<>~'- ~ ; .17v~%,, ~ - : (
,,j))
'~ ~~c'FT, Q': 7:7 ::7~7"{::;~ " <
,>':.737~;75;=., . . ,
=!
i=
oE
i=l., o
~8
18
I-
.,,-i
~
Sarangani ridge
~
ll~
~
~+,a
_ ~
,
i
126"E
~a~*aOE
...... I..
++i! Z
~,+
+
+
}::.<:I:~4P{G#--_:~--~:~?!~.!+;::;:?:.#-~ .
Pujada ridge . ~ i ~ !
" Line 19
:-.:':.:.+:.!=+ <:::!'-"+:~,i;+++.+.''..:, :-5,:~
~g{~:>+V.~:>-';?Lv.+_.:-.::i~:~-:
~g~{i~{i+{i~!!-<~
~-~+2-!*+'.+.:--+-?:
{+i++;~
,'+,+,ar+
Fig. 6. Single channel seismic profiles and interpreted line drawings across the Sangihe forearc basin. Note increasing deformation in the basin from south to north.
6 -:: :.?.:.~:::57:i..:,7.?:7:-.!:.; .:&~>:-~:::.i .... : ,:. :: +. +i!<e.:?~?~ ' ~!{~:~{~>~e~:¢~.{g>;~}~-~`..{r.~...?f~:~-+~.7c~<:-~:~.{:#5~::~{~i~g ~>+gi~+??:.>~£÷:#:.:_;g_.~_~?4---!g-:3--..:~ _~.-; ]-? :::]].::::7'7"r.::':i?i#:i ?Si~::'" :' :i?i ': :;: : " ""':(:': e:+!!).i`~{>!e{~!)5#~{>{~#{~{;i~i{~{~{{!;~.!~z~.~:~i:;f¢!~:¢:~`:~}~:gi{{~# ~~~a~%##i~v~+~!i>{i::~':::::':~-+~i> ,:'{;v<:7'v:~:?5~2:;~:; L in e 2 1
g:~:~:~.~`~:`...:~:~4?.:::~.`~?~:`!i:~5~:~¢?>:::..?.#~i!~`yy?:~.~!~:s#~ :?:-;~:'.+.k~~
.......
]4i'7"ie:-:':)'i:::!!:'ili!{i:+'{!i:-.' +~ ~!":-'~;
~+~-i~::~
6ON-
Sarangani ridge ~
+
M l'ta r ;a-=
+~.~
Ii+i+~++ii+++++++ :+++i+++++! ++iiii+i!iiiWi~
w
,..
125"E
........t................. 1
t.aa -.dl
38
C. RANGIN E T A L .
the Sangihe trough (see below). These lineaments may represent a conjugate fault/joint set within the basement. The eastern side of the Pujada Ridge is locally steeper than the western side, but includes a number of irregular terraces which become wider southwards. Medium to low reflectivity and seismic profiles indicate that there is a sediment cover although the thickness is not great (< 1 s twt). Over both the Pujada and Miangas ridges, the free-air anomaly map is well correlated with the bathymetry (Fig. 7). The slope descending from
126 E
the Sangihe arc is characterized by a regular gradient of the free-air anomaly with values decreasing towards the Sangihe forearc basin. The minimum (-80 mGal) is situated at the northern end of this basin at 5°50'N and 125°50'E, above an area of thrusting, where the sedimentary thickness is probably larger. The Pujada and Miangas ridges are underlain by north-south aligned positive anomalies separated by a low corresponding to the Central Depression. The Miangas ridge, which is topographically higher than the Pujada ridge, is
127 E
I
7N
+
6N
5N
%
4N
+
128 E
i
,~
3N
o? Fig. 7. Free-air gravity anomaly map of the surveyed area.
i
,
-~\9~IIT!
I | ..,.
+
COLLISION AND FAULTING IN THE N MOLUCCA SEA
associated with a smaller anomaly. Northwards, the two anomalies merge into a single one, as the ridges do. On the eastern edge of the Miangas ridge, a steep gradient (more than 15 mGal/km) can be correlated with its steep eastern flank.
The Sangihe forearc The Sangihe forearc basin (Figs 3 and 5) is located east of the Sangihe volcanic arc and its northern extension in Mindanao, the Sarangani peninsula, and west of prominent ridges (Pujada and Miangas) formed by ultramafic and mafic material (Fig. 3). This basin pinches out to the north in the Davao Gulf (Fig. 5), between the Sarangani and Pujada peninsulas, where we term it the Sarangani-Davao depression (Fig. 3). Between 5°37'N and 5°45'N there is a marked change in orientation of the eastern slope of the Sangihe volcanic arc, from 155 ° in the south to the 015 ° in the north. At this latitude the basin becomes wider and structurally simpler, and has a different orientation.
The Sarangani-Davao depression. With a central plain at depths of about 3200 m, the basin has an arcuate shape parallel to the Sarangani peninsula. Field work in Mindanao (Pubellier et al. 1991) has shown this arcuate peninsula is a large recumbent fold verging towards the east. The SaranganiDavao depression corresponds to the northern extension of the Sangihe Basin and sediments in the basin are folded and thrust in the Davao Gulf. This is related to the active compressive tectonics affecting the island of Mindanao (Pubellier et al. 1991; Quebral et aL 1995). At the western edge of the basin, the slope of the eastern edge of the Sarangani peninsula is very steep (Fig. 5) and is interpreted as the manifestation of a buried thrust. In the central part of the basin, three broadly northsouth trending, arcuate, and west-verging thrust anticlines are well marked in the bathymetry (Figs 3 & 5). Reverse faults or thrusts are observed along their eastern flanks. The free-air anomalies deduced from six gravity profiles across the basin (Fig. 7) show an asymmetric form with a clear minimum (-80 reGal) close to Sarangani shoreline on the west side of Davao Gulf. This anomaly is correlated with the thickness of sediment, suggesting a flexure of the basin due to loading, by overthrusting from the west, of the Sarangani peninsula. The Sangihe basin. Seismic profiles across this basin show at least 2 s twt of well-stratified sediments (Fig. 6). Buried folds with a clear westvergence are observed on seismic profiles, with an axial trend of 155 ° . This direction is parallel to bathymetric features at the basin margins. South of 5°25'N the seabed surface in the basin changes from a flat smooth surface to an irregular surface
39
cut by numerous channels. Silver & Moore (1978) extended their East Sangihe Thrust through this part of the Sangihe trough. Data presented here do not support the presence of any west-verging thrust along the western edge of the forearc basin.
Sangihe backarc and Cotabato Trench East of the Sangihe arc, the survey was extended into the Celebes Sea and along the southeast termination of the Cotabato Trench. At about 5°N the Cotabato Trench bends from north-south to an approximately ENE-WSW orientation parallel to the southern coast of Mindanao. In the Celebes Sea basin the ship track crossed ODP sites 767 and 770 (Fig. 8). South of 4°50'N the feature corresponding to the continuation of Cotabato Trench is a broad flatbottomed area, sloping gently northwards. A large meandering canyon running south to north down to 5°N probably originates from the large island of the Sangihe group indicating the trench dies out morphologically further south. To the east volcanoes are aligned north-south on the western flank of the Sangihe arc. At 4°40'N a conical seamount, probably belonging to the nearby Sangihe volcanic arc, rises 3500 m above the sea floor. Between 4°50'N and 5°20'N the Cotabato Trench bends from an orientation of 000 ° in the south to 120 ° in the north. The trench floor deepens from 5100-5900 m as it approaches the Sangihe arc. At the northern end of the north-south trending section (north of 4°50'N), which is marked by a 100 reGal negative gravity anomaly, a narrow NNE-trending accretionary wedge is observed at the foot of the steep western flank of the Sangihe arC.
At the trench bend there is an intermediate zone which trends 145 ° and is separated from the Sangihe arc by a non-linear ridge and a plateau which could be the physiographic equivalent of an outer arc high. A NNW-SSE trough 5100 m deep, is parallel to these on their landward side. Here, the trench-fill sediment sequence thickens significantly up to 2 s twt. A re-entrant in the wedge associated with bending of an anticline axis is interpreted as the effect of a subducted asperity (ridge or seamount) on the sea bottom. This is supported by the presence of a 1600 m high curved scar observed in the bathymetry of the wedge at 5°20'N, 124°50'E, which could be attributed to the collapse of the margin following the subduction of a seamount. Along the Cotabato active margin, which trends at 120 °, the lower slope consists of a series of anticlines, 20-30 km apm't, with curved axial traces which are concave upslope, forming a wide foldand-thrust belt typical of an accretionary wedge.
40
C. RANGIN ET AL.
The trend of the anticlines is roughly parallel to the trench axis with an en echelon pattern in several places. An outer arc high, 400-500 m in elevation, separates the accretionary complex from what is suspected to be a forearc basin with an upper surface dipping slightly towards the southeast. A large gravity low is located over the accretionary wedge probably reflecting a thick sedimentary prism. The 4600-5500 m deep abyssal plain of the Celebes Sea basin shows several ridges and seamounts sitting on the oceanic crust. On seismic profiles the basin floor has a very rough undulating surface (amplitude of a hundred metres) which possibly reflects the original grain of the oceanic crust overlain by no more than 0.5 s twt thick sediment cover. The first sediments deposited onto the basement were dated at 43 Ma during ODP Leg 124 (Rangin et al. 1990b). In the vicinity of the trench no clear normal faults were observed on seismic lines. The bathymetric map (Fig. 8) shows lineaments, ridges and troughs trending 110 ° in the northwestern part of the area surveyed. An elongated positive feature, 1.5 km above the sea floor and without a magnetic signature, interpreted as a group of seamounts, trends WSW, parallel to the magnetic lineations proposed by Weissel (1980) in the basin. Another major feature, trending northsouth, is more than 100 km long and l0 km wide and rises 1 km above the sea floor.
Ridge and its lower slopes form two terraces (Fig. 9) which occupy a triangular area which is narrow in the north and widens southward. The eastern margin of the West Talaud Bank is bounded by west-verging thrusts following the base of the cliffs of the Talaud ridge where a second OBS temporary network has been deployed (Fig. 9). The West Talaud Bank is thus thrusted towards the west above the undeformed Sangihe trough.
Talaud ridge The Talaud ridge forms a high area between the 2000 m bathymetric contours. Little of the ridge appears on the map (Fig. 9) because of the very shallow depths of the ridge summit. In the northern part of the area surveyed the Taland and Nanusa Islands are part of the same wide ridge, but further to the south the Talaud ridge narrows and swings southeastward. These trends are reflected by the shape of the islands' coastlines. On seismic profiles the ridge lacks any continuous reflectors and a very thin sediment cover was observed above basement rocks, possibly ophiolites or arc volcanic rocks, as indicated by the geology of the Talaud Islands on land (Sukamto et al. 1980; Moore et al. 1981). Both east and west flanks are fault controlled. These are most likely to be thrust faults because the fault contacts are sub-parallel to the curved isobaths.
East Talaud Bank
2. Talaud Islands to Morotai Basin A second transect, orientated NW-SE, south of the Talaud Islands, was surveyed in the deep waters between the Morotai Basin and the Sangihe trough. It was chosen to be perpendicular to the regional morphology. South of 5°N, the Miangas ridge merges into a very shallow area covered by reefs around the Nanusa Islands (Fig. 9). To the east, the Snellius Ridge which flanks the Philippine Trench between 4°50'N and 5°30'N, becomes very shallow south of 4°50'N. Consequently, most of the area at the latitude of the Talaud Islands was not surveyed because it is too shallow. Various morphological provinces were identified from west to east.
Sangihe trough Only a small part of the undeformed Sangihe forearc basin (Sangihe trough on Fig. 9) was surveyed in this area (Fig. 9). The flat part of the trough is much narrower here than in the north. The area deeper than 3500 m is less than 25 km wide. The trough is slightly deeper in its central parts and reaches depths of 4000 m. The West Talaud Bank rises from the Sangihe trough towards the Talaud
The eastern margin of the Talaud ridge is a wide area, named the East Talaud Bank, which connects northwards to the middle slope of the Philippine Trench inner slope(Figs 3, 9, 10). South of 3°55'N, the East Talaud Bank is very wide (about 35 km) and has an irregular surface which is convexupwards. North of 3°55'N, it becomes considerably narrower (about 18 km wide), and forms a northsouth trending feature between depths of 1500 and 3000 m. South of 4°N (Fig. 9), the bank is separated from the Talaud ridge by NW-dipping thrusts. Seismic profiles show that the East Talaud Bank resembles the West Talaud Bank which is interpreted as the deformed margin of the Sangihe Basin to the west. Along the east flank of the bank, a major low angle thrust can be traced southwards from 4°40'N to 4°00'N where it has a north-south trace; it then turns, first towards 140 ° , and then further south towards 030 °. There are at least 3 s twt of sediments in the slice above the thrust plane, which carries the East Talaud Bank eastwards onto the Snellius Ridge, an observation reported previously by Silver & Moore (1978). The Talaud ridge, and the East and West Talaud Banks on each side, are cut by 140 ° scarps that
41
COLLISION AND FAULTING IN THE N MOLUCCA SEA
*d
? 0
E
O
if}
O. @ @ 0
:1:0
--,
o
i,f l ¢,i o
c
© l (1) k_
/
~
o/ ¢)
/
o
s
II.~l
,
~~
-i ~
¢i
°~
"a
z
•
©
¢. RANGIN ET AL.
42 E126 40
~J
E127 20
E128
,,,
N a n u s a Island ( Q k , ~ . .
~-~
~S'~'-""¢,"tJ,ie '
~-
_
~,
......
i)i:~
~
Philippine i Trench N420
~
!_/2.
?,
/ "¢ .,~( 4,
• .
:
~,
,
• N3 20
...... I?:!L..:i. 1
_
!
I
San
Fig. 9. Bathymetric map of the area around the Talaud Islands showing the main interpreted structural elements.
could be strike-slip fault zones. This is supported by the evidence of earthquakes in this area with strike-slip focal mechanisms (McCaffrey 1991). The progressive offset of morphological units towards the northwest suggests left-lateral motion on these faults.
Snellius Ridge East of the main thrust separating it from the East Talaud Bank, the Snellius Ridge is a large submarine plateau sloping to the west with a gradient of 3-6 °. South of 3°40'N (Fig. 9) the plateau is dissected by SW-facing fault scarps trending 140 °. These features also dissect the northern margin of the plateau (Fig. 9) and are interpreted as the trace of major strike-slip fault zones (see above). Well
developed reflectors at the top of the Snellius Ridge could be platform carbonates capping the plateau. The positive gravity anomaly over the Snellius Ridge suggests a dense crust compatible with an ophiolitic or arc origin for the basement. In the southern and deepest end of the Snellius Ridge is the Morotai Basin. This basin has an elongate shape with its major axis orientated 070 ° , and has a maximum depth of about 3500 m. Along its northwestern margin, it is thrust westwards onto the Snellius Ridge, and folds trending 075 ° defornl the basin sediments. These structures terminate abruptly to the west at the curved thrust zone where the Morotai Basin margin is thrusted on top of the Snellius Ridge. Further south the basin is filled by more than 1.5 s twt of almost undeformed sediments. There is no evidence for significant
COLLISION AND FAULTINGIN THE N MOLUCCA SEA deformation at the southern side of the basin close to Morotai Island within the area surveyed, although a little further south the sea bottom rises rapidly from depths of more than 3000 m to emerge at the steep fault-controlled coasts of the island. The southwest end of the Snellius Ridge is situated between two convergent thrust fault zones, and is beneath the East Tataud Bank and the Morotai Basin (to the NW and SE respectively). The superficial convergent thrusts in this area are thus compatible with interpretations at depth that show two convergent subduction zones and an inverted U-shape of the subducted Molucca Sea oceanic slab. More precise correlations need to be done between the superficial data collected during the cruise and structures at depth.
Discussion During the MODEC cruise in the northern Molucca Sea it was possible to trace the Sangihe forearc
E 1 2 3 20.
E124
E125
43
continuously over 500 km from north to south. Between the eastern edge of the forearc and the Philippine Trench, three major crustal units or terranes were also mapped almost continuously from north to south (Fig. 10), and are interpreted to be separated by major lithosphere-scale boundaries.
Miangas-Pujada-Talaud unit The Miangas-Pujada-Talaud unit is probably formed by ophiolite slivers and melanges; it can be traced from the Pujada peninsula to the Talaud Islands, with a major offset between Miangas and Talaud north of the Nanusa Islands. The PujadaMiangas ridge is onlapped westward by the sediments of the Sangihe forearc basin; further south at the southern end of the Talaud ridge there are west-vergent thrusts moving the ridge onto the edge of the Sangihe forearc basin. This ridge could be interpreted as the backstop of the Sangihe subduction zone, implying it is underlain by
E127 r
.
.
.
.
.
E128 r
E1281 4 0
PHILIPPINE I BASIN
i
CELEBES SEA
¶.
0
E123 20
El 4
E125
E126
E;27
EI~8
~I~
40
Fig. 10. Overall interpretation of the northern Molucca Sea. The shaded area is the Miangas-Pujada-Talaud unit probably formed by ophiolite slivers and melanges. Immediately east of this is the middle slope unit.
44
C. RANGIN ET AL.
Sangihe forearc basement rather than oceanic crust of the subducted Molucca Sea. The MiangasPujada-Talaud unit is separated from the Philippine Trench middle slope terrace by a major discontinuity that could be a lithospheric-scale thrust. Seismic focal mechanisms associated with this thrust are clearly compressive with east-west P axes. This thrust can be traced south at least to 3°N at the foot of the Talaud ridge. Middle slope unit The middle slope unit, lying immediately to the east of the Pujada-Miangas ridge, is formed by low density material as indicated by the free-air gravity data. The middle slope unit can be traced southward into the East Talaud Bank which is thrusted eastwards onto the Snellius Ridge, almost certainly a fragment of the Halmahera arc terrane. North of 5°30'N, this thrust fault approaches the Philippine Trench. A broad gravity low trending north-south, about 100 km long and with minimum values of -260 mGal is centred on the middle slope unit (Fig. 7) and is clearly shifted to the west with respect to the Philippine Trench (-220 mGal). This anomaly cannot therefore be attributed to the classical gravity low associated with deep trenches. The juxtaposition of the low density material of the middle slope unit in the north and the high density material of the Snellius Ridge in the south, could be explained by the presence of a major crustal discontinuity trending 140 ° . The orientation of isogals on the free-air gravity map above the middle slope unit changes from north-south to NW-SE approaching the northern edge of the Nanusa Islands and the Snellius Ridge (Fig. 7). The discontinuity is also marked by a pronounced NW-SE steep gravity gradient. It is also subparallel to the orientation of the Philippine Trench south of 6°N (Fig. 10). Such a gradient and discontinuity is also clear in recent satellite FAA data. This supports the presence of a major crustal-scale fault zone separating two contrasting terranes. The middle slope unit could be composed of accreted sediments, and be part of the Sangihe accretionary wedge, with the Snellius Ridge representing part of the Halmahera arc. In this interpretation the boundary between the colliding arcs would be the thrust at the base of the middle slope and East Talaud Bank. An alternative interpretation is to consider this unit as a part of the Halmahera arc like the Snellius Ridge to the south. The middle slope unit would then represent an intra-arc basin. Such contrasting features are known within the Halmahera arc system to the south (Hall et al. 1988a, b; Nichols & Hall 1991). In this case the major west-dipping thrust located east of the Miangas-Talaud ridge would be the
main tectonic boundary between the Sangihe and Halmahera arcs. The major crustal discontinuity trending I40 ° which offsets the Pujada-Miangas ridge relative to the Talaud ridge is interpreted as a left-lateral strike-slip fault zone. Philippine Trench The dog-leg shaped Philippine Trench has two distinct segments. The north-south trending segment north of 6°N is the trace of the recently established west-dipping subduction zone as indicated by the short subducted slab. The trend of this trench segment is completely independent of the dominant oblique NW-trending structures observed in the inner slope. In contrast, south of 6°N, the N W - S E trending segment of the Philippine Trench is parallel to the interpreted left-lateral strike-slip fault zones observed in the whole forearc area. The southern segment of the Philippine Trench is tentatively interpreted as the trace of a major left-lateral strike-slip fault zone dissecting the Halmahera terrane. The small faults in Riedel position observed along this part of the trench support this hypothesis. However, thrust fault focal mechanisms of seismic events reported by Ranken et al. (1984) along this trench segment reveal this strike-slip fault zone is now turning into a thrust (Nichols et al. 1990). This suggests a migration of the Philippine Trench towards the south along pre-existing strike-slip fault zones. It is also possible that, although focal mechanisms support a thrust interpretation of this fault zone, the main long-term motion is strike-slip, possibly aseismic.
Conclusion The area surveyed in the northern Molucca Sea reveals the presence of two distinct geodynamic settings east of the Miangas-Pujada-Talaud ridge. A major NW-SE trending tectonic boundary separates these two areas. South of 6°N, the Snellius Ridge, a volcanic plateau capped by carbonates, is subducting below the Sangihe forearc terrane. The buoyancy of this plateau, a fragment of the Halmahera arc terrane, has induced deformation in the upper and lower plates (Talaud ridge and Morotai Basin). Its buoyancy also induces incipient subduction along a new plate boundary, the Philippine Trench. This nettectonic feature is developed along a former strikeslip fault zone cross-cutting the Halmahera terrane. North of the major NW-SE crustal discontinuity crossing the area surveyed, a wide accretionary prism is present within the inner slope of the present Philippine Trench. The thickness of this wedge cannot be explained by subduction along
COLLISION AND FAULTING IN THE N MOLUCCA SEA this trench because it is very recent (Cardwell et al. 1980; Quebral et al. 1995). This wedge could be interpreted either as an accretionary wedge developed at the deformed leading edge of the Sangihe forearc or be part of a former intra-arc basin which was part of the colliding Halmahera arc terrane. There is no evidence from gravity data of dense material comparable to the Snellius Ridge below this wedge. If the ridge extended originally north of 6°N, it could have been already subducted to some greater depth, or be considerably displaced left-laterally, as suggested by the presence of a similar buoyant terrane in central Mindanao (Pubellier et al. 1991). This survey also reveals no clear evidence for
45
accreted oceanic crust belonging to the recently subducted Molucca Sea. The M i a n g a s - P u j a d a Talaud Ridge is clearly in a back-stop position for the Sangihe arc, and the ophiolites of this ridge could therefore have very different origin from either the Molucca Sea or the Halmahera arc. This work was possible due to the co-operation programs between France, the Philippines and Indonesia. We want to thank particularly E. Domingo in the Bureau of Mines and Geosciences in Manila and Dr Zen of BPPT in Jakarta. The active co-operation of the L'Atalante crew and Genavir team was particularly appreciated. We thank particularly Captain G. Tredunit for his enthusiasm for the project and his active participation in contributing to the success of this cruise.
References
ALLEN, C. R. 1962. Circum-Pacific faulting in the Philippine-Taiwan region. Journal of Geophysical Research, 67, 4795-4812. AURELIO,M. A., BARRIER,E., RANGIN,C. & MULLER,C. 1991. The Philippine Fault in the Late Cenozoic evolution of the Bondoc-Masbate-Leyte area, central Philippines. Journal of SE Asian Earth Sciences, 6, 221-238. CARDWELL,R. K., ISACKS,B. L. & KARIG,D. E. 1980. The spatial distribution of earthquakes, focal mechanism solutions and subducted lithosphere in the Philippine and northeastern Indonesian islands. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 1-35.
HALL, R., AUDLEY-CHARLES, M. G., BANNER, E T., HIDAYAT,S. 8,~ TOBING, S. L. 1988a. The basement rocks of the Halmahera region, east Indonesia: a Late Cretaceous-Early Tertiary forearc. Journal of the Geological Society, London, 145, 65-84. & --. 1988b. Late Paleogene-Quaternary geology of Halmahera, eastern Indonesia: initiation of a volcanic island arc. Journal of the Geological Society, London, 145, 577-590. HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078. HAWKINS,J. W, MOORE,G. E, VILLAMOR,R., EVANS,C. & WRIGHT, E. 1985. Geology of the composite terrane of east and central Mindanao. In: HOWELL, D. G. (ed.) Tectonostratigraphic Terranes of the CircumPacific Region. Circum-Pacific Council for Energy and Mineral Resources, Earth Sciences Series, 1, 437-463. HOLLOWAY, N. H. 1982. The stratigraphic and tectonic evolution of Reed Bank, North Palawan and Mindoro to the Asian mainland and its significance in the evolution of the South China Sea. AAPG Bulletin, 66, 1357-1383. KRAUSE, D. C. 1966. Tectonics, marine geology and bathymetry of the Celebes-Sulu Sea region. Geological Society of America Bulletin, 77, 813-831.
LEWIS, S. D. & HAYES, D. E. 1983. A geophysical study of the Manila Trench, Luzon, Philippines: forearc basin structural and stratigraphic evolution. Journal of Geophysical Research, 89, 9196-9214. MCCAFFREY, R. 1982. Lithospheric deformation within the Molucca Sea arc-arc collision: evidence from shallow and intermediate earthquake activity. Journal of Geophysical Research, 87, 3663-3678. - 1991. Earthquake and ophiolite emplacement in the Molucca Sea collision zone, Indonesia. Tectonics, 10, 433-453. --, SILVER, E. A. & RAITT, R. W. 1980. Crustal structure of the Molucca Sea collision zone, Indonesia. In: HAYES,D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 161-177. MITCHELL, A. H. G., HERNANDEZ, E t~ DE LA CRUZ, A. E 1986. Cenozoic evolution of the Philippine archipelago. Journal of SE Asian Earth Sciences, 1, 3-22. MOORE, G. E & SILVER,E. A. 1983. Collision processes in the northern Molucca Sea In: HAYES,D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 27, 360-372. - - - , KADARISMAN,D., EVANS,C. A. & HAWKINS,J. W. 1981. Geology of the Talaud Islands, Molucca Sea collision zone, northeast Indonesia. Journal of Structural Geology, 3, 467-475. NICHOLS, G. J. • HALL, R. 1991. Basin formation and Neogene sedimentation in a backarc setting, Halmahera, eastern Indonesia. Marine and Petroleum Geology, 8, 50-61. , --, MILSOM, J., MASSON,D., PARSON, L. ET AL. 1990. The southern termination of the Philippine Trench. Tectonophysics, 183, 289-303. PUBELLIER,M., QUEBRAL,R., RANGIN,C., DEFFONTAINES, B., MULLER, C., BUTTERLIN,J. MANZANO,J. 1991. The Mindanao, collision zone, a soft collision event within a continuous strike-slip setting, Journal of Southeast Asian Earth Sciences, 6, 239-248. QUEBRAL,R., PUBELLIER,M. & RANGIN,C. 1995. Eastern Mindanao, Philippines: transition zone from a
46
C. RANGIN ET AL.
collision to strike-slip environment. Tectonics, (in press). RANGIN, C., JOLWET, L. & PUBELHER, M. 1990. A simple model for the tectonic evolution of southeast Asia and Indonesia region for the past 43 m.y. Bulletin de la Soci(t( g~ologique de France, 8 VI, 889-905. , SILVER, E. A., VON BREYMANN,M. T. ET AL. 1990. Proceedings of the Ocean Drilling Program Initial Reports, 124. RANKEN, B., CARDWELL, R. K. & KARIG, D. E. 1984. Kinematics of the Philippine Sea Plate. Tectonics, 3, 555-575. ROEDER, D. 1977. Philippine arc system - collision or flipped subduction zone? Geology, 5, 203-206. SENO, T., STEIN, S. A. & GRIPP, A. E. 1993. A model for the motion of the Philippine Sea plate consistent
with NUVEL-1 and geological data. Journal of Geophysical Research, 98, 17 941-17 948. SILVER, E. A. & MOORE, J. C. 1978. The Molucca Sea collision zone, Indonesia. Journal of Geophysical Research, 83, 1681-1691. SUKAMTO,R., SUWAR~O,N., YUSUP, J. & MONOARFA,M. 1980. Geologic map of the Talaud Islands, 1:250 000. Geological Research and Development Centre, Bandung, Indonesia. WEISSEL,J. K. 1980. Evidence for Eocene oceanic crust in the Celebes basin. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 37--47. WILLIS, B. 1937. Geologic observations in the Philippine islands. Natural Resources Council of the Philippines Bulletin, 13.
Continental collision in the Banda arc A. N. R I C H A R D S O N 1 & D. J. B L U N D E L L
Department of Geology, Royal Holloway, University of London, Egham, Surrey TW20 OEX, UK 1 Present address." Shell Internationale Petroleum Maatschappij B. V., Postbus 162, 2501 A N Den Haag, The Netherlands Abstract: Two deep seismic reflection profiles across the zone of convergence between Australia and the Banda arc east of Timor reveal the structures formed during lithosphere deformation in response to continental collision. Australian continental crust is bent down to the north to form the lower lithospheric plate. The immediately overlying upper plate is made up of a former outlier of the Australian continental shelf, now squeezed between the Australian continent to the south and the Banda arc to the north. The ages of metamorphism and of a regional unconformity in Timor indicate that the outlier began to accrete to the upper plate in the late Miocene. Near the Timor trough, at the junction between the two plates, seismic reflection data show the upper plate to be cut by north-dipping structures interpreted as thrusts in an accretionary prism formed since the arrival of the Australian continental shelf at the collision zone at about 2.5 Ma. The northern part of the collision zone is dominated by structures dipping southwards, antithetic to subduction, which penetrate the lithosphere to depths of at least 50 kin. Earthquake data indicate that these are active thrust faults which divide the upper plate into imbricate slices. The seismic reflection sections, supported by gravity data and seismicity patterns, have been used to develop a model of plate convergence and continental collision which now consists of a 200 km wide zone in which both oceanic and continental material is being shortened, thickened and uplifted.
The Banda arc lies at the intersection of the SE Asian and Indo-Australian plates (Fig. 1). Over the past few million years, Australian continental material has arrived at and collided with the oceanic plate of the Banda Sea in eastern Indonesia. The collision has progressed to such an extent that continental lithosphere has subducted under oceanic lithosphere to a depth of over 100 km. The collision of continental material with the subduction zone has given rise to a 'collision complex'. This forms a ridge, in places rising above sea-level (e.g. Timor), sub-parallel to the volcanic arc. Over the years, three main structural models, resulting mainly from near-surface observations, have been proposed for Timor: (1) The Imbricate Model (Fig. 2a; Fitch & Hamilton 1974; Hamilton 1 9 7 9 ) i s based largely on marine geological and geophysical data (e.g. von der Borch 1979; Silver et al. 1983; Karig et al. 1987). In this model, Timor is interpreted as an accumulation of chaotic material imbricated against the hanging wall of a subduction trench, the Timor trough, and essentially forms a large accretionary prism. (2) The Overthrust Model (Fig. 2b) is probably the oldest model in which Timor was interpreted in terms of Alpine-style thrust sheets (e.g.
(3)
Wanner 1913). This model was dominantly based on surface geology where overthrust sheets of the Timor allochthon are well exposed. Subsequent workers (Carter et al. 1976; Barber et al. 1977; Barber 1979; Haile et al. 1979; Brown & Earle 1983; AudleyCharles 1981, 1986a, b; Price & AudleyCharles 1983, 1987; Harris 1989; AudleyCharles & Harris 1990) have made a clear distinction between allochthonous units of non-Australian origin and parautochthonous units derived from the Australian continent. The Rebound Model (Fig. 2c; Chamalaun & Grady 1978) suggests that the Australian continental margin entered a subduction zone in the vicinity of the Wetar Strait. Subsequently, the oceanic lithosphere detached from the continental part, resulting in the uplift of Timor by isostatic rebound on steep faults.
Combinations of these models have also been proposed (e.g. Charlton 1989; Charlton et al. 1991; Harris 1992) in which the parautochthon is divided into two parts, reflecting the step-wise nature of the collision. The main part, the 'underplated' parautochthon, is thought to have accreted in the early stages of collision by the sequential addition by underplating of imbricate thrust slices of the
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 47-60.
47
48
A. N. RICHARDSON • D. J. BLUNDELL o
o
o
South
0
10 ° 125 °
126 °
127 °
East
128 °
129 °
130 °
Fig. 1. Location map of the BANDA deep seismic reflection survey lines DAMAR and TIMOR, with shot point numbers shown in italic. Earthquake epicentres shown as stars; numbers relate to McCaffrey (1988), table l, letters a, b and c to events of 30 May 1988, 8 August 1988 and 21 May 1991 (Harvard centroid-moment tensor solutions). Barbed line represents the plate boundary with barbs in the direction of subduction. Arrow indicates relative plate motion. Bathymetry in metres.
outermost Australian continental margin to the base of the forearc complex. A second phase of collision is thought to have commenced in the Early Pliocene with the addition to southern Timor of younger parts of the Australian continental margin, the 'frontally accreted' parautochthon.
Data acquisition and processing The British Institutions Reflection Profiling Syndicate (BIRPS) joined with the Indonesian Marine Geological Institute to record two long, multichannel normal-incidence reflection profiles across the Banda arc west of Timor. The two subparallel traverses of the arc were designed to image the deep structure of the central part of the collision zone (Fig. 1). The proximity of the two lines to each other and to Timor and the geologically contiguous Minor Islands to the east increased confidence that observed reflections represent regional structural features consistent with surface
geology. The survey ship used was M/V GECOKappa which towed a 92-channel hydrophone streamer 4.6 km long. The airgun array of 7324 in 3 (120 litres) capacity was 70 m wide, designed to produce 97 bar m peak-to-peak pressure at frequencies of 3-62.5 Hz. Along the western line, 'TIMOR', shots were fired at 50 m intervals with 50 m receiver groups producing records up to 23 s two-way time (TWT) to show reflections at relatively high spatial resolution down to about 45 k m depth. The eastern line, ' D A M A R ' , was located where deep earthquakes are frequent and was therefore recorded to 37 s with 100 m shot intervals. This was designed to compare reflectors, resolved slightly less well than in the T I M O R profile, with earthquake hypocentres and mechanisms down to around 140 km depth. The seabed topography along both profiles is rugged and ranges from < 100 m to >4000 m below sea-level. This produced several sets of strong
CONTINENTAL COLLISION IN THE BANDA ARC NORTH
(a)
SOUTH
Imbricate Model Volcanic Arc
(b)
-13mor
Australian Continental Shelf
Overthrust Model Volcanic Arc
13mor
Australian
49
normal moveout curves for reflections at two-way travel times longer than around 10 s, equivalent to depths of around 25 km. For greater depths, indirect evidence from refraction experiments in the Timor trough and Banda Sea (Bowin et al. 1980), and from a general knowledge of p-wave velocities in the upper mantle had to be used for the depth conversion.
Interpretation Australian Continental Shelf (TIMOR SP 6400-8000, DAMAR SP 1-750, Figs 3 & 4)
(c)
Rebound Model
~
Volcanic Arc ~ ~
-timer
Australian Continental Shelf
Fig. 2. Models proposed for the structure of Timor (after Barber 1981).
seabed multiples, which partly obscure primary reflections from deeper levels, and a high level of random noise. The multiples were so easily recognized on the brute stack that they could be distinguished from primary reflections passing through them. Attempts to remove these multiples for the final stack sections were largely successful, but at the same time probably degraded the primary reflections. The brute stacks displayed the clearest images of primary reflections, multiples and diffraction patterns, and so were used more than the final stacks in preparing line drawing interpretations. Following the interpretation method of many other deep seismic experiments, discontinuous linear segments of reflections identified as primary on the seismic sections were traced on to a transparent film overlay. These line drawings more clearly delineate the overall structure and are treated as an interpretation of the geometries of the main reflections (Fig. 3). The principal reflections interpreted on the line drawings were depthcorrected and migrated in the plane of the section to give their depth geometry (Fig. 4). The only direct velocity inforhaation available was provided from the moveout analysis of reflected data from the 4.6 km long streamer. This offset was insufficient to discriminate between
The upper 2.1 s TWT beneath the seabed of the continental shelf portions of both sections consist of fairly parallel reflectors interpreted as sedimentary bedding. This is consistent with the tie of the TIMOR line with the Troubadour No. 1 oil exploration well (Fig. 1) which penetrated Recent to Triassic marine sediments lying unconformably on granite encountered at a depth of 3315m, equivalent to 2.16 s TWT. Extensional faults are evident in the upper 3 s TWT of both sections and appear to penetrate the crystalline basement as well as cut the entire sedimentary cover. These faults were probably initiated in the Jurassic when the Australian Northwest Shelf was last stretched (Bradshaw et al. 1988; Powell 1982). In the region of Sumba Island, where subduction is continuing at present, many faults that were originally related to rifting and spreading are still in active extension and are giving rise to earthquakes with extensional mechanisms. Some strong, north-dipping reflectors were also identified in the otherwise transparent upper crust (Fig. 3). These reflectors appear to sole out of the Jurassic faults to cut through the entire upper crust and continue into a reflective zone in the lower crust. The upper continental crust consists of crystalline rocks so it is most likely that these reflectors represent faults, following the interpretation of other deep seismic reflection surveys elsewhere (e.g. Matthews et al. 1990). Between 8 and 12 s TWT at the southern end of the TIMOR section and 7 and 14 s TWT at the southern end of the DAMAR section lies a zone of intense, sub-horizontal, laterally extensive, layered reflections. The base of this zone is consistent with an increase in interval velocity from 6.3-7.3 to over 8 km s-1 at this level on both profiles (Fig. 5) and corresponds well with the normal jump in p-wave velocity across the Moho to 8.1 km s-1 in the upper mantle (Fowler 1990). The velocity structure and character of reflection patterns in this zone is therefore much like that of the lower continental crust imaged by a number of
20
10
0
u
. . . . . . . .
I
I
I
2000
n
3000
n
2000 n
~
I
I
4000
n
I
1000
I
km
I
I
Shotpoint
I an
Shotpoint
I
6000
8000
S
Fig. 3. Line drawing interpretations of the BANDA seismic sections: (a) Timor; (b) Damar. Major reflectors and zones of high reflectivity are highlighted with dotted lines.
(b)
(s) 30
TWT
(a)
T W T (s)
1 0 ~
N
© Z
(/3
o
Z
CONTINENTAL COLLISION IN THE BANDA ARC
1000 I
'
'
'
I
2000
.
.
.
.
I
,
3000 ,
,
,
I
Shot Point 4000
,
,
,
,
I
5000
,
,
,
,
I
51
6000
,
.
,
.
I
7000
,
,
.
,
I
8000
,
,
,
,
0
q
50
km
I
100
NORTH
I
7o
I
8o
J
9*
SOUTH
50
km 1 O0
(a)
I
I
I
iI1~, 3000 4b, j I I I ~I
--
i
,
I
2500 I i i
Shot Point 2000 I
I
~ . - ~ - - - - - - ~ . _ . _
6
~
~ ,~~ a
| -
km
NORTH
1500 I
I
I
l
i
1000 i
i
,
I
,
500 ,
,
,
I
0 ,
,
,
,
I
................
~
100
150
I
-
~
50
~
100
" /
km
i
i
i
7°
8°
9°
150
SOUTH
(b)
Fig. 4. (a) Interpretation of the major structures recognized on the TIMOR profile (Fig. 3a) after depth correction and migration, and their association with earthquake data. Earthquake hypocentral locations, projected along the regional strike (073°E), are shown as black dots. Fault plane solutions are shown as hemisphere projections in the vertical plane of the profile. Upper crust: light grey; lower crust dark grey. (b) Interpretation of the major structures recognized on the DAMAR profile (Fig. 3b) after depth correction and migration, and their association with earthquake data.
other deep seismic reflection surveys from other areas (e.g. Mooney & Brocher 1987). The base of this zone of high reflectivity has been identified as the Moho (originally by Fuchs 1969) and it would be reasonable to apply the same interpretation here. This reflectivity within the lower continental crust has been associated with crustal extension, involving basaltic sill emplacement (Serpa & Voogd 1987; Warner 1990) and/or ductile shear zones (Reston 1987). Since this part of the Australian continent became a rifted margin in the Jurassic (Bradshaw et al. 1988; Powell 1982), it would be reasonable to assign the sub-horizontal reflectors in the lower crust to this age. On this basis the crustal thickness at the southern end of both sections is 35-40 km, consistent with a crustal thickness of 30-40 km determined in this area using seismic refraction data (Bowin et al. 1980).
The other observed deformation of the crust is a gentle flexure as it is bowed down beneath the Timor trough, bringing the Moho down to a depth there of 45-50 km.
Subduction Trench (TIMOR SP 6400-6450, D A M A R SP 700-750) Undisturbed, horizontally bedded sediments of 544 m thickness (0.48 s TWT) are present in the Timor Trough at the base of the trench on the TIMOR profile. Using a sedimentation rate of 480 m/Ma in the trough axis obtained from analysis of DSDP-262 (Johnston & Bowin 1981), situated SW of Timor, these sediments accumulated in a period of c. 1.1 Ma during which little convergence
20
15
10
(s)
T~q" I0
5
0
,
3000 !
2500
!
T
,
2000
Shotpoint
i
3000 I
i
1500
4000 i
Shotpoint
i
1000
5000 I
=
500
6000 i
7000
1
Fig. 5. Contour plots of interval velocity obtained from velocity analyses along (a) TIMOR and (b) DAMAR. Velocities in ms-1; contour interval 600 ms -1.
(b)
(a)
(s)
TWT
2000
1000
8000
L~h t~
CONTINENTAL COLLISION IN THE BANDA ARC has taken place. However, immediately south of the horizontal sediments, a small, north-dipping reflector observed on TIMOR at SP 6500 is interpreted as a thrust which has created a hanging wall anticline and delaminated the upper 0.6 s TWT of sediments, sliding them 200-400 m to the south. Other thrusts have formed beneath. A similar situation is observed on the DAMAR profile. Thus a stick-slip system of convergence appears to be operating SE of Timor in much the same way as has been observed in the Timor trough by Karig et al. (1987) SW of Timor. Interval velocities in excess of 5 km s-1 are encountered beneath the trench at 3 s and 2 s TWT below the seabed on the TIMOR and DAMAR profiles respectively (Fig. 5). This is in approximate agreement with refraction data (Bowin et al. 1980). Seismic velocities appropriate for granitic basement are thus encountered at about the same depth below the seabed both at the trench and on the continental shelf at the southern end of the two profiles. The crust, measured from the base of the highly reflective region, appears somewhat thicker at the trench than at the southern end of both sections (Fig. 4). Thickening could imply that the crust has shortened on, or just prior, to its arrival at the subduction trench. However, upper crustal shortening through inversion of extensional faults offsetting sedimentary layers is not evident, except within the trench. The zone of high reflectivity, interpreted as representing the lower crust, appears to terminate abruptly just north of the subduction trench in the DAMAR section. If the high reflectivity of the lower crust is interpreted as a structural fabric of Jurassic age, this would be destroyed if it were folded and sheared by compression. In this case, reflectivity would be drastically reduced, which could explain its abrupt termination.
Accretionary Prism (TIMOR SP 5800--6400, D A M A R SP 750-1100) A few kilometres north of the trench, reflectors within 4 s TWT below the seabed are less coherent. This is probably because of the disrupted nature of the sedimentary package. Deeper reflectors tend to be swamped by strong seabed multiples on the brute stack and noise on the final stack. Nevertheless, a series of southward-verging duplex structures is visible on both sections within what appears to be an accretionary prism (Fig. 3). This interpretation is consistent with the results of seismic reflection surveys to the west and southwest of Timor and south of Sumba (Karig et al. 1987; Van tier Werff et al. 1994) and southeast of
53
the Tanimbar Islands (Jongsma et al. 1989; Richardson 1994). Thus there is evidence that a proportion of the sedimentary sequence is accreted to the underside of the developing accretionary prism within a few kilometres of the trench. In this portion of both TIMOR and DAMAR profiles there is a general increase in seismic velocity from south to noah within 4 s TWT below the seabed (Fig. 5). This is probably the result of compaction and de-watering of the sedimentary package as it is broken up and accreted to the upper plate. North-dipping reflectors co-linear with sedimentary reflectors dipping beneath the trench appear to extend only c. 30 km north of the subduction trench on both profiles. North of this, regions of reflectors parallel to sedimentary reflectors dipping beneath the trench appear to occur in discrete 'packages' offset by prominent south-dipping reflectors (Fig. 4). Shallower packages appear to be offset to the north of deeper packages. The upper surface of these packages can be followed downwards in a relatively straight line to the base of the profiles and extrapolated to c. 140 km depth to join the northern (top) surface of the plate defined by earthquake hypocentres which has been defined in this area by McCaffrey (1989). There are three alternative interpretations for the actual physical nature of the reflectors within the packages. The north-dipping reflectors may represent: (1) remnant sedimentary material attached to the lower plate that has escaped accretion to the collision complex and has been carried to great depth; (2) faults or shear zones representing the zone of decoupling between the upper and lower plates; (3) accreted (underplated) sedimentary rocks. This last interpretation was adopted by Clowes et al. (1987) for similar deep reflectors dipping parallel to the subduction zone under Vancouver Island and by Moore et al. (1991) for the eastern Aleutian Islands, but the other two possibilities appear equally plausible. Whichever is preferred, the packages of reflectors may be interpreted as broadly representing the upper surface of the Australian plate beneath the collision complex. Therefore, it does not appear that the slab has detached above about 140 km depth.
Central Collision Complex (TIMOR SP 3200-5800, D A M A R SP 1100-2350) North of the accretionary prism, where the reflectors dip mainly to the north, both profiles pass into a region of predominantly south-dipping reflections. This geometry is similar to that seen on
54
A. N. RICHARDSON & D. J. BLUNDELL
high quality oil industry seismic reflection profiles in the Tanimbar Islands region made available by Union Texas (SE Asia) Inc. (Richardson 1994) and in the forearc region around Sumba, along strike from westem Timor (van der Werff et al. 1994). The transition zone between north-dipping and south-dipping reflectors appears complex in most cases. The south-dipping reflectors represent either sedimentary layers or low-angle faults because they do not have the characteristic shape of diffraction tails. They are unlikely to represent sedimentary layering because they appear fairly continuous to considerable depth whereas there is no sign of wellbedded sediments just below the seabed. On the TIMOR line (SP 32004400, 4700-5100), the patterns of the south-dipping reflectors are reminiscent of large-scale thrust sheets (Fig. 4). Similar features are observed on the seismic reflection profiles around the Tanimbar Islands where south-dipping reflectors terminate upwards in antiform folds. On the DAMAR profile, major south-dipping reflectors appear to offset northdipping packages of reflectors to the north (Fig. 3), consistent with north-vergent thrust faulting involving the subducted plate. In both the BANDA and Union Texas profiles, interval velocity inversions coincide with these south-dipping reflectors (Fig. 5). If a region with a normal velocity gradient were deformed into a stack of thrust sheets, velocity inversions could be expected to develop at the base of some thrust sheets where high velocity material overrides low velocity material. Shallow earthquakes occur in this area and McCaffrey (1988) has calculated hypocentres and focal mechanisms for a number of them (Figs 1 & 4). Because they have been projected up to 60 km onto the profile it is impossible to correlate particular earthquakes with individual reflectors. However, it is clear that there is a reasonable association and the dominantly thrust mechanisms of the earthquakes support the other evidence that the reflectors represent thrusts. The reflectors' appearance, their coincidence with interval velocity inversions and association with thrust mechanism earthquakes all suggest that they represent north-vergent thrusts which appear to reach depths of tens of kilometres, possibly penetrating into the mantle. Similar reflectors observed elsewhere have been interpreted as shear zones and ancient subduction zone traces (Matthews et al. 1990). The crest of the upper thrust sheet at SP 3900 on the TIMOR profile (Fig. 3) is a topographic feature of the seabed and forms the eastward, submarine extension of the island of Kisar. Kisar is offset from the trend of the main archipelago and consists of tightly folded felsic schists and mylonites which
can be correlated with the Aileu Formation of northern Timor (Richardson 1994). The presence of the thrust on the TIMOR profile in alignment supports the idea that Kisar may have been uplifted on an out-of-sequence thrust. However, the structural grain in Kisar runs approximately N-S (Richardson 1994), at right angles to the structural trend of the other islands. It is possible that Kisar has rotated by 90 ° during thrusting, or alternatively that it forms a lateral ramp. The seabed high at SP 3400 on the TIMOR section (Fig. 3) is mid-way between Kisar and the volcanic island of Romang. South-dipping reflectors with velocity inversions (Figs 3a & 5a) visible at SP 3200-3400 are interpreted as thrust faults. Interval velocities below the topographic high at TIMOR SP 3400 are similar to those below SP 3900 but are generally lower than the next topographic high to the north at SP 2900. The crustal thickness appears to be 18-23 km. The topographic high at SP 3400 on the TIMOR section is therefore interpreted to be similar in composition to Kisar and so may be the northernmost occurrence of continental rocks at the surface along this profile. Dating features identified on deep seismic reflection profiles is one of the most intractable problems within the field of deep seismics (Klemperer et al. 1990) unless a reflection can be traced to outcrop or there is other independent evidence. Since none of the south-dipping structures on TIMOR and DAMAR have been identified before, let alone recognised at the surface, dating them must rely on more speculative geometrical observations. Consequently, the cross-cutting geometric relationship of the south-dipping reflectors with the packages of north-dipping reflectors is used as evidence for their age. The packages of north-dipping reflectors are interpreted as representing the upper surface of the Australian plate beneath the collision complex. The amount of subduction is thought to have been extremely limited south of western Timor for the last 0.6 Ma (Karig et al. 1987). Horizontal sediments at the trench axis on the TIMOR profile imply that subduction has been inactive there for the last 1 Ma apart from a few hundred metres of south-vergent thrusting. The shallowest cross-cutting, southdipping structure on the TIMOR and DAMAR profiles cuts the boundary of north-dipping reflectors about 30 km north of the trench axis. Assuming the rate of convergence to be decreasing (Johnston & Bowin 1981), it would have taken this part of the lower plate c. 1.4 Ma to have reached its present location. Hence, this thrust cannot be older than c. 2.4 Ma which is approximately the age of onset of the current collision derived from evidence in the Timor trough south of western Timor (Johnston & Bowin 1981).
CONTINENTAL COLLISION IN THE BANDA ARC The top of the north-dipping lower plate and the deepest of the south-dipping thrusts define the collision orogen as a triangular-shaped wedge (Fig. 4). Sedimentary material is being added to the southern end of the wedge by underplating on north-dipping thrusts. However, the northern portion of the collision orogen is deformed on south-dipping, north-vergent thrust sheets. The exact depth of the Moho is difficult to constrain for this portion of the profiles because the poor resolution of the velocity picks leads to significant variations between adjacent values of interval velocity. However, assigning the base of the crust to the boundary between interval velocities less than and greater than 8000 ms -l, the crust appears to be thickest directly under the topographic highs: SP 4600 (> 46 km), SP 4000 (c. 30 km) and SP 3400 (> 36 km) on the TIMOR section and SP 1600 (c. 50km) and SP 2150 (c. 20 km) on the DAMAR section (Fig. 5). A 2D density model was produced to match a gravity profile along the TIMOR line recorded in a previous survey by Woodside et al. (1989). Although poorly constrained, the model serves to delimit the approximate thickness of the crust in this region. Two end-member versions of the model were produced. In one (Fig. 6a), which defines a minimum value for crustal thickness, the crust of the central collision complex was made 30 km thick, the same as the continental crust of the Australian Northwest Shelf. A reasonable fit
55
between calculated and observed gravity was achieved. In the second (Fig. 6b), crust with 2 6 7 0 k g m -3 density representing the central collision complex was enlarged at the expense of an upper mantle region with 3 1 0 0 k g m -3 density, whilst maintaining the fit of calculated with observed gravity. The second model gives an upper limit for which a reasonable match can be made with the observed gravity data. In this, crust of the central collision complex attains a maximum thickness of nearly 60 km.
Volcanic Arc (TIMOR SP 1700-3200, DAMAR SP 2350-2900) At the northern end of both seismic sections, submarine volcanoes associated with the subductionrelated volcanic islands of the Banda arc form prominent topographic highs. The volcanoes are said to be built on oceanic crust on the basis of seismic refraction data (Bowin et al. 1980). Long sub-horizontal reflectors just below the seabed are interpreted as delineating lava flows (D. Snyder, pers. comm., 1995). The TIMOR profile crosses the main volcanic ridge at SP 2950. Volcanoes at SP 2000 and SP 2350 on the TIMOR profile lie to the north of the main volcanic ridge and have a steepsided, narrow trough between them at SP 2130 (Fig. 3a). This sharp trough lies along strike from the surface expression of the Wetar Thrust visible
Gravityo (mgal) /
•
t: 1::o°o °
"1°°I -200 2670
~
~
i
o
2800
i
2800
50
3300 \,........ lOO
(km) 150 2o0
/ ?
// /.s
3300 150
3
3200
2OO
/ 2~o
2~6
S 1~o Distance
~o
.=b
from Trench(km)
~
-~o
-lOO
loo Distance
3200 ~o
~
fromTrench(km)
-~o
250
Fig. 6. Gravity models to constrain estimates of crustal thickness in the central part of the Banda orogen. Densities of polygons are in kg m-3.
56
A. N. RICHARDSON 8Z; D. J. BLUNDELL
on sonar imagery (Masson et al. 1991). A southdipping reflector reaches the surface at the base of this trough and can be traced 50 km southwards to a depth of c. 20 km. Two earthquakes with thrust mechanisms project on to the TIMOR section within the horizontal range of this reflector (Fig. 4a) although with significantly greater hypocentral depths. It may be significant, however, that the south-dipping nodal planes of both these events dip at an angle close to the migrated dip of the reflector. This reflector is therefore interpreted as evidence for the along-strike extension of the Wetar Thrust, associated with current movement. A number of major, south-dipping reflectors are also present in this portion of the DAMAR section which extend close to the base of the seismic record (Fig. 3b). These structures therefore appear to be the dominant fabric of the upper plate on both profiles. Some shorter north-dipping reflectors are present on both profiles but none are as long or as continuous as the south-dipping reflectors (Fig. 4). The volcanic arc portion of the DAMAR profile includes the greater proportion of earthquakes less than 100 km deep along the transect (Fig. 4b). Most of these earthquakes have thrust mechanisms, with one nodal plane which has a south-dipping component, and plot close to the major reflectors. It is not possible to correlate specific earthquakes with individual reflectors because of the large envelope of uncertainty in hypocentral locations and the distance that they have been projected on to the section. However, the fact that earthquakes with thrust mechanisms are occurring in this region does suggest that the south-dipping reflectors may represent active thrust faults which penetrate the crust and upper mantle to depths of up to 100 km. On the TIMOR profile, the crustal thickness (defined where Win t > 8000 m s-1) varies from c. 12 km to over 20 km beneath the volcanoes. This is abnormally thick for oceanic crust, even including the volcanic edifices. Interval velocities > 8000 m s-1 are attained at depths of 8-19 km below sea-level on the DAMAR profile (Fig. 5b) which implies that the crust is thinner than on the TIMOR profile but still thicker than normal oceanic crust.
The backarc region (TIMOR SP 1-1700, D A M A R SP 2900-3200) A number of marine geophysical surveys have been conducted in the backarc region. From Lombok to Maopora, the backarc region is dominated by south-dipping reflectors beneath the volcanoes, sometimes overlain by an accretionary prism (Silver et al. 1983). Likewise, the south-dipping reflectors in this part of the BANDA profiles are
interpreted as thrusts because of their appearance on seismic profiles. The two most northerly volcanoes on the TIMOR section, at SP 400 and SP 1200, are isolated seamounts built on the oceanic crust of the Banda Sea. The northernmost edifice rises above sea-level to form the active volcano, Gunung Api. It is not clear why this volcano is active so far to the north of the subduction zone. The volcanoes traversed by the TIMOR profile appear to line up with Gunung Api and the seamount in between them (Fig. 1). The NNW-SSE lineation of seamounts suggests that they may have formed over a 'leaky' transform fault (D. Snyder, pers. comm. 1992). The DAMAR profile passes close to one of the seamounts at SP 2600 and directly over another at SP 2800, a few kilometres to the north. The basin areas in between the volcanoes show clear sedimentary layering near the surface and long, gently dipping reflectors down to the depth of the first multiple (Fig. 3). The velocity pick at TIMOR SP 1599 in the centre of one of these basins is interpreted as representing 1840 m of sediments (cf. oceanic lithosphere Layer 1, Vint = 1900-2400m s-1) and 5860 m of volcanics and intrusives (oceanic lithosphere Layers 2 and 3, Vin t - - 4 2 1 7 m s-1) overlying upper mantle material (oceanic lithosphere Layer 4, Vin t -- 8058 m s-1) at 7.7 km below the sea-bed. Most velocity picks at the extreme northern end of the DAMAR profile have velocities typical of the upper mantle (Vin t > 8000 m s-1) occurring 3 . 7 4 km below the seabed. All these velocity structures broadly agree with seismic refraction data (Bowin et al. 1980) and are therefore interpreted as representing normal oceanic crust. The northern part of the DAMAR line appears to traverse rather rough sea floor. However, one of the major south-dipping thrusts reaches the surface in this region and the rough topography may represent a small accretionary prism.
Summary The southern end of both profiles, which traverse the Australian Continental Shelf, is interpreted in a similar manner to deep seismic reflection profiles from other passive continental shelf areas. Wellstratified sediments overlie a largely reflectionfree upper crustal crystalline basement and a highly reflective lower crust. This arrangement is common in continental regions that have undergone extension. Structures at shallow depths under the northern slope of the trench (or the southern part of the collision zone) dip exclusively northwards and are interpreted to represent thrusts and tilted sediments in an accretionary prism. Deeper, north-dipping
CONTINENTAL COLLISION IN THE BANDA ARC structures are interpreted to be related to the interface between the lower and upper plates. This is supported by the extrapolation of this surface to greater depths where it corresponds well with the upper surface of the Australian plate as defined by earthquakes. The northern part of the collision zone is dominated by predominantly south-dipping structures which extend to sub-crustal depth. These are interpreted as thrusts along which the lateral shortening and crustal thickening have taken place. The thrusts are estimated to be no older than 2.4 Ma, the age of initiation of the collision of the Australian margin with the arc, and seem to be active at the present day at the northern end of the collision zone. A maximum crustal thickness of c. 60 km is calculated under the topographically highest part of the collision zone on the basis of calculated seismic interval velocities. If this crust had been constructed from rifted and thinned Australian margin material, the crust may now be more than double its original thickness which would imply over 50% (100-150 km) shortening in the continental portion of the collision zone. The oceanic crust also appears to be anomalously thick beneath the volcanoes on the basis of calculated seismic interval velocities. It appears likely that the oceanic crust has, like the continental crust, been shortened and thickened on predominantly south-dipping thrusts. The fact that these major thrusts are visible beneath and north of the volcanic arc suggests that the collision zone extends north of the non-volcanic arc and is in total about 200 km wide, comprising both continental and oceanic lithosphere, and may have accommodated over 200 km of shortening. The overall, large-scale character of the collision zone is dominated by two sets of divergent struc-
N
57
tures. The southern set is related to the subducting (lower) plate and dips in the same direction as subduction. The northern set, in the upper plate, is antithetic to this. Figure 7 is a cartoon to highlight these primary features.
The deep-seated nature of the orogen In considering the evolution of the collision complex as an orogen, three dates are significant: (1) The Banda allochthon in Timor was apparently uplifted at c. 37 Ma according to K-Ar dating (Sopaheluwakan 1990). (2) Accurate radiometric dating on the Aileu Formation on the north coast of Timor gives an age of metamorphism associated with gradual cooling of 8 Ma (Berry & Grady 1981) and a Mid-Miocene unconformity has been inferred across Timor (Audley-Charles 1968). (3) Pliocene deformation is consistently recorded by sedimentary sequences elsewhere in Timor (Audley-Charles 1968) and the start of the most recent collision has been dated at 2.4 Ma from evidence in the Timor trough (Johnston & Bowin 1981). The extent of crustal material in the central collision complex is defined by seismic velocity and gravity modelling. The crustal material has the form of an inverted triangle about 50 km (37-60) deep and 125 km (135-160) across on the TIMOR profile, which has a cross-sectional area of 3750 km 2. This thickness of crustal material could, conceivably, have been constructed by a number of means. However, it cannot have been constructed entirely from sediments accreted from the Australian plate since 2.4 Ma. Using the plate velocity vectors in DeMets et al. 1990 (Fig. 1),
Australian Continental S Timor Micro-Continent AccretionaryPrism ~li,helf _ Kisaj~ ~, iSubd ,.)., ~lt~ ctionTrench , , ~
Volcanic Arc ~1~ ~
~60 km
Moho ~"
/~5. /
~i~;~ ~ ; ~
0
km
1O0
I
I
I
]
Fig. 7. Cartoon structural cross-section of the Banda orogen based on our interpretation of the BANDA deep seismic profiles. Australian continental shelf sediments are shaded light grey. Highly reflective lower crust is shaded dark grey.
58
A.N. RICHARDSON t~ D. J. BLUNDELL
156 km (65 km/Ma for 2.4 Ma) of convergence normal to the arc could have occurred. This is an upper estimate as it does not take account of decreasing convergence velocity after collision. An upper estimate of the thickness of the sediments between the Australian coast and Timor from the Banda profiles is about 3 km. Assuming that this entire thickness of sediments is added to the growing orogen and that strike-slip movement has not removed material from the normal plane, a cross-sectional area of 468 km 2 would result (3 × 156 km). This is far less than the estimated crustal area of 3750 km 2. A collision event, adding crustal material to the orogen, must therefore have preceded the Pliocene collision event. Nor is it possible that the Banda orogen was constructed entirely from sediments accreted from the Australian Plate since 8 Ma. Using the same argument as above, a sediment volume accumulated over 8 Ma of convergence would amount to only 1560 km 2 in cross-sectional area, which is still substantially less than the estimated 3750 km 2 of crust. There must therefore be an additional piece of crystalline continental crustal material within the collision complex making up a major portion of its bulk. It is thus proposed that a micro-continental fragment, or perhaps an outer margin 'high', lay a few hundred kilometres to the north of the Australian Northwest Shelf and this was incorporated into the collision complex before the arrival of the Australian continental margin. The material separating the micro-continent from the Australian craton was probably oceanic in origin but could have been extremely stretched and attenuated continental lithosphere. This micro-continental fragment collided with the subduction zone at c. 8 Ma in the Miocene. This caused metamorphism
of the Aileu Formation at the leading edge of the micro-continental fragment ( ' m e t a m o r p h o s e d parautochthon' on Timor and Kisar) and folding in the unmetamorphosed rocks. The micro-continent was thrust beneath the forearc region, thus uplifting the Banda Allochthon, and may have induced Late Miocene inversion structures reported from the Timor Sea (MacDaniel 1988; Pattillo & Nicholls 1990). The Australian continental margin arrived at the collision zone 2.4 Ma ago and started downwarping into the subduction zone. The continental crust could not be subducted very far. Subsequently, the Australian continental margin has acted as a bulldozer, shortening and uplifting continental and oceanic parts of the collision zone on deep seated thrusts dipping antithetic to subduction. The oldest thrusts are presumably the more southerly ones. Since then the locus of activity has propagated northwards in a normal fashion until it has reached the backarc region where thrusts are active at the present day. Uplift is currently continuing as evidenced by several hundred metres of elevation of Pliocene coral reef terraces on Alor, Atauro and Wetar (Nishimura & Suparka 1986) and Sumba (Piranzolli et al. 1991). The authors wish to thank Dr D. Snyder for his major efforts in conducting the BANDA survey and supervising the data processing, and to him and their other BIRPS colleagues for ready access to the deep seismic reflection data. BIRPS is supported by the Natural Environment Research Council (NERC) and all BIRPS data are available, at the cost of reproduction, from the Marine Geophysics Programme Manager, British Geological Survey, Murchison House, West Mains Road, Edinburgh EH9 3LA. ANR is grateful to NERC for the award of a studentship during the period of this research.
R e f e r e n c e s
AUDLEY-CHARLES, M. G. 1968. The Geology of Portuguese llmor. Geological Society, London, Memoir, 4. 1981. Geometrical problems and implications of large-scale overthrusting in the Banda ArcAustralian Margin collision zone. In: MCCLAY, K. R. & PRICE, N. J. (eds) Thrust and Nappe Tectonics. Geological Society, London, Special Publications, 9, 407-416. - 1986a. Timor-Tanimbar Trough: the foreland basin to the evolving Banda Orogen. In: ALLEN, P. A. 8£ HOMEWOOD, P. (eds) Foreland Basins. International Association of Sedimentology Special Publication, 8, 91-102. - 1986b. Rates of Neogene and Quaternary tectonic movements in the southern Banda Arc based on micropalaeontology. Journal of the Geological Society, London, 143, 161-175. - - & HARRIS,R. A. 1990. Allochthonous terranes of the
South West Pacific and Indonesia. Philosophical Transactions of the Royal Society of London, A331, 571-587. BARBER, A. J. 1979. Structural interpretations of the island of Timor. South East Asian Petroleum Exploration Association Proceedings, 4, 9-12. 1981. Structural interpretations of the island of Timor, eastern Indonesia. In: BARBER, A. J. WIRYOSUJONO,S. (eds) The Geology and Tectonics of Eastern Indonesia. GRDC Special Publication, 2, 183-198. - - - , AUDLEY-CHARLES,M. G. & CARTER, D. J. 1977. Thrust tectonics in Timor. Journal of the Geological Society of Australia, 24, 51-62. BERRY, R. E & GRADY, A. E. 1981. The age of the major orogenesis in Timor. In: BARBER, A. J. & WIRYOSUJONO,S. (eds) The Geology and Tectonics of Eastern Indonesia. GRDC Special Publication, 2, 171-181.
CONTINENTAL COLLISION IN THE BANDA ARC BOWlN, C., PURDY, G. M., JOHNSTON, C., SHOR, G., LAWVER,L., HARTONO,H. M. S. & JEZEK,P. 1980. Arc-continent collision in the Banda Sea region. AAPG Bulletin, 64, 868-918. BRADSHAW, M. T., YEATES, A. N, BEYNON, R. M., BRAKEL,A. T., LANGFORD,R. P., TOTTERDELL,J. M. & YEUNG, M. 1988. Paleogeographic evolution of the North West Shelf region. In: PURCELL,P. G. & PURCELL, R. R. (eds) The North West Shelf Australia. Proceedings of the NW Shelf Symposium 1988, 29-54. BROWN, M. & EARLE, M. M. 1983. Cordierite-bearing schists and gneisses from Timor, Eastern Indonesia: P-T implications of metamorphism and tectonic implications. Journal of Metamorphic Geology, 1, 183-203. CARTER, D. J., AUDLEY-CHARLES,M. G. & BARBER,A. J. 1976. Stratigraphical analysis of island arccontinental margin collision in eastern Indonesia. Journal of the Geological Society, London, 132, 179-198. CHAMALAUN, E H. & GRADY, A. 1978. The tectonic development of Timor: a new model and its implications for petroleum exploration. Australian Petroleum Exploration Association Journal, 18, 102-108. CHARLTON, T. R. 1989. Stratigraphic correlation across an arc--continent collision zone: Timor and the Australian Northwest Shelf. Australian Journal of Earth Science, 36, 264-274. ~, BARBER, A. J. & BARKHAM, S. T. 1991. The structural evolution of the Timor collision complex, eastern Indonesia. Journal of Structural Geology, 13, 489-500. CLOWES, R. M., BRANDON,M. T., GREEN,A. G., YORATH, C. J., SUTHERLANDBROWN, A., KANASEWICH,E. R. & Spencer, C. 1987. LITHOPROBE - Southern Vancouver Island: Cenozoic subduction complex imaged by deep seismic reflections. Canadian Journal of Earth Science, 24, 31-51. DE METS, C., GORDON,R. G., ARGUS,D. E & STEIN, S. 1990. Current plate motions. Geophysical Journal International, 101, 425-478. FITCH, T. J. & HAMILTON,W. 1974. Reply to AudleyCharles & Milsom (1974). Journal of Geophysical Research, 79, 4982. FOWLER, C. M. R. 1990. The Solid Earth. Cambridge University Press. Fucns, K. 1969. On the properties of deep seismic reflectors. Zeitschriftfiir Geophysik, 35, 133-149. HAILE, N. S., BARBER, A. J. & CARTER, D. J. 1979. Mesozoic cherts on crystalline schists in Sulawesi and Timor. Journal of the Geological Society, London, 136, 65-70. HAMILTON,W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078. HARRIS, R. A. 1989. Processes of aUochthon emplace-
ment, with special reference to the Brooks Range Ophiolite, Alaska and Iimor, Indonesia. PhD thesis, University of London. 1992. Temporal distribution of strain in the active Banda Orogen: a reconciliation of rival hypotheses. Journal of SE Asian Earth Sciences, 6, 373-386. JOmqSTON, C. R. & BOWIN, C. 1981. Crustal reaction
59
resulting from the mid-Pliocene to Recent continent-island arc collision in the Timor region.
Australian Bureau of Mineral Resources Journal of Geology and Geophysics, 6, 223-243. JONGSMA,D., WOODSIDE,J. M., HUSON,W., SUPARKA,S. & KADARISMAN,D. 1989. Geophysics and tentative Late Cenozoic seismic stratigraphy of the Banda Arc-Australian continent collision zone along three transects. Netherlands Journal of Sea Research, 24, 205-229. KARIG, D. E., BARBER, A. J., CHARLTON, T. R., KLEMPERER,S. & HUSSON~,D. M. 1987. Nature and distribution of deformation across the Banda Arc-Australian collision zone at Timor. Geological Society of America Bulletin, 98, 18-32. KLEMPERER, S. L., HOBBS, R. W. & FREEMAN,B. 1990. Dating the source of lower crustal reflectivity using BIRPS deep seismic profiles across the Iapetus Suture. Tectonophysics, 173, 445-454. MACDAmEL, R. P. 1988. The geological evolution and hydrocarbon potential of the western Timor Sea region. Australian Petroleum Exploration Association Journal, 28, 270-284. MASSON, D. G., MILSOM, J., BARBER, A. J., SIKUMBANG, N. & DWlYANTO,B. 1991. Recent tectonics around the island of Timor, eastern Indonesia. Marine & Petroleum Geology, 8, 35-49. MATTHEWS, D. & THE BIRPS GROUP. 1990. Progress in BIRPS deep seismic reflection profiling around the British Isles. Tectonophysics, 173, 387-396. MCCAFFREY, R. 1988. Active tectonics of the eastern Sunda and Banda Arcs. Journal of Geophysical Research, 93, 15 163-15 182, 1989. Seismological constraints and speculations on Banda Arc tectonics. Netherlands Journal of Sea Research, 24, 141-152. MOONEY, W. D. & BROCHER, T. M. 1987. Coincident seismic reflection/refraction studies of the continental lithosphere: a global review.
Geophysical Journal of the Royal Astronomical Society, 89, 1-6. MOORE, J. C., DIEBOLD, J., FISHER, M. A., SAMPLe, J., BROCHER, T. ET AL. 1991. EDGE deep seismic reflection transect of the eastern Aleutian arc-trench layered lower crust reveals underplating and continental growth. Geology, 19, 420-424. NISHIMURA, S. & SUPARKA,S. 1986. Tectonic development of East Indonesia. Journal of SE Asian Earth Science, 1, 45-57. PATT~LLO, J. & NlCnOLLS, P. J. 1990. A tectonostratigraphic framework for the Vulcan graben, Timor Sea region. Australian Petroleum Exploration Association Journal, 30, 27-51. PIRANZOLLI, P. A., RADTKE, U., HANTORO, W. S., JOUANNIC, C., HOANG, C. T., CAUSSE,C. & BOREL BEST, M. 1991. Quaternary raised coral reef terraces on Sumba Island, Indonesia. Science, 252, 18341837. POWELL, D. E. 1982. The North West Australian continental margin. Philosophical Transactions of the Royal Society of London, A305, 45--62. PRICE, N. J. & Atn)LEY-CHARLES, M. G. 1983. Plate rupture by hydraulic fracture resulting in overthrusting. Nature, 306, 572-575.
60
A. N. RICHARDSON •
& -1987. Tectonic collision processes after plate rupture. Tectonophysics, 140, 121-129. RESTON, T. J. 1987. Spatial interference, reflection character and the structure of the lower crust under extension. Results from 2D seismic modelling. Annales Geophysicae, 5 (ser. B), 339-48. RICHARDSON, A. N. 1994. Lithospheric structure and dynamics of the Banda Arc, Eastern Indonesia. PhD thesis, University of London. SERPA, L. 8z VOOGD,B. DE 1987. Deep seismic reflection evidence for the role of extension in the evolution of continental crust. Geophysical Journal of the Royal Astronomical Society, 89, 55-60. SILVER, E. A., REED, D., MCCAFFREY,R. & JOYODIWIRYO, Y. 1983. Back arc thrusting in the eastern Sunda Arc, Indonesia: a consequence of arc-continent collision. Journal of Geophysical Research, 88, 7429-7448. SOPAHELUWAKAN, J. 1990. Ophiolite obduction in the
-
-
Mutis Complex, Timor, eastern Indonesia: an
D. J. BLUNDELL
example of inverted, isobaric, pressure metamorphism. PhD
medium-high
thesis, Free University, Amsterdam. VAN DER WERFF, W., KUSNIDA, D., PRASETYO, H. & WEERING, TJ. C. E. VAN 1994. Origin of the Sumba forearc basin. Marine and Petroleum Geology, 11, 363-374. YON DER BORCH, C. C. 1979. Continent-island arc collision in the Banda Arc. Tectonophysics, 54, 169-193. WANNER,J. 1913. Geologie von West Timor. Geologische Rundschau, 4, 136-150. WARNER, M. 1990. Basalts, water, or shear zones in the lower continental crust? Tectonophysics, 173, 163-174. WOODSIDE, J. M., JONGSMA,D., THOMMERET,M., STRANG VAN HEES, G. & PUNTODEWO. 1989. Gravity and magnetic field measurements in the eastern Banda Sea. Netherlands Journal of Sea Research, 24, 185-203.
Geophysical evidence for local indentor tectonics in the Banda arc east of Timor D A V I D B. S N Y D E R 1, J O H N M I L S O M 2, & H A R D I
PRASETYO 3
1BIRPS, Bullard Laboratories, University of Cambridge, Madingley Road, Cambridge CB3 0EZ, UK 2 Department of Geological Sciences, University College London, Gower Street, London WC1E 6BT, UK 3 Marine Geological Institute, J1. Dr. Junjunan 236, Bandung 40174, Indonesia Abstract: Two deep seismic reflection profiles and gravity measurements collected across the
Banda arc east of Timor help to characterize crustal and uppermost mantle structures in the region where arc-continent collision is thought to be furthest advanced. Reflectors beneath the Sahul Platform of the Australian shelf indicate geometries consistent with older extensional rift structures overprinted by more recent horizontal shortening. To the north, the negative Bouguer gravity feature associated with the southern parts of the accretionary complex is unusually broad and deep. Post-collisional sediments are thin where this feature is crossed by the seismic lines, implying that older sediments or the crust are anomalously thick. Still further north, the forearc basin is notably narrow near eastern Timor and contains few sediments, most undeformed. The backarc region to the north is remarkable for the presence of a N-S trending line of seamounts culminating in an active volcano, Gunung Api, which is situated 400 km above the WadatiBenioff zone. Reflection profiles across and along the seamount chain indicate underplating at Moho depths and a fault which appears to have acted as a conduit for basalt eruptions at the surface. Collectively, these observations imply anomalously thick and bouyant crust beneath the Banda arc east of Timor, and suggest two possible causes. Either a local promontory in the irregular boundary of the Australian craton was underthrust beneath the volcanic arc and forearc to 50-70 km depths or a Palaeozoic basin similar to the nearby Bonaparte Basin was underthrust and its former crustal structure inverted and thickened to form the buoyant crust. The new seismic reflection data help to locate the anomalously thick crust implied by gravity anomalies by defining the leaky fracture zone containing Gunung Api. It is inferred that the fracture propagates "ahead of the indenting wedge of underthrust, thickened Australian crust.
The Banda arc between Indonesia and Australia possesses some of the best examples of features associated with the collision of an arc and a continental margin (e.g. Hamilton 1979, 1988; McCaffrey 1988), and also displays distinctive segmentation along strike (Silver et al. 1983; Karig et al. 1987; Jongsma et al. 1989; Harris 1991). A pair of recent deep seismic reflection profiles were sited across one of these distinctive segments, immediately to the east of the island of Timor, in order to better characterize structural geometries throughout the lithosphere (Fig. 1). Focal mechanisms of large shallow earthquakes (McCaffrey 1988) and shallow seismic reflection profiles (Silver et al. 1983) and recent GPS measurements (Genrich et al. 1994) indicate that the Wetar thrust, a northward verging thrust on the northern edge of the inactive volcanic arc (Fig. 1), today absorbs most of the 75 mm a-1 (DeMets et al. 1990) relative convergence between Australia and the Banda Sea. In contrast, shallow seismic reflection profiles within the Timor Sea show reflector geometries in
Mesozoic and Cenozoic sediments that are typical of active subduction zone trenches worldwide (Karig et al. 1987). The Timor trough cannot simultaneously absorb plate convergence and transfer the motion of Australia to the volcanic arc. Loci of horizontal shortening have probably migrated northward across the 2 0 0 k m wide convergence zone during the past 2 Ma, with specific faults, active for short periods, acting as mechanisms of strain partitioning (e.g. McCaffrey 1992, 1996). A series of balanced cross-sections of the accretionary complex showed that m a x i m u m horizontal shortening, and hence the most evolved arc-normal convergence, has occurred in eastern Timor (Johnston & Bowin 1981; Harris 1991). Eastern Timor lies midway along the inferred convergence margin of the Australian continent with SE Asia (Fig. 1). Low 3He-4He ratios in volcanic rocks east of Flores were interpreted as due to the subduction and melting of continental crust at c. 150 km depths whereas to the west the
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 61-73.
61
62
D.B. SNYDERETAL
•
~
..
~
:,
~
~
"
,
,IRIAN JAYA
KAIIS. /
5o r,L.
xt,
,,
--) J
(
/
SEA
I
~
3
AUS
~
ff----
200km
10"
F f,,,I f
P /
,"it" ; L
/
15"S
/
t
q
// ,
/
,J
iiiii!i!!!!iii ......,.....o,..,.,..... •
C
ii~iiiiii!!::!i~ii::~i!iii~i~
AUSTRAUA
/ /
I
,r
I
'~
120*
I
I
130*E
Fig. 1. Regional location map showing the Archaean blocks on the Australian mainland (shading: ST = Sturt, K = Kimberley), principal landmasses, the 200 m bathymetric contour (short dash) that approximately defines the limit of the continental shelf, the 3 km contour (long dash) that approximates the limits of oceanic crust in the Indian ocean, the new seismic reflection lines (heavy solid lines), older refraction profiles (medium solid lines) with crustal thickness indicated in kilometres, and the 100, 300, and 500 km contours of the Wadati-Benioff zone (Cardwell & Isacks 1978). Large solid dots indicate the transition from subducted continental crust to subducted oceanic crust (to the west) within the subduction zone as determined from He isotope analysis of erupted lavas (Hilton et al. 1993). The vector labelled AUS---~SEA represents the local relative convergence vector between Australia and Eurasia (DeMets et al. 1990). C = Canning Offshore Basin, L = Londonderry High, A = Ashmore Platform, V = Vulcan sub-basin, P = Petrel sub-basin of the Bonaparte Basin, S = Sahul platform, M = Malita Graben, MS = Money Shoal Graben, B = Browse Basin, and WT = Wetar Thrust.
He source was inferred to be subducted oceanic (MORB) rocks (Hilton et al. 1992). The change in isotope ratios coincides with strike-slip faults at the surface (Breen et al. 1989) and crustal-scale fault planes defined by micro-earthquakes (McCaffrey et al. 1985) which are co-linear with the 3 k m bathymetric contour defining the N W margin o f the Australian continental shelf (Fig. 1). Thus defined, the western edge of the subducted continental margin lies c. 300 k m west of the study area. A similar distance along the strike of the volcanic arc
to the east of eastern Timor, the Wadati-Benioff zone (Fig. 1) curves sharply back onto itself to the N W (Cardwell & Isacks 1978). Here the gross Australia-SE Asia convergence vector (DeMets et al. 1990) is almost parallel to the volcanic arc, i m p l y i n g largely strike-slip d i s p l a c e m e n t s on structures parallel to the arc (see also McCaffrey 1996). If eastem Timor represents the most evolved segment of the Banda arc, local anomalously thickened crust or thick crust inherited from a
LOCAL INDENTOR TECTONICS IN THE BANDA ARC former promontory on the Australian shelf margin may be the cause. The current complex tectonic geometries associated with Irian Jaya prevent precise reconstruction of the former continental margin east of present-day Tanimbar, but New Guinea is thought to represent a large indentor wedge of Australian continental crust plowing into and bending the margin of SE Asia (Pigram et al. 1989). The east Timor region may contain a smaller promontory on the western margin of this larger indentor, but one that produced significant localized shortening across the Banda arc. Evidence for a promontory which has already subducted or underthrust the Banda arc must be sought using reflection and refraction profiles, gravity field anomalies and bathymetric observations. This paper describes and discusses this evidence. The variability in crustal structure observed in the unsubducted part of the Australian shelf south of Timor serves as the starting point, and it suggests similar complexity in subducted parts. Inversion of subducted sedimentary basin structures and thickening of sub-basin crust previously thinned during Devonian-Permian rifting provides an additional mechanism that can produce anomalously thick and buoyant crust east of Timor.
Structures of the northern australian shelf The Australian continental shelf south of Timor is covered by a relatively dense network of deep seismic reflection profiles plus gravity, magnetic and bathymetric measurements (e.g. O'Brien et al. 1993; AGSO North West Shelf Study Group 1994). Sedimentary basin structures are well understood, but crustal thickness is poorly constrained. The nearby Australian mainland contains two large Archaean cratons, the Kimberley and Sturt blocks (Fig. 1). Immediately offshore between these two blocks and continuing northwards to the Malita Graben lies the Bonaparte basin with its Petrel and Vulcan sub-basins. It, and the Canning Basin to the southwest, form composite Late Devonian to Early Carboniferous intra-cratonic rift basins with axes perpendicular to the Australian margin and associated with NE-SW extension (AGSO North West Shelf Study Group 1994). Mid-Carboniferous to Early Permian rifting that led to break-up of Gondwanaland and separation of 'Sibumasu' (Sengor 1987) from NW Australia and the opening of 'Neo-Tethys' (Veevers 1988), also initiated a series of generally NE-trending depocentres which constitute the Westralian Superbasin of Yeates et al. (1987). The superbasin's internal platforms and basins are orientated parallel to the continental margin and thus angularly super-
63
imposed on the older radial rift basins. In the study area these later features include the Sahul and Ashmore Platforms, which show little evidence of Mesozoic extension, and the Malita, Vulcan and Browse Basins (Fig. 1). The start of the break-up phase and the subsequent development of the continental margin was recorded as a major Lower-Middle Carboniferous unconformity. MidCarboniferous to Lower Permian basin fill may actually be a sag-phase sequence deposited following break-up and the 100-400% extension thought to have characterized this rift episode (Etheridge & O'Brien 1994). The Sahul Platform occupies c. 30 000 km 2 of the Australian shelf immediately south of eastern Timor. The Malita Graben separates it from offshore parts of the Sturt block of the Australian mainland (Fig. 1). Seismic reflection profiles and drill holes indicate that c. 6000 m of Carboniferous(?) and as much as 10 000 m of later sediments were deposited in the Malita Graben, confirmed by a core of 3300 m of post-Permian sediments on the Sahul Platform (AGSO North West Shelf Study Group 1994). The uniformity of sediment cover on the Sahul Platform suggests thermal subsidence rather than subsidence synchronous with rifting and implies little thinning of this crustal block. To the north, very young structures associated with convergence tectonics overprint older extensional structures (Figs 3 & 4). Although no evidence exists on seismic reflection sections to indicate thicknesses of post-Permian sediments as great as those observed in the Malita Graben, reflectors beneath the Jurassic and Permian levels indicate a possible older basin north of the Sahul Platform and currently underlying the Timor trough (Figs 3 & 4).
Crustal thickness estimates In contrast to the wealth of information available about the upper crustal and basin structure of the Australian continental shelf, its crustal thickness and that of the Banda arc itself are less well documented. Three refraction profiles provide general estimates for the thickness of Australian continental crust in this area, but these show variability and uncertainty commensurate with the basin structure described above (Jacobson et al. 1979). The three profiles were orientated parallel to the shelf margin, and two lie in the Timor trough (Fig. 1). Crustal thickness estimates range from 31-40 km within an area of 500 000 km 2, and profiles 150kin long show crustal thickness variations of 4-8 km (Bowin et al. 1980). A fourth profile within the Banda Sea, east of Wetar (Figs 1 & 2), indicated a typical oceanic crustal thickness of 11 km (Bowin et al. 1980).
64
D. B. SNYDER ET AL.
GUNUNG
4
B A N D A SEA
API
I I I
t-
,~.~.% .~-""~ /
i© !
i
5O IO4
"1
i
I
/
WETAR
....~
-/ /
" BANDA
"".
l /
I !
i
\
SER~ATA
KISAR
:I
I
I
.500o
I
I I I I I l
BANOA
+
l
127*
ARC
, lOOO
+9"s ! 29" E
0~~
~i"
searnount
island
D
~, l q # '
Troubadour Fig. 2. Location of major known submarine features and the new seismic lines TIMOR, API, and DAMAR. Dashed lines indicate older scientific expeditions in this area. The Troubadour well (solid dot) on the Australian shelf margin (Australian Geological Survey Organisation, unpublished data) is located at SP 7995 of the TIMOR line. Onland areas shaded in black contain exposures of ophiolites. M 12 is a refraction profile (Bowin et al. 1980).
By themselves, these thickness estimates are too sparse to characterize the crust, but crustal thicknesses can be extrapolated or inferred over the region by using reflection profiles. Penetrative reflectivity within extended continental margins generally disappears below the Moho (e.g. Matthews 1986). The intersecting reflection and refraction profiles in the study area (Fig. 1) demonstrate that this relationship applies here, at least
locally. By assuming that it applies for the entire northern Australian shelf, a gradual eastward thickening of the crust from about 30-40 km is inferred between eastern Timor and Tanimbar. This pattern of crustal thickening is not locally correlated with the known basin structure unless the Sahul Platform continues to the NE as argued by Charlton et al. (1991) using stratigraphic criteria. Unrecognized basins, similar to the Malita Graben
65
LOCAL INDENTOR TECTONICS IN THE BANDA ARC N
S
o--
_ --
,.
-"
~
_
.
-_-_~.-.
r .
.
.
.
.
.
.
"-~.-~÷-=~.-.-:-~----~"-~_--~2=_-~-.~.'-~~
.
.
.
--_-
,,
,, ~ ~
~~~--~,--:
.
~:..~-~.
6500
.
~-
• .
-~
-~-~-~--~:
~
SP 6 0 5 0
:
-~.~.-~_-_~.~.._~
~
~
7000
~
7500
7975
Fig. 3. Migrated seismic section from the TIMOR profile in the vicinity of the Timor trough. Post-Permian shelf sediments that can be tied to the Troubadour well (top right) are clearly seen beneath the sea floor on the inner trough wall, more recent undeformed sediments lie in the bottom of the trough. One arcuate reflector is indicated (arrow at SP 6500) in the uppermost basement beneath the shelf sediments; it may represent an unrecognized DevonianCarboniferous basin, or side-swipe from a fault plane parallel to the seismic profile and bounding such a basin. Moho lies at about 15 s beneath the trough. Numerous lower crustal reflectors occur between 8-12 s on the right half of the figure; these appear offset with a thrust sense (half arrows) along south-dipping thrust shear zones near SP 7250.
0
N
S
s
~
"~ 10 E
~
--
~
~ ~..-~ax.~..~.-~-..~3.,¢.,~~
sP 755
~ ~
_
. ~
~ ~
~
:
~ ' ~ ' ~ ; ~ ~ ' = ~ - ~ ~ - " ~ - ~ " ~ - - ~.~,,...::-.~_~--~:---z~g.r~,.~..-r,L.~,~,.~7.~;~---.:
500
~-~-.-,-~.-~.~.~-~.~=,~.~-.~.-.~_-~
250
120
Fig. 4. Migrated seismic section from the DAMAR profile in the vicinity of the Timor trough. Post-Permian shelf sediments are clearly seen near the sea floor on the inner trough wall, folded beneath the trough floor. Two northward-dipping reflectors indicated (arrows) in the uppermost basement beneath the shelf sediments are similar to the one described in the previous figure. Moho lies at 13-15 s.
66
D. B. SNYDER ET AL.
or Petrel sub-basin, lying north of the Sahul Platform and beneath the Timor trough may explain these relationships, but cannot readily account for the observed regional along-trench crustal thickness variations. The AGSO North West Shelf Study Group (1994) note that one important consequence of the concentration of CarboniferousPermian extension in the lower crust is that extensional structures appear only as subtle reactivation features within the upper crust and basin fill, often displaced or offset from areas of lower crustal thinning. Parts of the Australian shelf overthrust by the Timor accretionary prism may therefore contain additional NE-trending sag basins like the Malita Graben, or more probably, northern parts of the north-trending rifts associated with the Petrel sub-basin.
Gravity field observations and modelling The Australian continental shelf south of Timor is now covered by a relatively dense grid of marine
gravity measurements. To the north, coverage is less complete but recent surveys onshore have mapped the gravity field on nearly all the outer arc islands between Timor and Tanimbar (Richardson 1993). The gravity field of the Banda Sea region is dominated by a steep gradient which forms a continuous arcuate feature running from northern Timor past Tanimbar and the Kai Islands to southern Seram (Bowin et al. 1980). Typically, values of Bouguer gravity increase from the slightly negative, as in the south coast region of Timor, to greater than +150 mGal as at points on the north coast of Timor (Fig. 5). Attempts to model this gradient used profiles crossing the arc at various locations. Of these, the model of McBride & Karig (1987) incorporates measurements from an anomalous region of western Timor where contour trends are locally transverse to, rather than parallel to, the strike of the arc, and the model of Milsom & AudleyCharles (1985) is unsatisfactory because of the lack of seismic control. The model of Jongsma et al. (1989; see also Richardson 1993) (Fig. 6a) was
Fig. 5. Bouguer gravity map of the study area. Reduction density is 2.2 Mg m-3. Contour interval is 10 mGal. Solid lines indicate the (roughly N-S) deep reflection profiles and (roughly E-W) modelled transects. Note the particularly negative values associated with the accretionary complex between the seismic profiles south of the island of Moa and the north-south gravity 'notch' over the shelf immediately to the south. Stippled borders are located near inflection points in east-west transects of the gravity field and define the possible areal extent of an anomalous low-density body in the crust.
LOCAL INDENTOR TECTONICS IN THE BANDA ARC
~o0~,,_ I¢~. L ~
II~I: ouguerAnomaly
2°°1
~
Wetar
A'J q
C
I
TIMOR
Australianshelf
0 • ee • 100 ~
Moho
• %1 •
"f": .. 200
• :
y"
../
o
~o
•
300
~ ~ ~
/ x
,oo~,,
oontinentacrust oceanic oust
aomet=onarycomplex volcanic arc
(modified lrom Woodside et al 1989)
~
•
mantle earthquakes
(a)
.,oL > C
I ¢''~
~°I-~ =< I~ ~ ' ~ r °-'t-
~
~ ~
.
.
.
.
.
_.- --:~-I
/
....
-I
C' ~"=" /
~ ~ - ~ /
calcula~
WEST
o .
,
~ observed,ons h e l f ~
observed~,onprisrn 1
" - - /
-I
"
EAST .
.
.
.
.
.
.
.
.
.
.
~2o 41-
.
.
~
t~
3.2
60~- MANTLE
(b)
;
,o
8'o
,"
,6'°
20'0 210 2'°
3'°
D,s,ANcE. K,
Fig. 6. (a) Cross-strike (N-S) gravity profile and crosssection (Woodsideet al. 1989). Location coincides with the western deep seismic profile shown on Fig. 5. (b) East-west profiles across the 'Moa anomaly' and along the shelf margin. The simplistic gravity model shows that density contrasts at two crustal levels, one associated with a sedimentary basin and one with thickened lower crust, can produce the observed gravity anomalies. Profiles are located on Fig. 5, on which stippling indicates the possible extent of the anomalous crustal structure modelled here. Vertical bars on the observed curves indicate where the gravity profiles intersect.
67
based on gravity, magnetic and seismic data along a profile which nearly coincides with the TIMOR line (Fig. 2). The model shows thick continental crust in the downgoing plate almost as far north as the north coast of Timor, with low Bouguer anomalies in the south produced by a thick, lowdensity accretionary complex and the high values in the north due to the presence of what was originally sub-oceanic crust and mantle beneath the forearc basin and the volcanic arc. Subduction, rather than limited overthrusting, is hypothesized at the Wetar thrust and a mantle block is therefore shown as virtually isolated by the action of the opposing subduction zones (Fig. 6a). Of the more localized features of the gravity field, one of the most remarkable is the negative free-air and Bouguer anomaly feature centred directly south of the outer arc island of Moa (Figs 2 & 5). This feature was first indicated by measurements on a few widely spaced lines of marine survey, but was later substantiated by free-air gravity anomalies derived from satellite altimetry and measurements on Moa and the adjacent islands. The lowest Bouguer values occur 40 km to the south of the islands and the free-air minimum lies further south in deeper water. Bouguer anomalies, used in preference to free-air anomalies in Fig. 5, are less influenced by bathymetric variations and therefore better at resolving deep structures. On this map the Moa feature appears nearly elliptical, 40 km across from north to south (measured between half amplitude points) and 120 km from east to west, with an amplitude of c. 70 mGal. It is superimposed on a more elongated, lower amplitude feature that just reaches into the southern part of eastern Timor. Directly south of this 'Moa anomaly', the Bouguer anomalies are 30-50 mGal less than on nearby parts of the Australian shelf margin, and NNW of the feature Bouguer anomalies are 30-50 mGal less than in other parts of the forearc basin (Fig. 5). Collectively, these form a NNW trending band of lower gravity. The two deep seismic profiles cross the long axis of the 'Moa anomaly' on either side of its centre, at about the location of the two half-amplitude points. The profiles also lie near the base of local east-west gravity gradients on the Australian shelf (Fig. 5). Other than the discordantly dipping reflections indicated by arrows in Figs 3 & 4, neither seismic section appears to show basins or shelf features significantly different from those imaged on numerous unpublished seismic profile sections further east, where the regional gravity minimum has lower amplitude. These discordant reflectors have geometries consistent with out-of-plane reflections from dipping reflector surfaces (faults?) striking nearly parallel to the seismic profiles and offset by 10-15 km.
68
D . B . SNYDER E T A L
Two gravity profiles orientated along structural strike (Figs 5 & 6b) illustrate the similarity in the along-strike gravity trends along the prism and shelf by eliminating the more dominant northsouth gravity gradients associated with the subduction zone. One simple model indicates that a combination of anomalous, low densities at the base of the crust and at the top of basement can successfully match observed Bouguer values (Fig. 6b). Three sources of low density material are considered here. The first possibility, of a deep post-orogenic slope basin of the type known to occur south of Timor (e.g. Karig et al. 1987), is not supported by the seismic stratigraphy inferred from the seismic profiles (Fig. 3). Similarly, no direct evidence exists for a significantly thicker accretionary complex east of Timor, nor for anomalously lower density material within it, although Woodside et al. (1989) showed it as a 50 km thick tapering wedge. Thickening of the basement of the downgoing crustal layer by thrusts does seem plausible, based on duplex thrusts suggested by lower crustal reflector geometries (Figs 3 & 4) and on the variations in crustal thickness reported from the NW shelf of Australia (Bowin et al. 1980; O'Brien et al. 1993). Because thickening would effectively add low density material at Moho depth, it would have to be very local to produce a gravity field anomaly as localized as that near Moa. If thickening of the downgoing crustal layer is a valid explanation, syn-orogenic thrust duplexing seems more probable than isostatically uncompensated variations in crustal thickness that existed prior to collision. Elsewhere on the Australian shelf such variations are compensated. Thickening due to thrust duplexing could be accomplished as an independent tectonic mechanism or by restoration of crust previously thinned during the development of Palaeozoic sag or rift basins to its normal thickness whilst preserving or thickening the overlying basin strata. The resulting bi-level thickening is illustrated by the density model shown in Fig. 6b. Inherited crustal structure localizing and intensifying the thickening effect of syn-orogenic thrusting seems the most probable mechanism to produce the anomalously low Bouguer anomalies associated with the Australian shelf margin in this region.
Forearc crustal structures The forearc within the study area is unusual in the narrowness of the forearc basin and in the presence of the island of Kisar. Between the two deep seismic profiles the forearc basin is typically only 15-20 km wide and contains only tens of metres of sediment (Fig. 7). The basin appears to not have
acted as a depocentre, suggesting that it has been structurally higher than other parts of the forearc, such as the Savu Sea, enabling along-arc ocean currents to keep the sea floor relatively free of sediments during the past few million years. Recent mapping on the island of Kisar has revealed, from south to north, low-grade metamorphic rocks such as greenstones, amphibolites, and bands of thick mylonitized quartzites interlayered with greenstones and high-grade pelites (Dropkin et al. 1993; Richardson 1993). Some of these rocks bear a strong similarity to the Aileu (Grady & Berry 1977) or Lolotoi/Mutis Complex Formations on Timor and Moa (interpretations of Richardson 1993 and Dropkin et al. 1993 respectively), and are interpreted as dismembered ophiolites thrust onto thinned Australian shelf or oceanic crust or the volcanic arc. Kisar thus is related to the accretionary complex rather than to the volcanic arc, which emphasizes the unusual narrowness of the forearc basin in this segment of the Banda arc. Southward dipping reflections observed on the deep reflection sections (Fig. 7) may represent backthrusts that have translated the forearc (accretionary prism) northward over the forearc basin and thus narrowed it. The entire forearc block from Kisar to Sermata may represent a large rectangular klippe (Fig. 2).
Backarc features: a seamount chain including Gunung Api Anomalous features also occur north of this segment of the volcanic arc. Although the magnetic patterns are very poorly constrained, Lapouille et al. ( 1 9 8 5 ) suggested transform offsets of c. 50 km of anomalies identified as M7-M10 along a NW-SE orientated line passing through Gunung Api. More recent work (e.g. Rehault, pers. comm. 1995) suggests that the anomalies have been incorrectly identified, but the evidence for offset remains valid. The new seismic reflection profiling has revealed a number of previously undocumented sea-floor volcanoes (seamounts) along the TIMOR line and coincident with this inferred transform fault. The northern part of one of these seamounts is shown in Fig. 8. The API cross-line (Fig. 2) showed that between the seamounts lies a low ridge of extruded basalt separating two levels of the Banda Sea ocean floor (Fig. 9). Sediment thicknesses appear to be similar on both sides of the ridge, which therefore coincides with a down-tothe-east offset of 350-400 m in the sea-floor basement. Arcuate reflectors within the ridge suggest either folds and thrusts or primary extrusive structures among the basalt flows. Gunung Api, the seamounts and the inferred ridge basalt flows
LOCAL INDENTOR TECTONICS IN THE BANDA ARC
69
-0 u0
o5 E b-
10 SP 2135
2000
1800
1600
Fig. 7. Part of the DAMAR seismic reflection section within the forearc basin. This section was migrated but has not been depth converted. Approximately 200 m of relatively undeformed sediment lie within the 10 km wide forearc basin at the centre of the section. Greater sediment accumulations can be observed (arrows) in slope basins both north and south of the forearc basin. Curved arrows indicate possible backthrust.
N M12
-"
2
-~
~ G
. . . . GAPI
5 K
~.sJ
~!. . . . . . . . . . . . . . . . . . .
.
.
S ~--
API LINE
.
.
.
m
1.48
.
.
.
!
--I "
I
~
~
,
~.~
.z
..
,
"' ~ ~
-
~ ~
, -
.
~~
.
~'
.
~% " -""~ ....
~ .........
,,
~.,~:;,.~~.~.
/ /
~
.
.~.~'~-
~-~-~ .~e~- 3.6
I-I 0
~..-:-~. r~ ~,~,~ SP 500
'//YtiI~'/Z" ~,,,,'IZ~/., ".~, %,"..¢:'~3"~,,\~,~\3~&~,',,,*\'~;~,~..~,'~ 1
~
,.:~,~:~,,,, ~k-~"-'.: 1200
Fig. 8. Part of the TIMOR seismic reflection section just south of Gunung Api. This section was migrated and then depth converted using the velocities shown at the right in km s-1. In this depth section the Moho appears as a bright reflection (arrow) at a depth of 10 km where this line crosses the API line. If refraction velocity function M12, of Bowin et al. (1980), shown for comparison at the left, was used instead for the depth conversions, the bright reflection occurs at 14-15 km depths, and no obvious feature is associated with the Moho. Numerous convex-upward reflections (M) are migration noise due to the incomplete removal of the sea-floor multiple before migration and should be ignored.
70
D . B . SNYDER E T AL. TIMOR
LINE
[E
,
w
.,.
!
• ~..~.,:;
5~ , 5KM ~ I ~ ! ~ . . . .:~:~"~ ' ! ~ ~~ N ~
>.
:~-'~.=_-:-.:,~--=-7-.'~'~:Y~,-.?]-=':":'-~:.:.'":L':.'I)Z~.:"T::.
'
......7-:~...:::"17
Z
ZZ'_,;.::'~L~";:-~" - ~:-"-~-:,~ - ~"~:->.--.:,:'-':.--,~.;Z :- :- -, " , _ - -:: .7: "=r_.'--..::.-'.." ~ ..... . _ ~ . , . . , ; . . . 7........ ".7-,~'-"-"7.'%-:.:.." , - ' , . ~ ' . . ; . . . - ",~',,.-- .'...~.,,.' ,~ . . . . ., " ~ . , ,i'-
7='-.'7-'..,: .'~:,,.-'::/7:". ,-" - Z "--' .:,,7"--':" ~. " ..7. . . . . "."
"..:: 5
....
.......
,7.;. ::7":d.~'::-:.-::-'-2-~---".~,~-"÷i'.i-':!~-"-~==:?~.;'~-~Q~~ ~
": -:"-::::~"£ ~ ~-~_~7"'_':~-=.--'S7 ~ ' 7~~: :c'~L:--:L~-~:Cg=~
: . . . . . . . a . - .~:T,~;:..:._~:...,~_.~..~.:~:.~:.:::,~_:~7.,.~_~,_7. ~?.~:" *.. ...--_:. - . - ~ : , * . . . - - ' r . . , : ' , . . ' - : : . ; --...=~.-,,:.-'.-,~.,.-...'2".,-a,: ~. .L'-
"-- -::.'----..-' "~:,,:; . . . .
.'
',.7::.;~,~,, . q ' ~ . ~ " - . - - : ' 2 . ~ - : " .".':,~:"_-'-Z
~iee~'-L'.~,~..~Tr.~£"~'=.:~:~"~.:'~
7~--=--""- ' :!:-~'"5::";"".7.."-:~::-.~',,.~-.~::':.--=-::?:.:: -:~=::..:.:i::,c-;~-".:,.:'£7:~%:). ;:",~'~'.,'.~:~:;-:L:-~::~;~-.":.',.-.7~'Z-~i~-t~:'-.::~ •, ,', "--Y'.:"~-":.." : 2 . i ::~:5-.~:S.'%~!!.<======================: :'::..-:.:-::q::-'-7".':.. : .....:;>~d,a.?7_'. " . J " /7:5"- ":--~~:>::% a7":.7"~-.:,-:~-:::i / L I = "i ........ '......... " " ' " "......") ---"-' T................................... "' ~"' ........{"'/~" . . . . . . . . . . . . . . . . . . . . . . ) - °" SP 3 5 0
I
250
1 S0
50
Fig. 9. Part of the API seismic reflection section just south of Gunung Api where it crosses the TIMOR line. This
section was migrated using the same velocity function as used for the TIMOR line shown in Fig. 8. The reflection observed at c. 9.3 s (10-15 km depth) on the TIMOR line (see Fig. 8 caption) is not as bright here. A sea-floor ridge separates oceanic crust, overlain by < 200 m of sediments, at two distinct depths. Reflectors beneath the ridge include arcuate and short horizontal segments on the flanks that probably represent submarine basalt flows, possibly later folded (inset).
indicate that the fracture has acted as a m a g m a conduit. Together these features imply a leaky transform fault zone nearly coincident with the T I M O R seismic profile. A bright horizontal reflector is o b s e r v e d on the T I M O R line where it intersects the API line (9.3 s travel time on Fig. 9; 10.5 k m depth on Fig. 8), although only a w e a k w e s t w a r d dipping reflector is observed on the API line (Fig. 9). Nearby refraction studies indicated that the M o h o occurs at 10.3 k m depth (M 12 profile of B o w i n e t al. 1980). If the refraction velocities from the M 1 2 profile are used to migrate and depth convert the reflection section, the bright reflector occurs within the mantle at 1 4 - 1 5 k m d e p t h s and no o b v i o u s r e f l e c t i o n s coincide with the M o h o at 10.3 k m depth. Either the refraction velocities derived along the nearby profile M 1 2 are inappropriate here or else the mantle is m o r e reflective than the oceanic crust along this profile. T h e bright reflections m a y m a r k the location o f m a g m a rather than a change in rock lithologies or mineralogy. M o r e probably the bright reflector represents underplated m a g m a w h i c h produces a thicker crust i m m e d i a t e l y beneath the sea-floor ridge. A regional fracture zone, or leaky transform
fault, orientated nearly normal to the regional plate m a r g i n w o u l d s u g g e s t that the d e f o r m a t i o n associated with plate c o n v e r g e n c e m a y have here propagated further into the overlying plate than elsewhere along the plate boundary. The location of this fracture m a y be due to an older zone of weakness in the B a n d a Sea lithosphere or due to unusually thick or buoyant structures on the subducted plate (e.g. Cross & Pilger 1982).
Discussion:
appropriateness
an indentor
model
of
T h e part o f the B a n d a arc i m m e d i a t e l y east of T i m o r is characterized by several features not generally observed in other parts of the arc. A n active volcano within the backarc basin 400 k m above the W a d a t i - B e n i o f f zone is rare, and here occurs along a line o f seamounts orientated nearly normal to the c o n v e r g e n t margin. Uplift rates are high on both the volcanic arc and forearc accretionary c o m p l e x islands, bringing parts o f eastern T i m o r nearly 3 k m above sea-level. T h e forearc basin is only 10 k m wide where it was crossed by the D A M A R seismic line. A n o m a l o u s l y
LOCAL INDENTOR TECTONICS IN THE BANDA ARC low Bouguer gravity anomalies are observed in the corridor defined by the two deep seismic profiles, anomalies more negative than anywhere else along the Banda arc. Taken together, these features suggest that unusual lithospheric strains are associated with the convergence that occurs within this part of the arc. Explanations proposed involve three major lithospheric plates (e.g. Charlton 1986), but the authors prefer to discuss the local features in terms of a single two-plate boundary. Unusually thick parts of the Australian shelf entering the subduction zone are one possible explanation. Crust not greatly thinned during Jurassic rifting, crust thickened as it entered the modern convergence zone, or unusually thick sedimentary basin fill could all contribute relatively buoyant, low density continental crustal material to this segment of the arc. The greater amount of low density material satisfies the observed gravity anomalies. Its associated buoyancy, compared to oceanic crust, would enhance uplift. A thickened wedge of subducting material resists further subduction by deforming internally to form a core of continental material. This core may partly underthrust and dome the arc, and partly impinge upon and strain the oceanic crust of the backarc (Fig. 10). This form of localized strain is here proposed to have caused the fracture associated with the line of seamounts including Gunung Api. Unsubducted parts of the Australian shelf provide some evidence that thick sedimentary basins
GUNUNG API ~
71
may have underthrust the arc in the area of the deep seismic profiles. About 500 km to the south, more than 6000 m of Palaeozoic and younger sediments occur in the Petrel sub-basin (Schltiter & Fritsch 1985; AGSO North West Shelf Study Group 1994) between the offshore parts of two Archaean blocks (Fig. 1). The northern continuation of the Petrel sub-basin is poorly defined, so that it or a similar Carboniferous-Jurassic basin may continue beneath the Sahul platform or align with the band of lower Bouguer anomalies, including the 'Moa anomaly', that trend nearly orthogonal to the convergence margin (Fig. 5). Crustal thinning associated with the crustal extension that formed the Petrel subbasin and other similar basins to the SW has estimated stretching factors of 100-400% (O'Brien et al. 1993; Etheridge & O'Brien 1994). These estimates are consistent with the lower crustal thicknesses of 31 km modelled at the western ends of the two refraction lines, within the Moa gravity feature, when compared with the 35-39 km thicknesses (Fig. 1) reported at the eastern ends (Bowin et al. 1980). Thinner crust beneath the relatively undeformed shelf in this region is not obviously consistent with the proposed model of anomalously thickened continental crust doming and indenting the arc and backarc, respectively. Thrust duplexing within the lower crust could partly restore the pre-rift crystalline crust's thickness while retaining the full overlying basin's thickness, and result in local mass
DAMAR
BAN
....
Fig. 10. Block diagram showing the proposed local promontory of thickened Australian shelf crust and sediments. This promontory becomes a potential indentor wedge during subsequent convergence by lifting and offsetting the overlying crust across faults inferred to segment the accretionary complex that includes Timor and to propagate northward into the backarc Banda Sea. The model contains elements intended to explain the various observed anomalous features in the bathymetry, gravity field and crustal thickness in the vicinity of the new deep seismic reflection profiles. The cross pattern indicates crystalline basement within the Australian crust.
72
D.B. SNYDER ET AL.
deficits such as that m o d e l l e d for the ' M o a anomaly' (Figs 6b & 10). Shear thrusts within the relatively weak lower crust need not directly underlie the rift basins. O v e r l y i n g these deeper, N N E trending Devonian-Carboniferous basin structures are the NE trending platforms and graben structures, such as the Sahul Platform and Malita Graben, that are associated with the Carboniferous-Permian rifting of Gondwanaland. A small outboard platform, similar to the A s h m o r e and Sahul Platforms (Fig. 1), and effectively forming a promontory on the pre-collisional Australian margin, may have already been subducted and contributed to the low mass a n o m a l y of thicker continental material beneath the Moa gravity anomaly. Insufficent details of crustal thicknesses and structures are available to resolve more clearly the various superposed pre-collision extension-related structures from the recent collisional ones. The coincidence of a north-south orientated boundary between Archaean crustal blocks and the midPalaeozoic Petrel (Bonaparte) Basin with a N E - S W orientated late-Palaeozoic rifted m a r g i n and Neogene convergence zone permits many possibilities for crustal structures. Some of the older structures may nucleate subsequent strain. The presently preferred explanation for high uplift rates on Timor and other anomalous features within the Banda arc immediately east of Timor is
partial restoration of crustal thicknesses beneath a northward continuation of the Petrel sub-basin of the Bonaparte rift by thrust duplexing within the lower crust during recent collision. Dynamically depressed crust is inferred from the gravity field east of Timor. In eastern Timor the thickened crust has rebounded isostatically to produce rapid uplift and neutral gravity anomalies. In both areas, overcompensated crust within the subducted slab has acted as a buoyant mass that jams subduction and translates convergence to the north where it loads and fractures the oceanic lithosphere of the Banda Sea. We would like to especially thank K. Bartram, A. Macfarlane, B. Situmorang, S. Suparka and M. Takim, for their interest and support in organizing the seismic project in Indonesia. The BANDA seismic reflection profiles were acquired by GECO Singapore Pte. and processed by Prakla-Seismos, and are available at the cost of reproduction from the British Geological Survey (Marine Geophysics Programme Manager), West Mains Road, Edinburgh EH9 3LA, UK. This research was done under the auspices of the Indonesian Institute of Sciences (LIPI) and the Natural Environment Research Council (NERC) of the UK. BANDA was funded by the British Natural Environment Research Council and BIRPS' Industrial Associates Program (Amerada Hess Ltd, Amoco, BP Exploration Operating Company Ltd, Chevron (UK) Ltd, Conoco (UK) Ltd, Mobil North Sea Ltd and Shell (UK) Ltd). This paper is Cambridge Earth Sciences contribution number 4176.
References AGSO NORTH WEST SHELF STUDY GROUP. 1994. Deep reflections on the North West shelf: changing perceptions of basin formation. Proceedings Western Australian Basins Symposium, Perth, 63-76. BOWlN, C., PURDY, G. M., JOHNSON, C., SHOR, G., LAWYER, L., HARTONO,H. M. S. & JEZEK, P. 1980. Arc-continent collision in the Banda Sea region. AAPG Bulletin, 64, 868-915. BREEN, N. A., StaYER, E. A. & ROOF, S. 1989. The Wetar backthrust belt, eastern Indonesia: the effects of accretion against an irregularly shaped arc. Tectonics, 8, 85-98. CARDWELL,R. K. & ISACKS,B. L. 1978. Geometry of the subducted lithosphere beneath the Banda Sea in eastern Indonesia from seismicity and fault plane solutions. Journal of Geophysical Research, 87, 2825-2838. CHARLTON, T. R. 1986. A plate tectonic model of the eastern Indonesian collision zone. Nature, 319, 394-396. , DE SMET, M. E. M., SAMODRA,H. 8z KAYE, S. J. 1991. The stratigraphic and structural evolution of the Tanimbar Islands, Eastern Indonesia. Journal of SE Asian Earth Sciences, 6, 343-358. CROSS, T. A. & PILGER, R. H. 1982. Controls of subduction geometry, location of magmatic arcs, and
tectonics of arc and back-arc regions. Geological Society of America Bulletin, 93, 545-562. DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S. 1990. Current plate motions. Geophysical Journal International, 101, 425--478. DROPKIN, M. J., HARRIS, R. A. & ZEITLER, P. K. 1993. An Oligocene forearc crustal flake exposed in a contemporary arc-continent collision: Timor, Indonesia. Geological Society of America Annual Meeting, Abstracts with Programs, 25, 6, A--482. ETHERIDGE, M. • O'BRIEN, G. W. 1994. Structural and tectonic evolution of the Western Australian margin basin system. Australian Petroleum Exploration Association Journal, 34, 906-908. GENRICH, J., BOCK, Y., MCCAFFREY, R., CALAIS, E., STEVENS, C. E T A L . 1994. Kinematics of the eastern Indonesian island arc estimated by global positioning system measurements. EOS abstracts, 75, 162. GRADY, A. E. & BERRY, R. E 1977. Some PaleozoicMesozoic stratigraphic-structural relationships in east Timor and their significance in the tectonics of Timor. Journal of the Geological Society Australia, 24, 203-214. HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078.
LOCAL INDENTOR TECTONICS IN THE BANDA ARC 1988. Plate tectonics and island arcs. Geological Society of America Bulletin, 100, 1503-1527. HARRIS, R. A. 1991. Temporal distribution of strain in the active Banda orogen: a reconciliation of rival hypotheses. Journal of SE Asian Earth Sciences, 6, 373-386. HILTON, D. R., HOOGEWERFF,J. A., VAN BERGEN, M. J. & HAMMERSCHMIDT, K. 1992. Mapping magma sources in the west Sunda-Banda arcs, Indonesia: Constraints from helium isotopes. Geochimica et Cosmochimica Acta, 56, 851-859. JACOBSON,R. S., SHOR, G., KIECKHEFER,R. M. & PURDY, G. M. 1979. Seismic refraction and reflection studies in the Timor-Aru Trough system and Australian continental shelf. In: WAT~NS, J. S., MONTADERT, L. & DICKERSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 209-222. JOHNSTON, C. R. & BOWlN, C. O. 1981. Crustal reactions resulting from the mid-Pliocene to recent continentisland arc collision in the Timor region. Bureau of Mineral Resources Journal of Australian Geology & Geophysics, 6, 223-243. JONGSMA, D., WOODSIDE, J. M., HUSON, W., SUPARKA,S. & KADARISMAN,D. 1989. Geophysics and tentative late Cenozoic seismic stratigraphy of the Banda Arc-Australian continent collision zone along three transects. Netherlands Journal of Sea Research, 24, 205-229. KARIG, D. E., BARBER, A. J., CHARLTON, Z. R., KLEMPERER, S. & HUSSONG, D. M. 1987. Nature and distribution of deformation across the Banda Arc-Australian collision zone. Geological Society of America Bulletin, 98, 18-32. LAPOU1LLE,A., HARYONO,H., LARUE, M., PRAMUMIJOYO, S. & LARDY, M. 1985. Age and origin of the seafloor of the Banda Sea (eastern Indonesia). Oceanologica Acta, 8, 379-389. MATrHEWS, D. H. 1986. Seismic reflections from the lower crust around Britain. In: DAWSON, J. B., CARSWELL, D. A., HALL, J. & WEDEPOHL,K. H. (eds) The Nature of the Lower Continental Crust. Geological Society, London, Special Publications, 24, 11-21. MCBRIDE, J. H. & KARIG, D. E. 1987. Crustal structure of the outer Banda arc: new free-air gravity evidence. Tectonophysics, 140, 265-273. MCCAFFREY, R. 1988. Active tectonics of the eastern Sunda and Banda Arcs. Journal of Geophysical Research, 93, 15163-15182. 1992. Oblique plate convergence, slip vectors, and forearc deformation. Journal of Geophysical Research, 97, 8905-8915. 1996. Slip partitioning at convergent plate boundaries of SE Asia. This volume.
--,
73
MOLNAR, P., ROECKER, S. W. & JOYOD1WIRYO, Y. S. 1985. Microearthquake seismicity and fault plane solutions related to arc-continent collision in the eastern Sunda Arc, Indonesia. Journal of Geophysical Research, 90, 4511--4528. MILSOM, J. & AUDLEY-CHARLES, M. G. 1985. Postisostatic readjustment in the southern Banda Arc. In: COWARD, M. P. & RIES, A. C. (eds) Collision tectonics. Geological Society, London, Special Publications, 19, 353-364. O'BRIEN, G. W., ETHERIDGE, M. A., WILCOX, J. B., MORSE, M., SYMMONDS, P., NORMAN, C. & NEEDHAM, D. J. 1993. The structural architecture of the Timor Sea, northwestern Australia: implications for basin development and hydrocarbon exploration. Australian Petroleum Exploration Association Journal, 33, 258-278. PIGRAM, C. J., DAVIES, P. J., FEARY, D. A. & SYMONDS, P. A. 1989. Tectonic controls on carbonate platform evolution in southern Papua New Guinea: passive margin to foreland basin. Geology, 17, 199-202. RICHARDSON, A. N. 1993. Lithospheric structure and dynamics of the Banda Arc, eastern Indonesia. PhD thesis, University of London. SCHLt2rrER, H. U. & FRITSCH, J. 1985. Geology and tectonics of the Banda Arc between Tanimbar Island and Aru Island (Indonesia) - results of R/V Sonne Cruise SO-16. Geologisches Jahrbuch Reihe E, 30, 3-41. SENGOR, A. M. C. 1987. Tectonics of the Tethysides: orogenic collage development in a collisional setting. Annual Reviews of Earth & Planetary Science, 15, 213-244. SILVER, E. A., REED, D., MCCAFFREY, R. & JOYODIRWmYO, Y. 1983. Back arc thrusting in the eastern Sunda arc, Indonesia: a consequence of arc-continent collision. Journal of Geophysical Research, 88, 7429-7448. VEEVERS, J. J. 1988. Morphotectonics of Australia's northwestern margin - a review. In: PURCELL,P. G. & PURCELL, R. R. (eds) The North West Shelf, Australia. Proceedings of Petroleum Exploration Society Australia Symposium, Perth, 1988, 19-27. WOODSIDE, J. M., JONGSMA,D., THOMMERET,M., STRANG VAN HEES, G. & PtYNTODEWO 1989. Gravity and magnetic field measurements in the eastern Banda Sea. Netherlands Journal of Sea Research, 24, 185-203. YEATES, A. N., BRADSHAW,M. T., DICKINS,J. M., BRAKEL, A. T., EXON, N. E Er AL. 1987. The Westralian Superbasin: an Australian link with Tethys. In: MCKENZIE, K. G. (ed.) Shallow Tethys 2. Proceedings of Intenational Symposium on Shallow Tethys 2, Wagga Wagga, 199-213.
Detailed processing of seismic reflection data from the frontal part of the Timor trough accretionary wedge, eastern Indonesia B. D. H U G H E S 1'3, K. B A X T E R TM, R. A. C L A R K 1 & D. B. S N Y D E R 2
1 Department of Earth Sciences, The University of Leeds, Leeds LS2 9JT, UK 2 BIRPS, Bullard Laboratories, Madingley Rise, Cambridge, CB30EZ, UK 3 Present address: ABG EXPLORATION, 12 Howard Way, Cromwell Business Park, Newport Pagnell, Bucks MK16 9QR, UK 4 Present address: CSIRO (Division of Exploration & Mining), Floreat Park Laboratories, Wembley, Perth, WA, Australia 6014 Abstract: The DAMAR deep seismic line across the Banda arc in Indonesia provides a valuable insight into the Late Cenozoic collision between the northward-moving Australian plate and the Southeast Asian plate. Previous geophysical investigations across the accretionary wedge have only imaged relatively shallow structures due to the complex structure of the wedge which remains poorly resolved following conventional processing techniques. The design of a detailed processing sequence and its application to a 35 km section of the DAMAR line to include the Australian Northwest shelf, the Timor trough and the frontal part of the accretionary wedge, has allowed an improvement in resolution of structures both within and below the frontal part of the wedge. The Australian margin to the south of the trough contains a number of normal faults which show inversion features. The reprocessing of the accretionary wedge has allowed an improved interpretation of the internal structure in the frontal part of the collision zone and this consists of a number of high angle thrust structures. This has shown that shortening in the frontal part of the accretionary wedge has developed coherent thrust slices imbricated from the subducting Australian margin, rather than an incoherent melange.
The Timor region in eastern Indonesia is often quoted as a present-day fold-and-thrust belt in its earliest stages, and is one of the youngest collisional belts presently active. The study of the Timor region not only provides an understanding of the active collision between the Australian plate and Southeast Asia, but allows a view of the early stages of mountain belt evolution. The development of the region may be divided into two main phases:
of these data to assess the deformation styles, both within the accretionary wedge and on the Australian shelf immediately adjacent to the trough. Accepted models that assume that deformation has recently ceased at the Timor trough are also evaluated in the light of this work. Similar processing techniques to those described may be applied to similar datasets from other areas to increase the resolution of structures both within and below areas of complex geological structure.
• break-up of Gondwanaland and the development of the Australian passive margin; • development of subduction north of Australia with eventual collision of the Australian margin and development of the collision zone presently exposed on Timor. In 1992 the British Institutions Reflection Profiling Syndicate (BIRPS) and the Indonesian Marine Geological Institute commissioned two deep seismic lines across the Timor-Tanimbar trough to the east of Timor, the BANDA survey (Fig. 1). This paper discusses the development of a processing sequence to improve the resolution obtained by conventional processing on part of the DAMAR line (shotpoints 600-950), and the interpretation
Evolution of the Timor trough region One of the most extensive features in Indonesia is the linear, northerly dipping subduction zone delineated by the Java Trench-Timor trough which marks the collision of the northerly moving IndoAustralian plate with the Southeast Asian plate. During the break-up of Gondwanaland the Australian margin progressively rifted southwards with continental break-up on the Northwest Shelf by the Oxfordian (160 Ma) (Pigram & Panggabean 1984). Following rifting, the drift phase of the margin during the Cretaceous and Tertiary was characterized by the deposition of dominantly carbonate sediments. During the Neogene the
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution o.fSoutheastAsia,
75
76
B.D. HUGHES ET AL.
0 I~
o ~ ~'~
TIMOR
0
I
I
Banda Sea
GunungApi
km 10,01 ' 000/ 4 /L) ° /
DAMAIq~
0
o
Kisar
I
South
o 5 m r n a -I T r o u b a d o u r Well N-°1%
Australian Northwest Shelf 10 ° 125 °
126 °
127 °
East
128 °
129 °
130 °
Fig. 1. Part of the Banda arc in eastern Indonesia showing the location of the BIRPS TIMOR and DAMAR deep seismic lines. The barbed line indicates the trace of the Timor trough, volcanoes are shown as circles (black where active) and values on bathymetric contours are in metres. The section detailed in this paper is shown by the thickened line, located in the immediate vicinity of the trough axis on the DAMAR line, shotpoints 600 to 950.
northerly drift of the Indo-Australian plate produced the development of oceanic subduction to the north of Australia and ultimately the collision of the leading edge of the Australian plate with the Banda arc subduction complex. This resulted in the development of a fold-and-thrust belt exposed on a number of islands, the largest of which is Timor. Subduction is presently active along the Java Trench with decreased activity on the Timor trough. Recent geodetic measurements by McCaffrey (1996) suggest that the Timor collision zone has recently locked on to the northward-moving Australian plate with cessation of subduction at the Timor trough.
Marine geophysical studies in the Timor trough Numerous previous workers have studied the Australian margin and the Timor trough region using marine geophysical techniques (e.g. Bowin et al. 1980; Karig et al. 1987; Charlton, 1988; Masson et al. 1991). Cross-sections across the
Timor trough have previously been imaged using seismic reflection data, and a number of common features have been observed (Fig. 2): • an outer slope increasing in gradient towards the trough axis, underlain by Australian shelf and slope strata (A & B, Fig. 2). Normal faults are present, mainly dipping northwards; • a trough axis with a flat-topped wedge of trough fill sediments (F, Fig. 2) representing sediments trapped between the outer slope and the deformation front; • a deformation front marking the southerly extent of the Timor fold-and-thrust belt; • an accretionary wedge containing individual thrusts and folds representing deformation of the subducted Australian margin. The complex seismic response of accretionary wedges causes the geologically important structures within and beneath them to remain poorly imaged after conventional 2D processing, as seen in many previously published sections. An ideal processing sequence, always data-specific, should
SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH
77
(a)
TW'T (s)
(b) Inner slope Trough fill < >< ><
Outer slope ,,-
>
Deformation front
4
6 TWT
(s)
--"~
4
6
- -/ '
'
'
Fig. 2. Seismic reflection profile (Gulf line AU-38F) across the Timor trough and outer slope south of Timor showing the major morphological features. Adapted from Karig et al. (1987).
deliver optimum signal-to-noise ratio (SNR) while representing subsurface geometries faithfully and with optimum vertical and lateral resolution. These objectives are sometimes compromised against practical limits such as computer resources, but this study has developed a more detailed processing sequence which (compared with conventional processing) has resulted in a significant improvement in SNR, resolution, and hence geological interpretation.
Data processing Seismic reflection data processing techniques are well described in many texts (e.g. Yilmaz 1987). However, 'deep' reflection surveys (i.e. deep relative to hydrocarbon exploration, for which reflection methods are most developed) encounter particular difficulties: data from the Timor trough are quite typical in this respect. They show strong multiple activity, all reflectors are relatively deep, and numerous very strong diffractions are visible.
Migration is an essential process in such heavily faulted sections; with marked variations in reflector dip across the section, choice of suitable migration velocities was critical, but, as is typical of deep reflection surveys, velocities were difficult to measure accurately because of the very limited amount of moveout on the reflections. The data were provided by the BIRPS group already preprocessed and sorted into CMP gathers. The considerable water depth (2-4 s twt) meant that water-bottom multiples in fact only appeared deeper in the section than the zones of interest here: their removal was straightforward and multiple suppression was carried out as part of standard preprocessing as described by Snyder et. al. (1996). Conventional processing was then completed, to ascertain the overall geometry of the accretionary wedge. Velocity analyses were made at 100 CMP (2.5 km) intervals across the section. Given the simple structures south of the Timor trough, this was felt acceptable for that part of the line, but further north, velocities were difficult to determine
78
B.D. HUGHES ET AL.
reliably. The stacked section so produced was migrated using a finite-difference time migration, to allow lateral variations in velocity to be accommodated. Even so, internal structure of the accretionary wedge and the subducting Australian shelf beneath it were poorly resolved and a more detailed processing strategy was imperative for this part of the line. Shotpoints 600-950, spanning the toe of the wedge, are shown conventionally processed in Fig. 3. Key elements of the improved processing sequence (Fig. 4), which yielded the section in Fig. 5 (which shows the same shotpoint range a~Fig. 3), are now discussed briefly. Detailed processing
South of shotpoint 748, no revisions to velocity analyses were thought necessary. As the section becomes structurally more complex northwards, within the accretionary wedge, selection of velocities becomes more important. Instead of regular spacing, velocity analyses were located at the apices of all prominent diffractions, resulting in identification of more accurate velocities. The average separation between velocity analyses thus reduced to 1.6 km; still larger than would have been preferred, but the effect of the strong diffractions was to make additional velocity analyses elsewhere of little value. To refine velocity estimates further, bias from dip effects (inherent in any moveout-based velocities for dipping reflectors) was reduced by an iterative sequence of velocity analysis, DMO (dip moveout), and repeat velocity analysis. Common-offset FK-based DMO (Hale 1984) was used; because of limitations on computer storage,
600
I
700 I
|
DMO was applied separately to near, intermediate, and far offsets and the full dataset then reconstituted. Revised velocities were some 5-10% different from the original in and below the accretionary wedge but effectively unchanged for the shallow-dipping Australian shelf to the south. Successful migration of the resultant stacked section was critically dependent of the use of accurate, true, medium velocity throughout. Below the seabed, optimum migration velocities were determined empirically as percentages of stacking velocities, but that of the water column is well known and was not allowed to vary. In the complex parts of the line, test panels from faulted areas at either end of the shelf slope and from within the accretionary wedge were used. Optimum percentages (typically 80-90%) were chosen and applied to a finite-difference time migration of the full line (shown in Fig. 5). For structurally simple parts of the line, optimum percentages could be seen to correlate clearly with dip (see Hughes 1994 for a full discussion). A depth conversion was done postmigration, using geologically-constrained interval velocities.
Interpretation of the DAMAR line An interpretation of the DAMAR seismic l i n e involved an iteration between geological observations and the processing sequence. Structures identified were examined and additional processing was applied if required to redefine the velocity field allowing an improvement in resolution and interpretation. An interpretation of the reprocessed section is shown in Fig. 5b.
800 I
0 |,,,
900
I
I
SP
10 km I
TWT (s)
-5
Fig. 3. Example of part of the DAMAR line (shotpoints 600-950) processed using a conventional processing sequence. Note the poor lateral and vertical resolution which is attributed to the low signal-to-noise ratio resulting from conventional processing. Compare this with the same section in Fig. 5, reprocessed to give better definition, which has allowed an improvement in the geological interpretation.
SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH
79
PREPROCESSING Carried out by BIRPSgroup independently. Incorporated multiple suppressionand sortingto CMP gathers
I "
VELOCITY ANALYSIS Incorporates analysis of diffractions ......
re-sort to CMP~athers
[ NMO CORRECTION, CMP STACK
acceptable velocities,I gathers & stack? no / \yes
/
I sort pre-NMO CMP gathers DMO to constant-offset gathers: apply Hale's FK DMO
POST-STACK FILTERING AGC, bandpass filters, etc.
1 MIGRATION VELOCITY TESTING On several separate test panels along line
1 l
TIME MIGRATION Constant velocity in water column; variable percentage of stacking velocities sub-seafloor.
I DEPTH CONVERSION
[
Fig. 4. Outline of the processing sequence used to produce the section in Fig. 5. Emphasis is placed on determination of accurate and appropriate velocities. Key elements only are shown: background information on processes may be found in standard texts such as Yilrnaz (1987), and data-specific details are given in Hughes (1994).
The Australian margin slope The Australian margin slope (shotpoints 600-780) dips gradually north into the Timor trough at an angle of approximately 3.5 ° . Reflectors on the adjacent TIMOR line have been tied to the Troubador No 1 Well (see Hughes 1994), located at its southern end, and have been compared and verified with seismic data from the Timor gap to the west of the BANDA lines (see Baxter 1993). Reflectors on DAMAR have been directly correlated with those on the TIMOR section, thus making it possible to estimate that they relate stratigraphically to Mid-Late Tertiary (A), Mid-Base Tertiary (B), Mid-Late Jurassic break-
up unconformity (C) and Top Permian? (D) ages. The upper sediments represent the ongoing post-rift subsidence of the Australian passive margin and are Late Cretaceous to Miocene pelagic mudstones and shelf carbonates (Karig et al. 1987). These are laterally consistent across the section with no prominent angular erosional surfaces. Normal faulting on the DAMAR line is apparent in two main areas (shotpoints 608-635 and shotpoints 740-778) and these faults appear to bound a small graben structure. Some of these faults dip to the south, rather than north towards the trough as commonly described by previous workers. The faults show growth (indicated by thickening of sediments in the hanging wall) during Cenozoic
oo
~-.-,--,=-.--r---.---~
~
... .
~
.
~
.
.
.
-..-.~
.
l u o m o l l O O O ¢ I ~ - - - ~ - ] - ~
~- ~=-~Z.~~
:~-""
. . . .
--~--'z~ = "
~
~
~
:'~ " - . -
-
-
'
-.-~
_
.
:
." , ~
-"'.--~----~ -'~ _ ~ . ~ _~'...-_-"-~---,C-- ~ . .:-~-:,-_.~-.:.~ : . .- ~ ~ ~_' = - = ~~._ - - - ~',~- ~, .-. ,
='-~.~--'
~:.--7~-~?.~ ' ~ - ~ '
~
~ ~_,.-.~.~ - . . . " "-" , : -~"~'-..-..-~'=.=c.-.'~r,'=.=,' . 7 _Z , z : .. ,:.=..~ " -
~
006
.
.
'=
-
.
.
.
.
~' ~
.....
11:1
"
,_
- -~=
-'
. . . . . . . . . . . .
.>.=:
- ~ - .~_ -~- - . , . ~- ,.
.
~ - ; ~
~
.,-.=..
.
~
. .~. . . . .
- .-:,.~-~.
--~'L~:-L-~:,~-~;.7.~_
,._
-
~ -
00L
|
OOL
.
=.~.=.
-
._
-
. . . . . . . .
.
-
~
" - ~ - ~ - ~ -
-
~
.
......
_ _
"
S
_
"
3r
_
(q)
009
(~)
009
,4-,Z:
.---.-~
.
~_~
-~.~,'Z"-~'--"
-,,~_'.."-- -~.,--'~-~'_-,,..~_'_---,.-.~
~-~==~-~
--
"~
~-----------------~" ---
- . - _ - - =--3,T_ ~ "~.-~__ .... ~.--~:~-
:1
-==~-':~.~_7~ -~=--:
~
0~9
I
0~9
=-,. . -=
odol st Jl°qs aolno uN.mmttm!Ie'°'snv
-
_-_ _.~,--'-~'~--___ " - z - z - - - ~ - ~ - . . - . - - - ~ ~ ~ ~ ~ - ~ . . , . . = , , , , . ~ _
.
I
luo~J uoql~tuaoJo(I
I
I
0~L
O~L
008
|
~,~.
a . . . . .
-_:--~-~_.----_-
. .
~
- ~-~.~.
--
.
-
"-:
~
':"
"-.L~',
"'---'7 ~
~
= ' ~ - ' , ~
"-
0~:8
008
~_..~..~
i
-~-~"--~--~
0~8
- . . . ~ - ~ - -.
I
~
006
•~..~,~- -
. ttl)l~ ~dolsiol.I tu.o~po.,~A6 l, /~UO'I'IOIooVI dS
~
....
~
•OU!l pzqsep ~o~oIql e £q po3lxetu st o~po~ Otll ol luotuzlIOOOp Ieseq oq,L "l[not.D!p sT o*pOA~ Lmuo!loazoe oql uTtIl!ax saolooIJoz jo uo!lelO~O o "sou!i poqsnp Xq po:~a~m ore sllne~I "((I) Lue.[tu~od do& pue (D) Xl.ttu~ojuooun dn-]t-eozq otsseanf ole'-l-P.[IA[ '(~/) ~d~!laO.L os~fl-PtIAI '(V) ~e!lao.L ~le~I-Pt.IA[ ol ole[o~ ut~amtu ue!lezlsn V oql uo szoloolJOH "uoTl~lo.Idaolu! posodm.uodns e (q) pue uo!loos possooo.tdo.t oql (~) ~u.~taoqs (uo!leoo[ .tol [ "~!~I oos) .tom!,L Jo lseo 'q~noal aotuLL oql sso~oe ou![ ~IVIA[V(I oql jo uo!)oos possoooad oqJ., "~ "B!,tI
/
/
•
I ~
~01
~-
/~- ~-, i~W-~.-'-~_~-~.~~~ .LA~,L II
£A~.I.
N
dS
SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH sediment deposition and in places reach the sea floor. Inversion structures are apparent on the southernmost, north-dipping, fault (shotpoint 610) with a compressional anticline developed on reflector B. Testing this structure using different migration velocities has shown that this is not an undermigration effect (Hughes 1994). Some thickening of the sediments just below this reflector on the fault can be seen with steep up-thrown sediments immediately overlying low-angle dipping strata, possibly indicating a small thrust, although this is not well imaged. Some complicated zones are apparent on the section and are thought to correlate to normal faults (e.g. in the vicinity of shotpoint 747). The excessive displacement gradient generated by interpreting single faults suggests that this area may contain additional fault terraces, increasing the appearance of drag on the fault zones. This faulting also creates the topographic bulge in the sea floor at shotpoint 760. At shotpoint 635, a south-dipping fault with no apparent inversion similarly shows thickening of sediments in the hanging wall and steepening of reflectors immediately adjacent to the fault. This is also difficult to interpret as a single structure and probably involves terraces. It should be noted that the presence of inversion across this region and potential strike-slip activity may also give out-ofplane-of-section complications which are not considered here.
The trough axis The axis of the trough, marking the surface transition between the shelf slope and the accretionary wedge, is characterized by a change in the regional topographic dip from north to south. However, a significant thickness of trough sediment, as seen on the adjacent TIMOR line and in sections published by previous workers (e.g. Fig. 2) is absent on the DAMAR line. Karig et al. (1987) also noted a change in the geometry of the trough fill along axis with ranges of 0-10 km in width and 0 to >1 km deep. They proposed that this is due to the wedge thrust front being laterally discontinuous and that in some areas the trough fill may have been accreted into the toe of the wedge. Both the section in Fig. 2 and the adjacent TIMOR line show similar features of overthrusting of the trough sediments by the wedge front. From the reprocessing beneath the frontal part of the accretionary wedge, a pre-existing thickness of sediment in the trough does not appear to have been overridden by the wedge (as seen on the TIMOR line), unless it has been fully accreted to the wedge by its recent advancement. The top of the downgoing Australian plate lies at a depth of
81
3.8 km on the TIMOR line beneath a trough sediment fill of c. 630 m, and at a depth of c. 3.2 km at seabed on the DAMAR line at shotpoint 780. Therefore, the greater seabed depth of the TIMOR line may have produced a preferential site for trough sedimentation in comparison with the lack of sediment fill present on the DAMAR section. It may be possible that the inversion of the normal fault beneath the trough axis on the DAMAR line at shotpoint 760 has also contributed to the reduced bathymetry in the trough, although this is speculative.
The accretionary wedge The accretionary wedge is characterized by a change in the surface topography as the Timor collision zone overrides the Australian margin. The sea floor above the wedge shows a relatively smooth profile, with a lack of penetration of thrusts at the wedge front and only minor penetration further back in the wedge. This suggests that the frontal part of the wedge is inactive, with minor out-of-sequence thicknening further back producing internal thickening. This may be due either to the cessation of subduction at the Timor trough in this region, or that the frontal part of the wedge is inactive as internal thickening attempts to return the wedge to a critical state and allow propagation of the wedge front. The internal structure of the wedge is seismically complex. Picking structural and lithological reflectors is obviously difficult in such a complex zone which is likely to contain a high amount of diffracted energy which may not have been migrated back to a source reflector. The use of migration velocity analysis during this study has, however, led to the conclusion that a number of dipping events within the wedge are real reflectors and not diffraction effects. These reflectors are shallow in dip at the toe of the wedge, increasing in dip southwards and are most likely to be thrust faults developed during forward propagation of the thrust front, developing a back-steepening of the thrusts. These show typical thrust geometries with rollover anticlines in the hanging wall above a lowangled detachment. Correlation of reflectors from the Australian margin is difficult in such a highly deformed zone. A strong seismic reflector is defined by a high acoustic impedance contrast at the contact between two lithological units. The lithological units occupying the intervals between the strong reflectors A-D are considered to be homogeneous throughout as indicated by their relatively transparent internal seismic structure, although some internal layering is apparent. These units are
82
B.D. HUGHES ET AL.
thought to be similar to those observed from the Troubadour 1 well located nearby, and consist of." a Recent-Maastrichtian sequence of calcarenites and marls from seabed to 1.55 km depth; a 550 m thick claystone unit (Turonian-Cenomanian); Jurassic sandstones and claystones (2.1-2.8km depth); and Triassic limestones and claystonesiltstone-sandstone sequence (2.8-3.3 km depth). Therefore, during thrusting, imbrication which structurally juxtaposed two of the main lithologies could be expected to produce a reflector with acoustic impedance and phase similar to a stratigraphic contact. The strong convex-upward nature of many reflectors is thought to correlate to hanging wall geometries within thrust sheets rather than being from faults but it is unclear whether the lithological contacts picked correlate with the original stratigraphic reflectors on the Australian margin or whether these are due to lithological juxtapositions created during faulting.
Imaging of structures beneath the accretionary wedge Previous studies across the Timor trough have not generally imaged the structure of the Australian shelf beneath the accretionary wedge. Improved processing during this study has allowed some structures to be imaged beneath the frontal part of the wedge. The decollement itself is not strongly imaged, although the truncation of stratal reflections within the uniform shelf sediments allows an interpretation of its location. Beneath the wedge in the area of shotpoint 820 there appears to be an upwarping of reflectors C and D. This could be either a processing anomaly or a geological feature. Similar anomalous features are well known during processing and are associated with inaccurate velocity determination of the overlying structure, particularly if this is of a high velocity, resulting in a velocity pull-up. However, the feature at shotpoint 820 appears to be of high amplitude and short wavelength, which is inconsistent with any possible velocity anomaly within the overlying wedge. Therefore, this feature is tentatively interpreted as geological. The upper part appears to be truncated by the wedge decollement with no apparent continuation into the wedge. Thus, this structure must pre-date wedge imbrication of this part of the shelf. The shape of the feature on reflector D is similar to that beneath the zones of structural complexity seen elsewhere on the section (shotpoint 635 and shotpoint 747) and may be associated with similar fault structures, although resolved less clearly due to the overlying wedge.
Conclusions Following the design of a reprocessing sequence, an improvement of SNR and resolution of the DAMAR line has been achieved, with the improvement in data quality allowing an increased confidence in the interpretation. Application of similar, data-specific, processing should significantly improve other datasets from areas of complex geological structure. Recent geodetic measurements reported by McCaffrey (1995) have suggested that subduction at the Timor trough has almost stopped. However, observations by Charlton (1988) that the wedge thrust front in the Timor trough to the south of Timor has not advanced significantly in the last 450 000 years have been interpreted as indicative of episodic propagation of the thrust front during continuing subduction by Masson et al. (1991). From a detailed examination of the frontal part of the accretionary wedge on the DAMAR line, it is difficult to resolve this debate. The lack of recent sediment deposition to the south of the thrust front does not allow any estimation of timing of the most recent advance of the wedge and from Charlton's observations both scenarios are possible. A further indication of possible recent thrust activity within the wedge is the identification of active faults either at the thrust front or within the wedge, affecting the more recent sediments which overlie the wedge. Such faults will produce sediment thickening and sharp topographic variations on the wedge top. The reprocessing of the DAMAR line has allowed an improved interpretation of the internal structure of the accretionary wedge which should allow similar structures to be imaged. However, although some faults appear to outcrop at the seabed with the development of minor breaks in topography (e.g. at shotpoints 785, 810, and 922) and may indicate minor internal thickening, the thrust front and much of the internal part of the wedge appear to be recently inactive with continuous, undeformed sediment drape across the outer slope and the wedge front. The lack of a recent significant sediment thickness in the trough axis is attributed to a lack of deposition, rather than due to recent wedge movement imbricating any pre-existing trough fill. From the improved processing of the DAMAR line it is possible to identify the large-scale internal deformation mechanism within the wedge. This is dominated by high-angle thrusts which have imbricated sediments from the upper part of the subducting Australian passive margin. This indicates that the deformation mechanism within the frontal part of the accretionary wedge consists of the imbrication of coherent blocks rather than by a shearing of the subducting Australian margin sediments into an incoherent melange.
SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH
83
References BAXTER, K. 1993. Quantitative modelling of continent
collision: Application to the Timor region, Eastern Indonesia. PhD thesis, University of Liverpool. BowrN, C., PURDY, G. M., JOHNSTON, C., SHOR, G., LAWVER, L., HARTONO,H. M. S. & JEZEK, P. 1980. Arc-continent collision in the Banda Sea region. AAPG Bulletin, 64, 868-915. CHARLTON, T. R. 1988. Tectonic accretion and erosion in steady-state trenches. Tectonophysics, 149, 233-243. HALE, D. 1984. Dip moveout by Fourier transform. Geophysics, 49, 741-757. HUGHES, B. D. 1994. Reprocessing, modelling, and
interpretation of complex seismic reflection data from accretionary wedges: Application to the Timor Trough, Indonesia. MSc thesis, University of Leeds. KARIG, D. E., BARBER, A. J., CHARLTON, T. R., KLEMPERER, S. & HUSSONG,D. M. 1987. Nature and distribution of deformation across the Banda Arc-
Australian collision zone at Timor. Bulletin of the Geological Society of America, 98, 18-32. MASSON, D. G., MILSOM, J., BARBER, A. J., SIKUMBANG, N. & DWIYAYrO, B. 1991. Recent tectonics around the island of Timor, eastern Indonesia. Marine and Petroleum Geology, 8, 35-49. MCCAFFREY,R. 1996. Slip partitioning at convergent plate boundaries of SE Asia. This volume. PIGRAM, C. J. & PANGGABEAN,H. 1984. Rifting of the northern margin of the Australian continent and the origin of some of the microcontinents in eastern Indonesia. Tectonophysics, 107, 331-353. SNYDER, D., PRASETYO, H., BLUNDELL, D. J., PIGRAM, C. J., BARBER, A. J., RICHARDSON, A. & TJOKOSAPROETRO, S. 1996. Style of crustal deformation across the Banda Arc continent-arc collision zone as observed on deep seismic reflection profiles. Tectonics, in press. YtLMAZ, O. 1987. Seismic data processing. Investigations in Geophysics, No. 2. Society of Exploration Geophysicists.
Extension,
collision and curvature
in the eastern Banda
arc
JOHN MILSOM, STEVE KAYE & SARDJONO
Research School of Geological and Geophysical Sciences, Birkbeck College and University College London, Gower Street, London WC1E 6BT, UK A b s t r a c t : The Kai Islands occupy the region of maximum curvature in the east of the Banda arc, where the Aru trough has been regarded as the surface trace of past subduction and present arccontinent collision. Eocene to Pleistocene sediments on Kai Besar, the easternmost island, have not been deeply buried or imbricated but have experienced large-scale extensional faulting. The associated Bouguer gravity high of more than +200 mGal requires upfaulting of the accretionary complex, the attenuated Australian continental crust on which it rests and the underlying mantle at the western side of the Aru trough. Seismic reflection surveys show the deformation front within the Aru trough SE of the Kai Islands but entirely to its west further north. Instead of continuing NNE to an offset near the coast of New Guinea, the collision trace passes through the narrow and shallow strait between Kai Besar and the other islands, and thus mimics the relatively smooth curve of the Bouguer gravity contours, rather than the discontinuities of the bathymetric troughs. The continuity in deep and shallow structures is strong evidence for the existence of the outer arc as a single geological unit prior to the present phase of arccontinent collision.
The B a n d a arc, in eastern Indonesia, is the site o f the collision b e t w e e n the Australian continental margin, which is m o v i n g north with the IndoAustralian plate, and the B a n d a Sea which lies on the SE Asia plate (Fig. 1). The collision has been
]~2*e
:::
/ Fadol
t~3*~
~.'..'"]].
:
.~w
..... ~ - - , = !
¢
":-
.............
i':,-.~J ~ . J
.....
TI sal ya anndds u
-
I
.......... 6°s
I
:
" Weber I ........................ Basin
....... :!................ *J~:................ ~ o
Banda
"
::
Sea
"
~ ; "'/e "S i.." ,i#<--~ °-: - ,I: ~ 6.'.'~!.
"~iml>,,_,"
. Timor
-.
~
Ke,:n
........................
.
extensively studied in Timor, where the processes involved are relatively simple and where transfer o f imbricated material from the Australian margin to SE Asia has been well d o c u m e n t e d (cf. B a r b e r 1979; Charlton et al. 1991a). Similar processes
.........,i?.
...........~
'
o-"
• ..-.ii["
~ii:" a,-u
o.>'~..
~6"~"°~ 2 0 0 Km
............. t ~
-: ;.ll(d~ 124°F.,
.............
128°E
~
i .............
lO°S
1,52<>E
Fig. 1. The eastern Banda Sea. Inset shows details of the Kai Islands.
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 85-94.
85
86
J. MILSOM ET AL.
operate further to the east, where impressive images of continental crust underthrusting the arc have been provided by multichannel seismic surveys near the islands of Kisar and Damar (Richardson & Blundell 1996) and Tanimbar (Schltiter & Fritsch 1985). The similarity of the seismic sections from these two areas is surprising since the collision trace swings markedly to the north in the Tanimbar region to an orientation almost parallel to the present-day SE AsiaAustralia convergence vector. The problem is further compounded on the northern side of the arc, where the island of Seram has been described as geologically a mirror image of Timor (AudleyCharles et al. 1979). The very similar seismic reflection sections from the Seram Trough (Hamilton 1979: Jongsma et al. 1989a) presumably
record collision between Southeast Asia and the Bird's Head peninsula of New Guinea. The Banda arc thus appears to achieve the impossible by being involved in collision simultaneously, and with the same plate, to the north and to the south. As noted by McCaffrey & Abets (1991), either the Australia-New Guinea plate or the SE Asia plate, and quite possibly both, must be undergoing considerable internal deformation. The nature of the tectonic regime in the eastern part of the Banda arc has been controversial from almost the earliest days of plate tectonics. The Timor, Tanimbar, Am and Seram troughs which separate the islands of the outer Banda arc from the Australian continental margin are generally less than 50 km across and only 1.0-2.5 km deep. The Am trough, east of the Kai Islands, is exceptional,
~. .-.-.~.. ~"
,?
"7..~7".
/. / / ~ . ~ _ _ _ \
~b
L_z-" ~',,"
.
' // /
5 °S-
/
Weber '\ i B a s i n /'
~,~
)(:e-xi:N:
7ooo~/~
/ /
.....-"
• IJT-B -.~.
.
' o
-
" i
../
. //"
?G
/;//
\J
/
/
/
i/
/
/,
'/S"
/
/
/ /.
• ~. "~II-,,-a
/ F-
,
.
7os -
Arafura Sea
I'i/ Ooo
0
100 Km
ARU TROUGH Bathymetry Contours at 500m intervals ]OE/~
l~oF]
and at 200m,
Fig. 2. The Kai-Tanimbar region, showing locations of seismic sections interpreted or discussed in the text. Bathymetric contours at 500 m intervals, and at 200 m. Bathymetry reproduced from GEBCO sheet 5.10 compiled by D. Monahan, R. K. H. Falconer & M. Tharp and made available through the GEBCO Digital Atlas published by the British Oceanographic Data Centre on behalf of the IOC and IHO, 1994.
87
EXTENSION, COLLISION ~ CURVATURE, E BANDA ARC
being about 100 km across at its widest point and reaching a maximum depth of more than 3.5 km (Fig. 2). Fitch & Hamilton (1974) identified this trough as the subduction trace a view contested by Audley-Charles & Milsom (1974), who cited Bowin et al. (1980) as their authority for an extensional origin for the Aru trough and suggested the 7 km deep Weber Basin west of the Kai Islands as an alternative collision site. Although the reason for the great depth of the Weber Basin is still uncertain (but see McCaffrey 1988), later authors have generally followed Hamilton (1979) in placing a collision trace in the Am trough, often transferring motion from its northern end to the Scram trough by strike-slip motion along the Tarera-Aiduna fault in the extreme southwest of the Bird's Head (Fig. 3a). Geophysical data from the Kai Islands favour a third possibility which was hinted at by SchRiter & Fritsch (1985) in one of their illustrations (Fig. 3b) but was not discussed in,their text.
Geology of the Kai Islands The Kai archipelago (Fig. 1 inset) is one of the major island groups of the outer Banda arc, occupying the region of maximum arc curvature. The onshore geology has been described by Achdan & Turkandi (1982) and by Charlton et al. (1991b). The latter authors divided the group into Eastern, Central and Western Provinces, of which the Eastern Province consisted of the high, rugged
and strongly N-S elongated island of Kai Besar. The Central Province included Kai Kecil and the Tayandu Islands and the Western Province comprised a number of small islands, including Kur and Fadol, close to the edge of the Weber Basin. Kai Besar, which contains the major exposures of non-coralline rocks in the archipelago, consists of heavily block-faulted sediments of the Australian margin belonging to the Eocene, deepwater, Elat Formation, the shallow water OligoMiocene Tamangil and Weduar Formations and the Pliocene Weryahan Formation which consists of shallow-water limestone and marl. Achdan & Turkandi (1982) showed the Pliocene rocks in depositional contact on all older formations but Charlton et al. (1991b) failed to find any evidence of this unconformity. All geological discussions of Kai Besar comment on the strong NNE-SSW faulting, which Charlton et al. (1991b) considered to be extensional, both on the basis of the regional geological setting at the edge of the Am trough and because of the consistent pattern of large eastfacing escarpments (reaching to a maximum of 800 m above sea-level) and gently west-tilted sediments. The faults are offset or terminated at WNW-ESE lineaments which were interpreted as transfer faults. Kai Kecil, which is separated from Kai Besar by Selat Nerong (Nerong Strait), consists mainly of slightly raised coral reef. The occasional outcrops of the Weryahan Formation may occupy the cores of gentle anticlinal upwarps. The other islands of the Central Province are generally similar and are
....... % ~
............
# (a)
(b)
Fig. 3. The Kai-Bird's Head triple junction, showing plate boundary solutions suggested by (a) Hamilton (1979) and (b) Schltiter & Fritsch (1985). Bathymetry as Fig. 2 but with values omitted.
88
J. MILSOM ET AL.
notable for the presence of the mud volcanoes which are a feature of accretionary complexes around the entire outer Banda arc (cf. Barber et al. 1986). There is no evidence of imbrication and the carbonate platform is therefore assumed to have developed on top of the accretionary complex after convergence had ceased. High grade metamorphic rocks (silicic schists, gneiss and migmatite) which are exposed on the small islands of the Western Province have been correlated by Charlton et al. (1991b) with the Kobipoto Complex of Seram.
Seismic reflection data Examples of multichannel seismic reflection sections obtained by the research vessel R/v S o n n e along lines which crossed the southern end of the Aru trough were published by Schltiter & Fritsch (1985), and sections obtained by the Dutch Snellius II project across the trough NE of the Kai Islands using a 4-channel system have been published by Jongsma et al. (1989b). BGR-93, the northernmost of the Sonne lines, crossed the Am trough only a little south of Kai Besar but did not extend as far west as the offshore prolongation of the island (Fig. 2). The portion of the section reproduced by Schltiter & Fritsch (1985) shows Australian continental margin sediments dipping beneath unstratified material, identified as melange, at a deformation front in the deepest part of the trough, but ends in the west in a region where water depths still exceed 2500 m and so does. not show any of the structures at the trough's western margin. A significant difference between the section obtained on this line and those ?rom lines further south is the upfaulting of the sediments of the continental margin by about 1 s twt beneath the melange wedge about 15 km west of the deformation front. Schliiter & Fritsch (1985) interpreted the fault as a thrust but the evidence for overstepping of the upper block on the lower is not strong; the fault seems to be high angle but could be either normal or reverse. The southernmost of the Snellius lines (Line A2) ran from east to west across the Aru trough about 30 km north of the northern tip of Kai Besar. In the west it fell just short of the prolongation of the axis of the island but it did reach the East Kai Ridge, a bathymetric high between Kai Besar and the Am trough which rises to within 500 m of the sea surface. Along Line A2, the ridge is in water about 2000 m deep and the floor of the trough is at a depth of about 3300 m (Jongsma et al. 1989b). Underthrusting Australian continental margin was not imaged; this may have been because the seismic system used lacked the penetration of the system mounted on the Sonne or because the trough fill is considerably thicker (approaching 1 km), but the
sediments are virtually horizontal and the section can be interpreted in purely extensional terms. Arcward-dipping and arcward:thickening sediments do occupy the shallower basin between the East Kai Ridge and the Kai Plateau but there is no evidence for convergence there and the boundary between the two appears to be a large normal fault. The Kai region has also been covered by a high quality seismic survey which included data recorded to 8 s twt along lines which passed close to the northern and southern ends of Kai Besar (respectively UT-A and UT-B of Fig. 2). Figure 4 shows line-drawing interpretations of these sections. The interpretation of line UT-B which passed just south of the Kai Islands may be compared in part with Line BGR-93. In the southeast, beneath the shallow waters of the Arafura Sea, sediments of the Australian margin have been subject to normal faulting with slight rotation. Faulting increases into the Am trough and the eastern flank of the trough is underlain by fault blocks, some of which are noticeably tilted. The deepest part of the trough appears to coincide with the outcrop on the sea floor of the deformation front and further west the continental margin sequence can be traced, deepening towards the west, as far as the steep NW wall of the trough. There is a very thin layer of flat-lying sediment above the seafloor trace of the deformation front but little or no sediment cover on the accretionary complex within the trough. The abrupt change of sea-floor slope at the foot of the NW wall of the trough strongly suggests normal faulting, and this interpretation is reinforced by the reappearance of the west-deepening continental margin reflectors within the upthrown block. The throw on the fault appears to amount to about 5 s twt. The seismic character of the material which overlies the dipping reflectors and which forms the southern extension of the Kai Plateau is similar to that of the accretionary wedge in the Aru trough, although a little more high-frequency content has been preserved. The implications are clear; the Aru trough is an extensional feature bounded on both sides by normal faults, with the main basin-controlling fault in the west. The Kai Plateau, where crossed by the line, is part of the accretionary prism and therefore equivalent to southern Timor and perhaps to a part of the inner slope down to the Timor trough, but has been displaced upwards along major normal faults which post-date the main convergent movements. The time available for extensional faulting has been very short; there are still volcanic eruptions in the eastern Banda arc and a considerable level of seismic activity although, as demonstrated by McCaffrey & Abers (1991), fault plane solutions suggest extension in the trough and strike-slip
EXTENSION, COLLISION & CURVATURE,E BANDA ARC o
1 O0
I
I
I
~ Kai Kecil
..
"-
.-
--~
."
6000
~lr
I
~ ~,\~Y:~:~'--
f..,.._
;
I
--
-
~
:
I
~ ,~
.
"
^'7"f'-.~-_
Besar
~
,
Km
I
East Kai Ridge
-~.
t--~ -
I
. r .
x~
.,,
(N) Line UT-A
,/_
-~F~--~'~
" -r~%" ~
"
'
-.~
W S W -"
~ "~-'=:
ENE
(S) 0
I
__
I
89
I
I
50 I
I
I
I
I
1 O0 I
I
Line UT-B I
i
: ..~. A r u Trough ~ ~ ' ~ - ; ~ . . . ~ / / ',' .~.~_r2~;~\ - . . . . . . . . . . . . . . . ~ - ~ F " . . . . . . ~ , " , ' 7 7 ~ --- ~ ' ~ = ~ ' - ~ _ ~ Australian I / 1 " " ,~/'/ " \ "~)g/'~, ~ ~ "--':"~ . . . . Shelf ,- Z //°?, / , .
Z000 ~
6000
Plateau ---~ -
150 Krn
I
i
.
-~'L :
' , . \' ~ " \ ' / , , ' f " , / ~ "~"-
\
// SE
NW
Fig. 4. Line drawing interpretation of seismic lines north and south of the Kai Islands. Vertical scale in milliseconds two-way travel time. Static corrections have been applied, replacing seawater with a layer of velocity 2000 m s-l, so that vertical scales also give bathymetric depths in metres. Original sections provided by Union Texas (Kai) Inc., with the permission of Pertamina. See Fig. 2 for line locations.
rather than thrust motion in the Banda Sea. Although the present Am trough is a young feature, the graben framework may well have been established in the Mesozoic, when the passive margin of northwest Australia was formed by rifting. The line drawing in Fig. 4 of the seismic section obtained just to the north of Kai Besar, shows a very different picture. The East Kai Ridge, at about 2000 m, is underlain by sediments which dip gently westward and are similar in seismic character to the Australian continental margin sediments on the southern line. There are no indications of an accretionary complex beneath the ridge, which is consistent with their absence from the Snellius seismic sections of Jongsma et al. (1989b), which were obtained across the whole width of the trough. However, an apparent accretionary complex can be seen beneath the Kai Plateau, overlying westdipping reflectors which emerge at the sea floor in the northern extension of Selat Nerong, the narrow and shallow seaway between Kai Besar and Kai Kecil. The region between this easternmost part of the Kai Plateau and the East Kai Ridge is dominated by diffraction patterns which were probably generated by closely spaced normal faults. Kai Besar is therefore on the Australian side of the deformation front and underlain by Australian continental crust. The deformation front is east of Kai Besar in the south but west of Kai Besar in the north.
Gravity
data
The very strong gravity gradients associated with the Banda arcs have been known ever since the pioneering marine surveys of Vening Meinesz (1932, 1954). The gradient has since also been defined on land, notably in eastern Timor where Bouguer gravity increases over a distance of 50 km from close to zero along the axis of the island to more than +100 reGal at points on the north coast (Chamalaun et al. 1976; Kaye 1990). Similar gradients between similar limiting values have been reported from Seram (Milsom 1977) and from Tanimbar (Kaye 1990). Apart from its size and steepness, the Banda gravity gradient is also notable for its continuity and smooth curvature in plan view, controlled largely by deep crustal structure and therefore seen most clearly on regional maps of Bouguer gravity (Fig. 5). The pattern appears to be little affected by the presence of island groups, confirming the source of the gradient as deep-seated. If the position of the steepest slope can be taken as a guide, then in the Tanimbar group the main island is part of the same geological zone as southern Timor and equivalents of northern Timor are to be found in the small islands closer to the Weber Basin (Kaye 1990). Land gravity measurements have been made throughout the Kai Islands, with the majority of
90
J. MILSOM E T AL.
Fig. 5. The Bouguer gravity field of the Banda Sea region. Contours, at 25 mGal intervals, after Bowin et al. (1980) with additional information from Woodside et al. (1989) and unpublished data. Stipple indicates regions of negative Bouguer gravity.
stations situated within a few metres of the shoreline which provided a ready, if approximate, height reference (Kaye 1990). The Bouguer gravity map (Fig. 6) has been prepared by combining the onshore data with data from lines UT-A and UT-B. The Bouguer gravity low departs from the bathymetric axis, which trends NNE into the Aru trough and instead curves smoothly into the Seram trough, passing to the west of the main islands of the group. The contour pattern is dominated by a high which peaks at more than +200 mGal at the east coast of northern Kai Besar and by a broader low, with a minimum value of less than -30 mGal, over the western islands of the Central Province (Fig. 6). From the low there is a moderately strong gradient eastwards across Kai Kecil and a steeper one westwards across Kur and Fadol towards the Weber Basin. This pattern suggests that the Tayandu Islands belong to the same geological zone as southern Timor and the basinal areas of northern Seram. The main island of the Central Province, Kai Kecil, is then gravitationally analogous to the Timor trough or even to its outer (Australian) flank, although there is no deep marine basin between it and the other islands of the group, either to the east or the west. The strong gravity high which is associated with Kai Besar (Eastern Province) has no parallel elsewhere in the outer Banda
arc or on the Australian shelf, and is exceptional, in global terms, for any area underlain by continental crust. Marine gravity work has shown that the high extends, with diminishing intensity, almost to the coast of New Guinea (Woodside et al. 1989). The gravity profile interpreted in Fig. 7 was extracted from the data shown in Fig. 6 along a line oriented approximately ESE which crosses the centre of the Bouguer maximum on Kai Besar. This is the area in which gravity interpretation has most to offer. A profile based on seismic line UT-B, which passed south of Kai Besar, could have been based on actual gravity observations for a greater distance but would have been less interesting since the interpretation would have closely resembled that already published by Schltiter & Fritsch (1985). On the other hand, complications would have arisen in modelling a profile based on line UT-A because it would run parallel to, rather than across, the trend of the Bouguer contours immediately north of Kai Besar. The profile selected has its origin at the boundary between the West and Central provinces (i.e. between the Tayandu group and Kur and Fadol), close to the start of the Bouguer gravity rise towards the Weber Basin. About 50 km east of the origin, the line passes close to the western end of
EXTENSION, COLLISION ~ CURVATURE,E BANDA ARC
91
13~2°E 1j~tk 180
IN
6°S 0 I
50 Km I
Contour interval 10 reGal I
Fig. 6. Bouguer gravity of the Kai Islands. Onshore data from Kaye (1990). Offshore data from profiles provided by Union Texas (Kai) Inc., with the permission of Pertamina.
UT-A, but diverges thereafter since the seismic line, necessarily, passes to the north of Kai Besar. Despite this, the seismic section provides an important constraint on the shallow structure used in gravity modelling. Deep structures are less well controlled but a satisfactory fit with observation is difficult to obtain without supposing that only the relatively thin outer margin of the Australian continental crust has been subducted and that further crustal thinning occurred during the formation of the Ayu Trough. The gravity low around the Tayandu Islands and the high over Kai Besar require lateral changes in density both at shallow depths, to account for the steepness of the gradients, and in deep structure, to explain the large differences in absolute levels. The peak value of +200 mGal on Kai Besar is typical of intra-oceanic arcs and oceanic islands but can be achieved in an area underlain by continental crust by raising the Moho to an extent which implies both that the crust is thin and that the Aru trough western boundary faults cut down to and displace the Moho. This interpretation is consistent with the seismic reflection data. The gravity data also provide direct support for the existence of an accretionary complex beneath the Central Province. The gravity low requires both crustal thickening, which is provided by the doubling up of crust implied b y subduction, and low density material, either young stratified sedi-
ments or melange, near the surface. There are no indications of significant thicknesses of young undeformed sediments on seismic sections across the Central Province but Charlton et al. (1989b) noted the presence of mud volcanoes, which are characteristic of melange terranes. The model therefore incorporates an accretionary wedge; its density and thickness can be varied widely and still satisfy the gravitational constraints but the essential features of the model are preserved through these changes. The models are most easily brought into agreement with the observed field if a density of 2.2 Mg m-3 or less is used for the main part of the wedge. It seems fair to use the term melange for what is obviously a very disordered assemblage, but geological mapping in the possibly equivalent terrane of southern Timor has shown that large blocks of sediment have retained their original internal structure (cf. Barber 1979). The model includes a number of layers of sediment on the downgoing plate, the youngest of which are incorporated into the melange at the deformation front while the older are preserved to considerable depths beneath the wedge. These features are introduced to maintain compatibility with the seismic sections and are not dictated by the gravitational constraints. Gravity modelling can say little about the degree to which sediments are subducted or are incorporated with little deformation within the melange, and the densities assigned
J. MILSOMET AL.
92 200 f' Bouguer gravity (reGal)
I (Reductiondensity 2.2 Mg,m~) J°°t
~
~
~
Calculated ~ - .
ot, W N W . ~
()bscrvcd
• •
ESE
5~
_ : ..."...,.,.,.o.:...e_ s~lsm°~"ni[,
0
50
100
Kilomctrcs 150
Australian continental
200
crust
250
Fig. 7. Gravity profile and model cross-section across northern Kai Besar. The upper part of the accretionary complex (above the dotted line) has been assigned a density contrast of-0.5 Mg m-3. with standard crust, corresponding to a real density of 2.17 Mg m-3, the lower part a density contrast of -0.3 Mg m-3. Tertiary sediment layers have density contrasts of -0.47, -0.42 and -0.3 Mg m-3 with standard crust. Arc crust, continental crust and Mesozoic sediments have been assigned standard density.
to the sediments are not critical to the success or failure of the model. Discussion There are three possible routes by which the deformation front defined by seismic reflection surveys south of Timor might link with its mirror image defined by Jongsma et al. (1989a) in the Seram trough. The seismic sections of Schltiter & Fritsch (1985) which imaged the deformation front in the eastern parts of the southern Am trough eliminated the western route through the Weber Basin as a possibility and could also be considered as support for the eastern route along the Am trough to a transfer at the Tarera-Aiduna fault. The absence of the front from the seismic sections presented by Jongsma et al. (1989b) is not strong evidence against this latter possibility since the weak source used lacked deep penetration in the thick sediments. However, the UT-A seismic section (Fig. 4) is interpreted as showing a westdipping thrust which separates stratified sediments on Kai Besar from an accretionary wedge in the Kai Central Province. This interpretation is supported by the fact that the Eocene to Pliocene sediments on Kai Besar have been tilted but not imbricated and
show no signs of deep burial. The faulting they have suffered has been interpreted as extensional by Charlton et al. (1991b) and this seems to be in accord with the seismic data. It is almost impossible to reconcile such an interpretation with the existence of a true deformation front beneath the sediments in the Aru trough since fault displacement of the order of l0 km followed by considerable erosion would be needed to expose this front on the upthrown block, and these processes would have had to have taken place in the short period since the subduction zone locked. The trace of past subduction must therefore pass through Selat Nerong and then curve into the Seram trough. The fault locations in Fig. 8, which illustrate this solution, are inevitably schematic, being constrained by seismic data along the seismic lines of Fig. 4, but elsewhere being drawn on the basis of bathymetric features which are themselves not well controlled. Almost certainly the true pattern is more complex, but such complexity would not affect the conclusion that the deformation front broadly parallels the curvature of the Bouguer gravity contours and does not follow the bathymetric trends along the Am trough The cause o f the extreme curvature of the Banda arc is one of the major unsolved problems of SE
EXTENSION, COLLISION •
93
CURVATURE, E BANDA ARC
5°S
J k.~
/
0
c,.ar
-
100 K m
...
1 °E~"
133°E
Fig. 8. The Kai-Tanimbar region, showing suggested interaction of extensional faulting with earlier thrust faulting. Strong black lines show thrust faults of the deformation front, narrower lines mark possible locations of faults of the Aru trough extensional system, as suggested by bathymetry. Because of the absence of major geological differences between northern and southern Kai Besar (cf. Charlton et al, 1991b), the narrow isthmus which links the two is attributed to transfer faulting associated with the extensional system, rather than with the passage of the thrust from the Aru trough to Selat Nerong.
Asian geology (cf. Linthout e t al. 1991). Evaluation of the processes involved has been hampered by the lack of a universally accepted model of arc evolution. There are three broad possibilities, one of which is that the curvature has been a feature of the arc since its inception. This has found little favour (i) because of the problems involved in incorporating a tightly curved arc into reconstructions of the area during the Neogene, (ii) because of the difficulty of transporting some Australian continental fragments to their present positions within the arc, (iii) because of the evidence for some 74 ° counter-clockwise rotation of Seram (Haile 1981) and (iv) because of the
evidence for continuing N - S compression and E - W extension in the Banda Sea (McCaffrey & Abers 1991). An alternative possibility is that the arc was assembled relatively recently from two or more elements with quite different previous histories. Such a solution is attractive since it poses few mechanical difficulties and places few constraints on Tertiary reconstructions, but it has always encountered an obstacle in the evidence for continuity in deep structures provided by regional gravity maps. If it could be shown that the subduction trace coincided with the Aru trough and was offset along the Tarera-Aiduna fault (Fig. 3a),
94
J. MILSOM ETAL.
this objection might be discounte&since the offset w o u l d imply a major discontinuity in the upper, as well as the lower, plate. The seismic reflection evidence that curvature o f the deformation front m a t c h e s the curvature o f the B o u g u e r gravity contours, and the detailed gravity interpretation w h i c h c o n f i r m s this c o n c l u s i o n , nullify this argument. Whatever the geometrical difficulties involved in bending not merely a crustal sliver but
an associated subducted slab, it seems that this is what has happened in the Banda arc. We thank Union Texas (Kai) Inc. for providing access to their unpublished seismic sections and marine gravity data and permitting reproduction of our line drawing interpretations of the seismic data. Sardjono publishes with the permission of the Director, Geological Research and Development Centre, Bandung.
References ACHDAN, A. & TURKANDI,T. 1982. Preliminary geologic map of the Kai (Tayandu and Tual) quadrangles, .Maluku, Indonesia (Scale 1:250,000). Geological Research and Development Centre, Bandung. AUDLEY-CHARLES, M. G. & MILSOM, J. 1974. Comment on 'Plate convergence, transcurrent faults and internal deformation adjacent to south east Asia' by T. J. Fitch. Journal of Geophysical Research, 79, 4980-4981. , CARTER, D. J., BARBER, A. J., NORWCK, M. S. & TJOKROSAPOETRO, S. 1979. Reinterpretation of the geology of Seram: implications for the Banda Arcs and northern Australia. Journal of the Geological Society, London, 136, 547-568. BARBER, A. J. 1979. Structural interpretation of the island of Timor. SEAPEX Proceedings, 4, 9-21. , TJOKROSAPOETRO,S. & CHARLTON,T. R. 1986. Mud volcanoes, shale diapirs, wrench faults and melanges in accretionary complexes, eastern Indonesia. AAPG Bulletin, 70, 1729-1741. BOWIN, C., PURDY, G. M., JOHNSTON, C., SHOR, G., LAWVER, L. HARTONO, H. M. S. & JEZEK, E 1980. Arc-continent collision in the Banda Sea region. AAPG Bulletin, 64, 868-915. CHAMALAUN, E H., LOCKWOOD, K. & WHITE, A. 1976. The Bouguer gravity field and crustal structure of eastern Timor. Tectonophysics, 30, 241-259. CHARLTON,T. R., BARBER, A. J. & BARKHAM,S. T. 1991a. The structural evolution of the Timor collision complex, eastern Indonesia. Journal of Structural Geology, 13, 489-500. , KAYE, S. J., SAMODRA,H. & SARDJONO. 1991b. The geology of the Kai Islands: implications for the evolution of the Aru Trough and the Weber Basin. Marine and Petroleum Geology, 8, 62-69. FITCH, T. J. & HAMILTON,W. 1974. Reply to comments on Audley-Charles, M. G. and Milsom, J. on 'Plate convergence, transcurrent faults and internal deformation adjacent to south east Asia' by T. J. Fitch. Journal of Geophysical Research, 79, 4982-4985. HALLE, N. S. 1981. Palaeomagnetic evidence and the geotectonic history and palaeogeography of Eastern Indonesia. In: BARBER, A. J. & WIRVUSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication 2, 81-87.
HAMILTON,W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078. JONGSMA, D., WOODSIDE,J. M., HUSON, W., SUPARKA,S. KADARISMAN, D. 1989a. Geophysics and tentative Late Cenozoic seismic stratigraphy of the Banda Arc-Australian continent collision along three traverses. Netherlands Journal of Sea Research, 24, 205-229. --, HUSON, W., WOODSIDE, J. M., SUPARKA, S., SUMANTRI, T. & BARBER, A. J. 1989b. Bathymetry and geophysics of the Snellius II triple junction and tentative seismic stratigraphy and neotectonics of the northern Aru Trough. Netherlands Journal of Sea Research, 24, 231-250. KAYE, S. K. 1990. The structure of eastern b~donesia: an approach via gravity and other geophysical methods. PhD thesis, University of London. LINTHOUT, K., HELMERS, H. & ANDRIESSEN,E A. M. 1991. Dextral strike-slip in central Seram and 34.5 Ma Rb/Sr ages in pre-Triassic metamorphics related to early Pliocene counterclockwise rotation of the Buru-Seram microplate. Journal of SE Asian Earth Sciences, 6, 335-342. M¢CAFFREY, R. 1988. Active tectonics of the eastern Sunda and Banda arcs. Journal of Geophysical Research, 93, 15163-15182. -& ABERS, G. A. 1991. Orogeny in arc-continent collision: the Banda Arc and western New Guinea. Geology, 19, 563-566. MILSOM, J. 1977. Preliminary gravity map of Seram, eastern Indonesia. Geology, 5, 641-643. RICHARDSON, A. N. & BLUNDELL,D. J. 1996. Continental collision in the Banda arc. This volume. SCHLt3TER, H. U. & FR~TSCH, J. 1985. Geology and tectonics of the Banda arc between Tanimbar Island and Aru Island (Indonesia). Geologisches Jahrbuch, E30, 3-41. VENING MEINESZ, F. A. 1932. Gravity expeditions at sea, Vol I. Netherlands Geological Commission, Delft. -1954. Indonesian archipelago-a geophysical study. Geological Society of America Bulletin, 65, 143-164. WOODS1DE,J. M., JONGSMA,D., THOMMERET,M., STRANG VAN HEES, G. L. & PUNTEDEWO. 1989. Gravity and magnetic measurements in the eastern Banda Sea. Netherlands Journal of Sea Research, 24, 185-203.
Pre-Cretaceous evolution of SE Asian terranes I. M E T C A L F E
Department of Geology & Geophysics, University of New England, Armidale NSW 2351, Australia Abstract: During the last decade, a wide range of geological and geophysical data has led to the recognition of various continental terranes in SE Asia which, on tectonostratigraphic, palaeobiogeographic and palaeomagnetic grounds are 'suspect' or allochthonous in nature. Some of the recognized terranes may be composite and there is some disagreement with regard to the number of terranes and their boundaries. The continental terranes are bounded by sutures (representing former oceans), by narrow mobile belts or major fault zones. Comparative studies of the stratigraphy, palaeontology, and palaeomagnetism of the various pre-Cretaceous continental terranes of East and SE Asia suggest that they were all derived directly or indirectly from Gondwanaland. Asian continental terranes that are placed on the India-Australian margin of Gondwanaland in the Early Palaeozoic include Tarim (here regarded to include the Kunlun and Ala Shan terranes), Qaidam, Indochina/East Malaya (which includes the Qamdo-Simao block of western China), North & South China, Sibumasu, Qiangtang, Lhasa, Kurosegawa, NW and SE Hainan, West Burma and the Woyla terranes. The evolution of Gondwanaland and Tethys during the Palaeozoic and Mesozoic involved the rifting of continental slivers/fragments from northern Gondwanaland, and the northwards drift and amalgamation/accretion of these to form proto-East and SE Asia. Three continental slivers were rifted from the northern margin of Gondwanaland in the Early to Late Devonian (North China, South China, Indochina/East Malaya/Qamdo-Simao, Qaidam and Tarim terranes); Early-Middle Permian (the Cimmerian continent including the Sibumasu and Qiangtang terranes and possibly NW and SE Hainan); and Late Triassic to Late Jurassic (Lhasa, West Burma and Woyla terranes). The northwards drift of these terranes was accompanied by the opening and closing of three successive oceans, the Palaeo-Tethys, Meso-Tethys and Ceno-Tethys. Assembly of Gondwanaland-derived Asian terranes began with the amalgamation of South China and Indochina/East Malaya along the Song Ma/Song Da zone during the Late Devonian/Early Carboniferous to form 'Cathaysialand'. Palaeomagnetic, climatic and bitgeographic data indicate that Cathaysialand and North China were located within the Palaeo-Tethys at low northern/equatorial latitudes during the Late Carboniferous and Permian. Palaeomagnetically determined palaeolatitudes are consistent with the development of tropical Cathaysian floras on these terranes. The Tarim, Kunlun, Qaidam and Ala Shan terranes accreted to Kazakstan/Siberia in the Permian. A major episode of rifting occurred on the northern margin of Gondwanaland in the Late Carboniferous-Early Permian and the Cimmerian continent separated in the late Early Permian resulting in the opening of the Meso-Tethys. Suturing of Sibumasu and Qiangtang to Cathaysialand occurred in the Late Permian-Triassic, closing a major branch of the PalaeoTethys. South and North China amalgamated and then accreted to Laurasia by Late Triassic-Early Jurassic times. The highly disrupted Kurosegawa terrane of Japan, possibly derived from Australian Gondwanaland, accreted to Japanese Eurasia, also in the Late Jurassic. The Lhasa, West Burma and Woyla terranes rifted from NW Australian Gondwanaland in the Late Triassic to Late Jurassic and drifted northwards during the Jurassic and Early Cretaceous as the Ceno-Tethys opened and the Meso-Tethys was destroyed by subduction beneath Eurasia. Accretion of these terranes to proto-SE Asia occurred in the Cretaceous. The SW Borneo and Semitau terranes were derived from the South China/Indochina margin by the opening of a marginal basin in the Cretaceous which was subsequently destroyed by southwards subduction during the rifting of the Reed Bank-Dangerous Grounds terrane from South China when the South China Sea opened. The NW and SE Hainan terranes, which formed part of Early Palaeozoic and possibly Late Palaeozoic Gondwanaland, reached their current positions, relative to South China, sometime in the Jurassic-Cretaceous.
The pre-Cretaceous evolution o f SE Asia has received m u c h attention during the last decade and has been the principal focus of two recent I G C P projects, IGCP 224 (Pre-Jurassic Evolution o f East Asia) and I G C P 321 ( G o n d w a n a Dispersion and
Asian Accretion). The evolution of East and SE Asian terranes, in the framework of the evolution of Gondwanaland, Laurasia, Pangaea and Tethys, has important scientific implications for both regional and global earth sciences. Detailed information on
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 97-122.
97
98
I. METCALFE
the movements (particularly cross-latitudinal) of continental blocks and the changing spatial distribution of seas and oceans provides a basis for investigating globally important problems. Past ocean currents and climatic patterns have been very important in determining biogeographical distributions of both marine and land organisms on the Earth and a knowledge of changing continent/ocean configurations helps us model these past climates and current patterns and leads to a greater understanding of the observed changes in both the distributions of organisms, and sedimentation patterns in the evolving seas and oceans. This paper forms a contribution to the lUGS Global Sedimentary Geology Program's Project Pangaea which is primarily concerned with modelling global climate during the existence of Pangaea. Of importance to this modelling is the continent/ocean configuration, and more specific to East and SE Asian terranes, was Tethys an empty open gap into Panthalassa or did it contain continental fragments, and if so what positions did these occupy relative to the rest of Pangaea? Biologists, botanists and zoologists are becoming increasingly interested in SE Asian terrane movements, particularly in the Mesozoic and Cenozoic, because the movements of continental fragments may have provided land 'rafts' or chains of islands that influenced land plant and animal evolution and dispersal in the region. Another rationale for detailed studies of terrane evolution in the Asian region is the increasing economic importance of these regional studies in understanding metallogenesis and the development and evolution of hydrocarbon and coal-bearing sedimentary basins in the region. It is now well established that continental SE Asia (together with large parts of eastern Asia) comprises a complex assembly of continental terranes (Figs 1 and 2) which are entirely allochthonous to central and northern Asia (Siberia and Kazakhstan). All the SE Asian terranes are interpreted to have been derived directly or indirectly from Gondwanaland and their Palaeozoic and Mesozoic history involved the rifting of terranes from the northern margin of Gondwanaland, their northwards drift and amalgamation/accretion to form proto-SE Asia (Seng0r 1979, 1984; Ridd 1980; Mitchell 1981; Stauffer 1983; AudleyCharles 1983, 1984, 1988, 1991; Metcalfe 1986, 1988, 1990, 1993, 1994a, c ; Klimetz 1987; Seng/3r et al. 1988; Burrett et al. 1990; Nie et al. 1990). The northwards translation of terranes involved the opening and closing of successive Tethys oceans. Previous workers have recognized a two Tethys ocean system comprising Palaeo-Tethys ('Old Tethys') and Neo-Tethys or 'New Tethys' (e.g. Seng6r 1979, 1984, 1989). The rift-driftsuturing (amalgamation/accretion) history of the
SE Asian terranes is here considered within a framework of a three Tethys ocean system following Metcalfe (1994a, c). It is proposed that three continental slivers rifted from the northern margin of Gondwanaland during the Palaeozoic to Mesozoic, opening and closing three successive Tethys oceans, the Palaeo-Tethys, Meso-Tethys and Ceno-Tethys. The Palaeo-Tethys is broadly equivalent to the Palaeo-Tethys of previous workers, although timings of the closure of the different branches of Palaeo-Tethys differ in detail. The Meso-Tethys and Ceno-Tethys are approximately equivalent in age to what has previously been referred to as 'Neo-Tethys' but are here regarded as successive oceans developed by the Triassic-Jurassic rifting of a third continental sliver from northeast Gondwanaland.
SE Asian pre-Cretaceous terranes and their origins The pre-Cretaceous continental terranes of East and SE Asia are shown in Figs. 1 and 2. Some of these terranes may be composite, but where the composite nature is not well established or is controversial, they are treated as single units. The Cretaceous and Cenozoic evolution of the small terranes distributed around the South China Sea and in eastern Indonesia (Fig. 2), including West Irian Jaya, Buru-Seram, Buton, Banggai-Sula, Obi-Bacan, North Palawan, Spratley IslandsDangerous Ground, Reed Bank, Luconia, Macclesfield Bank, Paracel Islands, KelabitLongbowan, Mangkalihat, Paternoster, West Sulawesi, East Sulawesi and Sumba has been discussed by Metcalfe (1990) and will not be repeated here. The pre-Cretaceous terranes discussed below include those which now constitute mainland SE Asia. Some discussion of other East Asian terranes which have a bearing upon the evolution of SE Asia is also included. South China
The South China block is regarded as composite by some authors (e.g. Hsu et al. 1988, 1989, 1990). Hsu et al. (1988; 1989) recognized component Yangtze and Huanan Blocks separated by the 'Xiangganzhe Mesozoic suture'. This interpretation has been challenged by Rogers (1989) and Rowley et al. (1989), and geochronological data from Anhui and Jiangxi Provinces suggest that the suture recognized by Hsu is in fact of Proterozoic age (Chen, J. et al. 1991). Hsu et al. (1990) and Chen et al. (1993) recognize four component blocks, the Yangtze, Xianggui, Cathaysia and Dongnanya blocks separated by three melange zones (Fig. 3).
SE ASIAN TERRANE EVOLUTION
99
TARIM NORTH CHINA
SOUTH CHINA
HT
WOYLA" TERRANES
KT SUTURES Q SongMa Q Aibi-Xingxing Q Xiliao-He G Kunlun G Qinling-Dabie (~) Jinshajiang Q Lancangjiang O Banggong O IndusYarlungZangbo Q Nan-Uttaradit G Raub-Bentong ShahBoundary Woyla Meratus G Boyan G Changning-Menglian (~ Ailaoshan
Fig. 1. Distribution of principal continental terranes and sutures of East and SE Asia. EM, East Malaya; WB, West Burma; SWB, South West Borneo; S, Semitau Terrane; HT, Hainan Island terranes; L, Lhasa Terrane; QT, Qiangtang Terrane; QS, Qamdo-Simao Terrane; SI= Simao Terrane, SG, Songpan Ganzi accretionary complex; KL, Kunlun Terrane; QD, Qaidam Terrane; AL, Ala Shan Terrane; KT, Kurosegawa Terrane.
Palaeomagnetic data suggesting a separation of the Yangtze and Huanan blocks by a latitude difference of 8.3 ° in the Triassic (Chen et al. 1993) have been used to support the pre-Triassic composite nature of South China. However, the shallow inclinations reported from the Yangtze and 'Huanan' blocks (-11.8 and -13.0 respectively) are not significantly different (even w i t h small o~95s), and do not confirm a Triassic separation of these regions. The occurrence of identical endemic Devonian fish faunas from both the Yangtze region and southeast South China, which define a South China Province (Young 1993) also indicate a unified South China block at that time. South China is here regarded as a single tectonic unit since the Palaeozoic following the arguments of Rogers (1989), Rowley et al. (1989) and Chen, H. et al. (1991). Its boundaries with North China and the Indochina block are the
Qin Lin-Dabei suture, and Ailaoshan and Song Ma Sutures, respectively. Lower Palaeozoic shallow marine faunas of South China have affinities with those of northeast Gondwanaland and belong to the Asia-Australian and Austral realms in the Cambrian and Ordovician respectively (Yang, J. 1994; Li 1994). These faunal affinities, together with stratigraphic comparisons suggest that South China had its origin on the Himalaya-Iran region of the Gondwanaland margin (Burrett et al. 1990; Nie et al. 1990; Nie 1991, 1994). The Lower Palaeozoic placement of South China adjacent to the Himalaya-Iran region is consistent with the palaeomagnetic evidence which places South China in mid southern palaeolatitudes in the Ordovician (Lin et al. 1985; Burrett et al. 1990; Fang et al. 1990; Metcalfe 1990). Hoffman (1991) also placed the East and SE Asian blocks
100
i. METCALFE
PHILIPPINE
zo.
SEA PLATE
SEA
LEGEND
"
0
500 km
"
X X X X Suture
~
,6. _¢~..,,x Thrust .,L
Inactive Trench
~
Active Trench
o
~!~i~iiii!i~!iii:i~i¸ ; i ~ii ¸ ::~,i ~i i: ¸
Terrane Boundary .... J
Inferred Terrane Boundary J OceanIc Crust
~
AccretedCrust
~ 1
Extended Continental Crust Continental Blocks and Fragments IOOE
Fig. 2. Distribution of continental blocks and fragments (terranes) and principle sutures of SE Asia (modified after Metcalfe 1990). 1. South China; 2. Indochina; 3. Sibumasu; 4. East Malaya; 5. West Burma; 6. SW Borneo; 7. Semitau; 8. Sikuleh; 9. Natal; 10. West Irian Jaya; 11. Buru-Seram; 12. Buton; 13. Bangai-Sula; 14. Obi-Bacan; 15. North Palawan; 16. Spratley Islands-Dangerous Ground; 17. Reed Bank; 18. Luconia; 19. Macclesfield Bank; 20. Paracel Islands; 21. Kelabit-Longbowan; 22. Mangkalihat; 23. Paternoster; 24. West Sulawesi; 25. East Sulawesi; 26. Sumba; 27. Banda Allochthon; 28. Qiongzhong and Yaxian terranes of Hainan; 29. Simao terrane.
es -~ ss.j s
-'--.
a~'~'~.
,, st" "",,,, _ S
adjacent to the NW margin of Australia in his Neoproterozoic supercontinent reconstruction. Recent comparisons of the Proterozoic basement of South China with Australia and Laurentia led Li et al. (1994) to propose an alternative model whereby South China is placed between Eastern Australia and Laurentia in the Late Proterozoic, forming part of Rodinia. Such a position would be at variance with an Early Palaeozoic position adjacent to Himalayan India, and Proterozoic palaeomagnetic data are required to test this proposition. The Upper Sinian and Palaeozoic tectonostratigraphy of South China is also not comparable with Eastern Australia. A northern Gondwanaland margin origin for South China is favoured here.
J
.,,-
...-......
[
./i
. ~ _ . . _ ~ Island
i
i
Fig. 3. Component blocks and melange zones of South China proposed by Chen et al. (1993).
Sibumasu
The Sibumasu block (Metcalfe 1984, 1988) is bounded to the west by the Shah Boundary Fault
SE A S I A N T E R R A N E E V O L U T I O N
and the Andaman Sea basin and to the southwest by the Woyla suture in Sumatra. Its eastern boundary is formed by the Raub-Bentong suture in Peninsular Malaysia, and the Changning-Menglian and Lancangjian Sutures in western Yunnan. The eastern boundary of Sibumasu in North Thailand and Burma is still not clear. The northwards extension of the Raub-Bentong suture has been regarded by many authors to be the Nan-Uttaradit suture in Thailand. This suture appears to bend northeastwards to be truncated by the Red River-Ailaoshan zone. In western Yunnan, the Changning Menglian suture zone (Wu & Zhang 1987; Liu Benpei et al. 1991; Wu Genyao 1993; Wu Haoruo et al. 1995) which separates the Baoshan/Tenchong blocks (disrupted parts of Sibumasu) from the Qamdo-Simao terrane (Fig. 4) appears to continue southwards into Burma and then possibly into the Chiang Rai region of north Thailand. Deep-marine ribbon-bedded cherts (Fang Chert) in the Chiang Rai region have been dated by radiolarians as Devonian to middle Permian (Sashida et al. 1993). These are associated with basic volcanics which may be subduction-related (Barr et al. 1990). Some authors have suggested that the Sibumasu terrane is composite and have recognized a western 'Mergui Platelet' (Cameron et al. 1980), 'Phuket terrane' (Cooper et al. 1989) or 'Mergui Group nappe' (Mitchell 1992) characterised by the presence of Carboniferous-Permian diamictites. There seems however to be little basis for separating the diamictite-bearing 'terrane' from Sibumasu and the implied thrust contact between the Mergui Group and equivalents and other Palaeozoic strata of Sibumasu has not been demonstrated. In fact, in northwest peninsular Malaysia, a stratigraphic contact can be unequivocally demonstrated (Hutchison 1993). However, new geochronological data from western Thailand (Ahrendt et al. 1993, 1994; Hansen et al. 1994) suggest that the high grade metamorphic rocks previously regarded as Precambrian basement in NW Thailand, have depositional ages of less than 600 Ma. New K-Ar, Rb-Sr and U-Pb ages and structural observations also indicate very young (Tertiary) cooling ages for the metamorphic 'core complex' but older (Mesozoic) ages for deformational ages in overlying low grade metamorphic rocks. This suggests the possibility of eastwards directed thin-skinned tectonics in western Thailand during the Tertiary. If this is the case, the geometry and boundaries of the Sibumasu terrane may have been disrupted and the Palaeozoic and Mesozoic strata may well have been detached from their original basement. Distinctive Gondwanaland faunas (Cambrian to Lower Permian) with NW Australian affinities on Sibumasu (Archbold et al. 1982; Burrett & Stait
|01
• Deqin
,. / 'TIBET t..
Lancangjlang Suture '
- ~
Late Permian-Triassic Volcanic Arc
~
Late Permian-Triassic Granite Ophiolitic rocks and oceanic sediments
i f i
MYANMAR
~
,i
Chongshan metamorphic belt
J
120 km I
D
~t i ~ s
i
Baoshan •
SOUTH CHINA TERRANE
Changning
BAOSI-IAN ".-- BLOCK
';
Ailaoshan Suture
/
Menglian Suture
SIMAO'
BLOCK • Simao v
z/
' MYANMAR
'
,
--.\~, ~,-
". , i i
VIETNAM
I_AOS "f i
Fig. 4. Principal blocks and suture zones of western Yunnan (after Wu et al. 1995).
1985; Metcalfe 1986, 1988, 1990, 1991, 1993; Burrett et al. 1990; Shi & Waterhouse 1991; Racey et al. 1994; Archbold & Shi 1995) strongly suggest a NW Australian origin for the Sibumasu terrane. This is supported by the presence of Upper Carboniferous-Lower Permian glacial-marine diamictites (Stauffer & Mantajit 1981; Stauffer & Lee 1989, Jin 1994a, b; Wopfner 1994), Lower Permian cool-water faunas and 8180 cool-water indicators (Ingavat & Douglass 1981; Waterhouse 1982; Rao 1988; Fang & Yang 1991) which indicate proximity to the Late Palaeozoic Gondwanaland glaciated region. A Glossopteris flora has also recently been reported south of Baoshan in western Yunnan (Wang & Tan 1994) and typical Lower Permian Gondwanaland spores, including P r i m u s p o l l e n i t e s levius and Schauringipollenites maximus, have been recovered from near Tenchong, western Yunnan (Yang, W. 1994). The palynoflora reported from the Tenchong block (Sibumasu) is also remarkably similar to palynofloras described from the Collie basin of Western Australia (Yang, W. 1994). Gross stratigraphical comparisons between Sibumasu and NW Australia, especially with the Canning Basin (Metcalfe 1992, fig. 3), also show similarities consistent with Sibumasu having been positioned
102
I. METCALFE
40 Sibumasu Block (Ref. at 18N, 95E)
30, LU a 20. E3 }-- 10.
}<
o
._1
O -lo,
4 USTRALI~---A"
LM < -20. ._1 ,c~ -30 n -40
Devonianl CarboniferouslPermianlTriassicl
Jurassic I
Cretaceous
I
-.~
Observations
:::
E. Sumatra E. Malaya
--0-
Predicted from South China
--dr
Predicted from Australia
Tertiary
Fig. 5. Palaeolatitude versus time plots for the Sibumasu Block (after Van der Voo 1993).
outboard of NW Australian Gondwanaland in the Palaeozoic. In addition, Palaeozoic palaeomagnetic data indicate palaeolatitudes (Fig. 5) consistent with a position off NW Australian Gondwanaland in the Devonian, Carboniferous and Early Permian (Bunopas 1982; Bunopas et al. 1989; Wu et al. 1989; Metcalfe 1990, 1994a, b; Huang & Opdyke 1991; Van der Voo 1993).
I n d o c h i n a - E a s t Malaya block This block is bounded to the northeast by the Song Ma and Ailaoshan suture zones and to the west by the Changning-Menglian suture in Yunnan, the Nan-Uttaradit suture in Thailand and the RaubBentong suture in Malaysia. The block probably extends northwards as the Qamdo-Simao or Simao block (Wu 1993; C h e n & Xie 1994) which is regarded as a separate, possibly South Chinaderived, tectonic unit by Wu et al. (1995). The Upper Palaeozoic and Mesozoic faunas and floras are Cathaysian/Tethyan types, have affinities to those of South and North China, and show no relationship to Gondwanaland (Metcalfe 1986, 1988). Lower Palaeozoic faunas are poorly known, and their affinities have yet to be clearly demonstrated. The Precambrian core of the block (Kontum Massif) comprises granulite facies rocks exposed in Vietnam, and it has been suggested that this may have originally formed part of the Gondwanaland granulite belt (Katz 1993). It seems likely that Indochina (including the East Malaya and Simao terranes) had its origin on the northern
Gondwanaland margin but this must remain largely a speculation until we have constraining palaeobiogeographical and palaeomagnetic data.
West Burma block Formerly known as the 'Mount Victoria Land' Block (Mitchell 1989), this terrane is here referred to as the West Burma Block to avoid confusion with Victoria Land in Antarctica. Mitchell (1993) has reinterpreted Western Burma (together with the Woyla Group of Sumatra and the Meratus ophiolite of Borneo as a lower Cretaceous mafic arc, produced by SW-directed subduction, which was then thrust on to the Asian margin in the late Early Cretaceous. Following Metcalfe (1990), the author prefers to regard Western Burma as a probable continental fragment now largely obscured by younger sediments and volcanic rocks. This block has a pre-Mesozoic schist basement overlain by Triassic turbidites and Cretaceous ammonite-bearing shales and limestones in the Indoburman Ranges and by an upper Mesozoic-Cenozoic arc association in the Central Lowlands of Burma. There is as yet no direct evidence for the origin of this block but it sutured to Sibumasu in Cretaceous times. The West Burma Block is considered (Metcalfe 1990) a good candidate for part of the continental sliver that provided a source for sediments derived from the northwest in Timor during the Triassic, and which must have rifted from Gondwanaland in the Late Jurassic. This interpretation however remains speculative.
SE ASIAN TERRANE EVOLUTION
S W B o r n e o a n d S e m i t a u blocks
The origins of these two blocks, separated by the Boyan suture, are poorly constrained. Upper Carboniferous and Lower Permian fusulinid faunas and other Lower Permian microfossils are similar to those of South and North China and do not have affinities with those of Malaya or Thailand (Vachard 1990). Triassic and Jurassic floras and faunas have affinities with South China, Indochina, Japan and the Philippines (Hayami 1984; Kimura 1984; Kobayashi & Tamura 1984). There does not appear to be any Gondwanaland connections in the faunas and floras of these two blocks. A South China/Indochina origin seems likely as proposed by Rammlmair (1993) but reconciliation between substantial progressive counter-clockwise rotation of SW Borneo during the Cenozoic and Cretaceous and younger equatorial palaeolatitudes with the various tectonic models is required. H a i n a n I s l a n d terranes
Hainan Island appears to represent at least two tectonostratigraphic terranes but the nature and boundaries of these are still controversial. Yu (1989) and Yang et al. (1989) recognise two small terranes, the northern Quiongzhong and southern Yaxian terranes bounded by the east-west Jiushuo-Lingshui Fault. Chen et al. (1993) however, interpreted the Island as a Mesozoic orogenic belt created by the collision of the Huannan (SE China) and Dongnanya (SE Asia) blocks. Gt~rtir & Sengrr (1992) interpreted the island as originally forming a part of the Sibumasu terrane, being separated from other Sibumasu components in the Mesozoic by strike-slip faulting. The Lower Palaeozoic faunas of the island indicate links with northeast Gondwanaland (Sun 1963; Lin & Jago 1993) and the presence of possible glacialmarine diamictites and Permian faunas with Gondwanaland affinities suggests a northeast Gondwanaland origin (Yang et al. 1989; Yu 1989). Recent field investigations in Hainan (Metcalfe et al. 1994) have confirmed the probable presence of two terranes on Hainan, but the boundary of these is interpreted to be the SW-NE-oriented Baisha Fault and not the east-west Jiushio-Lingshui Fault. The Late Palaeozoic Gondwanaland origin of these terranes remains problematic. Woyla terranes
Several small continental fragments, including the Sikuleh and Natal blocks, now located along the southwest margin of Sumatra were accreted to 'Sundaland' in the Cretaceous and were called the Woyla terranes by Cameron et al. (1980) and Pulunggono & Cameron (1984). Stratigraphic
103
studies of part of the Woyla terranes in the Padang area (Yancey & Alif 1977) reveal similarities with the stratigraphy of the Exmouth Plateau of the NW Australian shelf (Grriir & Seng6r 1992) and limited palaeomagnetic data from the Sikuleh block (Haile 1979) suggest palaeolatitudes of 26°S for the Late Triassic and 10°S in the Late Mesozoic which is consistent with a NW Australian origin of these terranes. K u r o s e g a w a terrane
The Japanese Islands are composed of a rifted Asian basement and a huge accretionary complex, produced by long-lived westwards subduction of the Pacific and Philippine Sea plates beneath Asia, over which the volcanic arc has migrated eastwards. A number of accreted terranes has been recognized within the accretionary complex including oceanic (seamount) and continental allochthonous terranes. The Kurosegawa con= tinental terrane is a highly attenuated and disrupted composite terrane of predominantly Palaeozoic age now sandwiched between two Mesozoic terranes in the Chichibu Belt of SW Japan. It comprises a Lower Palaeozoic basement of igneous and high-grade metamorphic rocks, Silurian-Devonian volcaniclastic sediments with carbonate blocks, serpentinite, a Upper Palaeozoic-Mesozoic chaotic complex and well-bedded Upper PalaeozoicLower Mesozoic cover sediments (Aitchison et al. 1991). Lower and Middle Palaeozoic faunas and floras have affinities with Australia and suggest a north Australian Gondwanaland margin origin for this exotic terrane (Yoshikura et al. 1990; Aitchison 1993). Studies of chromium spinels from serpentinites of the terrane (Hisada et al. 1994) suggest that the serpentinites were emplaced in a forearc setting during the Devonian. The precise location of the terrane in Silurian-Devonian times is at best speculative but a position adjacent to eastern South China, on the Gondwanaland margin, has been suggested (Saito 1992; Hisada 1994). The precise site of original attachment to Gondwanaland and the drift history of this terrane, however, remain largely unknown.
Rifting and separation of Asian terranes from Gondwanaland It is here proposed that three episodes of rifting occurred on the northern margin of Gondwanaland in the Devonian, Carboniferous to Early Permian and Late Triassic-Late Jurassic. Three continental slivers separated from Gondwanaland in the Late Devonian, late Early Permian and Late Triassic to Late Jurassic times (Fig. 6).
104
I. METCALFE in Australia (Fig. 7) proposed by Chen et al. (1993). Nie (1994) has also suggested that the conspicuous Devonian unconformity of South China, and subsequent Devonian to Triassic passive margin sequences along the southern margin of South China record a Devonian rifting episode and continental separation. In addition, a widespread Late Devonian-early Carboniferous unconformity is recognized on most of the Asian blocks and may represent the breakup unconformity related to continental separation. Carboniferous to Early P e r m i a n rifting
Fig. 6. Schematic diagram showing the three continental slivers/collages of terranes, rifted from Gondwanaland and translated northwards by the opening and closing of three successive oceans, the PalaeoTethys, Meso-Tethys and Ceno-Tethys.
D e v o n i a n rifting
Various Chinese continental blocks (North and South China, Tarim, Indochina, Qaidam) that are postulated to have been located along or close to the northern margin of Gondwanaland in the Early Palaeozoic, outboard of Sibumasu and the Tibetan blocks, were by Carboniferous-Permian times located equatorially, did not have Gondwanalandrelated faunas and floras, and could not have formed part of Gondwanaland. Close proximity or actual attachment of these blocks to Gondwanaland in the Devonian is suggested by the distribution of Devonian vertebrate faunas (Long & Burrett 1989; Young 1990, 1993; Ritchie et al. 1992). A Late Devonian separation of these blocks from Gondwanaland is therefore suggested. Late Devonian clockwise rotation of these blocks away from northern Gondwanaland would be consistent with the contemporaneous counter-clockwise rotation of Gondwanaland about an Euler pole located
There is now extensive evidence of extension and rifting along the northern Gondwanaland margin in the Carboniferous to Early Permian. Evidence of Early Permian rifting in Northwest Australia has long been known (Falvey & Mutter 1981; Powell 1976; Bird 1987). Recent deep seismic studies on the Australian NW shelf have confirmed this and suggest that in the mid-Carboniferous to Early Permian a major episode of extension and crustal thinning took place on the margin of NW Australian Gondwanaland, including the present Australian NW shelf, resulting in the development of the Westralian Superbasin (Colwell et al. 1994). Crustal thinning of the Australian NW shelf at this time was most pronounced in the lower crust and was concentrated northwest of the NW shelf Megashear. Separation of the Sibumasu block (as part of the Cimmerian continent) occurred in the late Early Permian. Similar rifting events have been recorded from Northern Pakistan and Afghanistan (Pogue et al. 1992; Boulin 1988), the North Indian margin (Baud 1994), the Malagassy Rift (Wopfner 1994), Oman (Pillevuit 1993) and Iran (St0cklin 1974). Late Triassic to Late Jurassic rifting
Both Permian and Triassic ages have been previously proposed for the separation of the Lhasa block from Gondwanaland. A Permian separation has been advocated, either as a part of the Cimmerian continent (All~gre et al. 1984; Metcalfe 1988, 1990) or as a 'Mega-Lhasa' block which included Iran and Afghanistan (Baud et al. 1993). Evidence for Permian rifting on the North Indian margin and in Tibet (Baud 1994) is here regarded as being related to the separation of the Cimmerian continental strip which included Iran, Afghanistan and the Qiangtang block of Tibet, but not the Lhasa block. A Triassic rifting and separation of the Lhasa block from Gondwanaland has been proposed by several authors including Chang et al. (1986) and Dewey et al. (1988). Recent sedimentological and stratigraphical studies in the Tibetan Himalayas and
SE ASIAN TERRANE EVOLUTION
105
Fig. 7. Reconstructions showing the rapid counter-clockwise rotation of Gondwanaland in the Late Devonian to Early Carboniferous (after Chen et al. 1993).
Nepal (Liu 1992; Liu & Einsele 1994; Ogg & von Rad 1994; von Rad et al. 1994) have documented the Triassic rifting and Late Triassic (Norian) separation of the Lhasa Block from northern Gondwanaland. This Late Triassic episode of rifting is also recognized along the NW shelf of Australia (Colwell et al. 1994) where it continued into the Late Jurassic resulting in the separation of Western Burma and the Woyla terranes (Metcalfe 1990, 1994a, b; G6rtir & Seng6r 1992).
SE Asian Pre-Cenozoic sutures The formation of present-day SE Asia involved the progressive suturing of terranes to each other during Late Palaeozoic to Cenozoic times and their subsequent disruption, principally caused by the collision of India with Eurasia. The ages of sutures in eastern and SE Asia become younger to the south and southeast. Reconstruction of the history of amalgamation and accretion in SE Asia requires constraints on the ages of the various sutures. Constraining data include: ophiolite obduction
ages; melange ages; ages of 'stitching' plutons; ages of collisional or post-collisional plutons; ages of volcanic arcs; major changes in arc chemistry or isotope chemistry; convergence of apparent polar wander paths (APWPs); loops or disruptions in APWPs; ages of blanketing strata; palaeobiogeography; biostratigraphy (e.g. ages of oceanic sediments now located in the suture); stratigraphy/sedimentology (e.g. provenance stitching); structural geology (eg. age of collisionrelated thrusting). It is rarely possible to acquire all types of constraining data for individual sutures. The various SE Asian sutures are shown in Figs 1 and 2. The region formed by nucleation around the South China block and the Palaeozoic and Mesozoic sutures and constraints on their ages are discussed below from oldest to youngest. Song M a and Song Da sutures
The Indochina and South China blocks are separated by a NW-SE-oriented mobile belt in North Vietnam. Two sutures are currently recognized
106
I. METCALFE
within this belt, the Song Ma and Song Da sutures of Palaeozoic and Mesozoic age, respectively. The age of the collision and amalgamation of Indochina and South China is still controversial with some authors favouring an early amalgamation along the Song Ma suture in the Late Devonian-Early Carboniferous (Gatinsky & Hutchison 1987; Hutchison 1989a, b) and others favouring a Late Triassic amalgamation along the Song Da suture (SengOr et al. 1988). Proponents of an early suturing have proposed that the Early Carboniferous event represents the collision between Indochina and South China along the Song Ma suture to form the 'East Asia Continent' or 'Cathaysialand'. Alternatively, those advocating a Mesozoic suturing regard this event as the collision between an arc fragment and an already accreted arc along a SW-dipping subduction zone and consider the final welding of Indochina to South China as having taken place in the Late Triassic along the Song Da suture. The large-scale folding and thrusting and nappe formation which took place in the Early to middle Carboniferous suggest continent-continent collision rather than an arc-continent collision and recognition of the Song Da Permo-Triassic sequences as probably representing rift basin sequences (Tran 1979; Hutchison 1989a, b), supports the amalgamation of Indochina and South China as occurring in the Early Carboniferous. In addition, middle Carboniferous shallow marine carbonates are reported to blanket the Song Ma suture in North Vietnam and palaeobiogeographical data show that pre-middle Carboniferous faunas on each side of the Song Ma zone are distinctly different whilst the middle Carboniferous faunas are essentially similar (Tran 1992). Newly discovered Carboniferous floras on the Indochina block in Northeast Thailand also indicate continental connection between Indochina and South China in the Carboniferous (Laveine et al. 1994). It is also interesting to note that Middle Devonian fishes recently recorded from central Vietnam (Indochina block) by Janvier et al. (1994) do not include typical South China block forms (including the faunas from North Vietnam). This supports an Early Carboniferous amalgamation of the Indochina and South China blocks. The Song Ma suture is here regarded as the probable zone of welding of the South China and Indochina blocks in the Early Carboniferous. The Song Da zone appears to represent a zone of Permo-Triassic rift basins which closed in the Late Triassic and which was reactivated as a major strike-slip shear zone in the Late Mesozoic to Cenozoic. Much more work is required to unravel the very complex geology and tectonic history of North Vietnam/Laos. Identification of the continuation of the Ailaoshan suture of Yunnan (see below)
in Vietnam or Laos, if this is contiguous with the Song Ma suture and hence diachronous in nature, is a major task for the future.
Ailaoshan suture
The Ailaoshan suture zone (Figs 1 and 4) is a narrow NW-SE orientated belt of ophiolitic rocks and oceanic sediments which forms the boundary between the Simao (Indochina) and South China blocks. The ophiolitic rocks of the zone comprise peridotite, gabbro-diabase and basalt. The peridotite includes weakly depleted plagioclase lherzolite and spinel harzburgite with Sm/Nd ratios of mid-ocean ridge type (Zhang et al. 1984; Wang & Zhong 1991). The ophiolitic rocks are associated with deep-marine sedimentary rocks including ribbon-bedded cherts that have yielded some Lower Carboniferous and Lower Permian radiolarians (Wang & Zhong 1991). Upper Triassic rocks (Norian limestones and Rhaetian sandstones) appear to blanket the suture in this region. The suture trends southeast into North Vietnam but its extension there has not been demonstrated. If the suture is Mesozoic in age it may well correlate with the Song Da suture. If it is Palaeozoic in age then it probably correlates with the Song Ma suture. Another possibility is that it is Mesozoic but correlates with the Song Ma suture, closure of the Palaeo-Tethys being diachronous, earlier in the east and younger in the west. It has also been suggested that the Ailaoshan ophiolitic zone possibly comprises two belts, one of which correlates with the Song Ma suture of Vietnam and one that correlates with the Nan-Uttaradit suture in Thailand (Zhang et al. 1984). Further constraints are required on the nature and age of the Ailaoshan suture in order to resolve these various possibilities.
Jinshajiang suture
This suture, which probably correlates with the Ailaoshan suture to the south (Figs 1 and 4), defines the northeastern limit of the Qamdo-Simao block and forms the boundary between this block and the Kunlun terrane and the Songpan Ganzi accretionary complex terrane. The suture has well developed ophiolites, and melange comprises Devonian, Carboniferous and Permian exotic blocks in a Triassic matrix. The ophiolites are regarded as Upper Permian to Lower Triassic and the age of the suture regarded as Late Triassic by Dewey et al. (1988). An older age for this suture has however been suggested recently by Chen& Xie (1994) and Bian (1991) who recognize Upper Permian to Jurassic sediments unconformably overlying Lower Permian ophiolites in the Hoh Xil Range.
SE ASIAN TERRANE EVOLUTION Lancangjiang suture
This suture forms the boundary between the Qamdo-Simao and Qiangtang terranes and is here regarded as representing a segment of the main Palaeo-Tethys ocean. The suture zone rocks include Devonian and Carboniferous turbiditic 'flysch', ultramafics, ocean-floor basic extrusives of Permian age and Carboniferous-Permian melange. These suture zone rocks are also blanketed by Middle Triassic continental clastics ( C h e n & Xie 1994). Carboniferous-Permian island arc rocks are developed along the west side of the suture (Fan & Zhang 1994) and Upper Triassic collisional granitoids are associated with the suture (Wang & Tan 1994). In Tibet and Qinghai, the suture runs along the Lancangjian Thrust and is covered by huge thicknesses of Jurassic sediments in the Tanggula Range. Scattered ultramafic rocks occur in the Muta region of Tibet and in Qinghai (Chen & Xie 1994). The suture appears to be similar in character and age to the Changning-Menglian suture of Western Yunnan and the Raub-Bentong suture in Peninsular Malaysia here considered to represent the main Palaeo-Tethyan ocean. Banggong suture
This suture forms the boundary between the Tibetan Lhasa and Qiangtang terranes. Well developed ophiolites have been recorded which are associated with radiolarian cherts and the age of the ophiolites and their obduction in Tibet is well constrained as Upper Jurassic (Girardeau et al. 1984; Dewey et al. 1988). The suture is blanketed in Tibet by Cretaceous and Palaeogene rocks and structural data indicate that collision took place along this suture around the Jurassic-Cretaceous boundary (Dewey et al. 1988; Fan & Zhang 1994). The extension of this suture in Yunnan is not well preserved due to shearing but some Triassic 'flysch' and ultramafic rocks are found in the Luxi area and blanketing strata are of Middle Jurassic age indicating an earlier closure of the suture in this region. Nan- Uttaradit suture
The Nan-Uttaradit suture zone in Thailand (Figs 1 and 2) has long been regarded as representing the Palaeo-Tethys destroyed by continent-continent collision between Sibumasu and Indochina. The suture has also been generally regarded as being contiguous with the Raub-Bentong suture in Peninsular Malaysia (see below), and more recently, the Changning-Shuangjiang (ChangningMenglian of this paper) suture of SW China (Barr & MacDonald 1987). The precise age of suturing
107
is still controversial with ages of Lower Carboniferous (Helmke 1985b), Permian (Helmke & Kraikhong 1982; Helmke 1983, 1985a; Helmke & Lindenberg 1983), Early Triassic (Cooper et al. 1989; Mitchell 1989; Metcalfe 1990) and Late Triassic (Ridd 1980; Mitchell 1981; Seng6r 1984) being proposed. This wide range of ages reflected paucity and poor quality of constraining data. The suture includes a belt of pre-Permian ophiolitic mafic and ultramafic rocks (Type II ophiolitic associations) with associated blueschists distributed northeast of Uttaradit (Barr et al. 1985; Barr & Macdonald 1987). Hada et al. (1994) have shown that the suture in southeast Thailand (Sra Kaeo-Chanthaburi) comprises a western chertclastic belt and an eastern serpentinite melange belt. The western chert-clastic sequence appears to form a stack of imbricate thrust slices and is dated as Middle Triassic by radiolarians. The serpentinite melange belt includes a wide variety of rock packages including rocks of oceanic, island arc and continental affinities of various ages. Mafic and ultramafic blocks in the melange comprise oceanisland basalts, backarc basin basalts and andesites, island-arc basalts and andesites and suprasubduction cumulates generated in Carboniferous to Permo-Triassic times (Panjasawatwong & Yaowanoiyothin 1993). Limestone blocks in the melange range from upper Lower Permian to middle Permian and a granitic lens has yielded a zircon U-Pb age of 486 _ 5 Ma (Hada et al. 1994). Structural studies of Carboniferous to Permian meta-greywacke and phyllite, which enclose the mafic-ultramafic sequence, in the Sirikit Dam area (Singharajwarapan & Berry 1993) indicate that these represent an accretionary complex. These rocks are associated with Permo-Triassic volcanic and volcaniclastic rocks of dacitic and rhyolitic composition and a relatively unmetamorphosed Lower Triassic sandstone-shale turbidite sequence. Structures indicate east-directed accretionary thrusting and hence westward directed subduction. The suture zone rocks are overlain unconformably by Jurassic redbeds and post-Triassic basaltic lavas which disconformably overly the suture are intraplate continental basalts (Panjasawatwong & Yaowanoiyothin 1993). A Permo-Triassic age for the suture is therefore indicated. Changning-Menglian
suture
The Changning-Menglian suture in western Yunnan (Fig. 4) is a narrow north-south orientated zone of dismembered ophiolites and associated deep-marine sedimentary rocks that are interpreted as representing a segment of the main PalaeoTethys in East Asia (Huang et al. 1984; Zhang et al. 1985; Wu & Zhang 1987; Liu etal. 1991; Wu 1993;
108
I. METCALFE
I " =0
II
Chan~Belt West Central
~,'~• ~~ .. ~ a . c- ~ - • .~ : ~ . . o. . .
~
- ~."~-.
'~
.-
O' . C L ' ~. C~
~:~
.
East .o
.'~
'
~
..
c:~
,o
-~
.o . . ~ . o' .c . i c~ o . ~' . o
U ~
?
0
f,
121
F~
Beddedchert/ Turbidites hallow-water carbonates
~
Redbeds
~
Volcanics
Bathyal sediments Shallow-water clastics Stratigraphic break
Fig. 8. Devonian to Jurassic stratigraphy of the Changning-Menglian suture zone belt (after Liu et al. 1991)
Fang et al. 1994; Wu et al. 1995). The suture can be traced from Menglian northwards through Laochong, Tongchangia to Changning and ophiolitic melange includes blocks of harzburgite, cumulate websterite, gabbro, basalt, limestone and chert in a mud-silt grade matrix. Associated basalts are of mid-ocean ridge and ocean-island types (Wu et al. 1995) and remnants of limestone capped seamounts have been identified in the zone (Liu et al. 1991). Oceanic ribbon-bedded chert-shale sequences have yielded graptolites, conodonts and radiolarians indicating ages ranging from Lower Devonian to Middle Triassic (Qin et al. 1980; Duan et al. 1982; Wu & Zhang 1987; Wu & Li 1989; Liu et al. 1991; see Fig. 8). Limestone blocks and lenses dominantly found within the basalt sequence of the suture and interpreted as seamount caps,
have yielded fusulinids indicative of Lower Carboniferous to Upper Permian ages (Duan et al. 1982; Wu et al. 1995). The suture is truncated to the north of Changning by the Chongshan metamorphic belt and almost certainly continues northwards as the Lancangjiang suture to the west of Deqin (Fig. 4). The southern extension of the Changning-Menglian suture is problematic. The suture can be traced to the China-Myanmar border but its extension into Myanmar has not been documented. It may be offset by major left lateral strike-slip faults then to continue down into Laos along the Dien Bien Phu line of mafic and ultramafic rocks in Laos and hence to connect with the Nan-Uttaradit suture in Thailand and the RaubBentong suture in Peninsular Malaysia. Alternatively, it has been proposed (Huang et al. 1984) that the Changning-Menglian suture extends southwards as the Chiang Mai volcanic belt in northern Thailand which had been interpreted as the remnants of a subduction-related volcanic arc by Bunopas & Vella (1978). The volcanics of this zone have, however, been shown to have formed in an extensional continental setting, possibly a backarc setting related to westwards subduction along the Nan-Uttaradit zone to the east (Macdonald & Barr 1978; Barr et al. 1990). Interestingly, mafic rocks in the Chiang Rai area may possibly have formed in a subduction environment and these are associated with a chert-clastic sequence known as the 'Fang Chert' which has yielded Lower Devonian graptolites, Carboniferous conodonts and radiolarian faunas of Lower Devonian, Carboniferous and Lower to middle Permian ages. Could this be another small disrupted segment of the main Palaeo-Tethyan suture? Extension of the Changning-Menglian suture to the south, rather than connecting it with the Nan-Uttaradit zone which was considered to terminate against the Ailaoshan suture, led Wu et al. (1995) to propose a 'Simao Terrane' of Cathaysian affinity, bounded by the ChangningMenglian suture to the northwest, the Ailaoshan suture to the northeast and the Uttaradit-Nan suture to the southeast (Fig. 9). The similarity in both nature and age of the Nan-Uttaradit and Changning-Menglian sutures lead me to favour their correlation and to here regard the 'Simao Terrane' to be disrupted part or parts of the Qamdo-Simao block which is regarded as an extension of the Indochina terrane. Raub-Bentong
suture
The Raub-Bentong suture of Peninsular Malaysia forms the boundary between the Sibumasu and Indochina/East Malaya terrane and represents the Palaeo-Tethys. The suture zone is exposed as a
109
SE ASIAN TERRANE EVOLUTION
30N ,' a .~ •
,'
j
Pi
*
"/ " '
CHINA
t ?
t
OKunming
D O
fi "~ Pe*, t" ".-I
BURMA ,' : :
,,
te,
Chiang
AND ...
- ....
"LADS .,...'.-
~
..'
,
.~CAMBODIA!
110E I
Fig. 9. Map showing the Simao terrane proposed by Wu et al. (1995) narrow north-south-trending zone extending from Thailand through Raub and Bentong to the east of Malacca. Its extension to the south and southeast is controversial (Fig. 10) but a southwards (disrupted)
extension into Sumatra as suggested by Tjia (1989) and Hutchison (1993) is here favoured. The suture is exposed as a c. 13 km wide zone in Peninsular Malaysia and comprises melange, oceanic sediments (including ribbon-bedded cherts), schist, and discontinuous, narrow, elongate bodies of serpentinized mafic-ultramafic rocks that have been interpreted as representing ophiolite (Tjia 1987, 1989; Hutchison 1989b). The melange has a mud/silt matrix and contains a variety of clasts including ribbon-bedded cherts, limestones, volcanic and volcaniclastic rocks. Basic subduction-related volcanic clasts (which occur commonly in the Nan-Uttaradit suture further north) have not however been reported from the Raub-Bentong suture. The precise age of suturing is still poorly constrained but a latest Permianearliest Triassic suturing age is here favoured. Harbury et al. (1990) presented structural evidence which seems to preclude a Late Triassic collision. Ages of limestone clasts in the melange include Lower and Upper Permian (Metcalfe 1989) which indicate a latest Permian or Triassic age for the melange matrix. The Main Range 'collisional' 'S' Type granites of peninsular Malaysia range from Upper Triassic (230 __+9 Ma) to lowermost Jurassic (207_+ 14 Ma) in age, with a peak of around 210Ma (Liew & Page 1985; Darbyshire 1988). Ages of oceanic deep-marine bedded cherts within the suture zone range from Upper Devonian to Upper Permian (Metcalfe 1992; Spiller & Metcalfe 1993, 1995; Metcalfe & Spiller 1994). These data support the existence o f Palaeo-Tethys between Sibumasu and Indochina/East Malaya from Late Devonian to Late Permian times and suggest collision of Sibumasu with Indochina in the latest Permian or Triassic.
Qinling-Dabie
Fig. 10. Map showing the southwards extension of the Raub-Bentong suture as proposed by various authors.
suture
This suture forms the boundary between the North and South China blocks. Estimates of the age of this suture, based on geological data, range from the Middle Palaeozoic (Zhang et al. 1984; Mattauer et al. 1985, 1991; Wang 1991) to the Permo-Triassic (Wang et al. 1982; Klimetz 1983; Hsu et al. 1987; Seng/Jr 1987; Yin & Nie 1993; Nie & Rowley 1994). Palaeomagnetic data (Lin & Fuller 1990) suggest that the amalgamation of North and South China along the Qin Ling suture had occurred by the Late Triassic-Early Jurassic. These data comprise convergence of APWPs, rapid rotation of the North China block in the Late Triassic-Early Jurassic, convergence of palaeolatitudes and a Jurassic-Cretaceous loop (stagnation point) in the APWPs. The data also seem to be consistent with stratigraphic data and ages of collision-related
110
~. METCALFE
plutons in the Qinling foldbelt (Hsu et al. 1987). Enkin et al. (1992) utilizing palaeomagnetic data, proposed that the suturing could be as late as Jurassic, a view challenged by Nie & Rowley (1994) who cogently argue for a Permo-Triassic collision of South and North China. A PermoTriassic suturing age is here favoured with closure of the suture commencing in the Late Permian in the east, and progressing westwards with development of the Tan-Lu transform fault, into the Middle Triassic.
Shan boundary
Woyla suture
This suture comprises Cretaceous ophiolites and accretionary complex material along the southwest margin of Sumatra and marks the zone of accretion and collision of the small Woyla terranes with Sibumasu in the Late Cretaceous (Wajzer et al. 1991).
Meratus
suture
This suture, also known as the Sagaing or Mandalay suture, forms the boundary between the West Burma and Sibumasu terranes. An Early Cretaceous age for the suture is indicated by Cretaceous thrusts in the backarc belt (Mitchell 1990) and the Upper Cretaceous age of the Western Belt tin-bearing granites.
suture
This suture, located in southeast Borneo, is characterized by subduction melange and ophiolite of middle Cretaceous age (Hamilton 1979) which is overlain by Eocene strata. The ophiolite was obducted in Cenomanian times during an arccontinent collision (Sikumbang 1986).
Boyan suture
The Boyan suture forms the boundary between the small Semitau terrane and the South West Borneo
50N 40N 30N 20N 1ON
10s 2os 3os 4os 50S 1 -~
21-1
15@
16 ©
3 X
17 O
4A
5+
18 ~ )
6I
19 [ ]
7®
20 [ ]
8 ©
9,&
21 O
10 •
22 X
11 ~
12'~
23 o.*o 24 +
13~
14~
25'~
26 •
Fig. 11. Palaeolatitude plots for the Tarim, North China, South China, Indochina, East Malaya and Sibumasu terranes from Devonian/Carboniferousto Tertiary. Sources of data are 1. Lin (1987a, b); 2. Lin et al. (1985); 3. McElhinney et al. (1981); 4. Opdyke et al. (1986); 5. Sasajima & Maenaka (1987); 6. Chan et al. (1984); 7. Achache et al. (1983); 8. Maranate & Vella (1986); 9. Achache & Courtillot (1985); 10. Bunopas et al. (1989); 11. McElhinney et al. (1974); 12. E. A. Schmidtke & M. Fuller (pers. comm. 1994); 13. Haile & Khoo (1980); 14. Haile et al. (1983); 15. Bunopas (1982); 16. Sasajima et al. (1978); 17. Halle (1979); 18. Lin (1987b); 19. Zhao & Coe (1987); 20. Cheng et al. (1988); 21. Bai et al. (1987); 22. Li et al. (1988); 23. McFadden et al. (1988); 24. Zhai et al. (1988); 25. Huang & Opdyke (1991); 26. Enkin et al. (1992). Modified from Metcalfe (1993, 1994c).
SE ASIAN TERRANE EVOLUTION block. The suture zone rocks are Upper Cretaceous melange (Williams et al. 1986; Williams & Harahap 1987) produced during short-lived SWdirected subduction which ceased after collision of the Semitau terrane and then recommenced behind this accreted terrane to produce the Lubok Antu melange and the Sarawak accretionary complex.
111
Terrane evolution Available data suggest that all the east and SE Asian allochthonous terranes formed part of, or were close to, the northern margin of G o n d w a n a l a n d in the Early Palaeozoic. The continental blocks that are placed adjacent to the
PALAEO-PAClFIC LAURENTIA _ SIBERIA
AUSTRALIA CAMBRO-ORDOVICIAN ,~ (TREMADOC)
ANTARCTICA \,
) |!!i!!!!!!~!!i
INDIA t
30S
30N
LATE D E V O N I A N
1
Fig. 12. Reconstructions of eastern Gondwanaland for the Cambro-Ordovician and Late Devonian, showing the postulated positions of the East and SE Asian terranes. NC, North China; SC, South China; T, Tarim; I, Indochina; Q, Qaidam; WC, Western Cimmerian Continent; Qi, Qiangtang; L, Lhasa; S, Sibumasu; WB, West Burma; GI, Greater India; SA, South America. Present day outlines are for reference only.
112
I. METCALFE
Greater India-NW Australian Gondwanaland margin in the Early Palaeozoic include North & South China, Indochina/East Malaya (which includes the Qamdo-Simao block of western China), Tarim (here regarded to include the Kunlun and Ala Shan blocks), Qaidam, Sibumasu, Qiangtang, Lhasa, NW & SE Hainan, West Burma, Kurosegawa and Woyla terranes. Palaeomagnetic and biogeographic evidence indicates that the North and South China, Tarim, Indochina/East Malaya and Qaidam terranes were in equatorial or low northern palaeolatitudes by Carboniferous times (Fig. 11) and their faunas had no affinities
with those of Gondwanaland. These terranes must therefore be placed outboard of Sibumasu, Qiangtang and Lhasa on the Gondwanaland margin in the Early Palaeozoic and must have separated from Gondwanaland in the Devonian (Fig. 12). The separation of these terranes is interpreted to have occurred in the Late Devonian when mainland Gondwanaland rotated rapidly counter-clockwise about an Euler pole in Australia (Chen et al. 1993), and the Palaeo-Tethys opened between the terranes and northern Gondwanaland (Fig. 7). By Early Carboniferous times, South China and Indochina had amalgamated along the Song Ma
EARLY CARBONIFEROUS (VISI~AN)
EARLY PERMIAN
~ ~
LATE PERMIAN
~ 7, • ~
~ ~0
LATE TRIASSIC
wBL
Fig. 13. Palaeogeographicreconstructions of the Tethyanregion for Early Carboniferous,Early Permian, Late Permian and Late Triassic. Present day outlines are for reference only. Symbols as for Fig. 12.
3o
SE ASIANTERRANEEVOLUTION suture to form 'Cathaysialand'. However, a narrow ocean between the Qamdo-Simao portion of the Indochina terrane and South China (now represented by the Ailaoshan suture) persisted until Triassic times (Fig. 13). Palaeomagnetic data indicate that South China was equatorial during the Early Carboniferous (Lin et al. 1985; Lin 1987a; see Fig. 11). No palaeomagnetic data are available for the Carboniferous of Indochina but Middle and Late Carboniferous and younger faunas and floras have strong affinities with South China. Palaeomagnetic and palaeobiogeographic data for the North China block suggest that it was not attached to either South China or Siberia in the Carboniferous and that it lay in low northern palaeolatitudes (Lin 1987a; Lin & Fuller 1990; Nie 1991; Fig. 11). In the Mid-Late Carboniferous, Gondwanaland rotated clockwise, collided with Laurentia to form Pangaea, and eastern Gondwanaland moved from low to high southem palaeolatitudes. The predicted palaeolatitude for Sibumasu, if placed adjacent to NW Australia, would be about 40°S in the Late Carboniferous and recent palaeomagnetic results from the Yunnan part of Sibumasu give a palaeolatitude of 43°S for the Late Carboniferous (Figs 5 and 11) which is consistent with the proposed reconstruction (Fig. 13). Palaeomagnetic data for the Tarim block indicate that it moved from southern to northern palaeolatitudes in the Devonian and Carboniferous (Fig. 11) and that since the Permian, observed latitudes concur with those predicted from Eurasia indicating that it had already sutured to Eurasia (Nie & Rowley 1994; Fig. 14). Data to constrain the palaeoposition of the Qaidam terrane are largely lacking for this time. A major rifting phase occurred on the margin of NE Gondwanaland in the Early Permian (Powell 1976; Falvey & Mutter 1981; Bird 1987; AudleyCharles 1988; Pillevuit et al. 1993; Pogue et al. 1992; Baud 1994; Wopfner 1994) indicating that a substantial continental fragment or fragments separated from Gondwanaland at that time. Audley-Charles (1988) suggested that these terranes were Iran, North Tibet (Qiangtang) and Indochina but later (Audley-Charles 1991) supported the view presented by the present author (Metcalfe 1986, 1988, 1990, 1991, 1992, 1993, 1994a, b) that the rifting terranes were in fact Sibumasu, and Qiangtang (along with other terranes constituting the Cimmerian continent of SengOr). Palaeomagnetic data indicate that Sibumasu travelled rapidly from southern to northern palaeolatitudes in the Permo-Triassic (Figs 5 and 11) supporting this view. Post-Middle Permian faunas of Sibumasu are Tethyan in aspect and show affinities with Indochina and South China and do not contain any Gondwanaland elements (Metcalfe
113
80
TARIM
70
Predicted
from
-
60 50 40 30
~ t t ObSerVed
20
1 i 1
10 0
Irving & Irving (1982) Van der Voo (1990) Besse & Courtillot (1991) Enkin et al. (1992)
---
-10 ~. -20
I
I
I
I
I
I
I
I
I
R
N
E
K2
K1
J3
J2
J1
T3
I
I
T2 T1
I
P2
Fig. 14. Palaeolatitude versus time plots for centre of the Tarim block. Solid black circles are calculated from measured data on the Tarim block as reported by Enkin et al. (1992). Other plots are the palaeolatitudes predicted from the Eurasian reference poles of Irving & Irving (1982), Van der Voo (1990) and Besse & Courtillot (1991), from Nie & Rowley (1994).
1988). Recent work on Permian brachiopod palaeobiogeography (Shi & Archbold 1993, 1994a, b; Shi et al. 1995) has shown that terranes of Sibumasu and other Cimmerian continent components formed part of a Gondwanaland province in the early Early Permian, but by the late Early Permian there was a separate Cimmerian province indicating separation of Cimmeria from Gondwanaland. Upper Permian faunas belong to the equatorial Cathaysian province. The rifting and separation of the Cimmerian continent from Gondwanaland was effected by opening of the Meso-Tethys and northwards destruction of PalaeoTethys beneath Eurasia and the North and South China and Indochina blocks. Continental palaeomagnetic data show that Cathaysialand, characterized by its palaeoequatorial Cathaysian flora, remained in low northern palaeolatitudes during the Permian (Fig. 11). Sibumasu sutured to the Qamdo/ Simao/Indochina/East Malaya portion of Cathaysialand along the Changning-Menglian, Nan-Uttaradit and Raub-Bentong sutures in Permo-Triassic times. The precise age of suturing is still controversial and may well have been diachronous but a latest Permian-Triassic suturing age is here favoured (see above). Thus, by Late Triassic times, Sibumasu had amalgamated with Cathaysialand and this amalgamated superterrane was located in low northern latitudes (Fig. 13). Initial contact between South and North China probably also occurred in the late Permian with
114
i. METCALFE
collision occurring first in the east with development of the Tan-Lu transform fault and final suturing along the Qinling-Dabei suture occurring by Late Triassic times (Nie & Rowley 1994). The presence of coesite- and diamond-bearing ultrahigh pressure metamorphic rocks in the Dabei and Sulu regions (Wang et al. 1992; Okay e t al. 1993; Maruyama et al. 1994) which were probably exhumed during the Triassic (Nie et al. 1995)
indicate that the Qinling-Dabei orogenic system produced vast quantities of sediment during the Triassic to the Songpan Ganzi basin to the west (Zhou & Graham 1993; Nie e t al. 1995). It is not yet known if the small Gondwanaland terranes of Hainan Island, which yield Lower Palaeozoic trilobites and other invertebrates endemic to Australia and Gondwanaland and which have Lower Permian diamictites interpreted by
UINE~.
60~.
I
iili ! ~ :i~:~• PACIFIC OCEAN
c. s ~. . . . '. . . . .:
I
-60
{
c) :
:;::::::
: i
..'~. -
:
L ATE
C R E T A C E O U S . . .'" ...r,.
Fig. 15. Palaeogeographic reconstructions for Eastern Tethys in (a) Late Jurassic, (b) Early Cretaceous and (c) Late Cretaceous times. SG, Songpan Gangzi accretionary complex; SWB, South West Borneo; SE, Semitau; Si, Sikuleh; N, Natal; M, Mangkalihat; WS, West Sulawesi; Ba, Banda Allochthon; ES, East Sulawesi; O, Obi-Bacan; Ba-Su, Bangai-Sula; Bu, Buton; B-S, Buru-Seram; WIJ, West Irian Jaya; Sm, Sumba; PA, Philippine Arc. M numbers represent Indian Ocean magnetic anomalies. Other terrane symbols as in Figs 12 and 13. Modified from Metcalfe (1990) and partly after Smith et al. (1981), AudleyCharles (1988) and Audley-Charles et al. (1988). Present day outlines are for reference only.
SE ASIAN TERRANE EVOLUTION some workers as glacial-marine, were accreted to South China in the Palaeozoic or Mesozoic. The Late Triassic to Late Jurassic saw renewed rifting on the northeast margin of Gondwanaland and initial rifting of the third continental sliver from G o n d w a n a l a n d opening the Ceno-Tethys. The Lhasa terrane separated in the Late Triassic, followed by the West Burma block and Woyla terranes in the Late Jurassic. The Lhasa block moved rapidly northwards in the Late Jurassic to Early Cretaceous, separated from the West Burma block and Woyla terranes by a transform fault (Fig. 15). The Lhasa block sutured to Eurasia in the Early Cretaceous and by Late Cretaceous times, the
115
West Burma block and the Woyla terranes had accreted to proto-SE Asia (Fig. 15). Continued work on the geological evolution of East and SE Asia is currently supported by an Australian Research Council Grant which is gratefully acknowledged. I also thank the Department of Geology & Geophysics, University of New England and the Department of Geology, Australian National University for facilities provided. Thanks go to M. G. Audley-Charles and R. W. Murphy for constructive reviews of the manuscript. This paper forms a contribution to IGCP Project 321 and GSGP Project Pangaea. It was written whilst a Visiting Fellow in the Department of Geology, Australian National University, Canberra.
References ACHACHE, J. & COURTILLOT, V. 1985. A preliminary Upper Triassic palaeomagnetic pole for the Khorat plateau (Thailand): consequences for the accretion of Indochina against Eurasia. Earth and Planetary Science Letters, 73, 147-157. - - , - - . 8z BESSE,J. 1983. Palaeomagnetic constraints on the Late Cretaceous and Cenozoic tectonics of Southeastern Asia. Earth and Planetary Science Letters, 63, 123-136. AHRENOT, H., CHONGLAKMANI, C., HANSEN, B. T. & HELMKE, D. 1993. Geochronological cross section through northern Thailand. Journal of Southeast Asian Earth Sciences, 8, 207-217. --, HANSEN, B. T., LUMJUAN, A., MICKEIN, A. & WEMMER, K. 1994. Tertiary d6collement tectonics in Western Thailand and its implications on stratigraphy and paleogeographic reconstructions. In: ANGSUWATHANA, P., WONGWANICH, T., TANSATHIAN,W., WONGSOMSAK,S. & TULYATID,J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand, 291. AITCHISON, J. C. 1993. Allochthonous Silurian volcaniclastic rocks of the Kurosegawa terrane: Longdispersed early Gondwana fragments. In Third International Symposium of 1GCP 321 Gondwana dispersion and Asian accretion, Abstracts of Papers. 15-17. --, HADA, S. & YOSmKURA, S. 1991. Kurosegawa terrane: disrupted remnants of a low latitude Paleozoic terrane accreted to SW Japan. Journal of Southeast Asian Earth Sciences, 6, 83-92. ALLI~GRE,C. J. ETAL. 1984. Structure and evolution of the Himalaya-Tibet orogenic belt. Nature, 307, 17-22. ARCHBOLD,N. W. & SHI, G. R. 1995. Permian brachiopod faunas of western Australia: Gondwanan-Asia relationships and Permian climate. Journal of Southeast Asian Earth Sciences, 11,207-215. , PIGRAM, C. J, RATMAN, N. & HAKIM, S. 1982. Indonesian Permian brachiopod fauna and Gondwana-South-East Asia relationships. Nature, 296, 556-558. AUDLEY-CHARLES,M. G. 1983. Reconstruction of eastern Gondwanaland. Nature, 306, 48-50. -1984. Cold Gondwana, warm Tethys and the Tibetan Lhasa block. Nature, 310, 165.
1988. Evolution of the southern margin of Tethys (North Australian region) from Early Permian to Late Cretaceous. In: AUDLEY-CHARLES,M. G. 8z HALLAM, A. (eds) Gondwana and Tethys. Geological Society Special Publication, 37, 79-100. 1991. Tectonics of the New Guinea area. Annual Reviews Earth and Planetary Sciences, 19, 17-41. - - - , BALLANTYNE,P. D. & HALL, R. 1988. MesozoicCenozoic rift-drift history of Asian fragments from Gondwanaland. Tectonophysics, 155, 317-330. BAI, Y., CHEN, G., SUN, Q., SuN, Y., LI, Y., DONG, Y. & StJN, D. 1987. Late Paleozoic polar wander path for the Tadm platform and its tectonic significance. Tectonophysics, 139, 145-153. BARR, S. M. & MACDONALD,A. S. 1987. Nan River suture zone, northern Thailand. Geology, 15, 907-910. - - - , MACDONALD, A. S., YAOWANOIYOTHIN, W. & PANJASAWATWONG, Y. 1985. Occurrence of blueschist in the Nan River mafic-ultramafic belt, northern Thailand. Warta Geologi, 11, 47-50. --, TANTISUKRIT, C., YAOWANOIYOTHIN, W. MACDONALD, A. S. 1990. Petrology and tectonic implications of Upper Paleozoic volcanic rocks of the Chiang Mai belt, northern Thailand. Journal of Southeast Asian Earth Sciences, 4, 37-47. BAUD,A. 1994. Late Permian sequence stratigraphy of the N Indian margin. In: International Symposium on Permian Stratigraphy, Environments & Resources, Guiyang, China, Abstracts. 1. --, MARCOUX, J., GUIRAUD, R., RICOU, L. E. & GAETANI, M. 1993. Late Murgabian (266 to 264 Ma). In: DERCOURT,J., RIcou, L. E. & VRIELYNCK, B. (eds) Atlas Tethys Palaeoenvironmental Maps. Explanatory Notes. Gauthier-Villars, Pads, 9-20. BESSE, J. & COURTmLOT,V. 1991. Revised and synthetic apparent polar wander paths of the African, Eurasian, North American and Indian Plates, and true polar wander since 200Ma. Journal of Geophysical Research, 96, 4020-4050. BrAN QIANTAO. 1991. On the tectonic characteristics and evolution of the Hoh Xil region. In: REN, J. & Xm, G. (eds) Proceedings of First International Symposium on Gondwana Dispersion and Asian Accretion - Geological Evolution of Eastern Tethys. China University of Geosciences, Beijing, 305-306. -
-
-
-
116
I. METCALFE
BIRD, E 1987. The geology of the Permo-Triassic rocks of Kekneno, west Timor. PhD Thesis, University of London. BOULIN, J. 1988. Hercynian and Eocimmerian events in Afghanistan and adjoining regions. Tectonophysics, 148, 253-278. BUNOPAS, S. 1982. Palaeogeographic history of Western Thailand and adjacent parts of Southeast Asia - a plate tectonics interpretation. Geological Survey paper, 5, Department of Mineral Resources, Thailand. -& VELLA, E 1978. Late Paleozoic and Mesozoic structural evolution of Thailand. In: NUTALAYA,P. (ed.) Proceedings 3rd Regional Conference on Geology and Mineral Resources of Southeast Asia. Bangkok, Thailand. Asian Institute of Technology, 133-140. , MARANTE, S. & VELLA, P. 1989. Palaeozoic and Early Mesozoic rotation and drifting of Shan-Thai from Gondwana-Australia. ln:4th International Symposium on pre-Jurassic evolution of East Asia, IGCP Project 224, Reports and Abstracts. 1, 63-64. BURRETT, C. & STAIT, B. 1985. South-East Asia as part of an Ordovician Gondwanaland. Earth and Planetary Science Letters, 75, 184-190. , LONG, J. & STAIT, B. 1990. Early-Middle Palaeozoic biogeography of Asian terranes derived from Gondwana. In: MCKERROW,W. S. & SCOTESE, C. R. (eds) Palaeozoic Palaeogeography and Biogeography. Geological Society, London, Memoir, 12, 163-174. CAMERON, N. R., CLARKE, N. C. G., ALDISS, D. T., ASPDEN, T. A. & DJUNUDDIN, A. 1980. The geological evolution of northern Sumatra. Ninth Indonesian Petroleum Convention, Jakarta. CHANG CHENGFAET AL. 1986. Preliminary conclusions of the Royal Society and Academia Sinica 1985 Geotraverse of Tibet. Nature, 323, 501-507. CHAN, L. S., WANG, C. Y. & Wu, X. Y. 1984. Palaeomagnetic results from some Permian-Triass rocks from Southwestern China. Geophysical Research Letters, 11, 1157-1160. CHEN BINGWEI& XIE GUANGLIAN. 1994. Evolution of the Tethys in Yunnan and Tibet. Journal of Southeast Asian Earth Sciences, 9, 349-354. CHEN HAIHONG,SUN SHU, LI JILIANG,HSU, K. J., HELLER, E, HAAG, M. & DOBSON, J. 1991. Early Triassic palaeomagnetism and tectonics of South China. In: Seventh Regional Conference on Geology, Mineral and Energy Resources of Southeast Asia (GEOSEA VII), Bangkok. Abstract Volume. Geological Society of Thailand, 24. CHEN, J., FOLAND, K. A., XING, E, Xu, X. 8~ ZHOU, T. 1991. Magmatism along the southeast margin of the Yangtze block: Precambrian collision of the Yangtze and Cathysia blocks of China. Geology, 19, 815-818. CHEN, Z., LI, Z. X., POWELL, C. McA. & BALME, B. E. 1993. Palaeomagnetism of the Brewer Conglomerate in central Australia & fast movement of Gondwanaland during the Late Devonian. Geophysical Journal International, 115, 564-574. CHENG, G., BAI, Y. & SUN, Y. 1988. Palaeomagnetic study on the tectonic evolution of the Ordos block, North
China [in Chinese with English abstract]. Seismological. Geology, 10, 81-87. COLWELL, J. B., STAGG, H. M. J., SYMONDS, P. A., WlLLCOX, J. B. & O'BRIEN, G. W. (compilers for AGSO North West Shelf Study Group). 1994. Deep reflections on the North West Shelf: changing perceptions of basin formation. In: PURCELL,R G. & PURCELL, R. R. (eds) The Sedimentary Basins of Western Australia. Petroleum Exploration Society of Australia, 63-76. COOPER, M. A., HERBERT, R. & HmL, G. S. 1989. The structural evolution of Triassic Intermontane Basins in Northeast Thailand. In: THAYASUTmPITAK,T. & OUNCrtANUM, P. (eds) Proceedings of International Symposium on Intermontane Basins: Geology and Resources. Chiang Mai University, Thailand, 231-242, DARBYSHIRE, D. P. E 1988. Geochronology of Malaysian granites. Natural Environment Research Council, Isotope Geology Centre, Open File Report 88/3. DEWEY, J. E, SHACKLETON, R. M., CHANG CHENGFA SUN YIYIN. 1988. The tectonic evolution of the Tibetan Plateau. Philosophical Transactions of the Royal Society of London, A327, 379-413. DUAN YANXUE, XIAO YINWEN, HONG Yu, LIo YILAI & ZHAO ZHANGMIN. 1982. 1:200,000 scale geological map (Menglian) and technical report [in Chinese]. ENKIN, R. J., YANG,Z., CHEN, Y. & COURTILLOT,V. 1992. Paleomagnetic constraints on the geodynamic history of the major blocks of China from the Permian to the Present. Journal of Geophysical Research, 97, 13 953-13 989. FALVEY, D. A. & MUTTER, J. C. 1981. Regional plate tectonics and the evolution of Australia's passive continental margins. BMR Journal of Australian Geology & Geophysics, 6, 1-29. FAN, C. & ZHANG,Y. 1994. The structure and tectonics of western Yunnan. Journal of Southeast Asian Earth Sciences, 9, 355-361. FANG, N. & YANG, W. 1991. A study of the oxygen and carbon isotope records from Upper Carboniferous to Lower Permian in Western Yunnan, China. In: REN, J. & XIE, G. (eds) Proceedings of First International Symposium on Gondwana Dispersion and Asian Accretion Geological Evolution of Eastern Tethys. China University of Geosciences, Beijing, 35-36. --, VAN DER Woo, R. & LIANG, Q. Z. 1990. Ordovician palaeomagnetism of eastern Yunnan, China. Geophysical Research Letters, 17, 953-956. --, LIU, B., FENG., Q. & JIA, J. 1994. Late Palaeozoic and Triassic deep-water deposits and tectonic evolution of the Palaeotethys in the Changning-Menglian and Lacangjiang belts, southwestern Yunnan. Journal of Southeast Asian Earth Sciences, 9, 363-374. GATINSKY, Y. G. & HUTCHISON, C. S. 1987. Cathaysia, Gondwanaland and the Palaeotethys in the evolution of continental Southeast Asia. Geological Society of Malaysia Bulletin, 20, 179-199. GIRARDEAU,J., MARCOUX,J. ALLt~GRE,C. J., BASSOULLET, J. P., TANG YOUKING, XIAO XUCHANG, ZAO YOUGONG & WANG XmlN. 1984. Tectonic environment and geodynamic significance of the
SE ASIAN TERRANE EVOLUTION Neo-Cimmerian Dongqiao ophiolite, BanggongNujiang suture zone, Tibet. Nature, 307, 27-31. GORI3R, N. & SENGOR, A. M. C. 1992. Palaeogeography and tectonic evolution of the eastern Tethysides: Implications for the northwest Australian margin breakup history. In: yoN RAD, U., HAQ, B. U., et al. Proceedings of the Ocean Drilling Program, Scientific Results, 122, 83-106. HADA, S., BUNOPAS,S., THITISAWAN,V., PASAJAWATWONG, Y., YANOIYOTHIN,W., ISHII, K. & YOSHIKURA, S. 1994. Accretion tectonics in Thailand along the Nan-Uttaradit Suture Zone. AAPG Bulletin, 78, 1143. HAILE, N. S. 1979. Palaeomagnetic evidence for rotation and northward drift of Sumatra. Journal of the Geological Society, London, 136, 541-545. & KHOO, H. P. 1980. Palaeomagnetic measurements on Upper Jurassic and Lower Cretaceous sedimentary rocks from Peninsular Malaysia. Bulletin of the Geological Society of Malaysia, 12, 75-78. , BECKINSALE, R. D., CHAKRABORTY,K. R., ABDUL, H. H. & TJAHJO, H. 1983. Palaeomagnetism, geochronology and petrology of the dolerite dykes and basaltic lavas from Kuantan, West Malaysia. Bulletin of the Geological Society of Malaysia, 16, 71-85. HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078. HANSEN, B. T., AHRENDT, H., LUMJUAN,A., MICKEIN, A. ~; WEMMER, K. 1994. Do the high grade metamorphic rocks in Thailand represent the basement for the Paleozoic strata? In: ANGSUWATHANA,P., WONGWANICH, T., TANSATHIAN,W., WONGSOMSAK, S. & TULYATID, J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand, 55. HARBURY,N. A., JONES, M. E., A UDLEY-CHARLES,M. G., METCALFE,'I. & MOHAMAD,K. R. 1990. Structural evolution of Mesozoic Peninsular Malaysia. Journal of the Geological Society, London, 147, -
-
1 1 - 2 6 .
HAYAMI, I. 1984. Jurassic marine bivalve faunas and biogeography in Southeast Asia. Geology and Palaeontology of Southeast Asia, 25, 229-237. HELMKE, D. 1983. On the Variscan evolution of central mainland Southeast Asia. Earth Evolution Science, 4, 309-319. 1985a. The Permo-Triassic 'Paleotethys' in mainland Southeast-Asia and adjacent parts of China. Geologische Rundschau, 74, 215-228. 1985b. The orogenic evolution (Permo-Triassic) of central Thailand. Implications on palaeogeographic models for mainland SE-Asia. Mdmoires de la Socidtg Gdologique de France, 147, 83-91. & I@,AIKHONG, C. 1982. On the geosynclinal and orogenic evolution of central and northeastern Thailand. Journal of the Geological Society of Thailand, 5, 52-74. & LINDENBERG, H. G. 1983. New data on the 'Indosinian' orogeny from Central Thailand. Geologische Rundschau, 72, 317-328. HISADA, K., ARAI, S. ~; NEGORO, A. 1994. Devonian serpentinite protrusion confirmed by detrital chromian spinels in outer zone od SW Japan. In: -
-
-
-
-
-
117
ANGSUWATHANA,P., WONGWANICH,T., TANSATHIAN, W., WONGSOMSAK, S. 8¢ TULYATID, J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand, 76-80. HOFFMAN, P. E 1991. Did the breakout of Laurentia turn G ondwanaland inside out? Science 252, 1409-1412. Hsu, K. J., SuN, S. & LI, J. 1989. Mesozoic suturing in the Huanan Alps and the tectonic assembly of South China. In: SENGOR, A. M. C. (ed.) Tectonic Evolution of the Tethyan Region. Kluwer Academic Publishers, 551-565. --, WANG, Q., L1, J., ZHOU, D. & SUN, S. 1987. Tectonic evolution of Qinting Mountains, China. Eclogae Geologie Helveticae, 71,611-635. --., L], J., CHEN, H., WANG, Q., SUN, S. SENGOR,A. M. C. 1990. Tectonics of South China: Key to understanding West Pacific geology. Tectonophysics, 183, 9-39. --, SUN, S., LL J., CHEN, H., PEN, H. & SENGOR, A. M. C. 1988. Mesozoic overthrust tectonics in South China. Geology, 16, 418-421. HUANG, J., CHEN, G. & CHEN, B. 1984. Preliminary analysis of the Tethys-Himalayan tectonic domain [in Chinese with English abstract]. Acta Geologica Sinica, 58, 1-17. HUANG, K. & OPDYKE,N. D. 1991. Paleomagnetic results from the Upper Carboniferous of the Shan-ThaiMalay block of western Yunnan, China. Tectonophysics, 192, 333-344. HUTCmSON, C. S. 1975. Ophiolite in Southeast Asia. Geological Society of America Bulletin, 86, 797-806. 1983. Multiple Mesozoic Sn-W-Sb granitoids of Southeast Asia. In: RODDICK, J. A. (ed.) CircumPacific Plutonic Terranes. Geological Society of America Memoir, 159, 35-60. 1989a. The Palaeo-Tethyan realm and Indosinian orogenic system of Southeast Asia. In: SENGOR, A. M. C. (ed.) Tectonic evolution of the Tethyan region. Kluwer Academic Publishers, 585-643. -1989b. Geological Evolution of South-East Asia. Clarendon Press, Oxford. 1993. Gondwana and Cathaysian blocks, Palaeotethys sutures and Cenozoic tectonics in South-east Asia. Geologische Rundschau, 82, 388-405. INGAVAT,R. & DOUGLASS,R. 1981. Fusuline fossils from Thailand, Part XIV. The fusulinid genus Monodiexodina from Northwest Thailand. Geology and Palaeontology of Southeast Asia, 22, 23-34. IRVING, E. & IRVING, G. A. 1982. Apparent polar wander paths for Carboniferous through Cenozoic and the assembly of Gondwana. Geolophyscial Surveys, 5, 141-188. JANVIER, P., THANH TONG-DZUY & TRUONGDOAN NHAT. 1994. Devonian fishes from Vietnam: New data from central Vietnam and their paleobiogeographical significance. In: ANGSLIWATHANA, P., WONGWANICH, T., TANSATHIAN,W., WONGSOMSAK, S. & TULYATID, J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand, 62-68.
-
-
118
JIN, X. 1994a. Extent and timing of the PermoCarboniferous glacio-marine deposits bearing units in southwestern China. In: CHO, M. & K ~ , J. H. (eds) IGCP 321 Gondwana Dispersion and Asian Accretion Fourth International Symposium and Field Excursion. Abstract Volume. 43-47. 1994b. Sedimenmtary and paleogeographic significance of Permo-Carboniferous sequences in Western Yunnan, China. Geologisches Institut der Universit~it zu KGln, Sonderveroeffentlichungen, 99. KATZ, M. B. 1993. The Kannack complex of the Vietnam Kontum Massif of the Indochina Block - An exotic fragment of Precambrian Gondwanaland? In: FINDLAY, R. H., UNRUG, R., BANKS, M. R. & VEEVERS, J. J. (eds) Gondwana 8 - Assembly, Evolution & Dispersal. (Proceedings Eighth Gondwana Symposium, Hobart, 1991). A. A. Balkema, Rotterdam, 161-164. KIMURA, T. 1984. Mesozoic floras of East and Southeast Asia, with a short note on the Cenozoic floras of Southeast Asia and China. Geology and Palaeontology of Southeast Asia, 25, 325-350. KLrMETZ, M. P. 1983. Speculations on the Mesozoic plate tectonic evolution of eastern China. Tectonics, 2, 139-166. 1987. The Mesozoic Tectonostratigraphic Terranes and Accretionary Heritage of South-Eastern Mainland Asia. In: LEITCH, E. C. & SCHEIBNER,E. (eds) Terrane Accretion and Orogenic Belts. American Geophysical Union Geodynamics Series, 19, 221-234. KOBAYASHI,T. & TAMURA,M. 1984. The Triassic Bivalvia of Malaysia, Thailand and adjacent areas. Geology and Palaeontology of Southeast Asia, 25, 201-228. LAVEINE, J. P., RATANASTHIEN,B. & SITHIRACH, S. 1994. The Carboniferous flora of Na Duang coal mine, Northeastern Thailand, Its paleogeographic interest. In: ANGSUWATHANA,P., WONGWANICH,Y., TANSATHIAN, W., WONGSOMSAK,S. & TULYATID,J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand. 83-90. LI, Y., McWILLIAMS, M., Cox, A., SHARPS, R., Ll, Y., ET AL. 1988. Late Permian palaeomagnetic pole from dikes of the Tarim craton, China. Geology, 16, 275-278. LI ZH1MING. 1994. Ordovician. In: YIN HONGFU(ed.) The Palaeobiogeography of China. Clarendon Press, Oxford, 64-87. LI, Z. X., ZHANG, L. & POWELL, C. McA. 1994. Were the East Asian continental blocks part of Rodinia? 12th Australian Geological Convention, Perth, 1994. Geological Society of Australia, Abstracts, 37, 248-249. Lmw, T. C. & PAGE, R. W. 1985. U-Pb Zircon dating of granitoid plutons from the West Coast Province of Peninsular Malaysia. Journal of the Geological Society, London, 142, 515-526. LIN, J. 1987a. Late Carboniferous paleogeographic reconstruction. In: llth International Congress of Carboniferous Stratigraphy and Geology. Abstracts. 1, 283. 1987b. The apparent polar wander path for the South China block and its geological significance [in -
-
i. METCALFE
-
-
Chinese with English abstract]. Scientia Geologica Sinica, 4, 306-315. -& FULLER,M. 1990. Palaeomagnetism, North China and South China collision. & the Tan-Lu fault. Philosophical Transactions of the Royal Society of London, A331, 589-598. , -& ZHANG, W. 1985. Preliminary Phanerozoic polar wander paths for the North and South China Blocks. Nature, 313, 444--449. LIN TIAN-RUI & JAGO, J. B. 1993. Xystridura and other early Middle Cambrian trilobites from Yaxian, Hainan Province, China. Transactions of the Royal Society of South Australia, 117, 141-152. LIu BENPEI, FENG QINGLAI & FANG NIANQIAO. 1991. Tectonic evolution of the Palaeo-Tethys in Changning-Menglian Belt and adjacent regions, western Yunnan. Journal of China University of Geosciences, 2, 18-28. LIU GUANGHUA. 1992. Permian to Eocene sediments and Indian passive margin evolution in the Tibetan Himalayas. Tubinger Geowissenschaftliche Arbeiten, Reihe A, 13. & EINSELE, G. 1994. Sedimentary history of the Tethyan basin in the Tibetan Himalayas. Geologische Rundschau, 83, 32-61. LONG, J. & BURRETT, C. 1989. Fish from the Upper Devonian of the Shan-Thai terrane indicate proximity to east Gondwana and south China terranes. Geology, 17, 811-813. MACDONALD, A. S. & BARR, S. M. 1978. Tectonic significance of a Late Carboniferous volcanic arc in northern Thailand. In: NUTALAYA, P. (ed.) Proceedings 3rd Regional Conference on Geology and Mineral Resources of Southeast Asia. Asian Institute of Technology, Bangkok, Thailand, 151-156. MCELHINNY, M. W., HALLE, N. S. & CRAWFORD, A. R. 1974. Palaeomagnetic evidence shows Malay Peninsula was not part of Gondwanaland. Nature, 252, 641-645. --, EMBLETON, B. J. J., MA, X. H. & ZHANG, Z. K. 1981. Fragmentation of Asia in the Permian. Nature, 293, 212-216. MCFADDEN, P, L., MA, X. H., MCELHINNY, M. W. & ZHANG, Z. K. 1988. Permo-Triassic magnetostratigraphy in China: northern Tarim. Earth and Planetary Science Letters, 87, 152-160. MARANATE, S. • VELLA, P. 1986. Paleomagnetism of the Khorat Group, Mesozoic, Northeast Thailand. Journal of Southeast Asian Earth Sciences, 1, 23-31. MARUYAMA, S., LIOU, J. G. & ZHANG, R. 1994. Tectonic evolution of the ultrahigh-pressure (UHP) and highpressure (HP) metamorphic belts from central China. The lslandArc, 3, 112-121. MATTAUER, M., MATTE, P., MALUSKI, H., Xu, Z., ZHANG, Q. W. & WANG, Y. M. 1991. Paleozoic and triassic plate boundary between North and South China; new structural and radiometric data on the Dabie Shan (eastern China). Comptes rendus de l'Academieies Sciences de Paris, Serie 2, 312, 1227-1223. - - , MALAVIEILLE,J., TAPPONIER,P., MALUSKI, H., Xu, Z., Lu, Y. & TANG, Y. 1985. Tectonics of the Qinling Belts, build-up and evolution of eastern Asia. Nature, 327, 496-500.
SE ASIAN TERRANE EVOLUTION METCALFE, I. 1984. Stratigraphy, palaeontology and palaeogeography of the Carboniferous of Southeast Asia. Mdmoires de la Socidtd Gdologique de France, 147, 107-118. 1986. Late Palaeozoic palaeogeography of Southeast Asia: some stratigraphical, palaeontological and palaeomagnetic constraints. Geological Society of Malaysia Bulletin, 19, 153-164. 1988. Origin and assembly of Southeast Asian continental terranes. In: AUDLEY-CHARLES,M. G. & HALLAM, A. (eds) Gondwana and Tethys. Geological Society, London, Special Publication, 37, 101-118. 1989. Triassic sedimentation in the Central Basin of Peninsular Malaysia. In: THANASUTHIPITAK, T. & OUNCHANUM, P. (eds) Proceedings of International Symposium on Intermontane Basins: Geology and Resources. Chiang Mai University, Thailand, 173-186. 1990. Allochthonous terrane processes in Southeast Asia. Philosophical Transactions of the Royal Society of London, A331, 625-640. 1991. Late Palaeozoic and Mesozoic palaeogeography of Southeast Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 211-221. 1992. Ordovician to Permian evolution of Southeast Asian terranes: NW Australian Gondwana connections. In: WEBBY, B. D. & LAURIE, J. R. (eds) Global perspectives on Ordovician geology. (Proceedings Sixth International Symposium on the Ordovician System.). A. A. Balkema, Rotterdam, 293-3O5. 1993. Southeast Asian terranes: Gondwanaland origins and evolution. In: FINDLAY,R. H., UNRUG, R., BANKS, M. R. & VEEVERS,J. J. (eds) Gondwana 8 - Assembly, Evolution & Dispersal (Proceedings Eighth Gondwana Symposium, Hobart, 1991.) A. A. Balkema, Rotterdam, 181-200. 1994a. Gondwanaland origin, dispersion and accretion of East and Southeast Asian continental terranes. Journal of South American Earth Sciences, 7, 333-347. 1994b. Palaeomagnetic research in Southeast Asia: progress, problems and prospects. Exploration Geophysics, 24, 277-282. 1994c. Late Palaeozoic and Mesozoic Palaeogeography of Eastern Pangea and Tethys. In: EMBRY, A. F., BEAUCHAMP, B. & GLASS, D. J. Pangea: Global Environments and Resources. Canadian Society of Petroleum Geologists, Memoir, 17, 97-111. & SPILLER,E C. P. 1994. Correlation of the Permian and Triassic in Peninsular Malaysia: New data from conodont and radiolarian studies. In: ANGSUWATHANA,P., WONGWANICH,T., TANSATHIAN, W., WONGSOMSAK, S. • TULYATID, J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand, 129. , SHERGOLD, J. H. & LI, Z. X. 1994. IGCP 321 Gondwana dispersion and Asian accretion: fieIdwork on Hainan Island. Episodes, 16, 443-447. MITCHELL, A. H. G. 1981. Phanerozoic plate boundaries in mainland Southeast Asia, the Himalayas and
-
-
-
-
-
-
-
-
-
-
119
Tibet. Journal of the Geological Society of London, 138, 109-122. 1989. The Shan Plateau and Western Burma: Mesozoic-Cenozoic plate boundaries and correlation with Tibet. In: SENGOR, A. M. C. (ed.) Tectonic Evolution of the Tethyan Region. Kluwer Academic Publishers, 567-583. 1990. Mesozoic and Cenozoic regional tectonics and metallogenesis in Mainland SE Asia. Geological Society of Malaysia Bulletin, 20, 221-239. 1992. Late Permian-Mesozoic events and the Mergui Group Nappe in Myanmar and Thailand. Journal of Southeast Asian Earth Sciences, 7, 165-178. 1993. Cretaceous-Cenozoic tectonic events in the western Myanmar (Burma)-Assam region. Journal of the Geological Society of London, 150, 1089-1102. Nm, Y. S. 1991. Paleoclimatic and paleomagnetic constraints on the Paleozoic reconstructions of south China, north China and Tarim. Tectonophysics, 196, 279-308. 1994. Devonian rifting of South China from Gondwana - a case study. In: 12th Australian Geological Convention, Perth, 1994. Geological Society of Australia, Abstracts, 37, 319. & ROWLEY, D. B. 1994. Comment on 'Paleomagnetic constraints on the geodynamic history of the major blocks of China from the Permian to the Present' by R. J. ENKINet al. Journal of Geo14hysical Research, 99, 18 035-18 042. --, & ZmGLER, A. M. 1990. Constraints on the locations of Asian microcontinents in Palaeo-Tethys during the Late Palaeozoic. In: McKE~ow, W. S. & SCOTESE, C. R. (eds) Palaeozoic Palaeogeography and Biogeography. Geological Society, London, Memoir, 12, 397-408. , Y~, A., ROWLEY,D. B. & JrN, Y. 1995. Exhumation of the Dabie Shan ultra-high pressure rocks and accumulation of the Songpan-Ganzi flyusch sequence, central China. Geology, 22, 999-1002. OGG, J. G. & VON RAD, U. 1994. The Triassic of the Thakkhola (Nepal). II: Paleolatitudes and comparison with other Eastern Tethyan margins of Gondwana. Geologische Rundschau, 83, 107-129. OKAY, A. I., SENGOR, A. M. C. & SATIn, M. 1993. Tectonics of an ultrahigh pressure metamorphic terrane: The Dabie Shan/Tongbai Shan orogen, China. Tectonics, 12, 1320-1334. OPDYKE, N. D., HUANG, K., XU, G., ZHANG, W. Y. & KENT, D. V. 1986. Paleomagnetic results from the Triassic of the Yangtze Platform. Journal of Geophysical Research, 91, 9553-9568. PANJASAWATWONG, Y. & YAOWANO[YOTHIN, W. 1993. Petrochemical study of post-Triassic basalts from the Nan Suture, northern Thailand. Journal of Southeast Asian Earth Sciences, 8, 147-158. PILLEVUIT, A. 1993. Les Blocs Exotiques du Sultanat d'Oman: Evolution paldogdographique d'une marge passive flexurale. Mdmoires de Grologie (Lausanne), 17. POGUE, K. R., DIPIETRO, J. A., SAID RAHIM KHAN, HUGHES, S. S., DILLES, J. H. & LAWRENCE, R. D. 1992. Late Paleozoic rifting in Northern Pakistan. Tectonics, 11, 871-883. -
-
-
-
120
i. METCALFE
POWELL, D. E. 1976. The geological evolution and hydrocarbon potential of the continental margin off north-west Australia. Journal of the Australian Petroleum Exploration Association, 16, 13-23. PULUNGGONO, A. & CAMERON, N. R. 1984. Sumatran microplates, their characteristics and their role in the evolution of the Central and South Sumatra basins. In: Proceedings of the Indonesian Petroleum Association 13th Annual Convention. 121-143. QIN DEHOU, HE CHANGXIANG,YANGJIAWEN,LI HANSONG, CHEN KUNHONG, ET AL. 1980. 1:200,000 scale geological map (Baoshan) and technical report [in Chinese]. RACEY, A., SMITH, A. B. & DAWSON, O. 1994. Permian echinoderms from Peninsular Thailand. In: ANGSUWATHANA,P., WONGWANICH,T., TANSATHIAN, W., WONGSOMSAK, S. 8z TULYATID, J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Dept. Min. Res. Bangkok, Thailand, 106-114. RAO, C. P. 1988. Paleoclimate of some Permo-Triassic carbonates of Malaysia. Sedimentary Geology, 60, 163-171. RAMMLMAIR, D. 1993. The evolution of the Philippine archipelago in time and space: a plate-tectonic model. Geologisches Jahrbuch, 81, 1-48. RIDD, M. E 1980. Possible Palaeozoic drift of S. E. Asia and Triassic collision with China. Journal of the Geological Society, London, 137, 635-640. RITCHIE, A., WANG SHITAO, YOUNG, G. C. & ZHANG GUORUI. 1992. The Sinolepidae, a Family of Antiarchs (Placoderm Fishes) from the Devonian of South China and Eastern Australia. Records of the Australian Museum, 44, 319-370. ROGERS, J. 1989. Comment on 'Mesozoic overthrust tectonics in South China'. Geology, 17, 671-672. ROWLEY, D. B., ZIEGLER, A. M. & GYOU, N. 1989. Comment on 'Mesozoic overthrust tectonics in South China'. Geology, 17, 384-386. SAITO, Y. 1992. Reading geologic history of Japanese Islands. Iwanami-shoten, Tokyo [in Japanese with English abstract]. SASAJIMA, S. & MAENAKA, K. 1987. A paleomagnetic aspect on the assemblage of East Asian fragmented continents. IGCP Project 224, Pre-Jurassic geological evolution of eastern continental margin of Asia, Report, 2, 139-150. , OTOFUJI, Y., HIROOKA, K. SUPARKA,SUWIJANTOt~ HEHUWAT, F. 1978. Palaeomagnetic studies on Sumatra Island and the possibility of Sumatra being part of Gondwanaland. Rock Magnetism and Paleogeophysics, 5, 104-110. SASHIDA, K., IGO, H., HISADA, K., NAKORNSRI, N. & AMPORNMAHA, A. 1993. Occurrence of Paleozoic and Early Mesozoic radiolaria in Thailand (preliminary report). Journal of Southeast Asian Earth Sciences, 8, 97-108. SENGOR, A. M. C. 1979. Mid-Mesozoic closure of PermoTriassic Tethys and its implications. Nature, 279, 590-593. 1984. The Cimmeride orogenic system and the tectonics of Eurasia. Geological Society of America Special Paper, 195. 1987. Tectonic subdivisions and evolution of Asia.
-
-
Bulletin of the Technical University of lstanbul, 40, 355-435. 1989. The Tethyside Orogenic System: An introduction. In: SENGOR, A. M. C. (ed.) Tectonic Evolution of the Tethyan Region. Kluwer Academic Publishers, 1-22. - - - , ALTINER,D., CIN, A., USTAOMER,T. & Hsu, K. J. 1988. Origin and assembly of the Tethyside orogenic collage at the expense of Gondwana Land. In: AUDLEY-CHARLES, M. G. (~ HALLAM, A. Gondwana and Tethys. Geological Society, London, Special Publication, 37, 119-181. SHI, G. R. & ARCHBOLD, N. W. 1993. Distribution of Asselian to Tastubian (Early Permian) CircumPacific brachiopod faunas. Memoir of the association of Australasian Palaeontologists, 15, 343-351. , 1995a. A quantitative analysis on the distribution of Baigendzhinian-Early Kungurian (Early Permian) brachiopod faunas in the western Pacific region. Journal of Southeast Asian Earth Sciences, 11, 189-206. --. 1995b. Permian brachiopod faunal sequence of the Shan-Thai terrane: biostratigraphy, palaeobiogeographical affinities and plate tectonic/palaeoclimatic implications. Journal of Southeast Asian Earth Sciences, 11, 177-188. t~ WATERHOUSE, J. B. 1991. Early Permian brachiopods from Perak, west Malaysia. Journal of Southeast Asian Earth Sciences, 6, 25-39. - - - , ARCHBOLD,N. W. & ZHAN, L. P. 1995. Distribution and characteristics of mixed (transitional) mid-Permian (Late Artinskian-Ufimian) marine faunas in Asia and their palaeobiogeographical implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 114, 241-271. S1KUMBANG, N. 1986. Geology and Tectonics of PreTertiary rocks in the Metalis Mountains, South-east Kalimantan, Indonesia. PhD thesis, University of London. SINGHARAJWARAPAN,S. • BERRY, R. E 1993. Structural analysis of the accretionary complex in Sirikit Dam area, uttaradit, Northern Thailand. Journal of Southeast Asian Earth Sciences, 8, 233-245. SMITH, A. G., HURLEY, A. M. & BRIDEN, J. C. 1981. Phanerozoic palaeocontinental world maps. Cambridge University Press. SPILLER, E C. P. & METCALFE, I. 1993. Late Palaeozoic radiolarians from the Bentong-Raub suture zone, Peninsular Malaysia - initial findings. In: Third International Symposium of IGCP 321 Gondwana dispersion and Asian accretion. Abstracts of Papers. 67-69. & -1995. Late Palaeozoic radiolarians from the Bentong-Raub suture zone and Semanggol Formation of Peninsular Malaysia: -Initial Results. Journal of Southeast Asian Earth Sciences, 11, 217-224. STAUFFER, P. H. 1983. Unravelling the mosaic of Palaeozoic crustal blocks in Southeast Asia. Geologische Rundschau, 72, 1061-1080. & LEE, C. E 1989. Late Palaeozoic glacial marine facies in Southeast Asia and its implications. Geological Society of Malaysia Bulletin, 20, 363-397. ,
-
-
-
-
SE ASIAN TERRANE EVOLUTION & MANTAJIT, J. 1981. Late Palaeozoic tilloids of Malaya, Thailand and Burma. In: HAMBREY,M. J. & HARLAND, W. H. (eds) Earth's pre-Pleistocene glacial record. Cambridge University Press, 331-337. STOCKLIN,J. 1974. Possible ancient continental margins in Iran. ln: BURK, C. A. & DRAKE, C. L. (eds) The Geology of continental Margins. Springer-Verlag, Berlin, 873-887. SUN YUN-CHU 1963. On the occurrence of Xystridura fauna from Middle Cambrian of Hainan Island and its significance. Acta Palaeontoloica Sinica, 11, 608-610. TJ1A, H. D. 1987. The Bentong Suture. In: SITUMONANG, B. (ed.) Proceedings Regional Conference on Mineral Hyd. Resources of SE Asia. 73-85. 1989. Tectonic history of the Bentong-Bengkalis Suture. Geologi Indonesia, 12, 89-111. TRAN, V. T. 1979. Geology of Vietnam (North Part). General Geological Department, Hanoi, Vietnam. 1992. Tectonic zonation of Indochina Peninsula. In: 29th International Geological Congress, Kyoto, Japan. Abstracts. 1, 276. VACHARD, D. 1990. A new biozonation of the limestones from Terbat area, Sarawak, Malaysia. CCOP Technical Bulletin, 20, 183-208. VAN DER VOO, R. 1990. Phanerozoic paleomagnetic poles from Europe and North America and comparisons with continental reconstructions. Reviews of Geophysics, 28, 167-206. -1993. Paleomagnetism of the Atlantic, Tethys and Iapetus oceans. Cambridge University Press. VON RAD, U., DI]RR, S. B., OGG, J. G. & WIEDMANN, J. 1994. The Triassic of the Thakkhola (Nepal). In: Stratigraphy and paleoenvironment of a north-east Gondwana rifted margin. Geologische Rundschau, 83, 76-106. WAJZER, M. R., BARBER,A. J., HIDAYAT,S. & SUHARSONO. 1991. Accretion, collision and strike-slip faulting: The Woyla Group as a key to the tectonic evolution of North Sumatra. Journal of Southeast Asian Earth Sciences, 6, 447-461. WANG, H. Z., XU, C. Y. & ZHOU, Z. G. 1982. Tectonic development of the continental margins on both sides of the Qinling marine realm [in Chinese with English abstract]. Acta Geologica. Sinica, 63, 270-280. WANG, X. M., LIOU, J. G. & MARUYAMA, S. 1992. Coesite-bearing eclogites from the dabie Mountains, central China: Petrogenesis, P-T paths and implications for regional tectonics. Journal of Geology, 100, 231-252. WANG YI & ZHON6 DALAI. 1991. Indo-Sinian deformation in Sanjiang area, Western Yunnan Province, China. In: Proceedings First International Symposium on Gondwana dispersion and Asian accretion. China University of Geosciences, 232-237. WANG ZONGQI. 1991. The origins and their tectonic implications of Late Proterozoic and Paleozoic conglomerates in the northern Qinling Mountains. In: REN JISHUN & XIE GUANGLIAN (eds) Proceedings First International Symposium on Gondwana Dispersion and Asian accretion - geological evolution of Eastern Tethys, Kunming, China. China University of Geosciences, Beijing, 249-250. -
-
-
-
-
-
121
WANG ZUGUAN & TAN, X. 1994. Palaeozoic structural evolution of Yunnan. Journal of Southeast Asian Earth Sciences, 9, 345-348. WATERHOUSE, J. B. 1982. An early Permian cool-water fauna from pebbly mudstones in South Thailand. Geological Magazine, 119, 337-354. WILLIAMS, P. R. & HARAHAP, B. H. 1987. Preliminary geochemical and age data from postsubduction intrusive rocks, northwest Borneo. Australian Journal of Earth Sciences, 34, 405--415. --, JOHNSON, C. R., ALMOND, R. A. & SIMAMORA,W. H. 1986. Late Cretaceous to Ealy Tertiary structural elements of West Kalimantan. Tectonophysics, 148, 279-297. WOPFNER, H. 1994. Late Palaeozoic climates between Gondwana and Western Yunnan. In: CHO, M. & KIM, J. H. (eds) IGCP 321 Gondwana Dispersion and Asian Accretion. Fourth International Symposium and Field Excursion. Abstract Volume. 127-131. Wu, E, VAN DER VOO, R. & QIZHONG LIANG. 1989. Devonian palaeomagnetism of Yunnan province accross the Shan Thai-South China suture. Tectonics, 8, 939-952. Wu GENYAO. 1993. Late Paleozoic tectonic framework and Paleotethyan evolution in Western Yunnan. Scientia Geologica Sinica, 2, 129-149. Wu HAORUO& LI HONGSHENG. 1989. Carboniferous and Permian radiolaria in the Menglian area, western Yunnan [in Chinese with English abstract]. Acta Micropalaeontologica Sinica, 6, 337-343. -& ZHANG QI. 1987. Carboniferous and Permian readiolarites of Western Yunnan - relict of the Paleo-Tethys. Compte Rendu, 3, 90-96. - - - , BOULTER, C. A., KE BAOJIA, STOW, D. A. V. & WANG ZHONGCHENG. 1995. The ChangningMenglian suture zone; A segment of the major Cathaysian-Gondwana divide in Southeast Asia. Tectonophysics, 242, 267-280. YANCEY, T. E. & AHF, A. 1977. Upper Mesozoic strata near Padang, West Sumatra. Geological Society of Malaysia Bulletin, 8, 61-74. YANG JIALU. 1994. Cambrian. In: YIN HONGFU (ed) The Palaeobiogeography of China. Clarendon Press, Oxford, 35-63. YANG SHUFENG, YU ZI-YE, Guo LtNGZHI & SHI YANGSHEN. 1989. The division and palaeomagnetism of the Hainan Island and plate tectonic significance [in Chinese with English abstract]. Journal of Nanjing University (Earth Sciences Edition), 1, 38-46. YANG WEI-PING. 1994. Lower Permian palynological studies of the Tenchong Block in Western Yunnan, China. In: International Syposium on Permian Stratigraphy, Environments & Resources, Guiyang, China. Abstracts. 44-45. Ym, A. & NIL, S. Y. 1993. An indentation model for the North and South China collision and the development of the Tan-Lu and Honam Fault systems, eastern Asia. Tectonics, 12, 801813. YOSHIKURA, S., HADA, S. & ISOZAKI, Y. 1990. Kurosegawa terrane. In: ICHIKAWA, K., MIZUTANI, S., HARA, I., HAOA, S. & YAO, A. (eds) Pre-
122
i. METCALFE
Cretaceous Terranes in Japan. IGCP Project, 224 Publication, Osaka, 185-201. YOUNG, G. C. 1990. Devonian vertebrate distribution patterns and cladistic analysis of palaeogeographic hypotheses. In: MCKERROW, W. S. & SCOTESE, C. R. (eds) Palaeozoic Palaeogeography and Biogeography. Geological Society, London, Memoir, 12, 243-255. 1993. Vertebrate faunal provinces in the Middle Palaeozoic. In: LONG, J. A. (ed.) Palaeozoic vertebrate biostratigraphy and biogeography. Belhaven Press, London, 293-323. Yu ZI-YE. 1989. Hainan Island - A terrane from Gondwana. In: 4th International Symposium on pre-Jurassic evolution of East Asia. IGCP Project 224 Reports and Abstracts. 1, 11-12. ZHAI, Y. J., ZHANG,Z. K., LI, Y. A., LI, Q., LI, Y. P., ETAL. 1988. A study of Upper Carboniferous palaeo-
magnetism for the Tarim block. Geosciences, 2, 43-56 [in Chinese with English abstract]. ZHANG, Q., LI, D. & ZHANC, K. 1985. Preliminary study on Tongchangjia ophiolite melange from Yun County, Yunnan Province [in Chinese with English abstract]. Acta Petrographica Sinica, 1, 1-14. ZHANG, Z. M., LIOU, J. G. & COLEMAN,R. G. 1984. An outline of plate tectonics of China. Geological Society of America Bulletin, 95, 295-312. ZHAO, X. t~£ COE, R. S. 1987. Palaeomagnetic constraints on the collision and rotation of North and South China. Nature, 327, 141-142. ZHOU, D. & GRAHAM, S. A. 1993. Songpan-Ganzi Triassic flysch complex as a remnant ocean basin along diachronous Qinling collisional orogen, central China. Geological Society of America, Abstracts with Programs, 25, 118.
Cenozoic SE Asia: reconstructing its aggregation and reorganization GORDON
PACKHAM
Department of Geology and Geophysics, University of Sydney, NSW, Australia 2006 Abstract: Reconstructing SE Asia in the Cenozoic is dependent on being able to model three major tectonic events and their consequences. They are the collision of India with Eurasia, the rotational history of the Philippine Sea plate and the ongoing collision of Australia with eastern Indonesia. Models of the Eocene India-Eurasia collision imply extrusion along major strike-slip faults or crustal thickening and block rotation. Kinematic modelling of the present day tectonics of central Asia indicates rotations and strike-slip faulting and an eastward displacement of crust. Prior to middle Miocene time tectonic activity was closely associated with the collision zone and extrusion of Indochina was supposed to have occurred. Analysis of the volume of the topography formed by the collision indicates eastward displacement and the loss of some crust by expulsion or into the mantle. Palaeomagnetic measurements have found northward displacement of Tibet and strong clockwise rotation east of the eastern syntaxis but unequivocal evidence of southward displacement is rare. Large Palaeogene to early Neogene displacements on strike-slip faults have been documented and related with confidence to the rotations that formed at least the Oligocene South China Sea floor. Early Oligocene docking of the Indoburman Flysch belt south of the eastern syntaxis is difficult to explain in extrusion models. Whether the rotations that have been documented are a response by internal deformation or block rotation is unresolved, as is the amount of extrusion of SE Asia to the south. In the eastern region, rotation and associated northward motion of the Philippine Sea plate as it developed puts constraints on possible models because of a maintained close proximity of its southern margin to the northern edge of the Australian craton. Assembling of continental fragments now in the Banda Sea and derived from eastern New Guinea appears to have occurred by Late Cretaceous time. Formation of the Banda arc may have isolated the Banda Sea from the Australia plate then collision of the Australian margin possibly rotated the Banda Sea counter-clockwise. Although westward movement of the Philippine Sea plate relative to Australia commenced in the early Miocene, convergent leftlateral shear was not possible until the late Miocene thus limiting westward tectonic transport of terranes and the transmission of the shear into the region west of the Banda Sea.
This paper reviews a n u m b e r of the tectonic arguments which have been used to arrive at SE Asian reconstructions. Ideally, reconstructions should be based on the rigorous application of kinematic parameters. While major plate motions provide the large scale k i n e m a t i c f r a m e w o r k for deriving reconstructions, simple rigid plate solutions have limitations because of the uncertainties of the locations of plate boundaries, non-rigid deformation and the progressive reordering of elements within the complex. Two major but contrasted C e n o z o i c collision events have f a s h i o n e d the region. Namely, the collision between the cratonic masses of India and Eurasia Asia in the west and the collision o f arcs and small continental blocks derived from Australia together with the Australian craton in the east and southeast. Accordingly, this review falls into two parts, one dealing with the India-Asia collision and the formation of the South China Sea where the emphasis is biased strongly towards Palaeogene tectonic events and the second with the P h i l i p p i n e - B a n d a - N e w Guinea region
where the focus of interest is on the Neogene. As well as reconstructions covering the whole region, other studies have focused on specific problem areas. The four major regional subsets are: (a) the India-Eurasia collision and its possible displacem e n t of the constituent blocks of Sundaland; (b) development of the South China Sea; (c) motion of the Philippine Sea plate and the Philippine Islands; and (d) the d e v e l o p m e n t of the Banda Sea region and northern New Guinea and arc collisions with the Australian craton.
The India-Asia collision and western SE Asia While there is general agreement that the collision has indented the margin of Eurasia and that strain has been absorbed in substantial crustal thickening, there is controversy about the amount and distribution of the remaining strain and its effect on SE Asian tectonics. In their well k n o w n extrusion
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 123-152.
123
124
G. PACKHAM model, Tapponnier and his co-workers argued that during the collision, continental blocks including Sundaland were extruded possibly some 1000 km southeast from the path of the Indian craton (e.g. Peltzer & Tapponnier 1988). In contrast Dewey et al. (1989) considered that residual strain was confined to extension of the Tibetan Plateau east of the eastern Himalayan syntaxis, accompanied by oroclinal bending and block rotation in SE Asia and west of the western Himalayan syntaxis (Fig. 1). Two of the major ingredients of kinematic tectonic solutions have been (a) estimations of rates of slip on active faults and convergence in zones of shortening, and (b) reconstructions, based on forward modelling of the collision of the Indian craton with Eurasia. This discussion reviews the present day tectonics of Asia and then goes on to examine kinematic models, the sequence of tectonic events, the contribution of palaeomagnetic studies, the implications of South China Sea spreading and some geological relationships along the western margin of SE Asia.
Fig. 1. Rotation model of the India-Asia collision from Dewey et al. (1989). Zones of sinistral transpression (oblique lines), zone of dextral shear (vertical lines), maximum area of extrusion of Tibet (horizontal lines), subduction accretion since 45 Ma (dotted) and motion of India with respect to Eurasia in cm a-~ (arrows).
Late
Cenozoic
(,0=0.86
A simplified kinematic model by Avouac & Tapponnier (1993) for central Asia (Fig. 2) which applies for the last ten or so million years, was based on estimates of slip-rates on faults from field
IO0*E '
~,, 2,, ~II:$1~I"IIA
+ 0.1~k~\
.+:::+.............
KAZAKHSTAN'. N , TILM)I/TariM+":::! ....... ~..
-~
\~'~
+t
%
;
\%'~
All m o v e m l w ~ t s in r a m s -1
+ ~
....
.....................~:~i~#:~!+~i~i~ii~i!i~.':~.:~:~::: ,.... ~ ® ,
-
-..:. - .... .:.:.~ ..... :..:.:.x.:.:~:.:..:.:r...~;~ ~ .~.~.............++:-.~.....:.:.:-"
~i~:~:~:.~;:~!~.-.::~:.,.~:.. - r l ~ ; ~ + . ; ~ : ~ : . ~ . ~
40ON
.......+:+,..~.~¢~.~'.:~-~;+~
~
++++++++++:.
-~!+:~.,..,,.~:::,:.,..,~.~. ................,+ + + + + + : , , + ~ .
~o +
N .+~+~+ ++'++: . + - . - .N -:++:.+.~+. '...~:. ..... . . . . . . . . . . . . . . . . . . . ......
i ::
,.
..,-.--
7
-+
/'/
60 = 0.65 _+ 0.30
~
~ ~ ~-' ~\ ' +
~-.
\
~ct
+: ~ o,,.
~. "t,,s+ z
+:m ~,,,+.'~
~
-.:::~:+.L+:+:+~..:~::.-.:+r~::::.+
........................... ~
,%~
.+++ ~ ~
.'~ "b.
~ '
+
""~.~
\ \
\
,8,-,- r
,Et --<~
I~!+i+t Z o n e s #
~
~.:.~...-~/
~,~"
++
il.
,l.ll,
,
_~
/_~.
. .+, +. ++++ / #
:-:.,~ +:.+.:.: ~ .+:
O U TH .,+S....
m
300N
, i
%+ s /
-+-'
:;,x
~:::!+~]~i;~i~
+X,,~ ~ I
z~::~++++~_++++~!+:~[l""~;:'+LN,.D. "~+:: O _ -
~ +
INDIA ol a c t i v e s h o r t e n i n 9 /
40*
•.-:-:.:.:.:-:-:-:-:+>:....
•
.,---
+
,
, . S . . ' - "--,
: ~ . ~
'::+~L.+~;--~::+@:':-:::m::::::::'~ :::::::::::::~;k++~:~.:+:+:+:+jt:+:~(:.~:+~+~:+:+~++¢+:+~`++~:+:~:+:~:+:+:::~:+~+$~:+' "'- ~:~:5:::~
s,,+,.-,,~ ,,,,,,,s
(/+, .+
....
+~.++
i
,.~
+.,--~
J
o.
++N+;+-...... ++o++ ::-'::::::::;:::-."~ : : ~;~:~.~:~'~:~f'~,\ ........ .~i,~'.z..'+.:.... -.+ .....~ , ~
30ON
~
,~,..
of Asia
Tectonic models
80*E
Tibet/Siberl. I ,t@~
tectonics
, t::~.;';;:::ilg!;::i::::#~t' U H I N A ~'~ili::!l~i~!ii+;::~il/ I • i -:.:.:+i ,:.:+.:'~:++:.:'
It
"~ +"
, i
'O//,~,,~ ~%~. - ~
+ 500kin
t ,
Fig. 2. Kinematic model for central Asia modified from Avouac & Tapponnier (1993) showing slip rates on faults and Euler poles with rotation rates. The values east of 95°E are not constrained by inversion.
N
CENOZOIC SE ASIA RECONSTRUCTIONS observations, compiled earthquake data and the India-Eurasia convergence rate from the De Mets et al. (1990) NUVEL-1 model. In the model, two stable blocks (Tibet and Tarim) are rotating relative to stable Siberia about poles indicated on Fig. 2. Motion between the blocks is taken up along faults and zones of active shortening although there are large uncertainties in slip rates. The model transfers the estimated 40 mm a-1 of northeastward movement of Tibet into clockwise rotation and extrusion of crustal blocks along curved left-lateral faults. After allowing for shortening at the Longmen Thrust, the South China Block is rotating clockwise and moving southeast at 10-15 m m a -1 with respect to Siberia. Tibet, rather than being a simple rigid rotating block as in the Avouac & Tapponnier (1993) model, is according to Molnar & Lyon-Caen (1989) extending east-west. They calculated from earthquake data that through strike-slip and normal faulting, northern Tibet is shortening north-south at roughly 5 mm a-1 and extending east-west at roughly 1 0 m m a -1. In south Tibet and the Himalayas north-south graben formation results in east-west extension of 18 _ 9 mm a-1. The eastwest extension in southern Tibet and the Himalayas is attributable to radial thrusting on the Himalaya front and in northern Tibet to gravitational collapse. Using radiometric age data, Harrison et al. (1992) suggested that the graben could have originated between 8 Ma and 4 Ma. Molnar and Lyon-Caen (1989) concluded that eastern Tibet has an eastward component of movement with respect to the Tarim Basin and India of around 30-40 mm a-1 but declined to estimate the effect on South China because they could not evaluate the shortening rate in Longmen Shan. Avouac & Tapponnier (1993) estimated that 50% of the motion between India and Siberia is absorbed by the extrusion of Tibet to the northeast (at 40 mm a-1 at the eastern end). Molnar and LyonCaen (1989) estimated it at about a third in their analysis. Neither study took Altai into account. Dewey et al. (1989) approached the problem by adding the estimated present-day shortening rates along slip lines for Eurasia-India convergence and concluded that the residual available for extrusion, which they ascribe to gravitational collapse of the thickened Tibet crust along conjugate strike-slip faults, was not significant. Updating their estimates for the central Himalayas to Tienshan transect, the total India-Siberia convergence is 45 mm a-l (NUVEL-1) less the sum of 13 mm a-1 in Tienshan, 6 mm a-1 in west Kunlun, 5 mm a-1 in central Tibet and 18 mm a -1 in the Himalayas; the remainder is 3 mm a-t. A similar transect from the eastern syntaxis to Altai starting with 55 mm a-1 from NUVEL-1 and deducting the sum of 5 mm a-1 for
125
central Tibet, 3 mm a-1 in Nan Shan, 11 mm a -1 for the Altyn Shan fault and 18 m m a -1 for the Himalayas, leaves a residual of 18 mm a-1. In their topographic analysis Le Pichon et al. (1992) demonstrated that Tienshan and Altai (northeast of Tien Shan), together form an eastward increasing topographic mass. It is therefore likely that the Altai convergence rate is greater than the 13 mm a-1 estimated for Tienshan. Assuming a value of 15 mm a-1 would again give a residual of 3 mm a -1. Although the estimates, apart from NUVEL-1, have large uncertainties this methodology does not provide compelling evidence for extrusion.
D u r a t i o n o f the p r e s e n t t e c t o n i c r e g i m e
The structures presently active may have originated at various times from the early Miocene onwards. The onset of the present tectonic regime in Tienshan has been broadly established as early to middle Miocene (16 +22/-9 Ma, i.e. 7-38 Ma) by Avouac et al. (1993). The post-lower Mesozoic displacement on the Altyn Tagh Fault was estimated as 500 km by Peltzer & Tapponnier (1988). In contrast, Dewey et al. (1989) consider that the long major strike-slip faults such as the Altyn Tagh, Kun Lun and Tang Ting (Xan Shui He), all have displacements of 150 km. Depending on the estimated displacement, the Altyn Tagh Fault the fault is between 5 and 17 Ma old using an estimated slip rate of 30 mm a-1 and ignoring the uncertainty. The Red River Fault (Fig. 6) on the boundary of the South China and Indochina Blocks has been a right-lateral fault since 4.7 Ma but was left-lateral until 17-19 Ma (Leloup et al. 1993). Although not at present active, Quaternary rightlateral slip rate was estimated by Allen et al. (1984) at 2-5 mm a-1. The Pliocene to Recent fightlateral slip on the Red River Fault in the Avouac & Tapponnier model is about 50-75 km, or as little as 10-25 km using the slip rate of Allen et al., Longmen Shan and Qilian Shan presumably absorbing any excess strain. Between 5 and 1719 Ma any residual extrusion of the South China block would have also rotated Indochina because the Red River Fault was inactive.
Kinematic models for the IndiaAsia collision The position of the Indian Craton from the onset of collision, has been forward modelled from plate tectonics by amongst others, Dewey et al. (1989) and Le Pichon et al. (1992). The size of greater India prior to the collision is a matter of conjecture,
126
G. PACKHAM
5~.~ ~
~"~
,-, ~o~'~
~QAIDA.~.~OC~
ie~6
,o~'~
<
q
N. L H A S A ~\~= BLOCK
S O U T H CHINA BLOCK
SHILLONG HILLS
BENGAL
SHAN PLATEAU
+6 N SHAN --20'N
/ --15'N
INDOCHINA BLOCK
--IO'N
MALAY BASIN
-5*N
Fig. 3. Simplified tectonic map of western Southeast Asia and adjacent regions of central Asia modified from Tapponnier et al. (1986) and Dewey et al. (1988) and unpublished sources, with Jurassic and Cretaceous palaeomagnetic deviations from expected palaeolatitudes and declinations. Data as in Fig. 8. Results from Jurassic rocks are circled. Heavy lines are sutures, large arrows are early to middle Cenozoic fault slip and small arrows are late Cenozoic fault slip.
CENOZOIC SE ASIA RECONSTRUCTIONS
but for a given size and the time of collision, the location of the boundary of Asia can be determined within fairly small limits.
127
limestones are the youngest preserved Himalayan marine sediments (Searle et al. 1987). Harrison et al. (1992), Tapponnier et al. (1986) and Peltzer & Tapponnier (1988) favoured a collision date of 50 Ma.
Time of collision
Besse & Courtillot (1991) pointed out that the convergence rate between India and Eurasia drops sharply between anomalies A22 and A21 (50.347.8 Ma on the scale of Harland et al. 1990). An Indian Ocean spreading ridge reorganization, shown on Fig. 4 as the change from Phase 2 to Phase 3 followed between A20 and A18 (43.941.1 Ma). Both Dewey et al. and Le Pichon date the collision at 45 Ma (early middle Eocene). Subduction-related volcanism north of the suture ends at about 50 Ma and related plutons range from 94 to 41 Ma (Dewey et al. 1988). Middle Eocene
Fig. 4. Sea floor spreading phases in the eastern Indian ocean. Phase 1 is from Late Jurassic to mid-Cretaceous, Phase 2 is from mid Cretaceous to middle Eocene (anomaly 20) and Phase 3 is middle Eocene to the present. Note the location of the Exmouth Plateau (EP) and the Cape Range Fracture Zone (CR) which may have been the northern limit of greater India before rifting off Australia and Antarctica.
Greater India
On the basis of the fit of India against Australia and Antarctica, and the shedding of continental clastics north onto the Exmouth Plateau, Powell et al. (1988) suggested that continental crust of greater India extended north of the West Australian margin to the Cape Range Fracture Zone (Fig. 5). Spreading patterns in the Perth Abyssal Plain and the Bay of Bengal indicate a common breakup date of Valanginian (Powell et al. 1988; Ramana et al. 1994). This proposed extension of greater India 1300 km, north of the reconstructed location of the eastern syntaxis, was used by Le Pichon et a/.(1992) in their model which is discussed below. A smaller greater India has been proposed Dewey et al. (1989), Harrison et al. (1992) and Treloar & Coward (1991). Dewey et al. argue that apart from erosion loss after collision the remainder of the crust is stored in the Himalayas. On this basis, Le Pichon et al. (1992) estimate a minimum size for greater India as 600 km north of the suture, whereas Dewey et al. (1989) give a range from 450-1000 km. In the Harrison et al. and Treloar & Coward models, collision commenced near the western syntaxis leaving a 700 km gap at the eastern syntaxis which closed in the early Oligocene over 10 Ma after the collision onset. Le Pichon et al. (1992) assumed simultaneous collision and located the collision boundary 1000 1200 km south of the Indus-Zangbo suture (Fig. 4). Dewey et al. (1989) and Tapponnier et al. (1986) also assumed simultaneous collision, placing the Asia boundary fifteen degrees further south at its eastern end. The fit of pre-rift India against Antarctica makes it unlikely that it extends east of 95°E. Its southeastern margin runs from east of the Shillong Hills under the eastern Bengal Basin to the foot of the slope parallel to the northeast trending coastline. Some recent authors (e.g. Zhao & Morgan 1985; Peltzer & Tapponnier 1988; Le Pichon et al. 1992) show an eastward prolongation of the craton including the Cretaceous to Palaeogene flysch deposits of the Indo-Burman Ranges on the Indian plate. This is not compatible with the pre-rift reconstruction or the geology of the Bengal Basin where a southeast deepening continent margin is identifiable from the Early Cretaceous onwards (Lindsay et al. 1991). It is argued later that the flysch was part of the Asian margin.
128
G. PACKHAM I 60*E
'
70° ,,...,'1
i
t
80*
90"
II-- " - ~
~
IO0"E
I
I
4
--10*N
-0"
i
1
I
\~
I
Fig. 5. Estimated locations of the Asian margin at the time of the Indian collision from Tapponnier et al. (1986), Dewey et al. (1989) and Le Pichon et aL (1992).
Topographic analysis
Le Pichon et al. (1992) attempted to assess whether significant extrusion occurred by estimating the area of the crust lost during shortening, assuming conservation of continental crust, and comparing it with the volume of the topography that can be attributed to the collision after adding an estimate of the volume of eroded sediment. The model uses the larger greater India indicated above, assumes a simultaneous collision along the craton edge, and extended further north than the Avouac & Tapponnier (1993) model to include Altai. Assuming an average pre-collision elevation of 500 m, Le Pichon et al. found there was a uniform loss of mass west of the east syntaxis and an excess east of it (Fig. 6). The excess accounted for one third to half of the loss to the west. The loss of non-thickened crust can be accounted for by some combination of lateral extrusion and loss of lower crust into the mantle, possibly by eclogitization. The loss would be reduced if the crust was thinner i.e. was at sea-level or below it. For a minimal greater India in which there was 700 km of oceanic crust north of the eastern syntaxis at the time of collision, the loss of continental crust is reduced by about 20%. The transfer of mass to the east of 103°E still remains, but the deficit is reduced to between one third and one eighth, or even less if the crust was thinner. The minimum estimate of Le Pichon et al. (1992) is not very different from the result obtained by the kinematic approach of Dewey et al. (1989). They summed the strain throughout the collision event, along slip circle traverses through the western syntaxis, central Himalayas and the eastern syntaxis, and concluded that the crustal volumes
nearly balanced. No extrusion was required beyond the thickening of the Tibetan Plateau east of the eastem syntaxis. The crust lost from the north of the Indian Craton was assumed by Dewey et aL to have been thin like that of the Exmouth Plateau off Western Australia. The essential differences between the models are the thickness, volume and fate of the greater India crust. Le Pichon et al. assumed that the upper crust was stored in the Himalayas and partly eroded off, but the lower crust was transferred to the mantle or emplaced north of the Indus suture. If all the crust has been transferred under the plateau, half of its elevation would be due to the transfer and the remainder to local shortening. The mechanism Le Pichon et al. favour is injection of Indian crust into the Tibetan lower crust as a low viscosity fluid, as proposed by Zhao & Morgan (1985). Le Pichon et aL rejected the thin crust hypothesis of Dewey et al. because there is no evidence of Eocene and Oligocene deep water sediments overlying the Indian crust fed into the system. A hint of thin (or thinned) crust comes from the Upper Jurassic to Lower Cretaceous section of the Tethyan Himalaya cover north of the Main Central Thrust in Nepal. A clastic deltaic to shelf clastic succession deposited on northward palaeoslope is overlain by post-breakup pelagic slope carbonates (Gibling et al. 1994). 3000-
2000-
ItVS y n t ~
o
IOO0
/o.
' Longitude 80"
;o.
100"
Fig. 6. Estimates of equivalent linear shortening between India and Eurasia, determined by Le Pichon et al. (1992), from the volume of the topography and corrected for erosion assuming conservation of continental crust. Average block elevations are plotted for sea-level and 500 m. The upper curve with filled triangles represents the linear shortening from kinematics for a maximum greater India, the lower curve (added here) with open triangles is for a minimum greater India. Note the eastward displacement of shortening determined from topography.
CENOZOIC SE ASIA RECONSTRUCTIONS
Phases of the collision and their expression in SE Asia Rotation or extrusion of the part of SE Asia west of the Red River Fault has been identified as a major driving mechanism for sea floor spreading in the South China Sea (e.g. Tapponnier et al. (1986) and Briais et al. (1993) between 32 Ma and 16 Ma. This event, resulting from India-Eurasia collision, divides the Cenozoic SE Asian tectonic history into three major phases, namely middle Eocene to early Oligocene, early Oligocene to middle Miocene, and middle Miocene to Recent. M i d d l e E o c e n e to E a r l y O l i g o c e n e
Dewey et al. (1989) proposed that between 45 and 30 Ma, convergence was taken up by the commencement of stacking of northern Himalayan thrust sheets and thickening of the Tibet crust to 70 km elevating it to 3 km by 30 Ma, along with progressive indentation of the Asian margin and bending the older sutures (Fig. 1). Because the Palaeogene and older strata in the Tibetan plateau are strongly deformed and Neogene sediments are essentially flat lying, Dewey et al. (1988) dated the major deformation of the plateau as pre-Miocene and estimated that the shortening was about 50%. Harrison et al. (1992) accepted the degree of deformation of some of the Palaeogene sediments but questioned its generalization over the plateau. They estimated shortening at 150 km (< 20%). Using radiometric data to determine cooling rates in the Quxu pluton near Lhasa, Harrison et al. (1992) suggested thickening, uplift and unroofing of the pluton commenced in the early Miocene (21-18 Ma). Later work by Copeland et al. (1995) related unroofing of granite bodies successively towards the north to uplift on the Gangdese thrust to the south commencing at 27 Ma. This does not conflict with the argument of Dewey et al. (1988) that Palaeogene crustal shortening elevated the Tibetan Plateau. A review by Mercier et al. (1987) of the climatological evidence for the date of plateau uplift from palaeobotanical studies reveals conflicting interpretations. Some authors consider that steady uplift occurred in northern Tibet from the end of the Cretaceous, and in the south (Gangdese Belt) from the middle Eocene, and suggest the Himalayan uplift was later. Other authors consider that the uplift was predominantly Plio-Pleistocene. In their well known plasticine model experiments, Tapponnier and his co-workers (Tapponnier et al. 1982; Peltzer & Tapponnier 1988), partitioned extrusion and crustal thickening equally throughout the collision and conserved all crust. The indentor representing the Indian craton deformed a striped
129
plasticine representing Asia, producing a remarkably good facsimile of the observed fault patterns. In their tectonic evolutionary scheme for Asia, Tapponnier et al. (1982, 1986) and Peltzer & Tapponnier (1988) identified two major phases of extrusion. The first phase between middle Eocene and middle Miocene, had extruded the region south of the Red River Fault a total of 800-1000 km in two stages. Extrusion in the second phase from middle Miocene to Recent employs a model similar to that of Avouac & Tapponnier (1993) for the present tectonic regime, in which extrusion occurred north of the Red River Fault, as discussed above. Tapponnier et al. (1986) suggest thickening of the Shan Plateau and the Lhasa Block may have taken place before the formation of the South China Sea. But in Peltzer & Tapponnier (1988), the strain produced by convergence was apparently partitioned between thickening and extrusion along the Wang Chao and Three Pagodas Faults (Fig. 3). The end of movement on the Wang Chao fault at about 30 Ma has been supported by radiometric data from Maluski et al. (quoted in Huchon et al. 1994). Peltzer & Tapponnier (1988) estimated displacements on both the Wang Chao and the Three Pagodas Faults at 300 km by matching the displacement of the Uttaraditt suture and upper Cretaceous magma belts respectively. The magmatic belts are not as clearly defined on geological maps as portrayed by Peltzer & Tapponnier (1988) and the displacements on the Three Pagodas Fault could well be substantially less than their estimates and may even be right-lateral if the displaced extension of the Uttaraditt suture on the Gulf of Thailand is aligned with the projected BentongRaub suture (Fig. 3). Harrison et al. (1992) explained the delay in movement on the Red River Fault until early Oligocene (35 Ma) by assuming a smaller greater India than Le Pichon et al. (1992) and that collision commenced at the western syntaxis at about 50 Ma leaving a gap of 700 km of oceanic crust at the eastern syntaxis. Movement on the Red River Fault started after the early Oligocene closure of the gap and collision in eastern Tibet. In this scheme, the duration of displacement on the Wang Chao Fault would have been short. E a r l y O l i g o c e n e to E a r l y M i o c e n e
In the second phase, according to Dewey et aL (1989), after 30 Ma when the Tibetan Plateau had achieved considerably buoyancy, India began to indent Asia, deflecting the Asian sutures and commencing block rotation (Fig. 1). Wrenching began along conjugate faults bounding the elevated area and deformation spread north with diminishing
130
G. PACKHAM
intensity, into Tienshan and south into the Himalayas elevating them along the Main Central Thrust and thrusts in the Lesser Himalayas. Uplift of the Gangdese belt in the early Miocene resulted from underthrusting of the northern Himalayas. In their model Peltzer & Tapponnier (1988) proposed that movement on the Red River Fault generated the South China Sea as the southern part of Sundaland moved to the southeast. They estimated the displacement on the fault as 500 km, by matching the extrapolated Jinsha and Uttaraditt Sutures (Fig. 3). The magnitude and direction of displacement on the Red River Fault is supported by Lacassin et al. (1993) who estimated the leftlateral strain in mylonitic gneisses as 330 _+60 km and radiometric dating by Leloup et al. (1993) has found that in the Diancang Shan region in Yunnan, left-lateral ductile shear terminated between 17 and 19 Ma, preceded by 7 km of unroofing at 23 and 19 Ma, due to large scale oblique shear. In the Harrison et al. (1992) variant on the Tapponnier model the main phase of movement on Red River Fault was from 35 to 20 Ma. Half to two thirds of the convergence was taken up by extrusion. The end of the phase is based on evidence of unroofing of the Quxu Pluton near Lhasa, suggesting that significant uplift of the Tibetan Plateau occurred at 18 to 21 Ma rather than after 17 Ma as suggested by Tapponnier et al. (1986). The unroofing is apparently related to uplift associated with movement on the Gangdese Thrust that commenced at 27 Ma (Copeland et al. 1995). Briais et al. (1993) made a kinematic study of the spreading history of the South China Sea and the displacement on the Red River Fault. Spreading commenced at 32 Ma with a ridge jump at 26 Ma and a change in ridge trend from east-west to northeast-southwest at 24 Ma with spreading ending at about 15.5 Ma soon after the cessation of movement on the Red River Fault at 17-19 Ma as discussed above (Fig. 7). Using poles of opening they determined for the South China Sea and assuming that the Indochina Block was attached to the drifting margin Briais et al. (1993) determined the displacement on the Red River Fault as 555 km for a point near the Uttaraditt suture; a good correlation with the estimate of Peltzer & Tapponnier. In the reconstructions of Briais et al. convergence increased across the fault inland and there was extension in the lower Red River region, correlating well with the study of Leloup et aL (1993). The implication is that the Indochina Block and the drifting South China Sea margin formed a single small plate and that spreading was driven by motion of the Indochina Block. This point will be returned to later. A stage pole determined by Huchon et al. (1994) describing the early stage of opening was located at 13°N and 97°E in the
Fig, 7. Interpretation of South China Sea structure modified from Tapponnier et al. (1986) showing: oceanic crust (cross hatched) and shallow areas of continental crust (dotted). LP, Luconia Province; MAB, Macclesfield Bank; NCB, Nam Conson Basin; PA, Palawan Block; PS, Paracel Shoals; RB, Reed Bank; SCAR, Scarborough Seamounts; WNB, West Natuna Basin; YB, Yinggehai Basin.
northeast Andaman Sea. At 30 Ma it would have been located east of the eastern syntaxis but well south of the Indus suture. The closeness of the pole to the Indochina Block suggests that rotation was more important than translation. Although the commencement of South China Sea floor spreading at 32 Ma fits well with the initiation of movement on the Red River Fault, pre-drift rift basins containing Eocene sediments occur on both margins of the South China Sea (Salvidar-Sali et aL 1982; P. Chen et al. 1993). In the case of the Pearl River basin on the south China margin and Reed Bank on the drifted margin, there are both Eocene and Palaeocene sediments (Taylor & Hayes 1980; Yu 1990). That is, South China Sea margin extension commenced earlier than the India-Asia collision and the supposed time of initiation of the Red River Fault.
Early Miocene
to R e c e n t
The uncertainty in dating the initial elevation of the Tibet Plateau was discussed above. Rather than occurring by crustal shortening from late Eocene to
CENOZOIC SE ASIA RECONSTRUCTIONS mid-Oligocene time as proposed by Dewey et al. (1989), Tapponnier et al. (1986) and Harrison et al. (1992), consider that the bulk of the uplift took place after movement on the Red River Fault. This requires transfer of lower crust from the India plate across the Indus suture because the Neogene sediments of the plateau are relatively undeformed. As described previously, the present tectonic regime commenced in middle to late Miocene time. During it some of the convergence was transferred to eastward movement of Tibet relative to the Himalayas and in part to South China as indicated by the early Pliocene commencement of rightlateral movement on the Red River Fault. Elevation of the Tibet Plateau to its present height was approximately coincident. Harrison et al. (1992) suggested it occurred between 8 Ma and 5 Ma and Dewey et al. at about 5 Ma. England & Houseman (1988) explained this last uplift as the result of delamination of the mantle lithosphere beneath the plateau.
Palaeomagnetic data: central Asia and mainland SE Asia The objective of palaeomagnetic studies in central and SE Asia has been to determine the amount of northward movement of sites north of the Himalayas and whether blocks in SE Asia have been rotated and moved south due to extrusion. The palaeomagnetic data have been recently summarized by Y. Chen et al. (1993) and Huang & Opdyke (1993) using different methodologies. Rotations and change in palaeolatitude have been determined by comparing measurements with the Eurasian polar wander path compiled by Besse & Courtillot (1991). The uncertainties taken into account are statistical errors arising from the measurements within and between sites, structural corrections, estimates of rock age, date of magnetiztion and uncertainty in the polar wander path location. Fortunately, the Eurasia polar wander path assembled by Besse & Courtillot (1991) indicates very little movement in the pole location in the Cretaceous between 130 and 80 Ma. Studies have been biased towards Cretaceous sediments but unfortunately many of them are in non-marine successions and hence dating is imprecise. Huang & Opdyke (1993) compared the results from separate sites with the pole for that magnetization age. The Y. Chen et al. (1993) compilation compared the averages of individual studies on Cretaceous rocks for each block with an average Cretaceous pole. Although the results from the two methods are in general comparable Y. Chen et al. lost fine detail. The dataset for Middle Jurassic to Upper Cretaceous sites from Huang &
131
Opdyke (1993) are replotted according to longitude in Fig. 8 and individual sites in the eastern part of the region have been plotted geographically on Fig. 3. With one exception, the Pamir measurements west of the western syntaxis show counterclockwise rotation and possible but minor southward displacement. Tamir poles indicate significant northward motion with greater displacement to the west. The Tarim block data have large errors, Huang & Opdyke summarize it as 1380 _+790 km but suggest several hundred kilometres without significant rotation is more likely. The Tibet data (Lhasa Block) suggest about 1500 km of northward displacement with counter-clockwise rotation increasing to the east. East of the east syntaxis, the Markam sites (eastern Qiangtang Block) are rotated clockwise and unlike Tibet may be displaced southward but the error bars overlap zero displacement. Sites on the northern Shan-Thai Block, southwest of the Red River Fault and west of the Uttaraditt suture, indicate clockwise rotation apart from two locations close to the Red River Fault which are rotated slightly counter clockwise (Figs 3 and 8). Five Shan-Thai sites indicate possible northward displacement. The three Jurassic sites (22, 23 and 25) apparently indicate a strong northward displacement. These may not be significant because the wander path of Besse & Courtillot (1991) has a strong shift in the palaeopole corresponding to a palaeolatitude change at these sites from 49°N to 20°N between 170 Ma and 130 Ma (Oxfordian to Valanginian) after being relatively stationary from 210 Ma (Late Triassic). The fourth site (R on Fig. 8) with possible northward shift is from the Jurassic~Cretaceous Kalaw Red beds in eastern Myanmar on the Shan-Thai Block (Richter et al. 1993). For ages of magnetization between 130 and 80 Ma this sampling locality yields a palaeolatitude displacement between 2.6 and 6.0 degrees northward within a _+6 ° error bar. If there is a northward displacement it is only small. A possible small northward displacement has also been measured at Site 24. Possible southward displacements were found at Cretaceous sites 17, 18 and 19 are (Fig. 8) but the errors make them inconclusive. Yang & Besse (1993) report a palaeolatitude of 23.1 +_ 1.8°N for the medium temperature magnetic component in sediments from the Khok Kruat Formation (Albian-Aptian) in the Khorat Basin on the Indochina Block. After including the uncertainty of the palaeopole location Huang & Opdyke (1993) quote the palaeolatitude change as -8.0 _+5.3 ° northward with a clockwise rotation of 14.4 _+6.1 °. No fold test was possible at the site because of low dips but older formations in the same sequence do give a positive fold test. The
132
G. PACKHAM
Indochina 2O Moved North
°i 80
21,
60
dochina
Pami
01
/--I
/'~ ~
-20
-40
14~,Jl/ Markam
(I,j-/~r 28
I
II
~/
8
r
t
70
'
'
'
'
I
80
. . . .
Ill'
90~
1'
'
'
!
100
'
'
'
,
I
110
. . . .
I
120
. . . .
i
130
Fig. 8. Plots of deviations of palaeomagnetic measurements from the expected palaeolatitude and declination from the equivalent aged Eurasian pole from Besse & Courtillot (1991) plotted against latitude. Data and point numbers are from Huang & Opdyke (1993) except for the point 'R' which is from Richter et al. (1993). The inner bar is the uncertainty of determination from the individual dataset and the outer bar includes the uncertainty of pole determination.
Cretaceous results south of the Red River Fault from the western side of the Shan-Thai Block suggest a small northward m o v e m e n t and a correspondingly small southward m o v e m e n t on the east. Declinations are generally rotated clockwise and
they are deflected around 70 ° clockwise from the Lhasa Block measurements (Figs 3 and 8). At face value, these results could indicate rotation of the Shan-Thai Block about a pole located within it. The Khorat Plateau rotated further
CENOZOIC SE ASIA RECONSTRUCTIONS
133
south could be part of the same domain. Variability in disturbance of declinations suggest that the Shan-Thai Block may not have behaved as a single entity but was internally deformed by displacement along the many faults through the region. Although the rotation can only be dated as later than Early Cretaceous, the event may be of middle to late Tertiary age, since upper Cenozoic sites in central and western Thailand are also disturbed and are mostly but not universally rotated clockwise (McCabe et al. 1993; Richter et al. 1993). Sites of similar age from the Indochina Block in eastern Thailand (Khorat) and Vietnam are not significantly rotated. The South China results indicate possible southward displacement in the eastern part, but in the west site 26 in the Sichuan Basin may have been displaced to the north (6 _+6°). The magnetization age is indicated by Huang & Opdyke as uncertain. Sites 27, 28 and 33, between the Xian Shui He and the Red River Faults, are in the central Hunnan Basin are close to their expected palaeolatitude. Site 33 in the central Yunnan Basin near the Red River Fault is, like the Markam sites, rotated significantly clockwise. In summary, there is significant northward translation of Tibet, counter-clockwise rotation in Pamir, counter-clockwise rotation of eastern Tibet, clockwise rotation of sites east of the eastern syntaxis but disturbed close to the syntaxis and near the Red River Fault. One good Indochina Cretaceous dataset indicates probable southward motion, the remaining Cretaceous results spread from west of the Dien Bien Phu Fault in the Shan-Thai Block are ambiguous but could indicate clockwise block rotation with little change of latitude.
it, was moving at a higher velocity obliquely away from the Indochina block, rapidly decreasing the length of the contact between them. Thus after 26 Ma, spreading was not driven by motion of the Indochina Block. Further, spreading ended about 15.5 Ma, later than the cessation of movement on the Red River Fault (17-19 Ma according to Leloup et al. 1993). To invert the argument, using South China Sea spreading to determine fault movement would substantially overestimate the fault displacement after 26 Ma. Of the 560 km of Red River Fault displacement calculated by Briais et al. (1993), only the 240 km which occurred between the commencement of spreading and 26 Ma can be attributed to driving South China Sea spreading. While fault displacement may have continued, spreading in the South China Sea does not provide a means of determining it. The amount of strike-slip movement may have been small in the last phase of movement on the fault from 23-19 Ma when oblique shear and unroofing occurred. The fault could have been active before 32 Ma driving some of the rifting which preceded South China Sea spreading, although as noted earier this started in Palaeocene time before the commencement of the collision of India. Extrusion of Indochina on the Red River Fault fits well with the Oligocene phase of spreading in the South China Sea, but is difficult to reconcile with published basin histories in the region. In the Nam Con Son Basin (Fig. 7) on the Indochina Block syn-depositional extension occurred (Matthews & Todd 1993) while in the West Natuna Basin to the south there was post extension subsidence (Ginger et al. 1993).
The driving mechanism for South China Sea spreading
The western margin of Sundaland
In the discussion of early Oligocene to early Miocene events, it was pointed out that South China Sea spreading had been used to estimate displacement on the Red River Fault on the assumption that the spreading was driven by extrusion of Indochina on the Red River Fault. During the first stage of opening from 32 to 26 Ma, spreading ridges were short and the Indochina block had a long contact with the drifting side of the basin, extending almost to Reed Bank (Fig. 7). After a ridge jump between anomaly 7 and 6b at 26-24 Ma, spreading began to propagate rapidly to the south. The ridge trend changed from 5-10°E of N to about 60°E of N. The southwestward extension stabilized at about anomaly 6 at 20 Ma. As the spreading pattern changed, the drifting block, consisting of an elongated strip of thinned continental crust with oceanic crust outboard from
The geology of the western margin of Sundaland provides some constraints on models of India-Asia collision which have been largely ignored. Some aspects are examined here. The Java-Sumatra
margin
In the Sumatra-Java margin, there is a Late Cretaceous to middle Eocene hiatus in volcanic activity. Possible explanations are that, prior to the middle Eocene a north facing arc lay to the south and collided with polarity reversal in the middle Eocene or that there was little or no subduction on the Sumatra-Java margin during the hiatus. The latter is supported by the work of Wajzer et al. (1991), who have interpreted the Upper Mesozoic Woyla Group in the Natal area of western Sumatra as a west-facing accretionary prism which has later been disrupted by strike-slip faulting.
134
G. PACKHAM
Packham (1990) suggested that between latest Cretaceous and middle Eocene time, sea floor spreading in the Indian Ocean (Phase 2 in Fig. 4), extended east only as far as its present preserved limit off southern Sumatra, 1300 km east of the Ninetyeast Ridge. East of it lay the older sea floor of the Australia plate, which at that time was moving slowly away from Antarctica (Royer & Sandwell 1989). With middle Eocene reorganization of sea floor spreading, between anomalies 20 and 18 (44-41 Ma), subduction under the Sunda margin commenced or increased from a low rate. Radiometric ages of an andesite and a quartz diorite from northern central Myanmar of 50 and 53 Ma respectively (Mitchell 1993) suggest that subduction-related volcanism occurred there during the Sumatra volcanic hiatus and that the IndiaAustralia plate boundary intersected the Asia margin between Sumatra and Myanmar. The Myanmar dates also fall in the time range of the Linzizong Volcanics of the Lhasa Block (Dewey et al. 1988) and presumably represent the eastern end of the Paleogene volcanic arc. Using the India-Asia reconstruction parameters of Le Pichon et al. (1992), the extrapolated middle Eocene position of the India-Australia plate boundary would intersect the central Sumatra coast 500 km west of its present position (Fig. 9). It is in a similar position in the 45 Ma reconstruction of Peltzer & Tapponnier (1988). For the Sumatra margin to be essentially passive until then requires the India-Australia plate boundary to have been west of it, i.e. Sumatra must have been about 500 km further east relative to South China at the time collision commenced. The distance may be reduced by about 100 km, subtracting movement of western Sumatra on the Burma plate since the inception of the Sumatra Fault.
M y a n m a r a n d the I n d i a - A s i a collision The tectonic position of the Indoburman flysch belt, the time of its docking and the distribution of post-docking strain all have important implications for reconstructions of western SE Asia. As indicated in the discussion of greater India the eastern boundary of the Indian craton runs from south of the Shillong Hills to the Bay of Bengal (Fig. 3). The Indoburman flysch belt has been accreted onto the craton margin after collision. Palaeogeographically, the flysch is an Asian margin accretionary wedge and the proximal Central Basin of Myanmar is a shelf deposit. The flysch is a thick sequence of Senonian to middle Eocene feldspathic turbidites (Bender 1983), which for the most part predate the collision of India with Asia. Their distribution down the Arakan coast to the west of the oceanic
EURASIA
35 \ ,= 4 5 × _ _ X - -..~( ''~
INOIA " ~ PLATE
oo
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ......... ,5
20os 6O"E
PLATE 80"
I 100*E
Fig. 9. Reconstruction of India and western SE Asia for 45 Ma from Peltzer & Tapponnier (1988). To it have been added the proposed India-Australia plate boundary prior to Phase 2 of Indian Ocean Spreading (Anomaly 20-44 Ma), and an X in Assam at the eastern syntaxis and another on the Eurasia plate at the corresponding point at the northern end of the Indoburman flysch belt on the Eurasia plate from the same set of reconstructions for 45, 35 and 15 Ma. Docking of the flysch belt at Assam is thought to have occurred in the Oligocene between 30 and 36 Ma.
crust of the Bengal Basin precludes their derivation from the Indian margin. West of the flysch the Central Basin of Myanmar contains a shelf to nonmarine Albian-Aptian to Pleistocene sedimentary succession. In the Naga Hills near the eastern syntaxis (Fig. 3), the Upper Cretaceous-Eocene flysch is thrust over Palaeocene to upper Eocene platform cover (Mitchell 1993) and both are overlain by the post-docking Oligocene to ?lower Miocene, Barail Formation (Roy 1986). The Barail Formation progrades south from fluvio-deltaic to marine. The stratigraphic relations indicate an early Oligocene docking of the flysch terrane on the India plate, probably within 10 Ma of the India-Asia collision, and movement north on the craton since that time. On a broader scale, the structural units of western Myanmar appear to be an eastward continuation of the Asian margin wrapped around the eastern syntaxis in agreement with palaeomagnetic declination rotations (Fig. 3). The uplift between the Central Basin and the flysch belt corresponds to the Indus suture as indicated by Dewey et al. (1988). At its northeastern extremity in Myanmar, the suture appears to end against a branch of the Sagiang Fault. It is apparently
CENOZOIC SE ASIA RECONSTRUCTIONS displaced 450 km to the south to a belt of ultramarie rocks 100 km north of Mandalay, thence it trends north-northeast, curving north-northwest around the northern end of the Central Basin where it is separated from the Indus suture by the Lohit Thrust. The presence of Triassic flysch in Myanmar on the western margin of the Central Basin links it with the southern Lhasa Block-Indus suture as does the Upper Cretaceous flysch. The basement rocks of Myanmar from the Shan Scarps westward form part of the Cretaceous to Eocene magmatic belt which Mitchell (1993) compares with the Gangdese plutons. From docking at the beginning of Oligocene time, say 36 Ma, until the commencement of Andaman Sea spreading at c. 13 Ma (Curray 1989), the northward transport of the Indoburman flysch belt on the Indian craton relative to fixed Asia would be 1500 km, using the kinematics of Le Pichon et al. (1992). If the Shan-Thai Block was extruded the relative displacement would be that much greater. Identification of such a dislocation is very speculative. Mitchell (1993) suggested 1100 km of right-lateral displacement of the Cretaceous outer magmatic belt exposed in highs within the Central Myanmar basin, relative to the Shan-Thai Block prior to Andaman Sea spreading. Assuming coupling to the India plate this would take 17 Ma. The fault would have to be located between the Central Basin highs and the Sagiang Fault. This would leave the remaining 400 km to be taken up west of the Cretaceous magmatic belt in the Indoburman Ranges, the Andaman arc and pre-rift extension prior to Andaman spreading. A later time of docking would reduce the residual by 64 km per Ma. The position of the northeast comer of the flysch belt has been plotted on the 45 Ma reconstruction of Peltzer & Tapponnier (1988) on Fig. 9 together with its 35 and 15 Ma locations from their other reconstructions. At 35 Ma, about the time of docking according to the geological evidence cited above, their reconstruction shows it about 2000 km to the north. At 15 Ma it is still around 900 km north of the corresponding docking point in the Naga Hills. At 30 Ma when movement ended on the Wang Chao Fault, relative to stationary Eurasia the eastern syntaxis was at 1 I°N and the northern edge of greater India as far north as 22 ° corresponding to the present day location of the western end of the fault. Andaman Sea spreading commenced at about 13 Ma (Curray 1989). Most of the 460 km of extension has been taken up onshore along the westem margin of the Burma plate by right-lateral slip on the Sagiang Fault on the western margin of the Shan Plateau (Fig. 3). The northward movement of the India plate relative to stable Asia over the last 13 Ma was 790 km,
135
leaving a residual of 330 km between the India and Burma plates which cannot be accounted for easily. Summarizing the foregoing, the Indoburman flysch belt was a forearc on the Asian margin at the time of its collision. After docking early in the Oligocene, the Indoburman flysch belt and possibly central Myanmar, were wrapped around the syntaxis and transported north, possibly 2300 km on the Indian Craton along the western SE Asian margin. Attenuation of the crust in the Andaman Sea region in early Miocene time, was followed by sea floor spreading in the Andaman Sea from the middle Miocene onwards accompanied by displacement on the Sagiang Fault. Some of the northward strain could have been translated into rotation of the Shan-Thai and Indochina Blocks and westward transport of Sumatra and possibly adjacent parts of SE Asia. Movement on the Wang Chao and Red River Faults was taking place as the flysch belt and central Myanmar were being transported north. Extrusion models would require the Shan-Thai and Indochina blocks and the flysch belt to be extruded from north or west of the eastern syntaxis in Eocene time, prior to collision at the eastern syntaxis. Alternatively, they could have been squeezed out behind the SE Asian margin after the northern end of the Myanmar flysch belt had docked in southern Assam early in Oligocene time.
Western SE Asia: conclusions Analysis of the present seismicity documents an eastward motion of Tibet and sets the stage for the interpretation of the earlier history. The model proposed by Avouac & Tapponnier (1993) gives a residual southeastward movement of South China of 10-15 mm a-1. At this rate during the last 4.7 Ma of fight-lateral displacement on the Red River Fault, a displacement of only 50-75 km is indicated. At a lower rate based on field studies by Allen et al. (1984) it would be 10-24 km. These latter are compatible with the Dewey et al. (1989) suggestion that most of the eastward movement has been absorbed in the Longmen Shan and that the strike-slip displacements are a fraction of those suggested by Tapponnier et al. (1986). Any pre4.7 Ma extrusion of South China would have moved Indochina as well. Transition to the present tectonic framework is likely to have occurred in late early to middle Miocene time after left-lateral movement on the Red River Fault ceased between 17 and 19 Ma (Leloup et al. 1993). Forward modelling of the India-Eurasia collision from its onset at about 45 Ma produces different results depending on the estimated volume of crust fed into the Himalayan system. Even conservative models result in some extrusion of crust into eastern Tibet. Dewey et al. (1989)
136
G. PACKHAM
assumed conservation of continental crust and a greater India with thin continental crust north of the present craton, and concluded that extrusion is limited to the eastward extension of the Tibetan Plateau east of the eastern syntaxis. With a larger and thicker greater India, Le Pichon et al. (1992) found that there was a deficit of topography between the syntaxes and an excess to the east, supporting the notion of eastward transport, but between a half and one third of lost crust remains unaccounted for. This can be explained by partitioning the loss between extrusion and transfer of lower crust to the mantle by eclogitization. If a smaller greater India is assumed, the deficit reduces to one eighth to one third and even less if the crust was thinner, a result similar to that of Dewey et al. Ahead of the advancing Indian Craton, Dewey et al. (1988, 1989) suggested crustal thickening elevated the Tibetan Plateau to 3 km between 45 and 30 Ma. Palaeogene deformation was suggested because Neogene sediments on the Tibet Plateau are relatively undeformed. India then began to indent Asia, deflecting the old sutures and rotating adjacent blocks as the Tibet crust became more buoyant. After 30 Ma deformation spread north of Tibet and south into the Himalayas. Elevation of the southern margin of the plateau discussed by Harrison et al. (1992) is related to thrusting on the Ghagdese Thrust which commenced at 27 Ma (Copeland et al. 1995) and is independent of dating the elevation of the whole plateau. If plateau elevation did not occur until around 20-25 Ma, then it is not possible to explain how over 20 Ma of shortening was distributed without invoking extrusion. In their models Harrison et al. (1993) and Peltzer & Tapponnier (1988) partitioned strain from convergence between crustal thickening and extrusion. They proposed that Indochina was extruded along the Red River Fault before the uplift of plateau. The latter authors proposed an Eocene to early Oligocene extrusion of the region south of the Wang Chao Fault. Left-lateral displacements of the order of hundreds of kilometres have been documented for the Wang Chao and Red River Faults and radiometric data support an early end of movement on the Wang Chao Fault. Harrison et al. (1992) suggested that uplift of the plateau commenced between 18-21 Ma. at the time of oblique shear on the Red River Fault. Neogene plateau elevation must be explained by some form of crustal underplating by Indian Craton crust across the Indus suture because Neogene sediments of the plateau are mildly deformed. Elevation was accompanied by movement on the Main Central Thrust in the Himalayas and the commencement of deformation to the north of Tibet, passing into the present tectonic regime.
Relating spreading in the South China Sea to determining displacement on the Red River Fault by Briais et al. (1993), although elegant, may be applicable only to the pre-Miocene fault displacement since after that time, the southeastern drifting side of the South China Sea was moving away from Indochina and cannot have been driven by its motion. Only about half of their estimated displacement of 520 km can be demonstrated this way, although some additional slip may have occurred in the pre-spreading rift phase on the China margin. The rift phase enigmatically had a Palaeocene commencement, prior to the India-Asia collision. Palaeomagnetic measurements have been quoted in support of extrusion. A review of the data here reveals that there is good evidence for counterclockwise rotation in the Pamir and northward translation of the Lhasa Block with some counterclockwise rotation. Stronger clockwise rotation predominates in the Indochina and Shan-Thai Blocks but regional changes of latitude consistent with extrusion have not been established. Southward motion of the Khorat Plateau is the only result in the east which falls outside the error bars. The data are compatible with clockwise rotation of the Shah-Thai block around a pole within the block. The dispersion of declination deviations could well be a response to complex deformation such as thrusting on the western margin of the Khorat Basin, and possible displacements on the many faults which cross the region rather than the simple extrusion of stable blocks. Evidence for the location of some entities on the western margin of Sundaland conflicts with extrusion reconstructions by Peltzer & Tapponnier (1988). A Late Cretaceous to middle Eocene hiatus in volcanic activity in Sumatra has been interpreted here to indicate that its middle Eocene position was around 500 km further east adjacent to the slow moving Australia plate (Fig. 9). For the Indoburman flysch belt and the Central Basin of Myanmar to dock at eastern Assam in early Oligocene time, they also had to be substantially east and south of the position reconstructed by Peitzer & Tapponnier, unless they were extruded with other parts of Sundaland prior to collision at the eastern syntaxis. If the docking predated Indochina extrusion, the extruded block was extruded from northeast of the Central Basin inboard from it. After docking, the flysch belt and most of central Myanmar have been transported possibly 2300 km north on the margin of the India plate and rotated clockwise. There are corresponding counter-clockwise rotations in the Lhasa Block. A major dislocation may exist beneath the western part of the Central Basin. Continued movement resulted in the middle Miocene formation of the Burma plate which is separated from the
CENOZOIC SE ASIA RECONSTRUCTIONS remainder of SE Asia by Andaman Sea spreading, the Sumatra and Sagiang Fault. Some of the earlier northward strain may have contributed to the rotation of the Shan-Thai and Indochina Blocks and suggested westward movement of Sumatra. The India-Asia Collision has been responsible for clockwise rotation of parts or all the Shan-Thai and Indochina Blocks, left-lateral slip on the Red River and probably for generation of at least the Oligocene part of the South China Sea floor. After accretion of the flysch belt onto the Indian craton it was carried north while displacement was taking place on the Wang Chao and Red River Faults. In the absence of definitive evidence of southward movement it is probable that, if there was extrusion, it was on a smaller scale than advocated by Tapponnier and his co-workers. While the present day tectonics of Asia are indicative of the largescale processes resulting from the India-Eurasia collision, determination of the time of elevation of the Tibet Plateau is crucial to resolving the kinematics of contending general models. If, in the first 20 Ma after the collision, most of the shortening was not taken up by thickening the crust of the plateau then extrusion must have occurred.
137
10 Ma
Rangin et al. (1990) t
,l
._.._I--I-
....... t
(a)
The Philippine Sea plate and AustraliaSE Asia collisions A number of regional reconstructions have treated the eastern region in some detail. Some of the diversity in interpretation is illustrated in the Miocene to basal Pliocene reconstructions from Rangin e t al. (1990), Daly e t al. (1991) and Packham (1990) in Fig. 10. The earlier regional study of the western Pacific by Jolivet e t al. (1989) has been largely incorporated in Rangin et al. (1990). The kinematics of the model of Rangin e t al. are illustrated in Fig. 1l. Reconstructions by Hall (1996) have made a major advance in the understanding of the regional tectonics. In the following discussion, the history of rotation of the Philippine Sea plate and the Philippine arc and the problem of the Celebes Sea is reviewed to provide background for the consideration of the origin and dispersal of the microcontinental blocks of Northern New Guinea and the Banda Sea, the impact of Philippine Sea plate rotation on Northern New Guinea and the Banda Sea, and the transmission of Pacific plate motion to Northern New Guinea and the Banda Sea.
(b) -
~
i:~?iii ~
~
~ ~ ~'~ ~
k,,%~*iiiiiiiiiii~i!N~
i;i!ii' ~/.~
o ',
,
The Philippine Sea plate and the Philippine archipelago Since 5 Ma the Philippine Sea plate has been rotating clockwise relative to Eurasia about a pole
(c) Fig. 10. Reconstructions of southeastern SE Asia: (a) for 10 Ma from Rangin et al. (1990); (b) for late Miocene time from Daly et aI. (1991); and (c) for 5 Ma from Packham (1990).
138
G. PACKHAM
Fig. 11. Kinematic diagram showing trajectories of the various blocks of Southeast Asia digitized and plotted in the reconstructions of Rangin et al. (1990). Plates are Australia (Aus), Caroline (Ca), Eurasia (Eu) and Philippine Sea (Ph).
located north of Japan at 48°N and 157°E at 5.45°/Ma (Seno et al. 1993). The Philippine archipelago straddles the convergent boundary between the Eurasia and Philippine Sea plates (Fig. 12). The left-lateral Philippine Fault which is no older than late early Pliocene (Aurelio et al. 1991), takes up the motion between the essentially anchored western flank of the islands and the oceanic plate to the east. The youthful Philippine Trench is a product of changing arc polarity. Palaeomagnetic data from islands east of the Philippines on the Philippine Sea plate and Ocean Drilling cores have a coherent pattern indicating a large scale clockwise rotation of the whole Philippine Sea plate (Hall et al. 1995b). Because the Philippine Trench is young, it is generally accepted that much of the archipelago was developed on the Philippine Sea plate rather the Eurasian margin. In his reconstructions Hall (1996) has separated Luzon and some of the islands to its south from the southern part of the Philippines and derived them from an extension of northern Borneo, assembling the island arc chain in early Miocene time by transferring Luzon to the Philippine Sea plate. This interpretation is based on the palaeomagnetic results of Fuller et al. (1991) who found a large counter-clockwise rotation in Luzon throughout most of the Tertiary with little change of latitude.
Hall et al. (1995a) estimated that the Philippine Sea plate has rotated c. 90 ° clockwise since 50 Ma. The pre-Pliocene rotation is in two stages. A rotation of 50 ° between 40 and 50 Ma about a pole located at 10°N and 150°E, followed by an interval with no significant rotation until 25 Ma when rotation recommenced, amounting to 34 ° about a pole at 15°N and 160°E (Fig. 13). These findings put severe constraints on the reconstruction of north New Guinea and the Banda Sea. The Miocene rotation may have been initiated by the collision of the northern edge of the Australian craton with the Philippine Sea plate. The general location of the Philippines has been plotted on the Fig. 13 on the assumption that they were on the Philippine Sea plate. The eastern part of the Philippine islands is substantially comprised of rocks built up by island arc processes from the Cretaceous onwards, on oceanic basement dating back to Late Jurassic (e.g. Geary et al. 1988). Continental basement of at least Jurassic age is present in the west around Mindoro and western Panay. It is an extension of the North Palawan Block on the South China Sea drifted margin. Pubellier et al. (1991) have identified a terrane in western Mindanao, including Zamboanga and the Daguma Range (Fig. 12), which may also be continental. The continental blocks, part of the Asia margin, were incorporated in the Philippines
CENOZOIC SE ASIA RECONSTRUCTIONS
Fig. 12. Major tectonic elements and sea floor anomalies in the Philippine-Banda Sea region modified from Hall et al. (1995b). Terranes of continental or suspected continental origin are west of the line in the Philippine Islands which includes Mindoro, west Panay and west Mindanao.
139
as the Philippine Sea plate rotated clockwise. The essentially transform boundary brought the south to southwest facing Philippine arc northward to collide obliquely with the Asian margin. A middle to late Miocene collision between the Palawan Block on the southern flank of the South China Sea has been interpreted differently by Holloway (1982), Sarewitz & Karig (1986), Rangin (1991) and Bird et al. (1993). Transpression and strike-slip continued from latest Miocene time to the present (Bird et al. 1993). The eastern and western terranes of Mindanao were juxtaposed before early Pliocene time by strike-slip movement (Pubellier et al. 1991). Middle to upper Miocene volcanic rocks in western Mindanao have been interpreted by Pubellier et aL (1991) as an east-facing arc on the western side of the Molucca Sea. The arc may have continued south through the Sangihe arc to similar age volcanic rocks in eastern SW Sulawesi (Coffield et al. 1993). Rangin et al. (1990) and Packham (1990) both show this as subduction of the Philippine Sea plate under the Sangihe arc. Daly et al. (1991) indicate subduction of Celebes Sea crust from the west (Fig. 11). An alternative and perhaps a better model, as discussed later, is that before the Banda arc was initiated, the Molucca Sea was part of a northward prolongation of the Australia plate as illustrated by Daly et aI. (1991) in their early Miocene reconstruction.
~ 1~,~ 5M '~ a ~ PII
•
L ~ ~ EURASIA - C.~s
•
~'
,0,
oo.
RotationPole 40-50 Mo _
/IgWm~";~..~:: ~.I~ ,k7 ~ S E A
? ~ y/ p l . ] ~ ~
~
PLATE
~
BASIN
m
j
~'~,,~ ~ - A U S T R A~L I A E 3 ~
1~
~"
~ . ~ . : _ ~ . ' ~
-,~--
~
~
~,.
SOLOMON
~
\1
Fig. 13. Schematic reconstructions of relationships in the Northern Australia region based on Philippine Sea plate reconstructions from Hall et al. (1995b) for 40, 25 and 15 Ma. Locations are: MLA, the Melanesian arc; PA,a possible Palaeogene Australia-Pacific arc between the Philippine Sea and Australia plates; MRA, the Maramuni arc; BA, the Banda arc; PI, the location of the Philippine Islands; and CB, the Banda Sea Continental Blocks.
140
G. PACKHAM
The Celebes Sea The Celebes Sea and the West Philippine Basin have overlapping ages (Fig. 12). Both contain middle Eocene oceanic crust with similar spreading rates. Weissel (1980) mapped magnetic anomalies 20 to 18 (46-52 Ma) which parallel the Sulu ridge and young to the south, into the North Sulawesi Trench. Anomaly trends differ because of the clockwise rotation of the Philippine Sea plate and possible counter-clockwise rotation of the Celebes Sea (Weissel 1980; Shibuya et al. 1989). The Celebes Sea could be part of a larger domain including southern Borneo and parts of the Malay Peninsular where Fuller et al. (1991) have also documented counter-clockwise rotations. The presence of contradictory results from Borneo and the dating of the rotation make interpretation controversial. Hall (1996) argues for the rotation of Borneo on geological and palaeomagnetic grounds. Silver & Rangin (1991) concluded that the Celebes Sea was either a marginal basin formed behind an arc or is a trapped fragment of a larger entity, including the Molucca Sea now separated from it by the narrow Sangihe arc. Similar sea floor ages are suggested by the lower middle Eocene cherts and limestones associated with ophiolites embedded in what is possibly a lower Miocene forearc terrane described by Moore et al. (1980) from Talaud Island in the Molucca Sea. Assuming the arc is not allochthonous the Celebes Sea has a backarc relationship to Palaeogene volcanic rocks in the north arm of Sulawesi (Ratman 1976) and the eastern part of the southwest arm (Sukamto 1982) where they have been broadly dated as lower Eocene to upper Oligocene. Hall (1996) has linked the Celebes and Molucca Seas to the West Philippine Basin to form a single basin behind a North Sulawesi-Philippine arc. Marginal basin formation was probably enhanced at anomaly 20 time when Australia increased its velocity away from Antarctica from c. 1 0 m m a-1 prior to anomaly 20, to about 30 mm a-I between anomalies 18 and 20 (Royer & Sandwell 1989).
Continental blocks of the Banda Sea and North New Guinea One obvious feature of the terrane distribution in northern New Guinea and the Banda Sea is that oceanic and island arc terranes are to the north of the continental terranes. Intervening between them are the West New Guinea Composite Block and the East Papua Composite Terrane. The Finisterre Terrane continues offshore into the uncollided Melanesian arc. East-west tectonic transport has resulted in duplication of the arc so that the Manus
segment overlaps the New Britain segment (Fig. 14). Three categories of continental blocks are found north of the Australian craton. (1) In the southern Banda arc from Timor to Tanimbar, parts of the Australian craton have been incorporated into the upper plate during collision. (2) North of the Australia plate boundary, in the Banda Sea region are rifted blocks including the basement of Sumba, Buton, Banggai-Sula, Buru, Seram, the Bird's Head, Obi and southern Bacan and submarine ridges south of Buru and Seram. This group could also include the metamorphic belt in the north, central and southeast arms of Sulawesi, and possibly some of the higher tectonic units in Timor. (3) Composite terranes comprising a mixture of continental, arc and oceanic terranes have been accreted onto New Guinea to form the West New Guinea Composite Block and the East New Guinea Composite Terrane. Apart from Sumba and central Sulawesi which are problematical the remaining blocks have an Australian derivation, but there is disagreement about their source, time of detachment from Australia and the means of their tectonic transport.
S u m b a a n d the central S u l a w e s i m e t a m o r p h i c belt
Sumba has a basement of Upper Cretaceous turbidites overlain unconformably by gently dipping Palaeogene shallow water sediments and volcanic rocks (Burollet & Salle 1981), and resembles the stratigraphy of the adjacent Asian margin in southwest Sulawesi and offshore east Java as described by Hasan (1991) and Bransden & Matthews (1992). Sumba is separated from them by the Flores Basin and was probably incorporated in the Banda forearc in later Palaeogene time. Van der Werff et al. (1994) and Rangin et al. (1990) also consider that Sumba was rifted off SE Asia and is a separate block from Timor. Another problematical unit, not usually regarded as one of the drifted blocks, is the metamorphic belt of central Sulawesi which contains blueschists and is overthrust on the east by the East Sulawesi ophiolite belt. A lower Cretaceous metamorphic date shows that these former continental margin deposits are older than the Upper Cretaceous turbidites of southwestern Sulawesi which overlie an ophiolite basement. Two phases of EoceneOligocene metamorphism have been identified in the blueschists (Parkinson 1991). The metamorphic belt extends down the west side of the southeast arm of the island where Helmers et al. (1990) have
I'
120OE
SE ASIA PLATE
Trerl~ _
I
126=
"
~
~,
BandaSea
1
ar•
Tanimb ~ ~J~
Tarera
I-~(---./l~a'rnahera %% ~i
PLATE
PHILIPPINE SEA
~J
AUSTRALIA
Ma
138"
PLATE
~
PLATE
West '
°
~.
L,
PLATE
PACIFIC
krn
CoralSea
150"
~:i~:i!ii!i!::~i~:.
~:~:i:i:?:i:i:i:?::::.:......-.. 156°E
/25 Ma~
/
!o/ ~ - .".etr~i~iiii;i:!:!:!:.'%..':i:::'.%.'.'.
iii:... PLATE
• BISMARCK
SOUTH
Finisterre E : ~ Manus ........ Terrane
,,..~ OnactiV°'
1,5Ma
Papuan ,~_ Basin
~ i : : : : . ~ ~::iii!ii~ ~ : : . : . . - .
~~,,ir
Torricelli Terrane
CAROLINE
I
500
Fig. 14. Plate tectonic framework of the New Guinea region modified from Struckmeyer et al. (1993) showing relative longitude of the former locations of the western edge of the Philippine Sea plate as reconstructed by Hall et al. (1995b).
~"s
<~umba
•4°N• ~.~ $ulawe~i
4~
142
G. PACKHAM
obtained an upper Oligocene metamorphic date. Palaeogene volcanic rocks on the eastern side of the southwest ann of the island reported by Ratman (1976) and faulting events recorded in the late Eocene and mid-Oligocene in the platform carbonate Tonasa Limestone to the west of the volcanic arc are compatible with a convergent boundary and a mid-Oligocene collision. Helmers et al. (1990) distinguish the event from the later collision of the Banggai-Sula Block in northern east Sulawesi which Davies (1990) dates as latest Miocene. It can also be distinguished from the mid-Miocene collision of Buton-Tukang Besi with southeast Sulawesi (Davidson 1991). Hall (1996) envisages transfer of the metamorphic belt and the ophiolite from the India plate and to the Eurasia plate at 25 Ma. The eastern and central Sulawesi terrane has general similarity to the West New Guinea Composite Block, which contains amongst other things non-metamorphosed, continental slope sediments, low to medium grade metamorphic rocks of continental affinity, blueschists and ophiolites (Struckmeyer et al. 1993). The reconstruction of Daly et al. (1991) shows eastern and central Sulawesi forming in the oceanic domain, south of the Philippines and colliding with western Sulawesi in the late Miocene (Fig. 10). Although explaining the late Miocene tectonic episode in western Sulawesi, it is incompatible with the continental margin origin and tectonic history of the metamorphic belt.
Microcontinental blocks o f the Banda Sea and the Bird's H e a d
Regionally, the northern Australian craton basement can be divided into two types. In Papua New Guinea there are deformed Palaeozoic sediments and Palaeozoic to Triassic acid intrusive rocks. West, in Irian Jaya and on the northern Australian margin, undeformed Palaeozoic sediments overlie Precambrian basement. The microcontinental blocks of the northern Banda Sea and the Bird's Head characteristically have metamorphic pre-Mesozoic basements and are of the eastern type. Above basement is a Triassic clastic or carbonate shallow marine succession, overlain by deeper water Jurassic to Palaeogene sediments, passing into Cretaceous to lower Palaeogene bathyal to pelagic deposits. The preserved upper Mesozoic to Neogene successions are remarkably continuous, indicating little tectonic disturbance. The stratigraphy of the blocks has been illustrated by Audley-Charles (1988). A lower Jurassic (Pliensbachian-Sinemurian) breakup unconformity in the west Papuan Basin indicates that Indian Ocean sea floor extended east
60N
30N
30S
60S-
120E
180E
240E
Fig. 15. Relative motions of the Australia, Pacific and Eurasia plates in the early Palaeogene (64-56 Ma) from Gordon & Jurdy 0988).
at least to eastern New Guinea at that time (Pigram & Symonds 1991). Because of the divergence between Australia and Pacific motions prior to 43 Ma (Fig. 15) determined by Gordon & Jurdy (1988), it is unlikely that Pacific Ocean spreading would have extended far west of Papua New Guinea. Consequently the Australia plate boundary probably ran west of north to the Asian margin or the southern boundary of the spreading systems of the Celebes and Molucca Seas and West Philippine Basin. Since Phase 2 Indian Ocean sea floor spreading (Late Cretaceous to middle Eocene) may have extended east to western Sumatra (Fig. 4) Jurassic to mid-Cretaceous crust of the Australia plate may have extended from western New Guinea to western Sumatra from the Late Cretaceous to the middle Eocene. Palaeomagnetic studies on the microcontinents suggest that they were separated from the Australian craton by Late Cretaceous time. Ali & Hall (1995) concluded that palaeomagnetic results from an Upper Cretaceous marl (86_ 3 Ma) at a site on Sula indicated a formation palaeolatitude of 19 _+6°S, well to the north of Australia. Wensinck et al. (1989) obtained similar results from palaeomagnetic measurements on Santonian and Maastrichtian sediments from Misool. The palaeolatitudes obtained were 18.5 _ 3.2 ° and 21.5 _ 3.6 ° respectively. Reconstructions by Struckmeyer et al. (1993) place the Late Cretaceous Papuan margin at a palaeolatitude of 30°S, roughly 10°S of the microcontinents at that time. The north-south
CENOZOIC SE ASIA RECONSTRUCTIONS
relationship to Australia has been substantially restored by shortening in the Banda subduction zone. The Sula site and the Santonian Misool sites both give counter-clockwise rotations of around 40 ° but the Maastrichtian site on Misool is not rotated. Ali & Hall (1995) also found strong counter-clockwise rotations in Upper Cretaceous, upper Oligocene and middle Miocene volcanic rocks from north Obi. They relate the rotations to left-lateral shear on the left-lateral Strong Fault Zone since as the sites are adjacent to, or between splinters of the fault. There is no clear evidence of latitudinal separation of the Kemum Terrane of the Bird's Head from the Australian margin (Giddings et al. 1993). By comparing palaeomagnetic results from the Kemum Terrane with the Australian polar wander path they suggested that the terrane, originally located in eastern New Guinea, was rotated 55 ° counter-clockwise between Early Jurassic and middle Eocene time, then in the late Eocene and early Miocene interval moved westward to near its present location. Finally there was 10° of postmiddle Miocene counter-clockwise rotation. Unfortuiaately Cretaceous sediments did not yield poles because of strong overprinting. Their data are open to an alternative interpretation. The westward movement of the terrane to the Bird's Head with a decreased of palaeolatitude and counter-clockwise rotation could have taken place between the Early Jurassic and middle Eocene time followed by a 20 ° late Eocene and early Miocene clockwise rotation. Because the microcontinental blocks and Bird's Head terranes were were not involved in Oligocene tectonism of the New Guinea orogen, they were probably located to the west of the cratonic part of Irian Jaya by early Oligocene time as indicated on Fig. 13. From the foregoing discussion of palaeomagnetic data, Jurassic rifting may have removed them from eastern New Guinea to reach the indicated location by Late Cretaceous time. Hall (1995), Struckmeyer et aL (1993) and Rangin et al. (1990) all use a generally similar location for the blocks in their reconstructions. Using the palaeomagnetic interpretation of Giddings et al. (1993), Struckmeyer et al. (1993) moved the blocks westward by a complex Late Cretaceous to Palaeocene spreading system, followed by southwest movement on a major leftlateral southwest strike-slip fault. It is difficult to relate these events to known plate boundaries and vectors. In their reconstruction Daly et al. (1991) rotated Seram and Buru counter-clockwise to their present position from a location on the northwest Australian margin adjacent to Timor extending north to the Bird's Head of New Guinea. Banggai-Sula derived from the Birdfs Head is
143
located immediately to the north (Fig. 10). In their tectonic evolution, Banggai-Sula is transported west in the late Miocene time. Southward directed deformation in the Bird's Head and left-lateral strike-slip on the Tarera Fault are responsible for rotation of the Burn and Seram, arriving at their present orientation as Pliocene underthrusting occurred in the Seram Trough. This scheme avoids the problem of moving the blocks west from the Papua New Guinea margin but their pre-Triassic metamorphic basement (Kemp & Mogg 1992) indicates affinities with the eastern Australian tectonic province. The location does not conform to the Cretaceous palaeolatitude data but the sense of rotation is in agreement with the measurements. Packham (1990) assumed that initial rifting of the blocks took place in the Mesozoic but that they remained close to the New Guinea margin as marginal plateaus away from clastic deposition. About the beginning of Miocene time, they were transferred to the Philippine Sea plate and started to move west on the northern margin of the developing New Guinea Orogen (Fig. 10). The relative continuity of deposition non-clastic deposition makes this process problematical and in the case of Buton an origin in northern New Guinea in the vicinity of the Bird's Head is assumed. This would involve an earlier episode of westward movement. Again the palaeolatitude determinations do not fit.
Models for North New Guinea and the Banda Sea Although the tectonics of the New Guinea-Banda are driven principally by the west-northwest convergence of the Pacific plate and the northwardmoving Australia plate, the Pacific plate and the Australia plate are not in direct contact in the region. They are separated by the southern end of the Philippine Sea and Caroline plates. The Caroline plate (Fig. 14) was formed in Oligocene time by north-south spreading when, in the Hall et al. (1995b) model, the Philippine Sea plate was stationary. The Caroline plate has a slow counterclockwise motion relative to the Pacific plate about a pole located about 5 ° north of it (Weissel & Anderson 1978). It has been rotating away from the Philippine Sea plate since about 12 Ma from near the apex of the southward opening Ayu Trough and is being slowly subducted under the Philippine Sea plate in the Palau Trench (Fig. 11). Early Miocene volcanism (20 Ma) reported from Palau by Haston et aL (1988) suggests that it may have been part of the Pacific plate before 12 Ma. Caroline plate motion is transmitted westward into the seismically active New Guinea Highland
144
G. PACKHAM
region (Abers & McCaffrey 1988) and thence westward into the Banda Sea. In the south Banda region the strain is principally absorbed by thrusting into the Seram Trough but some of it is transmitted westward along the Sorong Fault Zone into Sulawesi and eventually to the North Sulawesi Trench (Silver et al. 1983). West of the Bird's Head, the northward motion of the Australia plate is absorbed in the Tanimbar-Timor trough, backarc thrusting north of Wetar and north-south shortening of the Banda Sea (McCaffrey 1988). The present tectonic configuration, dominated by strong left-lateral shear has led to models in which terranes were transported westward under its influence. Hamilton (1979), Charlton (1986) and Packham (1990) considered that the blocks were driven west from northern New Guinea along with the development of the Sorong Fault. As discussed above models which propose that the microcontinental blocks of the Banda Sea extended west from the Bird's Head before the collision of arc elements with north New Guinea are more probable. Collision tectonics has dominated the explanations for the formation of the New Guinea Orogen (Pigram & Davies 1987; Struckmeyer et al. 1993) but recently extensional tectonics for the Late Palaeogene phase has been invoked by Hill et al. (1993). The Philippine Sea p l a t e a n d N o r t h N e w G u i n e a in the P a l a e o g e n e
In the Melanesian arc middle Eocene volcanic rocks resting on oceanic basement mark the onset of volcanism. The identified western limit of the arc is the Finisterre Terrane of northeast New Guinea (Fig. 14). Distinct from it to the west is the Torricelli Terrane, in which Cretaceous to Oligocene volcanic activity is recorded (Fig. 14). A similar late Cretaceous commencement of arc volcanism is also found nearby in Halmahera (Hall et al. 1991; Ali & Hall 1995) and further afield in the Philippines. Similar volcanic rocks may be preserved in the Arfak, Waigeo, Tosem, Biak/Superiori and Gautier Terranes now located on and adjacent to north New Guinea (Pigram & Davies 1987; Charlton et al. 1991; Struckmeyer et al. 1993). This volcanic assemblage, referred to here informally as the northern arc, may have formed a south-facing arc on the margin of the young Philippine Sea plate as reconstructed by Hall et aI. (1995b) and Hall (1996). The development of the arc commenced before the Eocene rotational phase of the plate and continued through the late Eocene and Oligocene non-rotational phase. The collision of the West New Guinea Composite Block with the New Guinea margin in mid-Oligocene time as Australia moved north has
been regarded as the first stage of the development of the North New Guinea Orogen (Pigram & Davies 1987, Struckmeyer et al. 1993). In their models it had previously been rifted off and collided over a north-dipping subduction zone. In view of the proximity of the southern boundary of the Philippine Sea plate to the northern New Guinea margin at the time, as depicted by Hall et al. (1995b), an alternative explanation is that emplacement of the West New Guinea Composite Block was the result of a collision between part of the Philippine Sea plate and the New Guinea margin. The Torricelli Terrane and the other volcanic terranes to the west listed above may have originated from the margin of the Philippine plate, some of them from the gap between the present Philippine Islands and Halmahera, being left behind during Miocene clockwise rotation of the Philippine Sea plate. Hill et al. (1993), taking a very different approach, interpreted the West New Guinea Composite Block as part of the continental margin in which extension has exposed mid-crustal rocks as core complexes. The Sepik and Ramu Basins to the north of the block were initiated as half-graben in late Oligocene time. The model requires continuous slow southward late Eocene to late Oligocene subduction under the New Guinea margin with some compressional deformation. Most of the convergence was taken up to the north, away from the New Guinea margin. This explains the development of the composite block before the late Oligocene docking of the arc terranes commenced of which the first was the Torricelli Terrane. M i o c e n e B a n d a Sea tectonics
As a consequence of Miocene clockwise rotation of the Philippine Sea plate, Ali & Hall (1995) have determined that at its contact with the northern New Guinea, the plate boundary was a left-lateral strikeslip fault. The northward component of the Philippine Sea plate vector approximately balances the northward Australia plate vector leaving a resultant left-lateral movement of the Philippine Sea plate initiating westward movement of blocks on the Sorong Fault. Until Philippine plate motion changed at the end of the Miocene, strike-slip relationship could not have persisted far west. Westwards along the Sorong Fault, the distance from the closely located Euler Pole of rotation of the Philippine Sea plate increases hence the velocity vector increases and has a more northerly azimuth while the Australia plate vector is nearly constant. The slip direction would swing towards the north. In a simple model of the Philippine Sea plate in
CENOZOIC SE ASIA RECONSTRUCTIONS which the boundary swung to the north, west of Halmahera as envisaged by Rangin et aL (1990) and Daly et al. (1991) the Molucca and Banda Sea floors were on the Australia plate. They could be ocean crust (Fig. 13) trapped when the Banda arc formed by eastward extension of the Sunda arc. The Banda Sea at least has heat flow levels and oceanic sea floor depths appropriate for a late Mesozoic age of the Molucca Sea as discussed above. A rough minimum estimate of the age of the Banda arc can be obtained from the 1000 km length of the subducted slab in the vicinity of Timor (McCaffrey 1989). Using a convergence rate calculated from 7 9 m m a -1 calculated from NUVEL-1, the age is over 12 Ma. Lavas on Wetar range from 7.7 Ma to 3 Ma but granodiorite and diorite intrusive rocks on the island are 12 Ma old (Abbott & Chamalaun 1981). The maximum age is possibly 15 Ma. Simple restoration of the 1000 km of convergence of the Banda arc on the west and 200 km on the east would place the microcontinental blocks in a line running west-northwest from the Bird's Head at the time of the onset of subduction (Fig. 13). In the simple model proposed above, collision of the Buton-Tukang Besi block in the middle Miocene could have triggered the westward extension of the Sunda arc, initiating the Banda arc creating a transitory Banda-Molucca plate. As subduction on its eastern extension was limited by the Australian margin, the crust of the Banda Sea was rotated counter-clockwise about a pole located near the western end of the Timor Trough, isolating the Molucca Sea. With two subduction boundaries and probably a transform southern boundary, the Molucca plate became a triangle of oceanic crust larger than at present, able to adjust to changes in convergence of the surrounding plates. Banda Sea rotation provides an alternative explanation for counter-clockwise palaeomagnetic rotations observed on the microcontinental blocks. Seram may have been further disturbed by thrusting into the Seram Trough. Banda Sea rotation was accompanied by subduction under eastern Sulawesi, so that Sula collided with the Sulawesi ophiolite belt in latest Miocene time, later than Buton. A difficulty with this rotation model is that the pre-rotation location of the microcontinental blocks (Fig. 13) overlaps the Molucca Sea crust shown to the west of Halmahera in the 25 Ma and 15 Ma reconstructions of Hall et al. (1995b). If the Philippine Sea and Australia plates had been a degree or so further apart or a slightly different shape, the overlap would be removed. Alternatively westward motion may have redistributed the blocks to some degree. A further difficulty is that it does not explain the upper Miocene volcanic rocks and granites in central West Sulawesi.
145
Some of the wide diversity in interpretation of the Banda Sea region is illustrated in Fig. 10. Rangin et al. (1990) reconstructed the Banda arc as an east-west arc, formed in the middle Miocene, attached at the eastern end to western New Guinea at the northern margin of the craton. The subduction zone was south of a rifted continental block which was incorporated in Timor. The western end of the arc was joined onto the eastern Sunda arc in the Pliocene as Australia moved north. In their reconstructions Daly et al. (1991) subducted the old crust Banda crust under eastern Sulawesi as it moved westward, driven indirectly by Pacific plate motion, to collide with western Sulawesi in late Miocene time. In their model the Banda arc was apparently formed after the collision. The change of direction of Philippine Sea plate motion to close to west at 5 Ma (Seno et al. 1993) introduced a new tectonic regime which will be discussed in the next section.
Caroline~Pacific plate motion and eastern Indonesia
The collision of the Ontong Java Plateau on the Pacific plate with the Melanesian arc in the Solomon Islands (Packham 1973; Vedder 1986) late in early Miocene time was an event remote from eastern Indonesia but its ultimate consequences extend east to the Banda Sea. The collision reversed the polarity of the Melanesian arc, transferring it from the Australia plate to the Pacific plate which transported it westward to collide with northern New Guinea in late Miocene time (Pigram & Davies 1987). With the western end of the arc now identified as the Finisterre Terrane fixed, westward transport ruptured the arc duplicating it for about 1000 km between New Britain and Marius (Fig. 14). This amount of overlap required early late Miocene rupturing to have occurred. Although the commencement of terrane docking has not been dated precisely, deformed upper Miocene sediments in the Ramu Basin are overlain by a thick lower Pliocene to Pleistocene successor basin deposit (Hill et al. 1993). The progressive exposure of northern New Guinea to convergence between the Caroline and Australia plates is illustrated in Fig. 14. At 15 Ma, the eastern edge of the Philippine Sea plate was near the future site of docking of the Finisterre Terrane. Before the Finisterre Terrain docked there was no significant coupling between the Philippine Sea and Australia plates except in the vicinity of the boundary (the Sorong Fault). As already mentioned, the northward component of the motion of the Philippine Sea plate and the Australia plate approximately balanced, leaving a residual left-
146
G. PACKHAM
lateral strike-slip. The motion of the Caroline (or Pacific) plate could not be transferred to the Australia plate until docking of the Finisterre Terrane commenced (at about 7 Ma) because the Melanesian arc was south-facing. From about 7 Ma combined crustal shortening and lateral translation extended f r o m the docked part of the Finisterre Terrain to the eastern boundary of the Philippine Sea plate. As the Philippine Sea plate moved west, the interface increased in length from around 600-800 km as the terrane docked to its present length of about 1200 km by 3.8 Ma when northern segment of the arc started moving west with the Caroline plate and extension commenced in the Manus Basin, isolating the South Bismarck plate from the Solomon plate (Taylor 1979). By 4.0 _+ 0.5 Ma, thrusting and uplift had exposed sediments in the Muller Anticline in the western part of the Papuan Thrust Belt (Hill 1991). Hill (1991) estimated the amount of shortening in the Papuan fold belt at c. 100 km. From the topography, Abers & McCaffrey (1988) suggested it was 6 1 k m east of 140°E and 44kin in the western highlands. These latter figures do not take account of loss from erosion. The convergence rate across the Australia-Caroline plate Boundary at 140°E and 2.5°S, is 97 mm a -1 towards 21°S of W (calculated from NUVEL-1 and the CarolinePacific pole of Weissel & Anderson 1978). This is equivalent to north-south and east-west components of 47 and 85 mm a-1 respectively. The amount of north-south shortening in the east is about 325 km. In a Finisterre-west Papuan Fold Belt transect there is probably c. 200 km of shortening across the fold belt and thrusting of the Finisterre Terrain onto the margin identified by Pegler et al. (1995). This accounts for about 4 Ma of shortening. But much of the thrusting in the eastern part had been completed by 4 Ma possibly indicating a start at 6-7 Ma. This leaves over 100 km of shortening unaccounted for. In eastern Irian Jaya shortening has been only possible for a shorter time, probably less than 5 Ma, suggesting a possible 240 km of shortening. As expected, the thrust belt narrows westwards and the highlands are of smaller mass. Even if the Abers & McCaffrey estimate of 44 km of convergence is doubled allowing for erosion, it still leaves about 150 km to be accounted for. The excess can probably be accommodated in underthrusting into the North New Guinea Trench and its former extension in the east. Seismic activity in the New Guinea orogen, indicates that much of the strain is taken up in the west by left-lateral strike-slip, especially near the southern margin and while in the east, thrusting is more common (Abers & McCaffrey 1988). In order to account for the thrust direction in the Seram
Trough, McCaffrey (1988) concluded that the Bird's Head is decoupled from the Australia plate and is moving west or southwest with respect to it, probably along the east-west left-lateral Tarera Fault, south of the Bird's Head. The western end of the fault forms a triple junction between the Bird's Head, Banda Sea and the Australia plate where it meets the Seram and Tanimbar Troughs. The left-lateral component of convergence between the Caroline and Australia plate distributed across the orogen, is c. 600 km in the east, if strain was effectively initiated at about 7 Ma and possibly 400-450 km in the west if commencement was c. 5 Ma. Some of the strain has been absorbed in the eastern Bird's Head in the Lengguru Thrust Belt where orogenic sedimentation commenced in the Pliocene (post-N18, c. 5 Ma) and thrusting continued till at least 2 Ma (Moffat et al. 1991). The remainder of the Bird's Head has behaved as a single entity throughout the Neogene, bounded on the north and south by the Sorong and Tarera left-lateral strike-slip faults. This is kinematically consistent with a post-middle Miocene 10° counter-clockwise rotation of Kemum Terrane identified by Giddings et al. (1993). Much of the strain was probably taken up by westward or southwestward movement of the Bird's Head into the Seram Trough where the seismic zone beneath Seram extends to 300 km (McCaffrey 1988). On stratigraphic evidence from Seram deformation commenced at 5 Ma and continued throughout Pleistocene time (Kemp & Mogg 1992). Possibly 150km of the westward movement of the Bird's Head is expressed in leftlateral displacement of the thrust front on the Tarera Fault. The displacement could have been achieved in less than 2 Ma if the strain was concentrated on the fault. Apart from thrusting in Seram associated with a strike-slip component on the Seram Trough, only an insignificant amount to the convergence southwest of the trough has been transmitted to the southern part of the Banda Sea. Shortening of over 400 km, commencing at c. 5 Ma, can be accounted for in the undeahrust slab beneath Seram plus the shortening in the Lenggum Thrust Belt to which could be added a contribution from oblique subduction into the Noah New Guinea Trench. North of the Seram Trough, shortening has been transmitted to the west into Sulawesi between the Seram Trough and within strands of the Sorong Fault, absorption can be identified in the thrusting in the Banggai-Sula block and East Sulawesi between 5.2 and 3.8 Ma (Davies 1990) and an estimated 250 km of left-lateral slip on the Matano-Palu Faults Sulawesi (Silver et al. 1983). Finally, it should be noted that large-scale tectonic transport of terranes from eastern New
CENOZOIC SE ASIA RECONSTRUCTIONS Guinea to the Banda Sea region could not have been driven by the Caroline plate interacting with New Guinea because the amount of convergence was too small. Further, the Buton-Tukang Besi Block had already collided with eastern Sulawesi and the Sula Block was close to doing so when interaction commenced close to the end of Miocene time.
Eastern SE Asia: conclusions The history of rotation of the Philippine Sea plate has been recently refined by Hall et al. (1995b). Its initial southem latitude location places constraints on reconstruction models of the Banda Sea and the development of the New Guinea Orogen. Two earlier phases of clockwise rotation have been identified prior to the present one. Rotation about the present pole, located northeast of Japan results in westward motion at the southern end of the plate. The rotations about the earlier poles located to the east of the Philippine islands total 90 ° . The first was between 50 and 40 Ma and the second between 25 and 5Ma. The second phase is the more important in regional reconstructions. The Miocene rotational phase has been responsible for the middle to late Miocene collision of the Philippine chain with the Asian Margin in Mindoro and western Mindanao. The colliding arc, with the possible exception of Luzon, developed on the margin of the Philippine Sea plate before Miocene rotation phase. Hall (1996) suggests Luzon developed on the margin of the Eurasia plate and was transferred to the Philippine Sea plate early in Miocene time. The crust of the Celebes and Molucca Sea may have been formed as an eastward Eocene extension of the West Philippine Sea spreading system, north of an arc system linking Sulawesi and the Philippines. There is inconclusive evidence that a domain including the Celebes Sea, Borneo and the Malay Peninsular underwent subsequent counterclockwise rotation. Although most, if not all of the continental rocks in the Timor-Tanimbar orogen have been sliced from the Australian northwest shelf margin during its Pliocene collision with the Banda arc, the basement of Sumba straddling the Banda forearc has probably been rifted off the adjacent Asian margin in the Palaeogene time. The nearby composite block of the central Sulawesi metamorphic belt and the East Sulawesi ophiolite may have originated on the Australia plate and been emplaced by an Oligocene collision with West Sulawesi. Microcontinental blocks of the Banda Sea and the Bird's Head have an intruded and metamorphosed PermoTriassic basement which suggests an origin in eastern New Guinea.
147
Late Cretaceous palaeolatitude differences of around 10 ° between New Guinea and Sula, Seram and Misool and an Early Jurassic breakup unconformity in the West Papuan Basin indicate the blocks were probably rifted off at the time of breakup. The breakup and the Mesozoic ocean crust off western Australia suggest that Phase 2 Indian Ocean crust of the Australia plate probably extended from eastern New Guinea to the Asian margin and west to Sumatra, and the microcontinental blocks were contained in it. The lack of tectonic disturbance in the upper Palaeogene to lower Neogene stratigraphy on the blocks indicates they were assembled from the Bird's Head westward before that time away from the developing New Guinea Orogen. The close proximity of the Philippine Sea plate to north New Guinea as reconstructed by Hall et al. (1995b) makes it geometrically improbable that the blocks were transported to the southwest on the Pacific plate after 43 Ma as proposed by Struckmeyer et aL (1993). The first tectonism in the New Guinea Orogen occurred in mid-Oligocene time when the Philippine Sea plate was stationary and well after middle Eocene increase in northward velocity of the Australia plate. According to Pigram & Davies (1987) the West New Guinea Composite Block comprising variously deformed Mesozoic continental margin sediments, granitic intrusions, metamorphic rocks, extrusive rocks and ophiolite masses collided with the New Guinea continental margin. Alternatively it could have been a collision between the New Guinea continental margin and the edge of the Philippine Sea plate. A third possibility suggested by Hill et aL (1993) is that the composite block is an extensional core complex structure developed during Oligocene to early Miocene time in rocks already located on the New Guinea margin. The Torricelli Terrane, north of the composite block, is part of what is interpreted here as the Northwest arc developed on the south facing edge of the Philippine Sea plate. The arc which includes the Philippines, Halmahera and the various terranes of north New Guinea from the Torricelli Terrane west has a volcanic record extending back to Cretaceous time. It appears that docking the Torricelli Terrane on the north New Guinea margin was a late Oligocene event. It may have been detached from the Philippine Sea plate when Miocene clockwise rotation commenced and leftlateral movement on the Sorong Fault was initiated. The present author's preferred pre-15 Ma model is for the Philippine Sea plate boundary to curve northwest along the arc along the strike-slip direction against the Australia plate extended north to include the Molucca and Banda Seas prior to the
148
G. PACKHAM
formation of the Banda arc. The 1000 km long subduction zone under the west Banda arc indicates it was formed at c. 15 Ma presumably as an eastward extension of the East Sunda arc, transferring the Molucca and Banda Seas to the Eurasia plate from the Australia plate. Subduction may have been initiated by the collision of Buton with southeast Sulawesi. It would have reduced the latitude differences between the microcontinents and Australia. At about 12 Ma the Molucca plate became a separate entity when eastward subduction under Halmahera started. Australian continental crust first impacted at the east end of the Banda arc and has been subducted only 400 km or so. Differential subduction would have rotated the Banda Sea counter-clockwise, colliding BanggaiSula with western Sulawesi. This model does not explain the late Miocene volcanic and plutonic event in West Sulawesi. The remote collision of the Ontong Java Plateau with the north facing Melanesian arc north of the Solomon Islands set in train events that spread west to Sulawesi. The Melanesian arc has a much shorter history than the western arc having commenced forming on oceanic crust in middle Eocene time. Polarity reversal of the Melanesian arc transferred it to the Pacific plate. Its eastern end eventually collided with northeastern New Guinea to be emplaced as the Finisterre Terrane possibly c. 7 Ma. With the elimination of the subduction zone(s) between them, the convergence between the Caroline and Australia plates was transmitted across the New Guinea margin for the first time. At 7 Ma the eastern edge of the Philippine Sea plate was at about 142°E and c. 600--800 km of the margin was exposed to shortening. This distance progressively increased to c. 1200 km at c. 4 Ma when the Philippine Sea plate boundary was at 136°E. Combined strike-slip and spreading then commenced in the Manus Basin carrying the Manus arc segment west on the Caroline plate. Folding and strike-slip faulting in the New Guinea Orogen spread from east to west. Assuming that convergence commenced at c. 7 Ma and 5 Ma in the east and west respectively the shortening components are c. 330 and 240 km respectively. This can be accommodated in thrusting under the Finisterre Terrane, thrusting in the orogen and underthrusting into the North New Guinea Trench. The strike-slip component in the west is about 430 km. Some of this was taken up in the Lengguru Thrust Belt and possibly by oblique subduction into the North New Guinea Trench and within the orogen. The remainder must have moved the Bird's Head underthrusting it into the Seram Trough and moving it west on the Tarera Fault. McCaffrey (1988) concluded that Bird's Head is at present moving west or southwest with respect to
Australia. North of the Seram Trough there may be partial coupling with the Philippine Sea plate, transmitting strain westward to the thrusts between Banggai-Sula and East Sulawesi and to the PaluMatano fault system where Silver et al. (1983) have suggested there is 250 km of left-lateral movement. More important than explaining the distribution of the strain is the observation that the total amount is insufficient to transport blocks from eastern New Guinea to the Banda Sea region.
Concluding remarks The way in which the strain identified in the neotectonics of central Asia is being distributed in eastern Asia is still uncertain and hence its applicability to the rather different Palaeogene to Early Neogene tectonic regime is yet to be resolved. Kinematics of models partitioning Palaeogene shortening between crustal thickening of Tibet and block extrusion will not be determined until the time of elevation of the Tibet Plateau is confidently known, although substantial displacements have been documented on the Red River and Wang Chat Faults. The style of regional deformation indicated by the dispersal of palaeomagnetic results suggests that simple block rotation does not apply in the northwest Indochina block but it could well be an important process further east of the syntaxis. The regional geology north and east of the of the eastern syntaxis has not yet revealed the link between the Shan-Thai block and the terranes of eastern Tibet and the continuity of the eastern Himalayas to northern Myanmar except in broad terms. The related late Palaeogene palaeogeographical problem of the Indoburman flysch belt could be resolved by a better definition of the time of its docking south of the eastern syntaxis as well as an assessment of its former location on the Asian margin. Although the two sections of this paper have been treated separately, they are interdependent in that reconstruction of the eastern part of SE Asia requires knowledge of the location of the eastern Asian margin from late Oligocene to early Miocene time. One of the key pieces in the jigsaw is Borneo which has been alluded to only in passing. The two major unresolved problems are firstly, the relationship between the South China Sea Cenozoic spreading and deformation of the Rajang flysch belt, and secondly, the implications of possible counter-clockwise rotation of a domain including the island and adjacent regions and it relationship to Sundaland basin histories. While significant advances in the elucidation of the history of the rotation of the Philippine Sea plate have been made, its position relative to Asia is not well constrained as evidenced by the recent
CENOZOIC SE ASIA RECONSTRUCTIONS revision o f the location of the present Euler pole of relative motion by SenD et al. (1993). Allied to the relationship o f the Philippines to the Philippine Sea plate and the Asian margin is the absence of a detailed tectonic model for the Philippine arc from the Cretaceous onwards. A lack of highly detailed stratigraphic and palaeomagnetic studies in the microcontinental blocks, volcanic and oceanic terranes in the Banda Sea and along the northern margin of N e w Guinea hampers a more complete interpretation of the convergence and collision
149
history, as do the resolution o f the origin o f the West N e w G u i n e a C o m p o s i t e Block, the central Sulawesi Metamorphic Belt and the age of the volcanic rocks of the Banda Sea and the Sangihe arc. I thank the University of London SE Asian Research Group for sponsoring my attendance at the 1994 conference on the Tectonic Evolution of SE Asia and R. Hall in particular for his assistance and the provision of unpublished manuscripts. Petroconsultants (Australia) Pty. Ltd. have funded the preparation of the text figures.
References ABBOTT, M. J. & CHAMALAUN, F. H. 1981. Geochronology of some Banda Arc Volcanics. In: BARBER, A. J. & WIRYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication, 2, 253-268. ABERS, G. & MCCAFFREY,R. 1988. Active deformation in the New Guinea fold and thrust belt: seismological evidence for strike-slip faulting and basement involved thrusting. Journal of Geophysical Research, 93, 13 332-13 354. ALl, J. R. & HALL, R. 1995. Evolution of the boundary between the Philippine Sea plate: palaeomagnetic evidence from eastern Indonesia. Tectonophysics, 251, 251-275. ALLEN, C. R., GILLESPIE,A. R., HAN YUAN, SLED,K. E., Zhuang Buchum & Zhu Chengnan. 1984. Red River and associated faults, Yunnan Province, China. Quaternary geology, slip rates and seismic hazard. Bulletin of the Geological Society of America, 95, 686-700. AUDLEY-CHARLES,M. G. 1988. Evolution of the southern margin of Tethys (North Australian region) from Early Permian to Late Cretaceous. In: AUDLEYCHARLES, M. G. & HALLAM, A. (eds) Gondwana and Tethys. Geological Society, London, Special Publication, 37, 79-100. AURELIO,M. A., BARRIER,E., RANGIN,C. & MULLER,C. 1991. The Philippine Fault in the Late Cenozoic tectonic evolution of the Bondoc-MasbateN. Leyte area, central Philippines. Journal of Southeast Asian Earth Sciences, 6, 221-238. AVOUAC, J. P. & TAPPO~IER, P. 1993. Kinematic model of active deformation in central Asia. Geophysical Research Letters, 20, 858-898. - - , BAI, M., YoY, H. & WANG, G. 1993. Active thrusting and folding along the northern Tien Shun and the Late Cenozoic rotation of Tarim relative to Dzunggaria and Kazakhstan. Journal of Geophysical Research, 98, 6715-6744. BENDER, F. 1983. Geology of Burma. Gebuder Borntraeger, Bedim BESSE, J. & COURTILLOT,V. 1991. Revised and synthetic apparent polar wander paths of the African, Eurasian, North American and Indian plates, and time polar wander since 200Ma. Journal of Geophysical Research, 96, 4029-4050. BIRD, P. R., QUINTON,N. A., BEESON, M. N. & BRISTOW, C. 1993. Mindoro: a rifted microcontinent in
collision with the Philippines volcanic arc; basin evolution and hydrocarbon potential. Journal of Southeast Asian Earth Sciences, 8,449-468. BRANSDEN, P. J. E. & MATTHEWS,S. J. 1992. Structural and stratigraphic evolution of the East Java Sea, Indonesia. In: Indonesian Petroleum Association, Proceedings 21st Annual Convention. 417-454. BRtAIS, A., PATRIAT,P. & TAPPONNIER,P. 1993. Updated interpretation of magnetic anomalies and reconstructions of the South China basin: implications for the Tertiary evolution of Southeast Asia. Journal of Geophysical Research, 98, 6299-6328. BUROLLET, P. E & SALLE, C. 1981. Tectonic framework of Eastern Indonesia. In: BARBER, A. J. & WIRYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication, 2, 49-52. CHARLTON, T. R. 1986. A plate tectonic model of the eastern Indonesia collision zone. Nature, 319, 394-396. --, HALL, R. & PARTOYO,E. 1991. The geology and tectonic evolution of Waigeo Island, NE Indonesia. Journal of Southeast Asian Earth Sciences, 6, 289-298. CHEN, P., CHEN, Z. Y. & ZHANG, Q. M. 1993. Sequence stratigraphy and continental margin development of the northwest shelf of the South China Sea. AAPG Bulletin, 77, 842-862. CHEN, Y., COURTILLOT,V., COGNE, J-P., BESSE, J., YANG, Z. & ENKIN, R. 1993. The configuration of Asia prior to the collision of India: Cretaceous paleDmagnetic constraints. Journal of Geophysical Research, 98, 21,927-21,942. COFFIELD, D. Q., BERGMAN, S. C. GARRARD, R. A., GuRrrNo, N., ROBINSON,N. M. & TALBOT,J. 1993. Tectonic and stratigraphic evolution of the Kalosi PSC area and associated development of a Tertiary petroleum system, South Sulawesi. In: Indonesian Petroleum Association, Proceedings 22nd Annual Convention. 333-356. COPELAND,P., HARRISON,T. M., PAN YUN, KIDD, W. S. E, RODEN, M. & ZHANG YUQUAN. 1995. Thermal evolution of the Gangdese batholiths, southern Tibet: A history of episodic unroofing. Tectonics, 14, 223-236. CURRAY,J. R. 1989. The Sunda Arc: a model for oblique plate convergence. In: Proceedings of the Snellius H Symposium, Theme: Geology and Geophysics of the
150
G. PACKHAM
Banda Arc and Adjacent Areas, Part 1. Netherlands Journal of Sea Research, 24, 131-140. DALY, M. C., COOPER, M. A., WILSON, I., SMITH, D. G. & HOOPER, B. G. D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine and Petroleum Geology, 8, 2-21. DAVIDSON, J. W. 1991. The geology and prospectivity of Buton Island, Southwest Sulawesi, Indonesia. In: Indonesian Petroleum Association, Proceedings 20th Annual Convention. 209-234. DAVIES, I. C. 1990. Geology and exploration review of the Tomori PSC, Eastern Indonesia. In: Indonesian Petroleum Association, Proceedings 19th Annual Convention. 333-356. DE METS, C., GORDON, R. G, ARGUS,D. E & STEIN, S. 1990. Current plate motions. Geophysical Journal International, 101,425-478. DEWEY, J. F., CANDE, S., PrrMAN, W. C. III. 1989. Tectonic evolution of the India/Eurasia collision Zone. Eclogae Geologicae Helvetiae, 82, 717-734. , SHACKLETON,R. M., CHANGCHENGFA(~ SUN YIYIN. 1988. The tectonic evolution of the Tibetan Plateau. Philosophical Transactions of the Royal Society of London, Series A, 327, 379-413. ENGLAND, P. & HOUSEMAN, G. 1988. The mechanics of the Tibetan Plateau. Philosophical Transactions of the Royal Society of London, Series A, 326, 301-320. FULLER, M., HASTON, R., LIN, JIN-Lu, RICHTER, B., SCHMIDTr,E, E. & ALMASCO,J. 1991. Tertiary paleomagnetism around the S China Sea. Journal of Southeast Asian Earth Sciences, 6, 161-181. GEARV, E. E., HARRISON, M. T. & HIERTZLER, M. 1988. Diverse ages and origins of basement complexes, Luzon, Philippines. Geology, 16, 341-344. GIBLING, M. R., GRADSTEIN, E M., KRISTIANSEN, I. L., NAGY, J., SARTI, M. & WEIDMANN, J. 1994. Early Cretaceous strata of the Nepal Himalayas: conjugate margins and rift volcanism during Gondwanaland breakup. Journal of the Geological Society, London, 151, 269-290. GIDDINGS, J. W., SANTANA, W. & PIGRAM, C. 1993. Interpretation of palaeomagnetic results from the Bird's Head, Irian Jaya: new constraints on the drift history of the Kemum Terrane. Exploration Geophysics, 24, 283-290. GINGEr, D. C., ARDAKUSUMAH,W. O., HEDLEY, R. J. & Pothecary, J. 1993. Inversion history of the West Natuna Basin: examples from the Cumi-Cumi PSC. In: Indonesian Petroleum Association, Proceedings 22nd Annual Convention. 635-678. GORDON, R. G & JURDY, D. M. 1988. Cenozoic global plate motions. Journal of Geophysical Research, 91, 12 389-12 406. , HALL, R. 1996. Reconstructing Cenozoic SE Asia. This volume. --, ALl, J. R. & ANDERSON, C.D. 1995a. Cenozoic motion of the Philippine Sea plate: palaeomagnetic evidence from eastern Indonesia. Tectonics, 14, 1117-1132. --, FULLER,M., ALl, J. R. & ANDERSON,C. D. 1995b. The Philippine Sea Plate: Magnetism and Reconstructions. In: TAYLOR,B. & NATLAND,J. H. (eds) Active Margins and Marginal Basins: A
Synthesis of Western Pacific Drilling Results. American Geophysical Union Monograph, 88, 371-404. --, NICHOLS, G., BALLANTYNE, P., CHARLTON, T. & ALl, J. 1991. The character and significance of basement rocks of the southern Molucca Sea region. Journal of Southeast Asian Earth Sciences, 6, 249-258. HAMILTON, W. 1979. Tectonics of the Indonesian region. United States Geological Survey, Professional Paper, 1078. HARLAND, W. B, ARMSTRONG, R. L., Cox, A. V., CRAIG, L. E., SMITH, A. G. & SMITH, D. G. 1990. A geological time scale 1989. Cambridge University Press. HARRISON,T. M., COPELAND,P., KIDD, W. S. E & AN YIN. 1992. Raising Tibet. Science, 255, 1663-1670. HASAN, K. 1991. Upper Cretaceous flysch succession of the Balangbaru Formation, Southwest Sulawesi. In: Indonesian Petroleum Association, Proceedings 20th Annual Convention. 183-208, HASTON, R., FULLER, M. & SCHMIDTKE, E. 1988. Paleomagnetic results from Palau, West Caroline Islands: a constraint on Philippine Sea plate motion. Geology, 16, 654-657. HELMERS, H., HEBEDA, E. H. 8(: LUSTENHOUWER,W. J. 1990. The enigmatic relationship between subduction caused by metamorphism, ophiolitic emplacement and collision on eastern Sulawesi, Indonesia (abstract). In: Orogenesis in Action, Conference Abstracts. Geological Society, London, 30. HILL, K. C. 1991. Structure of the Papuan Fold Belt, Papua New Guinea. AAPG Bulletin, 75, 857-872. - - . , GREY, A., FOSTER, D. & BARRETT, R. 1993. An alternative model for the Oligo-Miocene evolution of northern PNG and the Sepik-Ramu Basin. In: CARMAN, G. J. • CARMAN, Z. (eds) Petroleum Exploration in Papua New Guinea. Proceedings of the Second PNG Petroleum Convention. 241-259. HOLLOWAY, N. n. 1982. North Palawan block - its relation to Asian mainland and role in evolution of South China Sea. AAPG Bulletin, 66, 1355-1383. HUANG, K. (~ OPDYKE,N. D. 1993. Paleomagnetic results from Cretaceous and Jurassic rocks of south and southwest Yunnan: evidence for large clockwise rotations in the Indochina and Shan-Thai-Malay terranes. Earth and Planetary Science Letters, 117, 507-524. HUCHON, P., LE PICHON,X. & RANGIN,C. 1994. Indochina Peninsula and the collision of India and Eurasia. Geology, 22, 27-30. JOLIVET, L., HUCHON, P. & RANGIN, C. 1989. Tectonic setting of Western Pacific marginal basins. Tectonophysics, 160, 23-47. KEMP, G & Mock, W. 1992. A re-appraisal of the geology, tectonics and prospectivity of Seram Island eastern Indonesia. In: Indonesian Petroleum Association, Proceedings 21st Annual Convention. 521-552. LACASSIN, R., LELOUP,P. & TAPPONNIER,P. 1993. Bounds on strain in large Tertiary shear zones of SE Asia from boudinage restoration. Journal of Structural Geology, 15, 677-692. LE PICHON, X., FOURNIER, M. & JOLIVET, L. 1992.
CENOZOIC SE ASIA RECONSTRUCTIONS Kinematics, topography, shortening and extrusion in the India-Eurasia collision. Tectonics, 11, 1085-1098. LELOUP, P. H., HARRISON, T. M., RYERSON, E J, CHEN WENJI, Li Qi, t~r AL. 1993. Structural, petrological and thermal evolution of a ductile shear zone, Diancang Shen Yunnan. Journal of Geophysical Research, 98, 6715-6744. LINDSAY, J. F., HOLLIDAY,D. W. & HULBERT,A. G. 1991. Sequence stratigraphy and the evolution of the Ganges-Bramaputra delta complex. AAPG Bulletin, 75, 1233-1254. MCCABE, R., HARDER, S., COLE, J. T. & LUMADYO, E. 1993. The use of paleomagnetic studies in understanding the complex Tertiary history of East and Southeast Asia. Journal of Southeast Asian Earth Sciences, 8, 257-269. MCCAFFREY, R. 1988. Active tectonics of the eastern Sunda and Banda arcs. Journal of Geophysical Research, 93, 15163-15182. 1989. Seismological constraints and speculations on Banda Arc tectonics. In: Proceedings of the Snellius H Symposium, Theme: Geology and Geophysics of the Banda Arc and Adjacent Areas, part 1. Netherlands Journal of Sea Research, 24, 141-152. MATTHEWS, S. J. & TODD, S. P. 1993. A tectonostratigraphic model for the southern Nam Con Son Basin, offshore Vietnam. Warta GeologL Newsletter of the Geological Society of Malaysia, 19, 276. MERCIER, J-L, ARMIJO, R., TAPPONNIER, P., CAREYGAILHARDIS, R. & HAN TONG LIN. 1987. Change from Late Tertiary compression to Quaternary extension in southern Tibet during the India-Asia collision. Tectonics, 6, 275-304. MrrCHELL, A. H. G. 1993. Cretaceous-Cenozoic tectonic events in the western Myanmar (Burma) - Assam region. Journal of the Geological Society, London, 150, 1089-1102. MOFFAT, D. T., HENAGE, L. E, BRASH, R. A., HARAHAP, B. H. & TAUER, R. W. 1991. Lengguru, Irian Jaya: prospect selection using field mapping, balanced cross-section and gravity modelling. In: Indonesian Petroleum Association, Proceedings 20th Annual Convention. 85-106. MOLNAR, P. & LYON-CAEN,H. 1989. Fault plane solutions of earthquakes and active tectonics of the Tibetan plateau and its margins. International Geophysical Journal, 99, 132-153. MOORE, G. E, KADARISMAN,D. & SUKAMTO, R. 1980. New data on the geology of the Talaud Islands, Molucca Sea. Bulletin of the Geological Research and Development Centre, Bandung, 3, 5-12. PACKHAM,G. H. 1973. A speculative Phanerozoic history of the south-west Pacific. In: COLEMAN, E J. (ed.) The Western Pacific: island arcs, marginal seas and geochemistry. Western Australian University Press, 369-388. 1990. plate motions and Southeast Asia: some consequences for basin development. Proceedings of the Southeast Asian Petroleum Association, IX, 55-68. PARKINSON, C. D. 1991. The petrology, structure and geologic history of the metamorphic rocks of central -
-
151
Sulawesi, Indonesia. PhD Thesis, University of London. PE~LER, G., DAS, S. & WOODnOUSE, J. H. 1995. A seismological study of the eastern New Guinea and the western Solomon Sea regions and its tectonic implications. International Geophysical Journal, 122, 961-981. PELTZER, G. & TAPPONNmR,P. 1988. Formation of strikeslip faults, rifts, and basins during the India-Asia collision: an experimental approach. Journal of Geophysical Research, 93, 15 085-15 117. PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and accretion history of New Guinea. BMR Journal of Australian Geology and Geophysics, 10, 193-212. & Symonds, E A. 1991. A review of the timing tectonic events in the New Guinea Orogen. Journal of Southeast Asian Earth Sciences, 6, 307-318. POWELL, C. McA., ROOTS, S. R. & VEEVERS,J. J. 1988. Pre-breakup continental extension in East Gondwanaland and the early opening of the eastern Indian Ocean. Tectonophysics, 155, 261-283. PUBELLIER, M., QUEBRAL,R., RANGIN, C., DEFFONTAINES, B., MULLER, C., er AL. 1991. The Mindanao Collision Zone: a soft collision event with a continuous Neogene strike-slip setting. Journal of Southeast Asian Earth Sciences, 6, 239-248. RAMANA, M. V., NAIR, R. R., SARMA, K. V. L. N. S., RAMPRASAD, Z., KRISHNA, K. S., ET AL. 1994. Mesozoic anomalies in the Bay of Bengal. Earth and Planetary Science Letters, 121,469-475. RANGrN, C. 1991. The Philippine Mobile Belt: a complex plate boundary. Journal of Southeast Asian Earth Sciences, 6, 209-220. ~, JOLIVET, L., PUBELLIER,M. 1990. A simple model for the tectonic evolution of southeast Asia and Indonesia region for the past 43 m.y. Bulletin de la Socidtd gdologique de France, 8 VI, 889-905. RATMAN, N. 1976. Geological map of the Tolitoli Quadrangle, North Sulawesi, 1:250,000. Geological Survey of Indonesia. RICHTER, B., FULLER, M., SCHMIDTKE,E., TIN MYINT, U., TIN NGEW, U., MYA WIN, U. & BUNOPAS, S. 1993. Paleomagnetic results from Thailand and Myanmar: implications for tectonic rotations in Southeast Asia. Journal of Southeast Asian Earth Sciences, 8, 247-256. RoY, T. K. 1986. Petroleum prospects of the frontal belt and subduction complex associated with the Indian plate boundary in the northeast. In: Sixth Offshore South East Asia Conference. 513-527. ROYER, J-Y. & SANDWELL, D. T. 1989. Evolution of the eastern Indian Ocean since the Late Cretaceous: constraints on Geosat altimetry. Journal of Geophysical Research, 94, 13 755-13 782. SALVIDAR-SALI,A., OSTERLE, H. G. & BROWNLEE, D. N. 1982. The geology of offshore northwest Palawan, Philippines. In: Proceedings of the 2nd Conference of the Asian Council on Petroleum (ASCOPE) Exhibition, Manila. 99-123. SAREWrrz, D. R. & KARIG, D. E. 1986. Process of allochthonous terrane evolution, Mindoro Island. Philippines. Tectonics, 5, 525-552. SEARLE,M. P., WINDLEY,B. E, COWARD,M. P., COOPER, D. J. W., REX, A. J., ET AL. 1987. The closing of -
-
152
G. PACKHAM
Tethys. Bulletin of the Geological Society of America, 98, 679-701. SENO, T., STEIN, S. A., GRIPP, A. E. & DEMETS, C. R. 1993. A model for the motion of the Philippine Sea plate consistent with NUVEL-1. Journal of Geophysical Research, 98, 17941-17948. SHIBUYA,H., HSU, V. & MERRILL, D. 1989. Paleomagnetic results of Leg 124: Celebes and Sulu Seas. Los, 70, 1365. SILVER, E. A. & RANGIN, C. 199t. Development of the Celebes Sea in the context of Western Pacific marginal basin history. In: SILVER, E. A., RANGIN, C., BREYMANN,M. YON (eds) Scientific Reports of the Ocean Drilling Program, College Station Texas. 124, 39-49. --, MCCAFr.~EY, R. & SMITH, R. B. 1983. Collision, rotation and the initiation of subduction in the evolution of Sulawesi, Indonesia. Journal of Geophysical Research, 83, 1681-1691. STRUCKMEYER,H. I. M., YUENG,M. & PIGRAM,C. J. 1993. Mesozoic to Cainozoic plate tectonic evolution and palaeogeography of the New Guinea Region. In" CARMAN, G. J. & CARMAN, Z. (eds)'Petroleum Exploration in Papua New Guinea. Proceedings of the Second PNG Petroleum Convention. 261-290. SUKAMTO, R. 1982. The geology of the Pangkajene and western part of Watampone, Sulawesi 1:250,000. Geological Research and Development Centre, Indonesia. TAPPONNIER, P., PELTZER,G. & ARMIJO,R. 1986. On the mechanics of the collision between Indian and Asia. In: COWARD, M. P. & Rms, A. C. (eds) Collision Tectonics. Geological Society, London, Special Publication, 19, 1154-157. --, LE DAIN, A. Y., ARMIJO, R. & COBBOLD, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611-616. TAYLOR, B. 1979. The Bismarck Sea: evolution of a backarc basin. Geology, 7, 171-174. & HAYES, D. E. 1980. Tectonic evolution of the South China Sea Basin. In: HAYES, D. E. (ed.) The Tectonic Evolution of Southeast Asian Seas and Islands, Part 2. Geophysical Monograph Series, American Geophysical Union, 27, 89-104. 9
-
-
TRELOAR, P. J. & COWARD, M. R 1991. Indian plate motion and shape: constraints on the geometry of the Himalayan orogen. Tectonophysics, 191, 198-198. VAN DER WERFF, W., KUSNIDA, D. PRASETYO, H. & VAN WEERING, T. C. E. 1994. Origin of the Sumba forearc basement. Marine and Petroleum Geology, 11, 363-374. VEDDER, J. G. 1986. Summary of the geology and offshore resources of the Solomon Islands. In: VEDDER, J. G., POUND, K. S. & BOUNDY, S. Q. (eds) Geology and offshore resources of Pacific island arcs central and western Solomon Islands. CircumPacific Council for Energy and Mineral Resources, Earth Science Series, 4, 295-306. WAJZER,M. R., BARBER,A. J., HIDAYAT,S. • SUHARSONO. 1991. Accretion, collision and strike-slip faulting: the Woyla Group as a key to the tectonic evolution of North Sumatra. Journal of Southeast Asian Earth Sciences, 6, 447-463. WEISSEL, J. K. 1980. Evidence for Eocene oceanic crust in the Celebes Basin.'In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. American Geophysical Union, Geophysical Monograph Series, 23, 37-48. & ANDERSON,R. N. 1978. Is there a Caroline plate? Earth and Planetary Science Letters, 41, 143-159. WENSINCK, H., HARTOSUKOHARDIO, S. & SURYANA, Y. 1989. Palaeomagnetism of Cretaceous sediments from Misool, northeast Indonesia. Proceedings of the Snellius H Symposium, Theme: Geology and Geophysics of the Banda Arc and Adjacent Areas, part 1. Netherlands Journal of Sea Research, 24, 287-301. YANG, ZENYU & BESSE, J. t993. Paleomagnetic study of Permian and Mesozoic sedimentary rocks from Northern Thailand supports the extrusion model for Indochina. Earth and Planetary Science Letters, 117, 525-552. Yu, Ho-SING 1990. The Pearl River Mouth Basin: a rift basin and its geodynamic relationship with the southeast Eurasian margin. Tectonophysics, 183, 177-186. ZHAO, W. & MORGAN, W. J. 1985. Uplift of the Tibetan Plateau. Tectonics, 4, 359-369.
Reconstructing Cenozoic SE Asia ROBERT
HALL
SE Asia Research Group, Department of Geology, Royal Holloway, University of London, Egham TW20 OEX, UK Abstract: Reconstructions of SE Asia at 5 Ma intervals for the past 50 Ma are presented. They are constrained by new data from the Philippine Sea plate, which forms the eastern boundary of the region, by recent interpretations of the South China Sea and Eurasian continental margin, forming the western boundary, and by the known motions of the Indian-Australian plate to the south. An attempt is made to satisfy geological and palaeomagnetic data from the region. The implications of these reconstructions for the Tertiary evolution of SE Asia are discussed in the light of other new data from the region. There are two regionally important periods of change during the past 50 Ma. Both appear to be the expression of arc-continent collision and resulted in major changes in the configuration of the region and in the character of plate boundaries. At c. 25 Ma the collision of the Australian continent with the Philippine Sea plate arc caused major effects which propagated westwards through the region. At c. 5 Ma collision of the Philippine arc and the Eurasian continental margin occurred in Taiwan. This appears to be a key to the recent tectonics of the region. Principal features of the model include the following interpretations. Middle Tertiary counter-clockwise rotation of Borneo closed a large proto-South China Sea and led to the development and destruction of marginal basins north of the Celebes Sea. The rotation implies that much of the north Borneo margin was not a subduction, but a strike-slip boundary for most of this period. It also suggests that the central West Philippine Sea, the Celebes Sea and the Makassar Strait formed part of a single marginal basin which opened between late Eocene and mid Oligocene, and narrowed westwards like the present South China Sea. Luzon is suggested to have formed in an arc on the north side of the Celebes Sea-West Philippine Basin, whereas most of the other Philippine islands probably formed part of an arc at the southern edge of the Philippine Sea Plate before the Early Miocene. Arc-continent collision in the early Miocene caused plate boundaries to change and initiated the clockwise rotation of the Philippine Sea plate. Since then the Philippine fragments have moved in a very narrow zone, mainly as part of the Philippine Sea plate, with significant strike-slip motion of fragments at the plate margin. Most subduction under the Philippines was oblique, mainly at the western edge, and north of Mindanao. The Molucca Sea was a very wide area which formed part of the Philippine Sea plate before c. 15 Ma and originated as trapped Indian ocean lithosphere. It has been eliminated by subduction on its east and west sides. The present-day double subduction system never extended north of the present Molucca Sea into the Philippines. The Sulawesi ophiolite has an Indian ocean origin and was emplaced on the west Sulawesi continental margin at the end of the Oligocene. The major change in plate boundaries at the beginning of the Miocene following arc-continent collision of the Australian margin with the Philippine Sea plate arc caused initiation of the Sorong Fault system and led to westward movement of continental fragments which were accreted to Sulawesi during the late Neogene. The Sula platform and Tukang Besi platform formed part of a single large microcontinent with the Bird's Head before c. 15 Ma. They moved to their present positions after slicing of fragments from this microcontinent at different times and each was attached to the Philippine Sea plate for a few million years before collision. Most of the Banda Sea is interpreted to have an extensional origin and to have opened during the late Neogene. The reconstructions imply that there has been little convergence at the north Australian margin in Irian Jaya since the early Miocene and most convergence has occurred during the last c. 5 Ma. Movement of Philippine Sea arc fragments within the northern New Guinea margin along strikeslip zones probably accounts for the terrane character of this orogenic belt.
Plate tectonic reconstructions o f SE Asia have some obvious practical value for the region in helping to understand the d e v e l o p m e n t o f sedim e n t a r y basins, the history o f volcanic arc activity, and similar processes w h i c h are linked through tectonics to the distribution o f natural resources. Reconstructions are also a necessary precursor to
From Hall, R. & Blundell, D.
understanding m o r e fundamental processes that have acted. W h a t are the important controls on the tectonic d e v e l o p m e n t o f the region (e.g. the role o f indentor tectonics), what are the critical events (e.g. different types o f collision event), and what is the nature o f deformation (e.g. rigid plate versus distributed d e f o r m a t i o n ) ? H o w far can plate
(eds), 1996, TectonicEvolutionof SoutheastAsia, Geological Society Special Publication No. 106, pp. 153-184.
153
154
R. HALL
tectonics go in describing the development of the region and the behaviour of crust and lithosphere (e.g. what is the role of strike-slip faulting versus contraction in the development of young orogenic belts)? Similar, more general ~uestions have relevance to collision processes and to the understanding of ancient orogenic belts for which parts of SE Asia have often been used as analogues. During the Cenozoic the region which now forms SE Asia was bounded to the north and west by a Eurasian plate, and to the south by the IndianAustralian plate. The motions of these plates are reasonably well known and their positions provide limits to the zone within which the SE Asian collage of microplates and sub-plate fragments can be moved when attempting plate reconstructions. The boundaries provided by these plate motions have been used as limits in previous reconstructions of Cenozoic SE Asia (e.g. Rangin et al. 1990; Daly et al. 1991; Lee & Lawver 1994). There have been considerable differences in dealing with the eastern boundary of the region. At present, the eastern edge of SE Asia is bounded by the Philippine Sea plate but the motion of this plate has been difficult to link to the global plate circuit because it is surrounded by subduction zones. Its Tertiary movement history has proved controversial, as illustrated by previous reconstructions, and there are major uncertainties in the position of the eastern edge of the region. Rangin et al. (1990) accepted evidence for clockwise rotation of the plate whereas Daly et al. (1991) and Lee & Lawver (1994) did not. Hall et al. (1995b) used palaeomagnetic data from east Indonesia to estimate Tertiary poles of rotation for the Philippine Sea plate and made a new reconstruction of this plate based on these poles, incorporating the effects of marginal basin opening within the plate. Discontinuous clockwise rotation for the entire plate during the last 50 Ma leads to palaeolatitude predictions which closely fit all palaeomagnetic data and also satisfies constraints on rotation inferred from magnetic anomaly skewness and seamount magnetization studies from the Philippine Sea plate. This model has been used to define the eastern margin of SE Asia as the basis for reconstructing the region using the ATLAS computer program (Cambridge Paleomap Services 1993) for the last 50 Ma. Figurel shows the present geography of the region and identifies the principal tectonic elements used in the reconstructions. Approximately sixty fragments (the number changes with age) have been used in reconstructing the region. Mercator projections showing reconstructions of the area bounded by latitudes 20°S and 30°N, and longitudes 90°E and 160°E, are presented for 5, 10, 15, 20, 25, 30, 35, 40, 45 and 50 Ma as Figs 2-11. Details of rotation poles, and fragments used, are
listed in Tables 1 and 2 which are available as Supplementary Publication No. SUP18101 (5 pp) from the Society Library or the British Library Document Supply Centre, Boston Spa, Wetherby, W Yorks LS23 7BQ, UK. The reconstruction is also available as an animation, on floppy discs, which can be run on either a Windows-based PC or a Mac with adequate hard disc space. Contact the author for details.
Methods ATLAS model
The ATLAS model uses a palaeomagnetic reference frame with Africa as the reference fragment with its movement defined relative to magnetic north. Movements of other major plates relative to Africa are based on Cande & Leslie (1986), Cochran (1981), Fisher & Sclater (1983), Klitgord & Schouten (1986), Le Pichon & Francheteau (1978), McKenzie et al. (1970), Royer & Sandwell (1989), Sclater et al. (1981), S~goufin & Patriat (1980) and Srivastava & Tapscott (1986). In these reconstructions South China is used as a reference and this is fixed to Eurasia for the period 50-0 Ma. There has been little Cenozoic motion of Eurasia whichever reference frame is used (e.g. Livermore et al. 1984; Gordon & Jurdy 1986; Besse & Courtillot 1991; Van der Voo 1993) and therefore it remains in a similar position to the present-day in all the reconstructions. In the ATLAS model there are small movements of Eurasia owing to the plate circuit used, particularly in the last 5 Ma, and therefore there are minor differences compared to reconstructions which use a fixed Eurasia (Lee &Lawver 1994; Rangin et al. 1990).
Palaeomagnetic
data
The model attempts to include the important constraints imposed by palaeomagnetism. Palaeomagnetic data can help to put some limits on interpretation of geological data since in principle they provide indications of palaeolatitudes and rotations. Interpretation is not always simple, and in SE Asia it is particularly difficult to reach unambiguous solutions. Van der Voo (1993) discusses in detail the use and problems of using palaeomagnetic results, and provides a particularly clear summary of problems in SE Asia. Besides the obvious drawbacks of collecting data in predominantly tropical, remote and often harsh terrain, there are additional problems such as error limits, remagnetization, and equatorial ambiguities. Not least amongst these are the difficulties of deciding
INDIAN
Malay Basins
::,:: '
PLATE
Gulf of Thailand
;
............
Java Sea
....
....
Palawan
Borneo
' • Bm.t~
SUNDA SHELF
Natuna Basins
.........
......
~r"cles~e~d
So~ China
EURASIA
~ "
•
,~'.'
.~B,,,~_.:._..-.c--/
~/w
"_-#//
Basin
SOUth
Sula
. . Besi Platform ,~-B~da,~
Sulawesi
Taiwan
f
I
'
1
/
mJ
CAROLINE PLATE _
PHILIPPINE SEA PLATE
PamceVela Basin
-
MADE
RECONSTRUCTIONS
margins during the Tertiary. Principal marine magnetic anomalies are shown schematically for the Indian ocean, Pacific ocean, South China Sea, Philippine Sea and Caroline Sea. Circled Z identifies Zamboanga peninsula of Mindanao. Double lines represent active spreading centres. Narrow lines represent principal bathymetric features of the Philippine Sea plate and the Caroline Ridge; the margins of the oceanic parts of the South China Sea, the Celebes Sea and the Andaman Sea; and deeper parts of the Sulu Sea and the Makassar Strait. Complexities in the Bismarck-Solomon Sea regions are not shown.
Fig. 1. Simplified present-day tectonic configuration of SE Asia. Shaded areas represent mainly ophiolitic, arc and other accreted material added to Eurasian and Australian
Indochtna
..... ~%1:
k/I
156
R. HALL
which interpretations of palaeomagnetic results should be accepted. One major problem with palaeomagnetic data, often not emphasized by the palaeomagnetists, is determining the motion history of a region. This is particularly important in areas where, for example, there are results from Mesozoic or older rocks, but few or no results from Tertiary rocks, such as the Malay peninsula and Thailand. Similarities in declinations of similar age rocks may indicate a common regional rotation history but in some cases similar declinations may be the results of different Tertiary motion histories for separate blocks (cf. Celebes Sea counter-clockwise rotation and age, v. Borneo counter-clockwise rotation and age, which are discussed below). Sea floor magnetic data from the South China Sea, for the Celebes Sea and the West Philippine Sea central basin also provide limits, although in all cases there are uncertainties in anomaly ages and correlation. The distinction between local and regional rotations is far from clear in SE Asia, particularly because good palaeomagnetic data are sparsely scattered in time and space, and modelling such as that attempted here shows clearly the need for long term, systematic and integrated palaeomagneticgeological studies of the region. It is possible to avoid many difficulties with palaeomagnetic data by explaining apparently anomalous or controversial evidence of movements as a consequence of local tectonics. This is often sensible, and Surmont et al. (1994) provide an excellent SE Asian example of how previously inferred large regional rotations are better interpreted as very local consequences of tectonics. However, at the same time they show that smaller systematic regional rotations can be recognized, and we must accept that much essential evidence for movements, for palaeogeographic reconstructions, and certainly for rotation about vertical axes, can only be acquired from palaeomagnetism and therefore to ignore all palaeomagnetic data is a position of despair. The present author has tried to distinguish regional scale movements from local movements but in several cases, principally Sundaland-Borneo and the Philippines, a choice has had to be made between different views on the basis of inadequate data. In these cases decisions were based on regional geological arguments, recognizing that other solutions are possible. But once a critical decision has been made, for example in this model that of accepting a large counter-clockwise rotation of Borneo, the reconstructions become both a development, and a partial test, of the decision; can reasonable reconstructions then be made that are consistent with the geological data set for the whole region? Thus, the reconstructions presented here
should be seen as a way of distinguishing local and regionally important datasets, identifying targets for future work, and a possible model of the region which provides a different, albeit sometimes controversial, interpretation of the development of SE Asia.
Principles and tests The fragments were left at current size in all reconstructions in order that they remain recognizable. This is broadly satisfactory for Neogene reconstructions, except for some areas of volcanic arcs. The present shape of fragments has been maintained although this may contribute to overlap of blocks if significant new crust has been created, which must be the case in the Sunda, Banda and the Philippine arcs. Before the Neogene many of the fragments may have had quite different sizes and shapes or simply may not have existed. This is true for many areas in which the extent or age of the basements is uncertain, for example, in parts of Philippines, Sangihe arc, Sulu arc, Java and Banda arc. In some cases, this has been incorporated by omitting the fragment before a certain time. For instance, most of the Bonin forearc did not exist at c. 50 Ma and probably grew during a period of rapid volcanism in the Eocene (Stem & Bloomer 1992; Taylor 1992). Thus, some fragments may not appear on all reconstructions between Figs 2 and 11. In moving fragments Occam's razor has been used in an attempt to find the simplest possible motion histories. Fragments were first attached, or partially coupled to, a major plate with known motion; most minor fragments can be linked in this way to a major plate (Australia, Eurasia, Philippine Sea). If more complex movements were required, experiments were made with transferring fragments from one major plate to another. During the reconstruction process possible solutions were tested by asking (1) are the palaeomagnetic data satisfied? (2) are the geological data satisfied? and (3) is there overlap of fragments during movements? As discussed below, both (1) and (2) require subjective interpretations and judgements; readers will inevitably have their own views of the value of these. The most important sources are cited and the reader may refer to regional compilations and reviews such as Hamilton (1979); Hutchison (1989) and Van der Voo (1993) for overviews of geology and palaeomagnetism. Below are summarized the main features of the data used and the reconstructions made for principal sub-areas within SE Asia. The implications of the judgements and interpretations made for the development of the region are then discussed.
RECONSTRUCTING CENOZOIC SE ASIA
Regions considered Indochina and South China Sea
In the reconstructions South China is fixed to stable Eurasia. The Indochina block south of the Red River fault and north of the Malay peninsula has been moved using the model of Briais et al. (1993). They suggest c. 550 km of movement on the Red River fault system with left-lateral motion between 32-15 Ma and some dextral movement since the late Miocene. Their average small circle pole (5.3°N, 66.3°E) results in significant overlap of Indochina and South China and this model therefore uses a small circle rotation pole (20.9°S, 61.5°E) further from the Red River fault, as Briais et al. (1993) suggest should be the case, which results in a smaller overlap but still provides the history of contraction and extension that they summarize. The movements on the fault estimated by them yield an approximately linear agedisplacement relationship and the model assumes regular displacement in the interval 15-32 Ma. The area between the Indochina coast and the present north Borneo coast is underlain by continental crust and the north Borneo margin has been fixed to Indochina to provide an indication of the northern edge of the proto-South China Sea created by removing the extrusion of the Indochina block. Taylor & Hayes (1980, 1983) identified ocean floor ~magnetic anomalies and used these to interpret the opening history of the South China Sea. This interpretation has been modified by Briais et al. (1993) and their model of opening, using their calculated poles of rotation, has been used in the reconstructions without change except for reassigning anomaly ages to the Harland et al. (1990) timescale. Palawan and Mindoro have been moved with Reed Bank in the reconstructions. The islands of Palawan and Mindoro are considered to include continental crust of south China origin and the bathymetric contour marking the north side of the Cagayan ridge is assumed to mark the southern limit of continental crust. The northwest part of the Sulu Sea is thus considered to be underlain by continental crust (Hinz et al. 1991).
Borneo
The basement of Borneo of western and interior Borneo consists of Palaeozoic and Mesozoic igneous, sedimentary and metamorphic rocks and this area behaved more or less as a craton during the middle and late Tertiary. To the north are younger additions to this continental core which have been interpreted as subduction accretionary complexes (Hamilton 1979) although this view is not universally accepted. In this work north and
157
south Borneo have been separated at the Lupar line, which is a zone of Mesozoic ophiolites and south Borneo has been treated as a single rigid fragment back to 50 Ma. West Sulawesi separated from east Borneo in the Tertiary, resulting in opening of the Makassar Strait and the development of large sedimentary basins in east Kalimantan. West Sulawesi has a long Tertiary history of igneous activity and the present eastern margin of the Sundaland block appears to have been an active margin for much of the Tertiary. There are two principal large-scale tectonic views of Borneo: one advocates a large counterclockwise rotation of the island, the second argues for no rotation of Borneo. Palaeomagnetic results are reported by Haile et al. (1977), Haile (1979), Schmidtke et al. (1990), Wahyono & Sunata (1987) and Lumadyo et al. (1993). These results are reviewed by Fuller et al. (1991) and Lee & Lawver (1994); the former favour a counter-clockwise rotation of the island and the latter favour no rotation. It is clear that the existing palaeomagnetic data are inadequate to reach a conclusion and those who reject the rotation of Borneo (Lumadyo et al. 1993; Lee & Lawver 1994; Rangin et al. 1990) emphasize the problems with the data. However, although not all the evidence points in the same direction there are also regional geological arguments that favour rotation and this model accepts a counter-clockwise rotation of southern Borneo. The major obstacle to incorporating the rotation in a regional tectonic model is determining the position of the rotation pole. The chosen pole is close to the northwest corner of Borneo (I°N, ll0°E). This allows Borneo to remain part of a Sunda block while permitting the rotational movement to be absorbed within the north Borneo accretionary complexes by closing a proto-South China Sea. It implies some extension between Borneo and the Malay peninsula and allows the southern boundary of Sundaland to rotate northwards. Because the pole is so close to the northwest corner of Borneo it requires no major deformation of the Sunda shelf to the northwest, although minor deformation would be expected. The earliest inversion event in the West Natuna Basin (Ginger et al. 1993) is Early Miocene, which is consistent with the timing chosen for the rotation (see below). The movement requires counter-clockwise motion of west Sulawesi with little latitude change, for which there is some evidence. It fails to account for the similarity of the Thailand and peninsula Malaysia counter-clockwise rotations reported by McElhinny et al. (1974) and Schmidtke et al. (1990). However, poles further from Borneo which could account for these data result in a much larger proto-South China Sea and also require large latitude changes which are not seen in the palaeomagnetic data.
!
\
e
~
~ '
Inner B*a'ndaarc
tw
"/
Sula platform
Okinawa trough
,pr°pagaresi
p.,,DIII~ ~dllb~l~4 j" Ambon
"
:: j}
trench
/
Convergence begins befween Caroline plate •~ andNew Guinea
PLATE ~\
CAROLINE .~
Clockwise rotation of Philippine Sea Plate
20°S
10°S
L~ O0
RECONSTRUCTING CENOZOIC SE ASIA Data from SW Sulawesi indicate that the SW arm was close to its present latitude in the late Jurassic (Haile 1978) and late Palaeogene (Sasajima et al. 1980) but has since rotated counterclockwise by about 45 ° . Directions recorded by late Miocene volcanic rocks of the SW arm are indistinguishable from those of the present field. The palaeomagnetic results from SW Sulawesi are very similar to those from Cretaceous rocks of the Malay Peninsula (McElhinny et al. 1974) and in Borneo (Fuller et al. 1991) there are similar counter-clockwise rotations since the late Mesozoic and before the late Miocene. The limited evidence from west Sulawesi and Borneo suggests that counter-clockwise rotation occurred before the late Miocene and sometime during the late Palaeogene to middle Miocene. This model therefore uses a rotation of 45 ° between 20 and 10 Ma. Thai-Malay
Peninsula
There are a number of faults at the southern boundary of Indochina on which there is likely to have been Tertiary movement although currently the movement histories of these faults are not well known. The Indochina block is separated from Thailand and peninsula Malaysia on a line representing the Three Pagodas and Wang Chao faults. This is undoubtedly an oversimplification and movement of the Malay blocks in the reconstructions cannot account fully for the formation of the sedimentary basins of the western part of the South China Sea. The movement history of the ThaiMalay region is difficult to determine. Offshore, in the western South China Sea, are several large sedimentary basins with complex trans-tensional histories (e.g. Ngah et al. 1996; Tjia & Liew 1996). The land area largely lacks Tertiary rocks and palaeomagnetic results do not provide a clear
159
picture. Post-Cretaceous clockwise rotations are recorded in Thailand and northern Malaysia (Schmidtke et al. 1990; Fuller et al. 1991) whereas counter-clockwise rotations (McElhinny et al. 1974; Halle et al. 1983; Schmidtke et al. 1990; Fuller et al. 1991) are reported from Tertiary and older rocks further south. Therefore a north Malaya block has been separated from a south Malaya block at the Khlong Marui fault. There is no evidence for a major suture separating the Malay peninsula from west Borneo and the counterclockwise rotations recorded are similar from both regions. However, moving both the Malaya blocks with south Borneo results in major overlap of the peninsula and Indochina. Hutchison (1989) observed, based on the compilation of Haile & Briden (1982), that although the declinations recorded in rocks from peninsula Malaysia and Borneo are similar, there are large differences in inclinations; these differences are also seen in more recent results (Fuller et al. 1991). Thus, it is possible that the counter-clockwise rotations of Borneo and Malaysia are of different ages. The model rotates the south Malaya block counterclockwise about the same pole as that used for south Borneo but reduces its counter-clockwise rotation to 15 ° . In contrast, clockwise rotations could be explained by extrusion of Indochina. The north Malaya block is rotated clockwise relative to south China using a pole close to the fragment. This keeps the Indochina and north and south Malaya blocks close together and implies late Tertiary trans-tensional faulting in the zone of the Three Pagodas fault where the Indochina and north Malaya fragments overlap before 20 Ma, Rotations are assumed to have occurred between 20 and 10 Ma in order to maintain the separation of the peninsula and south Borneo and to maintain the northern and southern parts of the peninsula as a
Fig. 2. 5 Ma reconstruction of SE Asia. At this stage the Philippine Sea plate was rotating clockwise about a pole close to its north edge (48°N, 157°E; outside the figure) and the Luzon arc collided with the Eurasian margin in Taiwan. On this and the following figures Eurasian blocks and blocks forming part of Sundaland, with areas accreted to its continental core before the early Tertiary, are shown in yellow. The Sunda Shelf and its extensions are shaded in pale yellow. The present areas of the Indian and Pacific plates are coloured blue. Blocks of Australian continentalorigin are shown in red. Areas shaded in pink are shallow and deep parts of the Australian continental margin. Submarine parts of Sula, Buton-Tukang Besi, and Bird's Head-related fragments are also shaded with pink. Areas shown in green are mainly volcanic arc, ophiolite and accreted material of the Ryukyu islands, the Philippines, north Moluccas, north Borneo, Sulawesi and northern New Guinea. The volcanic islands of the inner Banda arc are shown in orange. Areas within the Philippine Sea plate filled with magenta are remnant arcs. Thin black lines are used to show principal marine magnetic anomalies of the Indian ocean, Pacific ocean, South China Sea, Philippine Sea and Caroline Sea. Thin light blue lines represent marine bathymetry outlining at different stages the present limits of the Philippine Sea plate, Caroline Ridge, Caroline Sea, Sulu Sea, Andaman Sea, margins of the Makassar Strait, and the Java-Sunda, the Izu-Bonin-Mariana-Yap-Palau, Negros and Manila trenches. Complexities in the Bismarck-Solomon regions are not shown. Red lines with short paired arrows represent active spreading centres. Half arrows represent strike-slip motion. Thick black and red lines represent major faults or fragment sutures used in the reconstructions. Long arrows indicate motion directions of major plates. Circular arrows represent rotations.
160
R. HALL
broadly continuous fragment. Some extension is also predicted by the model from about 32 Ma as a result of Indochina extrusion. S u m a t r a a n d the A n d a m a n
Sea
North Sumatra is fixed to the south Malaya block for all the reconstructions. Because of the rotation of south Malaya discussed above, the Sumatran margin before 20 Ma would have been closer to N-S and sub-parallel to the motion vector for the Indian plate. In this configuration it is possible that the partitioning of convergence into an orthogonal subduction component and a parallel strike-slip component would not occur. South Sumatra is fixed to north Sumatra before 15 Ma. As the counterclockwise rotation of the Malay peninsula, Sumatra and Java proceeded, the angle between the Sumatran margin and the Indian plate motion vector would have become less oblique leading to formation of the dextral strike-slip system. Dextral motion along the Sumatran Fault zone is incorporated between 15 and 0 Ma. Rotating north Sumatra with the Malay peninsula can also account for extension in the Andaman Sea. The Andaman region has been included in the reconstructions but the model is probably over-simplified. The bathymetry of the Andaman Sea is complex (Curray et al. 1979) and very simplified bathymetric contours on the east and west sides of the sea are used as markers, fixed to north and south Sumatra respectively. This suggests that before c. 10 Ma there was a small amount of orthogonal extension. After c. 10 Ma extension was greater but highly oblique. This is broadly consistent with the age of the oldest oceanic crust in the Andaman Sea (c. 11 Ma) and the recent pattern of opening (Curray et al. 1979). Java Java is included largely for completion and has been rotated counter-clockwise by 30 ° between 20 and 10 Ma. This is a compromise between the rotations chosen for Sumatra and south Borneo. There is no evidence for great extension of the Java Sea during this period (e.g. Bishop 1980; Van der Weerd & Armin 1992), but if Java is rotated rigidly with south Borneo there is too much overlap of Java and Sumatra. The difference in the amounts of rotation for south Borneo and Java would permit some extension, and probable strikeslip faulting, of the east Java Sea between 2010 Ma. Because the pole of rotation is so close to the Borneo, Java and the north Sumatra blocks there is no significant change in their relative positions, although there is some overlap of Java and Sumatra. This could be accounted for by
assuming arc-parallel extension since both Java and Sumatra are likely to have been smaller than their present-day outlines, as indicated by the extensional histories of Neogene sedimentary basins of Sumatra, and the volume of Neogene arc volcanic rocks. The counter-clockwise rotation gives an assessment of the likely pre-middle Miocene orientation of the subduction zone south of Sumatra and Java which is closer to NW-SE than present-day. The rotation also has implications for the tectonic history of Java (e.g. oblique subduction could have caused strike-slip motion) and for the timing of volcanism. The model predicts important changes in volcanic and tectonic history beginning at c. 20 Ma for both Java and Sumatra. The subduction boundary south of Java must have changed eastwards to a more complex link into the Pacific. Sulawesi
Sulawesi consists of four principal tectonic belts: the west Sulawesi volcano-plutonic arc, the central Sulawesi metamorphic belt, the east Sulawesi ophiolite belt, and the continental fragments of Banggai-Sula, Tukang Besi and Buton. This configuration has been widely interpreted in terms of collision between the eastern micro-continental fragments and the western volcanic arc (e.g. Audley-Charles et al. 1972; Katili 1978; Hamilton 1979; Silver et al. 1983b) resulting in ophiolite emplacement and metamorphism. However, more recent work shows that this apparent simplicity is partly a reflection of incomplete knowledge of the region. There is evidence of several episodes of subduction beneath the west arm of Sulawesi since at least the late Cretaceous (Hamilton 1979) and this formed part of the east Sunda margin since the early Tertiary. Collision has played a significant part in the Tertiary development of the island but the micro-continental fragments arrived later than initially thought (e.g. Davies 1990; Parkinson 1991; Smith & Silver 1991). There has been little palaeomagnetic work on Sulawesi. The earliest results by Haile (1978) from the SW and the SE arms indicated that these arms originated in different regions during the late Jurassic-early Cretaceous (Audley-Charles et al. 1972; Katili 1978). Data from SW Sulawesi indicate that it was close to its present latitude in the late Jurassic (Haile 1978) and late Palaeogene (Sasajima et al. 1980) but rotated clockwise by c. 45 ° between the late Paleogene and late Miocene (Mubroto 1988). These results from SW Sulawesi are very similar to those from Cretaceous rocks of the Malay Peninsula (McElhinny et al. 1974) and Borneo (Fuller et al. 1991). Sasajima et al. (1980) reported Eocene-early Miocene clockwise rotation
1UO °1-
}
INDIAN PLATE
(3
Bomeo rotation complete
\
l l U~1-
,
~
~
f
OUlU
l ZU °1-
Seram I
..~ - •
l ;JU~1-
~4
~
140°~
f
150°E
New Guineastrike-slipsystem
~ f
CAROLINE
rotation ~ , of Philippine Sea Plate <
MoluccaSea doublesubduction f established ~ 1 "- ,e, Ayu trough .,~.~%,,...=._ spreading -_ N ~ t - . - -
,. Suluarc activity,,
• Sea,,
i
-~~~ Clodo~ise
20°S J
Fig. 3. Reconstruction of the region at l0 Ma. Between 5-25 Ma the Philippine Sea plate rotated about a pole at 15°N, I60°E. Counter-clockwise rotation of Borneo and related rotations of Sundaland were complete.
9U ~1-
iJ
\
rotationcomplete:
\
Subduction at Manila
PACIFIC PLATE
162
R. HALL
of the east part of the north arm whereas Otofuji et al. (1981) suggested no significant latitude change but clockwise rotation of more than 90 ° by the north arm between the Eocene-early Miocene. There are problems with interpretation of the Otofuji etal. (1981) data (J. C. Briden, pers. comm. 1994), and Surmont et al. (1994) show that 20-25 ° clockwise rotation of the whole north arm has occurred since the Miocene but that larger rotations are related to local shear zones. These rotations are consistent with the late Neogene tectonic history of Sulawesi proposed by Silver et al. (1983a) who reconstruct the island by removing the movement on the Palu, Matano, Tolo, Lawanopo and Kolondale faults. This paper follows their reconstructions and assumes, as they suggest, that this deformation occurred between 5 and 0 Ma. Palaeomagnetic work shows that lavas of the east Sulawesi ophiolite have a clear southern hemisphere origin (Mubroto et al. 1994) and formed at a latitude of 17 _+4°S. This is in marked contrast to similar Cretaceous and Tertiary rocks in the Halmahera islands where palaeomagnetic results from ophiolitic and associated rocks indicate sub-equatorial latitudes of formation (Hall et al. 1995a; Ali & Hall 1995). Palaeolatitudes of Sulawesi rocks are north of Cretaceous and the early Tertiary palaeolatitudes for the northern Australian margin but similar to Cretaceous palaeolatitudes for Sula (Ali & Hall 1995) and Misool (Wensinck et al. 1989). The age and origin of the east Sulawesi ophiolite is uncertain. Simandjuntak (1986, 1992) interprets the ophiolite as formed at an early Cretaceous spreading centre. K-Ar dates reported by Simandjuntak (1986) include 93--48 Ma gabbros and 54-38 Ma basalts. K-Ar ages on lavas by Mubroto et al. (1994) range from 79-16 Ma. Dating and geochemical studies by Girardeau et al. (1995) suggest a 44 Ma age and a backarc origin for part of the ophiolite suggesting that the ophiolite could be composite. Since the ophiolite and west arm were juxtaposed by the early Miocene (Parkinson 1991), the ophiolite is fixed to west Sulawesi from 25-0 Ma and before 25 Ma moved with the Indian plate. Makassar Strait
Geological similarities of east Borneo and west Sulawesi suggest that they have moved apart since the middle Palaeogene (Hamilton 1979) although the timing is not well constrained. The Makassar Strait is thought to be underlain by attenuated continental crust (Dtirbaum & Hinz 1982) and stretching occurred between early Palaeogene and Early Miocene (Situmorang 1982). The Makassar Strait was closed by fitting bathymetric contours on the west and east sides of the north and south
Makassar basins although this is only an approximation due to the thick Neogene sediments of the Mahakam delta, and young deformation in west Sulawesi (Bergman et al. 1996). The best fit is achieved using a pole NE of the north end of the strait at 6°N, 128°E requiring a rotation of 6 °. Because the rotation of Borneo and the Philippine Sea plate results in an alignment of the West Philippine central basin, the Celebes Sea and the Makassar Strait, the author suggests that these basins opened as part of a single basin, probably by spreading which propagated westwards from the West Philippine central basin, as discussed further below. The period of extension is assumed to be the same (44-34 Ma) which is consistent with the stratigraphic interpretation of Situmorang (1982, 1987).
Philippine Sea plate
The Philippine Sea plate provides an eastern boundary for the reconstructions. At present the plate is rotating clockwise about a pole near its northern edge (Seno et al. 1993). However, the Philippine Sea plate is now surrounded by subduction zones which separate it from the oceanic ridge system and consequently its earlier motion with respect to other major plates is difficult to determine. Subduction at the Philippine Trench is young (Cardwell et aI. 1980) and hence much of the Philippines must have been attached to the plate before the late Neogene. At the southern edge of the plate the Indonesian islands of the north Moluccas still form part of the plate. Reconstruction of the Philippine Sea plate and estimation of its past position are discussed by Hall et al. (1995b) with details of the data, the blocks, and the rotation poles used in reconstructing opening of the marginal basins. For the period between 5 and 0 Ma the present Eurasia-Philippine Sea plate pole (Seno et al. 1993) has been used. The rotation poles for the plate used for the period 50-5 Ma are based on palaeomagnetic data (Hall et al. 1995a) collected from the east Indonesian islands of the HalmaheraWaigeo region which contain a good Mesozoic and Tertiary stratigraphic record and indicate a long arc history. These palaeomagnetic data indicate that the Philippine Sea plate has rotated clockwise in a discontinuous manner since the early Tertiary with c. 35 ° clockwise rotation between 25 and 5 Ma, no rotation between 40 and 25 Ma and c. 50 ° clockwise rotation between 50 and 40Ma. Reconstructions based on these results, and including opening of marginal basins within the plate, show that other magnetic data from the plate are consistent with this rotation model (Hall et al. 1995b).
15Ma
\
rotation ~
!
i i
,~
Tukang
activel,
~ ~
k
<-4.
OCIr~a~ti~ise# /
©
PACIFIC PLATE
150°E
Terranesmovingin north
NewGuineastrike-slipsystem
MoluccaSea subductionbegins beneathHalmaheraL & ~ " •~ AyuTrough ~..~ ~ ~ - ~ ~ ~ ~ .~_~ %,. spreadingbegins --~
Bird'sHead "~" q microcontinent in 'baconslicer'. ~ . • ~
/ .~--x
Cagayanfidge collisionw#h north Sulu Palawan (
f
Spreadingends \ in Shikoku "~ basin
20°S
20°N
Fig. 4. Reconstruction of the region at 15 Ma. Rotation of Borneo and parts of Malaya, Sumatra and Java were underway. Strike-slip motion at the southern boundary of the Philippine Sea plate fragmented the Bird's Head microcontinent and moved blocks west in the plate boundary zone. Similar motions were occurring in the north Philippines. The backarc Sulu Sea began to close after collision of the Cagayan ridge with the Palawan margin,
INDIAN
Sumatran fault active
~, Malaya
Clockwise rotation of north
MiddleMiocene
7 EuRAsI,
o~
164
R. HALL
Celebes Sea
ODP Leg 124 results indicate that the Celebes Sea formed in the middle Eocene, probably not far from its present latitude (Silver & Rangin 1991). The palaeomagnetic inclinations indicate palaeolatitudes similar to the present-day latitude; no errors are quoted by Shibuya et al. (1991) but the data allow movement of up to 19 ° according to Silver & Rangin (1991) implying relatively large errors. The palaeomagnetic results of Shibuya et al. (1991) also indicate a counter-clockwise rotation of c. 60 ° between 42 and 20 Ma. The reconstructions in this paper imply a connection between the Celebes Sea and the West Philippine central basin. The suggestion that these two basins were linked was discussed by Silver & Rangin (1991) who argued against it without excluding it, based on apparent inconsistencies of spreading history, rates and palaeomagnetism between the two basins. In fact, there is no major difference between spreading rates estimated from spacing between anomalies 18 and 20 in the Celebes Sea (Weissel 1980) and the West Philippine central basin (Hilde & Lee 1984), and the small difference is consistent with a basin narrowing westwards. The stratigraphy and sedimentology of the two basins are similar (Nichols & Hall 1995) based on drilling by DSDP Legs 31, 59 and ODP Leg 124. The reconstructions also show that the apparently different rotation histories could lead to a sub-parallel alignment of their magnetic anomalies at the time of basin formation. In this model the Celebes Sea basin is suggested to have originally formed an extension of the West Philippine central basin which opened between 44 and 34 Ma, based on the ages of anomalies identified by Hilde & Lee (1984). The model therefore opens the basin and includes a 45 ° counterclockwise rotation between 44 and 34 Ma implying that the rotation occurred during opening. The magnetic anomalies identified by Weissel (1980) indicate the spreading centre was south of the present southern edge of the basin suggesting that part of the ocean has been subducted at the north Sulawesi trench since the late middle Miocene (Rangin & Silver 1991). The model assumes symmetrical spreading and eliminates the southern half of the ocean between 10 and 0 Ma at the north Sulawesi trench. The subduction is interpreted to have resulted largely from rotation of the north arm of Sulawesi (Surmont et al. 1994).
Sulu Sea-Cagayan
ridge
The Sulu Sea is a marginal basin thought to have opened as a backarc basin during the early Miocene (Holloway 1982; Hinz et al. 1991; Rangin & Silver 1991) south of the Cagayan ridge, although the
northern part of the Sulu Sea is underlain by continental crust as noted above. The Cagayan ridge is interpreted as a volcanic arc active for a short period in the early Miocene which collided with the south China margin at the end of the early Miocene (Rangin & Silver 1991). Part of this arc may also be present in Mindoro and Tablas (Marchadier & Rangin 1990). Rangin & Silver (1991) offer two scenarios for the history of this region. Their scenario A has been modelled by southward subduction of a proto-South China Sea beneath the Cagayan ridge between 20 and 15 Ma forming the Sulu Sea. Collision of the Cagayan ridge with Palawan at 15 Ma then resulted in development of a new subduction zone and southward subduction of part of the Sulu Sea beneath the Sulu arc between 15 and 10 Ma.
Philippines
With the exception of Palawan, Mindoro, Zamboanga and nearby parts of the west Philippines, the Philippine archipelago is composed of largely ophiolitic and arc rocks of Cretaceous and Tertiary age. The present Philippine fault is young, < 5 Ma, (Aurelio et al. 1991 ; Quebral et al. 1994) but there is evidence of older strike-slip faulting in the northern Philippines (e.g. Rutland 1968; Karig 1983; Karig et al. 1986; Stephan et al. 1986) possibly dating from the early Miocene. There is widespread evidence of volcanic activity throughout the Neogene implying subduction. West Mindanao east of Zamboanga is omitted before 5 Ma since almost all of this block consists of very young arc material, although it may include some basement of Eurasian continental affinities (Ranneft et al. 1960; Pubellier et al. 1991). It is likely that central Mindanao, although shown on most of the reconstructions, was also much smaller. The Philippines are widely considered to have formed part of an arc system at the edge of the Philippine Sea plate before the Pliocene (e.g. Rangin et al. 1985, 1991; Rangin 1991). South of Luzon palaeomagnetic results indicate clockwise rotations consistent with movement as part of the Philippine Sea plate. However, the Philippines are still palaeomagnetically insufficiently known to attempt a detailed plate tectonic model, even assuming this were possible, since so many fragments would be required. However, the general plate tectonic evolution is known (Rangin et al. 1990) and indicates that most of the Philippines have moved from the south and have collided with the Eurasian margin during the Neogene, although the relative importance of collision and strike-slip tectonics is uncertain. The tectonic evolution of this complex region has been simplified by localising
RECONSTRUCTING CENOZOIC SE ASIA all strike-slip movement on the present Philippine Fault and by moving the Philippines south of Luzon with the Philippine Sea plate. On the whole this avoids overlap of fragments, except between 105 Ma, and provides some limits on the large-scale tectonic setting of the region. The motions for the Philippines south of Luzon in the model are consistent with the palaeomagnetic data which predict clockwise rotations and northward movement (McCabe & Cole 1989; Fuller et al. 1991). The history of Luzon is more controversial. McCabe & Cole (1989), Fuller et al. (1991) and Van der Voo (1993) review the different interpretations of the palaeomagnetic data. This model accepts the Fuller et al. (1991) preferred interpretation of a largely counter-clockwise rotation history with a small latitudinal change for most of the Tertiary. They interpret relatively young clockwise rotations to indicate late Neogene movement with the Philippine Sea plate. Fuller et al. (1983) suggest that the palaeomagnetic results from Luzon indicate tectonic models should involve counter-clockwise rotation of Luzon since mid-Miocene, no important northward motion (_+ 500 km) since then, northward motion of the Zambales complex from equatorial latitudes with counter-clockwise motion since the Eocene, and no significant rotation in the Plio-Pleistocene. If Luzon is moved with the Philippine Sea plate and the rest of the Philippines for the whole of the period 50-0 Ma, most of these conditions are not met and there are major overlaps of Luzon, Sulawesi and the Celebes Sea. However, with the exception of the early counter-clockwise motion of the Zambales complex (which could be incorporated in a more detailed model) all of these constraints are satisfied if Luzon is positioned on the north side of the Celebes Sea before the Neogene. For the Neogene, Luzon was moved using the Philippine Sea plate rotation poles (Hall et al. 1995b) but at slightly lower rates, which implies a partial coupling to the plate, between 200 Ma, and assumed 40 ° counter-clockwise rotation between 25-20 Ma. The position of Luzon in the reconstructions is different from that normally assumed (e.g. Rangin et al. 1990) but can also satisfy many of the geological data, as explained below.
Halmahera
Reconstructions incorporating the rotation of the Philippine Sea plate and removing the effects of subduction at the Philippine Trench can be tested in part against present-day observations in the Molucca Sea region. The Philippine arcs terminate in the south in the Molucca Sea collision zone where the Halmahera and Sangihe arcs are actively
165
converging. West of Halmahera the Molucca Sea plate has an inverted U-shaped configuration and is dipping east under Halmahera and west under the Sangihe arc. Regional seismicity indicates that there are c. 200-300 km of subducted lithosphere beneath Halmahera. On the other side of the Molucca Sea, the Benioff zone associated with the west-dipping slab can be identified to a depth of c. 600 km beneath the Celebes Sea. In central Halmahera a fold-thrust belt forms the boundary between the ophiolitic eastern basement and arc volcanic western basement. Balanced crosssections indicate at least 60 km east-west shortening between east and west Halmahera in the fold-thrust belt and within the east arms (Hall & Nichols 1990). The age of thrusting is between 3 and 1 Ma. This movement is incorporated in the model and probably reflects intra-plate deformation associated with the change in motion of the plate. The western arms of Halmahera are covered by late Neogene to Recent volcanic .products related to eastward subduction of the Molucca Sea Plate. The volcanic arc was initiated at c. 12 Ma and subduction probably began at c. 15 Ma. An older phase of arc volcanic activity commenced in the late Eocene and terminated in the early Miocene. Stratigraphic similarities and the relative position of the Halmahera islands and the east Philippines suggest that they formed part of the same arc system before 5 Ma. The Halmahera islands were moved with the southern Philippine Sea plate in all the reconstructions.
Bird's Head
The Neogene movement of the Bird's Head has been estimated from constraints imposed by reconstruction of the Molucca Sea. Progressively restoring the oceanic crust subducted at the Sangihe and Halmahera Trenches requires a wide ocean and the Bird's Head needs to be moved further south than the northern Australian margin in order that the Bird's Head is south of this ocean. McCaffrey (1996) suggests the Bird's Head is currently moving south relative to Australia and this is incorporated by a small movement in the past 0.5 Ma. Before this time the movement of the Bird's Head has been modelled by assuming leftlateral strike-slip motion along a boundary parallel to the Aru Basin edge between 0.5 and 2 Ma. A small counter-clockwise rotation of the Bird's Head has been incorporated between 4 and 8 Ma based on palaeomagnetic results of Giddings et aI. (1993). A further small strike-slip motion is incorporated between 8 and 12 Ma, again constrained by reconstruction of the Molucca Sea. Before 12 Ma the Bird's Head is fixed to Australia.
20 Ma
"
basins
C"
crust thrust 'h Sulawesi
S~lu Sea opens
F~al spreading of South China Sea
Bird's Head microconfinent dismemberedby Sorong fault splays . . . .
.__~
Cagayan ridge separates f r o m /
ProtoSouth China Sea
/
/I
o
L~'---....~..
Molucca Sea forms part of PhilippineSea Plate
/.~~y/
.,
Spreadingin Shikoku \ basin
I
? ?
Sorong Fault systeminitiated
cA~o,,.~\
lO.
20°N__
20°S
10°S
-~\ 1
y, lSre''r'e
of Philippine Sea Plate
PACIFIC PLATE
I
Fig. 5. Reconstruction of the region at 20 Ma. Rotation of Borneo and parts of Malaya, Sumatra and Java began. Subduction of the Proto-South China Sea caused arc splitting in the Sulu arc and the separation of the active arc of the Cagayan ridge. South China Sea opening propagated SW into the Sunda shelf.
INDIAN PLATE
Early Miocene
E ~ PLATE
I
RECONSTRUCTING CENOZOIC SE ASIA
Sula The Sula platform is a fragment of continental crust widely considered to have been transported west by the Sorong Fault. Its origin is uncertain. Many authors consider it to be a piece of New Guinea that was detached from western Irian Jaya in late Cenozoic time (e.g. Visser & Hermes 1962; Audley-Charles et al. 1972; Hamilton 1979; Silver & Smith 1983). In contrast, Pigram et al. (1985) have proposed that it originated c. 1000 km further east, and was detached to form an independent micro-continent from central Papua New Guinea before the early Cretaceous. All these suggestions are based on stratigraphic features discussed by Pigram et al. (1985). The Sula platform is now attached to east Sulawesi and has a thrust contact with the east Sulawesi ophiolite. It was originally suggested that the collision of the west Sulawesi island arc and the Sula platform resulted in ophiolite emplacement in the east arm (Kiandig 1956; Hamilton 1979; Silver et al. 1983b). However, recent evidence indicates that this suggestion is incorrect. The ophiolite was obducted westward onto west Sulawesi at the end of the Oligocene (Parkinson 1991), whereas thrusting of the ophiolite onto the western edge of the Sula platform occurred in the latest Miocene (Davies 1990) indicating collision of the Sula platform with east Sulawesi must have occurred at c. 5 Ma. Sula is therefore moved with east Sulawesi from 0--5 Ma. Hamilton (1979) shows the Sula platform as a fragment moving along the north side of the Sorong Fault, implying it was attached to the Molucca Sea or Philippine Sea plates. By moving the Sula fragment with the Philippine Sea Plate before 5 Ma it fits closely to the Bird's Head at c. 1011 Ma. Charlton (1996) shows that the Tomori and Salawati basins, both of which are sharply truncated, would have formed the northern and southern parts of a single large basin if the Sula platform were moved back to this position. This need not preclude a Mesozoic separation of a Bird's Head microcontinent from further east on the north Australian margin.
B u t o n - T u k a n g Besi The Tukang Besi platform is a micro-continental fragment that collided with Sulawesi during the Miocene, although the timing of the collision is interpreted differently by different authors. Davidson (1991) suggested Buton to be a microcontinental fragment which collided in the early Miocene with SE Sulawesi, before the Tukang Besi platform. Here, both have been treated as parts of a single block, following Smith & Silver (1991), who consider Buton as part of the Tukang Besi platform.
167
This difference may be explicable if both formed part of an extended Birdfs Head microcontinent, as discussed later. Smith & Silver (1991) argue that collision of the Tukang Besi platform with Sulawesi was complete before the late Middle Miocene, since ophiolitic conglomerates and sandstones of this age (N13-N14) overlie deformed basement rocks. They interpret Lower Miocene conglomerates (N7-N9) that lack ophiolitic debris to indicate uplift associated with initial detachment of the platform from the northern New Guinea margin. If this interpretation is correct, initial uplift associated with detachment from New Guinea occurred between c. 15 and 17 Ma and collision of Tukang Besi with Sulawesi must have been complete by c. 11 Ma. To model this history Tukang Besi was fixed to west and central Sulawesi between 0 and 11 Ma. By attaching the platform to the southern edge of the Philippine Sea plate before that time Tukang Besi returns to a position, west of, and adjacent to the Sula fragment and the Bird's Head by 14 Ma. To obtain the best fit requires rotation of the Tukang Besi block, and this is incorporated by a clockwise rotation of 40 ° . The author therefore interprets the sequence of events to be: at c. 15 Ma a strand of Sorong fault propagates west, south of Tukang Besi; by 14 Ma Tukang Besi is fully attached to Philippine Sea plate; at 11 MaTukang Besi collides with Sulawesi, locking this strand of the Sorong fault and requiring a development of a new fault strand which caused the detachment of Sula. It is interesting to note that the development of Molucca Sea subduction beneath Halmahera begins at c. 13-15 Ma, indicated by K-Ar ages of volcanic rocks and reset ages (Baker & Malaihollo 1996), as well as biostratigraphic ages. Thus, a locking of subduction at the west side of the Molucca Sea, requiring initiation of a new subduction system on its east side, is temporally linked to development of the Sorong fault splay. Seram and Buru Hamilton (1979) suggested there is only shallow seismicity associated with the Seram Trough whereas Cardwell & Isacks (1978) argued that a deep slab was bent round the entire Banda arc. McCaffrey (1989) attempted to reconcile these differences on the basis of an increased number of better located seismic events, and concluded that there are two slabs, one subducted at the Timor trough, and a second subducted at the Seram trough. The Seram slab extends to no more than 300 km whereas the Indian ocean slab is continuous to over 600 km indicating that they record different histories of subduction. The bend in the Indian ocean slab implies that Seram has moved
168
R. HALL
eastwards while Australia has moved north. During the late Neogene the Banda arc has migrated east since the volcanoes become younger and the length of the subducted slab decreases eastwards. This configuration was modelled by moving Seram northwards to its present position relative to the Bird's Head between 4 and 0 Ma and moving Seram eastwards relative to the Bird's Head from 12-4 Ma. This is consistent with, although simplified from, the present tectonic configuration in the Banda Sea inferred by McCaffrey (1989). Buru is currently situated between strands of the Strong Fault system and Hamilton (1979) suggests it may have moved westward from New Guinea. The present author has allowed a small amount of westward movement between 4 and 0 Ma which fits Seram, Buru and the Bird's Head closer together but this is merely a guess.
Caroline plate The Ayu Trough opened during the middle and late Miocene (Weissel & Anderson 1978) and spreading may be continuing at the present day. There has been no subduction at the Caroline-Philippine Sea plate boundary and little convergence at the Caroline-Pacific boundary c. 25 Ma. The model uses the Caroline-Philippine Sea plate rotation pole and rate of Sent et al. (1993) for the period 5-0 Ma and then moves the Caroline plate with the east side of Ayu Trough before this. The Caroline plate is omitted from the reconstructions before 25 Ma.
Implications Timing of major changes The animated reconstructions show clearly that, during the interval 50-0 Ma, although there are important changes in movements and locally there are significant manifestations of these changes, there are two truly regionally important periods of change. Both of these appear to be the expression of arc-continent collision and resulted in major changes in the configuration of the region and in the character of plate boundaries. At c. 25 Ma the collision of the Australian continent with the Philippine Sea plate arc in New Guinea (Fig. 6) caused major effects which propagated westwards through the region. The Philippine Sea plate began to rotate clockwise requiring development of new subduction systems at its western edge. This led to the assembly of the Philippines archipelago, initiated new arc systems from north Sulawesi through to the Philippines, and led to the growth and partial destruction of marginal basins such as the Sulu Sea. The continued rotation of the Philippine Sea plate ultimately resulted in elimina-
tion of the Molucca Sea plate and accretion of fragments from the northern Australian margin into the SE Asian margin, notably in Sulawesi. At c. 5 Ma collision of the Philippine arc and the Eurasian continental margin occurred in Taiwan (Fig. 2). This appears the key to the recent tectonics of the region. Once again, the Philippine Sea plate motion changed, and new subduction systems were initiated, such as those currently active on the east and west sides of the Philippines. Most deformation now seems to be concentrated in the region between the Banda Sea and Taiwan. Summarized below are the major implications of the reconstructions for different parts of SE Asia.
Borneo Accepting rotation of Borneo has some important consequences for the reconstruction of the region, although the amount of rotation could be reduced to about 30 ° without seriously affecting the model. It means that there was initially a very wide protoSouth China Sea (Fig. 1 l) which began to close from c. 44 Ma (Fig. 10). Alternative reconstructions (e.g. Rangin et al. 1990; Lee & Lawyer 1994) which do not incorporate this counter-clockwise rotation, appear to lack the space required for the opening and closing of ocean basins such as that north of the Cagayan ridge and the Sulu Sea. Although the closure of the proto-South China Sea may have been partly driven by extrusion of the Indochina block the reconstructions suggest it was achieved by subduction either at the southeast side of the ocean or within the ocean (Figs 8 and 9). Once subduction was established slab-pull force would have caused extension of the IndochinaSouth China margin leading to formation of oceanic crust and extension of Eurasian continental crust (Fig. 7). Thus, extension in this region may not have been driven by strike-slip-related extrusion. In fact, the model suggests that subduction may have been underway before South China Sea opening began. Opening of the Celebes Sea required northward motion of Luzon because of its position north of the spreading centre, thus implying a subduction zone on the south side of the proto-South China Sea (Fig. 9). The Oligocene reconstructions are thus very similar to those of Taylor & Hayes (1983) except that south Borneo is rotated further. Many of the differences between this model and that of Rangin et al. (1990), which otherwise have many resemblances, result from the new constraints imposed by reconstruction of the Philippine Sea plate. If Borneo is not rotated there are problems in accounting for the evidence of continental crust in west Sulawesi in the early Miocene (see below; Bergman et al. 1996). The extended Bird's Head
110OE
Extension ~ dueto slab-puff
~
\
C ~
b.
f
"'"~' ....
Ophiolite collision
1120~E ~
Bird'sHead ~' microcontinent ~
1130°E~
Trapped Indian ocean
Openingof South kChinaSeasouth \ o f Macclesfield \ Bank
,b ~t
in.Sulawesi
Proto-
____1,
"l
1
1
4
0
°
E
withAustralia
~
r
? ?
Halmahera
/
d5OOE
CAROLINE~ C PLATE
Spreadingin PareceVela k basin
PACIFIC PLATE
20oS
10°S
10°N
20°N
Fig, 6. Reconstruction of the region at 25 Ma. Collision of the Australian continental margin in New Guinea, and the Bird's Head microcontinental block in Sulawesi, with the arc from north Sulawesi to Halmahera caused major reorganisation of plate boundaries. The active ridge jumped south in the South China Sea.
!
25 Ma L=eO,gocene
EURASIAN PLATE
I
/
~
1100°E
INDIAN PLATE
/
Dextral~\ motionot ree Pagodas system
extruded
Indochina
~
II10°E
Extensiondriven byslab-pufl ,andIndochina extrusion
J
EURASIAN PLATE
I
I1~
Bird's Head microcontinent
SulawesiwestArm
Subduction decreases westwards
ProtoSouth China Sea
~ /Pala N°wa~
Openingof outh ChinaSea northof Macclesfield Bank
.4
f
i
??
Openingof Parece Vela basinbegins
PACIFIC PLATE
[150°E
?
20°S
10°S_
0o
10°N_.
20°N_
Fig. 7. Reconstruction of the region at 30 Ma. South China Sea opening was driven by pull of the subducted Proto-South China Sea slab, and possibly by extrusion caused as India indented Eurasia. Arc splitting began in the eastern margin of the Philippine Sea plate with spreading in the Parece Vela basin.
/
|
movement on Red R/vet fau/t
Sin
Mid Oligocene
30 Ma'
I
¸--,,3 0
RECONSTRUCTING CENOZOIC SE ASIA
microcontinent would have been much further north than shown in the reconstructions and hence would have occupied the site of the Molucca Sea. For most of the Oligocene and early Miocene (Figs 6 and 7) the Sundaland margin in Borneo would have been orientated NW-SE and would have been a strike-slip dominated region and therefore the sedimentary basins of northwest Borneo should record an important component of this strike-slip history. This configuration also suggests that much of the sediment now found in north Borneo may have been fed southwards across the Sunda shelf from Indochina. This resolves a problem of the sediments now found in north Borneo which could not have a south Bomeo provenance because of their volume and because most of south Borneo was below sea-level during much of this period. The timing of subduction suggested by the model is similar to that deduced from the north Borneo stratigraphic record by oil company geologists (e.g. T a n & Lamy 1990) and the present author suggests that their 'Deep Regional Unconformity' is one manifestation of Australia-Philippine Sea plate collision and consequent reorganization of plate boundaries. Subduction became increasingly important in NE Borneo between c. 20 and 10 Ma (Figs 3-5) and the proto-South China Sea was entirely eliminated before 10 Ma. For Malay blocks west of Borneo, the reconstructions (Figs 4 and 5) can be regarded only as an approximation. The model suggests greater extension than that recorded in the Malay basins, and the timing of the rotations in the model needs to be improved. The similarities in rotations recorded through this part of Sundaland may be no more than coincidental, but if Bomeo and Java have rotated, there must have been some movement of Sumatra and the Malay peninsula, otherwise there is major overlap of fragments. The implications of rotations of blocks for the wider region are a problem which the palaeomagnetists often neglect, and therefore the model has some value, even if wrong, in attempting to face up to these implications, which in turn emphasize the need for more data to identify timing and to separate local and regional rotations. There can be no doubt that at present the palaeomagnetic data from Sundaland do not on their own convincingly demonstrate regional rotations. Origin o f the Celebes Sea
If palaeomagnetic evidence for the rotation of Borneo is accepted, there is a strong case to be made that the central West Philippine Sea, the Celebes Sea and the Makassar Strait formed part of a single marginal basin (Figs 8 and 9) which
171
opened between late Eocene and mid Oligocene, and widened eastwards like the present South China Sea. This interpretation has been incorporated in the model and the palaeomagnetic data from the Celebes Sea and Philippine Sea can i~e reconciled in such a model. The author follows Hamilton (1979) and many others in interpreting the Makassar Strait as an extended area of crust, probably without oceanic crust, although this need not preclude the Neogene convergent tectonics suggested by Bergman et al. (1996) for its west Sulawesi margin. The reconstructions are least convincing for the period between 44 and 40 Ma when the Philippine Sea plate was rotating rapidly clockwise and the Celebes Sea was opening with counter-clockwise rotation. However, this difficulty may have more to do with the inadequacy of the data on which the timings of rotations are based. For the Philippine Sea plate all we can say is that there was rapid rotation between 50 and 40 Ma (Hall et al. 1995b) and more data from Palaeogene rocks are needed to determine when and at what rate the rotation occurred. It may well have been fully complete by 45 Ma in which case there would be no difficulty with smooth opening of a basin narrowing west. For the Celebes Sea the position of the basin is consistent with a backarc setting related to northward subduction of Indian ocean lithosphere beneath west and north Sulawesi. Further east this setting appears less convincing because of the very large distance between the basin axis and the subduction zone (Figs 8 and 9). There were at least three major basins which opened in SE Asia with a similar east-widening geometry: a Mesozoic proto-South China Sea, the Eocene--Oligocene Philippine-Celebes basin and the Oligo-Miocene South China Sea. The South China Sea opening has been interpreted as linked to India indentation (Tapponnier et al. 1982) and as suggested here may be related to subduction, but this explanation seems unlikely for the PhilippineCelebes basin. Perhaps these three basins reflect some other lithospheric mechanism related to the long-term subduction of lithosphere east and south of SE Asia. Sulawesi collisions
The apparently simple tectonic configuration in Sulawesi of arc-ophiolite-continent is not the result of a single arc-continent collision (e.g. Silver et al. 1983b) but is a consequence of multiple collision events. The east Sulawesi ophiolite has an Indian ocean origin (Mubroto et al. 1994). Emplacement of the ophiolite on the west Sulawesi continental margin occurred at the end of the Oligocene (Parkinson 1991) and was followed by a change in plate boundaries at the beginning of the
i 100°E
~4
~, i
\ k
Sea
/
i
PACIFIC PLATE
I
l
140°E
Halmahera arc
~ f,,-
P
? ?
I150°E
~~l/estPhilippineSea andCelebesSea ____I openingalmost ~ complete , ~
EastPhilippinesarc
...d,
~
~
~
INDIAN-AUSTRALIANPLATE
Sulawesi arc
"/Z
Trenchadvancewith~ slabroll-backf . Subducti°n'~7~.Z~'(" ,:~/,~ ~ dies o~ tow~st.,"X "£
ProtoSouth China Sea
~,,1~
Extensiondriven bypullof subducted proto-SCSslab
• EURASIAN PLATE
i
4--
9
20°S
10°S
oo_
10 ° N
20°N_
Fig. 8. Reconstruction of the region at 35 Ma. Extension in the Sunda shelf and Eurasian continental margin was driven by pull of the subducted Proto-South China Sea slab. Active spreading of the Celebes Sea-West Philippine Sea basin ended at 34 Ma.
I 90°E ~
/
Early Oligocene
35 Ma
I
....j
/
1100°E
]110°E
~~,~~~/~'~,~1~
EURASIAN PLATE
120°EL ~1111~,.--,.,,.,~
1130°E
Arc activity at south edge of Philippine SeaPlate
1140°E
.p
10°N
20°N__
1150°E
20°S
10°S
NorotatiOnphilippineseaPlateOf
West Philippine Sea spreading extends to Celebes Sea
PACIFIC PLATE
~sea. ~'~~'le;es..~I'~~' ~lhA~ ~,~ '~--"'~[~~~~tSea " ~IL ~.a
Subduction of
ProtoSouth ChinaSea
Izu
~..,,.~.~....
Fig. 9. Reconstruction of the region at 40 Ma. Subduction of the Proto-South China Sea began as a trench became active north of Zamboanga-Luzon, caused by rapid opening of the Celebes Sea-West Philippine Sea. Between 40-50 Ma the Philippine Sea plate rotated clockwise about a pole at 10°N, 150°E.
90°E
-
Eocene
t.,o
174
R. HALL
Miocene (Fig. 6). There is isotopic evidence from geochemistry of igneous rocks for very old continental crust beneath west Sulawesi (Coffield et al. 1993; Priadi et al. 1993; Bergman et al. 1996), probably of Australian origin, because of the extreme isotopic compositions. The ages of the oldest igneous rocks reported (Bergman et al. 1996) imply early Miocene underthrusting of Australian lithosphere. Since the early Miocene there have been at least two further collisions, in SW and east Sulawesi, as fragments of continental crust have been sliced from the Bird's Head microcontinent and transported west for brief periods on the Philippine Sea plate or the Molucca Sea plate which was partially coupled to the Philippine Sea plate. Bird's Head microcontinent
The reconstructions suggest that the Sula platform and Tukang Besi platform formed part of a single large microcontinent with the Bird's Head at c. 15 Ma (Fig. 4). The model shows that movement of small continental fragments to their present positions can be explained easily if they were sliced from this microcontinent at different times and each moved with the Philippine Sea plate for a few million years before collision (Figs 2-5). Thus, the driving force for these motions was the Philippine Sea plate. Temporary locking of the strike-slip system at the southern edge of the Philippine Sea plate required development of new splays of the fault which resulted in the transfer of continental fragments to the Philippine Sea plate. As each fragment docked at the western end of the fault system in Sulawesi, a new splay developed and a new fragment began to move. The cartoons of Hamilton (1979) for the Neogene development of the region are remarkably similar to the predictions of the model although the timing is somewhat different, mainly because of new information from east Indonesia. This model has been described informally as a 'bacon slicer' and this does seem a very appropriate description of the mechanics of the process. The present author differs in one important respect from many interpretations of this region in suggesting that these fragments, although colliders, are not the major causes of contractional deformation associated with their docking. They are merely passive participants riding on the Philippine Sea plate whose major role is to lock a fault strand on their arrival at the Sulawesi margin. The reconstructions of Tukang Besi and Sula with the Bird's Head do not account for the evidence for Australian crust beneath west Sulawesi by the late Early Miocene (Bergman et al. 1996). The extent of this continental crust was estimated and used to outline an additional frag-
ment (Figs 5-9). If this is assumed to be beneath west Sulawesi by 17 Ma but is moved with the Philippine Sea plate from 21-17 Ma, it fits back to the western edge of the Bird's Head microcontinent (Fig. 6). It is suggested here that this fragment records the earliest effect of the 'bacon slicer' and was the first block detached from the Bird's Head microcontinent. It separated as a result of the development of the Sorong Fault system at the southem edge of the Philippine Sea plate and, like the other fragments, moved with the Philippine Sea plate for about 5 Ma. On the basis of the 25 Ma reconstruction, it can be speculated that the Bird's Head microcontinent was a single block at the end of the Oligocene formed by separation from Australia during the Mesozoic, to which had been added at least part of the east Sulawesi ophiolite. The reconstructions do not answer the question of the ultimate origin of the Bird's Head microcontinent, but they do require that the microcontinent has been moving with about the same motion as, and has remained in a similar position relative to, Australia for the past 25Ma. Palaeomagnetic results suggest that the microcontinent was at least 10°N of the north Australian margin in the late Cretaceous (Wensinck et al. 1989; Ali & Hall 1995) but its early Tertiary position is not well constrained in the model. The pre-Neogene geology of the region suggests there may have been several important events in the development of the Banda Sea region (e.g. ophiolite emplacement and metamorphism described by Sopaheluwakan (1990) from Timor), which could record relative movement between Australia and the microcontinent in passage to their 25 Ma positions. Molucca Sea
The model suggests that the Molucca Sea was a very wide area formed by trapping of Indian ocean lithosphere between the north Australian margin and the Philippine Sea plate arc when collision occurred at 25 Ma (Fig. 6). Thus, its age should be pre-Tertiary. It was eliminated by subduction on its east and west sides. Subduction probably began soon after the 25 Ma change in motion of the Philippine Sea plate on the west side of the Molucca Sea in the north Sulawesi-Sangihe arc, consistent with ages of arc volcanic rocks in north Sulawesi (Dow 1976; Effendi 1976; Apandi 1977). The Molucca Sea formed part of the Philippine Sea plate up to c. 15 Ma (Figs 4 and 5) when eastdipping subduction began beneath Halmahera (Baker & Malaihollo 1996). Most of the subduction occurred on the west side so the Molucca Sea has remained partly coupled to the Philippine Sea plate. The double subduction system of the Molucca Sea
RECONSTRUCTING CENOZOIC SE ASIA probably never extended north of its present northern edge into the Philippines. However, there was a oceanic area which was effectively continuous with, and north of, the Molucca Sea which was formerly the central part of the Celebes SeaWest Philippine central basin. In the reconstructions this northern part of the ocean is eliminated between 25 and 12 Ma and the Panay sector of the Philippine arc arrives at the western subduction margin at about 12 Ma. Slightly earlier (15 Ma) the Cagayan arc and Luzon arrive at the Palawan margin (Fig. 4) resulting in collision in Mindoro (Rangin et al. 1985). It may be one or both of these events in the Philippines which caused initiation of the east-dipping subduction system beneath Halmahera and the concomitant development of splays of the Sorong Fault system at the southern edge of the Molucca Sea. The reconstructions of the Philippines are too imperfect to be confident of the relationships between cause and effect but the timing of collision and age of the ophiolites in Mindoro (Rangin et al. 1985) and events in Panay (Rangin et al. 1991) are consistent with the model. Philippines
The Philippines are difficult to reconstruct for a number of reasons. Our understanding of Philippines geology and evolution is still insufficient, although considerable advances have been made in recent years as a result of investigations in the region (see for example, Sarewitz & Karig 1986; Stephan et al. 1986; Rangin 1991; and references therein). Reconstructions using the outlines of present-day fragments can be only approximate since much new crust has been added by arc processes. In order to describe the Philippines more precisely many more fragments need to be used and this is currently beyond the capacity of the ATLAS program. However, it seems that rigid plate tectonics may be an inadequate tool to describe the evolution of the area. At present this is illustrated most clearly at the south end of the archipelago where plate boundaries are ill-defined probably because there is a significant amount of within-plate deformation (Rangin et al. 1996). Karig et al. (1986) have drawn attention to the way in which at present only a small part of the Philippines is completely coupled to the Philippine Sea plate, and how in the past it is likely that coupling was never complete across the entire Philippine system because of strike-slip faulting. Because of the situation of all the Philippine fragments at the edge of plates it is probably unwise to rely too heavily on the evidence of arc volcanism to infer arc continuity. At present it is clear that some volcanism is related to subduction on the west side of the Philippines, some may be related to sub-
175
duction at the Philippine Trench, and some may not be directly related to subduction. Switching of subduction from one side of the Philippines to the other has almost certainly occurred in the past (e.g. Schweller etal. 1983; Karig e t a l . 1986). Some arcs may also be eliminated completely during collision as one arc overrides another, a process that is underway in the Molucca Sea. However, the reconstructions do provide some useful limits on what is feasible. Most of the Philippine islands probably formed part of an arc at the southern edge of the Philippine Sea Plate before the early Miocene '(Figs 7-10) as shown earlier in the reconstructions of Rangin et al. (1990). The choice here for the position of Luzon (Figs 6-10) shows that the geological evidence for subduction can be satisfied in ways other than by including all the Philippines in this arc. Since the early Miocene (Figs 2-6) the Philippine fragments have moved in a very narrow zone, mainly as part of the Philippine Sea plate, within which there appears to have been a substantial component of strike-slip motion (Sarewitz & Karig 1986). Most subduction under the Philippines was oblique, mainly at the western edge, and north of Mindanao. The Philippines are an ephemeral feature. They will probably end up as a composite arc terrane smeared onto the Eurasian continental margin which will be impossible to unravel. The great depths of young sedimentary basins in close juxtaposition to areas of high emergent topography suggest that lithospheric processes operating in this region are not well described by current plate tectonic concepts, and in some ways the Philippines are reminiscent of Pre-Cambrian greenstone belts. There seems no evidence that the Philippines are the result of a collision of two opposed arcs, progressively zipping up southwards towards the present-day Molucca Sea as shown in some reconstructions (e.g. Lewis et al. 1982; Rammlmair 1993). In particular, there is no evidence that the west-facing Halmahera arc extended north of the present north edge of the Molucca Sea into Mindanao (Quebral et al. 1995). Banda Sea
The age and origin of the Banda Sea have long been the subject of dispute. Several workers have suggested it is relatively old, possibly Mesozoic, with the Banda Sea representing trapped oceanic crust (Katili 1975; Bowin et al. 1980; Lapouille et al. 1986; Lee & McCabe 1986) whereas others have preferred a much younger, late Tertiary, age (Carter et al. 1976; Hamilton 1979; R6hault et al. 1995). Silver et al. (1985) suggested that the North Banda Sea may include crust of Pacific origin and identified the Banda ridges as continental
I 100°E
I110°E
North Sulawesi
-, ~ ~
Zambales~ ophioliteformation
ProtoSouth China Sea
1120°E
INDIAN-AUSTRALIAN PLATE
EURASIAN PLATE
?
.=
i130°E
©
PLATE
PHILIPPINE SEA .
Lu~oo
1140°E
Subductionof NNG-Pacific
? ?
IO°N
20°N
I150°E
20°S
~,~1~ 4~mahera
Boninitemagmatism ,,~ in Izu-Boninand •~ Mariana • \ forearcs
PACIFIC PLATE
\
Fig. 10. Reconstruction of the region at 45 Ma. The North New Guinea-Pacific spreading centre was subducted causing forearc magmatism and massive extension of the NE margin of the Philippine Sea plate.
90°E
A
• Middle Eocene
45 Ma
k
o~
RECONSTRUCTING CENOZOIC SE ASIA fragments based on dredging, a conclusion supported by more recent dredging (Villeneuve et al. 1994). Since some of the magnetic lineations of supposed Mesozoic age cross the Banda ridges the inferences based on magnetic anomaly ages may be wrong. Many of the lineations are parallel to the structural fabric of these continental areas (Silver et al. 1985; Rthault et aI. 1991). Despite this, there still appears to be a widespread acceptance that the Banda Sea is floored by old oceanic crust (e.g. Hartono 1990). This is not a conclusion supported by the reconstructions in this paper. In these the Banda volcanic arc is fixed to west Sulawesi during the late Neogene and the arc-trench gap is assumed to have maintained its present width. Timor and the islands of the outer Banda arc are fixed to Australia for the same period. The Birds' Head microcontinent, including the island of Seram, has been moved, as described above, to satisfy the constraints imposed by reconstruction of the Molucca Sea and the distribution of the deep subducted slabs inferred from seismicity. The reconstructions that result show that before c. 10 Ma the gap between Seram and Timor (Fig. 4) was filled by oceanic crust, now subducted beneath the Banda arc, which could indeed have been of Indian ocean origin and Mesozoic age. Timor arrived at the trench at c. 4-3 Ma consistent with geological data from Timor (e.g. Carter et al. 1976; Audley-Charles 1986; Harris 1991). However, before c. 10 Ma the relative distance between Timor and Seram is maintained (Figs 4-6) suggesting there was no subduction in the eastern Banda Sea area. The reconstructions suggest subduction began at c. 10 Ma (Fig. 3), resulting in the eastward propagation of the volcanic arc consistent with the very young age of the volcanoes in the inner Banda arc (Abbot & Chamalaun 1981; McCaffrey 1989). The north Banda basin extended as Seram moved east and the arc propagated east. The eastward movement of Seram is consistent with tectonic inferences from present seismicity (McCaffrey 1989) and geological observations on Seram (Linthout et al. 1991). The arc propagated east to the longitude of Seram at c. 5 Ma (Fig. 2) which is close to the age of the well known ambonites of this area. The reconstructions also show the south Banda basin extending rapidly between 5 and 0 Ma. Therefore, both the north and south Banda Sea can be interpreted as having an extensional origin and to have opened during the late Neogene. These interpretations are consistent with the age of young volcanics dredged in the Banda Sea (R6hault et al. 1995). The great depth of the south Banda Sea is still a problem, although it is known that Parsons & Sclater's (1977) age-depth relationships often do not hold in small ocean basins. Van Gool et al.
177
(1987) recorded high heat flows in several NW Banda Sea basins and suggested that they are not isostatically and thermally compensated. Caroline plate and New Guinea
The consequences of fixing the Caroline plate to the east side of Ayu Trough for Caroline-Pacific boundary motion are that there has been predominantly left-lateral motion since 25 Ma, minor oblique convergence between 15 and 5 Ma and minor oblique extension between 5 and 0 Ma (Figs 2-6). The timing of convergence and extension are partly dependent on the model since the period of 15-5 Ma was chosen as the interval of the main opening of the Ayu Trough, based on Weissel & Anderson (1978). However, these conclusions are consistent with the evidence of young oblique extension in the Sorol Trough (Weissel & Anderson 1978). The northern New Guinea arc terranes have been omitted from the reconstructions for simplicity and because there are insufficient data to reconstruct them adequately. However, the reconstructions do predict that there would have been relatively little convergence between the northern Australian margin and the Caroline plate during the Neogene. Most of the convergence that is required should have occurred in the past 5 Ma (Fig. 2) and would have been oblique. At the present day it appears that much of the convergence is distributed (McCaffrey & Abers 1991; Puntodewo et al. 1994) implying that only a proportion need be absorbed by subduction. This is consistent with the shallow seismicity associated with the New Guinea trench and lack of active volcanoes (Hamilton 1979; Cooper & Taylor 1987), and the absence of Neogene volcanicity in Irian Jaya and the Bird's Head (Pieters et al. 1983; Dow & Sukamto 1984). The reconstructions require young (5-0 Ma) underthrusting on the sector of the New Guinea trench east of the Bird's Head as interpreted by Hamilton (1979) and Cooper & Taylor (1987) but not north of the Bird's Head, consistent with observations of Milsom et al. (1992). The apparent complexity of the northern New Guinea margin (cf. Pigram & Davies 1987) appears to be due less to multiple collisions than to strike-slip faulting and fragmentation of the arc which collided with the Australian margin in the early Miocene. No reconstruction has been attempted here of the SW Pacific north and east of New Guinea, and the reader is referred to Yan & Kroenke (1993).
Final comments The reconstructions presented here differ significantly from earlier attempts to describe the
~
'\
1100°E
Indochina
~
II10°E
~
EURASIAN PLATE
(,,/
Zamboanga
~ 1120°E
Southand ast Sulawesi
INDIAN-AUSTRALIAN PLATE
• West Sulawesi
ProtoSouth China Sea
North ..~ Palawan/
Mindom
Taiwan
1130°E
Fig. 11. Reconstruction of the region at 50 Ma. The early rotation of the Philippine Sea plate began.
90°E
'~/
End Early Eocene
50 Ma
SouthChina
I
1140°E
NORTH NEW GUINEA PLATE
t
1150°E
20°S
o
...j oo
RECONSTRUCTING CENOZOIC SE ASIA development of SE Asia. Daly et "al. (1991) and Lee & L a w v e r (1994) neglect the rotation of the Philippine Sea plate and this accounts for major differences in our interpretations of the eastern parts of the region. Rangin et al. (1990) modelled clockwise rotation of the Philippine Sea plate and their reconstructions are quite similar to those here for the Philippines region but the models differ on the position of the Philippine Sea plate, the rotation of Borneo and the position of Luzon. As noted earlier, these differences are partly a consequence of different interpretations of inadequate data. Thus, even if these new reconstructions are rejected they do at least serve to draw attention to the need for many more good quality palaeomagnetic data from the region, in addition to g e o l o g i c a l data, particularly on the timing, chemistry and character of volcanic activity. Biogeographical data could also be especially useful in testing the different interpretations of the region and there is a need for a joint approach from botanists, zoologists and geologists. If the large frame is used it also becomes clear how much one is limited by the area, by the extent of possible oceanic crust, etc.; these limitations are much less obvious in local reconstructions where it is easier to move problems outside the area of immediate interest. The reconstructions suggest that the indentation of Eurasia by India has played a much less important role in the development of SE Asia than often assumed. They suggest that collision of the Australian continent has driven major rotations, and that the movement of smaller plates, such as the Philippine Sea plate and the Borneo microplate, is the result of the northward movement of Australia. Events which are thus 'caused' by movements of such plates may therefore be the consequence of more fundamental regional movements. Thus, caution is necessary in correlating events of similar age and interpreting their causes. The difficulties of reconstructing the Philippines are partly due to our still inadequate knowledge of this region but may
179
also reflect our limited understanding of arc processes, particularly at the lithospheric scale. The intra-oceanic history of the region leaves a poor record; arcs are ephemeral features on the geological timescale and may disappear completely, a process currently underway in the Molucca Sea. Probably the most difficult feature to incorporate is the role of strike-slip faulting. Like Karig et al. (1986) and Rangin et al. (1990) the present author considers that strike-slip faulting has played a major part in the development of the Philippines and the reconstructions confirm that there is limited space for the convergent motions often inferred from the arc volcanic record. These reconstructions also point to the importance of strike-slip faulting in other parts of the region such as Borneo and New Guinea, and suggest a need for re-examination of data and interpretations based on purely convergent collision models. I am particularly grateful to Simon Baker who has provided me with invaluable support, not only with the work involved in digitizing the fragment outlines used in the reconstructions, but with discussion and ideas at many stages. I also thank the numerous people with whom I have discussed the geology of SE Asia over many years who have contributed criticism, ideas and comments, including D. A. Agustiyanto, J. R. Ali, C. D. Anderson, M. G. Audley-Charles, P. D. Ballantyne, S. J. Baker, A. J. Barber, J. C. Briden, T. R. Charlton, E. Forde, R. Garrard, N. Haile, A. S. Hakim, S. Hidayat, T. W. C. Hilde, M. Fuller, Kusnama, J. A. E Malaihollo, R. McCabe, J. S. Milsom, A. H. G. Mitchell, S. J. Moss, R. Murphy, G. J. Nichols, C. D. Parkinson, M. Pubellier, R. Quebral, C. Rangin, H. Rickard, S. J. Roberts, N. Sikumbang, T. O. Simandjuntak, M. J. de Smet, B. Taylor and S. L. Tobing. Financial support has been provided at various stages by NERC award GR3/7149, grants from the Royal Society, the University of London SE Asia Research Group, and the University of London Central Research Fund. Work in Indonesia was facilitated by GRDC, Bandung and Directors including H. M. S. Hartono, M. Untung, R. Sukamto and I. Bahar. I thank R. J. Bailey, Gordon Packham and Claude Rangin for reviews and many helpful comments on the manuscript.
References ABBOTT,M. & CHAMALAUN,F. H. 1981. Geochronology of some Banda Arc volcanics. In: BARBER,A. J. & WmYosuJoNO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Indonesia, Special Publication, 2, 253-268. ALI, J. R. & HALL, R. 1995. Evolution of the boundary between the Philippine Sea Plate and Australia: palaeomagnetic evidence from eastern Indonesia. Tectonophysics, 251, 251-275. APANDI, T. 1977. Geologic map of the Kotamobagu quadrangle, North Sulawesi. Geological Research and Development Centre, Bandung, Indonesia.
AUDLEY-CHARLES, M. G. 1986. Rates of Neogene and Quaternary tectonic movements in the southern Banda arc based on micropalaeontology. Journal of the Geological Society, London, 143, 161-175. - - - , CARTER, D. J. & MILSOM, J. S. 1972. Tectonic development of Eastern Indonesia in relation to Gondwanaland dispersal. Nature Physical Science, 239, 35-39. AURELIO,M. A., BARRIER,E., RANGIN,C. & MULLER,C. 199 I. The Philippine Fault in the late Cenozoic tectonic evolution of the Bondoc-Masbate-N. Leyte area, central Philippines. Journal of SE Asian Earth Sciences, 6, 221-238.
180
R. HALL
BAKER, S. & MALAIHOLLO, J. A. E 1996. Dating of Neogene igneous rocks in the Halmahera region: arc initiation and development. This volume. BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. & GARRARD, R. J. 1995. Tertiary tectonic and magnetic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for a Miocene continent-continent collision. This volume. BESSE, J. & COURTrLLOT,V. 1991. Revised and synthetic apparent polar wander paths of the African, Eurasian, North American and Indian plates, and true polar wander since 200Ma. Journal of Geophysical Research, 96, 4029-4050. BISHOP, W. E 1980. Structure, stratigraphy, and hydrocarbons offshore southern Kalimantan, Indonesia. AAPG Bulletin, 64, 37-58. BowrN, C. O., PURDY,G. M., JOHNSTON,C., SHOR, G. G., LAWVER, L., er AL 1980. Arc-continent collision in Banda Sea region. AAPG Bulletin, 64, 868-915. BRIAIS, A., PATRIAT, P. 8z TAPPONNIER,P. 1993. Updated interpretation of magnetic anomalies and seafloor spreading stages in the South China Sea: implications for the Tertiary tectonics of Southeast Asia. Journal of Geophysical Research, 98, 6299-6328. CAMBRIDGE PALEOMAP SERVICES. 1993. ATLAS version 3.3. Cambridge Paleomap Services, CO. Box 246, Cambridge, UK. CANDE, S. C. 8z LESLIE, R. B. 1986. Late Cenozoic tectonics of the southern Chile trench. Journal of Geophysical Research, 91, 471-496. CARDWELL,R. K. 8£ ISACKS, B. L. 1978. Geometry of the subducted lithosphere beneath the Banda Sea in Eastern Indonesia from seismicity and fault plane solutions. Journal of Geophysical Research, 83, 2825-2838. , - & KARIG, D. E. 1980. The spatial distribution of earthquakes, focal mechanism solutions and subducted lithosphere in the Philippine and northeastern Indonesian islands. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 1-35. CARTER, D . J . , AUDLEY-CHARLES,M. G. 8z BARBER,A. J. 1976. Stratigraphical analysis of island arccontinental margin collision in Eastern Indonesia. Journal of the Geological Society, London, 132, 179-198. CHARLTON, Z. R. 1996. Correlation of Salawati and Tomori basins, eastern Indonesia: a constraint on left-lateral displacements of the Sorong Fault Zone. This volume. COCHRAN, J. R. 1981. The Gulf of Aden: structure and evolution of a young ocean basin and continental margin. Journal of Geophysical Research, 86, 263-287. COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A., GURITNO, N., ROBINSON,N. M. & TALBOT, J. 1993. Tectonic and stratigraphic evolution of the Kalosi PSC area and associated development of a Tertiary petroleum system, south Sulawesi, Indonesia. In: Proceedings Indonesian Petroleum Association 22nd Annual Convention 1993. 678-706. COOPER, P. & TAYLOR, B. 1987. Seismotectonics of New
Guinea: a model for arc reversal following arccontinent collision. Tectonics, 6, 53-67. CURRAY, J. R., MOORE, D. G., LAWVER, L. A., EMMEL, F. J., RArrT, R. W., e~r AL. 1979. Tectonics of the Andaman Sea and Burma. In: WATKINS, J., MONTADERT, L. & DICKENSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG, Memoir, 29, 189-198. DALy, M. C., COOPER,M. A., WILSON, I., SMITH, D. G. & HOOPLa, B. G. D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine and Petroleum Geology, 8, 2-21. DAVIDSON, J. W. 1991. The geology and prospectivity of Buton island, S.C. Sulawesi, Indonesia. In: Proceedings Indonesian Petroleum Association 20th Annual Convention 1991, 209-233. DAVIES, I. C. 1990. Geological and exploration review of the Tomori PSC, eastern Indonesia. In: Proceedings Indonesian Petroleum Association 19th Annual Convention 1990, 41-67. Dow, D. B. & SUKAMTO, R. 1984. Western Irian Jaya: the end-product of oblique plate convergence in the late Tertiary. Tectonophysics, 106, 109-139. Dow, J. 1976. Porphyry copper exploration in Block II, Sulawesi. Geological Research and Development Centre, Bandung Indonesia, Newsletter, 8 (33), 6. Dt)RBAUM, H.-J. & HINZ, K. 1982. SEATAR-related geophysical studies by BGR in the southwest Pacific. In: Transactions of the Third CircumPacific Energy and Mineral Resources Conference. 129-133.
EFFENDI, A. C. 1976. Geologic map of the Manado quadrangle, North Sulawesi. Geological Research and Development Centre, Bandung, Indonesia. FISHER, R. L. & SCLATER,J. G. 1983. Tectonic evolution of the southwest Indian ocean since the midCretaceous: plate motions and stability of the pole of Antarctica/Africa for at least 80 Myr. Geophysical Journal of the Royal Astronomical Society, 73, 553-576. FULLER, M., HASTON, R., LIN, J-L., RICHTEr, B., SCHMIDTKE, E. & ALMASCO, J. 1991. Tertiary paleomagnetism of regions around the South China Sea. Journal of SE Asian Earth Sciences, 6, 161-184. --, MCCABE, R. WILLIAMS,I. S., ALMASCO,J., ENCINA, R. Y., ZANORIA, A. S. & WOLFE, J. A. 1983. Paleomagnetism of Luzon. In: HAVES, D. E. (ed.) The Tectonic and Geologic Evolution of South-East Asian Seas and Islands, Part 2. American Geophysical Union Monograph, 27, 79-94.
GIDDINGS, J. W., SUNATA, W. & PIGRAM, C. 1993. Reinterpretation of palaeomagnetic results from the Bird's Head, Irian Jaya: new constraints on the drift history of the Kemum Terrane. Exploration Geophysics, 24, 283-290. GINGER, D. C., ARDJAKUSUMAH,W. O., HEDLEY,R. J. & POTHECARY, J. 1993. Inversion history of the West Natuna basin: examples from the Cumi-Cumi PSC. In: Proceedings of the Indonesian Petroleum Association 22nd Annual Convention. 635-658.
GIRARDEAU,J., MONNIER,CH., MAURY, R., VILLENEUVE, M., SOETISMA,D. • SAMODRA,H. 1995. Origin of the east Sulawesi ophiolite. In: Abstracts, Eighth
RECONSTRUCTING CENOZOIC SE ASIA
Regional Conference on Geology, Minerals and Energy Resources of SE Asia (GEOSEA 95). 51-52. GORDON, R. G. & JURDY, D. M. 1986. Cenozoic plate motions. Journal of Geophysical Research, 91, 12389-12406. Hamz, N. S. 1978. Reconnaissance palaeomagnetic results from Sulawesi, Indonesia, and their bearing on palaeogeographic reconstruction. Tectonophysics, 46, 77-85. --. 1979. Rotation of Borneo microplate complete by Miocene: palaeomagnetic evidence. Warta Geologi,
Geological Society of Malaysia Newsletter, 5 19-22. • & BRIDEN, J. C. 1982. Past and future palaeomagnetic research and the tectonic history of East and Southeast Asia. In: Palaeomagnetic research in Southeast and East Asia. UN/ESCAP, CCOP Technical Publication Bangkok, 13, 25-46. ., MCELHINNY, M. W. & McDOUGALL, I. 1977. Palaeomagnetic data and radiometric ages from the Cretaceous of west Kalimantan (Borneo) and their significance in interpreting regional structure. Journal of the Geological Society of London, 133, 133-144. , BECKINSALE,R. D., CHAKRABORTY,K. R., HUSSEIN, A. H. & HARDJONO, T. 1983. Palaeomagnetism, geochronology and petrology of the dolerite dykes and basaltic flows from Kuantan, west Malaysia• Bulletin of the Geological Society of Malaysia, 16, 71-85. HALL, R. & NICHOLS, G. J. 1990. Terrane amalgamation at the boundary of the Philippine Sea Plate. Tectonophysics, 181, 207-222• , ALI, J. R. ANDERSON,C. D. & BAKER, S. J. 1995a. Origin and motion history of the Philippine Sea Plate~ Tectonophysics, 251, 229-250. , FULLER, M., ALI, J. R. & ANDERSON,C. D. 1995b. The Philippine Sea Plate: Magnetism and Reconstructions. In: TAYLOR, B. & NATLAND, J. H. (eds) Active Margins and Marginal Basins: A
Synthesis of Western Pacific Drilling Results. American Geophysical Union Monograph, 88, 371-404. HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078. HARLAND, W. B., ARMSTRONG,R. L., Cox, A. V., CRAIG, L. E., Smith, A. G. & Smith, D. G. 1990. A geologic time scale 1989. Cambridge University Press. HARRIS, R. A. 1991. Temporal distribution of strain in the active Banda orogen: a reconciliation of rival hypotheses. Journal of SE Asian Earth Sciences, 6, 373-386. HARTONO, H. M. S. 1990. Late Cenozoic tectonic development of the Southeast Asian continental margin in the Banda Sea area. Tectonophysics, 181, 267-276. HILDE, T. W. C. & LEE, C-S. 1984. Origin and evolution of the West Philippine Basin: a new interpretation. Tectonophysics, 102, 85-104. HINZ, K., BLOCK,M., KUDRASS,H. R. 8z MEYER,H. 1991. Structural elements of the Sulu Sea. Geologische Jahrbuch, A127, 883-506. HOLLOWAY, N. n. 1982. The stratigraphic and tectonic
181
evolution of Reed Bank, North Palawan and Mindoro to the Asian mainland and its significance in the evolution of the South China Sea. AAPG Bulletin, 66, 1357-1383. HUTCHtSON, C. S. 1989. Geological Evolution of SouthEast Asia. Oxford University Press. Oxford Monographs on Geology and Geophysics, 13. KARIG, D. E. 1983. Accreted terranes in the northern part of the Philippine archipelago. Tectonics, 2, 211-236. --, SAREwrrz, D. R. & H~a~CK, G. D. 1986. Role of strike-slip faulting in the evolution of allochthonous terranes in the Philippines. Geology, 14, 852-855. KaTILZ, J. A. 1975. Volcanism and plate tectonics in the Indonesian island arcs. Tectonophysics, 26, 165-188. 1978. Past and present geotectonic position of Sulawesi, Indonesia. Tectonophysics, 45, 289-322• KLITGORD, K. D. & SCHOUTEN,H. 1986. Plate kinematics of the central Atlantic. In: VOGT,P. R. • TUCHOLKE, B. E. (eds) The Geology of North America. Volume M. Geological Society of America, 351-378. KT3NDIG, E. 1956. Geology and ophiolite problems of east Celebes. Verhandelingen van het Konin-
klijk Nedertandsch Geologisch-Mijnbouwkundig Genootschap, Geologische Serie, 16, 210-235 LAPOUILLE, A., HARTONO, H. M. S., LARUE, M., PRAMUNIJOYO, S. & LARDY, M. 1986. Age and origin of the seafloor of the Banda Sea (Eastern Indonesia). Oceanologica Acta, 81, 379-389. LE PICHON, X. & FRANCHETEAU,J. 1978. A plate tectonic analysis of the Red Sea-Gulf of Aden area. Tectonophysics, 46, 369-406. LEE, C. S. & MCC~E, [t. 1986. The Banda-Celebes-Sulu Basin: a trapped piece of Cretaceous-Eocene oceanic crust? Nature, 322, 51-54. LEE, T-Y. & LAWVER,L. A. 1994. Cenozoic plate tectonic reconstruction of the South China Sea region. Tectonophysics, 235, 149-180. LEWIS, S. D., HAYES, D. E. & MROZOWSKI, C. L. 1982. The origin of the West Philippine Basin by inter-arc spreading. In: BaLCE, G. R. & ZANORIA, A. S. (eds)
Geology and Tectonics of the Luzon-Marianas Region. Philippines SEATAR Committee, Special Publication, 1, 31-51. LrNTHOUT, K., HELMERS, H. & ANDRIESSEN, P. A. M. 1991. Dextral strike-slip in central Seram and 34.5 Ma Rb/Sr ages in pre-Triassic metamorphics related to early Pliocene counter-clockwise rotation of the Buru-Seram microplate (E. Indonesia). Journal of SE Asian Earth Sciences, 6, 335-342. LIVERMORE, R. A., VINE, E J. & SMITH,A. G. 1984. Plate motions and the geomagnetic field, II, Jurassic to Tertiary. Geophysical Journal of the Royal Astronomical Society, 79, 939-961. LUMADYO, E., MCCABE, R., HARDER, S. & LEE, T. 1993. Borneo: a stable portion of the Eurasian margin since the Eocene. Journal of SE Asian Earth Sciences, 8, 225-231. McC~a3E, R. & COLE, J. 1989. Speculations on the late Mesozoic and Cenozoic evolution of the Southeast Asian margin. In: BEN-AVRAHAM, Z. (ed.) The Evolution of the Pacific Ocean Margins. Oxford University Press.
182
R. HALL
MCCAFFREY, R. 1989. Seismological constraints and speculations on Banda arc tectonics. Netherlands Journal of Sea Research, 24, 141-152. 1996. Slip partitioning at convergent plate boundaries of SE Asia. This volume. & ABERS, G. A. 1991. Orogeny in arc-continent collision: the Banda arc and western New Guinea. Geology, 19, 563-566. MCELHINNY, M. W., HALLE, N. S. & CRAWFORD, A. S. 1974. Palaeomagnetic evidence shows Malay Peninsula was not part of Gondwana. Nature, 252, 641-645. MCKENZrE, D. P., DAVIES, D. & MOLNAR, P. 1970. Plate tectonics of the Red Sea and east Africa. Nature, 226, 243-248. MARCHADIER,Y. & RANGIN, C. 1990. Polyphase tectonics at the southern tip of the Manila trench, MindoroTablas islands, Philippines. Tectonophysics, 183, 273-287. MILSOM, J., MASSON, D., NICHOLS, G., SIKUMBANG, N., DWIYANTO, B., ET AL. 1992. The Manokwari trough and the western end of the New Guinea trench. Tectonics, 11, 145-153. MUBROTO, B. 1988. A palaeomagnetic study of the East and Southwest Arms of Sulawesi, Indonesia. DPhil. Thesis, University of Oxford. , BRIDEN, J. C., MCCLELLAND, E. & HALL, R. 1994. Palaeomagnetism of the Balantak ophiolite, Sulawesi. Earth and Planetary Science Letters, 125, 193-209. NGAH, K, MAZLAN, M. & TJIA, H. D. 1996. Role of preTertiary fractures in formation and development of the Malay and Penyu basins. This volume. NICHOLS, G. J. & HALL, R. 1995. Stratigraphic and sedimentological constraints on the origin and tectonic history of the Celebes Sea basin. In: Tectonics of SE Asia Conference abstracts. Geological Society, London 1994, 40. OTOFUJI, Y., SASAJIMA,S., NISHIMURA,S., DHARMA, A. & HEHUWAT, F. 1981. Palaeomagnetic evidence for clockwise rotation of the northern arm of Sulawesi, Indonesia. Earth and Planetary Science Letters, 54, 272-280. PARKINSON, C. D. 1991. The petrology, structure and geologic history of the metamorphic rocks of central Sulawesi, Indonesia. PhD Thesis, University of London. PARSONS, B. & SCLATER, J. G. 1977. An analysis of the variation of ocean floor bathymetry and heat flow with age. Journal of Geophysical Research, 82, 803-827. PIETERS, P. E., PIGRAM, C. J., TRAIL, D. S., Dow, D. B., RATMAN, N. & SUKAMTO,R. 1983. The stratigraphy of western Irian Jaya. Geological Research and Development Centre Bulletin, 8, 14--48. PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and the accretion history of the New Guinea orogen. BMR Journal of Australian Geology and Geophysics, 10 193-211. --, SURONO & SUPANDJONO,J. B. 1985. Origin of the Sula Platform, eastern Indonesia. Geology, 13, 246-248. PRIADI, B., POLVI~, M., MAURY, R. M., BELLON, H., SOERIA-ATMADJA, R., E T AL. 1993. Tertiary and -
-
Quaternary magmatism in central Sulawesi: chronological and petrological constraints. Journal of SE Asian Earth Sciences, 9, 81-93. PUBELLIER, M., QUEBRAL, R. D., RANGIN, C., DEFFONTAINES, B., MULLER, C., E T AL. 1991. The Mindanao collision zone: a soft collision event within a continuous Neogene strike-slip setting. Journal of SE Asian Earth Sciences, 6, 239-248. PUNTODEWO, S. S. O., MCCAFFREY, R., CALAIS, E., BOCK, Y., RAIS, J., ET AL. 1994. GPS measurements of crustal deformation within the Pacific-Australia plate boundary zone in Irian Jaya, Indonesia. Tectonophysics, 237, 141-153. QUEBRAL, R. D., PUBELLIER,M. & RANGIN, C. 1995. The onset of movement on the Philippine Fault in eastern Mindanao: a transition from collision to strike-slip environment. Tectonics, in press. RAMMLMAIR, D. 1993. The evolution of the Philippine archipelago in time and space: a plate tectonic model. Geologisches Jahrbuch, B81, 3-48. RANGIN, C. 1991. The Philippine Mobile Belt: a complex plate boundary. Journal of SE Asian Earth Sciences, 6, 307-318. & SILVER, E. A. 1991. Neogene tectonic evolution of the Celebes-Sulu basins: new insights from Leg 124 drilling. In: SILVER, E. A., RANGIN, C., VON BREYMANN,M. T., er AL. Proceedings of the Ocean Drilling Program, Scientific Results, 124, 51-63. ., JOLIVET,L. & PUBELLIER,M. 1990. A simple model for the tectonic evolution of southeast Asia and Indonesia region for the past 43 m.y. Bulletin de la Socidtd gdologique de France, 8 VI, 889-905. --., STEPHAN, J. F. & MI3LLER, C. 1985. Middle Oligocene oceanic crust of the South China S e a jammed in the Mindoro collision zone (Philippines). Geology, 13, 425-428. --, BUTTERLIN, J., BELLON, H., MI3LLER, C., CHOROW1CZ, J. & BALADAD, D. 1991. Collision n6og~ne d'arcs volcaniques dans le centre des Philippines: stratigraphie et structure de la chMne d'Antique 01e de Panay). Bulletin de la Socidtd g~ologique de France, 162, 465-477. ., DAHRIN, D., QUEBRAL, R. & THE MODEC SCIENTIFIC PARTY. 1996. Collision and strike slip faulting in the northern Molucca Sea, (Philippines and Indonesia): preliminary results of a morphotectonic study. This volume. RANNEFT, T. S. M., HOPKINS, R. M., FROELICH, A. J. & GWlNN, J. W. 1960. Reconnaissance geology and oil possibilities of Mindanao. AAPG BuUetin, 44, 529-568. RI~HAULT, J-P., MALOD, J. A., LARUE, M. BURHANUDDIN, S. & SARMILI, t . 1991. A new sketch of the central north Banda Sea, eastern Indonesia. Journal of SE Asian Earth Sciences, 6, 329-334. --, VILLENEUVE, M., HONTHAAS, C., BELLON, H., MALOD, J. A., er AL 1995. New geological samplings along a transect across the southeast Banda Sea basin (Indonesia). In: Tectonics of SE Asia Conference abstracts. Geological Society, London 1994, 49. ROYER, J-Y. & SANDWELL,D. T. 1989. Evolution of the eastern Indian Ocean since the Late Cretaceous: -
-
RECONSTRUCTING CENOZOIC SE ASIA constraints from Geosat altimetry. Journal of Geophysical Research, 94, 13,755-13,782. RUTLAND, R. W. R. 1968. A tectonic study of part of the Philippine Fault Zone. Quarterly Journal of the Geological Society of London, 123, 293-325. SAREWrrz, D. R. & KARIC, D. E. 1986. Geologic evolution of western Mindoro island and the Mindoro suture zone, Philippines. Journal of SE Asian Earth Sciences, 1, 117-141. SASAJ1MA, S., NISHIMURA,S., HIROOKA, K., OTOFUJI, Y., VAN LEEUVEN, T. & HEHUWAT, E 1980. Palaeomagnetic studies combined with fission track datings on the western arc of Sulawesi, east Indonesia. Tectonophysics, 64, 163-172. SCHMIDTKE, E. A., FULLER, M. D. & HASTON, R. 1990. Paleomagnetic data from Sarawak, Malaysian Borneo, and the late Mesozoic and Cenozoic tectonics of Sundaland. Tectonics, 9, 123-140. SCHWELLER,W. J., KARIG, D. E. & BACHMAN,S. B. 1983. Original setting and emplacement history of the Zambales ophiolite, Luzon, Philippines, from stratigraphic evidence. In: HAVES, D. E. (ed.) The
Tectonic and Geologic Evolution of South-East Asian Seas and Islands, Part 2. American Geophysical Union Monograph, 27, 124-138. SCLATER,J. G., FISHER,R. L., PATRIAT,P., TAPSCOTT,C. & PARSONS, B. 1981. Eocene to recent development of the southwest Indian ridge, a consequence of the evolution of the Indian ocean triple junction. Geophysical Journal of the Royal Astronomical Society, 64, 87-604. S~GOUFIN, J. & PATRIAT,P. 1980. Existence d'anomalies mrsozoiques dans le bassin de Somali. Implications pour les relations Afrique-AntarctiqueMadagascar. Comptes Rendus de l'Acaddmie Sciences, France, 291B, 85-88. SENO, T., STEIN, S. A. & GR1PP, A. E. 1993. A model for the motion of the Philippine Sea plate consistent with NUVEL-1 and geological data. Journal of Geophysical Research, 98, 17,941-17,948. SHIBUYA,H., D. L./VIERRIL, V. HSU & LEG 124 SHIPBOARD SCIENTIFIC PARTY. 1991. Paleogene counterclockwise rotation of the Celebes Sea - orientation of ODP cores utilizing the secondary magnetisation. In: SILVER, E. A., RANG1N, C., VON BREYMANN, M. T., er AL. Proceedings of the Ocean Drilling Program, Scientific Results, 124, 519-523. SILVER, E. A. & RANGIN, C. 1991. Leg 124 tectonic synthesis. In: SILVER, E. A., RANGIN, C., VON BREYMANN,M. T., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 124, 3-9. & SMITH, R. B. 1983. Comparison of terrane accretion in modern Southeast Asia and the Mesozoic North American cordillera. Geology, 11 198-202. , MCCAFFREY, R. & SMITH, R. B. 1983a. Collision, rotation and the initiation of subduction in the evolution of Sulawesi, Indonesia. Journal of Geophysical Research, 88, 9407-9418. --, JOYODrWlRYO, Y. & STEVENS, S. 1983b. Ophiolite emplacement by collision between the Sula platform and the Sulawesi island arc, Indonesia. Journal of Geophysical Research, 88, 9419-9435.
183
., GILL, J. B., SCHWARTZ, D., PRASETYO, H. & DUNCAN, R. A. 1985. Evidence for a submerged and displaced continental borderland, north Banda Sea, Indonesia. Geology, 13, 687-691. SIMANDJUNTAK,T. O. 1986. Sedimentology and tectonics
of the collision complex in the East Arm of Sulawesi, Indonesia. PhD Thesis, University of London. -
1992. New data on the age of the ophiolite in Eastern Sulawesi. Bulletin Geological Research and Development Centre, Bandung, Indonesia, 15, 38--44. SITUMORANG,B. 1982. Formation, evolution, and hydrocarbon prospect of the Makassar basin, Indonesia. -
In: Transactions of the Third Circum-Pacific Energy and Mineral Resources Conference, 227-231. -
1987. Seismic stratigraphy of the Makassar basin.
-
Scientific Contribution on Petroleum Science and Technology Lemigas Sumbangan Ilmiah, 1/87, 3-38. SMITH, R. B. & SILVER,E. A. 1991. Geology of a Miocene collision complex, Buton, eastern Indonesia. Geological Society of America Bulletin, 103, 660678. SOPAHELUWAKAN, J. 1990. Ophlolite Obduction in the
Mutis Complex, Timor, Eastern Indonesia: an Example of Inverted, Isobaric, Medium-High Pressure Metamorphism. Vrije University Press, Amsterdam. SRWASTAVA, S. P. & TAPSCOTT, C. R. 1986. Plate kinematics of the north Atlantic. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The Geology of North America. Volume M. Geological Society of America, 379-404. STEPHAN, J-E, BLANCHET,R., RANGIN, C., PELLETIER,B., LETOUZEY, J. & MULLER, C. 1986. Geodynamic evolution of the Taiwan-Luzon-Mindoro belt since the Late Eocene. Tectonophysics, 125, 245-268. STERN, R. J. & BLOOMER, S. H. 1992. Subduction zone infancy: examples from the Eocene Izu-BoninMariana and Jurassic California arcs. Bulletin of the Geological Society of America, 104, 1621-1636. SURMONT,J., LAJ, C., KISSEL,K., RANGIN,C., BELLON,H. & PPdADt, B. 1994. New palaeomagnetic constraints on the Cenozoic tectonic evolution of the north Arm of Sulawesi, Indonesia. Earth and Planetary Science Letters, 121, 629-638. TAN, D. N. K. & LAMY, J. M. 1990. Tectonic evolution of the NW Sabah continental margin since the Late Eocene. Bulletin of the Geological Society of Malaysia, 27, 241-260. TAPPONNIER, P., PELTZER, G., LEDAIN, A., ARMIJO, R. & COBBOLD, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611-616. TAYLOR, B. 1992. Rifting and the volcano-tectonic evolution of the Izu-Bonin-Mariana arc. In: TAYLOR, B. & FUIJIOKA, K., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 126, 627-651. & HAYES, D. E. 1980. The tectonic evolution of the South China Basin. In: HAYES, D. E. (ed.) The -
-
Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 89-104.
184
R. HALL
& 1983. Origin and history of the South China Basin. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands, Part 2. American Geophysical Union Monograph, 27, 23-56. TJtA, H. D. & Lmw, K. K. 1996. Changes in tectonic stress field in northern Sunda shelf basins. This volume. VAN OER WEERD, A. A. & ARMIN, R. A. 1992. Origin and evolution of the Tertiary hydrocarbon-bearing basins in Kalimantan (Borneo), Indonesia. AAPG Bulletin, 76, 1778-1803. VAN DER Woo, R. 1993. Paleomagnetism of the Atlantic, Tethys and lapetus Oceans. Cambridge University Press. VAN GOOL, M., HUSON, W. J., PRAWlRASASRA, R. & OWEN, T. R. 1987. Heat flow and seismic observations in the northwestern Banda arc. Journal of Geophysical Research, 92, 2581-2586. VILLENEUVE, M., CORNICE,J. J., MARTINI, R., ZANINETTI, L., RI~HAULT, J-P., ET AL. 1994. Upper Triassic shallow water limestones in the Sinta ridge (Banda Sea, Indonesia). Geo-Marine Letters, 14, 29-35. VISSER, W. A. & HERMES, J. J. 1962. Geological results of the exploration for oil in Netherlands New Guinea. Koniklijk Nederlands Mijnbouwkundig
Genootschaft Verhandlingen, Geologische Serie, 20. WAHYONO, H. & SONATA, W. 1987. Palaeomagnetism along transect VII. In: Joint CCOP/IOC Working Group on SEATAR 1987: Preliminary Report of the Jawa-Kalimantan Transect (Transect VII). Geological Research and Development Centre, Bandung, Indonesia, 1-10. WEISSEL, J. K. 1980. Evidence for Eocene oceanic crust in the Celebes Basin. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 37-48. & ANDERSON, R. N. 1978. Is there a Caroline Plate? Earth and Planetary Science Letters, 41, 143-158. WENSINK, H., HARTOSUKOHARDJO, S. ~; SURYANA, Y. 1989. Palaeomagnetism of Cretaceous sediments from Misool, northeastern Indonesia. Netherlands Journal of Sea Research, 24, 287-301. YAN, C. Y. & KgOENKE, L. W. 1993. A plate tectonic reconstruction of the SW Pacific 0-100 Ma. In: BERGER, T., KROENKE, L.W., MAYER, L., ET AL. Proceedings of the Ocean Drilling Program, Scient~c Results, 130, 697-709.
Contrasting tectonic styles in the Neogene orogenic belts of Indonesia T. O. S I M A N D J U N T A K 1 & A. J. B A R B E R 2
1 Geological Research and Development Centre, Jalan Diponegoro 57, Bandung 40122, Indonesia 2 Southeast Asia Research Group, Department of Geology, Royal Holloway, University of London, Egham TW20 OEX, UK Abstract: The recent compilation of a new tectonic map of Indonesia as part of the Geotectonic Map Project of East Asia has prompted a reassessment of the contrasted tectonic styles represented by currently developing orogenic belts. These orogenic styles provide a range of models illustrating the diversity and complexity of tectonic processes which may provide the key to the interpretation of other orogenic belts elsewhere in the world. The following distinctive types of orogenies have been recognized within Indonesia. 1. Sunda Orogeny in Java and Nusa Tenggara: involving subduction of oceanic crust with normal convergence, producing an orogenic belt of Andean type with trench, accretionary complex, forearc basin and Quaternary magmatic arc with active volcanoes built on the margin of the Sundaland continent. 2. Barisan Orogeny in Sumatra: with strongly oblique convergence and major strike slip transcurrent fault movement within the magmatic arc, along which a segment of continental crust is being displaced northwards along the western margin of Sundaland. 3. Talaud Orogeny in the northern Molucca Sea: convergence of the Sangihe and Halmahera oceanic magmatic arcs as the Molucca Sea Plate subsides beneath them. 4. Sulawesi Orogeny in eastern Sulawesi: collision of microcontinental blocks with subduction systems along the eastern margin of Sundaland. 5. Banda Orogeny in the southern Banda Arc between Sumba and Tanimbar: collision of the northern margin of the Australian continent with the subduction system along the southern segment of the Banda Arc. 6. Melanesian Orogeny in Irian Jaya and Papua New Guinea: a more advanced stage in the collision of the northern margin of the Australian continent with a magmatic arc on the Philippine Sea plate, commencing in the early Miocene with a partial reversal of polarity and the subduction of the Caroline plate beneath the collision zone. The present phase of orogenic activity in most of these occurrences commenced in midMiocene times and orogenic processes are still in progress.
Since the commencement of the first five year National Development Plan of Indonesia (Pelita I) in 1970 the geology of the whole country has been systematically and intensively investigated both by government institutions and by the private sector. These investigations have included the systematic geological mapping of the Indonesia archipelago, initially by the Geological Survey of Indonesia and later by the Geological Research and Development Centre (GRDC), partly in collaboration with the United States Geological Survey (USGS), the British Geological Survey (BGS), the Australian Bureau of Mineral Resources (BMR) (now the Australian Geological Survey Organisation: AGSO) and the Japanese International Cooperation Agency (JICA). At the same time geological and geophysical research has been continually in progress in many parts of Indonesia in collaboration with geoscientists from the University of London, UK; University of California, Santa Cruz, USA; Institute of Geophysics, Hawaii, USA;
INSU, France; the Snellius Programme, Netherlands; and JICA, Japan. Geological and geophysical investigations have also been carried out by the Indonesian government institutions and the private sector, including both domestic and foreign companies, in the exploration for energy and mineral resources. The result of these investigations has been an immense increase in the knowledge and understanding of the geological and tectonic development of the archipelago. The present paper is a review of this current understanding of the tectonic development of the Indonesian archipelago through Neogene times to the present day. The review arose out of the responsibilities of one author (TOS) in the compilation of the new Geotectonic Map of Indonesia as a member of the Working Group on the Geotectonic Map of East Asia organized by the Commission for the Geological Map of the World (CGMW) and the Committee for Coordination of Joint Prospecting for Mineral Resources in Asian
FromHall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, Geological Society Special Publication No. 106, pp. 185-201.
185
186
T.O. SIMANDJUNTAK t~ A. J. BARBER
Offshore Areas (CCOP) as part of the NW Quadrant Circum-Pacific Map Project. Tectonic
setting
of Indonesia
The present physiographic and tectonic configuration of Indonesia is considered to have developed since late Neogene times due to the interaction of three of the Earth's major lithospheric plates: the NNW-moving (c. 10 cm a-]) Philippine Sea plate; the NNE-moving (c. 8 c m a -1) Indo-Australian plate and the stationary, or slowly SE-moving (c. 0.4 cm a-1) Eurasian plate (Minster & Jordan 1978). The Indonesian archipelago therefore represents an immensely complicated triple junction, involving a complex pattern of small marginal ocean basins and microcontinental blocks bounded by subduction zones, extensional margins and major transcurrent faults. On the basis of geological and geophysical characteristics five regions of crust of different origins can be distinguished in the Indonesian region (Fig. 1). (1) Southeastern promontory of the Eurasian plate forming the Sundaland continental craton in Sumatra, West Java and Western Kalimantan; (2) Philippine Sea oceanic plate in the northeast; (3) Australian continental craton extending into the Indonesian region in Irian Jaya and the Arafura and Sahul
Platforms; (4) Indian oceanic plate in the southwest; (5) a transition zone, marking the zone of current plate interaction with active seismicity and volcanism extending through western Sumatra, eastern Java and the Banda arcs to northern Irian Jaya and through Sulawesi and the Moluccas to Mindanao in the Philippines in the central part of the region. In this zone subduction is still active, with the development of thrust, transcurrent and extensional faults. This region is also characterized by allochthonous cohtinental microplates with Tertiary and Mesozoic sediments overlying Palaeozoic basement, juxtaposed against Cretaceous and Tertiary terranes to form collision complexes.
Tectonic development of Indonesia P r e - N e o g e n e tectonics Western Indonesia. A concentric accretionary model for the tectonic evolution of western Indonesia, with the addition of successive orogenic belts to a continental core in Sumatra and Kalimantan from the Palaeozoic through the Mesozoic and Tertiary to the present day, was proposed by Katili (1973) and Gage & Wing (1980). Basaltic to andesitic volcanic rocks and granites dated at 276-298 Ma, and volcaniclastic
130 ° I'
j -.
~
CRATONS
...... PHILIPPINE. . "SEA ..... I ~ -i ...... PLATE ....... i"]
I~ ............
EURASIAN
125° "
MARGINALSEAS
~ .
OCEANPLATES
i.i South iii China ::: Sea :.:.:.:..
ITRANSITIONCOMPLEX
EI=I:~SUNDALAND: [27;-iCelebes :I ii:!:i:Sea.::::::;;
iiiiiii ...... ............... iiiii ................. ............... iiiiii 2o :....... CAROLINE PLATE.......... .......... ---_ t
,,; ,,; ,,: ,,." • ,,, ,, ; ,' ,, 5o,,'
:,'
:
%
" .................. ................
" ...........
--_
- .........
O°
.'i'
:
,.
ioo / INDIAN OCEAN • , PLATE. ./' ,,:' ;:" •,', .', /~Java ,. T
~y~
,(.:.i.: Banda
Sumba% Ti r ,"e /' n/' * .;~
,.'l,0o,,: ,,",,, ./." ]is°." ' ,?0o .'j " ,y :'"' 95°'''' ;'"'/ ,::'"100o'': ,'": ,," Io5*' ,.' .,,,': .::',,,,"" :: :",y",y
Fig. 1. The structure of the Indonesian archipelago and its tectonic setting in SE Asia.
,,.NewGuinea
-AUSTRALIAN
NEOGENE OROGENIC BELTS OF INDONESIA sediments associated with limestones containing a Permian fauna in northern and central Sumatra, have been interpreted as representing a magmatic arc with associated forearc or backarc regions (Katili 1973; Silitonga & Kastowo 1975; Rosidi & Santoso 1976; Cameron et al. 1980; Simandjuntak et al. 1981). A similar assemblage of rocks was described from West Kalimantan (Gage & Wing 1980; Pieters & Supriatna 1990; GRDC 1992), from Sarawak (Tate 1991) and from the Main Range of peninsular Malaysia (Hutchison 1973; Gage & Wing 1980). Pupilli (1973) and Gage & Wing (1980) suggested that these Permian magmatic arc systems were related to an Asia-ward subduction system in Sumatra and an Indian Ocean-ward subduction system in Kalimantan. A similar opposed double subduction system is proposed by Hutchison (1973) and Pupilli (1973) for the Triassic, with the Malaysian-Indonesian Tin Belt representing the magmatic belt related to subduction from the Asian directed subduction system, while the opposing system is represented by the Serian Volcanics in Sarawak. Cretaceous accretion complexes related to the Asian-ward subduction system can be traced through the melange wedges of this age in North and West Sumatra and Bengkulu, in Ciletuh and Karangsambung in Java, the Meratus Mountains of SE Kalimantan and in Bantimala and Wasapondo in South and central Sulawesi (Asikin 1974; Simandjuntak 1980; Sukamto 1986; Wajzer et al. 1991; GRDC 1995). Cretaceous magmatic arcs related to this system are represented by granites, basaltic-andesitic volcanics and volcaniclastics at Gumai and Garba Mountains in South Sumatra, the Meratus Mountains of SE Kalimantan and in SW Sulawesi (Musper 1937; Katili 1973; de Coster 1974; Sukamto 1975; Sikumbang & Heryanto 1986; GRDC 1992). These rocks together with associated sediments may represent forearc, trench slope, volcanic arc and backarc assemblages. The opposing Indian Ocean-ward directed system is represented by volcanic rocks and associated granites in the Anambas, Tambelan and Natuna islands in the South China Sea (Katili 1973; Pupilli 1973), the Schwaner Mountains of West Kalimantan and the Kuching-Sibu zone in Sarawak (Haile 1972). Palaeogene subduction systems can be traced by melanges through the outer arc islands of Nias, Pagai and Sipora, offshore Sumatra and in S. Java (Katili 1973; Karig et al. 1978; Hamilton 1979) and by volcanics and intrusives, the 'Older Andesites' of van Bemmelen (1949) associated with volcaniclastic, volcanic and hemipelagic sediments in Sumatra, Java and W Sulawesi (Katili 1973; de Coster 1974; Djuri & Sudjatmiko 1975; Sukamto 1975; Koesoemadinata et al. 1978; Cameron et al.
187
1980; Sukamto & Simandjuntak 1983), extending into N. Kalimantan, Sabah and Sarawak (Hutchison 1988; Pieters & Supriatna 1990; Tongkul 1991). Palaeogene volcaniclastic, siliciclastic and carbonate sediments on the eastern flanks of the Barisan Mountains in Sumatra and in S. Java, SE Kalimantan and South Sulawesi, followed by shallow marine to terrestrial deposits with economic coals and hydrocarbon source rocks, may represent interarc and backarc sequences related to the Palaeogene subduction systems of western Indonesia (de Coster 1974; Koesoemadinata et al. 1978; Cameron et al. 1980; GRDC 1995). I n d o n e s i a . In the Neogene a number of microcontinents, e.g. Buton, Banggai-Sula, collided with subduction systems along the eastern margin of Sundaland to form the present tectonic framework of eastern Indonesia. The pre-Neogene history of eastern Indonesia relates therefore to the origin and earlier history of these microcontinents before these collisions took place. All the microcontinents are considered originally to have formed part of the northern margin of the Australian continent and to have separated during Mesozoic or Palaeogene time. The history of break up and the subsequent development of these microcontinents is represented by sedimentary sequences on the northern margin of Australia, in northern Irian Jaya, the Arafura and Sahul platforms and within the microcontinents themselves (Dow 1977; Pieters et al. 1983; Pigram & Panggabean 1984; Simandjuntak 1986; Dow et al. 1988; Garrard et al. 1988). Thermal doming, volcanism, rifting and subsidence to form sedimentary basins commenced during the Palaeozoic on the northern margin of the Australian continent, which at that time formed part of Gondwanaland (Veevers 1984), and intensified during the Permian and Triassic (cf. Bird & Cook 1991 from Timor). The exact timing of separation of the microcontinents from Australia is still very much in dispute. Pigram & Panggabean (1984), based on the model of breakup sequences proposed by Falvey & Mutter (1981), suggested that the separation of microcontinents from central Papua New Guinea occurred in the early Jurassic between 141 ° and 145°E longitude. Rifting with the deposition of red beds in Irian Jaya and Papua New Guinea was followed by the deposition of a passive margin sequence which continued into the early Tertiary. It was earlier suggested that the microcontinents were detached from Irian Jaya in the mid-Tertiary and displaced westwards by movements along the Sorong Fault and/or its subsidiary traces (Visser & Hermes 1962; Krause 1965; Katili 1971; Gribi 1973; Hamilton 1979; Silver & Smith 1983). Eastern
188
T.O. SIMANDJUNTAK8z A. J. BARBER
However, the Jurassic passive margin sequence preserved in the microcontinents is overlain paraunconformably by Upper Cretaceous pelagic calcilutites; the Early to mid-Cretaceous was a period of non-deposition in the microcontinental blocks (Pigram et al. 1985; Simandjuntak 1986; Garrard et al. 1988). This depositional hiatus, followed by the abrupt deepening of the margins of the microcontinents in the Late Cretaceous, has been used to argue that the microcontinents were already detached and displaced from the northern margin of Australia by this time (Simandjuntak 1986, 1994; R6hault et al. 1991).
contribution to the development of five orogenic types in Indonesia (Fig. 2). Sunda orogeny
The Late Neogene Sunda orogeny affected the segment of the Indonesian arc between West Java and the islands of Nusa Tenggara as far east as Flores (Fig. 3). In this segment of the arc convergence between the Indian Ocean and SE Asian plates is normal to the subduction trace in the Java Trench with a rate of c. 7 cm a-1. The subduction system comprises an accretionary complex composed of offscraped Indian Ocean floor materials in the Java forearc ridge, a forearc basin developed on extended continental crust and containing late Palaeogene to Recent sediments. The volcanic arc which forms the backbone of Java and forms the islands to the east is constructed on continental crust, in West Java, on Mesozoic accretionary complexes in Central and East Java and on oceanic crust in Sumbawa and Flores. To the north of the arc in Java and in the Java Sea Tertiary backarc basins have developed on continental crust in the Sunda Shelf and on oceanic crust north of Bali and Flores. The basins on the Sunda Shelf formed in the late Palaeogene as rift basins in a terrestrial environment and subsided in the Neogene to be transgressed and covered by marine sediments. In the Late Neogene this system was affected by compression associated with the Sunda orogeny.
Neogene orogenies in Indonesia The present physiography of the Indonesian archipelago can be attributed directly to Neogene orogenic events. These events included plate convergence with subduction beneath the margins of Sundaland to produce a Cordilleran type of orogeny, a unique arc-arc collision in the Moluccas, collision of arcs and microcontinents and major continental blocks with the construction of mountain belts, transpression and transtension along major transcurrent belts, the construction of foreland fold and thrust belts, back-arc thrusting and reversal of subduction polarity. Examples of all of these types of orogenic event can be found among the islands of the Indonesian archipelago. These events will be described in terms of their
5° r
~
I°°
"P$
~
INDIAN
'S-~
]fig. 3
.
.
)f~
~OC,EN¥
.
.
.
TALAUD
~ Fig6-
~
% [ ~
°
~
BANDA[
~,~,~_~ -
95°
l(~o
105°
0o
115°
]
| ~..__ OROGENY ~ c-q~J
1210°
~25 °
]:0 °
Fig.8
~-I0°~
q]
~
AUSTRALIA ~x. 135~
11400/
,5o|
14; ° s.~.,,m..,,I~I.1
Fig. 2. Location of Neogene orogenies in Indonesia. Detailed figs to illustrate the structure of these orogenies are specified.
NEOGENE OROGENIC BELTS OF INDONESIA
189
Java) and Pekalongan (Central Java) show that segments of the backarc thrusts are currently active (Kertapati et al. 1972). The causes of the Late Neogene Sunda orogeny, with a change from an extensional to a compressional regime across the subduction system, are not clear. Compression may develop when smooth subduction of the downgoing oceanic plate is interrupted by topographic irregularities on the sea floor. This explanation could apply in the East Java-Bali segment of the subduction system where the Roo Rise is beginning to enter the subduction trench (Fig. 3), but does not explain compression elsewhere in the arc. It may be that the compression is localized along the northern margin of the volcanic arc due to magmatic intrusion and the consequent uplift of the arc.
In N. Java Mio-Pliocene turbidites are deformed into tight, locally isoclinal folds, readily observed in the field and on aerial photographs, while in S. Java-Nusa Tenggara older volcanic sequences are folded, faulted and uplifted to form mountains more than 3500 m above sea level. This phase of folding is associated in time with the intrusion of acid plutons, the uplift of the volcanic arc, the development of a major thrust system in which the volcanic arc is overthrust towards the Java Sea to the north, and the subsidence of the backarc basins, with the deposition of fine siliciclastic sediments, marls and carbonates of Plio-Pleistocene age. In N. Java the trace of a major backthrust, the Barabis-Kendeng Thrust (Simandjuntak 1992) can be traced from the Sunda Strait eastwards across Java and through the Bali Basin into the Flores Thrust, north of Flores (Prasetyo 1988). The thrust may continue eastwards as the Wetar Thrust to the north of Timor. The Barabis-Kendeng Thrust has been imaged in seismic reflection profiles in the northern part of West Java (Supryanto & Ibrahim 1993) and offshore north of Flores (Hamilton 1979; Prasetyo 1988), whilst the Bouguer gravity anomaly pattern in the northern part of East Java indicates the location of the Kendeng Thrust. In Central Java the thrust is cut and disrupted by the Cimandari and Citandui Faults which have wrench components of movement (Dardji et al. 1994). Earthquakes recorded at Majalengka, Brebes (West
Barisan orogeny
The Late Neogene Barisan orogeny affected the segment of the Indonesian arc occupied by the island of Sumatra (Fig. 4). Convergence in this segment of the arc between the Indian Ocean and Southeast Asian plates at a rate of c. 7 cm a-1, is currently markedly oblique (50-65°). The subduction system includes an accretionary complex, exposed in the offshore islands such as Nias (Moore & Karig 1980), and a forearc basin. The volcanic arc is constructed on continental crust and
Sulawesi J
/ NMD
• / Christmas ~lz°s
/ [
I /
convergoncoA-~j
/
~/~~Lienti2f I
~-~
ava Tren,,h
I Roo Rise
I ~ S
~' ~ ~--~
/
~ J
Activevolcanoesofthe VolcanicArc
Quaternaryvolcanicdeposits I ExposedPre-TertiaryRocks Cm(Cimandari)Fault;Ct (Citandui)Fault S P SeribuPlatform, A-AD Asem-AsemDepression WF E WestFlorenceEmbayment,A D ArdjunaDepres: E F B EastFlorenceBasin, M D MasalemboDepression N M D NoithMaduraDepression
Fig. 3. Sunda orogeny in Java and Nusa Tenggara. Line of section illustrated in Fig. 9 is indicated.
190
T . O . SIMANDJUNTAK •
(Koesoemadinata et a/.1978; Koning & Aulia 1985; Wang et al. 1989; Situmorang et al. 1991). Continuation of movements along the fault led to the opening of the Andaman Sea (Curray et al. 1979; Harding 1983). Commencement of the Barisan orogeny was marked by the uplift of the Barisan Mountains and arc volcanism, signalled by the influx of volcaniclastic sediments and regressive sequences in the Sumatran back-arc basin in the mid-Miocene. Uplift was accompanied by intrusions in the volcanic arc and transpressive movements along the Barisan Fault System. Fault movement resulted in the imbrication, duplexing and stacking of crustal slices to form a large-scale positive flower
the volcanoes sit on an uplifted pre-Tertiary basement terrane which forms the Barisan Mountains running the whole length of Sumatra. Both the basement and the volcanic arc are dissected by the Barisan dextral transcurrent fault system. Extensive Tertiary backarc basins lie to the northeast, behind the arc. Sumatra forms part of the Sundaland continental craton. In the Palaeogene the region was affected by extension and subsidence, resulting in the formation of rift basins. Some of these basins, such as the coal-bearing Ombilin Basin of West Sumatra, lie within the Barisan Fault System and have been interpreted as pull-apart basins formed by transtensional movements along the fault
I
I
96°
98°
A. J. BARBER
0 N
at'0km
Lake ada~n~ '(~lt ~xSingkarak
~
panjang
\
P
~ andEXtensi°nallakes rifts I x~.~olok " , r ~t.~ n;t,..... J,
West Sumatra Lake Toba , Mt. Kerinci
~
Jambi ngaipenuh
Bengkulu Line of section
• 4° ~
Active volcano
()ff
Quaternary-Recent Volcanics Tertiary Sedimentary Basins
%
Pre-Tertiary Basement 6°
0
100
200
300
400
n
[
500km
•
,\\ • Krakatoa
•
96° J
98° ]
100° I
NN 102°
[
104°
[
Steve Moss, J 995.
106° I
I
Fig. 4. Barisan orogeny in Sumatra. Crossed lines in the back-arc basins indicate fold axial traces. Inset map shows extensional features in the section of the Barisan Fault between Bukittinggi and Lake Kerinci (location indicated on
main map). Line of section illustrated in Fig. 9 is indicated.
NEOGENE OROGENIC BELTS OF INDONESIA
structure, leading to the erosion of the cover and exposure of the basement rocks which now rise to nearly 4000 m above sea level. Uplift of the Barisan Mountains was accompanied by subsidence of the forearc and back-arc basins. Plio-Pleistocene transpressive movements along the Barisan Fault System are considered to have induced fold structures in the sediments of the back-arc basins on axes which trend at an angle of 20 ° to the main fault (de Coster 1974; Hamilton 1979). The folds are commonly associated with small-scale wrench faults and may be related to large-scale wrench faults in the underlying basement (Tiltman 1990). Pleistocene movements along the fault system have, at least locally been transtensional with the opening of pull-apart basins, often holding lakes, such as Laut Tawar, Toba, Singkarak, Diatas, Dibawah, Kerinci (Fig. 4, inset), Ranau or a Recent sedimentary fill as in the Semangko Valley. Recent earthquakes resulting in the cessation of sulphuric hot springs and the disappearance of ponds at Tarutung (North Sumatra), and at Padang Panjang in Central Sumatra, indicate dextral transpressional reactivation along the northern segment of the Barisan Fault System. In Lampung Province, S. Sumatra NW-SE extensional cracks formed during an earthquake near Liwa in early 1994, indicating sinistral transtensional movement along this segment of the fault. The Neogene Barisan orogeny is attributed to variations in the rate of subduction of the Indian Ocean plate and the response of SE Asia to the continuing collision of India with the southern margin of Asia and the adjustment of crustal blocks by movements along major transcurrent faults (Tapponnier et al. 1982). The obliquity of subduction is considered to be responsible for the development of the Barisan Fault and the detachment of the Sumatran forearc to form a 'sliver plate', which is partially coupled to the northward movement of the Indian Ocean plate (Fitch 1972; Curray 1989; McCaffrey 1991). Movements along the Barisan Fault are considered to be responsible for the uplift of the Barisan Mountains and the transtensional and transpressional effects seen along the fault trace. Results of break-out analysis of sediments from boreholes in the Sumatran backarc basins show that the effects of oblique subduction are also being transmitted across the Barisan Fault into the backarc (Mount & Suppe 1992). Talaud o r o g e n y
The Neogene Talaud orogeny of the N. Moluccas provides the only example of an active arc-arc collision in the world. The Molucca Sea is bounded
191
to the west by the Sangihe volcanic arc and to the east by the Halmahera Arc (Fig. 5). Seismicity shows that the Sangihe Arc is underlain by a westdipping Benioff zone extending to a depth of 700 km, while an east-dipping Benioff zone underlies the Halmahera arc to a depth of 200 km (Silver & Moore 1978; Hamilton 1979; Sukamto et al. 1981). The shallow Molucca Sea between the two arcs is considered to be underlain by the collided accretionary complexes of the two opposing arcs uplifted to form the central TalaudMayu Ridge, the underlying Molucca Sea plate having subsided back into the mantle between them. Hamilton (1979), on the basis of seismic reflection profiles, suggested that the volcanic aprons from the forearcs of the Sangihe and Halmahera arcs had been thrust backwards across their corresponding arcs as the result of the collision. Recent Seabeam and seismic reflection data from the northern part of the Molucca Sea indicate that the floor of the sea consists entirely of the Sangihe forearc which has been thrust eastwards over the Halmahera forearc (Rangin et al. 1996). To the south the collision complex is truncated by a major strand of the Sorong Fault System. The Sangihe subduction system may swing round to the west and continue as the Batui Thrust in the East Arm of Sulawesi. Frequent earthquakes show that all these structures are currently active (McCaffrey et al. 1983; Simandjuntak 1989; Kertapati et al. 1972). Sulawesi orogeny
Sulawesi is made up of three structural units: preCretaceous accretionary material in the west upon which later developed a Neogene volcanic arc; Central Sulawesi and part of the SE Arm are composed of metamorphic rocks; the E. Arm and the remainder of the SE Arm are made up of a major ophiolite complex. The metamorphic rocks in Central Sulawesi and in the SE Arm include materials of both continental and oceanic derivation. These rocks are locally affected by high pressure metamorphism forming blueschists, developed in an east dipping Paleogene midoceanic subduction zone. Subduction resulted in the obduction of the East Sulawesi Ophiolite westwards across the metamorphic belt, accompanied by the extrusion of ophiolitic melange (Parkinson 1991). Neogene orogeny in Sulawesi was initiated by the collision of the two microcontinental blocks of Buton-Tukangbesi and Banggai-Sula with the eastern part of the island (Fig. 6). These two microcontinental blocks, having separated from the northern continental margin of Australia, possibly from the region of central New Guinea as has been
192
T. O. SIMANDJUNTAK~ A. J. BARBER
- - I • Active Volcano ~
Quaternary Volcanic Deposits
~
Mindanao
PHILIPPINE SEA
Molucca Sea Collision Complex 6°N
CELEBES SEA
4°N
; / --MOLUC A
lO00m
3000m
Line of Section Morotai !°N
)00m
.....
Sulawesi
Halmahera
oo
7_7/
Batui Thrust
I
'~
12~
~.~
~
~
t
~
~
~
l
t
~u~-
!{ S O R O N G FAUI~T S Y S T E M
,~oE
[
,2U
Fig. 5. Talaud orogeny in the North Moluccas. Line of section illustrated in Fig. 9 is indicated.
previously discussed, were carried westwards along the Sorong transcurrent fault zone by movements of the Philippine Sea plate (Ali & Hall 1995) and collided with the eastern margin of the ophiolite complex. The collision caused the obduction of the ophiolite onto the microcontinental blocks and the shortening and thickening of the ophiolite by
imbrication. The leading edges of the ButonTukangbesi and Banggai-Sula microcontinents were thrust beneath the ophiolite, uplifting the tightly folded, faulted and imbricated ophiolite and its pelagic cover to heights of more than 3000 m above sea-level (Smith 1983; Simandjuntak 1986; Garrard et al. 1988; Fortuin et al. 1990; Davidson
NEOGENE
OROGENIC
T
193
BELTS OF INDONESIA
T 123 °E
125°E
2ON -
Minahassa T r e n c h - ~ ~/ ~
lO00m \ ~ ~
~
3000m
~/) North Arm :S~2~ 3000m
MOLUCCA SEA 0 o
s a'
(
Line o( East Arm ult 2°S
,
Central
B A N D A SEA Buru Southeast Arm 4°S
Buton
-,
i .k ng /
I \ Recent Deposits ~
i" ~;~
A
Continental Crust
Quaternary Volcanic Rocks ~
Metamorphic Rocks k , ~
Neogene Volcanic Rocks
Ophiolitic Rocks
~
,27°E
\"
Majene Fold Belt
I I S. Baker 19951
Fig. 6, Sulawesiorogeny. Line of section illustrated in Fig. 9 is indicated.
1991). Also as a result of the collision the metamorphic belt of Central Sulawesi was thrust westwards over West Sulawesi and uplifted to form mountain ranges of nearly 3000 m. Overthrusting resulted in the formation of a foreland fold and thrust belt in Tertiary sediments, the Majene Fold Belt, which continues to develop westwards to the present day, affecting Recent sediments in the Makassar Strait (Coffield et al. 1993; Bergman et al. 1996).
Associated with the collision, or following shortly after, was the development of the NNWSSE trending Palu-Koro sinistral transcurrent fault, along which eastern Sulawesi has been displaced northwards with respect to western Sulawesi. Simandjuntak (1993a) has discussed the effects of transtensional movements along this fault and related displacement of the island of Sumba to its present anomalous position in the Banda forearc. More recent transtensional movements during the
194
T.O. SIMANDJUNTAK t~ A. J. BARBER
Quaternary, continuing to the present time, are responsible for opening pull-apart basins, such as those of Lakes Poso, Matano and Towuti, as well as the Palu depression. Recent earthquakes along the Palu-Koro and related faults show that the system is currently active. Banda orogeny
The Neogene Banda orogeny is due to a continentarc collision where the northern margin of the Australian continent, moving NNE at c. 7 cm a-1 is colliding with the subduction system along the southern side of the Banda Arcs (Fig. 7). The deformation front, marking the southern limit of the collision complex, lies along the 2 km deep Timor trough. Seismic reflection profiling across the trough shows Australian passive margin sediments passing northwards down beneath the trough where, at the deformation front, the upper parts of the sequence have been uplifted on thrust surfaces to form fold ridges in an accretionary complex (Karig et al. 1987). The structure of the collision zone is exposed in
,;o\ _FLORES SEA
C
<
,;,,o
/"
*
the island of Timor (Barber et al. 1977; Charlton et al. 1991; Harris 1991). The island is composed of sediments of Permian to Pliocene age, of Australian affinity which have been folded, thrust and imbricated. The Australian sediments have been thrust beneath an ophiolite slab with an underlying metamorphic sole, representing the trace of the Benioff zone prior to collision, but now uplifted to heights of more than 3000 m above sea-level in Timor (Sopaheluwakan 1990). Most of this uplift occurred in the Neogene, as Miocene, Pliocene and Pleistocene forearc sediments resting on the complex show a history of deposition shallowing upwards from bathyal facies through shallow water deposits to Pleistocene coral reefs (de Smet et al. 1990). These facies changes indicate two periods of rapid uplift, one 2 Ma and the other c. 100 000 years ago, representing different stages in the collision. The collision between the Australian margin and the Banda Arc subduction system appears to have commenced in the region that is now East Timor, and the collision is at its most advanced stage in this segment of the arc. Here the forearc between
*
ACTI
SUMBA \ /
I
" ~ -
"----~ " , L P - ~ " ~ t - ~
f
12o ~t . , ~ Y / ) / ~ ~
~ f
................
o HiberniaReef O AshmoreReef O.~-v.-'l
q i l l l ] I I I I l I I I I I I I I I J
. A c ~ O ~ : a qff+H-H -
I,'x,,3
~#" I~/"~]KTAI]IA De'l'q~ I.I'~.J.L~II"51d-/-~IX.I/~
PLATEAU/~
o
t
Scott Reef
300
,,oo/2oo
120°
/
122°
I
I
I
400
500 . 124 ° I
,"~
~_ "tk I Y,,~xK.tr,.) , ¢ . ~ ~ " - ~ ~[-'~ , 4 " l - l - l - [ - l - l ~ q~r-I" i i i i i i i .
r~T TT |III1.1 v ~ - .
.
.
.
.
1"~ ~ .
I F I I I I I I I I I I I I I I I I t.d III I I1111111 II 91 I I I I I I I | I I I I ! I I n ~ I I I I t I 1 I I I I I t I I I /[111 I I I 11 I I11 I I II I I I l l l l l l l l l l l l l l l l l l l ¢11111111111111111ll i I 1 l l l l l l l I I l l l l 4 C J'l II |l IIII IIII1111111 I i ii ii ii11 i iii II II I I
~ /'t./J-l"Ifl I II I 1 1 I 1 I /"~l./t,.,]l I I~"XL.tl/"II I I I I I I I I1 ~["11 I I I I I | I I I I I I I J I I"~ -kl.I I I I I I I I I I ~ I IdJ126°l I I 1 I 1280 . ~ c ~ .,, ,.72130 ° I I ] i I I I ~t.,•, "~7 I I /1111111111LI ff I I I I X . . . ~ I ~ 'xl III I 111 '~l¢~ t~"~ I | I I i 1 i i i i i i i i I t ' l 1 1 I I I )~'1"-~ I'T I t 1q I I I I
I I I II I I I I I I IJl I I I I I 11 1132 ° I I II11 I I I I I I I i i
Fig. 7. Banda orogeny: the collision zone between the southern Banda Arcs and Australia near Timor, based on Hamilton (1979). Horizontal ruling, collision zone with islands in black; triangles, active volcanoes; small v, associated with Quaternary volcanics; larger v pattern, Neogene volcanics. Line of section illustrated in Fig. 9 is indicated.
NEOGENE OROGENIC BELTS OF INDONESIA the volcanic arc and the deformation front has been reduced to a width of less than 100 km, compared with over 400 km for the forearc east of Sumba where collision has commenced only recently. Continued northward movement of the Australian continent has been accommodated by the uplift of the collision complex in Timor and lateral extension along conjugate faults (Charlton et al. 1991). Compression of the forearc has virtually eliminated the forearc basin which is only 40 km wide to the north of East Timor. Furthermore, shallow seismic activity is absent beneath East Timor suggesting that the downgoing plate and the collision complex have become locked in this segment of the arc. Volcanic activity in the islands of Alor and Wetar to the north of East Timor ceased about 3 Ma (Abbott & Chamalaun 1981) and the volcanic arc now rests on a southwarddipping thrust plane, the Wetar Thrust. The development of a small accretionary complex to the north of the thrust (Breen et al. 1989) suggests that the whole of the forearc and the volcanic arc are being carried northwards with the movement of Australia, and are overthrusting the Banda Sea floor to the north. These may be the first indications of reversal of the polarity of the subduction system, with the Banda Sea floor passing down southwards beneath the northern margin of Australia. Recent deep-seismic reflection profiles to depths of up to 100 km, a short distance to the east of East Tirnor, show reflections representing thrust surfaces dipping both northwards and southwards beneath
the collision complex which has been uplifted as a wedge by the convergence of the Australian and the Banda Sea lithosphere (Snyder et al. 1996). Melanesian orogeny
The Neogene Melanesian orogeny in Irian Jaya and Papua New Guinea is considered to be the result of oblique convergence of the Australian and Philippine Sea (and Caroline) plates (Dow & Sukamto 1984). In the late Palaeogene or very early Neogene, the northern promontory of the Australian Continent collided with an oceanic island arc constructed on the southern margin of the Philippine Sea plate (Hall et al. 1995). The remnants of this arc are now distributed through the northern part of New Guinea (Pigram & Davies 1987), disrupted by transcurrent fault movements to form discrete terranes (Fig. 8). The Central Ranges of New Guinea are composed of a major ophiolite complex, here interpreted as the oceanic lithosphere which underlay the oceanic island arc, subsequently uplifted as the result of the collision with Australia. To the south of the ophiolite the Australian passive margin sediments are deformed by thin-skinned thrust tectonics into a foreland fold and thrust belt, with folding, imbrication and duplexing. At deeper levels seismicity indicates that the underlying Australian continental crust is also involved in the thrusting (Abers & McCaffrey 1988). The whole collision belt is being thrust southwards over the northern
t ~0. r ~ ~ Wai e°*~'~-SOR~Q~ FA UL~ ' ~
~
e
-----,.,,.~%~ 7~4"s, "~ " ~
,~~' ~
a t...,.-
"~.
~/~ ~ \~'-~~r~. ~ ~ ~
d /~
~
"~.\~"" "
~ / ~~f ~ / y / ~ / -~//~///
_
~ rY
. jty~,aom ~: . 8s m A~.(,Tanlmbar 1301*E
1321"E
13~'E
G~nea ~
i
1361E
i
PACIFIC OCEAN ? b t t ~
A
~ ~
~
~LAEOGENF'~"..~~ FORELAA,_ ~
g------J~~ '--' ~
1381°E
1401. . . .
Active Volcanoes
QuaternaryVolcanics PalaeogeneVolcanics Ophiolites& related
"~
"-,4N/'c~ . ~. k
B A S I ~ ' ~ ~ . -." . ~
i :
~
~E
Thrusts
Nx&
\
pMERAUKE 1pM~LAFUR~ TRA Lineof~/FORELANI) 2 Section~....~-" '
ARAFURA SHELF
195
" I, -x_
1441~E
I~4~'[Or~E
Fig. 8. Melanesian orogeny in New Guinea, based on Hamilton (1979), with modifications after BMR (1972) and Dow et al. (1986). LFB, Lengguru Fold Belt; T-AF, Tarera-Aiduna Fault; WT, Waipona Trough. Line of section illustrated in Fig. 9 is indicated.
196
T. O. SIMANDJUNTAK & A. J. BARBER
margin of the Australian continent. The southern margin of the overthrust belt is marked by the Asmat Thrust (Fig. 8). The progress of the Banda and Melanesian orogenies in eastern Indonesia has been controlled by irrregularities in the northern margin of the Australian continent (Charlton 1986). Collision c o m m e n c e d in New Guinea. Using the k n o w n rate of northward m o v e m e n t of the Australian continent, Simandjuntak (1993b) postulated that at the time of the original collision, the subduction zone to the north of New Guinea was continuous with the subduction zone to the south of the Banda Arcs. Continued northward m o v e m e n t of Australia since the collision has displaced the collision zone northwards. Simundjuntak (1993b) also suggested that the western m a r g i n of the N e w G u i n e a promontory is marked by the Waipona Fault which transects the Asmat Thrust. He speculated that the northward m o v e m e n t of New Guinea relative to the Banda Arcs is responsible for the northward curvature of the Banda Arcs near Tanimbar and for the curvature of the Lengguru Fold Belt of the Bird's Head of Irian Jaya. Transcurrent m o v e m e n t along this fault is transpressive near Tanimbar and transtensional in the Aru Trough Following the collision, continued northward m o v e m e n t of Australia has thrust the Palaeogene arc, now attached to the northern margin o f Australia, over the Caroline plate to the north in a m u c h quoted example of polarity reversal (Dewey & Bird 1970; Johnson & Jaques 1980; Cooper & Taylor 1987). Recent geophysical studies in the Manokwari Trough immediately to the north of Irian Jaya, however, show that subduction is no longer in progress there (Milsom et al. 1992). Extensive outcrops of andesitic volcanic rocks of
mid- to late Miocene age throughout northern New Guinea indicate that southward subduction of the Philippine Sea plate beneath Australia occurred at that time. Quaternary and Recent volcanism is largely restricted to Papua New Guinea, and the potassic nature of the volcanic rocks suggests that they are related to extension rather than to subduction (Dow 1977). The northern part of New Guinea is transected by a major E - W strike-slip fault system, the New Guinea Megashear, which continues westwards into the Sorong Fault System of the Moluccas, already mentioned. Some of the records of shallow earthquake activity, especially concentrated in central Irian Jaya, may be related to continued m o v e m e n t on this fault system. Most records, however, show a thrust sense of movement related to compression in the M o m b e r a m a Thrust Belt (Abers & McCaffrey 1988). These thrust movements may be the effects of transpression which may also be responsible for the uplift of the Central Ranges to a height of nearly 6000 m. Transtensional movements in other parts of the fault system may be responsible for the opening up of rifts such as the Markham Valley in Papua New Guinea (Hill & Gleadow 1989). Transcurrent fault movements in northern New Guinea are attributed to the c o n t i n u e d westward m o v e m e n t of the C a r o l i n e / P h i l i p p i n e Sea plate at a rate of 12.5 cm a -l relative to the Australian plate.
Conclusions The whole range of plate tectonic and orogenic processes of subduction, accretion, construction of volcanic arcs, back-arc thrusting, strike-slip faulting, formation of sliver plates, arc-arc,
Fig. 9. Comparative interpretative cross-sections across the Neogene Orogenic Belts of Indonesia, all at the same scale. (a) Sunda orogeny. Convergence between Indian Ocean and Eurasian plates is normal to the subduction trench. The system consists of an accretionary complex, forearc basin, volcanic arc and backarc basin. Compression across the system has resulted in backarc thrusting. (b) Barisan orogeny. Oblique convergence of the Indian Ocean and Eurasian plates. The system is modified by the development of a 'sliver plate' where the forearc is driven northwards along the Barisan strike-slip fault by the movement of the Indian Ocean plate. (c) Talaud orogeny. The Molucca Sea plate is subsiding between the colliding Sangihe and Halmahera forearcs. The Sangihe forearc has been thrust over the Halmahera forearc to form the Talaud Ridge. Subduction of the Sulawesi Sea plate has commenced only recently. (d) Sulawesi orogeny. The Banggai-Sula microcontinent has collided with the ophiolite in the East Arm of Sulawesi. Repercussions of this collision include strike-slip movements along the Palu-Koro Fault and overthrusting in the Makassar Strait. (e) Banda orogeny. The Australian craton has been subducted ~.eneath the accretionary and collision complex in the Timor Ridge. The ridge is composed of accreted Australian continental margin sediments and earlier collided microcontinents. Subduction has ceased in this section of the collision zone, but the collision complex is being driven over the volcanic arc, which is in turn being driven over the Banda Sea plate to the north, in the earliest stages of subduction reversal. (f) Melanesian orogeny. The Australian Craton has been subducted beneath a Palaeogene volcanic arc with the development of a foreland fold and thrust belt from its overlying sediments. Ophiolite and arc volcanics in the central Ranges represent the uplifted hanging wall of the subduction zone. The system has reversed its polarity with the subduction of the Caroline Sea plate but the system is truncated by movement along the Sorong strike-slip fault zone due to the westward movement of the Caroline Sea plate. Data are drawn from references quoted in the text of the paper.
NEOGENE
SUNDA
BELTS
OF
Forearc Ridge Basin : -ml
Jawa Trench
0 ;.
[ii:i:
JAWA Volcanic Arc ~.-_.d~,._~
.
__ ___
:::::::::::::::::::::::::::::::::::::
JAWA SEA Sunda Shelf Kendeng Thrust __ ~ ~ ~-..--
-
,
esozo,c Accret,onaryComplexes
.: ......................
197
INDONESIA
OROGENY
INDIAN OCEAN RoD Rise
S
OROGENIC
~
~-~r/)
....
-Jr"
v- ~
~- SUNDA~ND _,×7
X
F
N
Jr
'
L
X
..,11..
f
INDIAN OCE
_-F__ "/'._I
~o
:iii!!:!
(a)
BARISAN OROGENY SW
INDIAN OCEAN Nicobar Fan
o
.
~
~;~ ~# ~
.
NIAS Forearc
Sunda Trench
.
.
.
R dge
BARISAN MOUNTAINS Barisan Fault Toba Caldera
flas n
s
< :~:~;~::::.,:::....~....................
X 4i-
4X
MALACCASTRAIT
Zs u N D A L A N D 4" F
:::i::ii~.i}i;::~:~: PLATE
~ i:~i:~:~
!~.:~iiiii~::iii!::~i{i~!:::.~.... :::~!ii:: !~!i ii}?i:~ili ::ii i ii
:iii
4-
~+~t
~m
~o
NE
Backam Basn
ii i~:i::i~ i! ! EURAS AN PLATE
i
~,,~,,,::,~:~::::~'':'~i!i!i!~i'{~ii,ii~!ii,i'iiii',!'~i~:i ii~i!iiii' i{i!i!ii!i!iiiiigi~!::~
(b) TALAUD
OROGENY
SULAWESI SEA W O
L 1
~ 1
~
NORTH SULAWESI Sang~e Arc thrust ,..-
< ....
~\[~'~.~" '
1
-
~
"~
"
MOLUCCA SEA Talaud Ridge Mavu
". . . . ,..
". . . .
...
HALMAHERA Arc
PACIFIC OCEAN Halmahera-Waigeo Ophiolite Terrane
thrust -
~
. . . .
\\~-"-
.~'_'~. " ~ - 5 - ~ _ ' % ~ .~:}.~
\
\
E
- - -
\
\
~-~>\ \
4
. . . . . . . . . . .
, \
\
\
\
\q
Imbricatedophiolite . \ \ \ \ \ 1
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ _ ~
....;:~:~::~ii~::::::ii::i!!:::::,i::: ii::ii!iiill~ili::i}~ii:.i!iiiii i i::::ii::i~ii~7.if!~i:ii:~ :i i iii~ili!i!i!ii~ii:::iii:!i!~!i::ii~ili:.i::: i :.i!i iilliii:: :~:
,o
'.iiii:,?i~:~)!:!iiiii~iiii! iii':::i! i
(c) SULAWESI OROGENY CENTRALSULAWESI EASTARM SULAWESI MAKASSAR STRAIT MagmaticArc MetamorphicBelt OphioliteBelt thrust belt Palu-KoroFault Batui Thrust
W
--
b++ sU.~A~D
km
~"
• 7L
.
+
~"
r--;l~z:-q-'--. " ~'",~L_-7L_T~--------~r.,
~
9
C
O
.
N
+
T
%
~
50
'
+
~
,
~. ' + ' ~
SULA E
"-x -+- ,
",
""~
BANGGAI
N
..................................................
".~ 4-
~
_~ ' +
>:~' '
, "'a
-x 4-- .4.
X
~ ./.
% '+
"~ '
~
x + + ",
I
X
× + ~1
> : : : : : : : :1:~:+ :"~*:~;~"~ ~ .................1:...................." : 1 1 : . . . . . . . .
((1) BANDA
VOLCANIC ARC (inactive) BANDA SEA Kisar Thrust WetarThrust Gunung Api
OROGENY
S TIMOR SEA Sahul Shelf
TIMOR RIDGE Timor Trough
°b +
N .........
k4 . . . . . . . + 4. × 4- ./_ kml" + AUSTRALIANCRATON jr x + 4- x
50
+
(e) MELANESIAN
L+
+
~+
CENTRAL RANGES Palaeogene Arc
OROGENY
Dolak MeraukePlatform
S
+
Thrust B e l t Asmat Thrust
ForelandBasin
,+
+
~+
+
~
~
~
~
.
~
PACIFIC OCEAN Sorong Fault Zone MeervlakteBasra New Guinea Trench N
.
kml-4.+ x if'- + . X AUSTRALIAN cRATON x if" ~ if" +_, X ~ L -/_ "/- ~ ~ ,..,. ...L "/- -r- + ./. -r .~ ~,.,, ~ .X'Y. -k'-"/. k ' - "/- ~ r - "/. -k'-"/. - X " - - × ' X
~
50 0
1O0
200
300
400
500
600
700
800
198
T.O. SIMANDJUNTAK t~ A. J. BARBER
micro-continental-arc and continental-arc collision, foreland fold and thrust belt formation, obduction and the uplift o f mountain belts m a y be f o u n d currently in progress in the I n d o n e s i a n archipelago. Our present understanding o f the structure o f the N e o g e n e - R e c e n t orogenic belts o f Indonesia is illustrated in a series o f cross-sections in Fig. 9. These present-day e x a m p l e s o f orogenic processes p r o v i d e models for the interpretation o f older orogenic belts such as the A l p i n e - H i m a l a y a n system, the H e r c y n i d e s and the Caledonides.
The authors are indebted to the Director of the Geological Research and Development Centre, Bandung, for permission to publish this paper. H. Faeni and Sudarto of GRDC, and Steve Moss, Simon Baker and Andy McCarthy of the University of London assisted in the preparation of the illustrations. Robert Hall and Derek Blundell invited TOS to attend the meeting on the 'Tectonic Evolution of Southeast Asia' organized at the Geological Society, and arranged sponsorship. Robert Hall, Tim Charlton and Steve Moss made many suggestions for the improvement of the paper.
References
ABERS, G. & MCCAFFREY,R. 1988. Active deformation in the New Guinea fold and thrust belt: seismological evidence for strike-slip faulting and basement involved thrusting. Journal of Geophysical Research, 93,13 332-13 354. ABBOTT, M. J. • CHAMALAUN, F. H. 1981. Geochronology of some Banda Arc volcanics. In: BARBER, A. J. & WIRYOSUJONO, S. (eds) The
Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Special Publication, 2, 253-268. ALI, J. R. & HALL, R. 1995. Evolution of the boundary between the Philippine Sea plate and Australia: palaeomagnetic evidence from eastern Indonesia. Tectonophysics, 251, 251-275. ASIKIN, S. 1974. Evolusi tektonik Java Tengah dan
sekitarnya ditinjau dari segi toori tektonik dunia yang baru. PhD Thesis, ITB Bandung. BARBER,A. J., AUDLEY-CHARLES,M. G. & CARTER,D. G. 1977. Thrust tectonics in Timor, Journal of the Geological Society of Australia, 24, 51-62. BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. & GARRARD, R. J. 1996. Tertiary tectonic and magnetic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for a Miocene continent-continent collision. This volume. BIRD, P. R. & COOK, S. E. 1991. Permo-Triassic successions of the Kekneno area, West Timor: implications for palaeogeography and basin evolution. Journal of Southeast Asian Earth Sciences, 6, 359-371. BMR, 1972. Geology of Papua New Guinea. Scale 1:1,000,000. Bureau of Mineral Resources, Geology and Geophysics, Canberra. BREEN, N. A., StaYER, E. A. & ROOF, S. 1989. The Wetar back-arc thrust belt, Eastern Indonesia: effect of accretion against an irregularly shaped arc. Tectonics, 8, 803-820. CAMERON, N. R. CLARKE, M. G. C., ALDtSS, D. T., ASPDEN, J. A. & DJUNUDDIN, A. 1980. The geological evolution of northern Sumatra.
Proceedings of the Indonesian Association, 9, 149-188.
eastern Indonesia. Journal of Structural Geology, 13, 489-500. COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A., GURITNO, N., ROBINSON,N. M. & TALBOT, J. 1993. Tectonic and stratigraphic evolution of the Kalosi PSC area and associated development of a Tertiary petroleum system, South Sulawesi. Proceedings of the Indonesian Petroleum Association, 22, 679-706. COOPER, P. & TAYLOR, B. 1987. Seismotectonics of New Guinea: a model for arc reversal following arccontinent collision. Tectonics, 6, 53-67. CURRAY, J. R. 1989. The Sunda Arc: a model for oblique plate convergence. Netherlands Journal of Sea Research, 24, 131-140. , MOORE, D. G., LAWVER,L. A., EMMEL, F. J., RAITT, R. W., ET AL 1979. Tectonics of the Andaman Sea and Burma. In: WATKINS, J., MONTADERT, L. & DICKENSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 189-198. DARDJI, N., VILLEMIN, T & RAMPNOUX, J. P. 1994. Palaeostress and strike-slip movements: the Cimandiri Fault Zone, West Java, Indonesia. Journal of Southeast Asian Earth Sciences, 9, 3-11. DAVIDSON, J. W. 1991. The geology and prospectivity of Buton Island, Southeast Sulawesi, Indonesia.
Proceedings of the Indonesian Association, 20, 209-234.
Petroleum
DE COSTER, G. L. 1974. The geology of Central and South Sumatra Basins. Proceedings of the Indonesian Petroleum Association, 3, 77-110. DE SMET, M. E. M., FORTU1N,A. R., TROELSTRA,S. R., VAN MARLE, L. J., KARMINI, M., Er AL S. 1990. Detection of collision-related vertical movements in the outer Banda Arc (Timor, Indonesia) using micropalaeontological data. Journal of Southeast Asian Earth Sciences, 4, 337-356. DEWEY, J. F. & BIRD, J. M. 1970. Mountain belts and the new global tectonics. Journal of Geophysical Research, 75, 2625-2647. DJURI & SUDJATMIKO. 1975. Geological Map of the
Petroleum
Majene and western part of the Palopo Quadrangles, Scale 1:250,000. Geological Survey
CHARLTON, T. R. 1986. A plate tectonic model of the eastern Indonesian collision zone. Nature, 319, 394-396. --, BARBER, A. J. & BARKHAM, S. T. 1991. The structural evolution of the Timor collision complex,
of Indonesia. Dow, D. B. 1977. A geological synthesis of Papua New Guinea. Bulletin Bureau of Mineral Resources, Geology and Geophysics, 201, 1-41. - & SUKAMTO, R. 1984. Western Irian Jaya: the end
NEOGENE OROGENIC BELTS OF INDONESIA product of oblique plate convergence in the late Tertiary. Tectonophysics, 106, 109-139. --, ROBINSON, G. P., HARTONO, U. t~ RATMAN, N. 1988. Geology of Irian Jaya: Preliminary Geological Report. Geological Research and Development Centre - Bureau of Mineral Resources, Canberra. FALVEY, D. A. & MUTTER, J. C. 1981. Regional plate tectonics and the evolution of Australia's passive continental margins. Bureau of Mineral Resources
Journal of Australian Geology and Geophysics, 6, 1-24.
FITCH, T. J. 1972. Plate convergence, transcurrent faults and internal deformation adjacent to Southeast Asia and the Western Pacific. Journal of Geophysical Research, 77, 4432-4460. FORTUIN, A. R., DE SMET, M. E. M., HADIWASASTRA,S., VAN MARLE, L. J. TROELSTRA, S. R. & TJOKROSAPOETRO, S. 1990. Late Cenozoic sedimentary and tectonic history of South Buton, Indonesia. Journal of Southeast Asian Earth Sciences, 4, 107-124. GAGE, M. S. & WING, R. S. 1980. Southeast Asian basin types versus oil opportunities. Proceedings of the Indonesian PetroleumAssociation, 9, 123-147. GARRARD,R. A. SUPANDJONO,J. B. & SURONO,1988. The geology of the Banggai-Sula Microcontinent, Eastern Indonesia. Proceedings of the Indonesian Petroleum Association, 17, 23-52. GRDC. 1992. Geological Map of Indonesia, Scale 1:5,000,000. Geological Research and Development Centre, Bandung. 1995. Geotectonic Map of Indonesia, Scale 1:5,000,000. Geological Research and Development Centre, Bandung. GRIBI, E. A. JR. 1973. Tectonics and oil prospects of the Moluccas Eastern Indonesia. Bulletin Geological Society of Malaysia, 6, 11-I6. HAmE, N. S. 1972. Confirmation of a late Cretaceous age for granite from Bunguran and Anambas Islands, Sunda Shelf, Indonesia. Geological Society of Malaysia Newsletter, 30, 6-8. HALL, R., ALI, J. R., ANDERSON, C. D. & BAKER, S. J. 1995. Origin and motion history of the Philippine Sea plate: evidence from eastern Indonesia. Tectonophysics, 251, 229-250. HAMILTON, W. 1979. Tectonics of the Indonesian Region. United States Geological Survey Professional paper, 1078. HARDING,T. P. 1983. Divergent wrench fault and negative flower structure, Andaman Sea. In: BALLY, A. W. (ed.) Seismic Expression of Structural Styles. AAPG, Studies in Geology, 15, 4.2, 1-8. HARRIS, R. A. 1991. Temporal distribution of strain in the active Banda orogen: a reconciliation of rival hypotheses. Journal of Southeast Asian Earth Sciences, 6, 373-386. HILL, K. C. & GLEADOW, A. J. W. 1989. Uplift and thermal history of the Papuan Fold Belt, Papua New Guinea: apatite fission track analysis. AAPG Bulletin, 75, 857-872. HUTCHISON,C. S. 1973. Tectonic evolution of Sundaland: a Phanerozoic synthesis. Bulletin of the Geological Society of Malaysia, 6, 61-86..
199
1988. Stratigraphic-tectonic model for Eastern Borneo. Bulletin Geological Society of Malaysia, 22, 135-151. JOHNSON, R. W. & JAQUES, A. L. 1980. Continent-arc collision and reversal of arc polarity: new interpretations from a critical area. Tectonophysics, 63, 111-124. KARIG, D. E., SUPARKA,MOORE, G. E & HEHANUSSA,P. E. 1978 Structure and Cainozoic evolution of the Sunda Arc in the Central Sumatra Region. United Nations, ESCAP-CCOP Technical Bulletin, 12, 87-108. --, BARBER, A. J., CHARLTON,T. R., KLEMPERER, S HUSSONG, D. M. 1987. Nature and distribution of deformation across the Banda Arc-Australia collision zone at Timor. Bulletin Geological Society of America, 98, 18-32. KATILI, J. A. 1971. A review of geotectonic theories and tectonic map of Indonesia. Earth Science Reviews, 7, 143-163. 1973. On fitting certain geological and geophysical features of Indonesian island arcs to the new global tectonics. In: Coleman, M. (ed.) The
Western Pacific: Island Arcs, Marginal Seas, Geochemistry. University of Western Australia Press, 287-305.
KERTAPATI, E. K., SOEHAMI,A. & DJUHANDA,A. 1972. Seismotectonic Map of Indonesia. Geological Research and Development Centre, Bandung. KOESOEMADINATA, R. P., HARDJONO, USNA, I. & SUMDIRDJA, H. 1978. Tertiary coal basins of Indonesia. United Nations, ESCAP-CCOR Technical Bulletin, 12, 43-86. KONING, T. & AULIA, 1985. Petroleum geology of the Ombilin intermontane basin, West Sumatra.
Proceedings of the Indonesian Association, 14, 117-137.
Petroleum
KRAUSE, D. C. 1965. Tectonics, marine geology and bathymetry of the Celebes Sea-Sulu Sea regions. Bulletin of the Geological Society of America, 17, 813-823. MCCAFFREY, R. 1991. Slip vectors and stretching of the Sumatran forearc. Geology, 19, 881884. --, SUTARJO,R., Er AL. 1983 Micro earthquake survey of the Molucca Sea and Sulawesi, Indonesia.
Bulletin Geological Research and Development Centre, 7, 13-33. MINSTER, J. B. & JORDAN, T. H. 1978. Present day plate motion. Journal of Geophysical Research, 83, 5331-5334. MILSOM, J., MASSON, D., NICHOLS, G., SIKUMBANG,N., DWIYANTO,B., ETAL. 1992. The Manokwari Trough and the western end of the New Guinea Trench. Tectonics, 11, 145-153. MooRE, G. E & KARIG, D. E. 1980. Structural geology of Nias Island, Indonesia: implications for subduction zone tectonics. American Journal of Science, 280, 193-223. MOUNT, V. S. & SUPPE, J. 1992. Present-day stress orientations adjacent to active strike slip faults: California, and Sumatra. Journal of Geophysical Reasearch~ 97, 11 995-12 013. MUSPER, K. A. E R. 1937. Toelichting bijblad 16 (Lahat).
200
T.O. SIMANDJUNTAK • A. J. BARBER
Geologic Map of Sumatra. Scale 1:2,000,000. Geological Survey of Indonesia. PARKINSON, C. D. 1991. The petrology, structure and geologic history of the metamorphic rocks of Central Sulawesi, Indonesia. PhD Thesis, University of London. ~ETERS, P. E. & SUPRIATNA, S. 1990. Geological Map of West, Central and East Kalimantan Area. Scale 1:1,000,000. Geological Research and Development Centre, Bandung. , PIGRAM, C. J., TRAIL, D. S., DOW, D. B., RATMAN, N. & SUKAMTO, R. 1983. The stratigraphy of western Irian Jaya. Bulletin Geological Research and Development Centre, 8, 14-48. PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and accretion history of the New Guinea orogen. Bureau of Mineral Resources Journal of Australian Geology and Geophysics, 10, 193-212. - & PANGGABEAN,H. 1984. Rifting of the northern margin of the Australian continent and the origin of some microcontinents in Eastern Indonesia. Tectonophysics, 107, 331-353. , SURONO & SUPANDJONO,J. B. 1985. Geology and regional significance of the Sula Platform, Eastern Indonesia. Bulletin Geological Research and Development Centre, 11, 1-13. PRASETYO, H. 1988. Marine geology and tectonic development of the Banda Sea region, Eastern Indonesia: a model of an 'Indo-Borderland' marginal basin. PhD Thesis, University of California, Santa Cruz. PUPILLI, M. 1973. Geological evolution of South China Sea area: tentative reconstruction from borderland geology and well data. Proceedings of the Indonesian Petroleum Association, 2, 223-242. RANGIN, C., DAHRIN, D., QUEBRAL, R. & THE MODEC SCIENTIFIC PARTY 1996. Collision and strike-slip faulting in the northern Molucca Sea (Philippines and Indonesia): preliminary results of a morphotectonic study. This volume. REHAULT,J. P., MALOD,J. A., LARUE,M., BURHANUDDINN, S. • SARM1LI,L. 1991. A new sketch of the Central North Banda Sea, Eastern Indonesia. Journal of Southeast Asian Earth Sciences, 6, 329-334. ROSIDI, D. (~ SANTOSO,B. 1976. Geologic Map of Painan Quadrangle, West Sumatera. Scale: 1:250,000. Geological Survey of Indonesia. SIKUMBANG, N. (~z HERYANTO, 1986. Geological Map of the Banjarmasin Quadrangle, SE Kalimantan. 1:250,000. Geological Research and Development Centre, Bandung. SILITONGA, P. H. (~ KASTOWO, D. 1975. Geological Map of the Solok Quadrangle, Sumatra. 1:250,000. Geological Survey of Indonesia. SILVER, E. A. & MOORE, J. C. 1978. The Molucca Sea Collision Zone, Indonesia. Journal of Geophysical Research, 83, 169 l - 1691. & SMITH, R. B. 1983. Comparison of terrane accretion in modern SE Asia and Mesozoic North American Cordillera. Geology, 11, 198-202. SIMUNDJUNTAK, T. O. 1980. Wasaponda Melanges. Proceedings of the Indonesian Association of Geologists, 8, 00-00.
1986. Sedimentology and tectonics of the collision complex in the East Arm of Sulawesi, Indonesia. PhD Thesis, University of London. 1989. Tunjaman ganda melahirkan jalur gunung api muda di Minahasa. Suara Pembaruan, 16 June 1989. 1992. Neogene tectonic development of the Indonesian Archipelago. In: IGCP 246 Conference. IGCP-IAGI-GRDC, Bandung. - 1993a. Tectonic origin of Sumba Island. Journal of Geology and Mineral Resources, 22, 10-20. 1993b. Neogene tectonics and orogenesis of Indonesia. Journal of Geology and Mineral Resources, 20, 2-32. 1994. Tectonic development of the Indonesian Archipelago and its beating on the occurrence of hydrocarbons. AAPG Bulletin, 78, 1162. • SURONO & BUHITRISNA, T. 1981. Geological Map of West Central and East Kalimantan Area. Scale 1:1,000,000. Geological Research and Development Centre, Bandung (Open File). SITUMORANG,B., GUNTUR,A., YULIHANTO,B., HIMAWAN, R. & GAMAL JACOB, H. 1991. Structural development of the Ombilin Basin, West Sumatra. Proceedings of the Indonesian Petroleum Association, 20, 1-16. SMITH, R. E. 1983. Sedimentology and tectonics of a Miocene collision complex and overlying Late Neogene orogenic clastic strata, Buton Island, Eastern Indonesia. PhD Thesis, University of California, Santa Cruz. SNYDER, D. B., PRASETYO,H., BLUNDELL,D. J., FhGRAM, C. J., BARBER, A. J., ET AL. 1995. Style of crustal deformation across the Banda Arc continent-arc collision zone as observed on deep seismic reflection profiles. Tectonics, in press. SOPAHELUWAKAN, J 1990. Ophiolite obduction in the Mutis Complex, Timor Eastern Indonesia: an example of inverted isobaric medium/high pressure metamorphism. PhD Thesis, Free University Press, Amsterdam. SUKAMTO, R. 1975. Geological Map of Indonesia Sheet VIII, Ujung Pandang. Scale 1:1,000,000. Geological Survey of Indonesia, Bandung. 1986. Tectonik Sulawesi Selatan dengan acuan khusus, ciri-ciri himpunan batuan daerah Bantimala. Doctoral Thesis, ITB, Bandung. - (~ SIMANDJUNTAK,T. O. 1983. Tectonic relationship between geological provinces of Western Sulawesi, Eastern Sulawesi and Banggai-Sula in the light of sedimentological aspects. Bulletin Geological Research and Development Centre, 7, 1-12. , APANDI, Z., SUPRIATNA, S (~ YASIN, A. 1981. The geology and tectonics of Halmahera Island and surrounding areas. In: BARBER, A. J. & WIRYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, 2, 349-362. SUPRYANTO (~ IBRAHIM, A. T. M. 1993. Pengembangan kemampuan penambangan lepas pantai, masa depan Indonesia. Proceedings of the Indonesian Association of Geologists, 2, 1162-1174. --
NEOGENE OROGENIC BELTS OF INDONESIA TAPPONNIER, P., PEWTZER,G., LE DAIN, A. Y., ARM1JO,R. t~£ COBBOLD, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611-616. TATE, R. B. 1991. Cross-border correlation of geological formations in Sarawak and Kalimantan. Bulletin of the Geological Society of Malaysia, 28, 63-95. TILTMAN, C. J. 1990. A structural model for North Sumatra. In: Lemigas Scientific Contribution on Petroleum Science and Technology. Special Issue, 24-44. TONGKUL, F. 1991. Tectonic evolution of Sabah, Malaysia. Journal of Southeast Asian Earth Sciences, 6, 395-405. VAN BEMMELEN,R. W. 1949. The Geology of Indonesia. Government Printing Office, The Hague VEEVERS, J. J. 1984. The Phanerozoic Earth History of Australia. Oxford University Press.
201
VISSER, W. A. & HERMES,J. J. 1962. Geological results of the exploration for oil in Netherlands New Guinea. Verhandelingen Koninklijk Nederlands Geologisch en Mijnbouwkundig Genootschap, Geologische Serie, 20, 1-265. WAJZER, M. R., BARBER, A. J., HIDAYAT,S. & SUHARSONO 1991. Accretion, collision and strike-slip faulting: the Woyla Group as a key to the tectonic evolution of North Sumatra. Journal of Southeast Asian Earth Sciences, 6, 447--463. WANG, J. E, BUDIJANTO, E M., JOHNSON, M. I. & SIRINGORIN60, R. S. A. 1989. Neogene seismic sequence and structural styles in 'B' and Peusangan Blocks, North Sumatra Basin, Indonesia. Proceedings of the Indonesian Petroleum Association, 18, 339-361.
Palaeomagnetism of the Sibumasu and Indochina blocks: implications for the extrusion tectonic model BRYAN RICHTER
& MICHAEL
FULLER
Department of Geological Sciences, University of California, Santa Barbara CA 93106, USA. email:
[email protected] Abstract: The Jurassic-Cretaceous Kalaw redbeds of Myanmar yield a prefolding Late
Cretaceous to Early Palaeogene magnetization which records c. 25-30 ° of clockwise (CW) rotation relative to the South China block. This corresponds to 10-15 ° CW relative to the Lower Cretaceous Khorat Plateau VGP. The data also show 5 ° of northward transport relative to the 100 Ma South China VGP or 12° relative to the Khorat Lower Cretaceous VGP. Similar CW rotations are measured in remagnetized Palaeozoic carbonates in Peninsular Thailand and Langkawi Island, Malaysia. These block motions most likely took place between the Late Cretaceous and the Late Oligocene. These and other recently published data have several implications for the extrusion tectonic model: (i) Sundaland has only rotated 25-30 ° CW relative to South China during the Tertiary; (ii) southeastward translation is only 300-500 kin; and (iii) Sundaland is composed of smaller sub-blocks, some of which have moved northward. This is interpreted to indicate that deformation of the Sibumasu block is dominated by the oblique subduction of the Indian Ocean Plate while deformation of the Indochina block is dominated by extrusion, in turn driven by convergence between the Indian Craton and Eurasia.
SE Asia has been located at the intersection of the Eurasian, Indian, Australian, Pacific and Philippine Sea plates throughout much of the Cenozoic (e.g. Hamilton 1979). These plate interactions culminated in the Palaeogene to present collision of India with Tibet and the Neogene to present collision of Australia with Sulawesi, Java and Borneo. The boundary conditions set up by the relative motions between these plates have had a profound impact upon the Cenozoic motions of the terranes and micro-plates which comprise SE Asia. Convergence between the Indo-Australia and Eurasian plates continues today and ultimately the terranes of SE Asia will form a broad orogenic belt between these two plates. It has become clear over the last decade that SE Asia has moved with respect to Eurasia during the Cenozoic. Field data which document these motions have been difficult to obtain and often more difficult to interpret. Thus, much of what is postulated about Cenozoic motions is derived from laboratory experiments (e.g. Tapponnier et al. 1986) suggesting that much of the deformation associated with the India-Eurasia collision has been accommodated by rigid plate-wide clockwise (CW) rotation and southeastward translation of the Indochina and Sibumasu blocks out of the collision zone. The available Mesozoic and Cenozoic palaeomagnetic data from the Indochina and Sibumasu blocks qualitatively support the predictions of these
laboratory models. However, ambiguities regarding the tectonic significance of these palaeomagnetic data remain. Many of the tectonic conclusions are based upon several key sampling localities or well preserved chronostratigraphic successions and very little is known about the intervening regions or time periods. With regard to the actual palaeomagnetic analyses, it has proved difficult to discriminate between primary and secondary magnetizations. A review of the literature on the Khorat Plateau, for example, shows the same basic dataset interpreted first as a primary magnetization, reinterpreted as a secondary magnetization, then finally reinterpreted as a primary magnetization (Achache et al. 1983; Chen & Courtillot 1989; Yang & Besse 1993). It is imperative, therefore, to reevaluate existing datasets as new data become available. Despite these difficulties, palaeomagnetism remains one of the few direct records of ancient plate motions available to us. Towards this end we review the available palaeomagnetic data from Sundaland and we present new palaeomagnetic data collected from the Kalaw Redbeds of the Shan Plateau of Myanmar and from Palaeozoic limestones located in Central and Peninsular Thailand. Moreover, we evaluate the tectonic significance of these palaeomagnetic directions. For example, are they caused by local or plate-wide rotations? Are they controlled by structural deformation? Finally, we conclude by evaluating existing tectonic models and propose modifications which make the models
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 203-224.
203
204
B. RICHTER • M. FULLER
more consistent with the measured rotations and translations.
Regional tectonic framework The data presented in this paper have been collected in Thailand and Myanmar from the Tertiary geological 'plate' known as Sundaland (Fig. 1). Sundaland is composed of several Palaeozoic and Lower Mesozoic terranes (Fig. 2). The origin of these blocks and their suturing ages have, among others, been studied by Metcalfe (1988, 1990). Hutchison (1989) provides an excellent summary of additional studies. The new data presented in this study were collected from the elongate N-S trending Sibumasu block (Shan-Thai). In our discussion we compare these data to published data from the Indochina block (Fig. 2). Sibumasu block
The Sibumasu block consists of portions of Myanmar, Thailand, northwest Malaysia, and Sumatra (Fig. 2). It is bounded to the west by the dextral Sagaing Fault, the Andaman Sea and the dextral Sumatran Fault Zone; to the east by the Bentong Raub line, the extension of the BentongRaub line into the Gulf of Thailand, and finally by the Nan-Uttaradit Ophiolite line in northern
30°N
20°N
10 ° N
Thailand; and to the north by the Red River Fault and the Himalayan Syntaxis (Metcalfe 1988). It may have originally been continuous with the Qiangtang or Lhasa blocks but intense deformation makes it difficult to confirm this. There is some agreement that the long axis of the Sibumasu block was originally orientated approximately E-W and was attached to the northern margin of Gondwanaland during the Early Palaeozoic (e.g. Hutchison 1989). It rifted from Gondwanaland in the Middle or Late Palaeozoic and began travelling northward. Collision timing along the Bentong-Raub line in Peninsular Malaysia is widely thought to be marked by the Late Triassic S-type granites of the Main Range Province (e.g. Cobbing et al. 1986). In contrast, Helmcke (1986), Barr & Macdonald (1991), and Barr et al. (1990) have shown that the proposed suture zones in Northern Thailand (Nan River Belt, Phetchabun Fold belt) have a more complicated collisional history which, although it includes the Late Triassic deformation seen elsewhere, began in the Upper or even Middle Palaeozoic. These contrasting collisional ages can be explained in several ways. (a) The Nan Suture may not correlate directly with the Bentong Raub suture and the Sibumasu block may have been broken into northern and southern halves during the Late Palaeozoic; (b) there may have been post-collision lateral translations along these suture zones which have obscured the original collisional relationships; (c) the northern part of the Sibumasu block may have collided with northern Indochina in the Late Palaeozoic but the collision did not go to completion, forming a partially amalgamated superterrane with a small trapped oceanic basin which finally closed in the Late Triassic; or (d) one or more marginal 'back-arc' basins may have formed and then been consumed prior to the final collisional episode in the Late Triassic. The last two hypotheses appear the most likely. Fortunately, almost all researchers agree that the major blocks which comprise the core of SE Asia had assembled and become sutured to South China (Eurasia) by the Late Triassic-Early Jurassic (Hutchison 1989 and references within). Indochina block
90 *
100 °
110 °
120 °
Fig. 1. Regional geography of Southeast Asia. Light circles show location of sites discussed in this study and heavy circles indicate the sites of Yang & Besse (1993).
The Indochina block is bounded to the north by the Red River Fault (Song-Ma and Song-Da suture zone) and to the east by the South China Sea marginal basin (Fig. 2). The southern boundary is poorly documented and it is not clear if the Indochina block has always been continuous with the Borneo block, which has similar Mesozoic igneous sequences, or if there is an E-W trending Late Palaeozoic suture zone near the Natuna and
I
_
_
.
.
_
/s_
~_~-~ga
_
i:::::i,,s ..~
_
-
.
.
.
.
Ratburi
,,',',
"'"'
Fm. g
_
Orouo
L/S -Le~yyin
i', ",",
"-'- Pindaya
_~ ~i~$il Krachan [ ~
',:i.:..:..-Kaen_g :i :i',,,s
::cCuupZg-E ~
.
-.-.---Semanggol ~ -. ~ Fm. -:.:......
.
t:-"--':-~n"
~uZp~n "L:-':-':':I~-Redbeds [~-.'-:-:--Lu [-':'[-'-'iLuang ang [ ~ F_.m .__ (~_F_m...=.
Shan Plateau, Myanmar ~Kala-w r.~-'-q Redbeds
~ ~ , ~ . ~T ~ ~ %" &~ ~
V
~~ng ~ o[~~a--.~ L'L~a~'°~\~" South Jfl I"' ~ _ ~ ~%:":"~..~ ,ndo( f ~t~'!, ,~ ~ C h i n a /~"% .-"P~%, E,st i~ i~
.
~
~es~sph'~n :'"'"'
~ H u a i Hin
[:-::---:lPhong Fro.
"-'-'-'" ~Nam
~
Chert
Khao Luak [' "Y'.'"| Fm.
nSai Fro. ;,','~ ;Saraburi L/S
~a
West Khorat Loei-Petchabun Kampuchea Plateau Fold Belt "--' J . ~ ---- UEvap°rite ~ v~-~Orite K '5:-.'.5'F:Gres J ::::::::::::::::::::: Superieurs t~'~J ~~ ~ _ _
INDOCHINA BLOCK
Diamictite ~ Carbonate~ Marine ~Continental ]~E-] Chert ~ Congl. [-~T] Volcani- ~ Evaporite rTFTT]Stratigraphic (GlacioShale, Redbeds clastics Break Marine?) Mudstone (S/S +Cong.) Fig. 2. Tectonostratigraphicterrane summaryfor SoutheastAsia showingPalaeozoicterrane boundaries and regional stratigraphiccorrelations.Tertiarystrata in peninsular Thailand and the Shan Plateau of Myanmar are only locallypreserved in small Oligo-Mioceneintermontanebasins. Diagrammodified from Metcalfe(1988).
C
.
--'.'."
l)
"
Tr
.
Tembelinq_ ~ . ~ Group
LangkawiPeninsular NW Malaysia Thailand -- -- TI-1I- . . . . . . . . . . . .
SIBUMASU BLOCK
O
206
B. RICHTER & M. FULLER
Mekong basins. The western boundary has traditionally been interpreted to be the NanUttaradit suture zone. This suture is thought to extend southward, beneath the Central Thai Basin, into the Gulf of Thailand and ultimately to the Late Triassic Bentong-Raub line in Malaysia (Metcalfe 1988; Hutchison 1989). In contrast, Barr & Macdonald (1991) conclude that the Nan suture actually marks a Permian suture between the Indochina block and the recently identified Sukhothai terrane. They place the Late Triassic suture zone farther west, along the eastern margin of the Northern Thai granite province. Unfortunately, the Late Palaeogene to Neogene Central Thai basin and the Pattani trough almost completely cover the suture zone from 16°N to 7°N and it is not yet possible to correlate definitively sutures along the length of the Indochina and Sibumasu blocks. Precambrian basement is exposed in the Kontum Massif of Vietnam and presumably forms the basement of the entire Indochina block. It may have originated on Gondwanaland, possibly along the eastern margin of Australia, but must have rifted away by the Early Carboniferous (Metcalfe 1988). Moderately to strongly deformed Palaeozoic sediments are widely distributed but the block is dominated by thick, relatively undeformed Mesozoic continental sediments of the postIndosinian Khorat Basin. Most palaeomagnetic studies in the Khorat Plateau are concentrated in this Mesozoic section.
Previous palaeomagnetic studies Bunopas (1982) examined the palaeomagnetism of a wide variety of rock types in Palaeozoic and Mesozoic strata throughout both the Indochina and Sibumasu terranes using orientated hand samples. His reconnaissance study formed the foundation for additional studies on Neogene basalts in Northern and Central Thailand (Barr et al. 1976; Barr & MacDonald 1979; McCabe et al. 1988); upon Upper Palaeogene to Neogene intermontane basins throughout Thailand (Richter et al. 1993); and upon Upper Palaeozoic and Mesozoic continental strata from the Khorat Plateau (Barr et al. 1978; Achache et al. 1983; Achache & Courtillot 1985; Maranate & Vella 1986; Chen & Courtillot 1989; Yang & Besse 1993). Upper Palaeozoic and Mesozoic o f the Khorat Plateau As noted above, the majority of palaeomagnetic data has been collected from Upper Palaeozoic and Mesozoic sediments of the Khorat Plateau. Many
of these early studies failed to recognize the potential significance of secondary magnetizations and their data are only presented in tectonically corrected coordinates. Yang & Besse (1993) further note that many of the demagnetization procedures may not have been sufficient for discriminating between primary and secondary magnetizations. Despite these concerns, the consistency between these studies is remarkable. Mean values of c. 3040 ° of CW deflection with positive inclinations of 300-40 ° are found in almost all of the Khorat studies. Furthermore, structural dips are low and corrected and uncorrected directions commonly overlap within their respective 0~95 error estimates. The primary concern with these earlier data is distinguishing between primary and secondary magnetizations and then assigning an age to that magnetization. Yang & Besse (1993) present new data from Upper Permian limestones, the Upper Triassic Huai Hin Lat formation, the Lower Jurassic Nam-Phong formation, the Upper Jurassic Sao Khua Formation, and the Lower Cretaceous Khok Kruat Formation. These data are of high enough resolution to allow an approximate chronology of magnetizations and rotations to be reconstructed. Their results are reviewed below and summarized in Fig. 3 and Table 2. The Permian Limestone samples yield a high coercivity CW deflected component which fails a fold test at the 95% level (Fig. 3). Synfolding analysis indicates that kappa is at a maximum at 30% unfolding and thus the magnetization is best interpreted as a synfolding remagnetization. At this locality, Upper Triassic continental sediments unconformably overlie steeply dipping, strongly deformed Permian units. Thus, the Late Triassic Indosinian orogeny is a strong candidate for the folding and initial remagnetization. The Upper Triassic Huai Hin Lat formation also gives a CW deflected direction. Based upon the interpretation of a positive fold test in the overlying Lower Jurassic Nam-Phong Fm they conclude that this is a primary magnetization and they present a tectonically corrected Late Triassic VGP at 52.1 °N, 169.8°E (o~95=7.3) (Table 2). This VGP is indistinguishable from the remagnetized Permian VGP. The Lower Jurassic Nam-Phong formation also yields a CW deflected magnetization. This is found in both positive and reversed polarities and passes a fold test at the 95% level (Table 2).The tectonically corrected VGP for this unit is located at 54.4°N, 175.6°E ((x95 =4.9) and is statistically indistinguishable from the Permian and Late Triassic VGPs. However, Yang & Besse (1993) caution that the folding may have occurred anytime between the mid-late Yanshanian orogeny and the Late Tertiary.
PALAEOMAGNETISM OF THE SIBUMASU BLOCK N
N
90 °
Ju and KI Redbeds Khorat Plateau, Thailand
207
The Aptian-Albian Khok Kruat Formation yields medium and high coercivity components which are quite close to each other (I, D = 37.2 °, 26.0 ° and 35.3 ° , 29.8 ° respectively). Bedding corrections are all less than 12° and the fold test is inconclusive. Despite this, Yang & Besse (1993) conclude that this is a primary Lower Cretaceous VGP. As above, we urge caution in assuming at Early Cretaceous age. The small ~95' the inconclusive fold test, and the lack of reversed polarities leave open the possibility of a younger remagnetization event.
Mesozoic of peninsular Thailand and peninsular Malaysia
N
90°
+ In Situ ++ + + +++++1-90 ° Permian Carbonates Khorat Plateau, Thailand
Fig. 3. Lower hemisphere equal area projections of site mean directions measured in the Khorat Plateau by Yang & Besse (1993). Solid symbols denote positive polarities while open symbols denote negative polarities. Ellipses are the 95% confidence interval about the mean (alpha95) but these have been omitted from the Ju + K1 plot for clarity. See Table 1 for exact values and text for discussion.
We do not follow them in assigning an Early Jurassic age to this pole because the age of the folding is so poorly understood. The interpretation of a positive fold test is reasonable, but without good age constraints on the age of folding we can only say that this is a pre-folding VGP with a maximum age of Early Jurassic. The Upper Jurassic Sao Khua Formation yields a high coercivity CW deflected direction (Table 2 and Fig. 3). This unit is poorly exposed and dips are very low. As such, fold tests are inconclusive. Polarity may be a useful argument in that one would expect to see an even distribution between normal and reversed polarities if the magnetization truly formed in the Late Jurassic. The lack of a conclusive fold test and the preponderance of normal polarities leaves open the possibility that this magnetization was acquired during the Cretaceous long normal period.
The southernmost extent of the CW rotations described above is presently not clear and it is useful to review the available palaeomagnetic data from peninsular Malaysia. The principal work from peninsular Malaysia is by McElhinny et al. (1974). They detected two distinct populations of palaeomagnetic data: (1) the Pengerang rhyolites, Singapore gabbro and dykes, Bentong Group, Singa Formation and Sempah Fm. displayed inclinations of 20-40 ° with CW rotated declinations of 20-33°; and (2) the Kuantan-Massai dykes and the Segamat basalts displayed inclinations of 30-40 ° with counter-clockwise (CCW) rotated declinations of 40 °. Haile & Khoo (1980) detected 20-30 ° of CCW rotation in the Maran, Teka and Kluang redbeds of Malaysia while Schmidtke et al. (1990) confirmed the CCW rotated declinations in mafic dikes at Kuantan and mafic flows at Segamat. McElhinny et al. (1974) argued that either the poles or the field directions of three of their four CW rotated sites passed a fold test. They estimated the age of folding to be Late Triassic and thus their CW rotated magnetization was assigned a preLate Triassic age. In the context of new data and statistical analyses, however, it is not clear that the CW rotation is a pre-folding magnetization and even if it is, it could be as young as Late Cretaceous. There are additional unpublished analyses conducted at the UCSB palaeomagnetism laboratory which bear on the review of Malaysian data. Schmidtke (pers. comm. 1991) measured CCW deflections of similar magnitude (30-45 ° ) in samples collected from the poorly dated Singapore and Lanchang dikes, the Jurassic-Cretaceous Raub redbeds, and the Permo-Triassic Kodiang Limestone. The Bukit Kemaman, Bukit Temiang, Lanchang and Tanjung Penyabong dykes display between 6 ° and 23 ° of CCW declination deflection while sites from the Mesozoic Raub and
9
-
-
34.8
18.5
1634.0 59.9
7.9
157.0 105.0
6.2
4.1 7.5
69.3
43.7 80.9 28.5
379.2 21.3 48.5 13,5
21.2 27.3
54.5 42.0 19.4 51.1 48.8 24.8 11.5 60.8 45.1 58.5 115.7 43.3 53.9 5.7
in situ k
-
-3.3
-3.7 -2.9
-5.2
-4.7 -5.0 -5.0
-13.4 -16.7 -11.2 -4.7
11.7 6.7
26.0 13.8 9.4 8.8 0.3 --44.5 20.9 32.3 5.9 13.4 39.3 9.4 -34.3 11.2
in situ A ° Lat.
-
-10.9 36.8 13.0
44.0
67.9 62.7 -35.5
35.4 59.4 19.8 ~).3
-
35.6 24.4 30.6
142.4 31.8 61.1 67.7
66.4 31.1 35.8 42.7
311.6 21.3
44.7
23.4
9.5 45.7
58.8 28.0 32.5 50.6 43.9 42.4 39.6 44.9 45.8 43.1 50.4 51.1 40.4
Corr. D
17.7 9.4 30.7 30.4 24.8 7.1 37.1 19.2 29.1 25.7 24.5 28.4 15.3
Corr. I
21.8
150.2
180.0
-
-
6.1
-
Corr. ct95
8.6
6.0
1.6
-
-
47.1
-
Corr. k
-
11.7 -14.3 -0.4
-32.7 -25.8 37.9 -7.5
-11.6 -29.3 -0.3 7.2
15.9 -6.4
8.5
11.7 16.1 4.3 4.3 7.6 17.1 0 10.9 5.2 7.2 7.9 5.5 12.9
Corr. A ° Lat.
E171.2 ° E174.5 ° E173.8 ° E173.0 °
N49.2 ° E172.4 °
N56.7 ° E181.3 ° N54.1 ° E 1 8 2 . 8 ° N55.4 ° E182.1 °
° ° ° °
E122.9 ° E138.5 ° E12.4 ° E27.4 °
N60.5 ° E164.3 °
N 5 2 . 6 ° E207.3 ° N 6 2 . 4 ° E156.3 ° N59.6 ° E187.3 °
N13.4 ° N52.9 ° N58.8 ° N67.7 °
N46.7 ° E 1 9 8 . 4 °
N53.1 ° E 1 8 0 . 6 °
N20.2 N42.7 N37.1 N33.4
N24.8 ° E172.0 ° N49.9 ° E137 °
N40.4 ° E354.9 ° N69.5 ° E164.3 °
N46.4 ° E190.6 °
Corr. VGP
N46.3 ° E 1 6 7 . 2 ° N41.8 ° E161.9 °
N47.3 ° E 3 5 7 . 4 ° N58.7 ° E 1 9 3 . 8 °
N50.6 ° E197.0 °
in situ VGP
values a n d italics indicate the p r e f e r r e d direction b a s e d u p o n interpretation o f fold tests.
n o r t h w a r d t r a n s l a t i o n s as p o s i t i v e a n d s o u t h w a r d t r a n s l a t i o n s as n e g a t i v e ; in s i t u , g e o g r a p h i c a l c o o r d i n a t e s ; C o r r . , s t r u c t u r a l l y c o r r e c t e d c o o r d i n a t e s . B o l d t y p e i n d i c a t e s m e a n
I, i n c l i n a t i o n (°); D, D e c l i n a t i o n (°); ¢~95, t h e 9 5 % c o n f i d e n c e i n t e r v a l a b o u t t h e m e a n d i r e c t i o n ; k, t h e p r e c i s i o n p a r a m e t e r k a p p a ; A L a t , p r e s e n t l a t i t u d e m i n u s p a l a e o l a t i t u d e , w i t h
Limestone VGP mean
33.5 36.1
19.2 17.8
14.9
60.8
Silurian Setul Limestone, Langkawi Island, Malaysia SM-88-62 N6.2 ° E99.8 ° 9 SM-88-63 N6.2 ° E99.8 ° 5 SM88-62, 63 N6.2 ° E99.8 ° 14
14.1 10.3 11.5
75.2 50.5 56.7
Permo-Triassic Lampang Limestone, Central Thailand T91-8 N18°15.75 ' E99°36.92 ' 4 40.3 T91-9 N18°15.75 ' E99°36.92 ' 4 40.7 T91-10 N18°15.75" E99°36.92 ' 7 40.7 T 9 1 - 8 , 9, 10 N18°15.75 ' E99°36.92 ' 15 41.0
3.1 11.4 8.8 18.9
43.2 48.3 27.1 37.2
Limestone, Peninsular Thailand N7°55 ' E99°55 ' 7 N10.95 ° E99.28 ° 9 N9.9 ° E99.1 ° 7 N7°36.43 ' E100°2.04 ' 6
14.9 24.1
9.2 11.9 17.8 8.5 11.1 12.4 17.1 11.9 11.5 8.0 7.1 10.3 8.3 19.1
in situ tx95
38.0 46.3 37.6 22.6
Permian Ratburi T90-10 T90-27 T90-45 T91-39
317.3 32.1
in situ D
Late Palaeozoic Plateau Limestone, Shan Plateau, Myanmar B 9 1 - 1 2 + 14 N20°48.89" E96°44.039 ' 6 17.6 B91-13 N20°40.89 ' E96°44.039 ' 3 26.5
N in situ (samples) I
58.5 24.1 27.2 42.9 38.3 352.9 40.1 54.3 40.4 41.7 48.0 44.8 341.1 39.3
Location
Juro-Cretaceous Kalaw redbeds of Shan Plateau (Shan-Thai) B91-8 N20°48.37" E96°26.98 ' 7 -10.2 B91-10 N20°47.98" E96°27.14 ' 5 13.8 B91-11 N20°47.98 ' E96°27.14" 5 21.9 B91-17 N20°38" E96°34 ' 8 22.8 B91-18 N20°38.95 ' E96°33.02" 6 36.6 B91-19 N20°42.24 ' E96°38.52 ' 7 77.0 B91-20 N20°45 ' E96°32" 8 -0.2 B92-54 N20°47.06 ' E96°27.83 ' 4 -22.1 B92-55 N20°45.40 ' E96°28.78 ' 5 27.9 B92-57 N20°41.75 ' E96°30.16 ' 7 14.3 B92-58 N20°42.24 ' E96°38.52 ' 5 -34.0 B92-60 N20°38.27 ' E96°33.43 ' 6 21,6 B92-81 N20°42.31 ' E96°30.55 ' 7 70.7 Mean N20°42 ' E96°33 ' 80 18.6
Site
T a b l e 1. S u m m a r y o f the n e w p a l e o m a g n e t i c d a t a p r e s e n t e d in this p a p e r oo
to
Formation
N in situ (samples) I
.
. 11.9
ct95
D
. 24.1
in situ
in situ
. 42.0
k
in situ
in situ
13.8
A° Lat.
pre-Khorat Group pre-Khorat Group
<) Devonian 0 Devonian
F o r m a t is the s a m e as Table 1.
OD+OP+OTr OD+OP+OTr
Khorat Group Khorat Group
0 Permian 0 Permian
$ Upper Permian $ Upper Permian
3 3
27 25
9 9
102 13
Palaeozoic of Khorat Plateau (Indochina Block)
73 88
56 61
Huai Hin Lat F m Nam-Phong F m Composite Sao Khua F m Khok Kruat Fm
$ $ $ $ $
Upper Triassic Lower Triassic Tru + J1 Upper Jurassic Lower Cretaceous
69 64
0 J + Tr 0 J + Tr
38.2 38.3
46.2 -39.4
33.4 37.9
39.0 39.4
31.7 25.4 117 38.1 37.2
36.2 37.2
29.7 32.3
39.5 183.6
25.5 27.6
24.8 33.0
36.9 39.0 27.9 27.1 26.0
40.8 32.3
3.1 8.1
11.3 12.5
14.4 15,0
12.6 6.0
13.3 15.3 38.2 3.1 2.5
15.9 7.0
22.0 25.0
120.0 99.0
42.0 68.0
17.6 75.4
33.8 14.0 9.8 247.4 388.5
19.0 120.0
-4.5 -4.5
-10.5 -5.3
-1.2 -4.3
-5.4 -5.7
-0.5 3.3 18.8 -4.8 -4.6
-3.1 -3.8
Mesozoic of Khorat Plateau (Indochina Block) 0 Chen & Courtillot (1989) $ Yang & Besse (1993)
Mesozoic of Western and Peninsular Thailand (Shan-Thai) Mesozoic Barr et al. (1978) 19 . Mesozoic Barretal. (1978) 5 13.8
Age
Table 2. Summary of published palaeomagnetic data
40.2 '~ 31.6
-12.1 8.0
31.4 31.7
31.8 39.4
44.4 40.1 41.8 37.3 40.5
50.0 46.9
38.8 32.8
Corr. Inc.
44.4 43.9
62.0 192.6
60.1 54.7
51.4 33.0
39.5 37.2 38.1 26.6 28.1
43.6 36.4
35.6 34.2
Corr. Dec.
4.3 17.1
11.3 31.6
15.4 18.5
12.6 6.0
8.3 6.6 4.7 2.6 2.4
14.1 9.3
12.0 9.6
Corr. c~95
11.0 6.0
120.0 16.0
36.0 45.0
17.6 75.4
85.0 71.6 77.5 339.2 398.5
23.0 70.0
31.7 11.8
Corr. k
-5.9 -0.1
23.1 21.0
0.0 -0.2
-0.6 -5.7
-9.4 -6.2 -7.4 -4.3 -7.0
-13.8 -11.1
-4.3 -0.3
Corr. A ° Lat.
N61.6 ° E179.1 ° N59.2 ° E179.3 °
N52.1 ° E169.8 ° N83.7 ° E134.9 °
N65.7 ° E186.2 ° N63.6 ° E179.3 °
N66.0 ° E174.9 ° N48.5 ° E176.15 °
N54.7 ° E185.5 ° N52.2 ° E191.4 ° N53.2 ° E189.2 ° not given not given
N51.3 ° E181.9 ° N59.3 ° E180.8 °
-
VGP
in situ
N48.0 ° E177.4 ° N48.1 ° E186.1 °
N24.5 ° E208.2 ° N65.6 ° E251.3 °
N32.8 ° E183.4 ° N37.9 ° E184.2 °
N58.45 ° E176.15 °
N52.1 ° E169.8 ° N54.4 ° E175.6 ° N53.6 ° E173.3 ° N64.8 ° E175.6 ° N62.7 ° E173.3 °
N48.1 ° E165.6 ° N54.8 ° E168.1 °
N56.3 ° E!76.6 ° N57.5° E184.0 °
Corr. VGP
tO
©
~Z
E ©
Z
> © E >
210
B. RICHTER & M. FULLER
Tembeling redbeds reveal a broad spectrum of directions between the CW and CCW rotated end members identified by McElhinny et al. (1974). These data have never been published because it has been difficult to evaluate the age of the magnetization and the influence of structural deformation and present field chemical weathering upon the magnetic signal. These unpublished data are presently being reevaluated. To summarize, the domain of CW rotations observed in Mesozoic rocks of the Shan and Khorat Plateaus only extends as far south as Langkawi Island, Malaysia. South of this the Mesozoic section shows a wide spectrum of CCW and CW rotated directions. The CW end members are very similar to the Shan/Khorat directions while the CCW end members are very similar to directions found in Borneo (Schmidtke et al. 1990). Thus, it appears that peninsular Malaysia is a transition zone between a CW rotating Indochina-Sibumasu block and a CCW rotating Borneo block. Cenozoic of Thailand
McCabe et al. (1988, 1993), building upon earlier work by Barr et al. (1976) and Barr & Macdonald (1979), detected small CW rotations in Neogene basalts from northern and western Thailand but detected no rotation of Neogene basalts located on the Khorat Plateau. These basalts are part of a tholeiitic to alkalic suite of Late Cenozoic basalts that are exposed throughout the Indochina block (Barr & Macdonald 1981). They conclude that volcanic activity began by at least 12 Ma and that there is no systematic relationship between the age and geographical distribution of volcanism. McCabe et al. (1988) initially interpreted the data as a Neogene CW rotation of western Thailand relative to both Eurasia and the Khorat Plateau while McCabe et al. (1993) reinterpreted the data as local block rotations driven by intracontinental deformation. Richter et al. (1993) presented reconnaissance data from several Neogene intermontane basins in Thailand. The Krabi and Nong Ya Pong basins of peninsular Thailand show small CW declinations comparable to those measured in Neogene basalts. In contrast, the reconnaissance data from the Mae Moh basin of north-central Thailand shows predominantly small CCW declinations. Although it is possible that a 15 ° rigid body CW rotation of peninsular Thailand took place during the Middle to Late Neogene it is very unlikely that a 20 ° CCW rotation of north-central Thailand took place concurrently. Thus, local rotations are a strong possibility. In contrast to McCabe et al. (1993), however, we conclude that the dispersion in Neogene palaeo-
magnetic data does not demonstrate that chaotic Neogene local rotations have obscured all Mesozoic motions. There is demonstrable consistency in Mesozoic data collected over a large region while all Neogene data have been collected from small discontinuous exposures which, in many cases, are structurally bounded. Thus, it appears that regional block motions are preserved in the Khorat and Shan Plateaus while data collected from the fold belts which bound these plateaus and from the Central Thai basin are probably complicated by Neogene strike-slip faulting.
Geology of palaeomagnetic sampling sites A complete review of the regional stratigraphy of the Sibumasu and Indochina blocks is beyond the scope of this paper but regional stratigraphic sections are presented in Fig. 2. The reader is referred to Hutchison (1989) for regional summaries; to Bunopas (1982), Snansieng et al. (1983) and Nutalaya et al. (1983) for Thailand; and Chibber (1934) and Bender (1983) for Burma for a more detailed stratigraphic discussion. M e s o z o i c continental s e d i m e n t s o f Shan Plateau, M y a n m a r
The Kalaw redbeds are an extensive sequence of continental sediments which are exposed along the western margin of the Shan Plateau east of the Pan Laung and Shan Boundary Faults (Fig. 4). They lie near the northern end of a linear belt of Upper Mesozoic continental deposits which include the Raub and Tembling redbeds of peninsular Malaysia and the Chumpon and Kanchanaburi redbeds of Thailand (Metcalfe 1988; Bender 1983). Mitchell (1992) indicates that they are coeval with the lower portion of the Khorat Plateau redbed sequence. Diagnostic fossils within the Kalaw sequence itself are lacking and the age is poorly constrained (Win Swe, pers. comm. 1992). The redbeds unconformably overlie the Lower-Middle Jurassic Loi-An (Pan Laung) Series and are unconformably overlain by Neogene to Quaternary non-marine sediments (Bender 1983). Most authors agree that the sequence ranges from Middle or Upper Jurassic to Cretaceous (Bender 1983; Wolfart et al. 1984; Myanmar Oil and Gas Enterprise, unpublished geological map) but it must be emphasized that geological mapping and dating within the Shan Plateau are still at a reconnaissance stage. The Kalaw sequence is moderately to strongly folded about sub-horizontal NNW-SSE trending axes. Folding is commonly thought to have taken place during the Middle to Upper Cretaceous
PALAEOMAGNETISM OF THE SIBUMASU BLOCK
95°E
21ON
--
15'
30'
Mt. Popa Chauk ~ ~ tO
-
_
45'
96°E
15' 20ON
_
_
_
_ _
Meiktilla
~ha
Taunggyi
Q Kalaw
Exolanation
"~ ~k
Q
QuaternaryCover
~/\
~
NeogeneBasalt
"~/~
~
Cretaceousto Lower Paleogene Intrusives
Pz
l/ ~,
Pz
Pz
Pt
)lnle
Jurassic-Cretaceous Kalaw Redbeds 45'
97OE
45'
. _ _ _ _.. , , o
a
_
30'
/ I \
45' 30'
15'
211
PaleozoicCarbonates Mogok Series Gneiss, Marble, Schist~
30'
\
Chaung Magyi Group
fJ
15'
Faults (air photo lineaments) / Roads
19ON
0
20
40
60
80
100 km
Fig. 4. Geological sketch map of the western edge of the Shan Plateau at the latitude of the Kalaw redbeds (modified from Bender 1983). Samples were collected at c. 400 m (0.25 mile) intervals along the Thazi-Kalaw road (dashed line). The dextral Sagaing fault is just off the western boundary of the map. (Wolfartet al. 1984; Win Swe, pers. comm. 1992) but Palaeogene sediments are absent at Kalaw and the age of the folding cannot be directly bracketed by the age of undeformed overlying sediments. Yang & Besse (1993) similarly caution that the folding of Upper Jurassic-Lower Cretaceous units in the Khorat Plateau may have occurred anytime between the mid-late Yanshanian orogeny and the Late Tertiary. There are several pieces of data which point towards a post-Cretaceous regional uplift and folding event. First, MacDonald et al. (1993) show that the Doi Inthanon metamorphic core complex in Northwest Thailand was formed and then uplifted between the Late Cretaceous (72 _ 1 Ma) and the Latest Oligocene epochs (25 _ 4 Ma). Second, Rangin (pers. comm. 1992) reports that Mesozoic continental sediments of the Shan Plateau are thrust over Eocene sediments of the Central Basin of Thailand. Third, thrust-cored anticlines of the
Chauk Oil Field in the forearc basin of Myanmar first formed during the Oligocene, as did the E-W trending cross basin highs which separate the forearc and back-arc basins into smaller sub-basins (Bender 1983). Fourth, a significant Oligocene unconformity is present throughout northern Myanmar (Bender 1983). Finally, the intermontane basins of Thailand as well as the Pattani trough began subsiding rapidly in the Late Oligocene (Polachan et al. 1991). Thus, there are several regional geological events which point to a Tertiary uplift and folding event.
Mesozoic and Palaeozoic carbonates of Thailand Upper Palaeozoic and Lower Mesozoic limestones were sampled throughout peninsular Malaysia, Thailand and Myanmar (Figs 1 and 2). These
212
B. RICHTER & M. FULLER
include the Ordovician to Silurian Setul Limestone exposed on Langkawi Island (Jones 1978), the Upper Palaeozoic Plateau Limestone of the Shan region of Myanmar (Bender 1983), the Lower to Middle Permian Ratburi Limestone exposed throughout peninsular Thailand (Nutalaya et al. 1983; Baird & Bosence 1993), and the Triassic Lampang Group of central Thailand (Nutalaya et al. 1983). The Palaeozoic units represent subtropical to tropical shallow marine shelf carbonates which dominated the Sibumasu block prior to the Indosinian orogeny while the more limited occurrences of Lower Mesozoic rocks record localized shallow marine deposition (Hutchison 1989). These limestones have almost certainly undergone the Late Mesozoic to Palaeogene deformation seen in the Kalaw redbeds but may also have experienced an older Late Palaeozoic or Late Triassic deformation (Jones 1978). Baird & Bosence (1993) conclude that the Ratburi Limestone in northern peninsular Thailand has undergone at least three episodes of diagenetic alteration. They argue that the high temperature isotopic signatures which characterize the final diagenetic event is likely the result of heating caused by the intrusion of Cretaceous granites. Silicification, tensional calcite veins and karstification characterize the Palaeogene evolution of the unit. Thus, the folding recorded in these carbonates could have taken place in either the Late Triassic-Early Jurassic Indosinian orogeny or subsequently during Late Mesozoic-Palaeogene subduction.
Palaeomagnetic sampling and analyses Sample locations are summarized in Table 1. Between six and ten individually orientated cores were drilled at each site using a portable gasoline powered diamond coring drill. These were orientated with a Pomeroy orientating platform and a Brunton compass and no correction for secular variation was required. None of the samples required the use of a sun compass. If exposure permitted, additional sites were collected several tens of metres up or down section. Each site collected during the 1991 and 1992 field seasons was located with a hand held Magellan GPS Navl000 Pro Global Positioning System unit and locations are accurate to _+30 m. Sites collected during the 1990 field season were located with local base maps and location errors can be as large as +_300 m.
Laboratory analyses Palaeomagnetic analyses were conducted at the University of California, Santa Barbara using either a 2-G Enterprises 3-axis cryogenic magnetometer
or a Molyneux Molspin spinner magnetometer. Three methods of demagnetization were employed depending upon the magnetic behaviour of individual samples: (a) alternating field de magnetization (AFD), (b) continuous thermal demagnetization (CTD) (Dunn & Fuller 1984), and (c) stepwise thermal demagnetization.
Statistical procedures Orthogonal vector plots (after Zijderveld 1967) were constructed for each sample in order to characterize its demagnetization behaviour. Principal component analysis was conducted using computer software written by S. L Gillette (© 1985) which utilizes the least squares regression technique of Kirschvink (1980). In addition, remagnetization circles were fitted to samples displaying two component behaviour using the Gillette software and directions were calculated after the method of Halls (1978). In all cases, site mean directions, confidence intervals (t~95), and precision parameters (k), were calculated after the method of Fisher (1953). Folded strata were analyzed using the methods of McElhinny (1964) and McFadden (1990).
New palaeomagnetic data Kalaw redbeds, Myanmar Palaeomagnetic samples were collected from 14 sites located along the Meiktila-Kalaw highway (Fig. 4). The traverse crosses several parallel folds and bedding attitudes vary considerably. Thus, it was possible to conduct a detailed fold test. Samples were collected primarily from medium to fine grained red sandstones which vary from weakly friable to well cemented. Hematite cements are present in all samples while the strongly cemented samples also contain as much as 30- 40% carbonate cement. In the latter case, hematite dust rims completely encircle the detrital quartz grains and appear to predate carbonate cementation. Conglomerates within the sequence were also sampled in the hope of obtaining conglomerate tests but the clasts are dominantly composed of very weakly magnetized limestones which exhibited variable behaviour during demagnetization. Orthogonal vector plots from demagnetization experiments are presented in Fig. 5. The magnetic behaviour of the medium to fine grained units within the Kalaw redbeds is excellent and in almost all cases, independent of grain size or degree of cementation, samples display single component behaviour above 150-250 ° C. The low temperature component is less rotated than the stable endpoint
PALAEOMAGNETISM OF THE SIBUMASUBLOCK
Thermal Demag B92-58-4C
W Up S [~L%I~36dC, , , ~\'%. i \'%458oc I \- 77oc
Thermal Demag B92-57-4B
AF Demag B92-57-4B
N
EDn] I i ~ ) 9 5 " _ [C~24°C
S
213
W Up
WUp :
:
:
:
:
N
19 0 0 Oe,, 1300 Oe
s I
-
: : 4 °C
~! "K"
:
:
N
oc
-,%
E Dn
/
...-.'1 K "~50 Oe
I.O'~,~ ~ , ~ 00e
0n] 1.o . \
",.2.90 oC
1900Oe |
500oC
Fig. 5. Characteristic orthogonal demagnetizationplots for samples collected from the Kalaw Redbeds, Myanmar. Closed symbols represent declination data plotted in the horizontal plane while open symbols represent the projection of inclination data upon the vertical N-S plane. Inset plots show the normalized intensity decay throughout a demagnetization experiment. Demagnetization steps are in degrees celsius or Oersteds and initial intensities are in Gcm3/g. Magnetic behaviour is excellent and after a soft present field component is removed the magnetization decays univectorially to the origin. Samples which did not display this behaviour were not incorporated into the site mean values. and most likely reflects recent chemical weathering in the present field direction. The remaining magnetization is extremely stable to 550°C above which intensity begins to drop off. In all samples a stable direction remains until temperatures of 640°C at which point magnetic behavior becomes erratic. These intensities have not been corrected for measurement at high temperature but the intensity drop at 580°C indicates that a fraction of the magnetic signal resided in magnetite. Thus, a combination of AF and thermal demagnetization was employed to determine qualitatively the relative importance of magnetite and hematite carriers. Sample B92-57-4b (Fig. 5) was first subjected to AF demagnetization in peak fields of 200 mT (2000 Oe) and was subsequently thermally demagnetized to 630°C. Approximately 60% of original intensity was still present after treatment in 200 mT fields, indicating that the majority of the signal resides in a mineral which is very resistant to AF demagnetization. In order to eliminate the possibility of an AF resistant pyrrhotite component, samples from sites B92-54, 55, 57, 58, 60 and 81 were thermally demagnetized in stepwise fashion to 400°C and then AF demagnetized in peak fields of 200 mT. The behaviour of these samples indicates that goethite or pyrrhotite, which are very resistant to AF demagnetization but not to thermal demagnetization, are not present. Hematite, therefore, is the primary magnetic carrier with a small portion of the signal residing in magnetite. There is no change in direction when the samples are AF or thermally demagnetized and thus both magnetite and hematite are interpreted to carry the same magnetization.
Thirteen of the fourteen sites show CW declinations of c. 40 ° with shallow positive inclinations (Fig. 6 and Table 1). The in situ mean inclination and declination are 18.6 ° and 39.3 ° respectively while the structurally corrected values are 23.4 ° and 44.7 °. This gives a corrected VGP at 46.4°N 190.6°E. After application of individual bedding corrections the o~95 decreases from 18.0 to 6.9 while kappa increases from 5.8 to 34.0. Statistical analysis of the significance of this fold test indicates that it is positive at the 99% level using the test of McElhinny (1964) and is positive at the 95% level using the SCOS1 procedure of McFadden (1990). In order to see if the Kalaw redbeds have a magnetic fabric which might have affected the remanence direction, we measured the anisotropy of magnetic susceptibility (AMS) for 16 representative samples. These data show a weakly anisotropic magnetic fabric with a strong clustering of maximum directions parallel to the regional fold axis and a girdle of minimum and intermediate directions perpendicular to the maximum direction (Fig. 7). The remanence direction is located approximately 80-90 ° away from the maximum anisotropy and does not appear to be controlled by it. All of the available data, therefore, indicate that the Kalaw redbeds possess a pre-folding magnetization which has not been altered by subsequent deformation.
Palaeozoic and Mesozoic carbonates A variety of Palaeozoic and Lower Mesozoic limestones also display c. 40 ° of CW deflection.
214
B. RICHTER & M. FULLER N
Kalaw R Shan P l a t e a u , Myanmar /
/
+ +
~
+
+ ~ ~ \ + ~ ~ ~ + + Corrected / ' + + + +++++Jr--~'-~r / Cooroinales y+-
90 °
270°--~-++++++++'~++ +++++++~90 °
Fig. 6. Positive fold test at the 99% level in the Jurassic-Cretaceous Kalaw redbeds of the Shan Plateau of Myanmar. The in situ and structurally corrected directions are 18.6°, 39.3° and 23.4 °, 44.7 °, respectively. Solid symbols denote positive polarities while open symbols denote negative polarities. Ellipses are the 95% confidence interval about the mean (alpha95). See Table 1 for individual site mean directions and fold test statistics.
Orthogonal vector plots for the Ratburi Limestones are presented in Fig. 8 and directional data are presented in Fig. 9 and Table 1. These samples display three types of demagnetization behaviour: (a) univectorial decay to the origin, (b) a strong low temperature overprint followed by univectorial decay of a high temperature component, and (c) multicomponent behaviour in which a high temperature signal cannot be isolated. In most cases the intensity of the signal is swamped by instrument noise before the Curie point is reached.
Nonh
270*
"
90*
AMSMinimum I+
~//
,, ,,MS,,,a=um {;
•
Axis 180 °
Fig. 7. Anisotropy of magnetic susceptibility (AMS) data for the Kalaw redbeds. See Fig. 3 for an explanation of symbols. The magnetic fabric is weakly developed with the maximum susceptibility clustered around the average fold axis for the region. However, the remanence direction does not appear to be controlled by it (see Fig. 6).
All three types of behaviour could be found at the same site in samples collected less than 1 m apart. This can be partially attributed to variable mineralogy (magnetite v. pyrrhotite), variable amounts of veining and recrystallization, and extremely weak initial magnetizations. Samples displaying multicomponent magnetizations have not been used in site mean calculations but there is still substantial dispersion in the mean directions. These sites are distributed between latitudes 6°N to 20°N (Fig. 9) and show a CW direction throughout much of the peninsular Thailand. Fold tests in the Setul Limestone of Langkawi Island, Malaysia, and the Lampang Limestone of north-central Thailand are negative (Table 1 and Fig. 10) and indicate that the magnetism was acquired after the youngest episode of folding (Fig. 11). Although the errors are large it is clear that the in situ poles cluster in a single region while the corrected poles are dispersed. As in the prior two single fold tests, this regional attitude test on the VGP'S indicates that the magnetization was acquired after folding.
Age of magnetizations The magnetization measured in the Palaeozoic and Mesozoic limestones is a secondary chemical remanence which is carried by pyrrhotite and low coercivity magnetite. There are two tectonic events which could have driven this remagnetization: the late Triassic Indosinian orogeny, and a combination of Late Cretaceous magmatic activity and the Late Cretaceous to Palaeogene? uplift event discussed above. The lack of a Mesozoic age progression of magnetizations makes it difficult to discriminate between these two possibilities and
PALAEOMAGNETISM OF THE SIBUMASU BLOCK
215
Thermal Demag T90-45-2 S
W Up : : '~78°C
:
:
:
N
- - ~ 85°C '~,
121 °C
Dn
C
Thermal Demag T90-27-6 W Up
t
o.ok
E Dn
500°C
T90-27-9Thermal Demag w up . . . .
-.
,. i ,
/f
_
~%
tu
23°C .~ o . - ~ - . -
( 10
.~ -'.~ 7 ~
282o(~ / .~ E Dn
500°C Fig. 8. Orthogonal demagnetization plots for continuous thermal demagnetization experiments conducted on samples from the Ratburi Limestones. These samples display three types of demagnetization behaviour: (a) Univectorial decay to the origin; (b) a strong low temperature overprint followed by univectorial decay of a high temperature component; and (c) multicomponent behaviour in which a high temperature signal can not be isolated. Samples displaying Type (a) or Type (b) behaviour were used to calculate site mean values. See Figs 3 and 5 for an explanation of symbols.
Fig. 9. In situ palaeomagnetic directions of Palaeozoic and Mesozoic limestones from Thailand and Malaysia. These show CW declination deflections similar to that seen in the Kalaw redbeds. These directions are not disturbed by the NE-SW trending Ranong strike-slip fault zone (RFZ). This indicates that at least some fault blocks in peninsular Thailand have not rotated with respect to each other. However, these data cannot distinguish between a plate wide rotation or a systematic 'domino' style rotation.
the errors overlap all of the VGPs reported from the Mesozoic of the Khorat Plateau. Thus, we can only confidently say that the pole measured in these limestones is no older than L a t e Triassic and probably no younger than Late Cretaceous. In view of the fact that the mean VGP coincides with the Late Triassic pole (Fig. 11) calculated for the Khorat Plateau by Yang & Besse (1993), a Late Triassic remagnetization is more likely.
216
B. RICHTER & M. FULLER
N N
90°
(a)
N
N +
90° +
Corrected/
+~~+ + ++ +7
90°
(b) Fig. 10. Negative fold tests in (a) Silurian Setul Limestones, Langkawi Island, Malaysia; and (b) the Early Mesozoic Lampang Limestone, North Central Thailand. See Table 1 for numerical data and Fig. 3 for an explanation of symbols.
Conversely, the age of the magnetization found in the Kalaw redbeds is clearly pre-folding. Thus the magnetization could be a primary depositional/ diagenetic magnetization which is as old as the stratigraphic age (Late Jurassic to Cretaceous) or it could be a much younger pre-folding chemical magnetization which is tied to the Late Cretaceous to Palaeogene uplift event described elsewhere in this paper. In order to refine the age of the Kalaw pole we have prepared tectonic reconstructions of South China, Indochina, the Shan Plateau, and the Lhasa and Qiangtang blocks at various Mesozoic time slices relative to the GAD and examined the reconstructed position of the Shah Plateau (Fig. 12). The reconstructions were prepared on an Evans and Sutherland workstation courtesy of G. Ramsayer and I. Norton of Exxon Exploration Company. These were made by directly restoring the palaeolatitude using the mean VGP position and then by empirically rotating the terrane along lines of palaeolatitude until space problems were minimized and 300--400 km of sinistral offset was present along the Red River fault (Table 3). A small range of solutions is valid as long as the 95% confidence ellipse for the VGP encloses the GAD axis. The positions of the South China and North China blocks are taken largely from Enkin et al. (1992) but have been modified by Ian Norton (pers. comm. 1994). The Lhasa and Qiangtang blocks are essentially held fixed to South China but compressional deformation across these blocks has been restored by using plate circuits to reconstruct the position of India at the time of collision and
r-----s ~______~: :COn-6dti~d: : i
/
"
i
:i:
J
180 °
Fig. 11. Virtual geomagnetic poles for the Ratburi, Triassic and Setul limestones of Thailand and Malaysia. This plot indicates a negative regional fold test but the errors are large. The corrected Khorat Tu, Ju, and K1 poles have been reproduced from Yang & Besse (1993). The mean in situ value for this study (open circle) coincides with the Khorat Tu pole. This suggests a Late Triassic remagnetization age for the carbonates measured during this study.
217
PALAEOMAGNETISM OF THE SIBUMASU BLOCK Table 3. Summary of the rotation poles, relative to the GAD, used to reconstruct the Shan and Khorat Plateaus against the reconstructed position of South China block (see Fig. 12)
Time slice (Ma)
Shan Plateau (Shan-Thai Block) Lat.
60 80 120 140 160 180
-2.0 -4.1 -13.1 -18.8 -30.7 -29.2
Long. Angle 99.6 100.5 95.6 93.1 86.9 88.3
43.5 43.8 44.7 46.0 50.6 50.4
Khorat Plateau (Indochina Block) Lat.
Long.
Angle
-17.4 -26.7 -35.4 -40.4 -51.9 -51.6
83.6 80.9 78.8 75.4 70.0 69.3
29.3 31.3 34.2 36.8 45.2 46.0
then by retrodeforming Lhasa and Qiangtang out to the reconstructed northern edge of India. This analysis shows that the Jurassic and Early Cretaceous reconstructions all have serious over-
Restored Khorat Plateau JI-Ke Pole Restored
Shan Plateau
laps or underlaps which are inconsistent with the geological data. The Late Cretaceous and Late Palaeocene reconstructions show the smallest space problems and also show the correct amount of displacement along the Red River Fault. This analysis suggests that the magnetization measured in the Kalaw redbeds is most likely to be Late Cretaceous to Palaeocene? in age.
Discussion of palaeomagnetic data Regional rigid block rotations v. chaotic local rotations As d i s c u s s e d above, U p p e r P a l a e o g e n e and Neogene palaeomagnetic data show a wide variety of CW, unrotated, and occasionally C C W rotated directions. McCabe et al. (1993) argue that the apparently chaotic N e o g e n e declinations in Thailand record local block rotations and that older terrane motions have been obscured. We present an alternative hypothesis. The Neogene rotations
Shan J-K pole I if held fixed to I Indochina _1 "t v~t~
x'--J I 00o
~ . ~ 120 ° E
120 Ma / Early Cretaceous
60 ° N 120 ° E
~60
Ma Late 1I Paleocene |
Fig. 12. Plate reconstructions of South China (Eurasia) relative to the geocentric axial dipole (GAD) at various time slices. For each time slice the VGPs for the Khorat and Shan Plateaus have been restored to the dipole axis and the respective terranes then restored along lines of palaeolatitude until space problems were minimized. The white Shan Plateau represents its position if held fixed to Indochina and then rotated CW relative to South China. The filled Shan Plateau is the position obtained by restoring the Shan VGP directly to the GAD. The white lines are the coastlines of South China and Indochina. Sinistral displacement along the Red River fault can be estimated by comparing these coastlines. See Table 3 for rotation poles and the text for additional discussion.
218
B. RICHTER & M. FULLER
appear to be limited to extensional and transtensional zones in the Central Thai province where strike-slip faults are well documented (Polachan et al.1991). The Central Thai province is bounded on either side by moderately to highly deformed fold belts and then by the relatively undeformed Shan and Khorat plateaus. Although faults have been mapped along the margins of these plateaus, the interiors are remarkably undeformed (Bender 1983; Hutchison 1989). Thus, it appears that these large crustal blocks have only deformed along their margins. Data collected from the marginal fold belts or the transtensional zones between these blocks probably record local rotations driven by transtension and strike-slip motion but data collected from the interior of these blocks probably record large-scale Late Mesozoic and Tertiary terrane motions. Lamb's (1994) discussion of the behaviour of the brittle crust in wide plate boundary zones provides a theoretical basis for this hypothesis. He proposes a simple two layer model of the lithosphere in which the brittle upper crust sits upon ductile lower crust (Fig. 13). The brittle crust consists of rigid blocks of varying sizes which can only deform by rotating and translating. Thus, the fine scale velocity field will be discontinuous with abrupt velocity changes across faults. When viewed at a large scale, these velocity variations are averaged out and match the velocity field of the underlying ductilely deforming layer. Thus, an overall CW rotation and southeastward translation in the lower crust could be recorded in the upper crust as a mixture of rotations. Sundaland is almost certainly broken into smaller crustal blocks than
have previously been recognized but there is ample evidence which indicates that terrane motions can still be measured.
Rotations The VGPs for the Khorat and Shan Plateaus and the APW path for Eurasia-South China (Enkin et al. 1992) are shown in Fig. 14. The mean Kalaw VGP is rotated 47.6 ° CW relative to the GAD, 30.4 ° CW relative to the Upper Cretaceous (100 Ma) Eurasia-South China reference pole, and 15.4 ° CW relative to the Khorat Lower Cretaceous VGP presented by Yang & Besse (1993). A similar CW rotation can be calculated relative to other Cretaceous or Palaeogene segments of the EurasiaSouth China APW path. Thus, we conclude that the Shan Plateau VGP is rotated c. 25-30 ° relative to South China and c. 10-15 ° relative to the Khorat Plateau (Indochina Block). Revisions in the ages or error estimates of the Shan and Khorat VGPs are unlikely to alter this basic conclusion. The palaeomagnetic data do not uniquely indicate when this rotation took place but the geological evidence indicates that it occurred during the Late Cretaceous and Palaeogene.
Latitude shifts Palaeolatitudes have been calculated for all of the data presented in Tables 1 and 2. These data show a clear pattern with the Khorat Plateau travelling slightly southward while the Shan Plateau has travelled northward by 8.5 ° +/- 3 ° relative to the
Rotating and
Fine scale
Large scale
translating
velocity field
velocity field
blocks
/.,,-J.,,,,..,,
I"'-" IP t
J'
"1-
•
1.
,,;.".,.,
""
""
,,
"
,,,'//// ,r.......Koo°,°°o..,,o.
f; :.:r Fig. 13. Velocity field model presented by Lamb (1994) in an attempt to explain brittle deformation in wide plate boundary zones. This model for continental deformation may be able to explain the variety of palaeomagnetic rotations observed in SE Asia. See text for discussion.
PALAEOMAGNETISM OF THE SIBUMASUBLOCK
.ur
219
J,
rat~ ~Khora(Tu(corr.)~l "-~Xo,.2_ ~/
9/
180*
Fig. 14. Composite VGP-APW plot which compares the motions of the South China block with the VGPs for the Indochina (Khorat Plateau), and Sibumasu (Shan Plateau) blocks. The APW path for South China is reproduced from Enkin et al. (1992) and the Khorat VGP positions are reproduced from Yang & Besse (1993). Despite the ambiguity in the age of the Shan Plateau pole (K1 to Tpal?) it is clearly rotated CW relative to the equivalent range of Khorat Plateau and South China positions. Likewise, it is clearly displaced northward. The exact amount of rotation and northward translation is dependent upon the inferred age of the pole.
GAD over the same time period. This corresponds to c. 12° northward motion relative to the Khorat Plateau Lower Cretaceous VGP and c. 5 ° relative to the 100 Ma Eurasia-South China VGE AMS measurements (Fig. 7) show that a strong bedding plane fabric is not present and thus shallow inclinations do not seem to be related to a depositional alignment of platy hematite. The fold axis is almost horizontal and thus shallow inclinations cannot be attributed to an overlooked plunge correction. These observations indicate that the small inclinations do not result from a problem with the palaeomagnetic signal. However, there are some potential structural complications. The Kalaw redbeds are exposed along the very western edge of the Shan Plateau within a zone of faulting that is bounded to the west by the N-S trending Sagaing-Shan Boundary-Pan Laung Fault system and bounded to the east by unnamed NW-SE trending faults (Fig. 4). It is possible that this triangular sliver is partially detached from most of the Shan Plateau. This leaves open the alternative hypotheses that a separate 'Kalaw block' has been tilted or has been translated northward relative to the Shan Plateau. We have collected extensively from Silurian, Devonian and Permian carbonates exposed near Taunggyi (Fig. 4), well away from the Pan Laung Fault. The inclinations from these samples should indicate if there is a separate Kalaw block which has moved northward relative to the rest of the Shah Plateau. Figure 12 shows the positions of the Shah Plateau, relative to the Khorat Plateau, assuming end member models of no relative northward motion and the total palaeomagnetically determined northward motion. As a cautionary note, the
error estimates on the palaeomagnetic data do not take into account external geological sources of error such as poor age control, limited outcrop, the uncertainty in recognizing remagnetizations and other factors. Thus the real errors are likely to be larger than shown. A more realistic reconstruction would restore the Shan Plateau to a position between the end members shown here. There is little doubt that oblique convergence has driven the Burma block almost 500 km northward during the Neogene (Curray et al. 1979). Furthermore, the Indian Ocean plate has been obliquely subducting beneath the western margin of Sundaland since the Late Cretaceous. Depending upon the timing of the CW rotation of Sundaland, this oblique subduction could have driven northward translation of all or part of the Sibumasu block during the Eocene and Oligocene epochs. What is presently unresolved is the size of these blocks and how displacement is distributed. Are the blocks just small slivers within the western edge of the Mergui shelf and Shan Plateau? Are they larger blocks bounded by faults such as the Three Pagodas and Wang Chao (Mae Ping) Faults? Is the entire 'rigid' Sibumasu block moving? Is displacement centralized on a master fault, similar to the present Sagaing fault, or is it distributed as small displacements along many faults? Additional research is needed to choose between these possibilities.
Review of tectonic models Extrusion tectonics
The extrusion tectonic model has evolved from the initial slip-line field theory experiments of Molnar
220
B. RICHTER & M. FULLER
& Tapponnier (1975) and Tapponnier & Molnar (1976) to scaled down mechanical experiments with plasticine (Tapponnier et al. 1982, 1986; see also Jolivet et al. 1990). Peltzer & Tapponnier (1988) added significantly to earlier studies in that they examined the impact of distance between indentor and the margin of the block as well as the impact of vertical and horizontal inhomogeneities upon extrusion. The type, location and timing of structures which developed during the plasticine extrusion experiments are remarkably similar to those mapped in SE Asia. These observations led Tapponnier et al. (1986) to conclude that the Eocene collision and subsequent penetration of India into Eurasia was accommodated primarily by more than 1000 km of eastward expulsion and c. 20 ° of CW rotation of Sundaland along the Red River fault during the Oligocene and Early Miocene. Beginning in the Middle Miocene the deformation front shifted northward and the South China block was likewise extruded eastward and rotated 20 ° CW between the Red River and Altyn Tagh faults. Thus, Sundaland is thought to have undergone a rigid body CW rotation of 40 ° since the Late Palaeogene (Fig. 15). A more detailed review of the extrusion model is presented in Richter (1995).
Time 2
Time 3
(a)
(b)
D e x t r a l m e g a s h e a r tectonics The dextral megashear model was elegantly summarized by England & Molnar (1990) who stated that 'the image of lateral transport on (east-west trending) faults, known also as continental escape, extrusion, or expulsion, is an illusion, and that instead the left-lateral slip on east-striking planes in eastern Tibet is a manifestation of north-striking right-lateral shear'. As this dextral shear evolves, the sinistral faults and the N W - S E trending blocks which they bound must rotate clockwise around local axes (Fig. 15). An important consequence is internal deformation which allows E - W shortening and N - S lengthening. Furthermore, Dewey et al. (1989) conclude that the India-Eurasia collision has been accommodated primarily by lithospheric thickening of Eurasia in a widening east-west trending zone which is bounded on the west by a sinistral megashear extending from Makran to Baikal and to the east by a dextral megashear which extends from Sumatra to the Tanlu Fault system. The tectonic evolution of eastern Tibet, Myanmar and Indochina would, therefore, be dominated by northsouth striking dextral shear and associated CW rotations around local vertical axes with little to no southeastward extrusion.
Extension in Central Thai Basins? ~
(c) Fig. 15. Schematic comparison of (a) the extrusion and (b) megashear model at the beginning, middle and end of motion. Compare this with (e) which attempts to integrate internal deformation into the extrusion model. In this model, Sundaland consists of large fault bounded blocks which rotate CW as a consequence of extrusion. But, as extrusion progresses, the blocks closest to the collision zone are displaced farther northward along N and NW-SE trending strike-slip faults (e.g. the Sagaing, Uttaradit or Three Pagodas faults). As motion continues, antithetic strike-slip faults will likely form and local zones of extension and compression will be created. Blocks caught within sinistral zones may rotate CCW while those caught in dextral zones may rotate CW. As the westernmost block (Shan Plateau) moves northward it may rotate CW, driving rifting in the Central Thai basins.
PALAEOMAGNETISM OF THE SIBUMASU BLOCK Polachan et al. (1991) have proposed a tectonic model for the evolution of the Tertiary basins of Thailand which is based upon extrapolation of the dextral simple shear strain ellipsoid presented by Wilcox et al. (1973) to the regional scale. This is based upon the systematic geometrical relationship between the north-south trending Tertiary basins and the conjugate NW-SE and NE-SW trending strike-slip faults of Thailand. The shapes and spatial distribution of these basins suggest that they formed as lazy-S and lazy-Z basins in direct response to strike-slip faulting. Within this framework the Red River, Mae Ping, Three Pagodas and Sumatra faults, which all trend NW-SE, represent the master dextral faults while the Northern Thailand, Uttaradit, Ranong and Klong Marui faults, which trend NE-SE, represent the conjugate sinistral faults. A genetic relationship between strike-slip faulting and basin formation is also supported by more detailed studies conducted by Remus et al. (1993) who have documented wrench features within the Phetchuban graben. The common theme of all three models is that southeastward extrusion is minimal, CW rotations are driven by dextral shear across plate boundaries, and Sundaland is broken into smaller sub blocks which have rotated in sympathy but which are separated by strike-slip faults. However, a simple dextral megashear model cannot explain northward transport of blocks along the western boundary of Sundaland.
Preferred tectonic model The extrusion model is an elegant way to link many of the geological features of Sundaland but data collected over the last decade indicate that the model must be refined (see Richter 1995). In summary, the new model must: (1) allow for crustal thickening and shortening within the collision zone (e.g. Dewey et al. 1989; Le Pichon et al. 1992; Houseman & England 1993); (2) recognize the importance of dextral shear along the western margin which may drive internal deformation within the extruding block (e.g. Dewey et al. 1989; Polachan et al. 1991; Richter et al. 1993); (3) reconcile palaeomagnetic data which show 20-30 ° of CW rotation of Indochina and Sibumasu relative to South China, rather than the predicted 40-50 ° (Yang & Besse 1993 and this study); (4) reconcile estimates of fault slip on the Red River fault which are far smaller (300-400 km) than the predicted thousands of kilometres (Tapponnier et al. 1990; Scharer et al. 1990; Lacassin et al. 1993; Leloup et a1.1993); (5) integrate basin analysis data which indicate that most or all of the fault motion on the Red River Fault is accommodated by transtension in the
221
Yinggehai basin and is not transmitted into the South China Sea (Qiming & Quanxing 1991; I. O. Norton, pers. comm. 1994); (6) explain Cretaceous and Palaeocene basin rifting ages along the South China shelf which predate extrusion (e.g. Chen & Pei 1993; G. Grabowski, pers. comm. 1994); (7) explain palaeomagnetic data which show CCW rotations in Borneo, Palawan, Sulawesi and the Celebes Sea (Schmitdke et al.1990; Fuller et al. 1991); and (8) account for interaction with the Australian and Philippine Sea plates along the southeastern margin of Sundaland (Richter et al. 1992; Hall et al. 1995). An alternative hypothesis is that Sundaland is composed of smaller sub-blocks which have rotated CW together but which have slid past each other (Fig. 15). Deformation of the Sibumasu block is dominated by the oblique subduction of the Indian Ocean Plate, resulting in CW rotations and regional dextral displacements along N-S and NW-SE trending faults. At this stage, we have not attempted a detailed structural analysis and it is expected that motions between smaller sub-blocks could be both dextral and sinistral. The model proposed by Lamb (1994) provides a good analogy (Fig. 15). The Indochina block, however, is too far behind the plate boundary to feel this motion and it deforms in response to extrusion, which is driven by convergence between the Indian Craton and Eurasia. These two modes of deformation meet in the Central Thai Basin, and could drive extension and uplift of the Doi Inthanon metamorphic core complex (Macdonald et al. 1993), the formation of intermontane basins, and the formation of the Central Thai basin and the Gulf of Thailand (Polachan et al. 1991). Finally, Sundaland is bounded to the southeast by the Australian plate and the Philippine Sea Plate (PSP). We hypothesize that the extruding block cannot simply override these plate boundaries and is partially or completely coupled to them. Thus, extrusion works with northward subduction of Australian oceanic crust and northward motion and CW rotation of the PSP to drive the CCW rotation of Borneo.
Conclusions (1) The Jurassic-Cretaceous Kalaw redbeds of the Shan Plateau of Myanmar yield a mean prefolding magnetization which is rotated CW 47.6 ° relative to the GAD. The age of this magnetization is interpreted to be Late Cretaceous or Early Palaeogene. The exact amount of rotation relative to South China varies with the inferred age but is approximately 25-30 ° CW over the range of possible ages. This rotation most likely took place between the Late Cretaceous and Early Miocene.
222
B. RICHTER & M. FULLER
(2) The mean Kalaw VGP is rotated 15.4 ° CW relative to the Khorat Lower Cretaceous VGE When both palaeomagnetic and geological errors are considered, this is consistent with a small CW rotation between the Khorat and Shan Plateaus. Such a rotation may be responsible for extension in the Central Thai basin and uplift and unroofing of the Doi Inthanon metamorphic complex in Northern Thailand. This rotation most likely began in the Middle or Late Oligocene. (3) Neogene and Upper Palaeogene palaeomagnetic data from Thailand show a variety of CW, CCW and unrotated declinations. We hypothesize that Sundaland is composed of large rigid blocks (the Shan and Khorat Plateaus, peninsular Thailand) which are separated by extensional and transtensional zones (Central Thai Basin, Pattani trough, Intermontane basins). The young scattered rotations appear to be limited to these extensional and transtensional zones where a strike-slip component of motion could easily drive local rotations. Within the deformation model proposed by Lamb (1994) the rotations recorded in the intermontane basins and Neogene basalts may record the fine scale velocity field while the rotations in the Shan and Khorat plateaus may record the large-scale velocity field. (4) The palaeolatitude data from the Kalaw redbeds show 8 ° of northward transport relative to the GAD, 12° relative to the Khorat Lower Cretaceous VGP, and 5 ° relative to the 100 Ma South China VGE The exact amount of northward motion varies with the inferred age of the respective VGPs. Part of this latitudinal motion can be explained by the extrusion model but plate reconstructions show that some or all of it must be accommodated by overall dextral motion between the Indochina and Sibumasu blocks or possibly within the western margin of the Sibumasu block. (5) CW rotations have been measured in Palaeozoic carbonates in peninsular Thailand and northern peninsular Malaysia. These were most likely remagnetized during the Late Triassic Indosinian orogeny but may have also been remagnetized during the Cretaceous. These data establish that a domain of CW rotation extends at least as far south as Langkawi Island, Malaysia.
South of this the pattern becomes more complex and mixed CW and CCW declinations are found in the same units. This zone of mixed CW and CCW grades southeastward into Borneo where the dominant sense of rotation is CCW. (6) The palaeomagnetic and geological data from Sundaland suggest several modifications to the extrusion tectonic model. Northern Sundaland has only rotated 25-30 ° CW with respect to South China during the Tertiary. The total southeastward translation of the Indochina block was only 300500 km. The extruding block is composed of smaller rigid blocks (the Shan and Khorat Plateaus, peninsular Thailand) which are separated by extensional and transtensional zones (Central Thai Basin, Pattani trough, Intermontane basins) which have accommodated internal deformation. As such, a regional CW 'domino' rotation is preserved in the large blocks while a variety of mixed rotations is found in the transtensional zones. Deformation of the Sibumasu block resulted from oblique subduction of the Indian Ocean Plate, whereas deformation of the Indochina block was dominated by extrusion driven by convergence between the Indian Craton and Eurasia. The extruding block could not simply override plate boundaries located to the south and east and was partially or completely coupled to them. Thus, extrusion worked with northward subduction of Australian oceanic crust and northward motion and CW rotation of the PSP to drive the CCW rotation of Borneo. This study would have been impossible were it not for the assistance of numerous colleagues throughout Thailand, Myanmar, and Malaysia. We would like specifically to thank S. Bunopas and V. Thitipawarn (DMR, Bangkok); U. T. Myint (MOGE); J. Murphy, T. Acomb, and N. Timble (Amoco Myanmar); C. Hutchison (University of Malaya, Kuala Lumpur); N. Halle (Petronas); A. Rb. Samsudin (University Kebangsan Malaysia); I. Norton, D. Maughn, G. Ramseyer, M. Fitzgerald, D. Leary and G. Grabowski (Exxon). Sample collection was assisted by C. Anderson and S. Cisowski. B. Dunn assisted in data analyses and interpretation. The study was funded by the National Science Foundation as well as by contributions by an industrial consortium consisting of Amoco, Chevron, Exxon and Unocal. Finally, we thank J. Briden and R. Hall for critical reviews which greatly enhanced the quality of the manuscript.
References ACHACHE, J. & COURTILLOT, V. 1985. A preliminary Upper Triassic palaeomagnetic pole for the Khorat plateau (Thailand): consequences for the accretion of Indochina against Eurasia. Earth and Planetary Science Letters, 73, 147-157. --, --. & BESSE, J. 1983. Paleomagnetic constraints on the Late Cretaceous and Cenozoic tectonics of southeast Asia. Earth and Planetary Science Letters, 63, 123-136.
BAIRD, A. & BOSENCE,D. 1993. The sedimentologica3and
diagenetic evolution of the Ratburi Limestone, Peninsular Thailand. Journal of Southeast Asian Earth Sciences, 8, 173-180. BARR,S. M. & MACDONALD,A. S. 1979. Paleomagnetism, age, and geochemistry of the Denchai basalts, northern Thailand. Earth Planetary Science Letters, 46, 113-124. & -1981. Geochemistry and Geochronologyof
PALAEOMAGNETISM OF THE SIBUMASU BLOCK Late Cenozoic basalts of Southeast Asia. Geological Society of America Bulletin, 92, 1069-1142. & -1991. Towards a Late Palaeozoic-Early Mesozoic Tectonic Model for Thailand. Journal of Thai Geosciences, 1, 11-22. & Haile, N. S. 1978. Reconnaissance Paleomagnetic Measurements on Triassic and Jurassic Sedimentary rocks from Thailand. Geological Society of Malaysia Bulletin, 10, 53-62. & REYNOLDS,P. 1976. Paleomagnetism and Age of the Lampang Basalt (Northern Thailand) and the Age of the Underlying Pebble Tools. Journal of the Geological Society of Thailand, 2, 1-10. ~, TANTISUKRIT, C., YAOWANOIYOTHIN, W. & MACDONALD, A. 1990. Petrology and Tectonic implications of Upper Palaeozoic volcanic rocks of the Chiang Mai belt, northern Thailand. Journal of Southeast Asian Earth Science, 4, 37-47. BENDER, E 1983. Geology of Burma - Beitrage zur Regionalen Geologie der Erde. Gebruder Borntraeger, Berlin. BUNOPAS, S. 1982. Paleogeographic history of Western Thailand and adjacent parts of Southeast Asia a plate tectonics interpretation. Geological Survey, Paper 5, Department of Mineral Resources, Thailand. CHEN, S. & PEI, C. 1993. Geology and geochemistry of source rocks of the Eastern Pearl River Mouth Basin, South China Sea. Journal of Southeast Asian Earth Sciences, 8, 1-4, 393-406. CHEN, Y. & COURTILLOT, V. 1989. Widespread Cenozoic(?) remagnetization in Thailand and its implications for the India-Asia collision. Earth and Planetary Science Letters, 93, 113-122. CHIBBER, H. 1934. The Geology of Burma. MacMillan, London. fOBBING, E., MALLICK,I., PITFIELD, P. & TEOH, L. 1986. The Granites of the Southeast Asian Tin Belt. Journal of the Geological Society of London, 143, 537-550. CURRA¥, J. R. et al. 1979. Tectonics of the Andaman Sea and Burma. In: WATK~S, J., MOrCrADERT,L., & DICKERSON,R W.(eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir 29, 189-198. DEWEY, J., CANDE, S. & PrrMAN, W. 1989. Tectonic Evolution of the India/Eurasia Collision Zone. Eclogae Geologica Helvitica, 82, 717-734. DUNS, R. & FULLER,M. 1984. Thermal demagnetization with measurement at high temperature (abs). LOS, Transactions of the American Geophysical Union, 65, 863. ENGLAND,P. & MOLNAR,P. 1990. Right lateral shear and rotation as the explanation for strike-slip faulting in eastern Tibet. Nature, 344, 140-142. ENKIN, R., YANG, ZHENYU,CHEN, Y. & COURTILLOT,V. 1992, Paleomagnetic Constraints on the Geodynamic History of the Major Blocks of China From the Permian to the Present. Journal of Geophysical Research, 97, 13 953-13 989. FISHER, R. A. 1953. Dispersion on a sphere. Proceedings Royal Society of London, 217, 295-305. FULLER,M., HASTON,R., LIN, J., RiCHTER,B., SCHMIDTKE, E. & ALMASCO,J. 1991. Tertiary Paleomagnetism -
-
,
_
-
-
223
of regions around the South China Sea. Journal of Southeast Asian Earth Sciences, 6, 161-184. HALLE, N. S. & KHOO, H. P. 1980. Paleomagnetic measurements on Upper Jurassic to Lower Cretaceous sedimentary rocks from Peninsular Malaysia. Bulletin of the Geological Society of Malaysia, 12, 75-78. HALL, R., FULLER, M., ALl, J. R. & ANDERSON, C. D. 1995. The Philippine Sea Plate: Magnetism and Reconstructions. In: Taylor, B. & Natland, J. (eds) Active Margins and Marginal Basins of the Western Pacific. American Geophysical Union Monograph, 88, 371-404. HALLS, H. 1978. The use of Remagnetization Circles in Paleomagnetism. Physics of the Earth and Planetary Interiors, 16, 1-11. HAMILTON,W. 1979. Tectonics of the Indonesian Region. United States Geological Survey Professional Paper, 1078, 345. HELMCKE, D. 1986. On the Geology of the Petchabun Fold-Belt (Central Thailand) - implications for the geodynamic evolution of Mainland S.E. Asia. Geological Society of Malaysia Bulletin, 19, 79-85. HOUSEMAN,G. & ENGLAND,P. 1993, Crustal Thickening Versus Lateral Expulsion in the Indian-Asian Continental Collision. Journal of Geophysical Research, 98, 12 233-12 249. HUTCHISON, C. S. 1989. Geological Evolution of Southeast Asia. Oxford University Press, Oxford Monographs on Geology and Geophysics, 13, 368. JOLIVET, L., DAVY,P. & COBBOLD,P. 1990. Right Lateral Shear along the Northwest Pacific Margin and the India-Eurasia Collision. Tectonics, 9, 1409-1419. JONES, C. 1978. The Geology and Mineral Resources of Perlis, North Kedah, and the Langkawi Islands. Geological Survey of Malaysia, District Memoir, 17, 257. KmSCHVrNK,J. 1980. The least-squares line and plane and the analysis of palaeomagnetic data. Geophysical Journal of the Royal Astronomical Society, 62, 699-718. LACASSIN,R., LELOUP,P. & TAPPONNIER,P. 1993, Bounds on strain in large Tertiary shear zones of SE Asia from boudinage restoration. Journal of Structural Geology, 15, 677-692. f LAMB, S. 1994, Behavior of the brittle crust in wide plate boundary zones. Journal of Geophysical Research, 99, 4457-4483. LELOUP, P., HARRISON, T., RYERSON, F., WENJI, C., Qt,L., ET AL. 1993. Structural, Petrological and Thermal Evolution of a Tertiary Ductile Strike-Slip Shear Zone, Diancang Shan, Yunnan. Journal of Geophysical Research, 98, 6715-6743. LE PICHON, X~ FOURNIER, M. & JOLIVET, L. 1992, Kinematics, Topography, Shortening, and Extrusion in the India-Eurasia Collision. Tectonics, 11, 1085-1098. MCCABE, R., HARDER, S., COLE, J. • LUMADYO,E. 1993, The use of paleomagnetic studies in understanding the complex Tertiary tectonic history of East and Southeast Asia. Journal of Southeast Asian Earth Sciences, 8, 257-268. ~ . , CELAYA,M., COLE, J., HAN, H., OHNSTAD,T., E T AL. 1988. Extension tectonics : The Neogene Opening
224
B. RICHTER & M. FULLER
of the North-South Trending Basins of Central Thailand. Journal of Geophysical Research, 93, 11899-11910. MACDONALD, A., BARR, S., DUNNING, G. & YAOWANOIYOTHIN, W. 1993, The Doi Inthanon metamorphic core complex in NW Thailand: age and tectonic significance. Journal of Southeast Asian Earth Sciences, 8, 117-125. MCELHINNY, M. 1964. Statistical significance of the fold test in paleomagnetism. Geophysical Journal of the Royal Astronomical Society, 8, 338-340. , HALLE, N. & CRAWFORD, A. 1974. Paleomagnetic evidence shows Malay Peninsula was not a part of Gondwanaland. Nature, 252, 641-645. MCFADDEN, P. 1990. A new fold test for paleomagnetic studies. Geophysical Journal International, 103, 163-169. MARANATE,S. & VELLA, P. 1986. Paleomagnetism of the Khorat Group, Mesozoic, Northeast Thailand. Journal of Southeast Asian Earth Sciences, 1, 23-31. METCALFE, I. 1988. Origin and assembly of south-east Asian continental terranes. In: AUDLEY-CHARLES, M. G. t~ HALLAM,A. (eds) Gondwana and Tethys. Geological Society, London, Special Publication, 37, 101-118. 1990. Allochthonous terrane processes in Southeast Asia. Philosophical Transactions Royal Society of London, A331, 625-640. MITCHELL, A. 1992. Late Permian-Mesozoic events and the Mergui Group Nappe in Myanmar and Thailand. Journal of Southeast Asian Earth Sciences, 7, 165-178. MOLNAR, E & TAPPONNIER, E 1975. Cenozoic tectonics of Asia: effects of a continental collision. Science, 189, 419-426. NUTALAYA,P. et al. 1983. Proceedings of the Workshop on Stratigraphic Correlation of Thailand and Malaysia, Volume 1, Technical Papers. Geological Society of Thailand and Geological Society of Malaysia. PELTZER, G.& TAPPONNIER, P. 1988. Formation and Evolution of Strike-slip Faults, Rifts, and Basins during the India-Asia Collision: An Experimental Approach. Journal of Geophysical Research, 93, 15 085-15 117. POLACHAN, S., PRADIDTAN,S., TONGTAOW,C., JANMAHA, S., [NTARAWIJITR, K. & SANGSUWAN, C. 1991. Development of Cenozoic Basins in Thailand. Marine and Petroleum Geology, 8, 84-97. QIMING, Z. & QUANXING,Z. 1991, A distinctive hydrocarbon basin - Yinggehai basin, South China Sea. Journal of Southeast Asian Earth Sciences, 6, 69-74. REMUS, D., WEBSTER, M. & KEAWKAN, K. 1993, Rift Architechture and sedimentology of the Phetchabun Intermontane Basin, central Thailand. Journal of Southeast Asian Earth Sciences, 8, 421-432. RICHTER, B. 1995. Tectonic Rotations of Southeast Asia The implications of New Paleomagnetic Data from -
-
Myanmar and Thailand upon the Extrusion Tectonic Model. MA Thesis, University of California, Santa Barbara. RICHTER, B., NORTON, I., SCHMIDTKE, E. & FULLER, M. 1992. Paleomagnetic Results from Southeast AsiaImplications for Paleogeographic Reconstructions (abs). LOS, Transactions of the American Geophysical Union, 73, 146--147. --, FULLER, M., SCHMIDTKE,E., U TIN MYINT, O TIN NGWE, ET AL. 1993, Tertiary Paleomagnetism of Myanmar and Thailand: Implications for the tectonic evolution of Southeast Asia. Journal of Southeast Asian Earth Sciences, 8, 1-4, 247-256. SCHARER, O., TAPPONNIER,P., LACASSIN,R., LELOUP, P., DALAI, Z. & SHAOCHENG, J. 1990. Intraplate tectonics in Asia: A precise age for large-scale Miocene movement along the Ailao Shan-Red River Shear Zone, China. Earth and Planetary Science Letters, 97, 65-77. SNANS1ENG, S., CHAODUMRONG, P., PRADIDTAN, S. BORIPATKOSOI1983. The Tertiary sedimentary rocks of Thailand. In: Proceedings of the Conference on Geology and Mineral Resources of Thailand, Bangkok, November 1983. 1-13. SCHM1DTKE, E., FULLER, M. & HASTON, R. 1990. Paleomagnetic data from Sarawak, Malaysian Borneo, and the Late Mesozoic and Cenozoic tectonics of Sundaland. Tectonics, 9, 1, 123-140. TAPPONNIER, P. & MOLNAR, P. 1976. Slip line field theory and large scale continental tectonics. Nature, 264, 319-324. --, PELTZER,G. & ARMIJO, R. 1986. On the mechanics of the collision between India and Asia In:COWARD, M. P. & RIES, A. C. (eds) Collision Tectonics. Geological Society, London, Special Publication, 19, 115-157. , LEDAIN, A., ARMIJO, R. & COBBOLD, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611-6. - - , LACASSIN,R., LELOUP,P., SCHARER,U., DALAI, Z., ET AL. 1990. The Ailao Shan/Red River metamorphic belt: Tertiary left lateral shear between Indochina and South China. Nature, 343, 431-437. WILCOX, R. E., HARDING,T. P. & SEELY,D. R. 1973. Basic wrench tectonics. AAPG Bulletin, 57, 74-96. WOLFART, R., WIN, M., BOITEAU,S., WAI, M., UK CUNG, P. & LWIN, T. 1984. The Tectonics of the Western Shan Massif, Burma. Geologisches Jahrbuch, 75, 285-294. YANG ZHENYU & BESSE, J. 1993. Paleomagnetic study of Permian and Mesozoic sedimentary rocks from Northern Thailand supports the extrusion model for Indochina. Earth and Planetary Science Letters, 117, 525-552. ZIJDERVELD, J. 1967. AC demagnetization of rocks: analysis of results. In: COLLINSON, D. W., CREER, K. M. & RUNCORN, S. K. (eds) Methods in Paleomagnetism. Elsevier, Amsterdam.
Timing of the Shan-Thai-lndochina collision: new evidence from the Pak Lay Foldbelt of the Lao PDR R O B E R T B. STOKES 1, 4, PAUL F. LOVATT S M I T H 2 & KO S O U M P H O N P H A K D Y 3
1 Surbiton Geological Services, 14 Dennan Road, Surbiton KT6 7RY, UK e Monument Resources (Overseas) Limited, 80 Petty France, London SW1H 9EX, UK 3 Department of Geology and Mines, Vientiane, Lao PDR 4 School of Geological Sciences, Kingston University, Penrhyn Road, Kingston upon Thames KT1 2EE, UK Abstract: The Pak Lay Foldbelt in the northwest of the Lao PDR is a product of the Indosinian orogeny, resulting from the collision of Shan-Thai and Indochina along a suture now marked by the Nan-Uttaradit ophiolite zone. Current hypotheses place the timing of this collision in either the Permian or the Triassic. Recent field work and subsequent laboratory analyses suggest that, in the area to the south of Pak Lay: pre-collision sediments are as young as Middle to Upper Jurassic; arc-related volcanism continued from the Triassic into the Late Jurassic; all these sediments and volcanic rocks form an imbricate zone; and the Cretaceous Khorat Group rests with marked unconformity above the imbricate wedge. From this we conclude that the Shan-ThaiIndochina suturing occurred in the Late Jurassic.
This paper presents a reinterpretation of the stratigraphy, structure and igneous activity in the southern part of Sainabouli Province in the northwest of the Lao PDR. It results from fieldwork, sampling and subsequent analyses undertaken by Monument Resources (Overseas) Limited and the Shlapak Development Company in the hydrocarbon exploration of their Contract Area. A total of six weeks of fieldwork was carried out during the months of November and December in 1992 and 1993 making a reconnaissance geological survey of the Contract Area from the Muang Kenthao region in the south to the Sainabouli Basin in the north. The area of the present study lies to the west of the Nam Khong (Mekong) River and to the south of Muang Pak Lay.
Current interpretations
Tectonic setting Workman (1972) described the Pak Lay region as the exposed part of a NNE-trending Indosinian linear foldbelt developed along the western margin of the Indosinia (Annamia) cratonic block. The folding was attributed to an early phase (approximately Late Triassic) of the Pacific Orogeny (Workman & Page 1967). In plate tectonic reconstructions, continental SE Asia has been shown to be a composite of allochthonous terranes which accreted to one another during the Palaeozoic and Mesozoic (Metcalfe 1988). There is general agreement that
they had all sutured to each other by the Late Triassic. Of these major terranes, only the ShanThai Block (also called Sibumasu) and the Indochina Block (also called Indosinia) are directly related to the Pak Lay Foldbelt. According to Cooper et al. (1989), NE-dipping subduction was initiated in the Permian beneath Indochina in addition to a possible southwest-dipping subduction beneath Shan-Thai. The two plates collided in the Early Triassic creating (inter alia) the Pak Lay-Petchabun Foldbelt. Extensional collapse of the Indosinian orogen during the Triassic allowed continental extensional basins to develop in northeast Thailand and adjacent parts of Lao, and then thermal subsidence, during the Jurassic and Cretaceous, allowed the red-bed Khorat Megasequence to be deposited over much of northeast Thailand and the Lao PDR. The suture between the Shan-Thai and Indochina blocks is generally accepted as being marked by the Nan-Uttaradit ophiolitic zone (Hutchison 1989, Barr & Macdonald 1991). This narrow zone extends northeast from Uttaradit along the Nan River (Fig. 1). It is believed to extend through the Lao PDR, past Luang Prabang, and into Vietnam as the Lay Chau Fault. Its location within Lao is uncertain.
Geological history of the Pak Lay Foldbelt The first geological interpretation of the Pak Lay area (Fig. 2) was produced from fieldwork by Bourret (1925) who summarized the stratigraphy,
FromHall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 225-232.
225
226
R . B . STOKES ET AL.
~'.,F ~
106 ° E
VIETNAM
-,
o
\ . "d'q ~it . °
i
'
~1~
SAKHON - NAKHON'~ BASIN
/
ET BASIN
KHORAT BASIN
1t
...... T H A I L A N D
.............................. ' .........
5
I
,P
14°N
C.i
,1
0
i
r]
Study Area
i
250km
i
Nan-Uttaradit Ophiolite Zone ~
----- Subcrop against Quaternary Alluvium D
Volcanics
r-] out crop of Indosinian Unconformity
Permo-Carboniferous Limestones
Triassic Huai Hin Lat Formation
n
- ~ Himalayan Folding
Fig. 1. Regional geological summary of the Monument/SDC Contract Area of the Lao PDR showing the area covered by this study.
PAK LAY FOLDBELT, LAO PDR
,, 18o15,N .......................... ........
•~
5 m
227
c. ,~..K~.
~
~!i~ ~ ~
THAILAND
18°00'N i~ii~i'~i~
C-J ED.R.
!~¸~'~~
17°45'N
THAILAND
17°30'N 101°00'E
E C-J
101 °15'E
101 °30'E
Phon Hong Group
Metamorphic Aureole
f
Major Faults
Khorat Group
Granite Cretaceous
f
Thrusts
Undifferentiated ? Late Palaeozoic to Late Jurassic Argillites, Triassic and Jurassic Volcanics, with local development of Mid-Permian Limestone in West
Granite Triassic Phalat Formation in East
• (~) Locality mentioned in text
20km I
Fig. 2. Simplified map of the geology of the Pak Lay Foldbelt including locations mentioned in the text.
228
R.B. STOKES ET
elucidated from superposition, as follows from base to top: (1) uncrushed crystalline basement; (2) terrain ancien: pre-Uralian shales and sandstones, Uralian limestones and cherts, very highly folded on N-S trends; (3) andesitic volcanic rocks: no trace of crushing in the border region where they are unconformable on older strata; (4) terrain rouge: gently dipping in the border mountain region; (5) gr6s sup6fieurs: horizontal or hardly dipping, except on the Nam Lay. Bourret's reasoning for this stratigraphic history can be summarized as follows. (i) The age of the granites is Permian or Early Triassic because they are post-tectonic and therefore younger than the intense folding which affected the region in preTriassic times. The granites pre-date the Mid Triassic 'terrain rouge' even though no clasts of granite could be found in the red-beds. (ii) The oldest sediments (the 'terrain ancien') are highly folded, often vertical, sandy shales and fine sandstones with calcareous shales, cut by numerous quartz veins. In the southern part of the province they occur only in the deepest parts of the valleys, because the higher parts are all formed by 'roches vertes'. The strike is always N-S and dips are usually very steep and sometimes vertical. In the area to the east of the Mekong (and to the west of Chiang Khan) these sediments are not metamorphosed, but elsewhere they are. Bourret dates these sediments as pre-Uralian and probably mainly Devonian. (iii) A mass of andesitic rocks forms an elevated plateau, between the Nam Sang
AL.
and Pak Leng, which is cut by deep W-E fiver valleys on the steep eastern scarp. According to Bourret (p. 86) 'we can observe everywhere in these valleys the superposition of the roches vertes on the 'ancient' shales'. Bourret found no evidence of bedding in the andesites. On the western margins of the plateau, however, stratified tufts showed moderate dips. (iv) The andesites are younger than the Uralian because they have locally 'digested' fusulinid limestones, and they are older than Triassic because pebbles of the andesites are included in breccias and conglomerates of the 'terrain rouge'. They can only be, therefore, either Permian or Lower Triassic. Bourret states that they are horizontally bedded in general. Workman & Page (1967; repeated in Page & Workman 1968) carried out a photogeological study following a field traverse of the area and established a stratigraphy which was very similar to that of Bourret. Their summary is reproduced here with ages taken from their text. (1) Deposition of Palaeozoic sediments (Upper Carboniferous to Permian argillites; thin beds of shale and siltstone, with occasional fine-grained sandstones. The shales are generally black or dark grey, often studded with crystals of pyrite). (2) Cimmerian 1 folding (approximately Upper Triassic). (3) Intrusion of granitic rocks. (4) Triassic volcanic episode. (5) Deposition of Khorat Group (Triassic to Jurassic). (6) Gentle epeirogenic folding and large-scale faulting. The end of the older Cimmerian 1 folding was
Ban Nam Pou Nong Phuk Heo Ban Kok Ke I , Nam Hoi [ ] I
NW
I
L_
'
~ i ~ . ~ .
~:~ii~i~i~
I
'~i~:~ii~#=~ ~'. !~=~i' ',~.,
'I
Nam Hueung ', SE
'I
'I, +
+
4-:4-
4-
4-
LEGEND Gres superieurs Jurassic Terrain rouge Triassic volcanics Triassic Granite intrusion Palaeozoic sediments
Fig. 3. Cross-section interpretation of the Pak Lay Foldbelt taken from Bourret (1925) and modified from information in Fontaine & Workman (1978).
J
PAK LAY FOLDBELT, LAO PDR marked by post-tectonic granodiorite intrusion, after which erosion and planation took place before the Triassic volcanic episode and the deposition of the Khorat group. In his review of 1972 (repeated in Fontaine & Workman 1978), Workman added (inter alia) the following points to the geology of the Pak Lay area. (1) The intrusive rocks, mostly with compositions in the range of granodiorite to tonalite, were emplaced in an early phase of the Indosinian orogeny, probably in the Early Triassic. (2) The thick volcanic formation in the Triassic at the base of the continental Mesozoic series conformably underlies the gently dipping red beds of the Upper Triassic to Jurassic. The group consists of andesite and trachyte lava flows interbedded with siliceous and feldspathic crystal-tufts and agglomerates. These are approximately Middle Triassic in age. The currently-accepted history of this foldbelt is that presented by Fontaine & Workman (1978) which can be deduced from the cross section in Fig. 3 and summarized as follows: (1) deposition of Palaeozoic sediments; (2) Indosinian orogeny (Early or Middle Triassic) resulting in isoclinal folding of the Upper Palaeozoic formations along N-S axes; (3) intrusion of granitic rocks (probably Early Triassic); (4) period of erosion and planation; (5) Triassic volcanic episode (approximately Middle Triassic); (6) deposition of the continental Khorat Group (Late Triassic to Jurassic) conformably overlying the thick volcanic formation; (7) gentle epeirogenic folding and large-scale faulting. This view of the history of the foldbelt is largely responsible for the acceptance of a Triassic age for the Indosinian orogeny and the related collision of Shan-Thai to Indochina.
New evidence from the Pak Lay Foldbelt
229
based essentially on foraminifera, come from the localities numbered 1-4 on Fig. 2. Their ages and lithologies are as follows: (1) Upper Murgabian to Lower Midian fossiliferous packstone; (2) Midian (or possibly higher i.e. Dzhulfian) fossiliferous boundstone; (3). Midian to Dzhulfian fossiliferous packstone and boundstone at Ban Buamlao; (4) yielded a wagenophyllid coral of probable Murgabian age. The remainder of the sedimentary rock samples are almost universally barren of fossils, but one exception is critical to our dating of the clastic sequence. Thinly-bedded, grey to dark grey claystones at Locality 5 have yielded a lean, moderately well preserved, terrestrially derived palynomorph assemblage most characteristic of a Middle to Upper Jurassic age. Taxa present include Classopollis spp., Calliatisporites dampierri, Cerebropollenites mesozoicus, Uvaesporites sp. and P. elatioides, along with some rare, reworked striate bisaccate forms of Permian to Triassic character.
Radiometric ages of the volcanic rocks Volcanic rocks have long been known from Permian and Triassic sequences in the Pak Lay and Loei-Petchabun foldbelts. Our data show that this vulcanicity continued into the Late Jurassic with the youngest age being obtained from volcanic rocks in the far west of the area. The field relationships of the Bajocian basalt from Locality 8 are unknown, but the Kimmeridgian alkali basalt from Locality 9 is part of a volcanoclastic sequence which dips 50°E with a strike of 175 °, conforming to the regional structure seen in the sedimentary sequences, and thus suggesting that it is extrusive. Details of the dates obtained are summarized in Table 1.
Biostratigraphy of the sediments Our samples of the 'Uralian' limestones in the Pak Lay region gave only Permian ages within the range Murghabian to Dorashamian. These ages,
Radiometric ages of the granites The Triassic age of the widespread granites in the Pak Lay and Loei Foldbelts is confirmed by our
Table 1. Ages of volcanic and plutonic igneous rocksfrom the Pak Lay Foldbelt determinedfrom whole-rock K-Ar analyses (localities indicatedon Fig. 2) Locality number 6 7 8 9 10 11
Lithology volcaniclastic lithic wacke crystal tuff basalt alkali basalt granite granite
Radiometric date ( M a )
Chronostratigraphic Age
210.4 _ 5.2 226.7 _+5.4 167.2 +_5.2 152.4 _+6.3 213.8 + 8.9 117.0 _+3.0
Late Triassic (Norian) Late Triassic (Carnian) Middle Jurassic (Bajocian) Late Jurassic (Kimmeridgian) Late Triassic (Norian) Mid-Cretaceous (Aptian)
230
R.B. STOKES ET AL.
dating of the Ban Pakthoun granite (Locality 10) which intrudes the (?)Upper Permian Phalat Formation in the east of the area. The dating of the Ban Samxong granite (Locality 11), previously regarded as Triassic, was made to test the Jurassic age for part (Locality 5) of the stratified sequence which it intrudes. The mid-Cretaceous age of the granite supports the biostratigraphy of the argillites. Details of the dates obtained are given in Table 1. Structure Relationship of the volcanic rocks to the Indosinian unconformity. The present authors have been unable to find any evidence in support of Bourret's conclusions, followed by all subsequent workers, that the volcanic rocks are unconformable above the Palaeozoic sediments, that they are conformable beneath the Khorat Group, and that they post-date the major episode of folding. On the contrary, both theoretical considerations and field evidence lead to the conclusion that the volcanic activity preceded the Indosinian orogenic event. The widespread basaltic to andesitic volcanic rocks have been interpreted, in their southern extension within the Petchabun belt in east central Thailand, to have formed above an east-dipping subduction zone (Bunopas & Vella 1983). This volcanism would, therefore, have pre-dated the collision event. Bourret's assertion that the volcanic rocks occupy the high ground and the sediments the low ground (and, therefore, that the volcanic rocks are unconformable above the folded sediments) is
demonstrably untrue. For example, the road from Ban Kengsao (on the Mekong) to Ban Bouamlao (in the Nam Gnang valley) cuts through the major NNE-trending ridge to the southwest of Pak Lay. The crest of this ridge exposes a thick sequence of claystones and fine to medium grained tufts. All the rocks dip steeply to the east. Passing westwards down from the ridge to the broad alluvial valley of the Nam Gnang, volcanic rocks predominate in the scattered outcrops. Imbrication in the pre-Khorat sequence. Bourret's assertion that the structure of this region is simple seems false. Mapping the distribution of the volcanic rocks is particularly difficult, as Whittle (1970) pointed out. With few exceptions, dips in the pre-Khorat rocks are relatively steep. The intimate association of the volcanic rocks, claystones and Permian limestones, and their apparent conformity suggest that they have been strongly tectonized and show either isoclinal folding or that the Pak Lay Foldbelt is a zone of structural imbrication which has juxtaposed sediments and volcanic rocks of varying ages. On the ground it was found that there were numerous unpredictable changes from argillites to volcanic rocks and back along our traverses. Where bedding is seen within the volcanic rocks, it is steeply dipping and of the same order of magnitude and orientation as is seen in nearby sediments. At Locality 12, volcanic tufts strike 036 ° and dip 75 ° to the SE; oil-stained siltstones and claystones roughly 1 km to the SSE at Locality 13 strike at 014 ° and dip at 76 ° to the
NW
SE
LEGEND
Imbricate zone of Late Palaeozoic to Late Jurassic sediments and Triassic to Late Jurassic volcanics
Granite and its MetamorphicAureole Khorat Group (Cretaceous) Late Jurassic Sediments Triassic to Late Jurassic Volcanics .
.
.
.
.
.
Mid Permian Limestones
Fig. 4. Diagrammatic cross-section showing the revised interpretation of the Pak Lay Foldbelt.
PAK LAY FOLDBELT, LAO PDR ESE. Due to a lack of dating of most exposures, imbricate thrusts can only be demonstrated where Permian limestones appear to overlie Triassic or Jurassic volcanic rocks in the northwest of the study area. Pre-Khorat
Group
(lndosinian)
unconformity.
There is clearly a major angular unconformity between the Upper Jurassic argillites and volcanic rocks, and the overlying Cretaceous Khorat Group. At Locality 5, the former dip 50°E with a strike of 008 °. The Khorat Group shows a consistent regional strike (140°-160 °) and gentle dip (c. 20°SW) in the area of Locality 14.
History of the Pak Lay Foldbelt Whilst individually none of the evidence above is unequivocal, these obervations together lead to the following conclusions concerning the geological history of the Pak Lay Foldbelt. The western part of the Pak Lay Foldbelt consists of (?)Upper Carboniferous to Upper Jurassic sediments and associated Triassic to Upper Jurassic lavas, agglomerates and tufts. The Triassic to Jurassic volcanic activity is interpreted as arcrelated and to have occurred during a period of subduction which preceded the collision event. The latter can, therefore, be no older than Late Jurassic. The Upper Palaeozoic to Jurassic sequence of the western zone was deformed by a compressional event at the end of the Jurassic which produced an imbricate, accretionary wedge with steeply dipping strata striking NNE. These structures result from the collision of the Shan-Thai and Indochina blocks during the Indosinian orogeny. Following a period of considerable erosion, the Cretaceous Khorat Group was deposited unconformably, overstepping the truncated strata of Carboniferous to Jurassic age. A major granite was intruded later in the mid-Cretaceous, probably associated with the basin inversion in the Khorat region before the deposition of the Upper Cretaceous Phon Hong Group (Lovatt Smith et al. 1996). Tilting to the SW and subsequent erosion of the uplifted Paklay-Petchabun Foldbelt has stripped the Cretaceous from much of the area revealing the structurally complex foldbelt. Our schematic reinterpretation of the line of section given by Bourret (1925) is presented as Fig. 4.
231
Discussion The results presented above relate to the area west of the Nam Khong (Mekong) river. The area to the east of the Mekong consists essentially of lesstightly folded Palaeozoic strata which were intruded by granites in the Late Triassic. The line of separation between these two major tectonostratigraphic terranes is difficult to fix. It was regarded as the Kenthao-Saiapoun fault zone by Workman & Page (1967). We have drawn it somewhat arbitrarily along a number of faults along the eastern margin of the C-J unit on Fig. 2. The relationship of these terranes with each other is obscure. The Late Triassic deformation and plutonic activity in the east represent an earlier event in the history of the western margin of the Indochina Block. The revised dating of the Ban Samxong granite resolves two anomalies noted by earlier workers. Firstly it explains why, contrary to his expectations, Bourret found no pebbles of the granite in the basal conglomerates of the Khorat Group. This is now to be expected, since the granite is of a later age. Secondly, the Cretaceous age explains why Workman & Page (1967) found it to be 'unlike any of the other intrusive rocks of the area .... ', since all the other granites which they examined are of Triassic age. Cooper et al. (1989) suggested that the Khorat Group was deposited during a phase of thermal subsidence following continental collision and extension during the Late Triassic. However, as Racey et al. (1996) point out, since the majority (if not the totality) of the Khorat Group is Cretaceous in age, a different mechanism for basin formation must be invoked. Our results suggest that the continental collision was in the Late Jurassic and that the thermal subsidence model is still acceptable. We thank the Directors of Monument Resources (Overseas) Limited, the Shlapak Development Company and the Department of Geology and Mines (Vientiane) for permission to publish this work. We are particularly pleased to thank the numerous villagers who gave help and hospitality during the field work. The paper has benefited from reviews by Charlie Bristow (Birkbeck College, London) and Ian Metcalfe (University of New England). All micropalaeontology, palynology and radiometric (K-As) datings were carried out by the Geochem Group, Chester.
References BARR, S. M. & MACDONALD,A. S. 1991. Toward a late Palaeozoic-early Mesozoic tectonic model for Thailand. Journal of Thai Geosciences, 1,
11-22. BOURRET, R. 1925. Etudes geologiques dans la region
de Pak-Lay (Moyen Laos). Bulletin du Service GEologique de l'Indochine, 24. BUNOPAS, S. & VELLA, P. 1983. Tectonic and geologic evolution of Thailand. In: Ntra'ALAYA, P. (ed.) Proceedings of a workshop on stratigraphic
232
R.B. STOKES ET AL.
correlation of Thailand and Malaysia, Bangkok. 1, 307-322. COOPER, M. A., HERBERT, R. & HILL, G. S. 1989. The Structural Evolution of Triassic Intermontane Basins in Northeastern Thailand. In: THANASUTmPITAK, T. & OUNCHANUM, P. (eds) Proceedings of the International Symposium on Intermontane Basins: Geology & Resources, Chiang Mai. 231-242. FONTAINE, H. & WORKMAN, D. R. 1978. Review of the Geology and Mineral Resources of Kampuchea, Laos and Vietnam. In: NUTALAVA, P. (ed.) Proceedings of the Third Regional Conference on Geology and Mineral Resources of Southeast Asia, Bangkok November 14-18, 1978. Asian Institute of Technology, Bangkok, 539-603. HUTCHISON, C. S. 1989. Geological Evolution of SouthEast Asia. Oxford University Press. LOVATT SMITH, P. E, STOKES, R. B., BRISTOW, C. & CARTER, A. 1996. Mid-Cretaceous inversion in the northern Khorat Plateau of Lao PDR and Thailand. This volume. METCALFE, I. 1988. Origin and assembly of south-east Asian continental terranes. In: AUDLEY-CHARLES, M. G. & HALLAM, A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publication, 37, 101-118. PAGE, B. G. N. & WORKMAN,D. R. 1968. Geological and Geochemical Investigations in the Mekong Valley, between Vientiane and Sayaboury and at Ban Houei Sai. Institute of Geological Sciences, Overseas Division, Report, 9. RACEY, A., LOVE, M. A., CANHAM, A. C., GOODALL, J. G. S., POLACHAN, S. & JONES, P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group North Eastern Thailand: Part 1, stratigraphy and sedimentary evolution. Journal of Petroleum Geology, in press. WHITTLE,G. 1970. Pak Lay Sheet. In: Notes on 1:250 000 Photogeological Reconnaissance Map of Laos. Institute of Geological Sciences, Overseas Division. WORKMAN, D. R. 1972. Geology of Laos, Cambodia, South Vietnam and the eastern part of Thailand a review. Institute of Geological Sciences, Overseas Division, Report, 19. -& PAGE, B. G. N. 1967. Interim Report on the regional geology and geochemistry of the Sanakham-Pak Lay area of western Laos. Institute of Geological Sciences, Overseas Division, Report, 8.
Mid-Cretaceous inversion in the Northern Khorat Plateau of Lao PDR and Thailand PAUL F. LOVATT SMITH l, R O B E R T B. STOKES 2, C H A R L I E B R I S T O W 3 & ANDREW CARTER 3
1 25 Tredegar Road, London E3 2EH, UK 2 Surbiton Geological Services, 14 Dennan Road, Surbiton, Surrey KT6 TRY UK 3 Department of Geology, Birkbeck College, University of London, Malet Street, London WC1E 7HX, UK
Abstract: Evidence is presented from the study area for the occurrence of a regional compressive tectonic event in the mid-Cretaceous(Aptian-Cenomanian). This is tentatively attributed to the effects of a distant continent-continent collision to the west. Pre-existing structural trends were reactivated parallel to palaeo-continental sutures to the northeast (Song Ma and Song Da) and west (Nan). The event interrupted the latest Jurassic to earliest Palaeocene subsidence and continental sedimentationof the Khorat Plateau Basin. The primary effects are shown by seismic data to be regional tilting, compressive folding, reverse faulting and inversion of the basin subsidence pattern. The data do not support the current view that such structuring was initiated during the Tertiary. However they do show that there was some reactivation of mid-Cretaceous structures during regional uplift in the Tertiary. This revision of timing of structuring suggests the hydrocarbon potential of the area may be greater than previously anticipated.
The study area comprises the MonumenffShlapak Contract Area in the Lao PDR and the western margins of the Khorat Plateau in Thailand (Fig. 1). The data presented in this paper results from Monument's hydrocarbon exploration programme in Lao PDR, and from sedimentological studies of the Khorat Group in Thailand. Geological fieldwork in the Lao PDR, totalling 19 weeks, was undertaken by PLS, RBS and others during the dry seasons of November-December 1991, January, March-April and November-December 1992, and November-December 1993. This resulted in the production of 30 geological maps at a scale of 1:100,000. Samples were analysed in the laboratory for petrography, geochemistry, biostratigraphy and isotopic dating. Geophysical surveys in the Vientiane Basin were conducted in April 1992 and March-May 1993 and resulted in 4 2 0 k m of seismic, 2146 km of airborne gravity, some 500 km of surface gravity and 6450 km of aeromagnetic data. Fission track analysis on samples from the Khorat Group cropping out in the Khorat monocline of Thailand was carried out by CB and AC.
Regional setting The Lao PDR study area is situated on the northern and western margins of the Khorat Plateau Basin. This is a large area of cropping out Mesozoic (mainly Cretaceous) continental deposits covering the northeast of Thailand, parts of Lao PDR and
Cambodia (Fig. 1). The basin system was once more extensive but has been reduced in area and thickness by significant regional Tertiary erosion. Present-day sub-basins of the Khorat Plateau Basin are defined by the distribution of Upper Cretaceous sediments. The Vientiane and Pakxan topographic plains are the Lao extensions of the Upper Cretaceous Sakhon Nakhon Basin. Around the northern margin of this basin the underlying Khorat Group (uppermost Jurassic to Aptian) is exposed in a pattern of eroded anticlines and outlying synclines (Fig. 2). The Phetchabun Foldbelt separates this eastern area from the NakhonThai/Sayabouri Basin where Mesozoic sediments belonging to both the Upper Cretaceous Phon Hong Group and the Khorat Group are also preserved. The similarity of the Cretaceous succession of this basin to that further to the east indicates that both areas can be considered as part of the same basin system. The surrounding massifs consist mainly of Precambrian to Upper Palaeozoic rocks, though some Upper Palaeozoic to Upper Jurassic sediments and volcanics are exposed in the Phetchabun Foldbelt. Two structural trends are present in the study area. An approximately north-south trend predominates in the west. In the Nakhon-Thai/ Sayabouri Basin, Cretaceous sediments are gently folded and faulted on this trend. Underlying sediments in this area and in the Phetchabun Foldbelt have been more severely tectonized by east-west
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, Geological Society Special Publication No. 106, pp. 233-247.
233
234
P.F. LOVATT SMITH ET AL. 1 103~E
i 108:E
~-~"N..
CHINA
.~j'X..'-'-/'" ~
" " ""o
250km
t
\ °.
L-.j"
VIETNAM 0 Hanoi
\-,~. -
o,'
%
~.°
Gu!f qt Tongkin
.?
Phrabang
20~S
.'~. f
~.0 E D.
Uttaradit 0
South China Sea
~:LL~ ).' SAKHON NAKHON• • . "%'%O/k~ "1,¢:.. "0" :.BASIN• " T l l a l ~ h e d ~ NAKHON THA ,'~.. ),-I7 L'udon••.~•i•i'••'.•.'. I. BASIN. '.•-~ t~2•.Thani. ~ • • • •. -. ~ ,,, ~.J J..v...... :•-.-.• .,,.,~ .,~/~:--•.... S4~tvanna~h~t~' ~t { ......... Phetchabuno :
i •
~s,
• : "
•
•
•
•
•:':'.•.-..
~'
z/c,),.
•
Da Nang
•
• ::::: ::::: K,.ORAT.BAS.~N.:.:.:.:.:...~ • i.l.i.i.i.i.i.i•i•i.i.i.i.i.i.i
Ubon&i.i.i~'\." ~.'~ ,~'" i.i•i.i&i.i.i.i.i.i.~.i.i.i.i.i•i.i.i.i.i, ili~"~lPakse '
15S
'.°.'.°•'."
THAILAND %. Bangkoko
"."
"°•••
• '.'
•°•
. . i • : • : • : • i • : - i .; . ; - . - ~ " - : - : ~ ' - - . = - - " ~ ~i ' - - ;- - . f L
".•.'
....
d
%. "~
"
AMBODIA
LEGEND •
Granites with Cretaceous K/Ar age dates
. . . . . .
~,-
103E~.x•/ I -
O
"
"I (~
~~:-:..-.::--,..- ~,0oml Gulf of Thailand
\
,-,"'/(:
I"'"..
'--"
•
X'\ (-o°
•
."
.,~1..~
..
•.
__.,i
,..j I
F.'.~, /.' .' .~
~'
~ ~.
t..
::,;•:~5.::: r-
, . .j
",
•
'~ I
Outcrop of Jurassic / Cretaceous sediments
,, - Subcrop of Jurassic / Cretaceous against Tertiary / Quaternary
,.i .r... ... .
.
~•
;'~, / - "
~. • . ",,108 "t I
Axis of major foldbelts /
Major Faults Study Area
Fig. 1. Simplified geological map of the Khorat Plateau region•
compression. Along the east side of the foldbelt Mesozoic sediments reappear in the Khorat Monocline, which defines the westem margin of the Khorat Plateau Basin. Within the central and
eastern part of this Basin a northwest--southeast structural trend becomes more dominant, parallel to the Annamitic Foldbelt. This trend is characterized by large wavelength anticlines which have topo-
CRETACEOUS INVERSION IN THE KHORAT PLATEAU ,
,
1
235
...) ,
,
,
f
PO. ,:.. XIENG-KHOANG
---~MASS,F
1 "~"°'~.
/!
V ! ET NA M
"....
~ :::::::::::::::::::::::::::
.J ~ : : : : : : : : : : : : : : : : : : : : : :
lo~k~
? .EGEND
(~
Phon Hong Group Outcrop
(~
Paleozoic or Basement Outcrop
Khorat Group Outcrop
,,If" Major Anticline
, ~ ' Major Syncline J
Major Lineament/Fault
Line of Section
Fig. 2. Lao PDR study area showing the major structural elements and location of seismic lines of Figs 5-8.
graphic expression. These include the Nam Leuk, Khammouan and Phu Phan Uplifts (Figs 1 and 2). The structures are often bounded by deep-seated, high angle reverse faults with large throws, often of more than 1 km.
Stratigraphy The stratigraphy of the Vientiane Basin is summarized in Fig. 3 and is based on that proposed by Ha Luong Tin & Do The Que (1991), being here modified and further subdivided in order to facilitate correlation with the better known sequences in Thailand. The maximum thicknesses indicated in Fig. 3 are based mainly on geophysical data supplemented by outcrop measurements. The chronostratigraphic section in Fig. 4 is based on an east-west seismic line across the Vientiane Plain. The Permian is present as marine platform and marginal facies, comprising limestones and clastic sediments. The Triassic crops out only in the west of the study area where it is present as extrusive volcanics and volcaniclastics. To the east its facies is assumed to be fluvio-lacustrine, similar to its equivalents in Thailand. The Khorat Group comprises mainly fluvial sediments. The Lao Khorat Group stratigraphy substantially follows that of Thailand, recently updated by Racey et al. (1994), except that the lowest unit, the Upper Norian to Rhaetian Phulekphey Formation (Thai Nam Phong Formation equivalent), has been assigned to the
pre-Khorat section. This is because a long hiatus between this formation and the overlying Nam Set Formation is indicated by the absence of LowerMiddle Jurassic sediments. The Phon Hong Group stratigraphic nomenclature is based on the results of potash exploration carded out between 1983 and 1986 by Vietnamese geologists, quoted by Ha Luong Tin & Do The Que (1991). In the Vientiane area the Thangon Formation reaches a maximum thickness of 700 m and consists of three evaporitic sedimentary cycles. A basal conglomerate has been noted in one locality in the west of the Vientiane Basin (Howlett 1993). The basal evaporite cycle contains a halite bed which reaches a maximum thickness of 340 m in boreholes. In the southeast part of the Vientiane Plain this salt has been mobilized into pillows and small diapirs. The evidence for an Albian to Cenomanian age for the Thangon (Maha Sarakham equivalent) Formation evaporites is based on biostratigraphic data from palynological analysis of cores taken in the Sakhon Nakhon Basin of Thailand (Sattayarak et aI. 1991). This formation was identified by Utha-Aroon (1991) as being deposited in a continental setting. The Saysomboun Formation is up to 150 m thick in the Vientiane area. It consists of predominantly red-brown fluvial claystones grading upwards into siltstones and fine sandstones. The claystones are dated by palynology as Upper Cretaceous both in the Vientiane by Ha Luong Tin & Do The Que
236
P.F. LOVATT SMITH ET AL.
MAX THICKNESS (KM) 0
VIENTIANE STRATIGRAPHY .°o..•l.O.o°.
A
ABBR.
THAI EQUIVALENT
SAYSOMBOUN FM.
K2Sb
PHU THOK FM.
THANGON FM.
K2Tg
MAHASARAKHAM FM.
BAN THALAT FM. CHAMPA FM.
K 1Bt K1C p
KHOK KRUAT FM. PHU PHAN FM.
0
BAN ANG FM. PHU PHANANG FM.
K 1Ba K 1Pp
SAO KHUA FM. PHRA WlHAN FM.
E ~ o z ~
NAM SET FM.
KINS
PHU KRADUNG FM.
PHULEKPHEY FM.
TPI
NAM PHONG FM.*
HUAI HIN LAT FM. LOWER TUFF / SST
THh Tv
DAT FA SHALE PHU HI SST.
PHALAT FM.
PPI
U. PERMIAN CLASTICS
NALANG FM.
PNI
PHA NOK KHAO FM.
A A A A A A A
LATE
A A A A ^ ^ A ~ ~
~
-
W 0 W
~
.,-. . .-~'. . . . . . o. .. ,. . . . . . .
o
EARN
.
.
.
.
.
711lllllll
I
1°o.......o...
? JUR. -
o.oV"" • i . e V . " ".'4V."
3
iTRIASSIC
LATE
,,i
,,,,,
• V"
~-
"V"
,-"V
•
•
LATE
s
6
,
LATEST
LOWER CLASTICS
tl
NAMURIAN -VISEAN
rr" 7
O ~
'm
"-" I
K
NAM THOM FM.
CNt
NASAFM.
DNs
NAM XOI FM.
O/S
,.-~
DEVONIAN z ,<
~
.
MIDDLE EARLY
i
J ~
i
J i
]
i
I
i
t
[
I
Metamorphic Basement
O .J
* The Nam Phong formation is currently included in the Khorat Group in Thai stratigraphy.
Fig. 3. Generalized stratigraphy, Vientiane area, Lao PDR.
(1991) and in the Savannakhet Basin of Central Laos (Huynh Xuan Dang and Boutheung Phengthavongsa 1989). In the Pakxan Plain the upper part of the Saysomboun Formation is preserved and contains massive beds of red-brown, cross-bedded sandstones. One of these sandstone units is over 180 m in thickness and forms a scarped ridge (Sayphou Khout) in the centre of the
Pakxan Plain. This represents the northern continuation of the Thai escarpment in which Phu Tok Formation sandstones were identified as having aeolian characteristics by Sattayarak (1983). In the Nakhon-Thai/Sayabouri Basin to the west this formation is substantially thicker than in the Vientiane and Pakxan areas. Some authors have assigned an Lower Tertiary age to this formation.
CRETACEOUS INVERSION IN THE KHORAT PLATEAU
0
WEST AGE
EOC
50
I
PHETCHABUN FOLDBELT
TECTONICS
km
237
EAST
I
[<
KHORAT
PLATEAU
q,
BASIN
Transtension and Uplift
~
KHORAT
~
MONOCLINE
VIENTIANE
(SAKHON
NAKHON)
BASIN
=~
NAM LEUK UPLIFT
PAL TERTIARY
EROSION
Flexural LATE
K 2 ~ . . . . . . . . . . . . . . . . .^.A
Subsidence
--
"'~
AA~
(J Collision? O W (5
EARLY Flexural Subsidence
I
~ K1p
,
~
K1Ns~_Ej~j_
-.
--
_
~
--
, --
,
~
--
--
--
]-
I
--
j
~ ~ --
.
.
-- -
--
- -
--
.
.
...~.._J_-I
--o--o
I
~6"-(~
--.._..= ...=
-- -
--
i ~
--
1
]
----
i 1
i
,
- -
o 0%
- -
--
.i
,i
Z
I
--
I
- -
--
DOGGERI
Quiescence
JURASSIC
HIATUS
LIAS
LATE L~
~
J PI Extension
t h r u st ~ ' . . _ . . - , . ~ . . . . ~ ~
-
MIDDLE
INDOSINIAN
OROGENY
Continental Collision EARLY
P~
LATE Active.,'
Passive LU
EARLY
Margin
o_
Fig. 4. Chronostratigraphic diagram for Vientiane area, Lao PDR.
The present authors prefer to place it in the Upper Cretaceous, based on the palynological evidence cited above. Neogene and Quaternary alluvial sediments occur locally, infilling abandoned river channels. In boreholes they reach thicknesses of up to 80 m. No older Cenozoic sediments are found in the Mesozoic basinal areas.
Current interpretations of postEarly Cretaceous inversion history There are currently two main interpretations of the post-Early Cretaceous structural history of the Khorat Plateau Basin, relating to the timing of inversion with respect to the deposition of the Upper Cretaceous Phon Hong Group. Cooper et al. (1989) interpreted the Khorat megasequences as being deposited in a basin formed by thermal subsidence during the Jurassic and Cretaceous, which was inverted prior to the deposition of the Maha Sarakham Formation evaporites. This inversion created a rimmed basin(s) within which the evaporites were deposited. The inversion was stated to be caused by the early stages of collision between India and Eurasia in the region of central Burma (Myanmar),
stress being fed to the Phetchabun Foldbelt via the Mae Ping strike-slip fault zone. They inferred a Late Cretaceous to Early Tertiary age for this inversion. Most other authors interpret the inversion as beginning after the deposition of the Phon Hong Group, during the Tertiary. Mouret et al. (1993) date the main inversion as beginning at c. 65 Ma (earliest Palaeocene). This is based on fission track data from outcrops in Thailand of the Phra Wihan Formation indicating that the maximum palaeotemperature, and therefore the onset of uplift, occurred at that time. These and other data are interpreted to demonstrate uplift and erosion of 3000 m for surface outcrops of Khorat Group sediments in the Phu Phra area of Thailand, which is on the northem margin of the Phu Phan Uplift. A similar age for the onset of inversion is also supported by Kozar et al. (1992).
Evidence of a two phase inversion history While our fission track data are in close agreement with those of Mouret et al. (1993) indicating that substantial uplift and erosion began at 60 Ma (midPalaeocene), other data show that this was the second phase of structuring to have affected the
238
P.F. LOVATT SMITH E T A L .
Khorat Group. The initial event occurred prior to the deposition of the Phon Hong Group in the midCretaceous and was responsible for regional tilting, compressive folding, reverse faulting and inversion of the basin subsidence pattern. The Tertiary event reactivated some of these trends but its main effect was to produce regional uplift and substantial erosion. This two phase model is consistent with the evidence published from Thailand mentioned above.~ The evidence in the study areas for this interpretation is listed below.
Unconformity at the base of the Phon Hong Group The Upper Cretaceous sedimentary cover has been more or less completely eroded in many areas of the Khorat Plateau due to Tertiary uplift, obscuring the evidence of mid-Cretaceous structuring. However an unconformity at the base of the Thangon (Maha Sarakham) Formation has previously been noted from seismic data in Thailand (Sattayarak et al. 1991). Fortunately, in the Vientiane area of Lao PDR. the preservation of a relatively thick section of Phon Hong Group sediments means that the character of the unconformity is readily discernible. The unconformity has been noted from shallow borehole data at the margin of the Vientiane Plain of Lao PDR by Ha Luong Tin & Do The Que (1991). A conglomerate containing clasts of siltstone, interpreted to be derived from the Khorat Group, has also been identified at the base of the Phon Hong Group at outcrop in the west of the Vientiane Basin (Howlett 1993). Seismic data from the Vientiane area show that this unconformity represents a more significant tectonic event than was previously realized. Four seismic lines (Figs 5-8) demonstrate this. Figure 5 shows the aspect of the Phon Hong Group towards the northern limit of its distribution. Horizontal Thangon Formation reflectors onlap uneroded Khorat Group both in the west (onto the Khorat Monocline) and in the east (at the base of the Nam Leuk Uplift). Although the shallowest reflectors may be confused by recent weathering effects it is apparent from these data that most of the Nam Leuk Uplift and Khorat Monocline structuring in this area occurred during the midCretaceous. The horizontal dip of the Phon Hong Group reflectors shows that subsequent Tertiary reactivation of these structures has been relatively minor. Figure 6 is a detail from a north-south line which intersects the line of Fig. 5. It demonstrates erosional truncation during the mid-Cretaceous of the upper Khorat Group reflectors on a reverse
fault-bounded horst. There is clear onlap by horizontal Thangon Formation reflectors over this feature. The reverse faults moved during or before the deposition of the Lower Salt Member of the Thangon Formation evaporites (AlbianCenomanian), since only the highest part of the Early Cretaceous horst has been eroded. The large reverse fault-bounded anticline seen on Fig. 7 is along a surface trend from the unfaulted Nam Leuk Uplift anticline seen in Fig. 5. The relief on both structures is of a similar order of between 1000 and 1400 m. The bulk of the reverse fault movement is therefore inferred to have occurred during the mid-Cretaceous event. It is interpreted that the movement occurred both before and during the deposition of the Thangon Formation, giving rise to a fault scarp and localized development of a syn-fault talus fan, against which the Thangon Formation evaporites were deposited. Some Tertiary reactivation is evident on this fault, shown by dipping Phon Hong Group reflectors on the upthrown side. Finally, Fig. 8 is a detail from the western margin of the Upper Cretaceous evaporite basin. Again, the shallowest horizontal reflectors in the Phon Hong Group may be confused by recent weathering effects but westward onlap at the base of the Phon Hong Group is strongly indicated. Together with the onlap seen along trend on Fig. 5 this indicates that the regional tilting of the Khorat Monocline was initiated in the mid-Cretaceous.
Sediment thickness and the distribution of halokinetic structures Figure 9 illustrates three seismic lines from the Vientiane Plain which demonstrate that an inverse thickness relationship exists in this area between the present day thicknesses of the Khorat Group and Phon Hong Group. The data show the presence of a complete section of Khorat Group sediments, thickening towards the northeast from 1900 m on line 1 to 4000 m on line 3. This demonstrates that during the Early Cretaceous there was an increase in the rate of subsidence and deposition from southeast to northwest. The subsidence pattern during the Late Cretaceous can be interpreted by examining the distribution of halokinetic structures in the Thangon Formation. The salt pillows and diapirs are assumed to have been generated by increasing overburden pressure of Upper Cretaceous sediments up to the time of maximum burial during the earliest Palaeocene. Line 1 shows that the most intense salt movement has occurred in the southeast, indicating that this was the area of thickest Upper Cretaceous palaeo-overburden. However
239
CRETACEOUS INVERSION IN THE KHORAT PLATEAU
TWO WAY TIME (SECONDS) LO o LO d -,~ ",~
TWO WAY TIME ( S E C O N D S ) ~'~ o Lo c5 ~" ',-
CO d ,
I
.
I
LL _J B=) III
,< Z ~f
-2
c4 =
o o ° ,,,,~
° o~
E
6 z o
g m
o
o
+
o T (W) H l d 3 0
co ~ 31Vl~lXOHddV
I-
o.
~
~
g
g r~
, (w) Hld'::lO 3 1 V I N I X O H d d V
240
E F. LOVATT SMITH ET AL.
TWO WAY TIME (SECONDS)
TWO WAY TIME (SECONDS)
~,
d
d
d
d
d I
d I
d I
d I
2
",,'ca_
~g
c4 d~
O
d
> d Z .~,
g, m
E •
g
g
g
g
r+
co '
i'.'
cq 7.
(LU) H l d 3 0 :llVl/~lXOklddV
g
I
I
~
~
g
,
,
,,7
(tu) H l d 3 a 31"v'lAIIXOt::lddV
I
241
CRETACEOUS INVERSION IN THE KHORAT PLATEAU
TwO wAY TiME (SECONDS)
TWO WAY TiME (SECONDS)
o o
o
o
o
o
d
04
~
qD
03
I
I
I
J
6
I
c5
c5
c5
I
c4
O O
O
6 Z O r~
I
+
cxl
,
(uu) H i d 9 ( ]
co
,
!
I
~
o
I
~_
:lLVl,gllXOk:lddV
(w) H/d=IQ =I/V~IXOEIddV ej . ,..,.
l.z
P 'D ~.
APPROXIMATE
I
I
_
--
DEPTH
I
I
,,.b
-h
(m) +
..~
APPROXIMATE DEPTH (m) ~ o
~" 3
o
I
I
I
+
o
I
I
o g\
Z 9
i
4~
_c
I
t~
I°
0
4~
1",3
~i!ii!ii~iiil
t:D
~,
I
I
t
J
o
o
o
9
S Q N O 0 : : : I S ) 31~lJ_ A V M O N L L
o
.o
.o
p
SGNOO3S) :::IIAII1 A r M O M /
YV , / 3 H.LIIAIS .LLVAO'-I "a "d
~
243
CRETACEOUS INVERSION IN THE KHORAT PLATEAU
Seismic Line I
i] km
E
VTN93-01 L o c a t i o n m a p of s e i s m i c L i n e s
Seismic Line 2
VIENTIANE BASIN, LAO P.D.R.
10
lkm
VTN93-04
Seismic Line 3 3000
2800
KEY Post - lower salt sediments (Phon Hong Group) Lower salt, Thangon Fm. (Phon Hong Group) Khorat Group
~Km
v i ~IU~
Fig. 9. Seismic montage showing relationship between Thangon salt diapirism and Khorat Group thickness.
244
P.F. LOVATT SMITH ET AL.
seismic lines 2 and 3 show that towards the northwest no salt movement has occurred, indicating a thinner Upper Cretaceous palaeo-overburden. The opposing thickness trends between Khorat and Phon Hong Groups are shown by this data to be subsidence-related and demonstrate a midCretaceous inversion of the subsidence pattern. It should be noted that in the Vientiane area gentle salt pillows are currently present at very shallow depths (Fig. 8), supporting fission track data which show that a substantial amount of Tertiary uplift and erosion occurred after maximum burial in the earliest Palaeocene. In the area of the Vientiane Plain the authors have calculated maximum Tertiary erosion of c. 1000 m from the character of the halokinetic structuring.
Abrupt facies change At the level of the mid-Cretaceous unconformity there is an abrupt change in facies from the Aptian Ban Thalat Formation of fluvial claystones, siltstones and sandstones to the Albian-Cenomanian Thangon Formation of restricted, continental evaporites. This facies change occurs throughout the Khorat Plateau Basin. It is difficult to envisage any other way that such a facies change could have been brought about other than by the effects of tectonism. These evaporites were suggested by Cooper et al. (1989) to have been deposited in restricted basins created by the topographic expression of syntectonic uplifts. Such potential drainage barriers are exemplified by the marginal uplifts visible on Figs 5-8. Three successively thinner evaporite cycles are present in the Phon Hong Group indicating three progressively more minor tectonic episodes.
these source rocks occurred during the Late Cretaceous burial. Thereafter only relatively minor quantities of hydrocarbons are likely to have been generated. Since trap formation nearly always pre-dates hydrocarbon generation these two timing constraints impose a mid-Cretaceous age of structural development for such gas-bearing traps, further supporting the evidence for mid-Cretaceous tectonism.
Lack of structural dip in Phon Hong Group At outcrop, the Phon Hong Group rarely dips at more than a few degrees, whereas the Khorat Group shows steeper dips and is even vertical or overturned adjacent to major faults. Although this could be explained by erosion of steeply-dipping Phon Hong Group sediments during the Tertiary, it is interpreted by us as supporting evidence for a mid-Cretaceous event.
Cretaceous granite intrusions K-Ar dating on granite intrusions from various parts of the Monument Contract Area gives ages of 78.3_+ 1.7 Ma (Campanian, Ban Boung Koang, No. 3 on Fig. 1), 94.9 + 2.1 Ma (Cenomanian, Keng Daeng, No. 1 on Fig. 1), l 1 7 + 3 . 0 M a (Aptian, Muang Kenthao, No. 2 on Fig. 1), and 137.2 +_3.0 Ma (Valanginian, Ban Phontiou, No. 4 on Fig. 1). Although no detailed geochemical analysis has been carried out, these granites are inferred to indicate crustal melting following the Indosinian continent-continent collision. The timing of their intrusion may be coincident with interruptions in the post-collision Cretaceous thermal subsidence of the Khorat Plateau Basin including, at c. 95 Ma, a mid-Cretaceous event.
Presence of trapped hydrocarbons The presence of thermally-derived gas in reservoirs within closed structural highs has been successfully demonstrated by many exploration wells in the Khorat Plateau of Thailand. Published seismic data over these structures show that they are defined by internally conformable, folded Khorat Group sediments (Kozar et al. 1992). This indicates that the structuring episode which formed these traps post-dates the Early Cretaceous. It can be assumed that the source rock for the gas lies in the pre-Cretaceous section since organicallyrich sediments within the Khorat and Phon Hong Groups are volumetrically insignificant. It is generally agreed that the most likely source rocks are within the Permian and Triassic section. The authors' burial history modelling indicates that the period of peak generation of hydrocarbons from
Fission track data from Thailand Data were obtained from seven outcrops of the Phra Wihan and Khok Kruat Formations on the Khorat Monocline in Thailand southwest of Khon Kaen. The data show a spread of ages all of which are younger than the depositional age (c. 145-112 Ma). This indicates that the fission tracks have experienced some annealing. Despite variations in mean age from 84 + 8 to 40 + 5 Ma, statistical modelling indicates that cooling began prior to 60 Ma and that the rock has experienced temperatures of around l l0°C. The model shows that the Cretaceous thermal history is poorly constrained because it has been overprinted by later heating. The dominant thermal event was at 60Ma, but there is no evidence to suggest that an episode of uplift did not occur in the mid-Cretaceous.
245
CRETACEOUS INVERSION IN THE KHORAT PLATEAU
Regional tectonic model The pre-Cretaceous tectonic history of the Indochina region is still not well understood. However there is agreement that a number of continental fragments including the Indochina Plate, on whose northern margin the Khorat Plateau Basin is situated, rifted from the northern margin of Gondwana during the Palaeozoic. These fragments drifted northwards, eventually accreting to the Eurasian margin during a series of continentcontinent collisions (Mouret 1994; Metcalfe 1996) of which the India-Eurasia collision is the latest. The timing of the collision of the Indochina Plate with its two neighbours, the South China Plate and the Shan Thai Plate still requires more research. Recent data from the Pak Lay Foldbelt show that tectonism occurred during the Late Jurassic (Stokes et al. 1996), indicating that the Shan-ThaiIndochina collision may have occurred later than previously thought. However it is evident that at the onset of Late Jurassic-Cretaceous basin subsidence in the Khorat Plateau and surrounding areas both of these sutures had already closed (Fig. 10). The Khorat Plateau and other continental sedimentary basins in the region (Fig. 1) may have formed as foreland basins as a result of flexural subsidence in front of the orogenic belts created by these sutures. Amongst the supporting evidence for this theory is that the depocentre of the Lower Cretaceous sedimentary basin in the Khorat Plateau runs parallel to the Song Ma-Song Da suture and Annamitic Foldbelt trend (N. Sattayarak, pers. comm. 1994) and that palaeocurrent data in the Khorat Group show a source from the Annamitic Foldbelt (Howlett 1993; Mouret et al. 1993). Also, there is no evidence of major syn-depositional faulting in the Khorat Plateau Basin, indicating that the stretching (13) factor was low. The onset of structuring and inversion of the Khorat Group as described in this paper can no longer be attributed to the collision of the India Plate with central Burma, as was proposed by Cooper et al. (1989), since this collision began during the Eocene (Daly et al. 1991). An alternative tectonic model must therefore be invoked. In the absence of a more convincing alternative, we suggest that mid-Cretaceous tectonism is best explained by a continental collision to the west, such as that of Western Burma with Shan-Thai which is stated by Metcalfe (1996) to have occurred in the Cretaceous. The constraining forces created by such a collision would have placed the entire region in compression, leading to the structural effects already described. The Nan suture runs parallel to the north-south Phetchabun Foldbelt trend, and the Song Ma-Song Da suture runs parallel to the northwest-southeast inversion trend
•
s c.,,A /
J
i
0
1000 km
Fig. 10. Pre-Cretaceous suture zones in SE Asia (after Cooper et al. 1989).
described above (Fig. 10). Both of these trends were compressionally reactivated during the mid-Cretaceous event in the study area. Subsidence resumed following this hiatus, initially in new evaporitic basins created by the inversion, and continued until the earliest Palaeocene. From this time onwards the Indochina area was subjected to a transtensional stress regime. Rifting occurred during the Oligo-Miocene both to the east (South China Sea) and west (Thailand Central Plain and Gulf of Thailand) of the Khorat Plateau (Daly et al. 1991), however there is no evidence of rifting in the Khorat Plateau itself. The evidence already cited indicates that the study area underwent some reactivation of mid-Cretaceous trends but that the main effect was to produce regional uplift and substantial erosion of much of the Upper Cretaceous sedimentary cover.
Implications for petroleum exploration Burial history models which assume a Tertiary age for the timing of post-Early Cretaceous inversion find it difficult to explain why, if trap formation post-dated the main period of hydrocarbon generation, most, if not all such inversion structures which have been tested by drilling have proven to contain gas. As discussed above the presence of hydrocarbons in structural traps defined at the base
246
P.F. LOVATT SMITH ET AL.
of the Khorat Group is much better explained by mid-Cretaceous timing of structural development. Tertiary reactivation of mid-Cretaceous structures has nevertheless enhanced the volume of some traps, which might explain why some structures are underfilled with gas (e.g. the Nam Phong gas field). The implication is that the petroleum exploration potential of the area is not only confirmed but may be greater than previously anticipated.
We thank Monument Resources Overseas Limited and Shlapak Development Company for permission to publish as well as colleagues in Monument for their help and advice. We also thank our numerous Lao and Thai colleagues, particularly in the Department of Geology and Mines, Ministry of Industry, Vientiane and the Mineral Fuels Division, Department of Mineral Resources, Bangkok, for their help, and the many Lao villagers for their hospitality during the fieldwork.
References COOPER, M. A., HERBERT, R. & HILL, G. S. 1989. The structural evolution of Triassic intermontane basins in northeastern Thailand. In" THANASUTHIPITAK,T. (ed.) Proceedings of International Symposium on Intermontane Basins: Geology and Resources, Chiang Mai. 231-242. DALY, M. C., COOPER,M. A., WILSON, ]., SMITH, D. G., HOOPER, B. G. D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine and Petroleum Geology, 8, 1-21. HA LUONG TIN & DO THE QUE. 1991. Determination of salt-bearing stratum depth and its structure in Vientiane region by geophysical data. In: TRAN VAN TRI (ed.) Proceedings Second Conference on Geology of lndochina, Hanoi. 2, 32-38. HOWLETT, P. 1993. Sedimentology of the Khorat Group. MSc Thesis, Birkbeck College, University of London. HUYNH XUAN DANG & BOUTHEUNG PHENGTHAVONGSA. 1989. Rock salt of the south of Laos. In: Geology of Kampuchea, Laos and Vietnam. Dept of Mines and Geology, Hanoi, 1, 28-30. KOZAR, M. G., CRANDALL, G. E, & HALL, S. F. 1992. Integrated structural and stratigraphic study of the Khorat basin, Rat Buri Limestone (Permian) Thailand. In: PIANCHAROEN,C. (ed.) Proceedings National Conference on Geologic Resources of Thailand: Potential for Future Development. Bangkok, 692-736. METCALFE,I. 1996. Pre-Cretaceous evolution of SE Asian terranes. This volume. MOURET, C. 1994. Geological history of northeastern Thailand since the Carboniferous. Relations with Indochina and Carboniferous to Early Cenozoic
evolution model. In: Proceedings International Symposium on Stratigraphic Correlation of SE Asia, Bangkok. 132-158. --, HEGGEMANN,H., GOUADAIN,J. & KRISADASIMA,S. 1993. Geological history of the siliciclastic Mesozoic strata of the Khorat Group in the Phu Phan Range area, Northeastern Thailand. In: THANASUTIPITAK,T. (ed.) Proceedings International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Palaeontology, Chiang Mai, 23-49. RACEY, A., GOODALL,J. G. S., LOVE, M. A., POLACHAY,S. & JONES, P. D. 1994. New age data for the Mesozoic Khorat Group of northeast Thailand. In: Proceedings International Symposium on Stratigraphic Correlation of SE Asia, Bangkok. 245-252. SATTAVARAK, N. 1983. Review of the continental Mesozoic stratigraphy of Thailand. In: NUTALYA,P. (ed.) Proceedings of Workshop on Stratigraphic Correlation of Thailand and Malaysia, Bangkok. 1, 127-140. --, POLACHAN, S. t~ CHARUSIRISAWAD, R. 1991. Cretaceous rocksalt in the northeastern part of Thailand. In: GEOSEA VII Abstracts. Bangkok, 36. STOKES, R. B., LOVATTSMITH,P. F. & SOUMPHONPHAKDY, K. 1996. Timing of the Shan-Thai-Indochina collision: new evidence from the Pak Lay Foldbelt of the Lao PDR. This volume. UTHA-AROON, C. 1993, Continental origin of the Maha Sarakham evaporites, northeastern Thailand. Journal of Southeast Asian Earth Sciences, 8, 193-207.
The 'Rajang accretionary prism' and 'Lupar Line' problem of Borneo CHARLES
S. H U T C H I S O N
c/o Department of Geology, University of Malaya, 59100 Kuala Lumpur, Malaysia Abstract: The Rajang Group, composed of the Belaga and Lupar Formations of Sarawak and the Embaluh Group and Selangkai Formation (in part) in Kalimantan, had a turbidite sedimentation history from Early Cretaceous to Late Eocene. These rocks generally young northwards. Inliers within the eastern Miri Zone have been mapped as the Kelalan and Mulu Formations. The Rajang Group was compressed into a steeply dipping quartz-veined phyllite-quartzite complex by the Sarawak orogeny and unconformably overlain locally in the north by the Upper Eocene continental to neritic Tatau Formation, extensively in the south by Middle to Upper Eocene basal continental sequences of the Ketungau, Mandai and Melawi basins, and widely in the north by the Upper Oligocene coastal to marine Nyalau Formation. In Sabah and East Kalimantan, the Upper Cretaceous to Upper Eocene Mentarang, Sapulut, Trusmadi, and possibly the East Crocker Formations, also belong to the Rajang Group. The West Crocker Formation demonstrates rapid facies changes into the more shaly Temburong Formation, and was deposited as sandy turbidites throughout the Oligocene. To the south their equivalents are the nearshore Kelabit and Long Bawang Formations. The West Crocker Formation was folded and uplifted in several Miocene pulses, resulting in regional unconformities and igneous events at Mount Kinabalu. The West Crocker Formation has not been metamorphosed, and dips are shallower than in the Belaga and Trusmadi Formations. Its provenance probably was from the uplifted Upper Cretaceous to Eocene Lurah and Kelalan formations of NE Kalimantan and East Sarawak. It is therefore proposed that the West Crocker and Temburong formations be excluded from the Rajang Group. Middle to Late Miocene Crocker Formation uplift and deformation, herein called the Sabah orogeny, was synchronous with spectacular basin inversion throughout the South China Sea and in the Meratus Mountains of Kalimantan. Uplift ceased in the Late Miocene and undeformed post-inversion formations unconformably overlie inverted folded structures. The Rajang Group flysch-belt may be interpreted as a north-facing accretionary prism. The Schwaner Mountains represent a subducfion-related Lower to Upper Cretaceous volcanoplutonic arc. Scattered Eocene volcanism, Miocene Sintang intrusives and Pliocene Metalung volcanics in Central Kalimantan and Sarawak post-date subduction. The Rajang Basin (Danau Sea) rapidly narrowed and by Eocene time subduction was transformed to collision as the Rajang Group was compressed between the Schwaner Mountain Zone and the Luconia-Balingian-Miri block. The Ketungau Basin is in sharp contact with the Rajang Group along the bounding Lupar Fault, which can be traced northwards into the East Natuna region. Palaeocurrents show that the Upper Eocene basal sandstones have a provenance in the metamorphosed Sibu Zone. The Melawi and Mandai basins of Kalimantan also unconformably overlie the flysch-belt, The basins are not forearc and were formed after transformation of the accretionary prism to a landmass formed of a collisional orogenic complex.
The models of tectonic evolution of the South China Sea region conflict with the d o c u m e n t e d geology of Borneo and especially with the Rajang Group. These models have then been applied to analyze the tectonic evolution of Borneo. The m o d e l of Taylor & Hayes (1980, 1983) wrongly shows active subduction of Mesozoic oceanic lithosphere continuously through the Palaeogene, b e c o m i n g extinct by the Middle Miocene when the Luconia Shoals were interpreted to have collided with the Northwest Borneo Trough. The model o f Holloway (1981) is not significantly different. The mistake can be traced back to Hamilton (1979), w h o described extensive tracts o f B o r n e o as
'm61ange', interpreted as s y n o n y m o u s with subduction-related accretionary wedges. The m o d e l of Ru & Pigott (1986) ignores the West Baram Line and wrongly extends the Northwest Borneo (Palawan) Trough through the Luconia and Balingian provinces (their fig. 1), and has active Late Eocene, southeastwards subduction. The model o f James (1984) perpetuated this mistake by not closing in the Late Eocene the Rajang Basin, or Danau Sea of Haile (1994). These models wrongly concur that subduction continued until the Early Miocene. The mistake of continuing the Northwest Borneo (Palawan) Trough along the southern margin o f the Luconia Province, ignoring
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, Geological Society Special Publication No. 106, pp. 247-261.
247
248
C.S. HUTCHISON
the West Baram Line, is also made by Tongkul (1990). All these models are flawed because they also ignore the peripheral, yet equally important region. Thus, Ru & Pigott (1986) and Williams et al. (1988) accept the palaeomagnetic evidence that West Borneo has rotated anti-clockwise (Schmidtke et al. 1990), but they ignore the evidence for clockwise rotation of Indochina (Achache et al. 1983). Williams et al. (1988) bring peninsular Malaysia and Sumatra together only in the Early Tertiary; an impossibility because they share the same Palaeozoic and Mesozoic stratigraphy and structure without displacement (Hutchison 1993). The Mid Oligocene to Early Miocene N-S extensional opening of the east sub-basin of the South China Sea, first detailed by Taylor & Hayes (1980), has been widely accepted and forms the basis of most subsequent models of the region, including contiguous Borneo. Using the data of Taylor & Hayes (1980, 1983) and Chen (1987), Briais et al. (1993) have produced a refined scenario of South China Sea spreading history, composed of three episodes: NW-SE extension from 32-30 Ma, followed by N-S extension from 30-26 Ma, then by NW-SE extension from 2316 Ma. The structural trends for the latest, Early Miocene, episode have been resolved in the Scarborough Seamounts region by Pautot et al. (1986) and Briais et al. (1989) to be 50" (magnetic anomalies) and 140° (transform faults). The choice of magnetic anomaly designations for the main eastern basin was influenced by the only direct geological constraint of volcanics dredged from the Scarborough Seamounts, radiometrically dated 10-15 Ma (Hrkinian et al. 1989). The volcanics are thought to have been extruded along the still weak relict spreading axis. Briais et al. (1993) admit that their interpretation is not unique. Some of their illustrated anomaly designations do not inspire confidence. Their scenario should prevail until the ODP programme can include the South China Sea Basin. The practice of designating anomalies without direct geological constraint has led to serious misinterpretations in the nearby Sulu Sea (Lee & McCabe 1986; Hutchison 1992a). More carefully constructed evolutionary scenarios of the Borneo margin of the South China Sea were painted by Tan & Lamy (1990) and Hazebroek & Tan (1993), which benefited from comprehensive petroleum company data. However their inferred subduction history from Late Cretaceous through Late Eocene, using the Lupar and Mersing lines, and from Late Eocene through Middle Miocene, using the Northwest Borneo (NW Sabah) trough, is not matched by identified subduction-related post Palaeocene volcanic arcs in Borneo.
The Rajang Group of Sarawak The Rajang Group consists of the Belaga and Lupar formations, which are isoclinally folded, with dips usually between 80 and 90° (Fig. 1). It extends across the border into Kalimantan as the Embaluh Group (Tare 1991). This monotonous flysch group is of interbedded sandstone and mudstone, is locally metamorphosed to lower greenschist facies phyllite and slate, and commonly veined by quartz (Kirk 1957; Liechti et al. 1960; Wolfenden 1960). The outcrop width of the group perpendicular to the E-W strike from the Lupar Line to the Tatau Horst is 200 km. Thickness cannot be determined and has been guessed to be in the neighbourhood of 15 km (Haile 1974), but the basis for his estimate is unknown. The Rajang Group constitutes the Sibu Zone of Sarawak (Fig. 1), outlined along the south by the Lupar Line ophiolite and along the north by the Bukit Mersing Line ophiolite (Hutchison 1989), but it also crops out north of this in the Tatau Horst and as various inliers within the Miri Zone, as the Kelalan and Mulu Formations. The Lupar Line ophiolite occurs as blocks in a 20-25 km wide mrlange zone, known as the Lubok Antu Mrlange. Along its SW side, the mrlange is in sharp contact with the basal Silantek Formation beds of the Ketungau Basin along the steep linear Lupar Fault. The Rajang Group has been interpreted as a deep marine fan deposit. Most outcrops have thin fine to medium grained sand layers which may be described as distal turbidite, but some outcrops show thicker sandstones. Most are characterized by prominent sole marks. Adjacent to the Lubok Antu Mrlange, the 5 km wide outcrop of the Lupar Formation is predominantly a turbidite sequence similar to the contiguous Layar Member of the Belaga Formation (Liechti et al. 1960; Tan 1979). It was mapped separately because it incorporates large bodies of ophiolitic rocks. Microfossils indicate an Upper Cretaceous age of Santonian to Maastrichtian. The Lupar Formation thus represents the oldest part of the Rajang Group of Sarawak (Table 1). From the south, adjacent to the Lupar Formation, to the Tatau Horst in the north, the Belaga Formation has been subdivided into the following members: Layar, Kapit, Pelagus, Metah and Bawang. The fossils indicate a progressive younging northwards from Upper Cretaceous Layar to Upper Eocene Bawang. The steep dips and the systematic northwards younging and ophiolite and mrlange association have led to the plausible interpretation of a northwards facing accretionary prism, related to Late Cretaceous (85 Ma) to Late Eocene (45 Ma) southwards subduction beneath the Kuching Zone of Sarawak of the Rajang Basin, named the Danau Sea by Haile (1994).
,~ ~
(~'"
/f~l
• PONTIANAK
E~MBALUH~
MELAWl
- " • e ,.BM .
MOUNTAINS
".
f....
•
:
•
%+
NIAH
-
M
-
+++
+
• WEST KUTEI
IVy'4
~'+~--~
--------
--~-÷' ' ~!%-!
//
~,,,,3M
+BRUNEI%
LABUANCC3
~
KOTA KINABALU
SEEANGKAI --'"-' • • -.~__FM_~-") . . "UPI~ER . . . .
L-f KAPUAE~_.GROUP
~. ~
SE
SCHWANER
ROU.
+ "- -------- ~_ + ~--~--+-.
• .. . . . . . . . . . .
O'~Z'-
~\,.
I
114"
MERIT'-~ ~ -PILA~_.J
'---'----..~___~--~(SIBU ZONE)
_~ ~----___--------
PENIAN
BINTULU
KE : KELALAN Fm
M = MULU Fm
K = KAYAN SANDSTONE
BM : BOYAN MI~LANGE
TRS : TRUSMADI Fm
BF = BELAGA Fm
i
~
I ~)PEDAWAN~
(KUCHING ZONE)
~b
"L/~
I
f
I
I
km
I00
0
200
Jurassic- Cretaceous PEDAWAN Fm ( in Sorowok)
orogenic flysch complex (RAJANG 8 EMBALUH GROUPS SELANGKAI Fm)
Lower Cretaceous to Upper Eocene
I
112"
!
116"
FM
°.
= SERANTAK VOLCS. = PI.YABUNG TUFF = MULLER VOLCS. = NYAAN VOLCS. = ARIP VOLCS.
Eocene P o s t Subduction Volcanism
International + + Border
I 2 3 4 5
5
SABAH
-5
Fig. 1. Sketch geological map of Sarawak, Sabah and contiguous Ka]imantan (Borneo) to emphasize the Rajang Group. The eastern Mid Zone contains several inliers of Kelalan and Mulu Formations, interpreted to be equivalent to the Rajang Group. The Sibu Zone contains several large Pliocene volcanic massifs. Within the Embaluh Group outcrop they are known as the Metalung Volcanics. Based on Tan (1982), Lim (1985), Pieters et al. (1987) and Pieters & Supriatna (1989).
-30
Iio°
Oligocene to Lower Miocene Crocker- Temburong flysch
continental basins
Tertiary, predominantly
I
t,O
©
~0 ©
E
t"
>
©
250
c . s . HUTCHISON
Table 1. Chronology of events discussed in the text Schwaner Post-subduction Rajang Group Rajang Group N. W. Sabah South China Sea Mountains Volcanism deposition deposition Crocker Ranges volcano& plutonism Sarawak& Sabah & plutonic arc W. Kalimantan E. Kalimantan 5 ..Plioeene ....................... Metalung Voles.................................................................................................... Post-inversion strata ................. 10 upper Mr. Kinabalu Shallow Regional 15 middle Deep Regional ~ SABAH OROGENYu" 20 Miocene Sintang 25 .... lower ............................. intrusives .... Deposition of the Melawi ........................................................ Deposition of the .................... 30 Oligocene Ketungau and Mandai Kelabit Fin. [W. Crocker Fm ] South China Sea 35 ..................................................................... basins ..................................... [ including the ] .............. basins .......................... 40 MiHler Tatau Fnt Batu Gadlng [ Temburong Fm l 45 Eocene ~ S A R A W A K OROGENY ~ ~ y ~ . 50 Nyaan, Piyubung /lk /l~ [ E. Crocker Fro. ]
Ma
55
NERIE,~
......................................................................
60 Palaeocene (60 ........................... 70 75 Upper 80 Cretaceous 85 90
:~:. ..............................................
"t'" .............. ~ ........ ~ ....... ~ .......................................................
Ma) ? Kayan Sstn ? 0llelaga Fro] 1 Trusmadi Fm V ...................................... [Lupar F m ] / ~ ........... [ Sapulut Fm l ............... I"................................................................... [Embaluh Group] [Mentarang F m ] |
G [potassic granite]
]
P A~ I 95 ........................... P ..................................... . ~ ........................... :4:" .................................................................................................................. loo r / / 105 Lower P 110 Cretaceous P 115 P 120 ........................... P ...................................... 125.. Valanginian .................................................
| |- -
[
. ]-{Pedawan Fm] ...... I............................................. .[... ..........................................................................
130
i
/ ~
/[Selaaghal Fm]""l
. 1/~ end ~ u n c o n f o r m i t y I l time V = volcanism [ [duration P=plutonism [ l o f turbidite G = granite
I | sedimentati°n
,
[ I [ ]
I
135 ..........................................................................................................................................................................................................................
Palaeocurrents in the Lupar Formation turbiditic sandstones indicate a major current direction towards 50-60 °, with a subordinate maximum towards 80-90 ° (Tan 1979). The provenance of the earliest part of the Rajang Group turbiditic sandstones was from the Kuching Zone, lying to the SW and W. Most of this zone formed an integral part of the extensive Late Cretaceous landmass known as Sundaland, upon which continental Eocene formations were widely deposited (Hutchison 1992b). The fluvial system, which brought sediments from the interior of Sundaland, must have reached the Danau Sea in the neighbourhood of what is now the Lupar Line. The geological record favours a narrow continental shelf as shown by outcrops of the Jurassic-Cretaceous Bau Limestone rapidly giving way to turbiditic Pedawan Formation of the Kuching Zone (Tang 1992), so it may be inferred that there was a rapid descent from land in the Kuching Zone to deep marine conditions even before the Lupar Line was reached. The lower, Upper Cretaceous, part of the continental Kayan Sandstone (Tan 1986) may be the on-land equivalent of the deep marine Lupar Formation and Layar Member of the Belaga Formation (Fig. 1). However, Tate (1991) and R. J. Morley (pers. comm. 1993) have expressed doubt about the older age palynological determinations by Mtiller (in Tan 1986) of the Kayan Sandstone and regard it as Upper Eocene.
Relationship to younger formations
One of the most spectacular Malaysian angular unconformities is exposed on the east end of the Tatau Horst (Wolfenden 1960; Halle & Ho 1991). The gently dipping basal conglomerate of the Tatau Formation rests on steeply dipping and complexly folded Upper Eocene Bawang Member turbidites of the Belaga Formation. The unconformity is palaeontologically dated as Upper Eocene. The Tatau Formation is an Upper Eocene to Oligocene sequence of conglomerate, sandstone and shale. It is continental to fluvio-marine, becoming neritic upwards, and includes the sub-aerial Arip Volcanics. From this informative outcrop, it may be inferred that the Danau Sea was eliminated and deformation, metamorphism,-uplift and peneplanation of the Belaga Formation to form a landmass was completed before the end of the Eocene. The author proposes to name this the Sarawak orogeny (Table 1). This information invalidates the tectonic model of James (1984), reproduced by Hutchison (1989). An equally spectacular unconformity occurs at the Bukit Besungai area of Batu Gading (Fig. 1) in the Middle Baram region (Adams & Haak 1962). Highly fossiliferous Upper Eocene limestone overlies with spectacular angular unconformity an Upper Cretaceous steeply dipping sequence of turbiditic Kelalan Formation, the eastern equivalent
RAJANG GROUP AND LUPAR LINE, BORNEO of the Belaga Formation (Haile 1962). This quarried location along the upper reaches of the Baram River (Fig. 1) is of identical significance to the Tatau Horst locality. The Tatau Formation is overlain by a thin (550 m) shale-sandstone sequence known as the Buan Formation (Wolfenden 1960). However, the major transgression over the landmass was accomplished by the Upper Oligocene to Lower Miocene Nyalau Formation, estimated to be about 5 km thick. It rests conformably on the Buan Formation and with marked unconformity on the Belaga Formation. The Nyalau Formation forms many outliers over the main Belaga Formation outcrops of the Sibu Zone (the largest at Merit Pila is shown on Fig. 1) and the unconformity is documented by Kirk (1957) and Wolfenden (1960). This is a very sandy formation in the west around Bintulu, becoming more muddy eastwards, where it is known as the Setap Shale. Detailed analysis by Shell indicates that a western landmass, known as the Penian High (Fig. 1), progressively gave way eastwards to a coastal plain rich in coal beds, then an inner neritic continental shelf. The outer shelf was characterized by a narrow NNW-SSE trending carbonate platform (Subis Limestone), seen at Niah. The edge of the shelf was marked by a continental slope, known as the West Baram Line, east of which bathyal conditions prevailed in a region which became the Baram Delta in the Late Miocene.
Southwards extension into Kalimantan An understanding of the Rajang Group is impossible without knowledge of its southwards extension, perpendicular to strike, into Kalimantan. The Lupar Formation and the Layar and Kapit Members of the Belaga Formation are known in northern Kalimantan as the Upper Cretaceous to Eocene Embaluh Group (Fig. 1; Table 1), which is not sub-divided (Tate 1991). This flysch sequence is strongly deformed. Bedding is in places overturned, and two fold phases have been identified. East of the Kapuas lakes, the main fold axis orientation is horizontal, trending 280 ° . The folding is asymmetrical, with southerly dips predominant (Williams et al. 1988). However, overall stratigraphic younging is towards the north, as in Sarawak. South-dipping thrust faults have been identified within the Embaluh Group. Southern outcrops along the Upper Mahakam River are reported to be unmetamorphosed (S. J. Moss pers. comm. 1995). The Embaluh Group is succeeded southwards in the Sintang District (Williams & Heryanto 1986) by the Lower Cretaceous (Valanginian) to Upper
251
Cretaceous (Turonian) Selangkai Formation, which youngs northwards and was first described by Zeijlmans van Emmichoven (1939). It has been restudied by Williams & Heryanto (1986), who described it as predominantly of calcareous mudstone with intercalations of pebbly and bouldery mudstone, graded sandstone and rare limestone and conglomerate. However, the formation is lithologically variable and some stratigraphic sections are dominated by turbiditic sandstone or other mass flow deposits. The structural and stratigraphic complexity has not been resolved, but Zeijlmans van Emmichoven (1939) suggested a formation thickness in excess of 3 km. The Selangkai Formation occupies a 45 km wide belt in which the main fold axes and outcrop trend WNW, with some cross folding and some strongly cleaved and thrust faulted zones. The lower part of the formation is in fault contact with the Boyan Mrlange, where it is tightly folded and cleaved. The dominant calcareous mudstone is a dark grey strongly sheared rock. It commonly contains, large blocks of fossiliferous limestone. Williams & Heryanto (1986) interpreted the sandstone beds, conglomerate and calcareous mudstone containing dispersed boulders as mass flow deposits. The sandstone beds have prominent sole marks and are turbiditic. The .-general appearance of the Selangkai Formation is that of a deep basin submarine fan into which blocks of limestone have been transported from a more stable shelf. Abundant volcanic detritus indicates a nearby contemporaneous volcanic arc, probably in the Schwaner Mountains (Williams & Heryanto 1986). The Upper Cretaceous (Turonian) Ingar Formation is in fault contact with the Selangkai Formation and both are folded and unconformably overlain by the Dangkan Sandstone, which forms the lower (Upper Eocene?) sequence of the Melawi Basin (Williams & Heryanto 1986). Unlike the Selangkai Formation, the Ingar Formation is predominantly muddy with an absence of turbiditic sandstone. It seems reasonable to conclude that the Danau Sea deep water fill is of Valanginian age in the south (Selangkai Formation), younging northwards through the Embaluh Group (= Lupar Formation and lower Belaga Formation), and eventually reaching its youngest (Late Eocene) age near the Mersing Line at the Tatau Horst. Nevertheless, the picture of a general northwards younging may be a gross oversimplification of this poorly known terrain. The Embaluh Group of the Upper Mahakam and Boh rivers of Kalimantan has yielded Middle Eocene planktonic foraminifera (Carter & Yeats 1973). The Selangkai Formation is similar both in its partial turbiditic character and age to the Pedawan Formation, which lies west across the Lupar Fault in the Kuching Zone.
252
C.S. HUTCHISON
The Lupar Fault appears from the Sarawak viewpoint to represent the most southerly termination of turbiditic rocks. However, the flysch belt extends southwards into the Sintang region of Kalimantan. The Lupar Line has therefore been given undue importance. It is the prominent fault margin between the Upper Eocene Silantek Formation of the Ketungan Basin and the Sibu Zone flysch belt. Indeed, within the Kuching Zone of Sarawak, there are abundant Lower to Upper Cretaceous turbidites known as the Pedawan Formation lying W and SW of the Lupar Line (Ting 1992). Ophiolitic m61ange zones are not confined to the Lupar Line. They also occur along the Bukit Mersing Line, and along the Semitau Ridge, known as the Boyan M61ange, although clasts in the latter are predominantly of quartz-rich sedimentary and metamorphic rocks (Williams et al. 1988). All of these m61ange belts lie close to and parallel to major faults which bound superimposed Tertiary basins. The Semitau Ridge margins are characterized by north dipping thrusts post-dated by Upper Oligocene sedimentary rocks.
the convexity would favour subduction from the south. The ophiolitic rocks of the Meratus Mountains have been dated as Lower Aptian (116 Ma) by Sikumbang (1986) and are associated with calcareous mudstone, sandstone and limestone. These ophiolites would seem to be ideal candidates to be paired to the Early Cretaceous volcanoplutonic arc of the Schwaner Mountains (Fig. 1). However, Sikumbang (1986) considered that there was no subduction relationship between them, and he concluded that the Meratus Mountains contain the record of a compressed marginal basin and paired volcanic arc. The Meratus ophiolite formed part of the Late Cretaceous landmass upon which diamond-bearing serpentinite conglomerate (the Pamali Breccia) was fluvially deposited (Bergman et al. 1987). However, the Meratus did not uplift to form the present day mountains until the Late Miocene (Hutchison 1992b). Before the uplift, the now separate Barito, Asem-Asem and Pasir basins were once one and the same. The Schwaner Mountains
Search for the paired volcano-plutonic arc Most authors (e.g. Hamilton 1979; James 1984) presume that the Upper Cretaceous through Upper Eocene Rajang Group of Sarawak (now modified to begin in the Lower Cretaceous by including Kalimantan) represents a northwards-facing accretionary prism, but there has been no clear identification of an appropriately positioned contemporaneous subduction-related volcano-plutonic arc. Gower (1990) concluded that the data from Borneo 'support the postulate of southeasterlydirected subduction of the South China Sea beneath Borneo during the early Tertiary.' At the same time he admitted that 'there are no convincing reports of early Tertiary magmatism, coeval with the postulated subduction.' The following discussion may clarify the situation of the appropriately distanced terrain lying 200-250 km to the south.
Convexity o f the arcs
The island of Borneo is dominated by a framework of arcuate structures, all convex towards the south and southwest (Fig. 1). This feature was emphasized by the earlier Dutch workers (Van Bemmelen 1949). Hamilton (1979) noted the importance of the Schwaner Mountains volcano-plutonic arc thus 'Presumably these igneous rocks are paired to the Cretaceous m61ange complexes that occur in both northwestern and southeastern Borneo'. He favoured subduction from the north, even though
The rigid 'core' of Borneo is formed by the Schwaner Mountains. Their northern boundary forms the southern margin of the Tertiary Melawi Basin (Fig. 1), and extends inland along an E-ESE trend, which cuts the equator east of Pontianak. The Schwaner Mountains extend eastwards more than 4 ° of longitude to disappear beneath the Cenozoic NW Kutei Basin, and they extend over 2 ° of latitude southwards to disappear beneath the Quaternary sediments of the coastal plains, which are underlain by the Billiton and Barito basins of SE Kalimantan. The region was divided by Van Bemmelen (1949) by an approximately E-W line along latitude l°S into the 'Schwaner Zone' and a 'southern region'. However, subsequent mapping (Amiruddin & Trail 1989; Pieters & Supriatna 1989) has shown that both zones form part of the Schwaner batholith, of which a major part is the Sepauk tonalite, with the Sukadana granite of the Ketapang batholith occurring in the SW corner. The Schwaner Zone, east of Pontianak, is mostly of 1-type (subduction-related) calc-alkaline plutonic rocks ranging from granodiorite to gabbro, dominated by the Sepauk tonalite, associated with a variety of metamorphic rocks mapped by Pieters & Supriatna (1989) as Palaeozoic, probably Carboniferous to Permian, though their age is indefinite. The 'southern region', lying east of Ketapang, is composed of Jurassic-Cretaceous flysch and contact metamorphosed limestone (the Ketapang Complex of Van Bemmelen 1949), and volcanic rocks (the Matan Complex of Van
RAJANG GROUP AND LUPAR LINE, BORNEO
Bemmelen), intruded by granitic batholiths (Pieters & Supriatna 1989). The Bunga basalt and Kerabai volcanics, formerly included in the 'Matan Complex', range from basalt, through andesite, to dacite and rhyolite, and represent a mature volcanic arc, which Tate (1991) paired with the Upper Cretaceous part of the Rajang Group, representing southwards subduction. The plutonic complex of the Schwaner Zone (Sepauk tonalite) has been dated by the whole-rock K-Ar method by Williams et al. (1988) and Bladon et al. (1989) at 104 (Albian) to 123 Ma (late Valanginian). Earlier, McDougall (in Haile et al. 1977), using biotite and hornblende mineral separates, determined K-At ages ranging from 75 (Campanian) to l l 5 M a (Barremian). The two datasets overlap and extend over the Lower and Upper Cretaceous. Williams et aI. (1988) report only 4 K-Ar dates from the potassic Sukadana granite of the 'southern region', lying within the range 86-91 Ma. However McDougall (in Haile et al. 1977) reported 10 mineral separate K-Ar dates from this province, eight of which lie within the range 78-86 Ma, but two granites gave considerably older ages of 127 and 153 Ma. Therefore this province contains important Late Cretaceous plutonic activity, dominated by alkaline granite, but there also occur some Lower Cretaceous to Upper Jurassic granitoids. The volcanic rocks of the 'southern region' were considered too weathered by Haile et al. (1977) for accurate dating. However, Pieters & Supriatna (1989) reported Upper Cretaceous K-Ar ages of 65-75 Ma (Maastrichtian-Campanian). In summary, most of the plutonic rocks of the Schwaner Mountains may be paired to the Selangkai Formation, which contains abundant volcanic detritus. Volcanism and high-level plutonism continued at least into the Maastrichtian, when the Embaluh Group, Lupar Formation and Layar Member of the Belaga Formation were being deposited. The Cretaceous part of the Rajang Group may therefore be reasonably interpreted as a northwards-facing accretionary prism, paired to the Cretaceous Schwaner Mountains volcano-plutonic arc (Table 1 and Fig. 2).
N o r t h o f the S c h w a n e r M o u n t a i n s
A narrow E - W trending zone through the Upper Kapuas region of central Kalimantan, with sporadic occurrences of Early Eocene acid volcanic and pyroclastic rocks (Fig. 1), of restricted 48-50 Ma age, has been interpreted by several authors to post-date subduction and be related to the early rift stages of formation of the Ketungau, Mandai, Melawi and West Kutei basins (Tate 1991; Doutch
253
1992; Van de Weerd & Armin 1992). They can be demonstrated to post-date subduction, because in several localities the Eocene volcanics unconformably overlie either the Boyan M61ange or the Embaluh Group. The Piyabung Volcanics, of gently dipping subaerial lithic and vitric tuff and agglomerate up to 200 m thick, extend 40 km E-W from 112 to 112 ° 30' within the Semitau Ridge at 0 ° 18'N (Pieters et al. 1987). A K-Ar whole rock age of 49.9 was obtained (Bladon et al. 1989). The Nyaan Volcanics, of 200-900 m of acid to intermediate tuff, agglomerate, ignimbrite and dacite, occur much farther east in the Mahakam area of the West Kutei Basin, at 114 ° 50% and 0 ° 50'N. They gave a K-Ar date of 48.6 Ma (Tate 1991) and 50 ___2.5 Ma (Pieters et al. 1987). In the Kelian district of the Upper Kutei, Van Leeuwen et al. (1990) describe Upper Eocene rhyolitic volcanic and pyroclastic rocks in a region dominated by Miocene volcanism. The Eocene igneous rocks are exclusively silicic, and have not been demonstrated to be subduction-related. In the same general region, Wain & Berod (1989) show Upper Cretaceous to Eocene volcanic rocks on their maps, but do not describe them. Near Singkawang, far to the west near the west coast at l°N, the Bawang Dacite gave a K-At date of 51.3 Ma. It is part of the Serantak Volcanics, which are less than 300 m thick. A much more extensive body of intermediate to basic volcanics lies entirely within the Mandai Basin in the Mtiller Mountains at 113 ° 20'E, 0 ° 30'N, K-Ar dated 40.9 Ma (Pieters et al. 1987). These intermediate to basic tufts and flows, many hundreds of metres thick, are interbedded with the basal sandstones along the Mandai River. The volcanic rocks were first documented by Molengraaff (1902). Zeijlmans van Emmichoven (1939) confirmed the existence of volcanic rocks, predominantly of basic-intermediate to basic composition in the region of the Mandai Palaeogene basin east of the Lake District (Kapuas), equivalent to the Mtiller Mountains of Molengraaff (1902). On the north side of the Rajang Group, the acidic Arip Volcanics and associated Piring Granodiorite of the Tatau Formation have not been dated but post-date the metamorphism of the Rajang Group. They are comparable to volcanics described above from the southern side of the orogenic flysch belt. It may therefore be concluded that there is no Eocene subduction-related arc to the north of the Schwaner Mountains, and that the predominantly silicic and pyroclastic eruptions are an integral part of the basin extensional process. It is therefore doubtful if the term accretionary prism should be applied to the whole of the Rajang Group, for only its Cretaceous part can be paired to a volcanic arc,
254
c . s . HUTCHISON
within the Schwaner Mountains. The PalaeoceneEocene part should not be classified as an accretionary prism, for subduction-related volcanoplutonic activity had ceased, and the situation was one of collision (Fig. 2). A province of abundant small epizonal stocks, sills and dykes extends from the Kuching Zone throughout the Ketungau Basin, with maximum development within the eastern Ketungau Basin (Williams & Heryanto 1986). They have been named the Sintang Intrusives (Williams & Harahap 1987). They range from pyroxene andesite to microgranite and are I-type. K-Ar dates from Sintang range from 30--23 Ma in the south and 18-16 Ma in the north. The stocks preferentially intrude the basinal strata, and have metamorphosed the Silantek Formation coals to anthracite (Tan 1979). Volcanic rocks overlie the Belaga Formation in the Usun Apau, Hose and other massifs (Kirk 1968). They are equated with the Metalung Volcanics, which overlie the Embaluh Group of Kalimantan (Doutch 1992). Although poorly studied, they appear to be Pliocene for they overlie outliers of the Nyalau Formation. In Kalimantan, they have been dated by the K-Ar method within the range 1.6-8.2 Ma (Bladon et al. 1989). They have been extrapolated much farther east, where they intrude the Mentarang Formation (BRGM 1982), and have yielded similar dates. However K-Ar ages as old as 25 Ma may not be from equivalents of the Metalung volcanics.
Ketungau Basin In Sarawak, the basal formation of the Ketungau Basin is known as the Silantek Formation. Only the lower part, which outcrops adjacent to the Lupar Line, can be dated by fossil content, but the age range is generally considered to be Upper Eocene to Miocene. The Upper Eocene Basal Sandstone Member outcrops on the Marup Ridge and is steeply dipping (Tan 1979). The strata flatten out southwards away from the Lupar Fault. The average down-dip palaeocurrent direction of the lower member is 228 ° from ripple marks and 194 ° from cross bedding (Tan 1979) indicating that the provenance was from the northeast. This is consistent with the petrology of the conglomerate and sandstone, which contain clasts of quartzite, schist, meta-greywacke and mafic volcanics, indicating provenance from uplifted and eroding Lupar Formation and Layar Member of the Belaga Formation. The base of the lower member is nearshore to shallow marine, but the Silantek Formation rapidly becomes continental upwards and has red-beds in the upper part, followed by the thick flat lying continental cross bedded Plateau
Sandstone. The Sibu Zone was therefore in the process of inverting to form a landmass during the early stages of Silantek Formation deposition. The steep dips of the lower Silantek Formation adjacent to the Lupar Line imply that the Rajang Group orogenic belt continued to uplift after deposition of the basal Silantek Formation, resulting in the boundary becoming the Lupar Fault. The rapid elimination of marine conditions from the Silantek Formation indicates continuing inversion of the Rajang Group flysch-belt. Accordingly it is wrong to refer to the Silantek Formation (Ketungau Basin) as forearc (Williams et al. 1988), since the Rajang basin had been eliminated and the Sibu Zone transformed from an accretionary prism into an orogenic belt welding together the Schwaner Mountains Zone of Sundaland onto the Balingian-Luconia Province (Miri Zone) (Fig. 2). The Rajang basin, or Danau Sea, was not related to the younger South China Sea and should not be referred to as the Proto South China Sea (Haile 1994). The Lupar Fault is the bounding margin of the Ketungau Basin, which developed by extensional tectonics affecting the Rajang Group orogenic belt. Studies to the north in the East Natuna area have shown that a similar situation persists, where Tertiary basins are unconformably superimposed on phyllitic basement (White & Wing 1978), which can be equated to the Rajang Group. The basin margins likewise commonly form prominent faults, similar to the Lupar Fault, but the trend has been bent into a N-S alignment northwards from the coast of Sarawak. During their early development, the Ketungau and Melawi basins were one and the same (Doutch 1992). They have subsequently become separated by uplift of the Semitau Ridge.
Rajang Group of Sabah and East Kalimantan The Rajang Group of Sabah, as presently defined, is of much longer depositional duration (Table 1) and its stratigraphy and structure have not yet been satisfactorily resolved. However, it can be subdivided into a 'Lower Part', whose sedimentation history extends from Late Cretaceous to Late Eocene, and an 'Upper Part', extending from Oligocene to Early Miocene (Lim 1985). It is the Lower Part only which extends into Kalimantan and Sarawak. The Embaluh Group appears to continue east of 115° as the Upper Cretaceous to Eocene Mentarang Formation, extensively exposed and of distal turbidite character, strongly folded and displaying slaty cleavage (BRGM 1982). Northwards along
255
RAJANG GROUP AND LUPAR LINE, BORNEO
Schwaner Barito Mountains [marginal [active volcanosea ?) plutonic arc)
Fore-arcbasin Selangkai Formation (in part)
0
Accretionary Trench prism / (SelangkaiFm in part, / + Embaluh G r o u p ~ /
~
Proto south turbidites+ China Barremian Sea / chert ~/'~
sea level
,0
KM 30 -4 t ~ r ~ J[ ~ ~ '
ou 11 i i
, ons¢
.4~." ~
"
~
or,.,.,',"
Ophl'olitenow seen inBoyan, Lupar and
~
,,,,o,',o
50 "I parIiy @~ ascending , ~ " {uplifted--A magma + .@~.
0
,4°
I 0 ~ 60
k,....
60 ]in Meratus ~volatiles , i f ~ ~r~" 70 ]...Mountains I ~'~'"Y I Present S
30 KM A,,
Mantle
70
I
1O0 KM
200
300
PresentN
(a)
rEmbaluhGroup RajangGroup [Indonesia] (Sarawak)
volcanic arc <~==extinct 0 10 -
final position of 'trench' at subduction-collision ¢ transition
---sea level - - - - - BALINGIAN _-- .... 10 -"---------: - : - - I - LUCONIA - 2 0 - ~ ~ ~ _ - - _ - _ _ - - / C O N T I N E N T I - - .... 20 -- 30 KM . . . . .
~
-
-
~
~
40 50 ~
" ....
50
60 ~ LEGEND: ~ ]
I 300 KM
(b) deformed turbidites basal strata overlying unconformity
I ~
i 200
Schwaner Early Mountains Melawi (buttress) Basin
I
I
400 ~
500
Early Lupar Ketungau Fault Basin ~
, .~.,,°n-.?~
Sibu Zone Uplift (Raiang Group)
60 600 KM from volcanoplutonic arc Tatau Horst
~..J,~ /,:...~/IA,'.4.~
Miri Zone
°°,..°
volcanics
plutons Piyabung
,,~,,-~
isostatic
..o.cc,
oceanic crust
,~'~" continental lithosphere
r.,
volcanism
sinkingslab
I
I
I
I
I
0
1O0
200
300
400
Present South
KM
500
Present North
(c) Fig. 2. Schematic N-S tectonic cross-sections through the Schwaner Mountains and Rajang Group. Vertical and horizontal scales are very approximate. (a) Late Cretaceous (c. 80 Ma) north-facing arc-trench system. (b) Palaeocene (c. 60 Ma) cessation of subduction, forwards migration of the 'trench' and incipient collision. (e) Late Eocene collision, compression and uplift of the Rajang Group between the Schwaner Mountains and the Balingian-Luconia Continent, followed by normal faulting and deposition of Late Eocene strata unconformably upon peneplained Rajang Group. The scenario of this figure is in strong contrast to that of James (1984), reproduced by Hutchison (1989, p. 85).
256
C.S. HUTCHISON
strike in SW Sabah, it has been mapped as the Sapulut Formation (Collenette 1965). No slaty cleavage was mentioned by Collenette, who described the mudstones as 'fissile and indurated', although previous workers (see Reinhard & Wenk 1951) referred to the Sapulut, as well as the Trusmadi Formation, as a "slate formation'. The farthest extension into Sabah is known as the Trusmadi Formation, well exposed in accessible outcrops from Ranau (near Mount Kinabalu) along the main road towards Telupid. The predominantly Palaeocene-Eocene Trusmadi Formation is a tightly folded, generally steeply-dipping, phyllitic distal turbidite, characterized by quartz veining, and identical in appearance to large parts of the Lupar and Belaga Formations of Sarawak, but in general more shaly. The upper parts of the Mentarang and Sapulut Formations show indications of shallowing of the sea, and limestones and jasper conglomerates are recorded in both (Hutchison 1989). In Kalimantan, the Mentarang Formation is unconformably overlain by the Upper Eocene Sebuku Formation basal conglomerate, followed by reefal limestone (BRGM 1982). Turbiditic sedimentation was therefore eliminated by tectonic uplift by Late Eocene, a situation identical to that in Sarawak (Table 1). A folding and metamorphic event clearly must separate the Trusmadi from the unmetamorphosed and less deformed Crocker Formation. This event is named here the Sarawak orogeny. However, only fault contacts have been observed between the two formations (Collenette 1965; Jacobson 1970). The very sandy Crocker Formation turbidite, containing several extremely thick sandstones, is usually considered as Palaeocene to Lower Miocene (Lim 1985). The upper western part has been mapped as the turbiditic West Crocker and Temburong Formations by Wilson (1964). The latter is very fossiliferous and has an age of Oligocene to Lower Miocene. The 'East Crocker Formation' is less well known and its reported Palaeocene and Eocene age is poorly constrained palaeontologically. Structural relations between the "West' and 'East' Crocker have not been resolved. The Sarawak equivalent of the Oligocene to Lower Miocene West Crocker and Temburong Formations is the Setap Shale. The unconformity at the top of the Trusmadi and Mentarang Formations indicates that the West Crocker and Temburong Formations should not be included as part of the Rajang Group of Sarawak. Palaeocurrent analysis on sole marks of the West Crocker and Temburong formations (Stauffer 1968) indicate northwards directed currents. The provenance is most likely to be the Lurah and Long Bawang formations, which were uplifted before the end of the Eocene; an interesting heavy mineral
provenance study remains to be carried out. These shallow marine formations are considered the time equivalent of the deep marine Mentarang Formation (BRGM 1982). If the Mentarang, Sapulut and Trusmadi Formation turbidites represent an accretionary prism, then the Late Cretaceous to Eocene paired volcanic arc, which cannot be seen in Sabah, needs to be sought in the northern arm of Sulawesi, which was contiguous with eastern Sabah before the Late Eocene opening of the Celebes Sea (Hutchison 1989). However, the volcanic arc activity of north Sulawesi began in the Early Miocene. Older rocks have been interpreted as ophiolite, with preTertiary metamorphic and intrusive rocks (Carlile et al. 1990). The Upper Cretaceous to Eocene Tinombo Formation in the north arm of Sulawesi contains only minor calc-alkaline and tholeiitic suites (Kavalieris et al. 1992). An outstanding problem remains, and that is the relationship between the shallow marine Upper Cretaceous to Eocene Lurah Formation (BRGM 1982) and the West Crocker-Temburong Formation. The areal distribution of formations indicates that while sediments of the West Crocker Formation were being transported from the south, there lay a landmass predominantly made of the Lurah Formation in that direction. Just across the border in Sarawak, the Upper Cretaceous to Eocene Kelalan Formation (Fig. 1) has been mapped by Haile (1962). He correlated it with the Belaga Formation, and outcrops at Batu Gading are of turbidite identical to the Belaga Formation. In the region south along strike from the West Crocker and Temburong Formations, Haile (1962) mapped the shallow marine coal-bearing Kelabit Formation, which has yielded good Oligocene to Lower Miocene fossils, and it is the proximal along-strike equivalent of the deeper water West Crocker and Temburong Formation turbidites. The Kelabit Formation is probably equivalent to the Long Bawang Formation, mapped in contiguous Kalimantan in a region probably underlain by evaporites, for all groundwater wells produce brine (BRGM 1982).
Uplift o f the Crocker Formation Deformation resulting in spectacular episodic uplift of the Crocker Formation is well documented in the offshore oil fields by several prominent unconformity surfaces (Levell 1987; Tan & Lamy 1990; Hazebroek & Tan 1993). The first is the Deep Regional Unconformity (15 Ma, Middle Miocene) and the latest is the Shallow Regional Unconformity (9 Ma, Upper Miocene). The tectonic front, or northwestwards limit of tectonic deformation and uplift, progressively moved ocean-wards
-
-
"
-
-
-
,?
I" /P~"
/
BARAM
EAST
,4 DELTA
/
PROVfNCEBOUNDARY
.
~5 °
~
~
, - / ~ / , ~ ~": X ~ x Ii
(a)
ANTPCLINE SYNCLINE " MAJORGROWTHFAULT
*+
/
¢-,
r
,//
SABAH
~/
116o
COUNTERREGIONAL
/
/
/"
GROWTHFAULT REVERSE/THRUSTFAULT , NORMALFAULT
~
//
/
/
×
KINABALU
6°
/
115o
o
....
....
j
,,60
s
/
/C
/(,
(b
/
//
EXTENT OF EROSION dVt::
SABAH
/
/ ' ,i
-/ "I ~
APPROX. . ,' Ovf" PALAEO- ENVIRONMENT ."" ~_~'.,@ ~ ' ,"~ ~ \ ~ 5 ~ BOUNDARIES ,"
SLUMP SCAR ON SLOPE C = CARBONATE
9Mo ,..,. u SRU
13Mo, 15MOv , v DRU
, , LIU
12Mo -T--~- U I U
--DRU
~:
(C)
I Oilfields .5}
IIKINABALU
%,°
t~16°
7o_
Fig. 3. (a) The main structural elements of northwest Sabah (after Tan & Lamy 1990; Rice-Oxley 1991). (b) Late Miocene palaeo-environments for stages IV E/F (after Rice-Oxley 1991), showing also the positions of identified slump scars along the Late Miocene continental slope (shelf edge), from Levell & Kusumajaya (1985) and Hutchison (1995). (c) Sea-ward limits of the important regional unconformities, which successively migrated away from the tectonic front of the uplifting West Crocker Formation fold-thrust belt (based on Levell 1987). Palaeo-environments: LCE lower coastal plain; C, coastal; HIN, holomarine inner neritic; HMN, holomarine middle neritic; HON, holomarine outer neritic; BAT, bathyal. Unconformities: DRU, deep regional unconformity; LIU, lower intermediate; U1U, upper intermediate; SRU, shallow regional unconformity. Oil fields: 1. Samarang; 2. Glayzer; 3. Erb West; 4. Ketam; 5. Tembungo; 6. St Joseph; 7. SW Emerald; 8. Furious; 9. Barton.
l ~
/"
/
PLaTfORM
tO
©
©
t"
[.-
>
©
> z
258
C.S. HUTCHISON
(Fig. 3). The unconformities extended quite close to the continental slope (tectonic front), so that the fluvial system brought sands to the edge of the very narrow continental shelf to spill over the slope into the bathyal area as submarine slumps, such as the one described by Hutchison (1995), and to feed turbidite fans. The persistence of outer neritic to bathyal marine conditions, without unconformities, west and northwest of the tectonic front has been documented by Rice-Oxley (1991). The Late Miocene tectonic front of the West CrockerTemburong orogen approximately coincides with the Jerudong Line, Morris Fault, and Transition zone between the Inboard and Outboard belts of Shell geologists (Fig. 3). The Mid to Late Miocene inversion of the West Crocker-Temburong basin to form the Crocker Ranges of Sabah has never been given a name, but the term Sabah Orogeny seems appropriate. Its cause remains obscure. The spectacular Late Miocene granitoid forming the core of Mount Kinabalu (4101 m) is a product of this orogeny, with whole rock K-Ar dates commonly as old as 10 Ma (Jacobson 1970).
The Sabah orogeny regionally This orogenic event was widespread throughout the South China Sea, but only in the Zhu Jiang Kou (Pearl River Mouth) Basin has it been given a name: Dongsha Movement ( S u e t al. 1989). However, this is not a good regional name, since the inversion there is not spectacular. The nearby Song Hong-Yingge Hai Basin contains very impressive Middle to Late Miocene inversion structures, well displayed in regional seismic records, but they have not been published. The best published account of spectacular basin inversion is by Ginger et al. (1993) in the West Natuna area, where the Anambas Graben became the Boundary High during Middle to Late Miocene inversion. The eroded anticlines are unconformably overlain by fiat lying latest Miocene-Pliocene Muda Formation. In the contiguous Penyu and Malay basins, inversion structures have not yet been analyzed and described because of a lack of seismic resolution at depth. However, the Upper Miocene unconformity has been well documented
in the southern end of the Malay Basin (Armitage & Viotti 1977; Nik Ramli 1988). Even though the Middle to Late Miocene inversion structures may not have been identified and described, the Upper Miocene unconformity is widely documented throughout Southeast Asia. All of the South China Sea basins rapidly returned to post-inversion subsidence after erosion of the inversion structures, so that the basins remain today beneath sea-level. Much farther south, the Late Miocene inversion is well displayed in the Meratus Mountains, which formed the basement to the Eocene through Miocene sequence, then rose up only in the Late Miocene to form a ridge which subdivided a formerly much more extensive basin (Hutchison 1992b). The post-inversion unconformity is also seen in Central Sumatra (Hutchison 1993). Both in the Meratus and Central Sumatra the uplifted basins have not subsequently subsided beneath sea-level.
Summary Although the interior of Borneo remains poorly known, the following broad evolutionary scenario may now be summarized. The turbiditic Rajang Group of Sarawak may be usefully extended into northern Kalimantan and central Sabah. It may be interpreted as an accretionary prism, compressed and uplifted between the Schwaner Mountains volcano-plutonic arc and the Luconia-Balingian-Miri microcontinent of Sarawak by the Upper Eocene Sarawak orogeny, to become an integral part of the extensive Upper Eocene Sundaland landmass upon which the Ketungau and other continental basins were developed. The Oligocene to Lower Miocene turbiditic West Crocker and Temburong Formations of Sabah should be excluded from the Rajang Group. Their nearshore equivalent has been mapped to the south as the Kelabit and Long Bawang Formations, suggesting provenance from eroding Rajang Group outcrops of the Sarawak Orogen. The West Crocker and Temburong Formations were deformed and uplifted by the Middle to Upper Miocene Sabah orogeny.
References ACHACHE, J., COURTILLOT, V. & BESSE, J. 1983. Paleomagnetic constraints on the Late Cretaceous and Cenozoic tectonics of Southeast Asia. Earth and Planetary Science Letters, 63, 123-136. ADAMS, C. G. & HAAK, R. 1962. The stratigraphical succession in the Batu Gading area, Middle Baram, North Sarawak. In: HAILE, N. S. The geology and mineral resources of the Suai-Baram area, north
Sarawak. Geological Survey, Department of the British Territories in Borneo. Kuching, Memoir, 13, 141-150. AMIRUDDIN. & TRAIL, D. S. 1989. Nangapinoh 1:25,000 quadrangle, West Kalimantan, Geological Data Record. Geological Research and Development Centre, Bandung. ARMITAGE, J. H. & VIOTTI, C. 1977. Stratigraphic
RAJANG GROUP AND LUPAR LINE, BORNEO nomenclature-southern end Malay Basin. In: Proceedings Indonesian Petroleum Association 6th annual convention. 69-94. BERGMAN, S. C., TURNER, W. S. & KROL, L. G. 1987. A reassessment of the diamondiferous Pemali Breccia, southeast Kalimantan, Indonesia: intrusive kimberlite breccia or sedimentary conglomerate? Geological Society of America Special Paper, 215, 183-195. BLADON, G. M., PIETERS, P. E. • SUPRIATNA, S. 1989. Catalogue of isotopic ages commissioned by the Indonesia-Australia Geological Mapping Project for igneous and metamorphic rocks in Kalimantan, Preliminary Report. Geological Research and Development Centre, Bandung. BRIAIS, A., PATRIAT,P. & TAPPONNIER,P. 1993. Updated interpretation of magnetic anomalies and seafloor spreading stages in the South China Sea: implications for the Tertiary tectonics of Southeast Asia. Journal of Geophysical Research, 98, 6299-6328. , TAPPONNIER,P. & PAUTOT,G. 1989. Constraints of data on crustal fabrics and seafloor spreading in the South China Sea. Earth and Planetary Science Letters, 95, 307-320. BRGM. 1982. Geological mapping and mineral exploration in northeast Kalimantan 1979-1982, Final Report. Rapport du Bureau de Recherches Grologique et Mini~res, 82, RDM 007AO, Orlrans, France. CARLILE, J. C., DIGDOWIROGO, S. & DARIUS, K. 1990. Geologic setting, characteristics and regional exploration for gold in the volcanic arcs of North Sulawesi, Indonesia. Journal of Geochemical Exploration, 35, 105-140 CARTER, E. G. & YEATS, A. K. 1973. The Lower Tertiary in Kattim Shell contract area, East Kalimantan: Results of 1972-1975 field surveys. Kaltim Shell NV Report. CHEN, S. 1987. Magnetic profiles. In: Atlas of Geology and Geophysics of the South China Sea, scale 1:2 000 000. Second Marine Geology Investigation Brigade of the Ministry of Geology and Mineral Resources, Guangdong, PR of China. COLLENETTE,P. 1965. The geology and mineral resources of the Pensiangan and Upper Kinabatangan area, Sabah, Malaysia. Malaysian Geological Survey, Borneo Region, Memoir, 12. DOUTCH, H . E 1992. Aspects of the structural histories of the Tertiary sedimentary basins of East, Central and West Kalimantan and their margins. BMR. Journal of Australian Geology and Geophysics, 13, 237-250. GINGER, D. C., ARDJAKUSUWAH,W. O., HEDLEY, R. J. & POT~tECARY, J. 1993. Inversion history of the West Natuna Basin: examples from the Cumi-Cumi PSC. In: Proceedings Indonesian Petroleum Association 22nd annual convention. 635-658. GOWER, R. J. W. 1990. Early Tertiary plate reconstructions for the South China Sea region: constraints from northwest Borneo. Journal of Southeast Asian Earth Sciences, 4, 29-35. HAILE, N. S. 1962. The geology and mineral resources of the Suai-Baram area, north Sarawak. Geological Survey Department of the British Territories in Borneo, Kuching, Memoir, 13.
--
- -
--
--,
259
1974. Borneo. In: Spencer, A. M. (ed.) MesozoicCenozoic Orogenic Belts: Data for Orogenic Studies. Geological Society, London, Special Publication, 4, 333-347. 1994. The Rajang accretionary prism and the TransBorneo Danau Suture. Abstracts Tectonic Evolution of SE Asia Conference, December 1994. Geological Society, London, 17. & Ho C. K. 1991. Geological Field Guide: SibuMiri traverse, Sarawak (24 September-1 October 1991). Petronas Petroleum Research Institute, Kuala Lumpur. MCELHINNY, M. W. & McDOUGALL, I. 1977. Palaeomagnetic data and radiometric ages from the Cretaceous of west Kalimantan (Borneo), and their significance in interpreting regional structure. Journal of the Geological Society, London, 133, 133-144.
HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey, Professional Paper, 1078. HAZEBROEK, H. E & TAN, D. N. K. 1993. Tertiary tectonic evolution of the NW Sabah continental margin. Geological Society of Malaysia Bulletin, 33, 195-210 Ht]KINIAN, R., BONTt~, P., PAUTOT, G., JACQUES, D., LABEYRIE,L., ETAL. 1989. Volcanics from the South China Sea ridge system. Oceanologica Acta, 12, 101-115. HOLLOWAY,N. H. 1981. The stratigraphy and relationship of Reed Bank, north Palawan, and Mindoro to the Asian mainland and its significance in the evolution of the South China Sea. AAPG Bulletin, 66, 1355-1383. HUTCHISON, C. S. 1989. Geological evolution of SouthEast Asia. Clarendon Press, Oxford, Oxford Monographs on Geology and Geophysics, 13. 1992a. The Southeast Sulu Sea, a Neogene marginal basin with outcropping extensions in Sabah. Geological Society of Malaysia Bulletin, 32, 89-108. 1992b. The Eocene unconformity in Southeast and East Sundaland. Geological Society of Malaysia Bulletin, 32, 69-88. 1993. Gondwanaland and Cathaysian blocks, Palaeotethys sutures and Cenozoic tectonics in South-east Asia. Geologische Rundschau, 82, 388-405. 1995. Mrlange on the Jerudong Line, Brunei Darussalam, and its regional significance. Geological Society of Malaysia Bulletin, in press. JACOBSON, G. 1970. Gunung Kinabalu area, Sabah, Malaysia. Geological Survey of Malaysia, Report, 8.
JAMES, D. M. D. 1984. The geology and hydrocarbon resources of Negara Brunei Darussalam. Brunei Museum and Brunei Shell Petroleum Company. KAVALIERIS,I., VAN LEEUWEN,T. M. & WILSON, M. 1992. Geological setting and styles of mineralization, North arm of Sulawesi, Indonesia. Journal of Southeast Asian Earth Sciences, 7, 113-129. KIRK, H. J. C. 1957. The geology and mineral resources of the Upper Rajang and adjacent areas. Geological Survey Department for the British Territories in Borneo, Kuching, Memoir, 8. -1968. The igneous rocks of Sarawak and Sabah.
260
c . s . HUTCHISON
Geological Survey of Borneo Region, Malaysia Bulletin, 5. LEE, CHAI-SHtNG & MCCABE, R. 1986. The BandaCelebes-Sulu Basin: a trapped piece of CretaceousEocene oceanic crust. Nature, 322, 51-54 LEVELL, B. K. 1987. The nature and significance of regional unconformities in the hydrocarbonbearing Neogene sequences offshore West Sabah. Geological Society of Malaysia Bulletin, 21, 55-90 - & KASUMAJAYA, A. 1985. Slumping at the late Miocene shelf-edge offshore West Sabah: a view of a turbidite basin margin. Geological Society of Malaysia Bulletin, 18, 1-29. LIECHTI, P., ROE, E W. & HALLE,N. S. 1960. The geology
of Sarawak, Brunei and the western part of North Borneo. Geological Survey Department for the British Territories in Borneo, Kuching, Bulletin, 3. LIM, P. S. 1985. Geological Map of Sabah, 1:500,000. Geological Survey of Malaysia, Kota Kinabalu. MOLENGRAA~, G. A. E 1902. Geological Explorations in Central Borneo (1893-94). E. J. Brill, Leyden. NIK RAMLI, 1988. Characteristics of J-sandstone (Tapis Formation) reservoirs in the southeastern part of the Malay Basin. In: Proceedings of the Southeast Asian Petroleum Exploration Society (Seapex). 8, 239-248. PAUTOT, P., RANGIN, C., BRIAIS, A., TAPPONNIER, P., BEUZART, P., ET AL. 1986. Spreading direction in the Central South China Sea. Nature, 321, 150-154. PIETERS, P. E. & SUPRIATNA, S. 1989. Preliminary
Geological Map of the West, Central and East Kalimantan area, 1:1 000 000. Geological Research and Development Centre, Bandung. , TRAIL, D. S. & SUPR1ATNA, S. 1987. Correlation
of Early Tertiary rocks across Kalimantan. In: Proceedings Indonesian Petroleum Association 16th annual convention. 291-306. REINHARD,M. & WENK, E. 1951. Geology of the colony of North Borneo. Geological Survey Department of British Territories in Borneo, Kuching, Bulletin, 1. RICE-OXLEY, E. D. 1991. Palaeoenvironments of the Lower Miocene to Pliocene sediments in offshore NW. Sabah area. Geological Society of Malaysia Bulletin, 28, 165 194. Ru, KE & PIGOTT, J. D. 1986. Episodic rifting and subsidence in the South China Sea. AAPG Bulletin, 70, 1136--1155. SCHMIDTKE, E., FULLER, M. & HASTON, R. 1990. Paleomagnetic data from Sarawak, Malaysian Borneo, and the Late Mesozoic and Cenozoic tectonics of Sundaland. Tectonics, 9, 123-140. SIKUMBANG, N. 1986. Geology and tectonics of pre-
Malaysia. Geological Survey of Malaysia, Report, 13, Kuching, Sarawak. 1982. Geological map of Sarawak, 1:500,000. Geological Survey of Malaysia, Kuching. 1986. Palaeogeographic development of West Sarawak. Geological Society of Malaysia Bulletin, 19, 39-49. -& LAMY, J. M. 1990. Tectonic evolution of the NW. Sabah continental margin since the Late Eocene. Geological Society of Malaysia Bulletin, 27, 241-260. TATE, R. B. 1991. Cross-border correlation of geological formations in Sarawak and Kalimantan. Geological Society of Malaysia Bulletin, 28, 63-95 TAYLOR, B. & HAYES, D. E. 1980. The tectonic evolution of the South China Basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. American Geophysical Union,
-
-
Geophysical Monograph, 23, Washington, 89-104. & -1983. Origin and history of the South China Sea Basin. In: HAYES, D. E. (ed.) The Tectonic and
Geologic evolution of Southeast Asian Seas and Islands, Part 2. American Geophysical Union, Geophysical Monograph, 27, Washington, 23-56. TING, C. S. 1992. Jurassic-Cretaceous palaeogeography of the Jagoi-Serikin area as indicated by the Bau Limestone Formation. Geological Society of Malaysia Bulletin, 31, 21-38. TONGKUL, E 1990. Structural style and tectonics of Western and Northern Sabah. Geological Society of Malaysia Bulletin, 27, 227-239. VAN BEMMELEN, R. W. 1949. The geology of Indonesia. Government Printing Office, The Hague. VAN DE WEERD, A. A. & ARMIN, R. A. 1992. Origin and evolution of the Tertiary hydrocarbon-bearing basins in Kalimantan (Borneo), Indonesia. AAPG Bulletin, 76, 1778-1803. VAN LEEUWEN, T. M., LEACH, T., HAWKE, A. A. & HAWKE, M. M. 1990. The Kelian disseminated gold deposit, East Kalimantan, Indonesia. Journal of Geochemical Exploration, 35, 1-6 I. WAIN, T. & BEROD, B. 1989. The tectonic framework and paleogeographic evolution of the Upper Kutei Basin. In: Proceedings Indonesian Petroleum Association 18th annual convention. 55-78. WHITE, J. M. JR. & WING, R. S. 1978. Structural development of the South China Sea with particular reference to Indonesia. In: Proceedings Indonesian
Petroleum Association 7th annual convention. 159-164.
STAUFFER, P. H. 1968. Studies in the Crocker Formation,
WILLIAMS, P. R. & HARAHAP, B. H. 1987. Preliminary geochemical and age data for post-subduction intrusive rocks, Northwest Borneo. Australian Journal of Earth Sciences, 34, 405-4 15. & HERYANTO, R. 1986. Sintang 1:250 000
Sabah. Borneo Region of Malaysia, Geological Survey Bulletin, 8, 1-13.
Quadrangle, West Kalimantan: Geological Data Record. Geological Research and Development
Tertiary rocks in the Meratus Mountains, south-east Kalimantan, Indonesia. PhD Thesis, University of London.
Su DAQUAN, WHITE, N. & MACKENZIE, D. 1989. Extension and subsidence of the Pearl River Mouth Basin, northern South China Sea. Basin Research, 2, 205-222. TAN, D. N. K. 1979. Lupar Valley, West Sarawak,
--,
Centre, Bandung. JOHNSTON, C. R., ALMOND, R. A. & SIMAMORA, W. H. 1988. Late Cretaceous to Early Tertiary structural elements of West Kalimantan. Tectonophysics, 148, 279-297.
RAJANG GROUP AND LUPAR LINE, BORNEO WILSON, R. A. M. 1964. The geology and mineral resources of the Labuan and Padas Valley area, Sabah, Malaysia. Geological Survey of the Borneo Region of Malaysia, Memoir, 17. WOLFENDEN, E. B. 1960. The geology and mineral resources of the Lower Rajang Valley and adjoining areas, Sarawak. Geological Survey, Department for
261
the British Territories in Borneo, Kuching, Memoir, 11. ZEIJLMANSVANEMMICHOVEN,C. P. A. 1939. De Geologie van het centrale en oostelijke deel van de Westerafdeeling van Borneo. Jaarboek van het Mijnwezen in Nederlandsch-lndie Verlhandelingen, 7-186.
Origin and tectonic significance of the metamorphic rocks associated with the Darvel Bay Ophiolite, Sabah, Malaysia S H A R I F F A. K. O M A N G 1 & A. J. B A R B E R 2
1 Department of Earth Sciences, Faculty of Science and Natural Resources, Universiti Kebangsaan Malaysia, Sabah Campus, Locked Bag No 62, 88996 Kota Kinabalu, Sabah, Malaysia 2 SE Asia Research Group, Department of Geology, Royal Holloway, University of London, Egham TW20 OEX, UK Abstract: Banded hornblende gneiss, foliated amphibolite, hornblende, chlorite and siliceous
schist form lenses in an 8 km wide belt within the Darvel Bay Ophiolite Complex. Foliation in the belt is generally steep to vertical, striking parallel to the trend of the belt and lineations are sub-horizontal. Mineral and geochemical studies show that the metamorphic rocks represent banded and isotropic gabbros, plagiogranites, doleritic and basaltic dykes, basaltic volcanics and cherts formed at a spreadingridge in a supra-subductionzone environment, which were deformed at high temperatures but low pressures along a transform fault. Incorporation of supracrustal cherts indicates that the transform extended for hundreds of kilometres between spreading centres. Garnet pyroxenites and amphibolites found as clasts in Miocene volcanic agglomerates formed at high pressures, and temperatures are interpreted as derived from a metamorphic sole underlying the complex, formed during subduction of ocean crust and the emplacement of the ophiolite complex on Sabah.
Ophiolite outcrops are distributed throughout the eastern part of Sabah, East Malaysia, from Banggi Island in the north to Ranau and Telupid and the Lahad Datu area in the south (Fig.l, inset). The Darvel Bay Ophiolite Complex in the south is the most extensive outcrop, extending 100 km westwards from Darvel Bay. The greater part of the complex consists of peridotite, largely serpentinized, but it also includes cumulate pyroxenites, layered and massive gabbros and diorites, metamorphosed to varying degrees. The complex is also cut by dolerite dykes, although these are never so abundant to be termed a sheeted dyke complex. These rock types are closely associated with outcrops of the Chert-Spilite Formation, composed of pillow basalt, banded ribbon chert, turbiditic sandstones, mainly volcaniclastic but with some rare quartz sandstones, and a few occurrences of massive limestone. The ophiolite complex has been well described in the publications of the Geological Survey of Malaysia (Reinhard & Wenk 1955; Fitch 1955; Dhonau & Hutchison 1966; Koopmans 1967). The complex has been interpreted as a segment of ocean floor, either of a Proto-South China Sea (Holloway 1981; Rangin et al. 1990) or of the Celebes Sea (Hutchison 1988). A wide range of K-Ar age dates has been obtained from the rocks of the ophiolite complex from 210Ma (Leong 1971) to 137 Ma
(Rangin et al. 1990). Cherts from the Chert-Spilite Formation have yielded radiolaria of Lower Cretaceous age (Leong 1977; Rangin et al. 1990; Aitchison 1994). Massive limestones associated with the Chert-Spilite Formation contain Cretaceous foraminifera (Leong 1974). The ChertSpilite Formation is interpreted as representing ocean floor sediments which were deposited on top of the ophiolite (Hutchison 1975), and carbonate cappings to seamounts. Since the underlying oceanic crust is unlikely to be much older than the oldest overlying sediments, the Early Jurassic K-Ar ages are regarded as spurious (Hutchison 1988). Fragments of ophiolitic rocks are found as clasts in Eocene sediments (Newton-Smith 1967; Rangin et al. 1990), suggesting that the ophiolite complex had been obducted onto Sabah either in the latest Cretaceous or earliest Palaeogene. In the early Miocene the ophiolite complex formed the basement to a volcanic arc, possibly related to continued subduction of the Proto-South China Sea. Mitchell et al. (1986) and Rangin (1989) have suggested that, as a result of continued compression, the complex was backthrust over the Celebes Sea floor to the south. The ophiolite complex is surrounded by chaotic melange deposits (Fig. 1) which contain fragments of all the rock units represented in the complex and the Chert-Spilite Formation, as well as
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 263-279.
263
264
s . A . K . OMANG (~ A. J. BARBER 0 r'
5
10
Km
Ophiolite Complex E ~ Felsite
~ ~'~] • " '* '° "~"* "*"*"*"° "° "° ". "° " . . . - , ~ '. ". "° '. ". ". "° ". ~ ' . ' ~ * * , ' i**: ~ ~ : : * : i * " []'EI , ,, ,, , , , ,, ,, -° "4 "° . "° "° .o .° .° ...... , ', ', ', ', ', ', "°,:*.:*.:*.:°.:*,:*,:°,:*,:*, ~ "°"°' ' °"°' l
Metamorphic rocks (foliated metabasites ) Sheared Serpentinite
~
RecentAIluvium
E~
Chert-Spilite Formation
D Neogene sediments Gabbros & Cumulates [ ~ M6langes Peridotite ~ Thrust fault (mainly serpentinised) - - " Fault :!:!:!:!:i: l: 8" 20' E. :::::::::::::::::::::::::::::::::::::::::::::::::::::
~!~:i :i !i:i:i:!...i.i~.:i:i:i!!ii i!i!i i i i i i i i !i!i i i!i!i !i i!::ii!i!~ii ~~:i.i:i:i:i!ii!i!i!iii!iiiiii!iiiiiiii!!!iii!!!iliiiiiiiiiii ====================================================
Gabbro
*Dolerite dyke PSI2 D~:: PS6
Itlr Pulau Sakar !
I
!
rl~l
~1 i i l.ll.l~.l~.ll. i i i i
: i
!
!
!
!
, ~ ~PT3
|
• ~. 1.1~.13.11.1~.1~ I~. ~ ~ i i i--.~.,1-.-~
p
p ,,.z.~
Pulau "~;K4
~ , ' %,• 9~l Katung- / ruiau Kalungan-.L Saddle/ % -.L Cumulate D C rocks
Darvel
Bay E t
SARAWAK 1 " ~ _/'f
/
Pulau Bohayan
~,
7
4* 58'N Pulau Tabauwan
~,
P
~ f
~':*'i*' 118" 10' E.
/ E --/- 118"20' E I
V~Silumpat Dolerite dyke
Fig. 1. Geological map of the Darvel Bay Ophiolite Complex near Lahad Datu showing location of samples used in this study. Inset map shows distribution of ophiolite complexes in Sabah. B, Banggi Island; R, Ranau; T, Telupid; DB, Darvel Bay.
y o u n g e r sediments and volcanic rocks. Clennell (1991) considered that the melanges were f o r m e d by p r o c e s s e s o f s e d i m e n t a r y s l u m p i n g and diapirism, coincident with the collision o f microcontinental blocks with the north Sabah margin and the extension o f the arc basement related to the d e v e l o p m e n t o f the Sulu Sea in late early and early mid M i o c e n e times. The ophiolite c o m p l e x and the melanges are overlain u n c o n f o r m a b l y by Late M i o c e n e to Pliocene sediments o f the 'circular basins' and in the Quaternary again f o r m e d the basement to a volcanic arc which extends from M i n d a n a o in the Philippines, through the Sulu archipelago to the Dent and S e m p o r n a peninsulas o f Sabah.
Metamorphic
rocks
A l t h o u g h the rocks f o r m i n g the Darvel B a y Ophiolite Complex are variably m e t a m o r p h o s e d throughout, d y n a m i c a l l y m e t a m o r p h o s e d rocks are concentrated in an E - W belt c. 8 k m wide extending westwards from Lahad Datu on the north side o f Darvel Bay (Fig.l) which can be traced for c. 40 k m to the west. Metamorphic rocks are well exposed in coastal sections around the shores o f Darvel Bay and particularly around the small islands within the bay, w h e r e they have been described previously by D h o n a u & H u t c h i s o n (1966) and Hutchison & D h o n a u (1969, 1971) f r o m the north side o f the bay, and by Koopmans (1967)
DARVEL BAY OPHIOLITE, SABAH
from the south. Inland exposures are poor, apart from road-cuts, quarries and scarce river sections. The metamorphic rocks have been described as metagabbros, amphibolites, hornblende gneisses and schists. Reinhard & Wenk (1955) described the dynamically metamorphosed rocks as 'Crystalline Schists' and Koopmans (1967) identified them as the 'Crystalline Basement Complex', with the implication that they were the oldest rock unit in Sabah. However, K-Ar dating reported by Dhonau & Hutchison (1966) showed that these metamorphic rocks are of Cretaceous age, the same age as the Darvel Bay Ophiolite Complex, and form an intrinsic part of the complex The structure of the dynamically metamorphosed belt has been fully described by Dhonau & Hutchison (1966). The metamorphic rocks are typically banded, schistose and lineated. Through most of the belt the banding and schistosity are vertical or steeply dipping and strike E-W, parallel to the general trend of the belt, swinging round to a more NW-SE trend in the islands of Darvel Bay. The lineation is sub-horizontal, lying in the plane of the schistosity and plunges at low angles either to E or W. From their mapping of the structure in the metamorphic rocks, Dhonau & Hutchison (1966) identified two large-scale open postmetamorphic monoclinal flexures on E-W axes affecting the foliation in the islands of Pulau Saddle, Pulau Bohayan and Pulau Silumpat. This paper gives an account of the field relations, petrography, mineral chemistry, geochemistry and K-At isotopic dating of the metamorphic rocks associated with the Darvel Bay Ophiolite Complex and discusses their origins and significance in the tectonic evolution of Sabah. A separate occurrence of metamorphic rock is found among Miocene volcanics at Tungku, 50 km to the east of Lahad Datu (Reinhard & Wenk 1955). Garnet amphibolite from this locality was also included in the present study.
Banded hornblende gneiss The dominant metamorphic rock type in the metamorphic belt within the Darvel Bay Ophiolite Complex is a banded gneiss, the 'Silumpat Gneiss' of Dhonau & Hutchison (1966) and Hutchison & Dhonau (1971), with alternating bands of white feldspathic and black or dark greenish ferromagnesian bands, well exposed on the north shore of Darvel Bay, west of Lahad Datu and in road cuts along the Silam Road. The bands range in thickness from tens of centimetres to millimetres. The banding is paralleled by a schistosity, with the alignment of feldspathic and ferromagnesian aggregates which may also be elongated to form a rodding type of lineation. The intensity of the
265
schistosity increases in localized shear zones. The banding is commonly folded into tight intrafolial isoclinal folds which may show the closed eyed outcrop pattern characteristic of sheath folds. Where folds are present the schistosity is parallel to the axial planes of the folds. In the more intensely deformed shear zones aligned hornblende and feldspar crystals define a mineral lineation, lying in the plane of the schistosity.
Amphibolite Amphibolites are commonly found associated with the banded gneisses in the field. These may form extensive outcrops, as on the northern side of Sakar Island or as smaller bodies which may show crosscutting, intrusive relationships to the hornblende gneisses. The grain size of these amphibolitic rocks is variable. Coarse grained amphibolites show the characteristic distribution of white feldspathic and black or green ferromagnesian minerals which indicate that they have been derived from gabbros (metagabbro). Medium and finer grained amphibolites may show relict ophitic textures in thin section and large euhedral plagioclase phenocrysts or polycrystalline aggregates pseudomorphing plagioclase which indicate an origin as hypabyssal intrusive rocks (metadolerite). Fine grained amphibolites without these features may represent basaltic volcanic rocks (metabasalts). Like the banded gneisses the amphibolites show varying degrees of deformation. Where they crosscut the banding in the gneisses, the amphibolites generally show a less intensely developed schistosity than the associated gneisses. In coarse grained amphibolites the gabbroic texture may be flattened in the foliation, in finer grained rocks minerals may be recrystallized in the schistosity to form hornblende schists. Where the coarser amphibolites cross-cut the banding of the gneisses, no clear chilled margins are evident and the schistosity crosses the contact between the gneisses into the amphibolites, although it is generally not as intensively developed in the latter. Finer grained amphibolites often have a dyke-like form, ranging in width from a few metres to tens of centimetres and show varying angular relationships with the banding of the gneiss. Sometimes they are concordant, and then generally have a well developed schistosity with a similar intensity to that in the adjacent gneisses. Where amphibolite dykes cut the banding at a high angle, the schistosity is generally less well developed than in the surrounding gneiss. Evidently some of the dykes were intruded into already deformed gabbros, but deformation continued after dyke emplacement, indicating that the dykes were intruded into a zone of active deformation.
266
s . A . K . OMANG t~ A. J. BARBER
twinning. Microprobe analysis shows that where plagioclase is associated with pyroxene it has a composition of An68-An74 (labradorite to bytownite) representing an igneous relic, but more commonly is An 6 -An32 (oligoclase to andesine) of metamorphic origin (Hutchison 1978). In the amphibolites, and especially in highly deformed rocks, the plagioclase is An 3 - A n 5 (albite). In many samples of hornblende gneiss and amphibolite the plagioclase has been completely pseudomorphed by low grade alteration products (saussurite). Epidote is a common constituent of the gneisses and amphibolites as an alteration product of plagioclase feldspars, but in occurrences at the western
Mineralogy In the less highly deformed and altered hornblende gneisses and metagabbros the ferromagnesian component includes relict crystals of pale green or colourless pyroxene. These may form the cores to brownish-green amphibole crystals and show alteration along their margins and along internal cracks to aggregates of small blue-green amphibole crystals. Microprobe analysis shows that they are clinopyroxenes of salite to augite composition (En40.5 Fst3 Wo47-En45.5 Fs]2.7 Wo28),with Xrvlg values of 90-97 and an Na20-content of 0.5 wt%. Plagioclase crystals have a grain size up to 4 ram, and form tabular crystals with albite
Tschermakite
Pargasite
II
Edenite Hornblende • dTremotite 0
(a) Glaucophane 2.0
0.5
1
1.8
Na[M4]
1.6 1.4
Metagabbro Metadolerite I 1.0
]4)Si(IV) = Ca(M4)Ai(IV)
t
.....................................................
Sodic-am~hibole
.......
II
1.2 Winchite a.0
Barroisite
aramite
Sodic-Calcicamphibole
0.8 .
.
.
.
Calcic-amphibole 0.2 ~
+ Metadolerite
I
0.0 0 Actinolite
(b)
2 Tschermakite
Hornblende AI[IV]
etal.
Fig. 2. (a) Composition of amphiboles in metagabbro and metadolerite from Darvel Bay on diagram of Deer (1966). (b) AI[IV] versus Na[M4] plot of amphibole compositions in metagabbro and metadolerite from Darvel Bay, isobars after Brown (1977).
DARVEL BAY OPHIOLITE, SABAH
end of Pulau Sakar and in Pulau Katung Kalungan there are amphibolites in which epidote is a major component of the rock. These rocks are distinguished in the field by their yellowish colour where they occur as bands among more normal amphibolites. In thin section they are foliated, with large crystals of epidote, colourless or pale pink in a matrix of blue-green amphibole and plagioclase. The plagioclase is albite or is represented by an aggregate of sericite and calcite. Microprobe analysis of the epidote crystals show that it is an iron-rich variety (Xps27_33). The Fe 3+ content is very high ( 1 . 5 - 2.0 atoms per unit cell) and the XFe 3+ (1.0) gives an estimated temperature of formation of c. 400°C (Nakajima et al. 1977). Amphiboles in the amphibolites occur either as large brownish-green or as small blue-green crystals which may be randomly oriented or aligned to form foliar and linear structures. The larger amphibole crystals are 2-4 mm in size and are greenish-brown in colour. Some of the brown hornblende has been interpreted by Hutchison (1978) on textural and mineral chemical grounds as of relict igneous origin. However, where they enclose pyroxene relics, or where they are aligned in the foliation and lineation, hornblende crystals are clearly of metamorphic origin. Electron microprobe analyses plotted on the (Na+ K)/A1TM diagram of Deer et al. (1966) show that the amphibole composition lies between the hornblende and pargasitic hornblende fields (Fig. 2a) (cf. Hutchison 1978, fig. 6). On the AlIV/Na (M4) plot of Brown (1977) compositions extend from the calcic amphibole towards the sodic-calcic amphibole field, corresponding to pressures of about 5 kbar (Fig.2b). On the AllY/A1vI plot of Fleet & Barnett (1978) the amphiboles lie in the low pressure field (Fig. 3a) and on the Na + K/A1TMplot of Jamieson (1981) follow the high temperature/low pressure trend (Fig. 3b). The Ti content of the amphiboles (0.04-0.3 atoms per formula unit) indicates that crystallization of the amphiboles took place in the temperature range 550-650°C under low pressure conditions in the lower amphibolite facies (Spear 1981). Using the semi-empirical geothermometer proposed by Plyusnina (1982) for plagioclase/ hornblende pairs, core compositions of crystals from the amphibolites give temperatures of 580-600°C and pressures of 3-4 kbars for R T conditions at the time of crystallization. Small blue-green amphibole crystals commonly occur in the banded gneisses and amphibolites marginal to pyroxene or brown hornblende crystals. In the study by Hutchison (1978), microprobe analyses of blue-green amphiboles from the Darvel Bay Ophiolite Complex showed that their composition lay mainly in the actinolitic hornblende field with some actinolite. In highly deformed
267
hornblende schists these actinolitic hornblendes form a major component of the rock and are aligned to form the schistosity and lineation.
Metamorphic histo:y As far as may be determined the primary mineralogical composition of the banded and isotropic gabbros, now found as banded hornblende gneiss and amphibolite, was clinopyroxene, hornblende (cf. Hutchison 1978), labradorite/bytownite and ilmenite. Olivine and orthopyroxene may have been present in some rocks, as olivine and noritic gabbros occur elsewhere in the Darvel Bay Ophiolite Complex (Omang 1993), but all evidence of their previous existence has been destroyed. The gabbros were recrystallized in a high temperature, low pressure metamorphic environment where pyroxene was replaced by brownish-green hornblende and plagioclase by oligoclase/andesine. Alignment of brown hornblende and plagioclase crystals with foliar and linear structures indicate that recrystallization took place in a high temperature dynamic environment; continued dynamic metamorphism under lower temperature amphibolite facies conditions are indicated where green amphiboles are aligned with the schistosity and lineation. Later recrystallisation under hydrothermal conditions at lower temperatures resulted in the local replacement of plagioclase by saussurite and the alteration of pyroxene and hornblende to chlorite.
Geochemistry Whole rock major element and trace element analyses of hornblende gneisses and amphibolites are shown in Table 1. On an AFM plot (Fig. 4) these metabasites lie in the tholeiite and oceanic gabbro (MORB) field of Kirst (1976). AlzO3]TiO2 ratios also fall in the MORB field (Fig. 5) of Sun & Nesbitt (1978). Trace element analyses show that high field strength (HFSE) elements (e.g. Zr, Y) have a high concentration relative to large ion lithophile (LILE) elements. The Cr-Y plot (Pearce et al. 1984b) and Ti-Zr-Y diagram (Pearce & Cann 1973) also suggest a MORB-like character. On the basis of spider diagrams the samples can be divided into three groups. The metadolerite dykes (Fig. 7a), are a suite of basaltic rocks with moderate LILE enrichment but are with a clear negative Nb anomaly relative to the light rare earth elements (LREE). One group of metagabbros (Fig. 7b) is very similar to the metadolerites, but with more primitive compositions, indicated by high Cr and Ni, and with much lower contents of
268
s.A.K. OMANG(~ A. J. BARBER Pargasite
2
High-pressure CT"amphib°lez°ne
~-'l'/
~i~'l~/ Ed~.
AI[IV] 1
Tschermakite
.............Homilende ........................................
Low-pressure ole zone
Actinolite 0
(a)
Glaucophane 0
2
AI[Vl]
Glaucophane
High-pressuretrend
2.01.51
Na+K Eckermakite 1 1 /
Ede ite
[ High-temperature 0.0 Tremolite/A~tinolit?
(b)
0.0
0.5
i
Pargasite
+
.=
~
,
1.0
1.5
Ai[IV]
d Tschermakite 2.0
Fig. 3. (a) AI[VI] v. AI[IV], and (b) AI[IV] v. (Na + K) of amphibole compositions in metagabbro and metadolerite from Darvel Bay; boundary between low and high pressure calcic amphiboles in (a) is from Fleet & Barnett (1978); high and low pressure trends in (b) from Jamieson (1981).
K, Rb and Ba. The other group of metagabbros (Fig. 7c) includes samples which have very spiky profiles. They are extremely depleted in most incompatible elements, particularly Nb, LREE, P and Zr; K and Rb are conspicuously low, and only Ba and Sr are at MORB levels in some samples. In view of their textures and the field relationships, the metabasic rocks are interpreted as part of a single ophiolite suite, with geochemical data indicating MORB-like compositions, but Nb deficient and show a mild LILE enrichment. The metadolerites represent an original liquid, with a chemistry suggesting a supra-subduction zone setting for their environment of formation. The group of metagabbros which resemble the metadolerites on spider diagrams are suggested to be former isotropic gabbros, while the other group of metagabbros probably represent cumulate rocks which have lost an evolved inter-cumulus liquid fraction enriched in incompatible elements.
Felsic rocks The hornblende gneisses and amphibolites are frequently cut by felsic veins on the outcrop scale. These veins may be concordant, parallel to the foliation and schistosity and are then boudinaged, or are cross-cutting. Cross-cutting veins may be straight or folded in a ptygmatic style, and undeformed veins may cut earlier folded veins. Where the veins are folded, the schistosity in the surrounding rocks has an axial plane relationship to the folds. These relationships provide clear evidence of multiple intrusion of felsic rocks into a zone of active deformation. At several localities within the metamorphic belt larger bodies of felsic rock are found. On the north shore of Pulau Sakar a felsic body, identified as trondhjemite (Specimen PS12c) has been intruded, with an irregular contact and without a chilled margin, into metadolerites. A weak schistosity is
9.33 8.83 c. 59
A1203/TiO2 CaO/TiO 2
71 364 336 57 n.d. 160.9 0.4 1 n.d. 84.5 1.2 34.6 2 9 10 100 97 219 17
9.36 6.99 c. 59
48.07 1.63 15.25 11.70 8.60 0.18 11.40 2.94 0.06 0.15 99.98 1.83
JS L
105 368 331 53 n.d. 168.6 0.3 1 n.d. 82.7 1.9 34.7 2 11 12 64 158 161 17
10.08 7.71 c. 59
47.55 1.55 15.62 11.55 8.59 0.19 11.95 2.75 0.04 0.13 99.92 1.80
JSp
127 487 228 60 n.d. 130.0 4.9 8 n.d. 25.0 0.4 15.6 0.3 0 2 5 53 359 12
22.86 21.63 c. 71
49.73 0.62 14.17 8.58 10.53 0.15 13.41 2.38 0.25 0.02 99.84 1.39
KS2
206 413 392 69 0.6 333.1 0.6 36 0.2 29.1 0.8 19.3 1 4 6 8 90 33 14
18.42 13.94 c. 64
46.42 0.97 17.87 9.83 8.83 0.18 13.52 2.01 0.12 0.02 99.78 2.56
PS2
159 243 120 41.6 n.d. 493.2 0.9 45 n.d. 6.7 0.1 8.8 1 1 4 32 42 n.d. 14
86.91 62.09 c. 75
48.82 0.23 19.99 5.58 8.50 0.14 14.28 1.96 0.12 0.01 99.63 2.73
PS 3a
128 886 214 63 0.3 103.1 0.2 3 n.d. 21.9 0.4 16.7 1 2 3 60 48 614 13
27.18 26.05 c. 75
50.11 0.57 15.49 6.57 9.96 0.13 14.85 2.07 0.03 0.02 99.79 1.59
PS6
103 269 303 41 0.6 283.6 0.4 8 0.4 78.8 1.0 27.4 2 9 7 17 69 104 18
11.76 11.52 c. 55
48.52 1.28 15.05 10.60 6.52 0.18 14.74 2.26 0.06 0.12 99.33 1.31
PS 12d
52 80 354 49 1.1 105.5 2.9 14 0.6 104.0 2.0 35.9 3 12 10 35 102 n.d. 16
9.24 5.25 c. 52
51.42 1.61 14.87 12.10 6.44 0.20 8.45 4.46 0.28 0.16 99.98 1.47
PS 11
45 74 345 49 0.4 170.6 4.1 32 0.7 105.3 1.8 36.0 2 6 8 64 93 44 15
8.72 5.49 c. 53
51.24 1.64 14.30 12.07 6.86 0.19 9.00 3.84 0.26 0.15 99.54 1.87
PS 11b
Oxides as wt%, trace elements in ppm. Data presented on a volatile-free basis; mg = M g [ M g +Fe2+], Mg =MgO/40; (Fe203* = Fe203 + FeOxI.III); n.d., below detection limit; LOI, loss on ignition at 1100°C.
Ni Cr V Sc Pb Sr Rb Ba Th Zr Nb Y La Ce Nd Cu Zn C1 Ga
115 338 355 56 0.7 155.6 0.6 10 n.d. 77.9 1.4 38.8 2 10 12 60 142 n.d. 17
44.22 1.70 15.86 12.38 9.13 0.22 15.01 0.99 0.03 0.13 99.67 3.44
SiO 2 TiO 2 A1203 Fe203* MgO MnO CaO N a2 0 K20 P205 Total LOI
mg
JS4a
Sample Element
Metagabbro
Table 1. Whole rock geochemical analyses for metagabbros and metadolerites of the Darvel Bay Metamorphic Complex
136 296 314 43 n.d. 126.4 1.5 24 0.3 110.4 1.7 32 3 10 10 18 98 n.d. 16
9.45 5.84 c. 62
48.82 1.63 15.40 11.37 9.29 0.19 9.52 3.12 0.15 0.14 99.62 2.02
PS 12b
92 273 295 46 0.3 109.0 2.0 26 0.3 107.7 1.7 35.1 2 11 11 83 99 33 13
9.11 5.08 c. 56
51.43 1.61 14.67 11.27 7.22 0.20 8.18 4.62 0.19 0.15 99.53 1.23
PS 13d
11 5 386 35 1.2 158.6 1.9 20 0.8 67.6 0.9 35.1 1 5 8 43 121 88 19
9.88 5.28 c. 40
51.42 1.55 15.32 13.94 4.64 0.23 8.18 4.10 0.20 0.21 99.79 2.60
PT3a
Fe203 ×0.9/72; total iron as Fe203
113 282 308 43 1.3 196.8 1.8 22 1.7 95.0 1.7 29.6 7 20 13 46 99 121 17
11.67 7.63 c. 61
49.28 1.34 15.64 10.96 8.45 0.17 10.22 3.58 0.18 0.17 100.0 1.51
PS 12a
Metadolerite
t,~
s.A.K. OMANG• A. J. BARBER
270
F
/ L I+gabr°s Me 1 M°ta"°l°rT
A
M I~ A
M
Oceanicgabbro(Kirst1976)~ AmphibolitesfromVema | FractureZame ] (blonnorezet at. 1984) J
Fig. 4. Metamorphic rocks from Darvel Bay plotted on an AFM diagram compared with gabbros and amphibolites from the Vema Fracture Zone, equatorial Mid-Atlantic; the tholeiite calc-alkaline boundary is from Irvine & Baragar (1971).
developed in the felsic body, parallel to the schistosity in the surrounding metabasites. Thin section study shows that the rock is fine grained with elongated crystals of quartz and feldspar defining the schistosity. Minor constituents are acicular amphibole, epidote, apatite, chlorite, sphene, zircon and Fe-Ti oxides. A block (1 m × 1.2 m) of felsic rock, identified as tonalite (Specimen PK4a), occurs enclosed in amphibolite on the foreshore on the south coast of Pulau Sakar. The block probably represents a xenolith incorporated in a basic intrusion, but now forms a boudin with a schistosity parallel to that in the surrounding metabasite. Thin section study shows that this rock is composed predominantly of quartz and plagioclase feldspar with subsidiary actinolitic hornblende. The rock has recrystallized, with the growth of porphyroblastic feldspar and hornblende crystals; granophyric texture in the quartz-feldspar matrix may represent an original igneous texture or be due to partial melting of the tonalite. The hornblende porphyroblasts contain trails of small inclusions of quartz, feldspar and Ti-oxides defining an internal schistosity. The porphyroclasts are enclosed in augen structures and the external schistosity in the matrix cuts across the
internal schistosity of the crystal at a high angle, indicating that the rocks have been subjected to multiple deformation (Omang 1993, pl. 5.7D). Whole rock major element and trace element analyses of felsic rocks are listed in Table 2, and are compared to ocean ridge granites (ORG) on the spider diagram of Fig. 7d. These analytical data show that although the felsic rocks are not true granites they are comparable to ORG in their incompatible element concentrations and resemble acid rocks fore other ophiolites, interpreted as co-genetic with basic rocks (Pearce et al. 1984a).
Metatuff Chloritic schists are exposed at Km 134 north of Kampong Silam on Jalan Silam, where they are in contact with sheared serpentinite, and near Kampong Lok Bikin on the coast of the mainland opposite Pulau Sakar where they are interbedded with metacherts. At the locality on Jalan Silam the schists are folded by small-scale open folds on ENE (060°-070 °) axes with a low angle of plunge (10-12 °) and steep axial planes indicating a southerly vergence.
271
D A R V E L BAY O P H I O L I T E , S A B A H
In thin section the schistosity is defined by fine grained chlorite flakes and small feldspar laths diverging around larger euhedral to subhedral plagioclase crystals showing albite and Carlsbad twinning (Fig. 8a). These schists have the composition of tholeiitic basalts and are therefore interpreted as tuffaceous rocks belonging to the ophiolite sequence, as hyaloclastic tufts forming part of the ocean floor assemblage, which has been intensely deformed and metamorphosed under greenschist facies conditions.
100
80 ophiolite
.C..•60 .<
40
MORB
2O
• Arc ",
!
!
1
2
1
3
4
TiO2 (wt.%)
(a)
Metachert
Thin bedded red-brown ribbon cherts occur extensively in the northern and southern areas of the Darvel Bay Ophiolite Complex (Fig. 1). In the northern part of the area cherts occur as blocks in the Kuamut Melange (Clennell 1991; Aitchison 1994). In the southern area bedded cherts folded and imbricated with basaltic pillow lavas, forming an accretionary complex, crop out along Jalan Silam. Near Kampong Diam, south of Lahad Datu, discontinuous exposures of greenish-white chert with a bedded appearance outcrop along the shore. In thin section these cherts are composed almost entirely of microcrystalline quartz. The quartz crystals are elongated and have a strong preferred orientation with a mylonitic texture (Fig. 8b). The apparent bedded appearance is a schistosity produced by deformation. Concordant quartz and epidotic veins extend along the schistosity which is cut by calcite veins.
80 70 60
ophiolite
'° ~ 3a
20 ~
~ Im
MORB
A •
10
Arc
4-__ ¥~I I
o o
~
(b)
I
!
2
3
4
TiO2 (wt.%)
Fig. 5. (a).TiO2 v. A1203frio2; (b) TiO2 v. CaO/TiO2 plots of Darvel Bay metamorphic rocks; fields from Sun & Nesbitt (1978).
TfflO0 .J.
1000 JDI
A
MORB: Mid-ocean ridge basalts
PK3a f
Cr
(ppm)
/
i/
100
II I I
\~ i
|
} I
I
: JD6
/
Zrt
!
(a)
/ (~
\
I + Metad°'eri e
~o'nAT~AAmphib°lites,~,3
h
~ IAT
10
IAATB~sla~d_alI~alt~°~e~iatesSalts
b
/
. . . . .
~
.i.'t Y(ppm)
, , ,
100 (b)
Fig. 6. Tectonic discriminant diagrams. (a) Y/Cr after Pearce et al. (1984b); (b) Zr-Ti/100-Y after Pearce & Cann (1973).
272
s.A.K.
OMANG
•
A.
J.
BARBER
1°°/
lO.00
1.0(3
0.00
o.1(
1.
0.0 Sr
i
0
i
!
i
i
i
i
!
0
i
a
K
Rb
Ba
Nb
La
Ce
Nd
P
Zr
Ti
Y
Sc
kssllsll cR PS12a
0.10 Cr
K
(a)
(a)
10.00
10.OO
Rb
Ba
Th
Nb
Ce
Zr
Y
Metagabbros (Cumulates)
1.0£ 0
0,10 V
0.01 Sr
o K
D Rb
i Ba
OlO]
• ~_~ I X JSp I &, PS12d I
•
o Nb
o La
o Ce
I Nd
o P
a Zr
I Ti
u Y
0011
Sc
(b)
Sr
Cr
~
T
•
q
i
I
I
I
I
I
I
I
I
I
K
Rb
Ba
Nb
La
Ce
Nd
P
Zr
Ti
Y
I Sc
Cr
(e)
Fig. 7. (a)-(c) MORB-normalized chemical plots for metadolerites and metagabbros from the Darvel Bay Ophiolite listed in Table I; normalized values from Pearce et al. (1984b). (d) ORG-normalized ocean-ridge granite chemical plot for felsic rocks listed in Table 2; normalized values from Pearce et al. (1984a).
Epidote crystals from the veins analysed by microprobe are of XpSl8_25. XFe 3+ (0.9) is relatively high, corresponding to temperatures of formation of 450-500°C (Nakajima et al. 1977), i.e. in the greenschist facies. These mylonitic quartz schists have been formed by the deformation of bedded cherts in an active shear zone and affected by calcium metasomatism in a hydrothermal system, in which the fluids were enriched in calcium carbonate, under greenschist facies metamorphic conditions. Garnet amphibolite
Fragments of high grade metamorphic rocks including garnet pyroxenites and garnet amphibolites occur in volcanic conglomerates in the Tungku and Pungulupi Rivers 50 km east of Lahad
Datu (Reinhard & Wenk 1955). Garnet amphibolites from the Tungku River were analysed during the present study. In thin section the amphibolites commonly show mylonitic textures with crystals of garnet, pyroxene and hornblende enclosed in augen structures, enclosed in a fine grained matrix of hornblende and plagioclase (Fig. 8c). P, T estimates from amphiboles and from garnet-pyroxene pairs in the present study (Omang 1993), and an earlier study of garnet pyroxenite from the same locality by Morgan (1974), give the conditions of formation of these rocks as T > 850°C at P > 5 kbar. Bulk rock geochemistry, rare earth and trace element geochemistry of the garnet amphibolites show that they are MORB tholeiites and represent oceanic crustal materials. These rocks were metamorphosed as pyroxene granulites and garnet
DARVEL BAY OPHIOLITE, SABAH
Table 2.
Whole rock geochemical analyses for felsic rocks from the Darvel Bay Metamorphic Complex
Sample No.
LD5
PK4a
Sit 2 Tit 2 AI203 Fe203* MgO MnO Cat Na20 K20 P205 Total LOI mg
60.14 0.438 17.90 5.16 2.01 0.120 8.89 4.73 0.333 0.250 99.98 4.73 c. 44
61.46 1.461 15.38 9.43 2.22 0.120 4.77 4.76 0.091 0.308 100.00 1.30 c. 32
Ni Cr V Sc Pb Sr Rb Ba Th Zr Nb Y La Ce Nd Cu Zn C1 Ga
19 13 75 10 3.4 537 7.9 206 2.9 273.3 3.2 22.9 19 40 21 84 56 8 22
5 2 94 24 3.4 169 1.2 23 0.3 169.6 3.9 53.9 6 20 17 4 15 35 19
PS 12c 67.27 0.370 16.60 2.89 1.06 0.044 3.52 7.68 0.044 0.149 99.62 0.73 c. 42 5 2 26 4 0.3 125 0.5 188 2.9 188.0 5.5 25.1 15 37 21 38 19 179 13
Oxides in wt%, trace elements in ppm. Total iron as Fe203(FeeO3 + FeOxI.III); mg = [100Mg/(Mg + Fe)], where Mg = MgO/40 and Fe = F203 x 0.9/72. Data presented on a volatile-free basis; LOI, loss on ignition at 1100°C.
amphibolites at temperatures and pressures characteristic of the upper mantle and were deformed and recrystallized with mylonitic textures in the amphibolite facies. These features are consistent with deformation and recrystallization of oceanic crustal rocks carried down in a subduction zone.
Origin of metamorphic rocks in the Darvel Bay Ophiolite Complex Protoliths of rocks which form the belt of metamorphic rocks in the Darvel Bay Ophiolite Complex extending westwards from Lahad Datu can be identified as mantle peridotites, cumulate pyroxenites and gabbros, isotropic gabbros, plagiogranites, doleritic and basaltic dykes, volcanics
273
and sediments. This association of rock types is characteristic of the ocean floor and its underlying mantle. The basic rocks have a tholeiitic composition and fall predominantly in the MORB field of discriminant diagrams, showing that they originated by partial melting of mantle peridotite in a mid-ocean ridge environment. Chemical analyses show depletion in HFS elements, enrichment in LIL and LREE, positive Sr and negative Nb and Ce anomalies, characteristic of island arc tholeiites, indicating that the mid-ocean ridge was generated above an active subduction zone. In map view (Fig. 1), in cross-section (Fig. 9) and also on an outcrop scale the rocks are distributed in lenticular slivers elongated parallel to the E - W trend of the metamorphic belt. The slab of mantle peridotite forming Silam Hill is juxtaposed along the northern shore of Darvel Bay against the 'Silumpat Gneiss', representing deformed cumulate gabbro, and with amphibolite and hornblende schist, representing isotropic gabbro or doleritic and basaltic dykes, from higher levels of the ocean crust. On the northern side of Silam Hill mantle peridotite is juxtaposed with felsic rocks, representing K-deficient silicic differentiates from tholeiitic magma which were intruded into oceanic crust at a high structural level. Chloritic schists at Jalan Silam and Lok Bikin, identified as deformed and recrystallized volcaniclastic rocks, represent tufts or hyaloclastics, and mylonitic quartz-schists, identified as banded cherts, represent the ocean floor and its sedimentary cover. All these rocks are now at the same structural level as the mantle peridotite. Most of the metamorphic rocks in the belt are foliated, sometimes intensely, to form schists, and frequently show mineral lineations. The foliation or schistosity is generally steeply inclined or vertical and the lineation is subhorizontal. Both foliation and lineation are orientated parallel to the E - W trend of the belt. Preferred orientation of minerals in the gneisses and schists defining foliation, schistosity and mineral lineation, and rotated inclusion trails in hornblende porphyroblasts in the tonalite from Pulau Sakar, indicate that the metamorphic rocks recrystallized syntectonically. Compositions of amphiboles and amphibole/plagioclase pairs indicate that this recrystallization occurred under high temperature but low pressure conditions in the amphibolite and greenschist metamorphic facies. No clear metamorphic gradation across the belt was recognized in our study, different facies being randomly juxtaposed (cf. Hutchison 1975). Basaltic dykes and felsic veins cutting the metamorphic rocks show varying relationships to the foliation and schistosity. Dykes and veins may be concordant to the foliation or schistosity in the country rocks, or may cut across these structures
(c)
(a)
1 mm 4
Fig.8. (a) Photomicrograph of chlorite schist, deformed crystal tuff, Jalan Silam Km 134, Darvel Bay, Sabah. (b) Photomicrograph of siliceous schist (metachert) with epidote bands and cross-cutting quartz veins, Kampong Diam, Lahad Datu, Sabah. (e) Photomicrograph of garnet amphibolite (Specimen EKc) showing a fractured garnet crystal with alteration to hornblende enclosed in a fine grained mylonitic hornblende-plagioclase matrix to form an augen structure. Pebble from Tungku River, Dent Peninsula, Sabah. Scale bar 1 mm.
~)
t~
275
DARVEL BAY OPHIOLITE, SABAH I SECTION
A-A'
]
Bukit Silam
South 3000
3000 Feet 2000
2000 1000
1000
0
0
Distance A-A'
~ 35 Km
] SECTION South 3000 " 2000- Silam Road 1000" ~
Bukit Silam ~ ~ ~
[
B - B' I
North _ SilamDam ~ ~ Sg. Telewas ~ e . . . . .
3000 Feet 2000 1000 0 Distance B-B' - 29 Km
I SECTION C-C' I South 1,,,~ tmo 0
P. Gifford
Kampong North Silam Kg. Sepagaya
South I000 0 Feet
0
i
SECTION D - D'[ dolerite dyke Pulau Sakar k, v~ L~.~"
North _ Veer 1000
o
D-D': ~ 17 Kml I Distance D-I3
[ Distance C-C': - 17 Km]
South [ SECTION E" E'i North 10IN) 1000 Feet 0 0 Sea-level [ Distance E-E' : ~ 11 Km I
Fig. 9. Schematic cross-sections across the metamorphic belt in the Darvel Bay Ophiolite Complex Sabah. Lines of section and key to ornaments are shown on Fig. 1.
with varying degrees of obliquity. Hutchison & D h o n a u (1971) reported a xenolith o f ' S i l u m p a t Gneiss' (i.e. foliated banded gabbro) e n c l o s e d in a basaltic 'sill' on Pulau Silumpat. Within the dykes, schistosity m a y be d e v e l o p e d with varying degrees of intensity; in general the m o r e concordant the d y k e the m o r e intense the schistosity within it.
These features indicate that s o m e of the dykes were intruded into rocks w h i c h had already been d e f o r m e d , but that d e f o r m a t i o n c o n t i n u e d after d y k e e m p l a c e m e n t . D e f o r m a t i o n o f the felsic veins is indicated by b o u d i n a g e where they are concordant, and folding w h e r e they are cross-cutting. B o u d i n a g e d and folded veins are frequently cut
276
s.A.K. O M A N G • A. J. BARBER
by veins which are completely undeformed. The relationships of basaltic dykes and felsic veins to the foliation and schistosity demonstrate that deformation and intrusion of igneous rocks were going on continuously during the development of the metamorphic belt. Metamorphic rocks associated with ophiolite complexes from many parts of the world have been attributed either to processes of deformation and metasomatism which affect ocean floor materials shortly after their formation in the mid-ocean ridge environment, or to the processes of subduction which lead to the emplacement of the ophiolite on a continental margin. It has been suggested that foliated rocks may be formed at the crust-mantle boundary in the region of a mid-ocean ridge by gravitational spreading of the ridge (Smewing et al. 1984; Gibbons & Thompson 1991). Foliated metamorphic rocks within ophiolites have also been interpreted as having originated along transform fault zones (Karson & Dewey 1978; Saleeby 1978; Simonian & Gass 1978; Prinzhofer & Nicholas 1980; Karson 1984). Foliated metamorphic rocks which occur along the basal thrust of an ophiolite complex have been interpreted as a 'metamorphic sole' formed during subduction and emplacement of the ophiolite on a continent (Davies 1971; Williams & Smyth 1973; Jamieson 1980; Searle & Malpas 1980; Spray & Williams 1980; Ghent & Stout 1981; Moores 1982). On the basis of the characteristics detailed above, the rocks of the metamorphic belt within the Darvel Bay Ophiolite are interpreted as having formed along a transform fault zone. Throughout the oceans, mid-ocean ridge systems are offset along transform faults at intervals of a few hundred kilometres. In these zones segments of ocean crust are moving past each other between active ridge segments. As the two crustal slices move past each other, oceanic materials along the fault are deformed, producing foliated peridotites, banded gneisses from cumulate gabbros and foliated and schistose amphibolites from gabbros and sheeted dykes. In transform fault zones at the time of their formation the foliation surfaces will be set vertically, parallel to the trend of the fault zone and lineation will be sub-horizontal representing the direction of movement along the fault. In an active fault zone crustal slices on either side of the fault will be of different ages. On one side of the fault the crust will have been formed recently at the ridge axis and will be at a high temperature; on the other the crust will be older and will have cooled and subsided as it moved away from the ridge axis. Deformation and shearing of the oceanic crustal rocks along the fault, with the ingress of water, will induce recrystallization of hydrous phases which are aligned to form schistose and linear structures
parallel to the sense of movement within the fault zone. At the time of formation the grade of metamorphism will decrease across the fault zone, from pyroxene and hornblende granulite against the hotter younger crustal segment, declining to greenschists and unmetamorphosed rocks towards the older colder segment and vertically towards the ocean floor. As has already been stated, no simple gradation of metamorphic facies has been recognized in the metamorphic belt at Darvel Bay. However, in an active fault zone any such simple arrangement is likely to be disrupted by continual movement along the fault, juxtaposing unrelated fault slices of different metamorphic grades and from different crustal levels, as has been described from Darvel Bay. No sedimentary infills, composed of clastic debris flows formed by submarine erosion, as reported from the Arakapas Transform in Cyprus (Simonian & Gass 1978; McLeod & Murton 1993) have been recognised as associated with the Darvel Bay Transform, suggesting that several kilometres have been removed by erosion since the emplacement of the ophiolite on Sabah. The mylonitic quartz-schists described from Kampong Diam, and identified as deformed banded cherts, are an unexpected component of an assemblage of rocks incorporated in a fossil transform fault zone. Cherts are deposited on the ocean floor only after it has subsided beneath carbonate compensation depth, some millions of years after the ocean crust formed at the mid-ocean ridge. The presence of chert among the deformed rocks indicates that a much older segment of ocean crust was juxtaposed against an active spreading ridge, implying that the transform fault extended over several hundreds of kilometres. Karson & Dewey (1978) came to a similar conclusion regarding the Coastal Complex transform associated with the Bay of Islands Ophiolite in Newfoundland, and pointed to present day examples where very long transform faults separate short ridge segments, in the Gulf of California and the Andaman Sea. Continued movements over a long distance through transtensional and transpressional fault segments increases the likelihood of the juxtaposition of uplifted mantle peridotite, with downfaulted crustal materials including supracrustal volcanics and metasediments, as seen in Silam Hill and along Jalan Silam. The garnet pyroxenites and garnet amphibolites from the Tungku and Pungulupi rivers described earlier in this account are not compatible with a transform fault origin. As previously concluded the high pressures and temperatures of formation of these rocks, combined with their mylonitic textures are consistent with their formation in a subduction zone. These rocks are therefore considered to
DARVELBAY OPHIOLITE, SABAH
277
Tectonic setting and emplacement of the Darvel Bay Ophiolite
subduction of Proto-South China Sea crust led to the obduction and uplift of the ophiolite by underthrusting of Upper Cretaceous and Lower Palaeogene turbidites of the Crocker Formation. Hutchison (1988), from the occurrence of Oligocene granitoid intrusions in the Long Laai area SW of Darvel Bay, suggests that the subduction of continental crust may also have been involved. Fragments of ophiolitic rocks in Eocene conglomerates show that the ophiolite had been obducted, uplifted and was subject to erosion by Eocene times (Newton-Smith 1967). It has been demonstrated in the foregoing account that the major occurrence of metamorphic rocks associated with the Darvel Bay Ophiolite represents a fossil transform fault, while fragments of high pressure metamorphic rocks in overlying volcanics indicate that the ophiolite is underlain by a metamorphic sole related to subduction and the obduction of the ophiolite in Sabah.
Evidence given in this account confirms the interpretation of the Darvel Bay Ophiolite Complex as a segment of oceanic crust and upper mantle originating at a mid-ocean spreading ridge. Evidence has also been given that the spreading ridge was developed above a subduction zone in a back-arc basin. The belt of metamorphic rocks, which occurs within the complex and extends westwards from Lahad Datu, is interpreted as a fossil transform fault. Radiometric ages and biostratigraphic ages from radiolarian cherts and foraminiferal pelagic limestones in the Chert-Spilite Formation indicate a Lower Cretaceous age for the origin of the ophiolite. Shallow-water limestones with Upper Cretaceous fossils, interpreted as the carbonate cappings to seamounts, indicate that the ophiolite still formed part of the ocean floor at this time. Hamilton (1979), Holloway (1981) and Rangin et al. (1990) have suggested that the Darvel Bay Ophiolite, together with other ophiolite fragments in eastern Sabah, originally formed part of a ProtoSouth China Sea crust. In this scenario, southward
The work described in this paper was presented as a thesis by SAKO for the award of the PhD degree of the University of London, sponsored by the Universiti Kebangsaan Malaysia (UKM) and the Government of Malaysia. Bulk rock geochemistry was determined in the geochemical laboratories at Royal Holloway, University of London using a Philips PW1480 XRF Spectrometer under the supervision of Drs M F Thirlwall and G F Marriner and Mr Ceil Jenkins. Mineral analyses were carried out on a JEOL Superprobe 733 electron microprobe at Birkbeck College under the supervision of Prof Robert Hall, University College and mineral compositions were recalculated using a suite of programs developed by Prof Hall who also gave general advice on the interpretation of the chemical data. K-Ar isotopic dating was carried out at the NERC Isotope Geosciences Laboratory at Keyworth, Nottingham under the supervision of Dr C. C. Rundle. Mr David Lee, Director of the Geological Survey of Malaysia, Kota Kinabalu, provided the samples of garnet amphibolite from the Tungku River used in this study. Attendance at the Conference on the 'Tectonic Evolution of Southeast Asia' held at the Geological Society, London was sponsored by UKM. Presentation has been greatly improved by careful and constructive reviews by Prof. C. S. Hutchison and Dr J. E. Dixon.
represent fragments of a metamorphic sole which must be present at depth below the Darvel Bay Ophiolite of the Dent Peninsula. A K-Ar age of 7 6 + 21 Ma obtained from garnet amphibolite during the present study coincides with the Late Cretaceous-Palaeogene age of subduction beneath the Darvel Bay Ophiolite inferred from the stratigraphic evidence. These metamorphic sole rocks, which must underlie the whole of the ophiolite complex, were intersected by Miocene volcanics and carried to the surface, to be incorporated with andesitic fragments, ophiolitic rocks and sandstones in volcanic breccias and conglomerates during the formation of the Sulu Volcanic Arc.
References AITCHISON, J. C. 1994. Early Cretaceous (pre-Albian) radiolarians from blocks in Ayer Complex melange, eastern Sabah, Malaysia, with comments on their regional tectonic significance and the origins of enveloping melanges. Journal of Southeast Asian Earth Sciences, 9, 255-262. BROWN,E. H. 1977. The crossite content of Ca-amphibole as a guide to pressure of metamorphism. Journal of Petrology, 18, 53-72. CLENNELL, M. B. 1991. The origin and tectonic significance of mrlanges in Eastern Sabah, Malaysia. Journal of Southeast Asian Earth Sciences, 6, 407--429.
DAVIES,H. L. 1971. Peridotite-gabbro-basalt complex in eastern Papua: an overthrust plate of oceanic mantle and crust. Bulletin of the Australasian Bureau of Mineral Resources, Geology and Geophysics, 128. DEER, W. A., HOWIE, R. A. & ZUSSMAN, J. 1966. An Introduction to the Rock-forming Minerals. Longman, London. DHONAU,T. J. & HUTCHISON,C. S. 1966. The Darvel Bay area, east Sabah, Malaysia. In: Geological Survey of Malaysia, Borneo Region, Annual Report for 1965. 141-160. FITCH, E H. 1955. The geology and mineral resources of part of the Segama Valley and Darvel Bay area,
278
s . A . K . OMANG • A. J. BARBER
colony of North Borneo. Geological Survey, Department of British Territories in Borneo, Memoir, 4. FLEET, M. E. & BARNETT, R. L. 1978. AliV/A1vi partitioning in calciferous amphiboles from the Frood Mine, Sudbury, Ontario. Canadian Mineralogist, 16, 527-532. GHENT, E. H. & STOUT,M. Z. 1981. Metamorphism at the base of the Semail Ophiolite, southeastern Oman Mountains. Journal of Geophysical Research, 86, 2557-2571. GmBONS, W. & THOMPSON,L. 1991. Ophiolitic mylonites in the Lizard Complex: ductile extension in the lower crust. Geology, 19, 1009-1012. HAMILTON,W. 1979. Tectonics of the Indonesian Region. United States Geological Survey Professional Paper, 1078. HOLLOWAY, N. H. 1981. The North Palawan Block, Philippines; its relation to the Asian mainland and its role in the evolution of the South China Sea. Geological Society of Malaysia Bulletin, 14, 19-58. HONNOREZ, J., MEVEL, C. & MONTIGNY, R. 1984. Occurrence and significance of gneissic amphibolites in the Vernh Fracture Zone, equatorial MidAtlantic Ridge. In: GASS, I. G., LIPPARD, S. J. & SHELTON, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 16, 77-94. HLrrCHISON, C. S. 1975. Ophiolite in Southeast Asia. Geological Society of America, Bulletin, 86, 797-806. 1978. Ophiolite metamorphism in northeast Borneo. Lithos, 11, 195-208. 1988. Stratigraphic-tectonic model for Eastern Borneo. Geological Society of Malaysia Bulletin, 22, 135-151. & DHONAU, T. J. 1969. Deformation of an alpine ultramafic association in Darvel Bay, East Sabah, Malaysia. Geologie en Mijnbouw, 48, 481-494. & -1971. An alpine association of metabasites and ultrabasic rocks in Darvel Bay, East Sabah, Borneo. Overseas Geology and Mineral Resources, Institute of Geological Sciences, Report, 10, 289-308. IRVINE, T. N. & BARAGAR,W. R. A. 1971. A guide to the chemical classification of the common igneous rocks. Canadian Journal of Earth Science, 8, 523-548. JAMIESON, R. A. 1980. Formation of metamorphic aureoles beneath ophiolites - evidence from the St.Anthony Complex, Newfoundland. Geology, 8, 150-154. 1981. Metamorphism during ophiolite emplacement - t h e petrology of the St. Anthony Complex. Journal of Petrology, 22, 397-443. KARSON, J. A. 1984. Variations in structure and petrology in the Coastal Complex, Newfoundland: anatomy of an oceanic fracture zone. In: GASS, I. G., LIPPART, S. J. ~¢ SHELTON, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publication, 13, 131-143. & DEWEY, J. F. 1978. Coastal Complex, western Newfoundland: An Early Ordovician oceanic -
-
-
-
-
-
-
-
-
-
fracture zone. Geological Society of America, Bulletin, 89, 1037-1049. KIRST, P. 1976. Petrology of the metamorphic rocksfrom
the equatorial Mid-Atlantic Ridge and Fracture Zone. PhD Thesis, University of Miami. KOOPMANS,B. N. 1967. Deformation of the metamorphic rocks and Chert-Spilite Formation in the southern part of the Darvel Bay area, Sabah. Geological
Survey of Malaysia, Borneo Region, Bulletin, 8, 14-24. LEON6, K. M. 1971. Peridotite gabbro problems, with special reference to the Segama Valley and Darvel Bay area, Sabah, East Malaysia. Geological Society of Malaysia, Newsletter, 28, 4-13. -1974. The geology and mineral resources of the
Darvel Bay and Upper Segama area, Sabah. Geological Survey of Malaysia, Memoir, 4 (revised). 1977. New ages from radiolarian cherts of the Chert-Spilite Formation of Sabah. Bulletin of the Geological Society of Malaysia, 8, 109-111. MCLEOD, C. J. & MLrRTON, B. J. 1993. Structure and tectonic evolution of the southern Troodos Transform Fault Zone, Cyprus. In: PRITCHARD, H. M., ALABASTER,T., HARRIS, N. B. W. & NEARY, C. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publication, 76, 141-176. MITCHELL, A. H. G., HERNANDEZ, E • DE LA CRUZ, A. E 1986 Cenozoic evolution of the Philippine Archipelago. Journal of Southeast Asian Earth Sciences, 1, 3-22 MOORES, E. M. 1982. Origin and emplacement of ophiolites. Review of Geophysics and Space Physics, 20, 735-760. MORGAN, B. A. 1974. Chemistry and mineralogy of garnet pyroxenites from Sabah, Malaysia. Contributions to Mineralogy and Petrology, 48, 301-314. NAKAJIMa, T., BANNO, S. & SUZUKI, T. 1977. Reaction leading to the disappearance of pumpellyite in low grade metamorphic rocks of the Sanbagawa metamorphic belt in central Shikoku, Japan. Journal of Petrology, 24, 263-284. NEWtON-SMITH, J. 1967. Bidu-Bidu Hills, area, Sabah, East Malaysia. Geological Survey of Malaysia, Kuching, Report, 4. OMANG, S. A. K. S. 1993. Petrology, geochemistry and -
-
structural geology of the Darvel Bay Ophiolite, Sabah, Malaysia. PhD Thesis, University of London.
PEARCE, J. A. & CANN,J. R. 1973. Tectonic setting of basic volcanic rocks determined using trace element analysis. Earth and Planetary Science Letters, 19, 290-299. , HARRIS, N. B. W. & TINDLE, A. G. 1984a. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25, 956-983. - - , LIPPARD,S. J. & ROBERTS,S. 1984b. Characteristics and tectonic significance of supra-subduction zone ophiolites. In: KOKELAAR, B. P. t~ HOWELLS, M. E (eds) Marginal Basin Geology: Volcanic
and Associated Sedimentary and Tectonic Processes
DARVEL BAY OPHIOLITE, SABAH
in Modern and Ancient
Marginal Basins.
Geological Society, London, Special Publication, 16, 77-94. PLYUSNINA, L. E 1982. Geothermometry and geobarometry of plagioclase-hornblende bearing assemblages. Contributions to Mineralogy and Petrology, 69, 33--47. PRINZHOFER, A. & NICHOLAS, A. 1980. The Bogota Peninsula: a possible oceanic transform fault. Journal of Geology, 88, 387-398. RANGIN, C. 1989. The Sulu Sea, a back-arc basin setting within a Neogene collision zone. Tectonophysics, 161, 119-141. ~, BELLON, H., BERNARD, E, LETOUZEY,J., MULLER, C. & SANUDtY, T. 1990. Neogene arc-continent collision in Sabah, Northern Borneo (Malaysia). Tectonophysics, 183, 305-319. REINHARD, M. & WENK, E, 1955. Geology of the Colony of North Borneo. Geological Survey, Department of the British Territories in Borneo, Bulletin, 1. SALEEBY,J. 1978. Kings River ophiolite, southwest Sierra Nevada foothills, California. Geological Society of America, Bulletin, 89, 617-636. SEARLE, M. P. & MALPAS,J. G. 1980. Structure and metamorphism of rocks beneath the Semail Ophiolite of Oman and their significance in ophiolite obduction. Transactions of the Royal Society of Edinburgh, 71, 247-262.
279
SIMONIAN, K. O. & GASS, I. G. 1978. Arakapas fault belt, Cyprus: A fossil transform fault. Geological Society of America, Bulletin, 89, 1220-1230. SMEWING,J. D., CHRISTENSEN,N. I., BARTHOLOMEW,1. D. 8,: BROWNING, P. 1984. The structure of the oceanic upper mantle and lower crust as deduced from the northern section of the Oman ophiolite. In: GASS, I. G., LIPPARD, S. J. & SHELTON, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publication, 13, 41-53. SPEAR, E S. 1981. An experimental study of hornblende stability and composition variability in amphibole. American Journal of Sciences, 281, 697-734. SPRAY, J. G. • WILLIAMS,G. D. 1980. The sub-ophiolitic rocks of the Ballantrae Igneous Complex, SW Scotland. Journal of the Geological Society, London, 137, 359-362. SUN, S-S. & NESBITT, R. W. 1978. Geochemical regularities and genetic significance of ophiolitic basalts. Earth and Planetary Science Letters, 44, 119-138. WILLIAMS,H. & SMYTH,R. 1973. Metamorphic aureoles beneath ophiolite suites and Alpine peridotites: tectonic implications with west Newfoundland examples. American Journal of Sciences, 273, 594-621.
Role of pre-Tertiary fractures in formation and development of the Malay and Penyu basins KHALID
NGAH,
MAZLAN
MADON
& H. D. T J I A
PETRONAS Research & Scientific Services, Lot 1026 PKNS Industrial Estate, 54200 Hulu Kelang, Malaysia Abstract: Major faults in Sundaland trend NNW to NW, WNW, N and E. Some of the NNW to NW and N-striking faults across Mesozoic areas of the Malay Peninsula were active until midEocene time. Small, fault-bounded Tertiary basins onshore may be pull-apart basins associated with such faults. Mainly from seismic data, NNW to NW, N and E-striking faults have been recognized in the pre-Tertiary basement of the Malay and Penyu basins off the east coast of the peninsula. These faults were reactivated before the Late Oligocene and during the Middle to Late Miocene. N-striking faults in pre-Tertiary areas are common throughout Sundaland. In the field, these faults are found to be the oldest (possibly Jurassic) regional fractures. The regional NNW-NW and WNW fractures are believed to have originated as strike-slip faults when the peninsula was subjected to late Mesozoic deformation. The onshore E-W faults were probably extensional fractures that developed as secondary structures associated with sinistral slip motions along NNW-NW faults. Upper Cretaceous dolerite dykes fill some of the E-W fractures. NWstriking basement faults of the Malay basin continue onshore SE Asia as the Three Pagodas fault zone. Initially these were sinistral basement wrench faults creating secondary E-W extensional fractures. In the Middle to Late Miocene the regional stress field changed, resulting in reversal of slip movement along major wrench faults and structural inversion of the sedimentary basins. This inversion is manifested as E-W anticlines located over half-graben. In the Penyu basin similarly striking half-graben probably developed in the same fashion. There, the NW-striking Rumbia fault divides the basin into two parts. Half-graben in the western part remained orientated E-W, but those in the eastern part became rotated clockwise by continued left-lateral slip along the Rumbia fault. After the Miocene the two basins continued to subside, developing an almost undisturbed blanket of post-Miocene sediments. Locally, residual stress caused some of the structures to grow.
The Malay and Penyu basins are Cenozoic sedimentary basins located in the South China Sea and are floored by continental lithosphere (Fig. 1). This southern portion of the South China Sea belongs to tectonically semi-cratonic Sundaland. Geologically, Sundaland comprises the entire Sunda Shelf and contiguous land areas of Borneo, Java, Sumatra and peninsular Malaysia. The margins of Sundaland have experienced young crustal movements and are tectonically transitional to the mobile belts bordering the region in the east, south and west. The Malay basin is 500 km long in a northwest direction, about 200 km wide, and contains more than 12 km of Oligocene and younger sediments. The older sediments are of terrestrial origin with minor marine intervals; holomarine conditions have prevailed only since the latest Miocene. Oil-prone areas are situated in the southern and central parts of the Malay basin, while the northern part is gas-prone. Hydrocarbon prospects have also been located in the Penyu basin and non-commercial oil discoveries have been made. This basin has a roughly sub-circular shape with a diameter of about 350 km and contains up
to 6 km of Cenozoic sediments that have been interpreted to have been deposited in an environment generally similar to those of the Malay basin. The N W to N N W elongation of the Malay basin is roughly parallel to the elongation of the Malay Peninsula to its west, while the more equant shape of the Penyu basin suggests its shape to be strongly controlled by fractures. The structural history of these two basins comprised a tensional period when the basins began subsiding before the Late Oligocene. This subsidence continued during the late Early Miocene to Late Miocene, accompanied by regional compression during the Middle to Late Miocene, and finally quiet tectonic subsidence in the Plio-Pleistocene. These generalized events have been described by Khalid Ngah (1975), Ng (1987) and Md. Nazri Ramli (1988). Figure 2 is a correlation table of the stratigraphy in the Malay, Penyu and the West Natuna basins simplified from various unpublished and published reports. The oldest known sediments within the Malay basin are Upper Oligocene. Their ages were determ i n e d from pollen from cuttings from wells located in the basin's border zones, while their
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 281-289.
281
282
K. NGAH ET AL.
VIETNAM
BASEMENT NORTH SUNDALAND CON SON SWELL
•
~iiiiiiii
TERENGGANU~
1
2
3
4
5
900 1200
;:: i . ? :
......
: :~::il~'ltl
O~iiiiiiiiiii!!
BASIN
I!i¸ iiiiiiiiiiiiiii iiii!!;iililii!iiii!i!ii?' iiii:iii i i i i~ii!i~i'!iiii~i i ! !
:
::
! F-~.UkSIN IA!
PENINSULAR MALAYSIA
:: i: :
SOUTH CHINA SEA i
DRM
Fig. 1. Outline of the Cenozoic Malay, Penyu and West Natuna basins of northern Sundaland, SE Asia. Pre-Tertiary basement depths are shown in two-way travel time, for which an approximate conversion curve is given in the inset. Heavy lines are major faults.
interpreted equivalents in the deeper part of the basin were extrapolated using seismic profiles. Seismic profiles across the basins suggest that a thick sequence of older, possibly mainly Cenozoic, sediments still exists beneath the interpreted Upper Oligocene strata. Felsic granitic rocks in the Natuna, Anambas and Tambelan islands yield Upper Cretaceous radiometric dates ranging between 90 Ma and 70 Ma (Katili 1973). The ages of granitic (granites to granodiorites; Areshev et al. 1992) basement of the Mekong basin and the adjacent shelf range from 9 7 M a to 178Ma, and most are around 100-110 Ma or Albian. This basement is also part of Sundaland. Upper Cretaceous-lower Palaeogene gramtes have been determined onshore in the Bana, Ankroet and Deo-Ka complexes, and Areshev
et al. (1992) believe that such younger intrusive rocks may well exist offshore but have yet to be discovered. Heatflow studies in the Malay basin and other well data from the West Natuna basin show above normal values with certain areas possessing high values over 140 mW/m 2 (Mohd Firdaus Abdul Halim 1993; Wan Ismail Wan Yusoff & Swarbrick 1994). One area of high heatflow coincides with the triple junction of the Malay, Penyu and West Natuna basins. Pre-Tertiary basement is topographically high in a wide region around the triple junction and has been postulated to represent the remnants of a Malay Dome (new name; Tjia 1995). It has also been suggested that the centre of the dome was situated above a mantle plume. The widespread occurrence of Upper Cretaceous felsic
FORMATION AND DEVELOPMENT OF MALAY & PENYU BASINS
PULAI NO, 1 T
EPMI
I PILONG
!
!
BEKOK
[
UPPER SANDCOAL
i
V IV
l
TAPIS ---
~--LOWER SAND-COAL
"
....
PETRONAS CARIGALI
i
and
J
i
PILONG
[
MIOCENE
CONOCO
PENYU BASIN VIII VII Vl
RECENT PLIOCENE
1
BASIN
MALAY
AGE (Approximate)
283
i I,
i MUDA
PILONG
J J
PARI i
K
WEST NATUNA BASIN
i
L
[
u
_i J i
PULAI
[- TERENG~:3ANU _-SHALE
SELIGI
l
TAPIS
liB IIA
TERENGGANU
i
................................... L
BARAT , UDANG GABUS
•
PENYU
OLIGOCENE
N
i
Fig. 2. Stratigraphic correlation table of the Malay, Penyu and West Natuna basins from published and unpublished data. EPMI; Esso Production Malaysia Inc.; Conoco, Continental Oil Company. Unpublished material is held by Petronas.
granites and their regional distribution suggest that the Malay Dome has a radius of c. 500 km. The present paper addresses the role that preTertiary structures have played in the formation, delineation, and development of the Malay and Penyu basins.
Fracture patterns in the pre-Cenozoic basement Figure 3 shows the fracture patterns onshore in peninsular Malaysia and offshore in the Malay and Penyu basins. On land in peninsular Malaysia, the north-south trending Bentong (-Raub) suture is the limit to a western region of Gondwana origin. This Gondwana fragment became attached to the eastern portion of the peninsula by Early Triassic time (Metcalfe 1988; Tjia 1989a). The eastern portion of the peninsula was already part of the Asian continent when the western region docked with it along the Bentong suture. The pattern of detectable strike-slip motions on the major onshore faults indicates that the latest motions were in response to regional compression with a maximum horizontal stress orientated between ENE-WSW and E-W. This is roughly perpendicular to the strike of regional structures onshore. Approximately N-NNE-striking faults (such as the Kelau fault in Pahang, vertical faults within the Bentong Suture
zone, and two unnamed north-south faults on Pinang Island) moved right-laterally. In contrast, the NNW, NW and WNW-trending faults are leftlateral strike-slip faults. These include the Lebir fault in Kelantan; the Lepar fault, Pahang; the Baubak fault, Kedah; the Bukit Tinggi and Kuala Lumpur fault zones in Selangor/Perak; and the Mersing fault bundle in Johor. Details of these faults have been discussed elsewhere (Tjia 1989b). Several small Tertiary basins in the peninsula are located close to or within large strike-slip fault zones. The Lawin basin in Perak seems to be associated with the Baubak (older name: Bok Bak; Burton 1965) fault zone and after detailed studies, Syed Sheikh Almashoor (pets. comm. 1994) has placed the basin within the fault zone. Another basin, the Batu Arang basin in Selangor, is tangential to the Kuala Lumpur fault zone. Pollen studied by PRSS biostratigraphers indicate Middle Eocene basal beds in this basin. A recent study of major peninsula faults by Zaiton Harun (1992) showed that cataclastic to mylonitized material in certain fault zones is Upper Cretaceous to Middle Eocene (53-46 Ma). The regional map prepared by the Geological Survey of Malaysia (1985) shows uppermost Cretaceous (67 Ma) granites at the Kelantan-Perak border, at Gunung Ledang (also known as Mount Ophir), Johor and a few smaller felsic intrusive bodies in the central belt of the peninsula and in southern Johor. The evidence cited
284
K.
NGAH
ET AL.
$
I 105
102
ci
:!ii~:, .
Plnang Island
FAULT
PATTERNS
l!'i '~
~'!' II "~,! ii'!!li[ .",
' ~
I
', ,i~ ' ' ~i ", , ' , iI , ~' i
'
, '
Basement contours in km below sea level
loo I
lol J
lO2 ~ ~ ,= !~: ,li,,l~,~i~ lo3 ,, II1]~~11 L,h ~'..... ,04 , :,llii::Hi [~I;U,I~Ui!;U~ ] 1,,, ~
~rzo,'v~
lo5 1
Fig. 3. Fault patterns in the pre-Cenozoic basement of peninsular Malaysia and offshore in the Malay and Penyu basins. Faults onshore are from Tjia (1989b); those in the offshore basins were derived from unpublished reports held by Petronas. The dashed lines within the Malay basin represent major basement fractures which have been inferred from structural styles in sediments overlying the basement and from trends of magnetic/gravity anomalies.
above strongly suggests that a regional diastrophic event took place in the region at about the Cretaceous-Tertiary boundary. At that time the entire peninsula, and most probably the entire area of Sundaland to which it geologically belongs, was already a cratonized basement. Upper Jurassic-Cretaceous sediments (Gagau Group) were not folded but were only tilted ten degrees or less, and regional compressive stresses became decoupled along major fractures. These major fractures moved mainly as strike-slip faults. The fractures in the offshore area shown on Fig. 3 were compiled from various unpublished sources held by PETRONAS. Major basement fractures within the Malay basin have been inferred from structural styles in sediments overlying the basement and from trends of magnetic/gravity anomalies. In general, different parts of the offshore area display different dominant fracture directions. In the Penyu basin, and in the south end and axial parts of the Malay basin, the east-west and NW-SE fractures are dominant, whereas in the north end of the Malay basin, north-south fractures are
dominant. NNW and NW-trending fractures are evident along the borders of the Malay basin, with some NE-trending fractures in its southeast corner. Figure 3 shows that strikes of major fractures in the offshore regions mimic those mapped onshore the peninsula, and the authors conclude that the offshore fault strikes are of pre-Tertiary origin.
Major structures in the Malay and Penyu basins
Malay basin and vicinity Folds in the Malay basin generally trend east-west. Exceptions are along its southwest margin and in its northwest corner where anticlines trend NW-SE or almost N-S, respectively. These strikes seem associated with the NW-striking Western Hingeline fault zone (WHL), and the north-striking Kapal-Bergading tectonic line, respectively (Fig. 4). Known major fault zones are the WHL that marks the southwest boundary of the basin; the
FORMATION AND DEVELOPMENT OF MALAY • PENYU BASINS
i
I
285
I
"%
o,.-'
•
..
%.
Ser gadlng
::......
~,,,
~ESAH
7"__
fz
~'OO~,'ff6" ........
iI
~%
° "
.
,outt zo~k ~uNm
nln )n,.,L
• L.arut
"'
1 i
!
~,
...... 4/,'%_ ~ ........ ~
~,\
.........
n,~,,~,
~
cC,,~.
;
l\
!
I:['~~i,, ~
wm_,,
I I
b ~
"%~,
i
gun
:.
~ !
Dungun
++
i
+
+ 104' I
/
!
' AA,~ n d idni ¢ng
/:
v'ooL iong ]
~1~ Hydrocarbonfields +++ Fetsic grani toids in basement
ii o
~Pulai
. . . . . . . .. . . . . . . .
Angsi ~ i ............... X ~
I /
W..W'... . . . . . . . . . . . . "q)
.;/l'~-
+ o+.°
~=_
7 / ..
,ooo (,
,.05" i
Fig. 4. Major fault zones and hydrocarbon areas in the Malay basin. Note the orientations of HC fields and their apparent displacements across north-south fault zones. Known occurrences of felsic granites in the basement are also shown. K a p a l - B e r g a d i n g tectonic line; north-trending Dulang, Bundi and Mesah zones; NNW-striking D u n g u n zone that includes the D u n g u n Graben; the N W Selambau zone; and an u n n a m e d NW-zone along part of the basin's northeast margin.
There are half-graben and graben along many of the offshore fault zones. Their shapes, orientations of bounding fractures, and the positions of basins along or within these major fault zones indicate that the basins are pull-apart basins produced by
286
K. NGAHETAL.
wrenching. The pull-apart basins became depocentres of thick sedimentary sequences. Strike-slip reversals along the major fault zones are demonstrated by the fact that the sediments within the pull-apart basins were compressed into en echelon folds whose orientations are consistent with lateral fault slip in directions opposite to those during the formation of the pull-apart basins (see Tjia & Liew 1996). In cross sections, change in the regional stress field from initially (trans)tensional to subsequent (trans)pressional is shown by inverted anticlines. The later stress field produced anticlinal bulges and domes where the sedimentary columns are thickest, that is, in the half-graben. In other
words, the orientation and position of the folds were pre-determined by those of the basement faults. Seismic profiles show that the E-W anticlines in the axial zone of the Malay basin are located over E-W half-graben. Figure 5 indicates that many of the half-graben are en echelon. Based on this observation, it is suggested that these staggered patterns are associated with NW-striking basement faults. Furthermore, these faults are postulated to represent a major NW-trending basement fault zone along the axis of the basin. Towards the northwest, this basement fault zone (named as the Axial Malay fault; Tjia 1995) most probably continues as the
( j
,< :
0 I-
.\ \
\
t00 km I
7~0~
\
\
6%5
\
7 [ \\
Ma laysia I
-
1
~
~
~
'
\
I
Fig. 5. En echelon faulted sub-basins in the axial zone of the Malay basin are interpreted to have developed through sinistral wrenching along the newly interpreted Axial Malay fault zone. Figure has been adapted from various sources held by Petronas.
FORMATION AND DEVELOPMENT OF MALAY & PENYU BASINS Three Pagodas fault. Sinistral slip along the Axial Malay fault produced the east-west half-graben before the Late Oligocene. Strike-slip reversal during the Middle to Late Miocene compressed the sub-basin fills into east-west anticlines. The major N-S faults of the Malay basin delineate tectonic domains of different structural styles and/or different deformation history, and have laterally displaced crustal blocks. The distributions of oil and gas fields (structural domes) across the Kapal-Bergading Line and across each of the other three N-S fault zones (Dulang, Bundi and Mesah) appear displaced dextrally by several tens of kilometres (Fig. 4). The position and orientation of these domes have been determined by those of the basement fractures. The stratigraphic record of the basin suggests that compressive tectonic deformation took place in the Middle to Late Miocene, during which time right-lateral wrench motion could have occurred along the major N-S fault zones. The regional maximum compressive stress was most probably orientated between NE-SW and ENE-WSW. This orientation of regional compression is also consistent with the observed strike-slip reversals along the WHL and Dungun fault zones.
287
Penyu basin The pre-Tertiary basement highs that frame the Penyu basin include (clockwise on Fig. 1): the Tenggol Arch (N and NE), Johor Platform (S), peninsular Malaysia (W) and Pahang Platform (NW). The Tenggol Arch separates the Penyu basin from the Malay and West Natuna basins. Sedimentary cover over the basement highs is thin, except for the Tenggol Arch that subsided somewhat and is capped by more than a kilometre of Tertiary sediments. Basement faults in NW-SE and E-W directions have compartmentalized the basin into ten half-graben. The basin is further divided by the NW-striking Rumbia fault into a west part and an east part. Half-graben and fault scarps strike E - W in the west, but strike WNW in the east part. The Rumbia fault is over 100 km long and its heave is more than 2 km. Left-lateral separation by the fault of once-contiguous structures is believed to exceed 20 kin. En echelon sigmoidal fractures near the fault support this left-lateral slip interpretation (Fig. 6). The E - W striking fractures and halfgraben are interpreted as secondary structures that resulted from left-lateral slip along the Rumbia fault and other, as yet undiscovered NW-trending
WRENCH
,,~!~..~,,, " ~ ~
DEPRESSION WITH DIRECTION OF DEEPENING
2;iii;iii;;;;(:~'El~Ttl~i;11i FAULT SCARP BASEMENT HIGH i!i~iii!i!ii~ii!i!!ii!~.
PARI 1 0
PENYU BASIN BASE OF TERTIARY
ii.~';iiiiiii i iiii iiii" ~
....... ~.~'!!i~i
~
.¢,:
~'~"~""
~"~i:
_.~,,.
10 km
Fig. 6. Plan of major basins and faults at the base of the Tertiary in the Penyu basin. The Rumbia fault separates E-W structures in the western part from WNW structures. En echelon sigmoidal fault pattern (top centre) suggests wrenching. Simplified from an unpublished map by Texaco (1992) and held by Petronas.
288
K. NGAH E T A L .
wrench faults in the basement. Continued left-slip movement along the Rumbia fault is postulated to have rotated the E-W faults and basins in the eastern part of the basin into their present W N W ESE orientation, with rotation of up to 25 ° (Fig. 7). There are anticlinal features over the deepest parts of the corresponding half-graben but the effect of compressional tectonics in the Penyu basin is less obvious compared to that in the Malay basin. A regional unconformity at the top of the Penyu Formation (see Fig. 2) represents tectonic uplift at the end of Oligocene time. Subsidence in Early Miocene time began with deposition of the Pari Formation consisting of alternating sandstone and shale. Unpublished palaeontological information from a study conducted by TEXACO in 1992 dates a transpressional event in mid-Miocene.
Discussion and conclusions The structural history and its relationship to regional tectonics are postulated to be as follows. (1) The Malay and Penyu basins originated during the Late Cretaceous as two of the three rift arms that developed above a continental crustal dome located over a mantle plume. The West Natuna basin is the third rift arm. (2) The main strike of the Malay basin follows the NNW-NW trending tectonic grain of this part
of Sundaland. The northern part of the basin is controlled by N-S fractures that became especially dominant farther north in the Gulf of Thailand. Geological evidence from onshore peninsular Malaysia indicates that the NNW-NW tectonic grain was established in Late Triassic-Early Jurassic time. In the mid-Eocene, parts of continental SE Asia experienced expulsion as a consequence of collision between the Indian plate with the Eurasian plate. Fractures parallel and subparallel to the tectonic grain in the basement of the basin moved as wrench faults. A bundle of long NW-SE fractures along the axis of the offshore sub-basins (the Axial Malay fault zone) became an extension of the Three Pagodas zone. Through left-lateral motion of the Axial Malay fault zone, en echelon east-west tension faults and half-graben developed. At that time dextral wrench movement along the WHL and its major fault splay, the Dungun zone, resulted in the formation of smaller pull-apart basins. (3) The Penyu basin is dominated by NW-SE and E - W structures. The NW-SE structures are sub-parallel to the Late Triassic-Early Jurassic tectonic grain of the Malay Peninsula. The Rumbia fault zone consists of NW-SE fractures and divides the basin into two parts. In the western part, E - W trending half-graben and faults are dominant, whereas on the east the dominant trend is W N W ESE. It is interpreted that this strike deviation is
SOUTHWEST
NORTHEAST
EFt
CheratingRumbia- 1
Rhu-
IIIA
1
E) c~ Alu - Alu E- I
Fig. 7. Schematic cross-section of the Penyu basin. Note half-graben, mild inversion structures at Rumbia-1 and at Cherating-1. From unpublished material by Texaco (1992) and held by Petronas.
FORMATION AND DEVELOPMENT OF MALAY & PENYU BASINS the result of clockwise rotation of initially E - W trending half-graben and other structures through continued sinistral slip along the Rumbia fault. These E - W structures are believed to have developed within and across fault-bounded, N W - S E elongated crustal blocks, which slipped sinistrally during differential expulsion of SE Asia. Expulsion began in mid-Eocene, when the Indian plate started to push into the Eurasian plate. (4) By the late Early Miocene, the Australian continent had approached SE Asia sufficiently closely as to obstruct free expulsion of its crustal fragments. In addition, the effect of the westward drive of the Pacific plate also began to be felt in this region. Before that time, active N - S spreading in the South China Sea basin and the Philippine basin apparently prevented stress generated by the Pacific plate motion to reach the SE Asian region. From the late Early Miocene to Late Miocene, the combined effect of northward and westward convergence of the Indian-Australian and Pacific plates manifested themselves in a regional compressive regime and reversals of slip sense along several of the N W - S E striking wrench faults. Some examples of the resultant effect of this movement are in the Malay basin, where dextral slip occurred along its Axial
289
Malay basement faults, accompanied by compression that developed anticlinal features forming oil- and gas-bearing structures. Unpublished data from EPMI-acreage in the southeastern part of the basin suggest thrusting towards the south. These thrust faults and the anticlines trend E-W, that is, the orientation determined by the basement fractures. Slip reversals along the Dungun fault zone and other N - S to N W - S E trending regional faults which inverted the pull-apart sedimentary in fills are also indicative of the movement. The axes of the inversion were orientated in a sense consistent with the lateral slip sense of the associated wrench fault. The Rumbia fault does not seem to have experienced slip reversal. On the contrary, continued sinistral slip appears to have rotated structures in its eastern part clockwise for 25 ° . Since the mid-Miocene, continued sinistral strikeslip on the Rumbia fault most probably negated the effect of most of the compressive stress exerted by the convergence of the plates. This stressdecoupling explains the milder deformation observed in the Penyu basin. We are grateful to C. S. Hutchison and R. E. Swarbrick for constructive comments on an earlier version of the paper.
References ARESm~V, E. G., TRAN LE DONG, NGO THUONG SAN SHNm, O. A. 1992. Reservoirs in fractured basement on the continental shelf of southern Vietnam. Journal of Petroleum Geology, 15, 451-464. BURTON,C. K. 1965. Wrench faulting in Malaya. Journal of Geology, 73, 781-798. GEOLOGICAL SURVEY OF MALAYSIA. 1985. Geological Map of Peninsular Malaysia, 1:500,000. Ipoh, Geological Survey of Malaysia, 8th edition. KATILI,J. A. 1973. Geochronology of west Indonesia and its implications on plate tectonics. Tectonophysics, 19, 195-212. KHALID NGAH. 1975. Stratigraphic and Structural Analyses of the Penyu Basin, Malaysia. MSc Thesis, Oklahoma State University. MD. NAZRI RAMLI. 1988. Stratigraphy and paleofacies development of Carigali's operating areas in the Malay Basin, South China Sea. Bulletin Geological Society of Malaysia, 22, 153-187. MOHD. FIRDAUSABDUL HALIM. 1993. Heat flow in the Malaysia sedimentary basins. In: Petronas Research & Scientific Services Sdn. Bhd., Proceedings Exploration Research Seminar. I, Progress Reports and Findings. 11-25. METCALFE, I. 1988. Origin and assembly of Southeast
Asian continental terranes. In: AUDLEY-CHARLES, M. G. & HALLAM,A. (eds) Gondwana and Tethys. Geological Society, London, Special Publication, 37, 101-118. NG. T. S. 1987. Trap styles of the Tenggol Arch and the southern part of the Malay Basin. Geological Society of Malaysia Bulletin, 21, 177-193. TJIA, H. D. 1989a. Tectonic history of the BentongBengkalis suture. Geologi Indonesia, 12, 89-111. 1989b. Major faults of Peninsular Malaysia on remotely-sensed images. Sains Malaysiana, 18, 101-114. 1995. Inversion tectonics in the Malay Basin: evidence and timing of events. Geological Society of Malaysia Bulletin, in press. & L~EW, K. K. 1996. Changes in tectonic stress field in northern Sunda Shelf basins. This volume. WAN ISMML WAN YUSOFF & SWARBRICK,R. E. 1994. Thermal and pressure histories of the Malay Basin, offshore Malaysia [abstract]. AAPG Bulletin, 78, 1171. ZAITON HARUN. 1992. Anatomi Sesar-sesar Utama Semenanjung Malaysia. PhD Thesis, Universiti Kebangsaan Malaysia, Bangi.
Changes in tectonic stress field in northern Sunda Shelf basins H. D. T J I A & K. K. L I E W
PETRONAS Research & Scientific Services, Sdn. Bhd., Lot 1026 PKNS Industrial Estate, 54200 Hulu Kelang, Malaysia Abstract: The Tertiary basins of the northern Sunda Shelf are underlain by normal and
attenuated continental crust that is characterized by moderate to high average geothermal gradients in excess of 5°C/100 m. In the Malay basin, upper Oligocene and younger sediments are more than 12 km thick; in the other basins, such sediments are between 4 and 8 km thick. The Malay, Penyu and West Natuna basins are aulacogens meeting at a triple junction that marks a Late Cretaceous hot spot in the centre of the Malay Dome. Sub-basins (commonly haif-graben) developed as pull-apart basins within regional, north to northwest-striking, wrench fault zones. The NW-striking Three Pagodas fault most probably extends farther southeast as the Axial Malay fault in the basement along the length of the Malay basin. Pre-late Oligocene sinistral slip along this fault developed east-west half-graben that fixed the position and orientation of inverted anticlines that later developed in the basin-filling sediments through strike-slip reversal along the fault. Initial basin subsidence took place during Eocene-Oligocene time. Regional tensional conditions prevailed until the early Miocene. During the middle to late Miocene, regional compressionai stresses caused reversals of the sense of motion on the major wrench faults, and structural inversion of the basin-filling sediments. On some of the north-striking wrench faults there are indications of up to 45 km of right-lateral displacement which is possibly postMiocene. The regional wrench faults have acted as domain boundaries with each tectonic domain characterized by different stress fields. The stress systems evolving during the Cenozoic are attributed to varying degrees of interference of plates coupled with changes in convergent directions and/or rates of motion of the Pacific plate, the Indian Ocean-Australian plate, and continued, differential extrusion of SE Asian crust following the collision of the Indian plate with the Eurasian plate.
Sundaland is the geological province in SE Asia which comprises pre-Tertiary basement rocks of the Sunda Shelf and contiguous land areas of western Borneo (or West Kalimantan plus westernmost Sarawak) and peninsular Malaysia. The northern Sunda Shelf extends south to the Johor Platform, and north to a line connecting Kota Bharu with the SW tip of Vietnam (Fig. 1). Most of Sundaland is tectonically stable, but possible Late Cenozoic vertical crustal movements have been detected in certain areas. One such area is onshore Johor where uplift reached several scores of metres and where block tilting has been recognized (Anizan Isahak 1993). The western edge beneath the petroleum basins of central and south Sumatra have experienced downwarping. On the shelf proper, net subsidence of several kilometres created the Malay, Penyu, West Natuna, Nam Con Son and Mekong basins. This subsidence was accompanied by stretching and attenuation of the continental crust. Average geothermal gradients of 51.8°C/km in the Malay basin are consistent with thinner than normal crust (Mohd. Firdaus Abdul Halim 1993). Tapponnier et al. (1982) suggested that the formation of the basins in continental SE Asia resulted from collision of the Indian plate with the
Eurasian plate. The collision that began 45 Ma ago has pushed out crustal blocks of SE Asia towards the southeast along major wrench faults. This extrusion took place in stages with one block moving earlier than the other. It is suggested that this extrusion and associated wrench-faulting created a tensional environment in continental SE Asia which resulted in the formation of basins floored by continental crust. By mid-Miocene time, the stress field changed into a compressional regime. The reason was, according to one suggestion, that the northward progression of the Indian-Australian plate blocked further expulsion. Harder et al. (1992) agreed that the continuing northward push by India had been one of the causes that changed the stress system in SE Asia throughout most of the Tertiary period. They further believed that changes also occurred through rotation of the extruded fault-bounded crustal blocks. We propose that the change from a tensional regime into a compressional stress field could also be attributed to the influence of the westward-moving Pacific plate. In the Eocene (at 43 Ma), the Pacific plate changed its direction from northwestward to westward convergence. It seems that the effect of the latter on the SE Asian region
From Hail, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 291-306.
291
292
~. D. TJIA & K. K. LIEW
1 ,
l'°i I
I0~
!
~',~
NIA
INDOS
i sT~R", .¢ .. .•
g
10"II
! ,' ! a I,]/ KHMER I
I
l
It llii.
, ~
....
II
L . ~ : : .:--.,: -, . .
" " '~° . .o¢, • ,
;~i'.' ".
• ..: • " . .
"
'..
.
- o .
,,j
I "~-SONGKHLA ~,
,
",
,.
.,"
' .
, .---~
blAM" CON SON' , /" #'-
.
"•
'--lJj
,
' | /
I •
(
.
.
-
.
,.•..
. . I
q
pt l
~
•
(
s
"J2 -.':'
- /" [
...~:.. • ! ~
Peninsular j
,
,
. Malaysia
.
,
:;:
"..
i..
;.Tenggol ' ' r':..Arch "- ' •
,.-- iOi i'
•
"
PENYI
"
t
~
w e s t
•
'
UNA
"
.
.
•
,
I
' '
"
<
.
,
'-"i
["-"
I
Not u(
~4
• . ".:
..
-
x
Fig. 1. The major Cenozoic basins o f the northern Sunda Shelf. The shelf area is stippled; its northern boundaries are only approximate• Basin outlines drawn after a map by C C O P (1991)•
only became felt by mid-Miocene time. Before that time, the westward push of the Pacific plate may have been buffered by north-south spreading in the West Philippine, Caroline and South China Sea basins. The present paper deals with the geometry of structures and tectonic domains in the northern Sunda Shelf. Known faults and new evidence for major faults and their kinematics serve to define the tectonic domains more accurately and establish the structural development of the Tertiary basins of the region.
Malay
basin
Among the Cenozoic basins in the northern Sunda Shelf, the Malay basin is a major hydrocarbonbearing basin on the continental shelf of the South China Sea. The basin strikes northwest over a distance of 500 km and is 200 km wide (Fig. 1). Oligocene and younger sediments within the Malay basin are at least 12 km thick• Average long-term subsidence rates are therefore in the order of 0.30.4 m/1000 years, which is one order lower than rates of vertical crustal movements in tectonically
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
293
mobile regions. One such region is eastern Indonesia where long-term uplift and subsidence rates of 1-4 m/1000 years are common (Tjia et al. 1974; Fortuin & De Smet 1991). The northwest strike of the Malay basin is of restricted extent. Towards the northwest, the basin adjoins northtrending basins of the Malaysia-Thailand Joint Development Area (MTJDA) and of the Gulf of Thailand. Towards the southeast the Malay basin is intercepted by the West Natuna basin that strikes east to north-northeast. In the southwest corner, the Tenggol Arch, a basement high, separates the Malay basin from the almost east-trending Penyu basin (Fig. 1).
mentary rocks and igneous bodies. The centre of the basin contains more than 12 km of sediments which progressively onlap onto the basin margins. Wells are generally less than 5 km deep. The oldest dated sediments are upper Oligocene. Recently acquired deep (12 seconds) seismic sections suggest that in the axial zone of the basin, layered sequences possessing seismic signatures similar to the oldest known sediments extend several kilometres deeper below the Upper Oligocene strata. In other words, the basin was initiated well before the early Oligocene.
Stratigraphy
Fractures and f o l d s
The Malay basin sedimentary sequences are most commonly classified using the EPMI (Esso Production [Malaysia] Incorporated) seismostratigraphic scheme which is based on information from the northeastern and southeastern parts of the basin. For the northern and southwestern parts, PETRONAS Carigali's nomenclature is preferred. Carigali's stratigraphic units are based on the recognition of regional shale and/or marine transgressive pulses, supported by seismic correlation (Md. Nazri Ramli 1988). Most EPMI seismic sequence boundaries correspond to erosional unconformities at the basin margins, except the top of Group I (uppermost lower Miocene) which is a maximum flooding surface. The groups are tied to the EPMI Tertiary eustatic sea-level chart. Recently, based on micropalaeontological markers, the EPMI stratigraphic scheme has been revised and this is used in the following explanation. The initial basin fill (Group M to Group J, Lower Oligocene-lowermost Miocene) was mainly nonmarine. It was followed by deposition of Group I to Group H (Lower Miocene-lower Middle Miocene) in a stable geological environment. Severe regression occurred in mid-Group H time, and this was succeeded by a major marine transgression during upper Group H and Group F times (Middle Miocene). Group E (upper Middle Miocene) was dominated by general regression. During Group D (uppermost Middle Miocene) times, transgressive conditions prevailed. A major unconformity marked the end of Group D time, but transgression continued into Group B and Group A (Upper Miocene-Pleistocene). A comparison of the various stratigraphic schemes used in the Malay, Penyu and West Natuna basins is shown in Khalid Ngah et al. (1996). The pre-Tertiary geology of the deeper part of the basin is poorly known. The basement is thought to consist mainly of Jurassic-Cretaceous metasediments, some upper Palaeozoic metasedi-
The five main fracture directions in the Malay basin are north, NW, NNW, NNE and east. The easttrending fractures are confined to the pre-Tertiary basement (Fig. 2). The NW-striking fractures are along the basin margins with a single example, the Selambau fault (9 on Fig. 2) in the axial zone of the basin. Northward, the Western Hinge-line (WHL: 5 on Fig. 2) fault zone curves to become parallel to the dominant north-striking faults of the MTJDA and the Gulf of Thailand regions. The Kuda-Ular fault (4) is one of north-trending zones. Another north-trending fracture is the Kapal-Bergading fault zone (5) that separates the Malay basin into a southern and a northern domain. The northern domain is marked by north-trending faults and folds, while the southern domain has east-west folds and variably striking major faults. Work in progress onshore peninsular Malaysia has identified southern extensions of these northerly striking faults. Onshore the Ping-Teris fault is aligned along longitude 102053 ' and offshore can be traced into the northerly-striking fault bundle on the western margin of the Malay basin. The Ping-Teris fault zone consists of several, up to 10 m wide, bands of contorted phyllonite alternating with parallel bands of less deformed phyllite totalling 30-100 m wide in outcrops near the Ping River and Teris Village in Terengganu State. The phyllonite and the phyllite bands strike 345-350 °, with foliations dipping west or east at more than 75 °. Slickensides on fault planes are subhorizontal and associated smallscale fault-plane features such as bruised steps, accretion spalls, and pluck steps (see Tjia 1972a, for descriptions and definitions) indicate fightlateral wrenching. Proprietary data indicate that in the MTJDA, a north-striking fault that may be the offshore extension of the Ping-Teris fault is associated with fight-stepping en echelon faults in Miocene sediments and suggests, therefore, sinistral wrenching of younger age. Three north-south fault zones are also prominent in the
Structures in the Malay basin
294
H . D . TJIA •
/
K. K. LIEW
- Faults -- Faults in basement ,- Folds in OligoceneMiocene sediments
A 6 eeoooeo
.... /
°e=
D I
ODOQQ
a'eoeeeooG
,
eooeoo
PENINSULAR[~
0
5O
I
/r ,
100
I
ILLI,IIIIIIII I IIIIIII
1o,,. I
I
\" los.1 /
I.
Fig. 2. Major structures in the Malay basin. Named faults are: 1. Dungun; 2. Western Hinge-line or WHL; 3. Tenggol; 4. Ular-Kuda; 5. Kapal-Bergading; 6, Dulang; 7, Bundi; 8. Mesah; and 9. Selambau. Note the apparent dextral separations of east-west fold axes in the central and southern part of the basin. MTJDA, Malaysia-Thailand Joint Development Area. Information derived from unpublished reports held by PETRONAS.
southern domain: (6) Dulang, (7) Bundi and (8) Mesah fault zones. These three fault zones appear to have dextrally offset the hydrocarbon fields of the Malay basin. The Dulang fault exhibits 45 km, and together the Mesah and Bundi faults causes 30 km of dextral separation (see Khalid Ngah et al. 1996). Within the southern tectonic domain, most of the east-west folds are interpreted to be associated with east-west basement faults that usually form the southern side of half-graben. These deeper structural trends have been based on gravity, magnetic and occasionally also on seismic interpretation. Another fracture trend in the basin is NNE, which is confined to the northeast and east sides of the Malay basin. The WHL is associated with NNW-striking fault zones: the (1) Dungun and (3) Tenggol fault zones (Fig. 2). The Upper Oligocene (and most probably older Tertiary strata) to Miocene basin-filling sediments are deformed into large open folds with limbs that commonly dip less than 15 ° . Steeper dips are clearly associated with faults. The Sepat anticline is at 35 km the longest. Within the central part of the Malay basin, fold axes strike east-west; they are north-south in the northem part of the basin, and
NW-SE to NNW-SSE along the basin margins. In the southern part of the basin the east-west folds are transected by shallow, N-S crestal faults. Seismic profiles show that the basinal east-west folds are located over half-graben and deeper parts of the main basin (Fig. 3). Three to four east-west trending parallel belts of anticlines and associated hydrocarbon fields can be distinguished (Fig. 2). Adjacent to and within the Western Hinge-line fault zone, the anticlines are parallel or almost-parallel to the fault zone. The major part of this fault zone strikes northwest. The fold trends resemble dragfold orientations caused by left-lateral wrenching. North to north-northwest trending folds occur to the west of the Kapal-Bergading north-south trending tectonic zone. Some rounded domal and basinal structures straddle this tectonic zone. These more rounded fold structures in the border zone most probably reflect the influences of different stress fields in adjacent tectonic domains.
Pull-apart depressions Distinct pull-apart basins in the pre-upper Oligocene basement occur along the Western Hinge-line fault zone in the northwest, and in the
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
SW
Inversi on
295
NE
~-Plio- Pleistocene
!
,~ ~
Miocene
/ IIIIIIIIII
r
__
i4.
~
I
I
:
11111111
'
/ iJJiilllll
I I
/11"IIIIIIIIIIIIIIIIIIIIIII
°SilTliit [11
li
lllltlllllillllr!!l]!!tlll
)re-Tertiory
[i1| pre-U. OLigocene
11 '/ 111111111
Fig. 3. Compressed composite cross-section across the Malay basin interpreted from seismic lines. Even at 12 km depth no basement is apparent in the axial zone which also has a large inverted anticline. Another structural inversion in the SW part of the section represents pull-apart basins within the Western Hinge-line fault zone (2 on Fig. 2). The deeper parts were interpreted from a seismic line reprocessed by PRSS.
south-central part of the Malay basin. Along the WHL fault zone, the long axes of the pull-apart basins trend north to NNW (Fig. 4). This orientation, combined with the general strike of the fault zone, suggests right-lateral slip along the WHL as the cause of the pull-apart basins, This transtensional stress field pre-dates the Upper Oligocene sediments. The anticlines formed in the Upper Oligocene-Miocene sedimentary fills of the basins trend sub-parallel to the elongation of the basins. This fold pattern resembles that of drag folds due to left-lateral slip on the WHL fault zone. The sense of this later wrench motion was opposite to that which produced the pull-apart basins, and developed when the region was subjected to transpression during the middle to late Miocene (Tjia 1995). On the basis of existing information the WHL fault zone can be traced southeastward only as far as the Penyu basin. In the area of the Laris fields, within the northstriking Mesah Fault zone, are at least three probable pull-apart basins (Fig. 5). The orientation of their faulted boundaries with respect to the regional strike of the main fault zone suggests leftlateral wrenching as the cause. However, the fault pattern within the northwest-striking Selambau fault zone consists of two medium-sized pull-apart structures that suggest dextral wrenching along the major fault zone (Fig. 5). The major east-west anticlines consisting of Miocene and older sediments of the Malay basin are offset by the Mesah fault zone in a right-lateral sense by about 30 km
(Fig. 2). This suggests that dextral fault displacements occurred in Late Miocene or even younger times. On the other hand, colleagues from EPMI (pets. comm. 1994) report that they have not encountered indications of large-scale horizontal displacements along the north-south faults they are familiar with. They further suggest that major strike-slip displacements on N-S faults probably took place before the Malay basin became filled with Tertiary sediments.
D u n g u n Graben The Dungun Graben is located to the northwest of the Tenggol Arch. This graben has rhomboid shape in plan view; its two longer sides trend northwest, while the two shorter sides trend north (Fig. 6). Liew (1993) studied the structural development of this graben. It is elongated in a NNW-SSE direction and continues towards the NNW as a 40 km long and 5-6 km wide fault zone. Fracture sets in the pre-Tertiary basement of the graben also distinctly display a pattern of longer NNW faults interconnected by shorter north-south faults. The basement deepens towards the northeast where a large NNW-fault forms its boundary. Along this fault, heave is 1.8 km and throw is more than 2 km. The west side of the graben has many NNWstriking step-faults. There is a NNW-plunging basement antiforrn in the south corner of the graben. The Tertiary base of this graben consists of three elongated structural units (west, central and east;
296
H . D . TJIA •
~,
K. K. LIEW
lo
104 O0
.
o E r-
!'(I ? ......
,
"%°,
_~
I
z
t---4
........
o
..
,..
°°. °"°
.
-6
r-
r-~,
... !
k
........ , ..... ~
.
.
.
.
.
i
°'- - . . . . . . . .
o... ....
~ . .... / / . . . .
.... .~
...
Isl.
\'~ % w ,..
! ,
! ,
........... ........
....
-:z ....
'~
i/I
f
~ ,
b
•
C d
........ ' .......... ,JjI
' .
.
".
.
"~'...
~
430.
•
~'x'~ f I 2 3
X
A to F Inverted pull-apart basins
"~(X
104"00
I Fig. 4. Structures in uppermost Oligocene to Upper Miocene sediments of the northwestern part of the Malay basin. Explanation of symbols: a. major fault, hachures on downthrown side where determined; b. smaller fault; c. axis of plunging anticline; d. fold axis, type not determined; e. synclinal axis; f. sense of slip on wrench faults; 1. pre-Oligocene; 2. middle-late Miocene; 3. probably post-Miocene. Faults and folds based on an EPMI map of 1991; newly interpreted are the wrench faults and their timing, and the pull-apart basins. Note that the KapalBergading tectonic line separates east-trending from north-trending folds.
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
104" .
297
10/.,°30
..,'..
f(": i:: DULANG .....=FAULTZONE
BUNDI .-'---,FAULT ZONE
i
"-;."'"'"'" """ I
.k..,
MESAH "": FAULTZONE,:"
'r l
6'30
(...
b04"
' "-----FAULT FOLDAXIS "~'..~i':: HCFIELDof PROSPECT ~ P U L L APART
\ i's
\ f Fig. 5. Structures in the Central Malay basin. The interpreted pull-apart basins within the Mesah and Selambau fault zones possess patterns suggesting right-lateral wrenching as their cause. Structures were interpreted from a 1991 EPMI map held by PETRONAS.
Fig. 6). The east unit contains the large NW fault set that forms the graben boundary. There the basement dips steeply west and is transected by a few NNW faults. The central unit is characterized by left-stepping en echelon WNW-striking faults and a basement that is gently inclined towards the NNW. Bifurcating NNW- and N-trending faults dominate the west unit of the graben. There, the basement deepens towards the ENE. The rhomboid shape of the Dungun Graben and its location along the NNW Dungun fault zone suggest that it was formed by dextral wrenching along the fault zone (Fig. 6). The Dungun-B and
Dungun-C graben may be dextrally-displaced portions of a single basin with an offset of c. 14 km. However, the NNW-trending fold axes of the Tertiary basement and overlying Tertiary sediments (not shown on Fig. 6) are interpreted as drag folds and suggest a later sinistral wrench motion along the Dungun fault zone.
Penyu basin The Penyu basin lies offshore of Kuantan, Pahang. It consists of two parts separated by the major NWstriking Rumbia fault (new name; see Khalid Ngah
298
H.D. TJIA & K. K. LIEW structures in the western basin from the W N W ESE structures to the east. This fault is more than 100 km long and possesses an average heave of 2 km. The structure also separates areas with different tilting directions. In some areas en echelon faults, sometimes sigmoidal in shape, suggest wrench fault movement. The Rumbia fault runs parallel to a zone of probable left-lateral faulting (see Khalid N g a h e t al. 1996). In other words, the Rumbia fault is perhaps also a left-lateral wrench. The structural history of the Penyu basin is as follows. Prior to the Late Oligocene, north-south extension/transtension formed east-west striking half-graben in the western and eastern Penyu basin. Alternatively, these half-graben may have formed as consequence of left-lateral wrenching along major NW-trending faults parallel to the Rumbia fault but located outside the Penyu basin proper. We suggest that during Late Oligocene to Middle Miocene continued left-lateral motion on the Rumbia fault zone rotated the half-graben of the eastern Penyu basin clockwise some 25 ° into WNW-striking structures.
West Natuna basin
Fig. 6. The Dungun fault zone runs NNW and contains the Dungun Graben and Dungun-B graben, Dungun-A and Dungun-C half-graben. Structures based on unpublished 1992 map of Western Mining Co., held by PETRONAS.
et al. 1996). It is bordered by the Tenggol Arch to
the northeast, the Pahang Platform to the northwest, and the Johor Platform to the south (Fig. 1). Khalid Ngah (1975) analysed the stratigraphy and structure of the basin. Proprietary data available since 1992 reveal further information. The western part is elongated east-west and possesses half-graben of similar strikes. The basement of these half-graben dips northward, except in the southeast corner, where the basement is inclined towards the southeast. The main east-west fault set of the western basin abuts against the NW-striking Rumbia fault. Half-graben in the eastern part of the Penyu basin trend WNW to NW. The basement in most of these half-graben inclines towards the southwest. Most fold axes within the Penyu basin are parallel to major faults. This suggests that these folds may be rollovers. The NW Rumbia fault separates the E - W
The West Natuna basin broadly curves from an east-northeast trend in the southwest to a northeast direction farther towards the east and north. The structural trend of the Malay basin continues into the west-central part of the West Natuna area (Figs 1 and 7).The following geological information on the West Natuna basin is from Wongsosantiko & Wirojudo (1984), Daines (1985), and Sujanto et al. (1986). The basin began to develop during the Early Oligocene when rifting and/or pull-apart processes produced NE-SW half-graben that began to be filled with non-marine sediments. Upper Oligocene-Lower Miocene sediments are of fluvial-lacustrine character and are considered hydrocarbon source rocks. In late Early Miocene, the area became subjected to compression and this resulted in reversal of slip sense on existing normal faults and inverted structures in the half-graben. Right-lateral displacements took place along east and ESE-striking fault zones, while left-lateral offsets developed along north-striking faults (see Daines 1985). Folding of sediments filling halfgraben produced anticlines with ENE trends. This fault pattern implies a maximum horizontal stress directed NW-SE. During the remainder of the Miocene, paralic-marine sedimentation took place in quiet tectonic surroundings. Late Miocene to Early Pliocene tectonic deformation affected the entire basin and developed an unconformity, especially on anticlinal crests. Overlying Pliocene to Recent marine sediments are relatively un-
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
--2" ~(~ 108"
299
of the Sunda Shelf (Figs 1 and 7). Wongsosantiko & Wirojudo (1984) pointed out that this basin has a different structural history compared with the West Natuna basin. Oligocene half-graben occur only in the northern part. The graben fills are thinner and suggest a smaller amount of subsidence, and consequently also a weaker rifting effect. Outside the graben area, the larger part of the East Natuna basin is covered by relatively thin (less than 1 km) Oligocene sediments. Miocene sediments are locally thick and are bounded by growth faults which began to develop in Early Miocene time. Most of the Miocene faults are extensional and can be grouped into NNW and NE-striking fractures. Only minor folds exist and these are associated with NNW-SSE faults. Some of these NNW faults suggest dextral-slip reflecting some lateral tectonic stress. In the East Natuna basin the dominant stress regime during the Miocene was (ex)tensional. During the Plio-Pleistocene, the basin was tectonically inactive except in its southern part where Pliocene faults are common, and in its southeast part which is characterized by clay diapirism. Marked thickening of Plio-Pleistocene sediments towards east is interpreted to indicate rapid subsidence coupled with sedimentary loading. These processes probably caused faulting and clay diapirism. NE-striking faults occur in the northernmost part and in the centre of the East Natuna basin. Elsewhere, faults strike NNW to NW. The boundary with the West Natuna basin is marked by two regional north-striking, weakly curved normal faults; the southern fault downthrows west and is concave in that direction, the northern fault downthrows east.
Mekong basin Fig. 7. Isopachs (in km) of Cenozoic sediments in some of the northern Sunda Shelf basins. North-south lineaments of isopach contours probably represent major faults. The Vietnam Shear is also indicated by the linear shelf edge. Note the en echelon arrangement of sub-basins within the Nam Con Son basin. Structures were interpreted from a CCOP map (1991).
disturbed. Pliocene faulting is essentially absent. Locally, total sediment thicknesses reach 4500 m. It is believed that except in the deeper parts, maturation and migration of hydrocarbons occurred only very recently.
East Natuna basin The north-south elongated East Natuna basin, also known as the Sokang/Soikang basin, lies to the east of the Natuna Arch and occupies the eastern margin
The Mekong and Nam Con Son basins are located in the northeast comer of the Sunda Shelf that adjoins the Indochina craton (Figs 1 and 7). The basins were formed in fractured and subsided preTertiary basement and are filled by more than 4 km and more than 6 km of continental sediments respectively, and locally also by peri-marine sediments (CCOP 1991). New proprietary data indicate that a large part of the basement includes upper Mesozoic granites that serve as important hydrocarbon reservoirs. The Mekong or Cuu Long basin strikes ENE and is located offshore of the Mekong Delta. It consists of a main basin where the maximum sediment thickness exceeds 6 km, and several smaller basins. All are within a 100 km wide zone trending NE over a distance of 450 km. In the basement rocks of the main basin, longitudinal faults dominate and are associated with most of the half-graben. Other halfgraben are associated with shorter east-west faults,
300
H . D . TJIA & K. K. LIEW
especially in the SW sector and in the NE part of the basin. Along the NE-orientated faults, downthrow directions are mainly symmetrical about the basin axis. The shortest faults are north-south. In the White Tiger field in the centre of the basin, these north-south faults seem to be associated with the NE-trending faults, suggesting left-lateral motion along the latter. Many faults with throws between 200 m and 1000 m occur in the northeast part of the basin (CCOP 1991). Other structural elements include basement highs, basins and monoclines. Unpublished cross-sections of the White Tiger and Dragon fields show that deformation intensity decreases from the Lower Oligocene to the Middle Miocene beds. Post-Miocene beds are essentially undisturbed (Areshev et al. 1992). The basin fill consists mainly of siliciclastic sediments intercalated with some basalt and andesite horizons in the Oligocene and Lower Miocene. The Palaeogene is generally of continental character, while the younger beds represent deltaic to shallow marine environments. Regional unconformities were mapped at the base of Palaeocene to Oligocene, in the Miocene, and in the Pliocene (CCOP 1991). The basement may be mainly block-faulted granitic rocks with K-Ar ages of 97-178 Ma (T. Minh et al. (1991) cited by Mauri et al. 1993). Geothermal gradients are moderate to high (3.9°C/100 m) and Recent volcanism in the northeastern part of the basin (Fontaine & Workman 1978) may be expected to contribute to the high heat flow.
Nam Con Son basin On its south side, the Mekong basin is separated by the NE-striking Con Son Swell from a basin complex that is collectively known as the Saigon basin, or Ho Chi Minh basin, or Nam Con Son basin (Figs 1 and 7). The Con Son Swell may be a regional horst bounded by NE-striking normal faults and is composed of Cretaceous granites (Mauri et al. 1993). Fault throws average 800 m (CCOP 1991). The Nam Con Son basin may be divided into an eastern part and a western part. The division is effected by a north-trending basement high that extends from the Natuna Arch to the Con Son Swell. The eastern part of the basin strikes NNE and contains two deeper en echelon troughs trending NE. Its eastern border is linear and strikes north (Fig. 7). The CCOP (1991) map shows that this lineament extends along the entire edge of the Sunda Shelf basement. We provisionally name it the Vietnam Shear and interpret it as a major fault zone with probable strike-slip motion. The pattern of NE-striking troughs in the eastern Nam Con Son basin in association with the Vietnam Shear suggests that the basins may be the result of right-
lateral wrenching along the north-trending Vietnam Shear. The western part of the Nam Con Son basin consists of several smaller troughs that in the south strike north and swing to NNE trends in the north. The change in strike can probably be attributed to the fact that the western Nam Con Son basin straddles a north-trending fault zone that also forms the western margin of the basement high that subdivides the greater Nam Con Son basin. Northtrending faults border the troughs of the western basin. The CCOP (1991) map shows 7 km of sediments in the basin. Matthews & Todd (1993) classify the basin as a Tertiary rift complex and recognize four major tectonostratigraphic units. The oldest unit was interpreted from seismic features and comprises Palaeogene rift-fill deposited in a gently divergent setting; seismic signatures of lower Miocene postrift sediments suggest regional thermal subsidence. A second pulse of extension is represented by middle-Miocene rift-fill that is capped by an onlapping marine sequence. Gentle compression associated with uplift of depocentres formed a major late-middle Miocene truncational unconformity. A third pulse of mild extension is indicated by minor faulting and major regional subsidence resulting in very rapid flooding during the late Miocene and Plio-Pleistocene.
Changes in stress field Recently, Tjia (1994) proposed that the Malay, Penyu and West Natuna basins originated as rift arms upon an upper Mesozoic crustal dome in the northern Sunda platform. The updoming was probably caused by a rising mantle plume. Major arguments for that concept include the following. (1) The three basins form a three-armed star and meet at a triple junction. (2) The triple junction is characterized by one of the higher heat flow areas in a region that already possesses above-average geothermal gradients. (3) The relatively high positions of pre-Tertiary basement represented by the Con Son Swell, The Natuna-Paus-Ranai High, the Johor Platform, the Tenggol Arch and peninsular Malaysia may be remnants of the crustal dome. (4) Late Cretaceous granitic intrusions are widespread in this region and define the initial age of the domal structure. (5) The tensional stresses that appear to have acted contemporaneously to create the three basins of varying strikes are most simply explained as result of rifting atop a lithospheric dome. In other words, the Malay, Penyu and Natuna basins are interpreted as aulacogens that developed upon a late Cretaceous Malay Dome (new name) possessing a diameter of 1000 km. At one time or another, previous researchers on the Cenozoic basins of this region have attributed
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF the basin origins to subsidence by crustal thinning, or to rifting, or to wrench-fault related processes without further explanation. From this study it has been concluded that these basins experienced multiple deformational processes of contrasting nature. Superimposition of deformational signatures has contributed to the confusion about their origin. For instance, the Penyu, West Natuna and East Natuna basins have been suggested to be extensional basins. However, en echelon arrangement of fractures and basins in their basement rocks also indicates that transtension caused by wrenching must have been important modifiers in basin development. Another example: east-west halfgraben in the western Penyu basin seem related to left-lateral wrenching along the NW-striking Rumbia Fault and/or other major faults of similar strikes. Similarly, the NE-trending half-graben in the northern part of the East Natuna basin may be associated with right-lateral wrenching along the major, north-striking Vietnam Shear that marks the remarkably linear western margin of the South China Sea basin (Figs 1 and 7).
Kinematic: Cretaceous-Eocene At the transition from the Mesozoic into the Cainozoic, the regional stress field for onshore Peninsular Malaysia consisted of a maximum horizontal stress orientated between E - W and ENE-WSW. This regional compression caused strike-slip motions in sinistral sense along WNW, NW and NNW major faults, and in dextral sense along faults striking between N-S and ENEWNW (Tjia 1972b, 1989). The Malay Dome, a late Cretaceous crustal dome of c. 1000 km diameter, developed near the present triple junction of the Malay, Penyu and West Natuna basins. These basins are interpreted to have developed as rift arms on that dome. Figure 8 shows the offshore faults in the northern Sunda Shelf that may have been active during this period and the strain ellipses illustrate how the regional compression direction was NE-SW in the northeast and changed gradually to an ENE-WSW orientation in the southwest.
Kinematics: Middle EoceneMiddle Oligocene The stress fields in the domains bordered by the major wrench faults are schematically shown by strain ellipses on Fig. 9. Opposed slips on subparallel wrench faults suggest that the faultbounded crustal blocks moved in directions indicated by the large arrows. The differential block movements are interpreted to have resulted from continued indentation of the Indian plate into Asia
301
combined with interference by local obstacles blocking or retarding certain block movements. The obstacles could consist of massifs or of faults locking where fault strikes changed. The Axial Malay fault is interpreted to exist in the deep basement in the Central Malay basin and to represent an extension of the Three Pagodas fault. Sinistral slip along the Axial Malay fault created roughly eastwest orientated half-graben and graben in the pre-mid Eocene basement. The Vietnam Shear most probably moved dextrally as a result of N-S spreading of the South China Sea basin. Northsouth faults bordering the western Nam Con Son basins may have moved in the same sense. Subsidiary structures generated by these movements could be the en echelon NE-trending subbasins in the Nam Con Son basins (see also Fig. 7).
Kinematics: end Miocene By the beginning of the Middle Miocene (17 Ma; Taylor & Hayes 1983) the South China Sea basin stopped spreading. As a result, the authors suggest that from then on the westward convergence of the Pacific plate was able to influence the Sunda Shelf region. By the end of the Miocene, the continental crust of the Indian-Australian plate had moved closer to continental SE Asia and thus also began to form an important barrier to southeastward extrusion of the crustal blocks. These external barriers combined with internal obstacles (massifs, locking faults) most probably caused slip reversals along several major wrench faults. The Ping-Teris fault and the Western Hinge-line fault became leftlateral, and the Axial Malay fault became rightlateral. The Mae Ping Fault may have become left-lateral again through the convergence of the Pacific plate. The stress fields consistent with these wrench movements and the directions of block extrusions are shown on Fig. 10. Slip reversal along the Axial Malay fault caused folding of the sediment fills of the east-west half-graben into eastwest anticlines and probably also dextral motions along the Dulang and Mesah faults.
Conclusions During the Late Cretaceous a hot mantle plume is interpreted to have risen, centred at the triple junction of the Malay-Penyu-West Natuna basins. This formed the Malay Dome, with a diameter of c. 1000 km, of which relics are represented by relatively high regions of basement rocks such as most of peninsular Malaysia, Natuna Islands and Con Son Swell. The Malay-Penyu-West Natuna basins and perhaps also the western Nam Con Son basin (see Fig. 1) began to develop as aulacogens.
302
H.D. TJIA • K. K. LIEW .':'-k
lU •
.
.
.
•
!
.
-o
J
,,,
Strike-slip
Strain
fault
10"--
ellipse VIETNAM
5
~o
8' "
"
;:'2"":'"
".
0
Z 0
4_ N : ! "
"':
0
100
I
I
200
Ikm 6"
WEST ' NA T U NA
IlIiI%~IIN..,,~
~,9 ~ 11 a°t'I 11'
...
1..
Fig. 8. Kinematics in the northern Sunda Shelf during the Cretaceous to Eocene. Strike-slip faults onshore peninsular Malaysia were determined through fieldwork; the offshore faults are discussed in the text. Strain ellipses show gradual change of regional compression (parallel to the short axes of the ellipses): in the northeast striking NE-SW and in the southwest trending ENE-WSW. The orientations of strain ellipses have been determined by aligning the short (parallel to compressional stress) ellipse axis at about 45 ° to the adjacent wrench fault and taking the slip sense into consideration. Inset shows the Mekong basin and part of the Nam Con Son basin. Faults are: KL, Kuala Lumpur; B, Bukit Tinggi; BA, Baubak; BB, Bentong-Bengkalis Suture; L, Lebir; P, Ping-Teris; M, Mersing zones; MO, Ma' Okil; D, Dungun; Du, Dulang; Me, Mesah.
The origin of the other basins, Mekong, eastern N a m Con Son and East Natuna basins, is not yet clear. The presence of thick sediments in these basins indicates that tensional stresses caused basins to form in the early Oligocene or earlier in
the area of the M e k o n g and East Natuna basins, and by the early Miocene also in the N a m Con Son area. The middle Eocene collision of the Indian plate with the Eurasian plate began p u s h i n g out elongated crustal slabs of continental SE Asia
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
I
303
I Strike slip faul t
G
Strain ellipse
r, :.'.:- (:i:i;'... • .,." • , ~'." , . '
!.i: ::::
....4"NATUNA
NA ~
I.
104" ,11
106' I
0
100
I
I
1"::i;.':", \'~" ""
20O Ikrn
I
Fig. 9. Kinematics between the Mid-Eocene to Mid-Oligocene in the northern Sunda Shelf. Faults: WHL, Western Hinge-line; DR, Dungun-Rumbia. Graben: Dg, Dungun; Tb, Tok Bidan. The shorter strain ellipse axis is parallel to horizontal compressional stress, consistent with sense of wrench motion on the adjacent fault. Dashed bold lines are N-S faults interpreted from lineaments of the Nam Con Son basin. Half-tipped arrows indicate sense of wrenching. Large arrows indicate direction of motion of fault-bounded crustal blocks. Basins are stippled.
towards southeast consistent with the Tapponnier et al. (1982) model of extrusion tectonics. These
crustal blocks are bounded by NW-trending wrench faults, of which the Mae Ping and the Three Pagodas faults and their splays play important roles in the tectonic development of the northern Sunda Shelf region. Kinematic analysis of structures in pre-upper Oligocene rocks of the basins shows that until late-early Miocene, regional tensional stress prevailed, and that the major faults and their splays acted as wrench faults. Alternate crustal blocks bounded by these faults moved in opposite directions (Fig. 9). The Ping-Teris, Western Hinge-
line and Mae Ping faults had a dextral slip sense whereas the Dungun-Rumbia and the Axial Malay faults had a sinistral slip sense. Pull-apart basins developed along the WHL, and east-west halfgraben formed in the Malay basin. The eastern part of the northern Sunda Shelf was influenced by dextral slip along the Vietnam Shear as result a of north-south spreading of the South China Sea basin. This major shear and other N-striking faults on the shelf probably created the NE-trending sub-basins in the Nam Con Son basin. By the middle Miocene, structural inversion began and the slip sense along wrench faults
304
H.D. TJIA • K. K. LIEW
J
Strike
slip
fault
S t r a i n ellipse g e n c e at ,•Conver
.'-~';.:'1 i'.":C .--' ".~.PLATE..... -:. -. -.: ~:.'""
....
"i
.':..,.........."
. ;
,;: --: ...::. .~ : : • .... :~|::'..'~
?:,,.~...:..... ,-~.• -.... : .;,/
;T/
•~.:'. :..." • ":~...;'.': ,
, . : . . ' , . ' . . . -.;
WEST , NATUNA I
NATUNA I. ' 25' Rotation of half groben in eastern Penyu Basin 106"
0
IO0
I
,,
2oo
~ km
Fig. 10. Kinematics at the end of the Miocene in the northern Sunda Shelf. Abbreviations as on Fig. 8 and 9. Other faults: D, Dulang; Me, Mesah. Continued sinistral slip along the Dungun-Rumbai fault caused 25 ° clockwise rotation of structures in the eastern Penyu basin. Half-tipped arrows indicate sense of wrenching. The stress fields consistent with these wrench movements are represented by strain ellipses. Shorter axes of strain ellipses are parallel to compressive stress; large arrows indicate directions of motions of fault-bounded crustal blocks.
reversed. We propose that the change into a general compressional stress regime for the region was the result of convergence by the IndianAustralian plate from the south, by the Pacific plate from the east, and by continued indentation of the Eurasian plate onto the Indian plate. The latter had continuously pushed out SE Asian crustal blocks towards the southeast, but by middle Miocene time this tendency became impaired. First, the IndianAustralian plate had reached a proximal position with respect to SE Asia. Second, before middle Miocene, the influence of westward convergence of the Pacific plate was absorbed or greatly reduced
by north-south spreading of the West Philippine basin (Eocene) and by that of the South China Sea basin (mid Oligocene-early Miocene). Reversal to dextral slip on the Axial Malay fault compressed the basin-filling sediments into large east-west anticlines, whose position and orientation were fixed by the pattern of half-graben. Sinistral slip along the W H L compressed the sedimentary fills of the pull-apart basins into anticlines trending subparallel to the long axes of the pull-apart basins. The crustal block bounded by the Dungun-Rumbai and Axial Malay faults together with the crustal block to the northeast of the Mae Ping fault moved
STRESS FIELD CHANGES, NORTHERN SUNDA SHELF NW-ward, while the other crustal blocks were extruded SE-ward (Fig. 10). After the Miocene, the development of shallow crestal faults orientated normal to the anticlines probably attests to further warping of the structures. This deformation may be a tectonic process or may represent differential isostatic adjustment in the Malay basin. A major outstanding problem is the timing of wrenching along the northerly trending Dulang, Bundi and Mesah faults. In the Malay basin, dextral offsets of folds associated with hydrocarbon fields suggest post-folding movement. The youngest folded sediments are upper Miocene. But colleagues from EPMI have shown that a northsouth fault which has not appreciably displaced
305
sediments is as old as middle Miocene (Md. Nor Mansor & K.W. Rudolph at the Petroleum Geology Seminar, Geological Society of Malaysia in Kuala Lumpur, 7 & 8 December 1993). Most of the unpublished information referred to in this article was obtained from PETRONAS Exploration Management Department. We thank the managements of PETRONAS EMD and PETRONAS Research and Scientific Services for permission to publish this article. Thorough reviews of an earlier version by R. W. Murphy and B. Clennell are very much appreciated and led to this rewritten paper. A travel grant from the organizers of the SE Asia Tectonic Evolution Conference enabled Tjia to participate in the meeting. Liew's attendance was supported by PRSS.
References ANIZAN ISAHAK. 1993. Geomorphology and soils of Southeast Johore. PhD Thesis, Universiti Kebangsaan Malaysia, Bangi. ARESHEV, E. G., TRAN LE DONG, NGO THUONG SAN & SHNtP, O. A. 1992. Reservoirs in fractured basement on the continental shelf of southern Vietnam. Journal of Petroleum Geology, 15, 451464. CCOE 1991. Total sedimentary isopach maps offshore East Asia. (6 maps with explanatory text) CCOP Technical Bulletin, 23. DAINES, S. R. 1985. Structural history of the West Natuna basin and the tectonic evolution of the Sunda region. In: Proceedings of the Indonesian Petroleum Association 14th Annual Convention, Jakarta. 39-61. FONTAINE, H. & WORKMAN,D. R. 1978. Review of the geology and mineral resources of Kampuchea, Laos and Vietnam. In: Geology and Mineral Resources of Southeast Asia, 3rd GEOSEA, Bangkok, Thailand. 538-603. FORTUrN, A. R. & DE SMET, M. E. M.1991. Rates and magnitudes of Late Cenozoic vertical movements in the Indonesian Banda Arcs and the distinction of eustatic effects. In: MACDONALD, D. (ed.) Sedimentation, Tectonics and Eustasy: Sea-level Changes at Active Margins. International Association of Sedimentologists, Special Publication, 12, 77-89. HARDER, S. H., MCCABE, R. J. & FLOWER,M. E J. 1992. A single mechanism for Cenozoic extension in and around Indonesia. In: Symposium on Tectonic Framework and Energy Resources of the Western Margin of the Pacific Basin, Kuala Lumpur Nov-Dec 1992. Programme & Abstracts. 34. KHALID NGAH. 1975. Stratigraphic and Structural Analyses of the Penyu Basin, Malaysia. MSc Thesis, Oklahoma State University. --, MADON, M. & TJ~A, H. D. 1996. Role of preTertiary fractures in formation and development of the Malay and Penyu basins. This volume. LIEW KIT KONG. 1993. Structural development at the west-central margin of the Malay basin (basement of Blocks PM 2 and PM 7). Geological Society of
Malaysia, Annual Geological Conference 1993, Warta Geologi, 19, 131-132. MATTHEWS, S. J. & TODD, S. E 1993. A tectonostratigraphic model for the southern Nam Con Son basin, offshore Vietnam. In: Geological Society of Malaysia, Petroleum Geology Seminar 1993, Programme and Abstracts. 22-23. MAURI, S., HARDER, S. & MCCABE, R. 1993. Gravity modelling of Cenozoic extensional basins, offshore Vietnam. In: FLOWER, M. E J. et al. (eds) 1992 Report to the Indochina Research Consortium (January 1993). MD. NAZR1 RAMLI. 1988. Stratigraphy and paleofacies development of Carigali's operating areas in the Malay basin, South China Sea. Bulletin Geological Society of Malaysia, 22, 153-187. MOHD. FIRDAUS ABDUL HALIM. 1993. Heat flow in the Malaysia sedimentary basins. Petronas Research & Scientific Services Sdn. Bhd., Proceedings Exploration Research Seminar. 1, Progress Reports and Findings. 11-25. SUJANTO, E X., HARTOYO, P. • SUGIHARTO, H. 1986. Hydrocarbon geology of producing basins in Indonesia and future exploration for stratigraphic traps. In: CCOP Technical Publication 17/ASCOPE Technical Publication, 7, 5-24. TAPPONNIER, P., PELTZER,G., LEDAIN, A., ARMIJO, R. & COBBOLD,P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611-616. TAYLOR, B. & HAVES, D. E. 1983. Origin and history of the South China basin. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands, Part 2. American Geophysical Union Monograph, 27, 23-56. TJIA, H. D. 1972a. Fault movement, reoriented stress field and subsidiary structures. Pacific Geology, 5, 49-70. 1972b. Strike-slip faults in West Malaysia. In: International Geological Congress, 24th Session, Montreal, Section 3. 255-262. 1989. Major faults of Peninsular Malaysia on remotely-sensed images. Sains Malaysiana, 18, 101-114.
306 -
-
n . D . TJIA & K. K. LIEW 1994. Origin and tectonic development of MalayPenyu-West Natuna basins. In: PETRONAS Research & Scientific Services Research for Business Excellence Seminar, Kuala Lumpur, June 20-21, 1994. 1995. Inversion tectonics in the Malay basin: evidence and timing of events. Geological Society of Malaysia Bulletin, in press.
, FUJII, S., KIGOSHI, K. & SUGIMURA,A. 1974. Late Quaternary uplift in eastern Indonesia. Tectonophysics, 23, 427-433. WONGSOSANTIKO, A. & WIROJUDO, G. K. 1984. Tertiary tectonic evolution and related hydrocarbon potential in the Natuna area. In: Proceedings Indonesian Petroleum Association 13th Annual Convention, Jakarta. 161-183.
Far-field and gravity tectonics in Miocene basins of Sabah, Malaysia BEN CLENNELL
Dept of Earth Sciences, University of Leeds, Leeds LS2 9JT, UK
Abstract: At the end of the early Miocene, at c. 17 Ma, oceanic spreading ceased in the South China Sea Basin, as a series of collisions were initiated between continental blocks derived from the Asian mainland and the northwestern margins of Borneo and Palawan. In the NW Sabah Basin this tectonic event caused a major period of uplift and erosion which produced the so-called 'Deep Regional Unconformity'. While these mainly compressional events were occurring at the southern margin of the South China Sea, the Sulu Sea was undergoing extension, with rifting in the northwest and oceanic spreading in the southeast. Recent seismic surveys in the Sabah offshore area show that the Deep Regional Unconformity can be traced through from the South China Sea into the Sulu Sea as a boundary between two contrasting depositional systems that are both represented in sedimentary sections onshore eastern Sabah. Eastern Sabah changed from an environment of deep marine clastic deposition in the Oligocene and early Miocene, to shallow marine and terrestrial sedimentation in the mid to late Miocene, with a major period of sedimentary mrlange formation occurring at the time of the Deep Regional Unconformity. Inversion of the Miocene sequence in eastern Sabah appears to be limited to the edges of basement blocks, which were moved by far-field tectonic stresses. Post Middle-Miocene basin evolution in the onshore (Central Sabah Basin) and adjacent parts of the Sulu Sea offshore (Sandakan Basin) has been strongly influenced by mud diapirism and gravitational sagging of progradational sand-rich sediments into underlying muds and mrlange units. These sequences appear to have been decoupled from later tectonic events by the great thickness of underlying muds and mrlanges.
SE Asia has been an area of diverse and rapidly changing tectonic and sedimentary environments throughout the Tertiary. It is generally agreed that large-scale plate motions provided the ultimate driving force of the main periods of Tertiary basin formation in SE Asia (e.g. Hamilton 1979; Hutchison 1989). For this reason, major tectonic events, such as phases of rifting or inversion, may be expressed synchronously in several basins across the region. Nevertheless, the timing and kinematics of basement movements vary from basin to basin. Within individual basins, the effects of far-field tectonic forces on the sedimentary architecture are augmented by gravity tectonics and autocyclic sedimentary processes. Only rarely can eustatic signals be resolved amid the noise of these competing processes, so that classical sequence stratigraphy can only play a limited role in the understanding of SE Asian basins. Sabah is a state of Malaysia that lies in North Borneo, at the geological junction between the cratonic core of Sundaland, the Sulu, Celebes and South China Sea marginal basins, and the volcanic arc systems of the Philippines. This paper focuses only on one time interval in the geological evolution of this complex region; namely the Early
to Middle Miocene transition. A simplified interpretation of the Early Miocene tectonics of the Sabah region is shown in Fig. 1. Areas discussed here are the Northwest Borneo Basin, the Southwest Sulu Sea Basin, and its onland extension into eastern Sabah, known as the Central Sabah Basin (Hutchison 1992). The offshore areas are k n o w n from recent deep-seismic surveys (Hinz et al. 1989, 1991), and commercial seismic surveys and drilling for petroleum exploration and production. The data from onland Sabah come principally from road cuttings associated with a major road improvement scheme that traversed eastem Sabah from 1988 to 1991 (Clennell 1992).
Western and northern Sabah and adjacent parts of the South China Sea Proceeding from N W to SE, the west of Sabah can be divided into several structural provinces, originally delineated by Hinz et al. (1989) and Hazebroek & Tan (1993); these are shown in a modified form in Fig. 2. A simplified stratigraphy of Sabah is shown in Fig. 3. The outermost part of the northwestern Sabah margin is floored by several blocks of attenuated
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 307-320.
307
308
B. CLE~'mELL
Fig. 1. Suggested Early Miocene (18 Ma) tectonics of the Sabah region. The following should be noted: active submarine and subaerial vulcanicity along the Cagayan Ridge extending into Sabah in the Sandakan area; cessation of spreading in the South China Sea Basin is linked to onset of compression in Northwest Borneo Basin, Western Sabah, Northwest Sulu Basin and Palawan; incipient spreading in the Southeast Sulu Sea, and associated rift-basin initiation in the Central Sabah Basin. The geological evidence suggests compression and extension were underway synchronously in the same region, while between these domains strike slip zones are postulated between the major lithospheric blocks (positions of inter-block transform faults are guides only). Abbreviations: NWSS, Northwest Sulu Sea; SESS, Southeast Sulu Sea; RB, Reed Bank; NP, North Palawan; SP, South Palawan; CAG, Cagayan Ridge; MI, Miri Platform; LS, Luconia Shoals; DG, Dangerous Grounds; ZG, Zamboanga; NWBB, Northwest Borneo Basin; PA, Panay; LU, Luzon; BSF, Balabac Strait Fault (postulated); WSSB, West Sulu Sea Basins; CSB, Central Sabah Basin. Data from Daly et al. (1991); Rangin et al. (1990a, b, c, 1991); Mascle & Biscarrat (1979).
continental lithosphere, which separated from the south Asian mainland during Palaeocene rifting and moved southwards as oceanic spreading occurred in the South China Sea from c. 4 5 - 1 7 M a (Holloway 1982; Ru & Piggott 1986; Briais et al. 1989; Ben-Avraham 1989). The continental fragment adjacent to Sabah is known variously as the Dangerous Grounds Block or the Northwest Sabah Platform (Zone I in Fig. 2). The NW Borneo-Palawan trough (Zone II) is a SE-facing half-graben at the edge of the NW Sabah platform that is overridden to the SE by shortened sediments of Neogene age (Zone III). Zone III is variously interpreted as the compressive toe of a subsided delta (e.g. Tan & Lamy 1990; Hazebroek & Tan 1993, in analogy with the Niger Delta described by Damuth 1994), or a tectonic thrust system in its own right (Hinz et al.. 1989). Details
of the structure suggest that the former interpretation applies, at least to the south, where Zone III passes gradually into the East Baram Delta proper (Zone VI). To the north in Zone IV, the lower compressive belt is overridden by a major thrust sheet system of mainly Palaeogene sediments that can be traced onland as the Crocker Formation and equivalents (Fig. 3). The D e e p Regional Unconformity
Figure 4 (A-A' on Fig. 2) is an interpretation of the seismic line BGR 86-06 (Hinz et al. 1989) showing the relationships between these two compressive belts. The top of the upper thrust slice is marked by a major unconformity, known as Horizon 'C' by Hinz et al. (1989, 1991), that is believed to mark the time of collision of the Dangerous Grounds
309
MIOCENE TECTONICS OF SABAH, MALAYSIA
I117°
}116°
115°
IVa
Balabac Strait
~" / J
t/.= B~ggi
VIII
7°
KUDAT
I~
•
/~
SOUTH; ~ CHINA / ,,~ S E A'~,,~
V,,J">-'-'-"" p-Y'" /
fz_
.t~-"
KOTABELUD'
r"
•
,,I'Xb'
_(
I
VII
.-~
..
I,
#.f,,," /
~, /
Baramg. Delta
Labu~. f
Kiias
,
,~....... i
KEY M~langes Ophiolite
I
l
°
......... Major
.........
BRUNEk BAY °,, ,
.....
,,
, ,
,,)~/&:/ .... .. .... ..,,~..- 1~ t S :"::..--/
Overlhrust
Synformal axis ""'x"
"~.% Reverse t~, fault
""'x"
..
..,I~ " Major
lineament
Anti formal ~....~... axis
Granitoids
,,:
-.:~--
i 115
-
I~l
4 ,
/
50 km
"-.~I[I~, RANAU ~il~'-j,,~.
:KINABALU • \
.,
yr t
5°
Mt. Kinabalu
•"r~:" ..'KOTA'" ~i "~
:
"-{
•.~:~.~ Strikeslip "
.. • Normal fault
Structural Province boundary
,,
,i
STRUCTURAL PROVINCES I
II
N.W. Sabah Platform (Dangerous Grounds)
IV
NW Borneo- Palawan Trough (Sabah Trough)
III Lower Thrust Sheets & Compressive Delta Toe
V
Major Thrust Sheet System (Imbricated Palaeogene & ? Ophiolite)
VII
Inboard Belt: reverse and wrench faulting X
VI
RedeformedMajor Thrust Sheet System pierced by mud dia.oirs
VIII
Banggi-Belabac Zone: thrust/wrench oomplex
East Baram Delta: growth faulting, diapirs wrench tectonics
IX
Crocker Belt (Palaeogene accreted complex)
Central Sabah Basin (onland extension of Sotu Sea Basin)
Fig. 2. Tectonic map of western Sabah and adjacent offshore area. Onland data from Lim & Yin (1985), B6nard et al. (1990) and Tongkul (1991). Offshore data from Hinz et al. (1989) and Rice-Oxley (1991).
platform with northern Borneo. This unconformity is extensive, and is more usually referred to as the 'Deep Regional Unconformity' (Bol & Van Hoorn 1980; Levell 1987). While the nature and origin of this unconformity can be appreciated in sections such as Fig. 4, it becomes more deformed, and less easy to interpret, further landwards. The overprint to the west in Zone IVa is predominantly compressional, while later extensional structures are dominant in Zone IVb. To the far north, in Zone V, the overlying sediments are decoupled from the underlying thrust
belt by listric extensional detachments (Levell & Kasumajaya 1985), and the entire sequence of structural units is pierced by NE-trending walls of diapiric shale. In the south, the sediments of zone VII and slivers of the underlying basement have been buckled upwards to form Labuan island and the Klias Peninsula, where mud-cored folds and mud volcanoes are found (Leichti et al. 1960; Wilford 1967; and personal observations). Transpression has affected this zone, inboard parts of Zone IVb and V, and is the dominant structural style in Zone VIII.
310
B. CLENNELL
Time N.W. Sabah Offshore W. & N. Sabah units N W Outboard Belt Inboard Belt S E [ W
E. Sabah & W. Sulu Sea Basins SandakanBasin S.E. Sabah
I
~ ~ ' . [:'::::-"-.-~~
~
/'.~ it,,, [ - i" ~
u. f : : : : : : - ~ : i : : : : : \
,,,
z I
°1 0
_
~
ueep I-legional
fC::k
Unconformity/ Horizon"C",~
[ ! - - 3 - : ' : : ~ ~
~~_::.:a..
M. =============================
,,, | Z I U. uJ
.
0
E
G a n d u m a n & T o g o p i Fms.
~
/" i -- i ~ i/'~;~:,,~:~
~ ~ ~ . . - . . ' . ' . ' . . ~
::::::
~¢/
: : : : : :""r:":["::" " !" :" :" :" :"'"
:"
Youn er shalowmarnecastcs£_ nc ~andakan Fm& Tanjong I--m
~
Ne
ene
' :,i<.=": KS".:', ;i..";~. "::: :::.; ... .":::: :"!i..:..:
and Ouatemary:
......',!:::::::::::::::::::
"o8| M ~
"
:::!ii::>:::
,,, | L .
, ,^ ~
~ Crocker Formation and equivalents: [~ ~',~%,1 Palaeogenedeep marine dastics
,~ 0
~
Cretaceousophiolite and related
.............................. ~
~
C
"
~
_
_
~
Fig. 3. Simplified stratigraphic section across Sabah from NW to SE. Modified from Tan & Lamy (1990) and Clennell (1992).
NORTHWEST
A
~--II 1-
SOUTHEAST
~I~'
TROUGH
III
J"
IVa
COMPRESSIONAL TOE
2 m
_ Plio-Pleistocene ,~
_
~
A
P
SLOPE: EXTENDING COVER OVERLIES BURIED THUST BELT Neogene~ - ~ ~ , e , , ' ' , ~ - A \
"" ^
3Seafloor
-- A ~
......., ...... ", ; : i,"./.-"?""
....... "---=., ~ , ~ "
..: j:..."~; . . . . . . . . . ",,,
t--; 5 _
• 6- ~ ?
[
".
"."%%%. "
....... " - , ~ : x ",, o ~'~,. "'-,., ",,
: ""
Major Thrust System (Lower Tertiary "Crocker"sediments & ? Ophiolite)
':" ,'6 " J " " --.'~ ""
. . . . . .". ........... - " - . . . . . . . . . . . . . . . . . . . . . . . . . . Lower Thrust" Shee L.-M. Miocene . . . . . . . . . . t System (Neo-en .. ""
D'-'-~ I 0 kn'l
7--
. . ......... .~¢~.o
............ " C " . . " . .
]
~ • sea~ments)
" " ~ Northwest Sabah Platform: rifted continentalmargin basement and ? Mesozoicto Palaeogenesediment
--
D ~
Fig, 4. Structural cross-section A-A" across northwest Sabah Margin, re-drawn from unmigrated deep seismic line BGR 86-06 in Hinz et aL (1989). Vertical exaggeration is c. 10 x. Horizon 'C', at c. 17 Ma, corresponds to the end of an early Miocene compressional phase in the South China Sea. Abbreviations: A, Early Pliocene unconformity; B, Late Miocene uncf.-Shallow Regional Unconformity of Levell (1987); C, Latest Early Miocene uncf. (Deep Regional Unconformity); D, top NW Sabah Platform (?Oligocene to L. Miocene carbonate); 5, Miocene/Pliocene boundary; 6, Base Upper Miocene.
MIOCENE TECTONICS OF SABAH, MALAYSIA
Onland and regional structure of western Borneo The onland structure in the west of Sabah is dominated by an uplifted belt of folded and faulted turbiditic sediments (comprising Crocker, Sapulut and Trusmadi Formations) believed to have been accreted to Borneo since early Tertiary times, during some form of plate convergence, (Hamilton 1979; Hutchison 1988, i989; Rangin et al 1990b; B6nard et al. 1990). Tongkul (1987) presents a detailed structural and sedimentological model explaining the Crocker Formation of the Kota Kinabalu area as a packet of trench sediments accreted to northern Borneo in Palaeogene to early Miocene times during subduction of an oceanic slab ('Proto-South China Sea'; Fig. 1). It remains unclear how much oceanic lithosphere, if any, was subducted during this accretionary/compressive episode (compare Holloway 1982; Taylor & Hayes 1983; Hutchison 1988; Rangin et al. 1990b; Clennell 1992) but field outcrops in the Mount Kinabalu and Telupid areas suggest that the Crocker sediments are floored by ophiolitic basement. Certainly, the Neogene history of the North Borneo and Palawan area has involved only collision of continental material rifted from south Asia rather than subduction of any oceanic lithosphere. The rifled continental material is divided into a series of blocks by what may be throughgoing structures linking to transforms at the South China Sea spreading centre (Tongkul 1990, 1994). Although the trajectories of the individual blocks were broadly from the north, channelled by the lineaments, they were slightly divergent in rate and direction. The main period of convergence at the Borneo margin occurred during the Early to Middle Miocene, at which time the Palaeogene sediments were uplifted to form the mountains of the Crocker Range, which reach heights of 1-2 km above sea-level, even in their present eroded state. The Deep Regional Unconformity (DRU) is not a time-line of regional extent, but rather marks the period during which collision had the maximum impact on the sedimentary record in a particular place. In some areas, there is a marked angular discordance and a considerable section of missing stratigraphy, while in other places, the DRU is represented by a brief hiatus. The controls on this pattern depend not only upon local rates and degrees of tectonic uplift, but also on tectonic and eustatic events that influenced the source areas and distribution paths of the sediments (Tan & Lamy 1990; Rice-Oxley 1991). Thus the architecture of the unconformity depends on the combined effects
311
of erosion and sediment dispersal patterns. There was a second culmination in convergence in the mid Late Miocene which buckled the unconformity surface in many parts of Zones III and IV (Hazebroek & Tan 1993).
The Central Sabah-Western Sulu Sea basin system According to Hutchison (1992) the Central Sabah Basin can be regarded as an onland extension of the Southeast Sulu Sea Basin (Fig. 5). Results of scientific drilling (Nichols et al. 1990; Rangin & Silver 1990a, b) suggest that the Southeast Sulu Sea Basin was initiated in latest Oligocene or earliest Miocene time, by rifting of a pre-existing island arc and ophiolitic terrain (Fig. 2) whose onland extensions are now exposed in the Upper Segama area of Sabah, and in the Philippine island of Panay (Hinz et al. 1991). Subsequently, a limited amount of oceanic spreading occurred along a NE-SW axis, opening up a marginal basin bounded to the north by a rifted and subsided arc (Cagayan Ridge) and ophiolite terrain (the Northwest Sulu Sea Basin and Palawan Island), and to the south by the volcanic arc of the Sulu islands. The close proximity of the pole of rotation to the marginal basin meant that eastern Sabah lay at the hinge zone of the spreading (Hutchison 1992). Lithospheric stretching near this pole was presumably not sufficient to generate new oceanic crust (White & McKenzie 1989), but there is scattered evidence of basaltic magmatism along the axis of the Central Sabah basin in the relevant time interval, i.e. from the late Oligocene to the middle of the Miocene (Kirk 1968; Tamesis 1990; Clennell 1992). It is evident that the eastern Sabah to southeastern Sulu Sea region experienced mainly NW-SE extension from Early to Middle Miocene time. At some time in the late Miocene, opening of the Southeast Sulu Sea Basin was extinguished by the subduction of the spreading ridge southwards and eastwards beneath the Sulu Archipelago, and the larger Philippine islands of Negros and Panay (Hinz & Block 1990; Hinz et al. 1991). At this time the tectonic development of eastern Sabah became disconnected from that of the Sulu Sea, probably by the growth of a transform fault system running from the Sulu archipelago towards the Balabac Strait (Hinz et al. 1991). The geometry and kinematic linkage of this system remains obscure, but wrench faulting has apparently affected a series of upper Neogene basins running from the Balabac Basin (Beddoes 1976) through to the Sandakan Basin (Fig. 5).
312
B. CLENNELL
D ~
BALABAC 1118°
/
KEY
I 119° E
t
/
SEDIMENTS
~ ] Sandakan & Tanjong Formations
BALABA
D'
C
STRAIT
Kulapis Formation
,;
~;
~7~ Labang Formation
IGNEOUS ROCKS
/
, ....
i/
/ /
anggi I.
/
........ ....
/
""--....,.
t
",,. /
[ ~ Granitoids Miocene to
/
BANCAUAN BASIN
Volcanics Quaternary
/
/ /
I ~ Late Oligocene to
Middle Miocene pyroclastics and lavas
I'-,, I ",~"" I
I
~1",,. ""
.I
i. "-::.2-,
OTHER UNITS •
~l~ Earlyto Middle
,,,,,.
/
...,S A 3
Miocene melange units
Manlang
~ j Ophioliticbasement and oceanic cover sediments
SANDAKAN
B,
Neogene oceanic
1
I
i
lithosphere
f Cf
, 7
/
.---" Lineament
:. C. E .N T. R. A.L " ~ ' , ~ "v"
/,,
",,~::: 5:-::-:-~:;::G .:::.- <,'-" • ~_-"
J
j /
Garam . .•. . .,," . . Bukit .. .~,~ //
I
/
"
l~ ~t'~-
-
.
.~..
. .-. . . , .
.
/"
/
.'
5°
m
"....
~. I , ,
, . . . ,,),IBASIN .
DARVEL BAY
""~':""::~":::'"":" ~" •
,~
~ ~
;:((I Bp!n :::I ~
O
,/
"
" ".......... ~ it'*-
I KALIMANTAN
C E L E B E S
~¢^^^~^%~2-
i~tTAWAU ^ ' ^ ' ^ ~ ~ - - ' v~
[ 117°
'~ ~ ' J , "~~ - ~ ' ~~"~ 1 ~1118°
S EA
-',^^^,
SEMPORNA
[ 119°E
Fig. 5. Simplified geological map eastern of Sabah, based on Lim & Yin 1985, Tongkul (1991, 1993) and Clennell (1991, 1992). Offshore information from Hinz & Block (1990), Hinz et al. (1991) and Wong (1993).
Earlier
sediments
the Labang
in eastern
and Kulapis
Sabah,
formations
In the Central Sabah Basin, Oligocene to Lower Miocene sedimentation produced two contrasting units known as the Labang and Kulapis formations. The relationships between these units are largely obscured by later deformation but their general nature can be assessed from the available exposures.
The Labang Formation (Collenette 1965) consists of grey-coloured deep marine turbidites (full Bouma a-e sequences, and c--d-e sequences are commonplace) and interbedded hemi-pelagic shales. The entire Labang unit is deformed by folding and faulting, but even so, a lower shaly part, and an upper part with thicker and more abundant sandstones, can be distinguished. The foraminiferal and trace fossil assemblages both suggest that the sediments were deposited in a bathyal environment.
MIOCENE TECTONICS OF SABAH, MALAYSIA The Kulapis Formation (Collenette 1965) is a particularly unusual unit, as it consists of predominantly red-coloured shales and sandstones that can be shown by occasional, well preserved foraminiferal and nannofossil assemblages to have been deposited in a fully marine environment (L. Gallagher pers. comm. 1991). The most distinctive facies consists of thickly bedded sandstones, up to 10 m thick, that occasionally show graded bedding and convoluted flute marks at the base, but more usually are structureless, and may have been deposited by large grain-flows. Elsewhere the Kulapis Formation consists of thick mudstones with packets of more thinly bedded sandstones and siltstones that show normal grading and flute marks. The facies associations and fauna together suggest that the Kulapis Formation was deposited in a marine outer shelf environment with unstable slopes.
East Sabah melanges A remarkable feature of eastern Sabah is the great extent of mud-matrix melange units and associated chaotic rocks (dark stipple on Fig. 5). The source of the melange matrix, and most of the sandstone blocks, can be traced to disaggregated Labang Formation sediments (the grey matrix melanges). Subordinate volumes of melange and broken formation are derived from slumping of Kulapis Formation sediments (red matrix melanges), and these include large intact rafts of the massively bedded sandstones. There is a small amount of ophiolitic material within the melanges, and to the south of Sabah, both the matrix and blocks are sourced increasingly by volcaniclastics. The nature, origin and tectonic significance of the East Sabah melanges have been described in considerable detail by the author (Clennell 1991, 1992), and it was concluded that melanges were mostly produced by submarine slope failures that were triggered by tectonic rearrangement of the Central Sabah Basin at the end of the Early Miocene. Parts of the melanges are demonstrably of olistostromal origin, such as those associated with slump folds. Large volumes of weakly foliated scaly clay melange may have originated as submarine debris flows. Tracts of the melange which have a broadly horizontal fabric of more intense shearing, marked by zones of scaly clay and cataclasites, are believed to have been produced by tectonic shearing and gravitational sliding. Relatively rigid blocks of sand-rich Kulapis Formation sediments slid over a substratum of muddier Labang Formation sediments as the whole sequence underwent stratal extension that detached along melange horizons. A relatively small volume of melange occupies piercement structures
313
emplaced into Middle Miocene and younger sediments. These melanges are demonstrably of diapiric origin, and are associated with effusive mud volcanism which continues to the present day in eastern Sabah and in the Sandakan Basin.
Successor basins, the Tanjong and Sandakan formations Sedimentation in the Central Sabah Basin continued throughout the Miocene, with input of quartz-rich sediments from the uplifting Crocker Formation to the west and volcaniclastic material from active magmatic arcs to the south. These sediments spilled into the axial part of the Central Sabah Basin, forming a system whose NE-SW trend parallels that of the Sulu Sea. It is therefore widely presumed that the sediments were funnelled into rifted basins associated with extension and spreading of the Sulu Sea (e.g. Hutchison 1988, 1992; Tjia et al. 1990). This appears to be the most likely explanation of the sedimentation patterns, but direct evidence of extensional structures onland over much of eastern Sabah is lacking. Some normal faults with the correct trend have been traced in the field and in remote-sensing images by Tongkul (1993), but he points out that many of the larger NE-SW trending faults have a considerable wrench component. The association with basaltic magmatism and melange formation, and the trend of the later basins and their later structural style, provides indirect evidence that the successor basins are rift-related. In the south and central part of Sabah, the quartzrich sedimentary sequence is known as the Tanjong Formation, and it consists of mainly shallow marine facies with some thickly-bedded sandstones intercalated with thinner-bedded sandstones, siltstones and shales. Coarsening-upwards and thickening-upwards sequences are common, occasionally culminating in terrestrial deposits, including coals. In places the underlying Kulapis Formation sediments grade upwards into the Tanjong Formation, the red colour of the earlier sediments gradually giving way to the grey and yellow colours of the Tanjong. Where the Tanjong Formation overlies the Labang Formation or the melange units, there is generally a marked unconformity, examples of which can be seen on the Kinabatangan River near Bukit Garam. The earliest Tanjong sediments are upper Lower Miocene, and extend up into the Middle Miocene. This means that the Tanjong must have been deposited in quiescent areas while melanges were still being formed in other parts of the basin. In the Sandakan Peninsula, the Tanjong depositional system is represented by slightly
314
B. CLENNELL
-]
B
WSW ...... ~
-
.....
1
ENE '
.....~. . . . . . . ~ . - ......._::
,~,~,
:;"~4.'~',.~:~-~;:'~.~:~':_~-YZ:~
.
'
.--~---. --,-' ..........'...... ..~........ ~., =_-_.' .:.......~'....,-,~--,-::.~-x,,~,~_..-. . . . . 1 ,=: ~ .....
- ......
-
" ......
_
~
.....
~'~';~-'~-~TW.~';~.'~*,~,~""4~.~
~:......
-~
'
.....
~
_
'
Y
" ~ ' ~ . , ~ ~ :
~
-"~:'~--~:-='-¢~°!!~:~:,~'=~'~'-~..~="
.
~
B
..,
'
~. ~z~-~
4
,.~-.--,~-: ..........
U
~ - - ~ , ~ ~
-
.
~
-
-
~
.
~
(a)
WSW
ENE v
~
"
i
-
B,
Unit 3 ( P l i o - P l e i s t o c e n e ) ~ _ - -
1 --
.~ Unit lm(Base M i i ~ i o c
~
~.1%
_~__..~
¢9
3--
Lower Miocene shales, ~ volcaniclastics and m61ange 4
5
I
km |
m
(b)
younger sediments, whose oldest fossils date from the early part of the Middle Miocene (Lee 1970; Ujiie 1970, 1977; Clennell 1992). For this reason they are known as the Sandakan Formation, but it is thought that the Tanjong and Sandakan were probably contiguous at one time, as the mineralogical composition of the sands in the two formations are petrographically very similar (Clennell 1992). The Sandakan Formation is everywhere unconformable on older rocks: either volcaniclastics, tuffaceous sediments that can be correlated with the Labang Formation, or melanges. The youngest fossils in these melanges are themselves of earliest Middle Miocene age, so the unconformity at the base of the Sandakan Formation can be dated to the early Middle Miocene.
Subsidence, diapirism and circular basins Tjia et al. (1990) suggested, on the basis of facies and palaeocurrent correlations, that the Tanjong sedimentation was originally extensive throughout the Central Sabah Basin. From facies analysis, the pattern emerges of a prograding sequence of shoreline and terrestrial deposits, which originated in the south and become younger to the NE (towards the Sulu Sea). The Tanjong deposits later became segmented during subsidence that produced a series of circular sub-basins, three of which (the Maliau, Malibau and Bukit Garam basins) are shown on Fig. 5. The structure of these circular basins cannot be explained by a far-field stress with a single orientation of the principal components. Normal faults
MIOCENE TECTONICS OF SABAH, MALAYSIA
C
315
WNW
ESE
WNW
ESE
:- C '
(c)
C
~
Unit 3 (Plio-Pleistocene) ~ -------~1 Unit2 ? . ----~~-_.~-~'~ /
.--""""--~~
-C'
~,.~
\ -4. Q
1-N ~ ~Unit 1 (Base MiddleMiocene J \ ~ to mid UpperM i o c e n e ) j / ' /
[..a
~ 2 - - Lower Miocene ~N. ,,~"~ ~ volcarnMi2~n: ~
-~
,
.
. ~
~
/
/ \ \Q. Mud diapir
~
8
3-
,
5 km
i
LowerMioceneshalesand m~lange 4--
(d) Fig. 6. (a, b) Excerpt of commercial seismic line SB6-27 orientated down dip across part of the Sandakan basin immediately offshore the Kinabatangan estuary (section B-B' on Fig 5.). (c, d) Part of line SB6-01 orientated across the strike of the Kinabatangan circular basin (section C-C' on Fig. 5). Unit l, a prograding wedge of Middle and Upper Miocene sediments (Sandakan Formation equivalents and younger) builds out over thick wedge of shales and ?melanges which extrude diapirically at the basin edges. The overlying Plio-Pleistocene sequence (Unit 3) is generally aggragational and shows onlap to the southwest. Unit 2, a second progradational sequence deposited during a lower sea-level stand at the end of the Miocene, is preserved only in central parts of the basin. Stratigraphic information from Wong (1993).
generally trend tangentially to the basin structure and wrench faults are often orientated radially. Tham (1984) and Lee & Tham (1989) explain this as the result of differential subsidence of neighbouring sectors of the basin. The basin edges have tilted more than the basin interior, leading to higher stratal dips at the periphery than at the
centre, where the sediments are almost flat lying. There is evidence that some of this subsidence and tilting was syn-sedimentary, as slump structures, mud injections and convoluted flute marks attest to rapid and unstable deposition (Lim 1990; Clennell 1992). Also, some fault block rotations generated topographic highs within the Tanjong basin system
316
B. CLENNELL
that were exploited by reef-building organisms. The subsidence is greatest where the Tanjong sediments overly a great thickness of mud-rich sediments. For example, the Bukit Baram basin has a central high area where it overlies Kulapis Formation sediments, and two curved lobes to the NW, where the Tanjong sediments have subsided into melanges, and to the south where the Tanjong overlies the Labang Formation. To the north, the Sandakan Formation shows some of the same features as the Tanjong basins. The main outcrop on the Sandakan Peninsula is founded on volcanic basement that has not permitted great subsidence, but has extended to form a series of gently tilted fault blocks. Outlying basins have mostly been eroded, but around the Manjang river a small circular basin of Sandakan Formation sediments can be identified (Collenette 1966; Clennell 1992). Although this is less than 5 km across, field exposures and aerial photographs show that it has the same arcuate shape and internal structure as the larger circular basins. In the adjacent offshore area, the Middle to Upper Miocene progradational sequence continues into the Sandakan Basin (Wong 1993). A seismic dip line from offshore the Kinabatangan area (Fig. 6a, b) shows the main depocentre, with a wedge of sandy sediment (Unit 1, base Middle Miocene to mid Late Miocene in age) prograding over a substratum with chaotic internal reflectors, interpreted as thick shales or possibly melanges, which are pushed up ahead of the sand to form a diapiric wall. The corresponding strike line (Fig. 6c, d) shows the arcuate section of the basin, with the reflectors dipping more steeply at the basin margins than in the centre. The lateral margins of the basin are interpreted as being constrained by walls of mud-rich melange or shale pushed up as the sandy sediments sank into them. If exhumed by later uplift, this basin would presumably resemble the circular basins currently exposed onshore. A number of other sagged features can be identified in the Sandakan basin, including some large basins several kilometres across and at least one smaller circular depression only c. 1 km across, that is, on the scale of the Manjang basin. Widespread mud diapirism is evident in the offshore area where mud volcanic islands and 'mud lump' islands can be found north of Sandakan, and also onshore, in the Dent Peninsula, where there are several large mud volcanoes (Rheinhard & Wenk 1951; Haile & Wong 1965; Wilford 1967; Tongkul 1989). The clay-mineral composition of the muds erupted from eastern Sabah mud volcanoes shows that they are derived from the Labang Formation or from the clay matrix of the melange units (Clennell 1992).
Relationship with Neogene tectonic events Both the Kulapis and Labang Formations are affected by intense tectonism, which produced large-scale folding and faulting across the Central Sabah Basin. Since the overlying Tanjong sediments are essentially undeformed except for synsedimentary structures (e.g. Tongkul 1993), the main deformational episode must have occurred at around the time of the melange formation, i.e. from late Early Miocene to early Middle Miocene. From direct geological and biostratigraphic correlation, the foregoing events can be tied to the opening and spreading of the Southeast Sulu Sea marginal basin (see Hutchison 1992). Therefore, the main sense of lithospheric strain in the region during the Lower to Middle Miocene could be assumed to be extensional and to operate in a NE-SW direction. However, exactly the same time interval is characterized in the northwest Borneo offshore by the major shortening and uplift that produced the Deep Regional Unconformity. The DRU can be traced continuously through the Balabac strait from the South China Sea into the Sulu sea, on the seismic data of Hinz et al. (1989, 1991); and on unpublished commercial data. It has already been argued that this unconformity is not a time line, but rather an 'event' line that marks the propagation of a major phase of structure formation across the region. To the west the structures produced are mainly compressional, with wrenching becoming increasingly important in the northeast. In the Balabac basin, the northernmost sub-basin of the Sulu sea (Fig. 7), the DRU is clearly distinguishable as the boundary between unreflective Lower Miocene sediments and the overlying stratified sequence of middle to late Miocene age. The basin is floored by oceanic crust which has been faulted and imbricated. The structure of the Balabac basin suggests that it may have formed originally as a rift basin, which was subsequently inverted by NW-SE compression (Hinz & Block 1990; Hinz et al. 1991). The DRU and overlying reflectors of up to Late Miocene age are buckled upwards in the immediate vicinity of the overriding basement, and draped by undeformed PlioPleistocene sediments. This brackets the age of a second phase of tectonism to the later Miocene. To the south of the Balabac Basin, the basement of the Banggi Ridge has been pushed over the Miocene sediments, so that it forms a 'pop-up' structure that verges to the NW in the north and to the SE into the Bancauan Basin. Contemporaneous shortening in eastern Sabah is illustrated by two roadside exposures. At Telupid ophiolitic basement overrides grey-matrix melange and, further west, sediments of the Kulapis Formation, in a high-angle imbricate sequence.
317
MIOCENE TECTONICS OF SABAH, MALAYSIA
NORTHWEST
D
<
PALAWAN HIGH
.......-./;
.~.,~ ~ N
_.~.-" ~
SOUTHEAST J.,, rl ~
Seafloor
BALABAC BASIN Plio-Pleistocene
Mid Miocene
~l ~ BANGGI RIDGE rl" / U. Miocene- L. Pliocene
,;v,:..~-~~--..::.
B ~";'.-:'.~,, .," . . . . . . . . )-.-- ._, . -. . . . - - . - - - - - ~ C a = . ~ ~ v , ~ Palaeo.gene and Lower Miocene deep marine clayey sediments
D
"'-. "','.'., ? Palaeo~ene and ""-..":'L-M. Miocene " , , ~ . . ~
~4
~6 I
I
? Palaeogene sediment
Fig. 7. Structural cross-section D-D' across Northwestern Sulu Sea (location shown on Fig. 5), redrawn from unmigrated deep seismic line BGR 86-32 in Hinz et al. (1991). Vertical exaggeration is c. 10 ×. The Balabac Basin is bounded to the northwest and southeast by compressional ridges of imbricated ophiolitic basement. The initial tectonic phase ceased in the latest Early Miocene and is marked by Horizon 'C', the equivalent of the Deep Regional Unconformity. Later compression, inverting the basin, occurred in early to mid Upper Miocene time. Abbreviations: A, Early Pliocene unconformity; B, base Late Miocene unconformity; C, Late Early Miocene unconformity (Deep Regional Unconformity of Levell 1987); D, top oceanic crust (Cretaceous ophiolite).
The transport direction is towards the NNW. The amount of shortening here is modest (probably less than 1 km), and dies out towards the east. The timing of this compression seems to tie with the age of possible inversion noted in the offshore; certainly it post-dates the melange formation and deposition of the Kulapis Formation. From south of the Segama River almost to the Kinabatangan River, the original unconformity surface between the Labang depositional system and the overlying sediments is almost completely lost among later compressional structures. Early to Middle Miocene sediments, which are the stratigraphical equivalent of the Tanjong Formation, but richer in volcaniclastic detritus, are imbricated with slices of Labang Formation sediments and melanges. This area is structurally and stratigraphically analogous to the higher levels of the seismic section in Fig. 7, where sediments overlying the DRU are involved in the shortening. While some areas of the Central Sabah Basin have experienced this later compressional phase, most of the Tanjong, Sandakan and equivalent sediments that were deposited north of the Kinabatangan River and in the Malibau and Meliau Basins are essentially undeformed. Minor inversion and wrench structures of mid-Late Miocene age are evident in the Sandakan and Bancauan basins (Bell & Jessop 1974; Tamesis 1990; Wong 1993; and unpublished commercial data). These movements are very modest compared with the structures shown in Fig. 7. It is suggested that these later events are expressions of far-field tectonic forces that are transmitted through rigid crustal
blocks, but only expressed strongly near to block margins (e.g. around the Telupid and Segama ophiolite blocks, and along the Banggi and Bancauan ridges shown on Fig. 5). As an additional factor, in the interior of the Central Sabah Basin, the cover sediments may have been effectively decoupled from the effects of tectonism by the thick shales and melanges which form the main substratum to the Tanjong depositional system. Tanjong/Sandakan sedimentation was therefore able to prograde continuously northeastwards along the axis of the Central Sabah Basin without any major interruptions, while some surrounding areas were being uplifted. In the sedimentary record this history is expressed as continuing sedimentation from the Middle Miocene to lower Pliocene without any marked tectonically produced unconformities. Indeed, the post Middle Miocene sediments of the Sandakan Basin are amenable to sequence stratigraphic analysis (Wong 1993) demonstrating that eustatic sea-level changes were large compared with tectonic movements. Many of the main structural features in the basin developed as a consequence of local gravitational instabilities rather than regional tectonic events (c.f. Talbot 1992).
Summary Both 'far field tectonics', i.e. stresses ultimately driven by motions of major plates, and gravity tectonics (diapirism and slumping) have interacted to influence basin process in the Sabah region. A major unconformity surface, the Deep Regional
?
318
B. CLENNELL
Unconformity, can be traced across the northwest Sabah offshore and into the onshore of eastern Sabah. The Deep Regional Unconformity marks a major reorganization of the regional stress and strain field that produced a widespread interruption in deposition. Rather than being a time line this surface is diachronous, and was formed largely by c o m p r e s s i o n in the n o r t h w e s t and mainly by extension in the southeast (Central Sabah and Sandakan Basins). In eastern Sabah this upheaval was marked by the generation of extensive mudmatrix melanges and a change from deep marine to shallow marine and terrestrial deposition. While basin initiation in Sabah was controlled by the main regional tectonic events, the internal architecture that developed through the Miocene was controlled by local patterns of subsidence and uplift of the substratum. Where this material is relatively rigid, the effects of tectonically-driven compression, extension and wrenching were transmitted directly to the basin sediments. This produced inversion structures that are localized to the margins of the main basement blocks, such as the Banggi Ridge and the Telupid ophiolite block.
Where the succeeding sediments prograded over thick shale-rich sequences or m u d - m a t r i x melanges, the basin interiors are largely decoupled from far-field tectonic influences, but have undergone gravity-driven sagging and m u d diapirism, producing a series of remarkable circular basins in the Central Sabah and Sandakan basins. This paper led from a PhD project supervised by Dr Tony Barber at Royal Holloway, University of London and Dr Felix Tongkul of Universiti Kebangsaan Malaysia, Sabah. Their continued help has been invaluable. The project was fully funded by Shell International Petroleum Co. Ltd. Further support came from the University of London SE Asia Research Group: sincere thanks go to its administrator, Diane Cameron. Professors K. Hinz and Claude Rangin provided much useful information. Seismic lines of the Sandakan Basin were provided by W.M.C. (Malaysia) SDN BHD. Particular thanks go to Terry Walker, and to Petronas SDN BHD for allowing access to, and publication of, this data. Additional help and ideas came from Herman Soediono, Lim Pen Siong, Michiel de Smet and Chris Talbot. This paper has benefited greatly from reviews by Chris Elders and Felix Tongkul.
References BEDDOES, L. R. 1976. The Balabac sub-basin, southwestern Sulu Sea, Philippines. SEAPEX Program, South East Asia Conference, Paper 15, 1-22. BELL, R. M. & JESSOP, R. G. C. 1974. Exploration and Geology of the west Sulu Basin, Philippines. Journal of Australian Petroleum Exploration Association, 14, 21-28. BI~NARD, E, MULLER, C., LETOUZEY,J., RANGIN, C. & Tahir, S. 1990. Evidence of multiphase deformation in the Rajang-Crocker Range (Northern Borneo) from Landsat imagery interpretation: Geodynamic implications. Tectonophysics, 183, 321-339. BEN-AVRAHAM,Z. 1989. Multiple opening and closing of the Eastern Mediterranean and South China Basins. Tectonics, 8, 351-362. BOL, A. J. & VAN HOORN, B. 1980. Structural style in western Sabah offshore. Bulletin of the Geological Society of Malaysia,12, 1-16. BRIAIS, A., TAPPONNIER, P. & PAUTOT, G. 1989. Constraints of Sea Beam data on crustal fabrics and seafloor spreading in the South China Sea. Earth and Planetary Science Letters, 95, 307320. CLENNELL, M. B. 1991. The origin and tectonic significance of mdlanges in Eastern Sabah, Malaysia. Journal of South East Asian Earth Sciences, 6, 407-429. 1992. The Mdlanges of Sabah, Malaysia. PhD Thesis, University of London. COLLENETTE, P. 1965. The Geology and Mineral Resources of the Pensiangan and Upper Kinabatangan area, Sabah. Malaysian Geological Survey, Borneo Region, Memoir, 12. 1966. The Garinono Formation, Sabah, Malaysia.
-
-
In: Malaysia Geological Survey, Borneo Region, Annual Report for 1965. 161-167. DALY,M. C., COOPER,M. A., WILSON, I., SMITH,D. G. & HOOPER, B. G .D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine & Petroleum Geology, 8, 3-21. DAMUTH, J. E. 1994. Neogene gravity tectonics and depositional processes on the deep Niger Delta continental margin. Marine & Petroleum Geology, 11,320-346. HALLE, N. S. & WONG, N. E Y. 1965. The Geology and Mineral Resources of the Dent Peninsula, Sabah. Malaysia Geological Survey, Borneo Region, Memoir, 16. HAMILTON,W. 1979. Tectonics of the Indonesian Region. United States Geological Survey, Professional Paper, 1078. HAZEBROEK, H. P. & TAN, D. N. K. 1993. Tertiary tectonic evolution of the NW Sabah continental margin. Bulletin of the Geological Society of Malaysia, 33, 195-210. HINZ, K. & BLOCK, M. 1990. Summary of geophysical data from the Sulu and Celebes Seas. In: RANGIN, C. StaYER, E. A. & VON BREYMANN, M. T. (eds) Proceedings of the ODP, Initial Report, Ocean Drilling Program, 124, 87-91. , KUDRASS, H. R., & MEYER, H. 1991. Structural elements of the Sulu Sea, Philippines. Geologisches Jahrbuch, A127, 483-506. - - - , FRISCH,J., KEMPTER,E. H. K., MANAFMOHAMMAD, A., MEYER, J., Er AL 1989. Thrust tectonics along the continental margin of Sabah, Northwest Borneo. Geologische Rundschau, 78, 705-730. HOLLOWAY, N. H. 1982. North Palawan block - its
MIOCENE TECTONICS OF SABAH, MALAYSIA relation to Asian mainland and role in evolution of South China Sea. AAPG Bulletin, 66, 1355-1383. HtrrCHISON, C. S. 1988. Stratigraphic-tectonic model for E. Borneo. Bulletin of the Geological Society of Malaysia, 22, 135-151. 1989. The Geological Evolution of S.E. Asia. Oxford University Press. 1992. The Southeast Sulu Sea, a Neogene marginal basin with outcropping extensions in Sabah. Bulletin of the Geological Society of Malaysia, 32, 89-108. KrRK, H. J. C. 1968. The Igneous Rocks of Sarawak and Sabah. Malaysian Geological Survey Bulletin, Borneo Region, 5. LEE, C. P. & THAM, K. C. 1989. Circular Basins of Sabah. Proceedings Geological Society of Malaysia Petroleum Geology Seminar, Kuala Lumpur, 1989 (abs.). Bulletin of the Geological Society of Malaysia, 29, 54. LEE, D. T. C. 1970. Sandakan Peninsula, Eastern Sabah, Eastern Malaysia. In: Geological Survey of Malaysia, Borneo Region, Report, 6, 1-75. LEICHTI, P., ROE, F. W. & HAILE,N. S. 1960. The Geology of Sarawak, Brnnei and the western part of North Borneo. British Territories of Borneo, Geological Survey Department, Bulletin, 3. LEVELL, B. K. 1987. The nature and significance of regional unconformities in the hydrocarbon-bearing Neogene sequence offshore West Sabah. Bulletin of the Geological Society of Malaysia, 21, 55-90. -• KASUMAJAYA, A. 1985. Slumping of the late Miocene shelf-edge offshore West Sabah: a view of a turbidite basin margin. Bulletin of the Geological Society of Malaysia, 18, 1-30. LIM, E.H. 1990. Geologi am kawasan timur Kota Kinabatangan, Sabah. BSc Thesis, Universiti Kebangsaan Malaysia, Sabah [in Malay]. LtM, P. S. & YIN EE HENG. 1985. Geological Map of Sabah, Scale 1:500 000. 3rd. Edition. Geological Survey of Malaysia. MASCLE, A. & BISCARRAT, P. A. 1979. The Sulu Sea: a marginal basin in South East Asia. In: WATKrNS, J. S., MONTADERT, L. & DICKERSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 373-381. NICHOLS, G. J., BETZLER, C., BRASS, G. W., HUANG, Z., L/NSLEY, B., ET AL. 1990. Depositional History of the Sulu Sea from ODP sites 768,769 and 771. Geophysical Research Letters, 17, 2065-2068. RANG/N, C. & SILVER, E. 1990a. Geological setting of the Celebes and Sulu Seas. In: RANG/N, C., SILVER, E. A. & VON BREYMAYN,M. T. (eds) Proceedings of the ODP, Initial Report. Ocean Drilling Program, 124, 35-41. & -1990b. Neogene tectonic evolution of the Celebes-Sulu Basins: New insights from Leg 124 drilling. In: RAYG/N, C., StaYER, E. A., & YON BREYMANN, M. T. (eds) Proceedings of the ODP, Initial Reports. Ocean Drilling Program, 124, 123-144 --, JOLIVET, L., PUBELLIER, M. & the Tethys-Pacific working group. 1990a. A simple model for the tectonic evolution of SE Asia and Indonesia region -
-
-
-
319
for the last 43 m.y. Bulletin de la Socidt~ g~ologique de France, 8, 889-905. - - - , BELLON, H., BENARD, F., LETOUZEY,J., MOLLER, C. & SANUDIN, Y. 1990b. Neogene arc-continent collision in Sabah, Northern Borneo (Malaysia). Tectonophysics, 183, 305-319. , SILVER, E. A., HINZ, K., KUDRASS,H. & ODP Leg 124 Scientific Party 1991. Structural setting of the Sulu Basin: New insights from drilling. ODP Scientific Results, 146. Ocean Drilling Program. , PUBELLIER,M., AZt~MA,J., BRIA1S, A., CHOTIN, P., eT AL. 1990C. The quest for Tethys in the western Pacific. 8 palaeogeodynamic maps for Cenozoic time. Bulletin de la Socidtd gdologique de France, 8, 907-913. RHEINHARD, M. & WENK, E. 1951. The Geology of the Colony of North Borneo. British Borneo Region Geological Survey Bulletin, 1, 7-11. RIcE-OXLEY, E. D. 1991. Palaeoenvironments of the Lower Miocene to Pliocene sediments in offshore NW Sabah Area. Bulletin of the Geological Society of Malaysia, 28, 165-194. Ru, K. & PIGOTT, J. D. 1986. Episodic rifting and subsidence in the South China Sea. AAPG Bulletin, 70, 1136-1155. TALBOT, C. J. 1992. Quo Vadis Tectonophysics? With a pinch of salt! Tectonophysics, 16, 1-20. TAMESIS, E. V. 1990. Petroleum geology of the Sulu Sea Basin, Philippines. SEAPEX Proceedings, Volume IX, (Eighth Offshore Southeast Asia Conference, Singapore). 45-54. TAN, D. N. K. & LAMY, J. M. 1990. Tectonic evolution of the NW Sabah continental margin since the Late Eocene. Bulletin of the Geological Society of Malaysia, 27, 241-260. TAYLOR, B. & HAYES, D. E. 1983. Origin and history of the South China Sea Basin. In: HAYES,D. E. (ed.) The tectonic and geologic evolution of SE Asian seas and islands, Part 2. American Geophysical Union, Geophysical Monographs Series, 27, 23-56. THAM, K.C. 1984. Geology of the Bukit Garam Area, Sabah, East Malaysia. BSc Thesis, University of Malaya. TJIA, H. D., KOMOO,I., LIM, P. S. ~¢ TUNGAHSURAT. 1990. The Maliau Basin, Sabah: geology and tectonic setting. Bulletin of the Geological Society of Malaysia, 27, 261-292. TONGKUL, F. 1987. Sedimentology and structure of the Crocker Formation in the Kota Kinabalu Area, Sabah, East Malaysia. PhD Thesis, University of London. 1989. Geological control on the birth of the Pulau BatH Hairan mud volcano, Kudat, Sabah. Warta Geologi, 14, 153-165. 1990. Structural style and tectonics of western and northern Sabah. Bulletin of the Geological Society of Malaysia, 27, 227-239. 1991. Tectonic evolution of Sabah, Malaysia. Journal of South East Asian Earth Sciences, 6, 393-406. 1993. Tectonic control on the development of Neogene basins in Sabah, East Malaysia. Bulletin of the Geological Society of Malaysia, 27, 95-103. -
-
-
-
320 --
B. CLENNELL
1994. The Geology of northem Sabah, Malaysia: its relationship to the opening of the South China Sea Basin. Tectonophysics, 235, 131-137. UJHE, H. 1970. Miocene foraminiferal faunas from the Sandakan formation, north Borneo. In: Contributions to the Geology and Palaeontology of S.E. Asia, University of Tokyo Press, 8, 165-185. 1977. New species and subspecies of benthonic foraminifera from the Miocene Sandakan Formation, north Borneo. In: Contributions to the Geology and Palaeontology of S.E. Asia, University of Tokyo Press, 18, 87-102.
WHITE, R. & MACKENZIE, D. E 1989. Magmatism at rift zones: the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research, 94, 7685-7729. WILFORD, G. E. 1967. The mud volcanoes of Sabah. Journal of the Sabah Society, 3, 12-21. Sabah Museums Service, Kota Kinabalu. WONG, R. H. E 1993. Sequence stratigraphy of the Middle Miocene to Pliocene southern offshore Sandakan Basin, E. Sabah, Malaysia. Bulletin of the Geological Society of Malaysia, 33, 129-142.
Mesozoic and Cenozoic plutonic evolution of SE Asia: evidence from Sumatra, Indonesia W. J. M c C O U R T 1, M. J. C R O W 1, E. J. C O B B I N G 2 & T. C. A M I N 3
1 British Geological Survey, International Division, Keyworth, Nottingham NG12 5GG, UK 2 Consultant Geologist, 25 Main Street, Radcliffe-on-Trent, Nottingham, UK 3 Geological Research and Development Centre, Bandung, Indonesia Abstract: New K/Ar mineral ages from the Barisan Mountains of southern Sumatra suggest
four main periods of plutonic activity: Miocene-Pliocene (20-5 Ma), Early Eocene (60-50 Ma), Mid-Late Cretaceous (l17-80Ma) and Jurassic-Early Cretaceous (203-130Ma). These and all other published ages from exposed plutons in western Sumatra indicate a further period of plutonic activity in the Permian (287-256 Ma). They also suggest either that Early Mesozoic activity began in the Late Triassic, or that there were two distinct magmatic cycles, one in the Late Triassic to Early Jurassic (220-190 Ma) and one extending from the Mid-Jurassic to Early Cretaceous (170-130 Ma). In addition, poorly controlled ages from eastern Sumatra indicate that the important Triassic to Early Jurassic (240-195 Ma) tin-belt magmatism of the peninsular Malaysia Main Range Province extends into that area. Preliminary geochemical studies on the Mesozoic granitoids of the Barisan Range of southern Sumatra confirm that they are calc-alkaline, I-type, metaluminous, subduction-related volcanic arc granites (VAG). They broadly correspond to the southerly extension of a combination of the Central Valley and Western Granite Provinces of Thailand and Burma, and underline the fact that there has been a history of subduction-related magmatism along the southwestern edge of Sundaland since earliest Mesozoic times. The plutonic suites are crudely arranged in subparallel, locally overlapping, NW-SE trending belts, focused along deep-seated faults that have acted as magmatic conduits. It is proposed as a preliminary model that breaks in plutonic activity broadly correspond to changes in approach angle and/or rate of subduction, and that in some instances at least they relate to periods of collision and accretion of allochthonous material (terranes, slivers or blocks) of both oceanic and continental character. At least two such events seem to have occurred during the Mesozoic-Cenozoic tectono-plutonic evolution of Sumatra. One in the early Middle Cretaceous reflects collision and accretion of the oceanic Woyla terranes, and one in the latest Cretaceous is possibly related to collision of a continental sliver/block, the West Sumatra terrane, to the Sundaland margin.
Plutonic rocks are widely exposed throughout the Barisan Mountains of western Sumatra, and locally exposed, through Tertiary-Quaternary cover sequences, in eastern Sumatra. In general terms the granitoid rocks of the Barisan Mountains, define a series of sub-parallel, N W - S E trending, southwesterly migrating plutonic belts of TriassicJurassic, Cretaceous and Late Tertiary age (Katili 1973; Gafoer & Purbo-Hadiwidjoyo 1986) and a less well-defined Eocene belt (McCourt & Cobbing 1993). The presence of these plutonic belts and their assumed calc-alkaline composition, has been extensively quoted as evidence that Sumatra has been located in a region of plate convergence, marked by continental margin subduction, since the Early Mesozoic and perhaps earlier (Katili 1973; Hamilton 1979; Cameron et al. 1980). As part of the Southern Sumatra Geological and Mineral Exploration Project (SSGMEP, 1989-94) a reconnaissance field study and sampling programme was undertaken in 1992 of the main
granitoid plutons of the Barisan Mountains of southern Sumatra. The aim of this study was to establish the geological and geochemical characteristics and the isotopic ages of the plutons in order to evaluate the role of plutonism in the tectonic evolution of this part of Sundaland. During this investigation thirteen intrusive bodies ranging from individual plutons to batholiths were examined; all exposed major lithologies from each were sampled for whole rock geochemical analysis, and selected samples dated using the K-Ar method on mineral separates. The geochemical and geochronological data, comprising major element oxides and trace element compositions, new K-Ar mineral ages and a summary of all published ages from plutonic rocks of Sumatra and the tin-islands, quoted in this paper are available as Supplementary Publication No. SUP 18098 (8 pp) from the Society Library or the British Library D o c u m e n t Supply Centre, Boston Spa, Wetherby, W. Yorks LS23 7BQ, UK. The study has established that the Barisan
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 321-335.
321
322
W. J. McCOURT
Mountains granitoids range in age from Late Triassic to Pliocene, and has confirmed that they are I-type in character.
Tectonic setting and regional geological framework Sumatra lies along the southwest margin of the Sundaland cratonic block, the SE Asian continental extension of the Eurasia Plate (Fig. 1). To the west, oceanic crust of the India plate is being obliquely subducted beneath Sumatra along the Sunda trench in a N20(E direction at a rate of between 6-7 c m a -1 (Curray et al. 1979; Hamilton 1979). This zone of oblique convergence forms part of the Sunda arc-trench system which extends for more than 5000 km from Burma to eastern Indonesia. The Tertiary-Recent Barisan volcanic arc of western Sumatra and the pre-Tertiary sequences are variably broken up into a number of NW-SE
ET AL.
trending slivers by major faults which are strands of the Sumatra Fault System (SFS), some of which are Mesozoic in age. The SFS as defined in this paper, incorporates all those (NW-SE) faults along and west of the eastern foothills of the Barisan Mountains and includes the Great Sumatra Fault (van Es 1919; van Bemmelen 1949; Westerveld 1952). Many of the faults of the SFS have a history of multiple movement and reactivation and are interpreted as deep-seated, subduction related, continental margin structures. To the east of the magmatic arc and the SFS, are the back-arc basins containing thick sequences of Tertiary to Quaternary sediments, underlain by probable Upper Palaeozoic continental basement (Gafoer 1990). The pre-Tertiary continental core of SE Asia is interpreted as a complex assembly of various tectono-stratigraphic terranes including allochthonous micro-continental blocks, island arcs and accretionary complexes bounded by regional
Subduction zones
'~X
Transcurrent Fault Direction of Plate movement 20-
Philippine Sea Plate 10 m
Ocean
Plate lOO I
/ 11o ~ I
12¢
/
Fig. 1. Tectonic setting of Sumatra on the SW margin of Sundaland (from Hutchison 1989).
MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
faults, some of which are interpreted as sutures, marked by slivers of ophiolite, melanges and volcanic-plutonic arcs (Mitchell 1981; Stauffer 1983; Audley-Charles et al. 1988; Metcalfe 1988, 1990, among others). The pre-Tertiary framework of Sumatra is similarly interpreted as a mosaic of oceanic and continental terranes (Fig. 2) assembled through processes of microplate convergence and accretion since the Middle Permian (Pulunggono & Cameron 1984; McCourt et al. 1993). Much of central and eastern Sumatra is believed to be underlain by continental crust and the oldest exposed units are Upper Palaeozoic (PermoCarboniferous) metasediments. Pulunggono & Cameron (1984) proposed that the basement comprised two continental terranes, the Mergui and Malacca microplates, separated by a Triassic suture complex, the Mutus Assemblage (Fig. 2). The dominantly Palaeozoic Mergui microplate comprises, 'older granitic basement', PermoCarboniferous clastic and carbonate metasedimentary sequences, Lower to Middle Permian volcanics and associated sediments and Middle to Upper Triassic marine sediments. The Malacca microplate, which is confined to subcrop in Sumatra, consists predominantly of quartzites (Eubank & Makki 198t; Koning & Darmono 1984). It is considered to be of Carboniferous to Middle Permian age based on lithological and structural correlations with the Kenny Hill Formation of peninsular Malaysia (Fontaine & Gafoer 1989) and accordingly is part of the Gondwana affinity terrane of peninsular Malaysia (Hutchison 1994). The Mutus Assemblage, described from well records by Eubank & Makki (1981), comprises a mixture of radiolarian chert, red-mauve shales, thinly bedded limestone, a sandstone-shale sequence and basalt. It was assumed to be Triassic by Pulunggono & Cameron (1984), but Hutchison (1994) has proposed that it represents a Palaeozoic subduction-accretion complex. The Permo-Carboniferous metasedimentary basement sequences of the Mergui microplate are exposed along the eastern flank of the Barisan Mountains and through the Tertiary cover of the present day back-arc zone. They comprise in broad terms, glacio-marine coarse clastic pebbly units, finer-grained turbiditic slates interpreted as distal equivalents, and warmer water marine shelf and slope deposits including massive limestones (Simandjuntak et al. 1991). The contact between the sequences of glacio-marine sediments (Bohorok-Mentulu and Kluet-Gangsal formations) and warmer water marine-shelf sediments (Kuantan-Terantam formations) is interpreted to be tectonic and the two are considered to represent separate terranes with differing Late Palaeozoic tectonic histories that probably united in the Middle
323
Permian (Gafoer 1990; McCourt et al. 1993). In south Sumatra this contact corresponds to the regional, NW-SE striking, fault separating (or juxtaposing) the basement sequences of the Tigapuluh and Duabelas Mountains (Fig. 2), structurally corresponding to the Medial Sumatra Line of Hutchison (1994). The Permo-Triassic basement units are exposed along the axial region of the Barisan Mountains. They comprise a lower sequence of Lower to Middle Permian andesitic volcanics, reefal limestones and carbonaceous sediments (Palepat, Silungkang and Mengkarang Formations) and an upper one of Middle to Upper Triassic pelagic sediments and limestones (Kuala and Tuhur Formations). In north Sumatra, the Permian volcanic arc association has recently been interpreted as a separate terrane from the PermoCarboniferous sequences (Wajzer et al. 1991) and this model is followed here. The contact of the Triassic units, where seen, is disconformable on the Permian volcanic association and unconformable on the Permo-Carboniferous sequences. Upper Mesozoic sequences occur along the western side of the Barisan Mountains in southern Sumatra, in tectonic contact with older units to the east. They comprise varying proportions of oceanic basalts, ultrabasic rocks interpreted as ophiolitic, andesitic lavas and volcaniclastics, pelagic sediments and cherts, fine-grained turbiditic units, shallow marine clastics, limestones, phyllitic rocks and schists. The majority of these are (?Mid- to) Upper Jurassic to Lower Cretaceous and comprise a series of allochthonous slivers that correspond in part to the Woyla terranes of Wajzer et al. (1991) and belong to the Woyla microplate of western Sumatra (McCourt et al. 1993). In broad terms the Tertiary sequences of Sumatra correspond to three distinct depositional environments: sediments and minor volcanics of the forearc zone, calc-alkaline volcanics and sediments of the magmatic arc zone and sediments of the back-arc zone. During the Quaternary, Sumatra was the scene of tremendous volcanic activity throughout the Barisan Range, and the modern physiography was established. Tectonism and uplift in the latest Miocene and Plio-Pleistocene caused inversion of the Cenozoic sedimentary basins, development of widespread NW-SE trending fold structures, and reactivation of deep-seated basement faults.
Granitoid plutonism of Sumatra Granitoid rocks are present throughout the Barisan Mountains and are particularly common along and close to the junction of the Mergui and Woyla microplates (Figs 2 and 3). Permian magmatism
324
W. J. McCOURT ET AL. 1. MALACCA MICROLATE 2, MERGUI MICROPLATE A. Bohorok Tigapuluh Terrane B. Kluet-Kuantan.DuabelasTerrane C, Palepat Terrane D. Kuala FM
:
3. WOYLA MICROPLATE
A. Woyla terrane B. Pasamar~ terrane C. Gumai-Garba terrane
4. CONTINENTAL FRAGMENTS
A. Sikuleh B. Natal C Bengkulu
2A
EQUATOR
0 I
250km J v
V
v
'~
v vv~
Fig. 2. Simplified pre-Tertiary microplate configuration of Sumatra (modified from Pulunggono & Cameron 1984).
is indicated by Rb-Sr ages of 264 + 6 (Sibolga granite) and 256 +_6 Ma (Ombilin granite), a K-Ar muscovite age of 287 _ 3 Ma (Ombilin granite) and by restricted exposures of andesitic volcanics. The Sibolga granite is a composite batholith recording several intrusive phases that intrudes metasediments of the Kluet Formation (Aspden et al. 1982). The main rock types are porphyritic K-feldspar biotite-hornblende granite, monzogranite, quartz diorite and diorite. In addition to the Permian age several younger dates are available suggesting mainly Late Triassic (219-206Ma) and Late Jurassic (147-144Ma) magmatic activity. The Ombilin granite is a foliated muscovite granite without clear field relationships, although Silitonga & Kastowo (1975) imply that it intrudes the meta-
sedimentary Kuantan Formation. The Permian volcanic arc appears to be restricted to a fault slice along the western margin of the pre-Mesozoic metasedimentary basement. Evidence for widespread 'Triassic' plutonism is seen in the abundance of intrusive ages ranging from 230-195 Ma from Sumatra and the adjacent tin-islands. This phase of plutonism appears to correspond to two separate belts: one in eastern Sumatra and the tin-islands, the southerly extension of the Main Range and Eastern Granite Provinces of peninsular Malaysia, and the other in the Barisan Mountains of western Sumatra. The latter comprises hornblende-bearing biotite-granitoids of 1-type character with a wide compositional range which are predominantly quartz diorites to grano-
MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
D
325
Tertiary/(Eocene and Mid-Pliocene) Plutonic Arc Mid-late Cretaceous Plutonic Arc
EZ] ( Late ) Triassic-Jurassic/Early Cretaceous Plutonic Arc
SIKULEH GRANITE
[ ~ Permian Plutonic-Volcanic Arc SIBOLGA NITE
•HATAPANG~X GR NIT
MANUNGGAL BATHOLITH.~
GRANITE OMBILIN GRANITE
BATHOLITH
PLUTON
Fig. 3. Distribution of the principal plutonic belts of (western) Sumatra and location of the main plutons.
diorites. This western Sumatra Triassic to Early Jurassic plutonic arc, is located close to the edge of the Mergui microplate intruding metasediments of the Kluet-Kuantan terrane and its trace coincides with a compositionally similar but more extensive Middle Jurassic-Early Cretaceous (170-130 Ma)
and granitic plutons which commonly intrude Upper Oligocene to Miocene andesitic volcanics to the west of the Semangko Fault.
arc.
In terms of their overall major element oxide geochemistry (Table 1 & Fig. 4), the granitoids are typical subduction-related, I-type granites (Chappell & White 1974). Plutonic rocks range in composition from gabbro to granite on all common classification-nomenclature plots, with a concentration of samples in the fields of quartz dioritetonalite, granodiorite and monzogranite. On the Streckeisen QAP plot they occupy the same field as granitoids from the Lima segment of the Coastal Batholith of Peru (Pitcher et al. 1985). They are calc-alkaline (Fig. 5a, b) cafemic, and dominantly metaluminous with the most highly differentiated felsic derivatives ranging to slightly peraluminous (McCourt & Cobbing 1993). Finally although
A Mid to Upper Cretaceous (117-80Ma) plutonic arc extends along the length of the Barisan Mountains, intruding oceanic rocks of the Woyla microplate and broadly contemporaneous Upper Mesozoic continental foreland sequences. This arc is focused along a major NW-SE striking fault zone interpreted to approximate to the Early Cretaceous Woyla Suture, and dominated by dioritic to granodioritic hornblende-bearing granitoids with subordinate K-feldspar megacrystic biotite granites of I-type character. Scattered Lower Eocene plutons (60-50 Ma) intrude earlier plutonic arcs, mainly the Upper Cretaceous arc. The Miocene arc (20-5 Ma) consists of granodioritic
Geochemistry
326
W.J. McCOURT ET AL.
Table 1. Summary of new K-Ar plutonic ages from the Barisan Mountains, southern Sumatra INTRUSION Jurassic to Early Cretaceous Sulit Air Suite Bungo Batholith Way Sulan Gabbro
203 _+6 169 ± 5 151 __4
Middle to Late Cretaceous Garba Pluton Sulan Pluton W Sekampung Diorite Branti Granodiorite Padean Pluton
117 113 89 86 82
±3 ±3 --!-3 _+3 ±3
Early Eocene Jatibaru Microgranite Lassi Pluton Bungo Batholith
60 ± 3 57 ± 2 54 _ 2
Miocene Way Bambang Pluton Lolo Pluton
20 _ 1 ll ± 1
183 +_ 13 156 ± 6
149 _+5 154 +_7
141 ± 5 153 ± 4
138 ± 5 148 ± 4
117 __ 2 111 __ 3 89 ± 2
115 ± 4
86 ± 3
8223
82 ± 2
8l ± 2
82 _+2
54 ± 2 54 ± 2
55 __ 2
53 ± 2
131 _+7
129 ± 4
89 ± 2
53 + 2
19 ± 2
5 ± 0.2
discussed here as a single population, the granitoids are not interpreted to be co-magmatic, since they r e p r e s e n t m a g m a t i s m e x t e n d i n g o v e r almost 200 Ma, but are considered to be products of a similar and l o n g - l i v e d m e c h a n i s m o f m a g m a genesis. On the Nb-Y and Rb-(Nb + Y) plots o f Pearce et al. (1984) the volcanic arc character o f the granitoids is clear (Fig. 6a, b) with only four samples falling outside of the VAG field, all o f which are highly differentiated and anomalous with
I
I
I //
I
/
5-
4-
respect to the bulk population (McCourt & Cobbing 1993). The VAG character o f the Barisan granitoids is also evident w h e n the granitic rocks (SiO 2 > 70%) are plotted on the ORG-norrnalised plots o f Pearce et al. (1984). A c c o r d i n g to these authors, LIL e n r i c h m e n t is a c o m m o n feature o f both s u b d u c t i o n - r e l a t e d volcanic arc granites and within-plate rifted granites. Within-plate granites, however, also typically show an enrichment in Ta and Nb, and have Zr, Y and Ce normalized values greater than or close to 1; volcanic arc granites on
1
00 0g °0 eo
~
3-
~
O
~o
%"
• 0°
•
g •
•
2-
QQ 1-
/. 0
I
0
1
.
2
3
t 4
_t 5
6
Na20 ( w t % )
Fig. 4. Alkali variation diagram for the south Sumatra granitoids. Line separates the fields of I- & S-type granites of Chappell & White (1974).
MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
Na20 + K20
327
MgO
(a)
~
6
//.
f
oee
•
0 0
ooo
ee~J ~
/ jJ II
J • o11 j • f J j J J
4 Z
J
o 50
(b)
/
I 60
L 70
80
SiO 2 (wt %)
Fig. 5. (a) AFM plot for the south Sumatra granitoids. (b) Total alkalis v. silica variation diagram for the south Sumatra granitoids; dashed line denotes calc-alkaline field.
the other hand have Ta and Nb normalized values close to 1 and Zr, Y and Ce values less than 1, depending on the nature of the arc setting. The entire population the Barisan granites are geochemically similar and Fig. 7 confirms their VAG characteristics of LIL enrichment, depletion in Nb, Zr and Y, and Ce values close to 1. Collectively these patterns closely resemble those of subduction-related, Andean I-type granites from Peru and Chile (cf. Pearce et al. 1984; Pitfield et al. 1986). On primitive mantle- or MORB-normalized spider diagrams, the subduction-related character of the Sumatra rocks, excluding the granites sensu stricto, is once again evident and the strongly negative Nb N and, to a lesser extent, TiN anomalies of the entire population are emphasized (cf. McCourt & Cobbing 1993; Fig. 7b). This
Ta-Nb-Ti, and associated Y-Yb depletion is a typical geochemical signature of calc-alkaline intermediate magmas formed in a subduction setting, most probably reflecting amphibole and garnet retention at the site of partial melting (Briqueu et al. 1984; Foley & Wheller 1990), and thus characteristic of volcanic arc granitoids. Low values of Rb/M (M = HREEs, Y, Zr, etc.) and high to very high values of K/Rb of the granitoids (McCourt & Cobbing 1993) underline their continental-arc, I-type character, and Rb/Zr ratios generally less than 1 and often less than 0.70, in conjunction with low Nb contents, suggest a primitive to normal arc setting (Brown et al. 1984). In summary the geochemical data clearly confirm the volcanic arc character of the Barisan granitoids and provide evidence for the existence
328
ET AL.
W . J. M c C O U R T
1000
z
WPG E
lOO
v
:
"
Z syn- COLG
3
o
ee
10
;+ / - - ./
= .
.
• •
~
eeo
I
•
/
ORG
/
(a)
,
1
,
, ,,
,,=l
~
I ~
,
~,+1 100
~
,
, = ~l~,l 1000
2000
Log Y ( ppm ) 2000 1000
i
i
i
i t I i I
i
+
~ ,','''~
J
/Y'
'~'~l
t
__ _
WPG
syn - C O L G
_ _
Et~ Q"
v
100
•
--
qbo •
-_
•
•
oo O
o
•
•0
•
_ _
_ _
ORG
VAG
(b)
1
I
I
I
I
I Ill+
I
I
I
I Illl
I
I
100
10
I
I
-
I 1}1]
1000
2000
Log Y + Nb (ppm)
Fig. 6. Tectonic setting-classification plots for the south Sumatra granitoids, following Pearce et al. (1984). VAG, Fields of volcanic arc granite; syn-COLG, syn-collisionalgranite; WPG, within-plate granite and ORG, oceanic ridge granite as defined by Pearce et al. (1984).
of an active margin and subduction along the Sumatran continental edge. A further important feature brought out by the data is the compositional similarity of the granitoids of different ages, as can be seen from the way that the entire sample population, representing Late Triassic to Pliocene plutonism, plots very close together on the normalized spider diagrams (Fig. 7a, b). The best interpretation of this similarity is that the mechanism of magma genesis has been a consistent and long-lived one throughout the last 200 Ma. It is considered that this mechanism can only be batch or partial melting in a subduction zone and mantle wedge environment below the continental margin. A crustal contribution does not seem to have been a factor in the geochemical evolution of the plutonic
arcs of western Sumatra, although this remains unproven in the absence of isotopic data for our dataset.
Geochronology
Material was collected for both Rb/Sr and K-Ar analysis but the granitoids whole rock geochemistry precludes their use for Rb/Sr geochronology and consequently the dating programme was restricted to the K-Ar method. Approximately 40 new mineral ages were determined on samples collected from the Barisan plutonic belt. All were fresh, undeformed, nonmineralized material collected at outcrop away
MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA Norm: OGR
100
,
,
,
!
i
,
,
i
i
,
!
i
w
"ALL" SAMPLES
,& d
.01
(a)
I
I
I
I
I
1
t
I
i
I
I
I
|
Sr
KzO
Fib
Ba
Th
Ta
Nb
Oe
HI
Zr
Sm
Y
Yb
Norm: MOFIB
1000
i
i
w
,
i
i
i
,
I
!
!
1
,
,
r
,
,
!
!
!
i
~
"ALL" SAMPLE S
K
.01
(b)
x
i
i
t
i
i
i
1D
l
-.~
i
l
i
i
l
i
,
i
l
|
|
i
R b B a T h U K N b L a C e S r N d H f Z r S m E u Ti G d B y Y
i
Incorporating all other published ages from exposed plutons in western Sumatra (Fig. 8), a further plutonic event in the Permian (?287256 Ma) is indicated. Furthermore, either the Jurassic plutonism began in the Late Triassic, or alternatively there were two distinct magmatic episodes, one in the Late Triassic to Early Jurassic (220-183 Ma) and one in the Mid-Jurassic to Early Cretaceous (170-130Ma). The latter suggestion is preferred because the ages correspond to similar cycles of plutonism in peninsular Malaysia, peninsular Thailand and Burma (Cobbing et al. 1992).
Episodic plutonism in Sumatra: a plate tectonic model Episode A (?287-256 Ma)
100
"--
329
t
i
E r Y b Lu
Fig. 7. Multi-element spider diagrams for the south Sumatra Barisan granitoids, normalized to (a) oceanic ridge granite and (b) mid-ocean ridge basalt. Area in between solid lines shows compositional field of all samples (n = 54).
from any obvious faults, shears, solution planes or sizeable dykes. The limitations and pitfalls of K-Ar dating are well known to the authors, in particular that plutons or intrusive bodies dated by a single sample should not be considered to be reliably dated. Most of the quoted ages are either from duplicate samples and/or mineral pairs and are considered reliable and interpreted as intrusive ages. In addition, the plutonic episodes discussed below (A-F) have ages which are in close agreement with those from other regional studies using Rb-Sr mineral and whole rock isochron dating techniques in addition to K-Ar methods (Cobbing et al. 1992). At least four periods of plutonic activity have been identified: Jurassic to Early Cretaceous (203-130 Ma), Middle to Late Cretaceous (11780 Ma), Early Eocene (?60-50 Ma) and MiocenePliocene (20-5 Ma) as outlined in Table 1.
A discontinuous Lower-Middle Permian volcanic arc is present as an elongate fault-bounded strip along the western edge of the Mergui microplate and is interpreted to be subduction related (Katili 1973; Pulunggono & Cameron 1984). Palaeontological evidence from sediments associated with the andesitic volcanics indicates a warm climate and Cathaysian affinities, in contrast to the Gondwana characteristics of the remainder of the Mergui microplate sedimentary sequences (Fomaine & Gafoer 1989). The volcanics and associated sediments are accordingly interpreted as a separate terrane, probably an oceanic arc, that was subsequently accreted to the Mergui continental margin (cf. Wajzer et al. 1991) through northerly directed subduction and the closure of a marginal ocean basin in the Late Permian or, more likely, Early Triassic. A single K-At age of 248 _+ 10 Ma from these Lower to Middle Permian volcanics in Sumatra (Nishimura et al. 1978) may approximate to the age of collision and accretion. The Sibolga and Ombilin granites, with ages of 287-256 Ma, may represent plutonism associated with this Late Palaeozoic subduction but their present position is probably exotic.
Episode B (224-180 Ma) The collision and accretion of the Permian (Peusangan-Palepat) volcanic arc is interpreted to be part of a major terrane amalgamation event, that included the collision of the continental East Malaya and West Malaya (Sibumasu) Blocks along the Bentong-Raub Line (suture) in the earliest Triassic. The extensive granitic, S- and I-type, magmatism of peninsular Malaysia, Thailand and the Indonesian tin-islands was a direct consequence of this collision (Mitchell 1977). Granite ages range from 250-195 Ma but the majority are c. 220 Ma
W.J. McCOURT E T A L .
330
O
A
MID - LATE PERMIAN ARC
B
( LATE ) TRIASSIC - EARLY JURASSIC ARC
C
MID - JURASSIC - EARLY CRETACEOUS ARC
D
MID- LATE CRETACEOUS ARC
E
EOCENE ARC
F
NEOGENI5ARC
30--
_Z 5:
(E) 2o
c~
(F)
~ o )
c>
::iiiiiiiiiiiYiiiSiiiiiJJ
~o
~i iiiiiiiiiiJ 1o
30
I 50
70
100
130
150
::::::::::::::::::::::: 200
(A)
I 250
PLUTON AGES ( M a )
Fig. 8. Histogram of available plutonic ages from Sumatra, incorporating this study and previously published data.
(Cobbing et al. 1992) suggesting that most of the plutonism was post-orogenic. In addition, a partly coeval plutonic arc of subduction related I-type granitoids can now be identified in western Sumatra (219 _+4 to 183 _+ 13 Ma). Plutons of this subduction-related arc intrude Palaeozoic metasediments of the West Malaya (Sibumasu) terrane deformed by the collision, supporting an Early Triassic suturing age. Thus it is tentatively suggested that the Early Mesozoic plutonism may in fact be made up of two separate magmatic events: an Early Triassic event dominated by collision-related crustal S- and I-type granites in peninsular Malaysia, Thailand and the Indonesian tin-islands, and a Late Triassic to Early Jurassic (220-180 Ma) post-collisional event. The latter is represented by an I-type plutonic arc in western Sumatra (Episode B) and coeval S-type and crustal I-type granitic magmatism in the Main Range Province, Indonesian tin-islands and eastern Sumatra (Episode B1). This magmatism was probably related to tectonic release and adiabatic decompression, with resulting anatexis, in the back-arc region, with granites channelled along deep-seated faults. A postulated change in the convergence angle of the oceanic plate resulted in a more oblique subduction regime in the Early Jurassic that brought the Episode (B-B 1) plutonism to an end and resulted in transpressional strike-slip along the Sundaland continental margin which was taken up along older fault structures.
Episode C ( 1 6 9 - 1 2 9 Ma)
Middle Jurassic to Early Cretaceous plutonism in Sumatra is represented by an extensive I-type, subduction-related belt (Bungo batholith, Sulit Air suite) focused along the western edge of the Mergui microplate broadly coincident with, but laterally more extensive than, the Episode B plutonic belt. The plutonism appears to have been channelled along the junction between the Permian volcanic arc and the Palaeozoic continental margin metasediments, interpreted as an Early Triassic suture. This period of plutonism is correlated with northwest-directed subduction beneath the Sundaland continental margin, in line with postulated northwestward spreading based on identified sea floor magnetic anomalies in the eastern Indian Ocean (Patriat & Achache 1984). This plutonic arc may also extend north into the Shan scarp region of Burma on the basis of limited geochronological evidence (Cobbing et al. 1992). Episode C plutonism terminated in the latest Early Cretaceous (or early Middle Cretaceous) following the collision, accretion and local obduction of the allochthonous Woyla terranes of southern Sumatra, some 125 Ma ago (McCourt et al. 1993). These ophiolitic rocks can be correlated with similar rocks in western Burma (Mawgyi andesites), Tibet (Donqiao ophiolite) and possibly SE Kalimantan (Alino Formation and Meratus ophiolite), and it is likely that they represent fragments of an oceanic arc system that
MESOZOIC---CENOZOIC PLUTONISM IN SUMATRA collided with, and was thrust over, the continental margin of Sundaland at this time (cf. Mitchell 1993).
Episode D (120-75 Ma) Subsequent to the accretion event, possible northdirected subduction was reestablished and I-type granitoid magmas were emplaced into the now cratonized Woyla microplate, with the majority of the plutons focused along the main suture line and related faults (Sikuleh granite, Manunggal batholith, Ulai intrusion, Garba pluton, Sulan pluton). Based on Mitchell (1993), it is probable that this subduction-related arc extends north into Burma west of the Sagaing Fault, where the oldest dated plutons are of mid-Cretaceous age. This western Burma Arc (?equivalent to the Central Valley Province of Cobbing et al. (1992)) is made up of I-type granodioritic to tonalitic plutons, with K-Ar ages of 106+7, 103+4, 9 8 + 4 , 9 4 + 4 and 91 __.8 Ma, and like its Sumatran equivalent intrudes a sequence of deformed oceanic rocks, basaltic andesites and basalt pillow lavas, the Mawgyi andesites (Mitchell 1993). Broadly contemporaneous plutonism is also recorded from the Western Province of Thailand and Burma (Cobbing et al. 1992) and corresponds to a mixed population of I- and S-type granites with high initial ratios indicative of a significant crustal component in most cases. Clarke & Beddoe-Stephens (1987) proposed that this belt of Upper Cretaceous S- & I-type crustal granites also extends into eastern Sumatra, as indicated by the 80 Ma Hatapang granite. It is suggested that this plutonism was related to anatexis, the result of crustal thickening accompanying thrusting that was contemporaneous with subduction and VAG, I-type magmatism in the Central Valley Province and western Sumatra. Middle to Late Cretaceous magmatism continued northwards through the Mogok Belt into Assam (Mitchell 1993) and reported ages of 113-82 Ma on the Gandise batholith in Tibet (Debon et al. 1986) may indicate a further extension of this plutonism. The general absence of plutonic rocks with ages in the range 75-60 Ma, coincides with the well documented latest Cretaceous deformation throughout this region, including Sumatra (de Coster 1974; Hamilton 1979; Cameron et al. 1980; Pulunggono & Cameron 1984). Exactly why plutonism ceased is not clear. The model proposed here involves a suggested change from high angle to oblique subduction along the continental margin, related to a change in oceanic spreading patterns and plate configurations, that resulted in the accretion of a continental sliver, the West Sumatra terrane. This terrane is now present as a series of fragments such as Sikuleh, Natal and possibly
331
Bengkulu (Fig. 2) that correspond to the southerly extension of the West Burma terrane of Metcalfe (1994). Much of the evidence for this event, however, has since been destroyed during tectonism and disruption related to the Early Tertiary collision of India and Eurasia.
Episode E (60-?50 Ma) Following the Late Cretaceous deformation event, a new subduction regime was established along the continental margin of Sumatra as evidenced by a short-lived but extensive plutonic episode from 60-50 Ma (Episode E). This I-type, VAG plutonic arc (Lassi pluton, Nagan granodiorite etc.) was superimposed on the earlier Cretaceous and Jurassic arcs via deep-seated older fault structures in the continental margin which acted as magma conduits. Limited regional evidence suggests that this mainly Early Eocene (57-52 Ma) plutonism extends into Burma and Thailand where it is of combined I- & S-type character (Cobbing et al. 1992; Mitchell 1993). It is suggested that this plutonic episode was brought to an end by the Middle Eocene collision of India and Eurasia at about 50 Ma, approximating to the timing of proposed ophiolite emplacement in the IndoBurman Ranges of western Burma (Mitchell 1993). A further important consequence of the collision of India and Eurasia was the indentation and related deformation of the Lower Tertiary margin of Asia and the probable extrusion and clockwise rotation of much of SE Asia, including Sumatra. The shape of this margin prior to collision was, as suggested by Tapponnier et al. (1986), a simple slightly convex line extending from Sumatra to the western Makran. As a preliminary model it is proposed that this margin was characterized by a series of subparallel, outwardly younging plutonic belts representing prolonged convergence and subductionrelated plutonism, along the margin, from the Early Mesozoic to the Early Tertiary.
Episode F (30-0 Ma) Subsequent to the India-Eurasia collision, and a related major reorganization of plate motions and spreading patterns in the Indian Ocean, NNE directed subduction was established along the Sundaland margin. Available plutonic ages from Burma, 38 + 1 Ma (Mitchell 1993) suggest that subduction-related activity was taking place along the margin by the Early Oligocene, although plutonism in Sumatra was apparently not established until the Early Miocene (Episode F). Wajzer et al. (1991) reported Late Oligocene ages (3028 Ma) from the Air Bangis granite of central Sumatra, but concluded that these plutons, and
332
W.J. McCOURT
ET AL.
/'), ./ /
~
J
! / (
•
\ v
x v
'/+~L .
+
.3 )
I I
GRANITE PROVINCES EASTERN, I - type of Permian - Triassic age r~
MAIN RANGE, S - type of dominantly Triassic age WESTERN, I & S-type of Jurassic - Cretaceous & Tertiary age
~]
q 0
ENTRAL VALLEY, I - type Mid - Late Cretaceous age
~
Q
0
L
400km
I
Fig. 9. The main granite provinces of SE Asia, modified from Cobbing et al. (1992) for Sumatra. Inset shows the principal tectonostratigraphic terranes of SE Asia based on Metcalfe (1988, 1990).
MESOZOIC---CENOZOIC PLUTONISM IN SUMATRA contemporaneous volcanics of the Langsat volcanic arc, had formed elsewhere along the Sundaland margin and were tectonically juxtaposed against the Woyla Group of Sumatra sometime prior to the Middle Miocene. Rock et al. (1983) proposed that the Langsat Volcanics were of Palaeogene age, a conclusion confirmed by Wajzer et al. (1991), who assigned them a Late Eocene to Early Oligocene age on the basis of whole rock K-Ar dates (4038 _+ 1 Ma). Thus the proposed age of the Air Bangis plutonism is almost identical to that noted above from Burma. The younger dates (30-28 Ma) from the Air Bangis granites could therefore reflect their collision with and accretion to the Sumatran margin and this could relate to the proposed midOligocene collisional event responsible for the recorded inversion in the forearc basins of Sumatra and Java as proposed by Daly et al. (1991). Following this mid-Oligocene event, widespread andesitic volcanism was established in Sumatra and the main Neogene magmatism of the Barisan arc was initiated. Subduction-related VAG plutonism was widespread by the end of the Early Miocene and in the Middle Miocene the entire Barisan arc became volcanic. Middle Miocene to Early Pliocene, I-type granitoid plutons are essentially
(EUR)ASIA
._4_
,~• ~
, .
~
-----<._× - - - . . - , < ,~
,~ •
333
confined to the western Barisan Ranges in Sumatra, focused along the Semangko Fault strand of the SFS, approximately coinciding with the exposed western limits of the Upper Mesozoic Woyla terranes and the main axis of the Upper Oligocene volcanic arc. The Neogene plutonic episode is also recorded from the Burma Arc and possibly the Mogok Belt (Mitchell 1993).
Conclusions It is proposed that the Mesozoic-Cenozoic plutonic evolution of Sumatra was episodic and that five plutonic episodes can be recognized: Late Triassic-Early Jurassic, Middle Jurassic-Early Cretaceous, Middle Cretaceous-Late Cretaceous, Early Eocene and Miocene-Pliocene. Each episode was dominated by I-type, subduction-related plutonism. Granitoids from different episodes are very similar petrographically and geochemically and are of VAG character. This is interpreted to mean that Sumatra has been the site of an active convergent margin and of long-lived subduction, albeit episodic, since the Early Mesozoic. The majority of the plutonic episodes identified from
X
N
[~
__
~
;"..... :"- .>----------~.. |
~, --.<.:-.>..'a.:~.: . : .-. , ",": ~ " . : '.:.' '....
MID- LATE "J CRETACEOUS ARC
~
...........
I
, ,-~!...:.;.,.....,
,/f
\k
"
Fig. 10. Cartoon, not to scale, illustrating a possible plate-tectonic setting along part of the Asian margin prior to the Eocene collision of India (modified from Metcalfe (1990). SE Asian terranes as follows: SC, South China; C, Qiangtang; I, Indochina/East Malaya; L, Lhasa; S, Sibumasu; WB, West Burma; WY, Woyla.
334
W.J. McCOURT ET AL,
Sumatra can be recognized, albeit in disrupted form, throughout m u c h of SE Asia (Fig. 9). Thus the Triassic to Early Jurassic plutonism correlates with the Eastern and Main Range Granite Provinces of Thailand and Malaysia, whereas the Middle Jurassic and Cretaceous plutonism can be correlated with a combination of the Western and Central Valley Provinces of Thailand and Burma. It is suggested as a preliminary model that prior to the Eocene collision of India the Sundaland margin was orientated approximately W N W and made up of a series of outwardly younging subductionrelated plutonic arcs (Fig. 10), some of which probably extended along the southern margin of the Asian plate. The present distribution and geometry of these arcs in SE Asia is the result of the effects of the collision of India and Eurasia, i.e. indentation, extrusion and strike-slip faulting, as predicted by the model of Tapponnier et al. (1982, 1986). The more recent major dextral strike-slip m o v e m e n t s along the SFS and related master faults outside of Sumatra, the onset of which coincided with the opening of the A n d a m a n Sea c. 11 Ma ago
(Curray et al. 1979), has further complicated this scenario. We also propose that breaks in plutonic activity correspond to periods of oblique approach, that, in some instances, relate to the collision and accretion of allochthonous terranes. We further suggest that one of the underlying factors that controls the development and siting of the various plutonic arcs is the availability of deep-seated faults along the continental margin, that probably extend down to the site of m a g m a generation at or close to the subduction zone. This paper is published with the permission of the Directors of the Geological Research and Development Centre, Bandung, and the British Geological Survey, Nottingham. The work in Sumatra was carried out as part of a bilateral technical cooperation project between the governments of Indonesia and the United Kingdom and funded jointly by the Indonesian Directorate General of Geology and Mineral Resources (DGGMR) and the Overseas Development Administration (ODA) of the British Foreign Office. We thank A. H. G. Mitchell and S. J. Moss for suggestions which improved the text.
References ASPDEN, J. A., KARTAWA,W., ALDISS, D. T., DJUNUDDIN, A., WHANDOYO, R., er AL. 1982. The geology of the Padangsidempaun and Sibolga Quadrangle, Sumatra (1:250,000). Geological Research and Development Centre, Bandung. AUDLEY-CHARLES,M. G., BALLANTYNE,P. D. & HALL, R. 1988. Mesozoic-Cenozoic rift-drift sequence of Asian fragments from Gondwanaland. Tectonophysics, 155, 317-330. BR1QUEU, L., BOUGAULT, n. & JORON, J. L. 1984. Quantification of Nb, Ta, Ti and V anomalies in magmas associated with subduction zones: petrogenetic implications. Earth and Planetary Science Letters, 68, 297-308. BROWN, G. C., THORPE, R. S. & WEBB, P. C. 1984. The geochemical characteristics of granitoids in contrasting arcs and comments on magma sources. Journal of the Geological Society, London, 141, 413-426. CAMERON, N. R., CLARKE, M. C. G., ALDISS, D. T., ASPDEN, J. A. & DJUNUDDIN, A. 1980. The geological evolution of northern Sumatra. In: Proceedings Indonesian Petroleum Association 9th Annual Convention. 149-187. CHAPPELL, B. W. & WHITE, A. J. R. 1974. Two contrasting granite types. Pacific Geology, 8, 173-174. CLARKE, M. C. G. & BEDDOE-STEPHENS, B. 1987. Geochemistry, mineralogy and plate tectonic setting of a Late Cretaceous Sn-W Granite from Sumatra, Indonesia. Mineralogical Magazine, 51, 371-387. COBBING, E. J., PITEIELD,P. E. J., DARBYSHIRE,D. E E & MALLICK, D. I. J. 1992. The granites of the Southeast Asian tin belt. British Geological Survey, Overseas Memoir, 10. CURRA¥, J. R., MOORE, D. G., Lawver, L. A., EMMEL, E J., RArrr, R. W., er AL. 1979. Tectonics of
the Andaman Sea and Burma. In: WATKINS, J., MONTADERT, L. & DICKENSON, P. W. (eds) Geological and Geophysical Investigations of Continental Margins. AAPG Memoir, 29, 189-198. DALY, M. C., COOPER,M. A., WILSON, I., SMITH,D. G. t~ HOOPER, B. G. D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine and Petroleum Geology, 8, 1-21. DEBON, E., LE FORT, P., SHEPPARD,S. M. E & SONET, J. 1986. The four plutonic belts of the TranshimalayaHimalaya: a chemical, mineralogical, isotopic and chronological synthesis along a Tibet-Nepal section. Journal of Petrology, 27, 219-250. DE COSTER, G. L. 1974. The geology of the Central and Southern Sumatra Basins. In: Proceedings Indonesian Petroleum Association 3rd Annual Convention. 77-110. EUBANK, R. T. & MAKK1,A. Ch. 1981. Structural geology of the Central Sumatra Back-Arc Basin. In: Proceedings Indonesian Petroleum Association, lOth Annual Convention. 153-196. FOLEY, S. E • WHELLER, G. E. 1990. Parallels in the origin of the geochemical signatures of island arc volcanics and continentaI potassic igneous rocks: the role of residual titanates. Chemical Geology, 85, 1 18. FONTAINE,H. & GAFOER,S. 1989. The pre- Tertiary fossils of Sumatra and their environments. CCOP Technical Secretariat Publication, Bangkok, Thailand. GAFOER, S. 1990. Tinjauan Kembali Tataantratigrafi Pratersier Sumatra Bagian Selatan. [A review of the pre-Tertiary sequences of Southern Sumatra. Abstract in English.] In: Prosiding Persidangan Sains Bumi dan Masyarakat Universiti Kebangsaan Malaysia, 9-10 July 1990.
MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA --
& PURBO-HADIWIDJOYO,M. M. 1986. The geology of southern Sumatra and its beating on the occurrence of mineral deposits. Bulletin of the Geological Research and Development Centre Bandung, 12, 15-30. HAMILTON,W. 1979. Tectonics of the Indonesian Region. USGS Professional Paper, 1078. HUTCHISON,C. S. 1989. Geological evolution of Southeast Asia. Clarendon Press, Oxford. 1994. Gondwana and Cathaysian blocks, Palaeotethys sutures and Cenozoic tectonics in Southeast Asia. Geologische Rundschau, 82, 388-405. KATILI, J. A. 1973. Geochronology of west Indonesia and its implications on plate tectonics. Tectonophysics, 19, 195-212. KONING, T. & DARMONO, F. X. 1984. The geology of the Beruk Northeast field, central Sumatra: oil production from pre-Tertiary basement rocks. In: Proceedings Indonesian Petroleum Association 13th Annual Convention, 385-406. McCOURT, W. J. & COBBING, E. J. 1993. The geochemistry, geochronology and tectonic setting of granitoid rocks from southern Sumatra, western Indonesia. Southern Sumatra Geological and Mineral Exploration Project, Report Series, 9, Directorate of Mineral Resources/Geological Research and Development Centre, Bandung, Indonesia. , GAFOER, S., AMIN, Z. C., ANDI MANGGA, S., KUSNAMA,ET AL. 1993. The geological evolution of Southern Sumatra. Southern Sumatra Geological and Mineral Exploration Project, Report Series, 13, Directorate of Mineral Resources/Geological Research and Development Centre, Bandung, Indonesia. METCALFE, I. 1988. Origin and assembly of Southeast Asian continental terranes. In: AUDLEY-CHARLES, M. G. & HALLAM, A. (eds) Gondwana and Tethys. Geological Society, London, Special Publication, 37, 101-118. 1990. Allochthonous terrane processes in Southeast Asia. Philosophical Transactions of the Royal Society, 331A, 625-640. 1994. Gondwanaland origin, dispersion and accretion of East and Southeast Asian continental terranes. Journal of South American Earth Sciences, 7, 333-347. MrrCHELL, A. H. G. 1977. Tectonic settings for emplacement of Southeast Asian tin granites. Bulletin of the Geological Society of Malaysia, 9, 123-140. 1981. Phanerozoic plate boundaries in mainland SE Asia, the Himalayas and Tibet. Journal of the Geological Society, London, 138, 109-122. 1993. Cretaceous-Cenozoic tectonic events in the western Myanmar (Burma)-Assam region. Journal of the Geological Society, London, 150, 1089-1102. NISHIMURA,S., SASAJIMA,S., HIROOKA,K., THIO, K. H. & HEHUWAT, E 1978. Radiometric ages of volcanic products in the Sunda Arc. In: SASAJIMA, S. (ed.) Studies of the physical geology of the Sunda Arc. University Press, Kyoto, 34-37. -
-
-
-
335
PATRIAT, P. & ACHACHE, J. 1984. India-Eurasia collision chronology has implications for crustal shortening and driving mechanisms of plates. Nature, 311, 615~521. PEARCE, J. A., HAm~tS, N. B. W. & TINDLE, A. G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25, 956-983. PITCHER, W. S., ATHERTON, M. R, COBBING, E. J. & BECKINSALE, R. D. 1985. Magmatism at a plate edge-the Peruvian Andes. Blackie, Glasgow & Halstead Press, New York. PITFIELD, P. E. J., COBBING, E. J., CLARKE, M. C. G., MALLICK, D. I. J. & TEOH, L. H. 1986. Granitic provinces in the Southeast Asia Tin Belt. In: Transactions of the Fourth Circum-Pacific Energy and Mineral Resources Conference Singapore. 575-589. PULUNGGONO, A. & CAMERON, N. R. 1984. Sumatran microplates, their characteristics and their role in the evolution of the Central and South Sumatra Basins. In: Proceedings Indonesian Petroleum Association 13th Annual Convention. 121-143. ROCK, N. M. S., ALDISS, D. T., ASPDEN, J. A., CLARKE, M. C. G., DJUNUDDIN,A., ETAL. 1983. The geology of the Lubusikaping Quadrangle Sumatra. Geological Research and Development Centre, Bandung. SILITONGA, P. H. & KASTOWO. 1975. Geological map of the Solok Quadrangle, Sumatra. Geological Survey of Indonesia, Bandung. SIMANDJUNTAK,T. O., SURONO,GAFOER, S. & AMIN, T. C. 1991. The geology of the Muarabungo Quadrangle, Sumatra. Geological Research and Development Centre, Bandung. STAUFFER, P. H. 1983. Unravelling the mosaic of Palaeozoic crustal blocks in Southeast Asia. Geologische Rundschau, 72, 1061-1080. --, LE DAIN, A. Y., ARMUO, R. & COBBOLD,P. 1982. Propagating extrusion tectonics in Asia, new insights from simple experiments with plasticene. Geology, 10, 611-616. TAPPONNIER, P., PELTZER, G. & ARMIJO, R. 1986. On the mechanics of the collision between India and Asia. In: COWARD, M. P. & RIES, A. C. (eds) Collision Tectonics. Geological Society, London, Special Publication, 19, 115-157. VAN BEMMELEN, R. W. 1949. The Geology of Indonesia. Martinus Nijhoff, The Hague. vAN Es, L. J. C. 1919. De tectoniek van het westelijk helft van den Nederlandsch Oost Indischen Archipel. Jaarboek Mijnwezen Nederlandsch Oost Indies, Verhandelingen, 5-143. WAJZER, M. R., BARBER,A. J., HIDAYAT,S. & SUHARSONO. 1991. Accretion, collision and strike-slip faulting: the Woyla Group as a key to the tectonic evolution of north Sumatra. Journal of Southeast Asian Earth Sciences, 6, 447--461. WESTERVELD,J. 1952. Phases of mountain building and mineral provinces in the East Indies. Report of the 18th International Geological Congress, London, 13, 245-255.
The Mentawai fault zone and deformation of the Sumatran Forearc in the Nias area M . A. S A M U E L
& N. A. H A R B U R Y
SE Asia Research Group, Research School of Geological and Geophysical Sciences, Birkbeck College and University College London, Gower Street, London W C I E 6BT, UK. Abstract: The Sumatran Forearc is the site of oblique plate convergence between the IndoAustralian and SE Asian-Eurasian Plates. It is generally accepted that the forearc sliver is not behaving as a rigid plate and that the rate of slip increases along the fight-lateral Sumatran Fault System from southeast to northwest. Two contrasting hypotheses have been invoked to explain this pattern of decoupling: by using slip vectors to suggest arc-parallel stretching; and a second major right-lateral strike slip zone, parallel to the Sumatran Fault System, to accommodate the oblique subduction. This zone, termed the Mentawai fault zone, lies just to the east of the outer-arc ridge and is indicated as intersecting with other faults on Nias Island. In this paper data are presented from Nias where a thick forearc succession is excellently exposed in three subbasins with half-graben geometries and broadly comparable Palaeogene-Recent histories. Mud diapirism has generated melanges which cut the Oligocene to Recent strata. Outcrop data, LANDSAT, Synthetic Aperture Radar and aerial photographs were used to determine fault offset patterns, and marked extension of the island along north-striking right lateral faults and ESE striking left-lateral faults is revealed. Extension was further accommodated along a set of ENE striking normal faults. In the region where the proposed Mentawai fault zone comes onto Nias there is an important, long-lived, basin-bounding fault. This structure had in excess of 5 km normal throw during the Oligocene-Miocene whilst minor contractional reactivation of the fault occurred during the Pliocene phase of uplift and deformation which affected all of Nias. Data from onshore and offshore the Nias region suggest that all the features visible from seismic sections collected over the Mentawai fault zone to the south of Nias can be explained in terms of inversion of originally extensional structures and mud diapirism. It is suggested that strike-slip motion is of limited importance along the 600 km long Mentawai fault zone. Rather this highly structured zone represents a deformation front whose origin can be explained by Pliocene to Recent subduction-driven inversion at the outer margin of the forearc.
The Sumat~an Forearc has been recognized as a type example of an obliquely convergent margin for over twenty years (e.g. Fitch 1972). In recent years two contrasting models of deformation of the Sumatran Forearc have been proposed (McCaffrey 1991; Diament et al. 1992). This paper outlines these m o d e l s and presents new data on the Sumatran Forearc, from the area of Nias, which allow a critical assessment of these two models. Furthermore the results lead to an improved understanding of the d e f o r m a t i o n processes acting within, and at the margin of, the forearc.
Deformation of the Sumatran Forearc The Sumatran Forearc forms part of the Sunda subduction system which extends from Sumba in the east to B u r m a in the north (Figs 1 and 2; Moore et al. 1980b; Curray 1989). Plate tectonic models predict convergence rates varying from 7.8 c m a -1 near S u m b a w a to 6 c m a -1 near the A n d a m a n Islands (Minster & Jordan 1978; DeMets et al. 1990) with the direction of convergence close
to n o r t h - s o u t h ( N e w c o m b & M c C a n n 1987; McCaffrey 1991). The Sumatran arc has a classic morphology of trench, accretionary prism, outer-arc ridge, forearc and volcanic chain with active andesitic volcanism (Karig et al. 1979) and seismicity displays a distinct Benioff zone (Page et al. 1979). Recent work has shown that the outer-arc ridge along Sumatra comprises the outer part of the forearc rather than forming part of the accretionary prism which is located further to the southwest (Samuel 1994; Samuel et al. 1995). The outer-arc ridge is wholly submerged off Java whereas the ridge gains sub-aerial expression in the Sumatran Forearc. A further important difference between the Javan and Sumatran forearcs is the subduction direction which varies from nearly orthogonal off Java to oblique west of Sumatra where the trench strikes N140°E. Fitch (1972) proposed that oblique convergence in the Sumatran area could be a c c o m m o d a t e d by the right-lateral Sumatran fault zone. This fault zone connects to the A n d a m a n Sea in the northwest
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 337-351.
337
338
M.A. SAMUEL • N. A. HARBURY
where extension and sea floor spreading is an extreme effect of oblique convergence (Fig. 2; Curray et al. 1979). To the southeast the zone extends to the Sunda Strait where further extension is occurring (Huchon & Le Pichon 1984; Harjono et al. 1991) and where it probably crosses the forearc (Curray 1989). A sliver plate (the Burma Sliver Plate; Curray et al. 1979) is thus decoupled from the Eurasian and Indo-Australian plates and moves northwest with respect to the Eurasian Plate. This plate has been called the Sumatran Sliver Plate or forearc sliver in the Sumatran area (Jarrard 1986; Diament et al. 1992). McCaffrey (1991) presented slip vector data that suggested the oblique convergence in the Sumatran area was not completely decoupled. The results indicated increasing decoupling northwestwards (Fig. l a). Modelling of the pattern of slip vectors by McCaffrey (1991) suggests that the forearc is not behaving as a rigid plate but is subject to arc-parallel stretching at a uniform strain rate of 3--4 x 10-8 per year. He showed that the northwestward motion of the forearc relative to the SE Asian plate should increase from near zero at the Sunda Strait to 4 5 - 6 0 m m a -1 in northwest Sumatra. McCaffrey (1991) acknowledged that the deformation processes acting to stretch the forearc were poorly understood. Shallow strike-slip earthquakes in the forearc have nodal planes that strike north and east across the forearc rather than parallel to it (McCaffrey 1991) and it was suggested the deformation of the forearc was most likely to occur on strike-slip faults crossing the margin than on arc-parallel strike-slip faults or arc-perpendicular normal faults. There are a number of geological indications that support the concept that deformation must be occurring in the Sumatran Forearc. For example there does not appear to have been sufficient movement along the Sumatran Fault Zone to account for the 460 km of opening of the Andaman Sea in the last 13 Ma (Curray et al. 1979). Indeed Bellier et al. (1991) suggest that the rate of slip along the southern end of the Sumatran Fault zone is only 6 mm a-1 which is considerably less than the 40 mm a-1 rate of opening of the Andaman Sea (Curray et al. 1979). Diament et al. (1992) proposed an alternative solution to explain the decrease in rate of slip to the southeast along the Sumatran Fault system. A single channel seismic reflection survey led to the identification of a complex set of structures which they correlated for 600 km along the outer margin of the Sumatran forearc and named the Mentawai fault zone (Fig. l b, c). These features were apparent on all seismic sections from southern Sumatra to Siberut although the detailed structure of the zone was highly variable; from north to
south, faulted anticlines, faulted blocks, horst and graben systems and flexures were all described. Diament et al. (1992) interpreted the structures as comprising a right-lateral strike slip zone (Fig. lc). They based their arguments on three points: (1) it is possible to interpret some features, identified on seismic lines, as positive flower structures; (2) the fault zone is straight and appears to run continuously for hundreds of kilometres; (3) a pronounced difference is commonly found in the depth to acoustic basement on either side of the fault zone. Diament et al. (1992) further suggested that displacement on the Mentawai fault zone may be relayed to the Batee fault which they tentatively map as linking with the Mentawai fault zone on Nias (Fig. lc). Nias is therefore of prime interest as it potentially lies at the intersection of two major fault zones. Furthermore it is the only island along the outer-arc ridge which is cut by the Mentawai fault zone (Fig. lc). Deformation in the Nias area Previous work by Moore & Karig (1980) suggested that Nias comprised part of an uplifted accretionary complex. Two main stratigraphic units were defined. The lower unit, the 'Oyo Complex' was interpreted as a tectonic melange that formed prior to the Early Miocene, when deep marine sediments were unconformably deposited above it in trench slope basins (Moore et al. 1980a). The Miocene to Pliocene sedimentary succession was reported as a shallowing-up sequence. Uplift was continuous, with compressional deformation occurring largely along arcward-dipping thrust and reverse faults. More recently Pubellier et al. (1992) have suggested that Nias comprises part of an Eocene Tethyan suture zone. Their interpretation is based largely on the apparent recognition of in situ Eocene rocks forming part of the non-ophiolitic sedimentary succession on the island. The research of Samuel (1994) and many previous workers (e.g. Douville 1912; Moore et al. 1980a) shows however that the rocks in question are of Upper Oligocene and Lower Miocene; they contain both a reworked Eocene fauna and younger microfossils. As part of an integrated study of the Sumatran Forearc by the University of London SE Asia Research Group the structural and stratigraphic evolution of the whole of Nias has recently been studied in detail (Samuel 1994; Samuel et al. 1995). The studies have led to a substantial reinterpretation of the sedimentological and structural setting of the island and the work has shown that Nias consists of three main sub-basins and a basement high (Fig. 2). These detailed sedimentological, biostratigraphic and palaeobathymetric studies of the successions on Nias have led to the construction
SUMATRAN FOREARC & MENTAWAI FAULT ZONE
339
Bengal Fan ................. ~ Stretching vectors (Forearc relative to SEA Plate) Slip vectors ---4> Plate vectors --
IO°N
--e
S.E. Asian Plate 3- j ~-~ --
Borneo
L
0o
Sunda Strait
Indian
-
-
East Java Sea Sumbawa
Ocean
10os
~
90OE
IO0OE
t
110OE
J
~ I \ Sin;kel Fault~ t 0-"~_~ [Singkel"%~.~ "~ -
~anyakl% ,[ basra
~
% / 5" I~__~
\
~,,~'e,>
I
~
I--Eql
I t - ) II ' '1
"-" . X " _
\
~a
o oeoo, o
"~"~ ~
Sumatra I \
,;^1
, ~.
p
Ao .,ha,,
I-'~'~'11
\ \
I ,'-, II ' '1
X \
\
\
I
}~,
Quaternaryandl activev°lcan°es-AI
I
•
India
\ ~ \
cean
t
l
I
Fig. 1. (a) Location of the Sunda Arc and major structural features in the region. Tectonic features, slip vectors and plate motion vectors after McCaffrey (1991). Only the azimuth of individual slip vectors is shown. These slip vectors show a clockwise rotation from the Sunda Strait northwestwards along Sumatra. Modelling of the slip vectors indicates that the forearc is subject to arc-parallel extension as shown by the stretching vectors. (b) Structural features in the Nias area. The location of the 'Mentawai fault zone' suggested by Diament et al. (1992) is shown and other offshore faults have been mapped from seismic profiles by Karig et al. (1980) and Matson & Moore (1992). The faults onshore Sumatra have been mapped during field studies (e.g. Page et al. 1979). (c) Model of Diament et aL (1992). Right-lateral motion is taken up both along the Sumatran Fault System and the parallel-orientated Mentawai fault zone (Mfz). Right-lateral movement is transferred via the Batee Fault (BF) which transects the forearc.
340
M . A . SAMUEL 8Z N. A. HARBURY
L [ •
t
Recent alluvium
1 °30'N
~ 1 II [i~
Elevated reef compl. . . . Upper Pliocene-Pleistocene Middle Miocene-Lower Pliocene
Y
Lower-Middle Miocene shallow marine sediments Oligocene-Lower Miocene deep marine sediments Sediment & Melange Complex Basement (B)
-~unungsitoli
Gido Line of X-section (Fig. 4 )
\ \°
\
\\
",.\
\,
_ 1ON
Major anticline
N Major synciine J
Faults Active mud volcanoes
"B
N A 30'E
0
~
20 km
Fig. 2. Simplified geological map of Nias showing major structures identified on the island. The map was compiled from field traverse data, aerial photographs, SAR (synthetic aperture radar) and LANDSAT images. The line of the 'flexure' in southeastern Nias is highlighted and the line of cross-section, shown in Fig. 4, is indicated. Key localities referred to in the text are also plotted. A large area of the Mujoi Sub-basin has been mapped as 'sediment and melange complex'. This complex is composed of deformed sedimentary sections, predominantly of Oligocene and Lower Miocene age, intruded by melanges. The individual melange intrusions cannot be shown accur~/tely on maps of this scale. Two active mud volcanoes on Nias are also located. The inset shows the position of the three sub-basins and the basement high on Nias.
341
SUMATRAN FOREARC 8,~ MENTAWAI FAULT ZONE
One of the fundamental differences in the new stratigraphy for the forearc compared, for instance, with the scheme of Moore e t al. (1980a), is the recognition that the Oligocene and Lower Miocene sediments on Nias form part of a single, stratigraphically continuous, sedimentary succession (Fig. 3). Workers such as Moore & Karig (1980) believed that the Oligocene deep marine sedimentary rocks only occurred in the melanges whereas this study, particularly in the central part of
of a new stratigraphy for Nias (Fig. 3). When these studies are combined with structural data collected over the entire island a new picture of the geological evolution of Nias emerges (Fig. 3). There is strong evidence that the Oligocene to Early Miocene history of the Sumatran Forearc was extensional/transtensional and below we outline some of the key points of this history before concentrating on specific aspects of the deformation of Nias.
AGE
N c)
NN ~,,D
T
Lahewa Sub-basin
I
Mujoi
Gomo
Sub-basin
I Sub-basin
I
Mola Basement High
I I
Structural events (7) Rapid uplift continues with
Quaternary
(6) Pliocene uplift
(5) Regional subsidence
(4) Early M i o c e n e uplift
(3) Extension of basement & development of basins under transtension
~
P18
I~ o
NP22
~ .~ ~ :. ~. . . .~.. . +,.~ ~. /.". ~. . . . .~ ..~ ~......
~
," -% .~
P16 Imr'~Lu
I
P"
i, ii~~?,'!i~'~/~i~ii,~,,~!~:~,, ;L,E! !i ~;;!; i;:!,/~, ¸i,,~' ~ i,ii! ;)i;~ i/i!~i i~;');~i:i~i!i~i~,,.i~!~!i ........
Mid
I ~ leo ~ ~
..........
. . . . . . . . .
s.:~..,,, j..:-...~.... <..~' . ./ . . "~ = ~ ................. .................
: ,
~!
! . : : i:'~
"" -- -- "
O,UnKnownor,9,n
(2) Emplacement & uplift of b a s e m e n t
ii
Formation ,1,hydrothermal alteration of ocean,corust and mantle
~[ = Ophio[itic basement, [I = Ohgocene-Lower Miocene deep marine sedimentary rocks, I I ] = Tufts, I V = Oligocene-Lower Miocene conglomerate marker, V = L o w e r - M i d d l e Miocene shallow marine limestone-dominated sediments, V[ = M i d d l e Miocene-lowest Pliocene shelfaI to upper bathya] sedimentary rocks, V I I = diatomites, V I I I = M i d d l e Miocene-lowest Pliocene tuff marker, IX = M i d d l e Miocene-lowest Pliocene limestone marker, X = U p p e r Pliocene-Pleistocene shallow marine siliciclastics, XI = U p p e r P]iocene-P]eistocene elevated reef complexes.
1
Fig. 3. Chronostratigraphy for Nias. The scheme is based on detailed sedimentological, palaeobathymetric and biostratigraphic data collected during four field seasons. Over 350 samples from across the island were dated. N, P, planktic foraminiferal zones (Bolli et al. 1985); NN, NP, nannofossil zones (Bolli et aL 1985); and T, far eastern benthic foraminiferal letter stages (Adams 1984). The large numbers refer to the structural events detailed in the right hand column. The Roman numerals refer to the relevant formations and marker horizons. Sub-basin bounding faults are marked by double vertical lines and the single vertical lines represent faults which transect the sub-basins. Both the SE-striking sub-basin bounding faults and the transecting north- and ESE-striking faults were important in controlling sub-basinal sedimentation, particularly during the Early Miocene.
342
M.A. SAMUEL (~ N. A. HARBURY
Nias, has identified sections where Oligocene rocks pass conformably up into the Lower Miocene, as suggested earlier by Elber (1939) and Paul (1941). Contrary to previous interpretations, detailed studies of the melanges on Nias have revealed that they intrude and include strata of Oligocene to Pleistocene age (Samuel et al. 1995). The melanges formed by mud diapirism, which continues to the present day as shown by active mud vulcanism in the north of the island (Fig. 2). The melanges do not therefore form basement to the sedimentary successions on Nias. The basement, in at least western and central parts of Nias, is ophiolitic in nature. This is deduced from the presence of apparently intact sections of ophiolitic material along the western coast of Nias and the occurrence of ophiolitic material in many of the diapiric melanges across the island (Fig. 2; Samuel et al. 1995). The youngest age for the ophiolite complex is MidEocene, as dated by radiolarian cherts. The ophiolite complex was therefore emplaced between the Mid-Eocene and Oligocene when deep marine sedimentary rocks were unconformably deposited above it (Fig. 3). The earliest structures recorded in these sediments are extensional and sedimentological, and stratigraphic criteria reveal that the Oligocene and Lower Miocene sediments were deposited in a set of half-grabens on a slope leading down to the Oligocene to Early Miocene trench to the southwest (Samuel et al. 1995). As extension/ transtension proceeded during the Early Miocene, topographic variations were produced by transtensional faults that cut across the strike of the sub-basins. This is recorded in the stratigraphy of Nias, since shallow marine Lower Miocene carbonates are found along-strike of coeval deep marine facies, particularly in the Gomo Sub-basin (Fig. 3). Facies in the carbonates such as P o r i t e s bindstones and floatstones, algal bindstones and oncolitic packstones prove that these carbonates must have been deposited in inner to mid-shelfal environments as suggested by a number of previous workers (e.g. Elber 1939; Paul 1941; Burrough & Power 1968; Bradley 1973; Harbury & Kallagher 1991). Localized uplift of some western parts of Nias occurred during the Early Miocene whilst sedimentation was uninterrupted in other areas (Fig. 3; Samuel et al. 1995). The uplift was driven by reverse movements on originally extensional and transtensional faults. Regional subsidence commenced during the Mid-Miocene; deposition was initiated above the Mola Basement High and carbonate sedimentation, in areas such as the Gomo Sub-basin, gave way to shelfal and upper bathyal siliciclastic-dominated sedimentation despite a global fall in sea-level (Fig. 3; Haq et al. 1987). The main phase of uplift and deformation on Nias
occurred during the Pliocene (Fig. 3). The majority of the faults that had controlled the earlier extensional/transtensional geometry on Nias were reactivated (Fig. 4). The depth to basement in easternmost Nias (the Mola Basement High) is about 2 km (Fig. 5). Carbonates, sitting directly above basement, can be identified on a number of seismic sections across the Mola Basement High and they can be tied to Middle Miocene carbonates recovered from the Suma Well (Fig. lb; Karig et al. 1980). Equivalent carbonates can also be examined in rivers such as the Gawo in eastern Nias. The Upper Miocene succession thickens to the southwest (Fig. 5) and both the Upper Miocene and the Pliocene unconformity can be directly traced to the surface (Fig. 6). Workers such as Moore & Karig (1980) have remarked on the steepening of the Upper and Middle Miocene sedimentary rocks from the northeast to the southwest (from the Mola Basement High to the Gomo Sub-basin; Figs 4 & 6) and have termed the feature a large homocline or flexure. This flexure has been interpreted by Moore & Karig (1980) as the surface expression of a reverse fault with a steep west dip. In contrast, Diament et al. (1992) suggest that the 'flexure' is a continuation of the Mentawai strike-slip fault zone they identified to the southeast. The results of this study strongly support the observations made by Moore & Karig (1980) that the feature is the surface expression of a major southwesterly dipping fault. Rather than having a net reverse throw, however, this fault must have a large extensional throw to accommodate the thick succession of Lower Miocene and Oligocene sediments to the southwest; a stratigraphic thickness of at least 3.3 km of monoclinally tilted northeast dipping Lower Miocene-Oligocene sediments is exposed in the Gawo River (Fig. 6). These sediments are underlain by a further c. 2 km of Oligocene sediments (Fig. 4; Samuel & Harbury 1996). This thick Oligocene to Lower Miocene succession is clearly not present above the Mola Basement High and this high does not continue into the region of the Gomo Sub-basin (Fig. 4). The fault, which bounds the sub-basins on Nias from the Mola Basement High, is a long-lived structure; considerable normal throw occurred across it during the Oligocene and Miocene extensional phase. There is a large difference between the depth to basement either side of the fault (2 km on the footwall and an estimated maximum of 5-6 km on the hanging wall). The fault was mildly reactivated in a contractional sense during the Pliocene phase of uplift and deformation. This reactivation is manifested by minor folding of Middle and Upper Miocene sedimentary rocks which can be traced across the flexure in
N17
N12 N9
N4
P22 ,
~
N4
N7 N l l N17
N22
Sediment& melange complex
0
V=H
Coast
10 Km,
Mola Basement High-----
Lower-Middle Miocene shallowmarine sediments
N4.5 N4-5 P22 N4
Gomo Sub-basin
position of critical dates is shown using the planktic foraminiferal zonation scheme (Bolli et al. 1985). This cross-section is necessarily interpretative at depth. A full justification for these interpretations is given in Samuel & Harbury (1996).
Fig. 4. Cross-section over Nias Island. The line of cross-section is portrayed on Fig. 2. The section is drawn at true scale from near continuous traverse information and the
Basement
MiddleMiocene-LowerPliocene
P21-22
Mujoi Sub-basin
Oligocene-LowerMiocene deep marine sediments
Upper Pliocene-Pleistocene ~
coast
Lahewa Sub-basin
4~
t'rl
;Z
0
()O 90
80
70
6O
|
\, "
30
20
20
,'
2 "~' <-Y ,-.'~
40
50
Gawo river
-J~
30
------! ?Reef? ~
10
Coast 60
~
70
90
-- ~
80
~
00
10
-
20
30
40
150
70
180
230 240 250
260
Shelf edge
~
"
270 280 290 300
~
~ _ ~ ~ - ~ : ~ " ~ ' ~ . ~
190 200 210 220
~ ~ - ~ _ _ _ - - - -
. txeer ~
60
-
- --
Two-
-4
2tl(t~e
- 1
310 Sholpoints
Northeast
Fig. 5. Line drawing of two seismic lines crossing easternmost Nias and running offshore. The vertical exaggeration is c. × 2 and the lines are unmigrated. These profiles are tied by further lines to the S u m a Well (Fig. Ib).
25 km
4O
i<~/, ~
reflectlons~
~
~
50
"~-.....,.)J-'~",
~
Pliocene unconformity ~
Prominent reflection
,
-~ . _ ~
30
4
.
~
40
f-,~
NOISE
80 170 160 150
3
'"21
Two-
II
250 240 230 220 250 210 20(3 190
Southwest
;> 7~
.z .>
;>
.>
345
SUMATRAN FOREARC t~ MENTAWAI FAULT ZONE
50 K m ]
2 Km
: .... : :iiliiiiii:!i~i:i~i: )iI~ ~i~:~i::~iii!~:~ .... ili! 97°45'E
L E G E N D
Upper pliocen L •. . . . . .
-.
-., -.
L o w e r - M i d d l e Miocene shallow marine sediments
,eistocono
with limestone marker horizon •
~
U p p e r Miocene-lowest Pliocene with limestone a n d tuff marker horizons C o n f o r m a b l e Contact Major Faults
L o w e r Miocene d e e p marine sediments with conglomerate marker horizon Oligocene d e e p marine sediments
D ~ m ~ m ~ ~
Unconformity
Fig. 6. Traverse map of the Gawo river in eastern Nias. 1:50 000 topographic maps were used as base maps and a contour interval of 25 m is shown. A 'flexural' steepening is apparent through the Pleistocene and Middle Miocene. Note also the localized folding of the Upper Miocene. Both these features are related to the reactivation of the main basin-bounding fault (Fig. 4).
southeastern Nias (Moore & Karig 1980; Samuel 1994). The folding can be seen in a number of along-strike traverse sections including that of the Gawo river (Fig. 6). Whilst it is possible to correlate broadly the Mentawai fault zone with the region where the 'flexure' and the associated basin-bounding fault occur on Nias, it is clear that, although of variable width, the Mentawai fault zone is not confined to a single fault. The Mentawai fault zone is represented by a number of faults within a zone of
variable width (4-25 km). Onshore Nias there are a set of southwest dipping faults, synthetic to the main basin-bounding fault (Figs 2 and 4). These faults have a similar history to the main basinbounding fault; they were extensional during the Oligocene to Early Miocene, became passive features during the Mid-Miocene and Late Miocene and were variably reactivated during the Pliocene. Fault zones associated with faults, such as that separating the Gomo Sub-basin from the Mujoi Sub-basin, are exposed in some parts of the island,
346
M.A. SAMUEL 8z N. A. HARBURY
such as in the Me river (Fig. 2; Samuel & Harbury 1996). Deformation in these sections is exclusively contractional with no indication of any strike-slip movements. Although the Mentawai fault zone may be identifiable as a distinct feature for 45 km along the strike of southeastern Nias, its location to the north of the Gido river is less clear (Fig. 2). The trace of the main basin-bounding fault would appear to run offshore and there is a marked change in the style of deformation to the north of the Gido river (Fig. 2). The flexural steepening is no longer apparent. Instead the sedimentary rocks are deformed into broad, upright and NE verging folds, whose geometry is apparent from the trace of featureforming lithologies (Fig. 2). Discrete areas of folding are separated by north and ESE striking faults which controlled the development of these structures. The orientation of these structures is the same as that of faults such as the Batee and Singkel faults that cut across the present-day forearc basin (Fig. 1B).
Fault patterns and displacements on Nias Examination of aerial photographs, LANDSAT and particularly SAR (Synthetic Aperture Radar) images (Fig. 7) has allowed the identification of the following four distinct sets of faults on Nias (Fig. 8).
SE-striking. Four major SE striking faults are identifiable. They are readily distinguishable from the stratigraphic ridges which generally run southeast to northwest (Figs 2 and 6) as they bound separate areas of Nias with markedly different morphological expression (Figs 7 and 8). These geomorphologically distinct areas, which correspond to the sub-basins, have been subjected to different intensities of deformation (Fig. 4). The easternmost, SE-striking fault, is the least well defined as it is only marked by its flexural surface expression.
North-striking faults are clearly identifiable over large areas of Nias (Fig. 8). These faults are observed in many cases to cross-cut and offset the southeast striking faults with a right-lateral sense of displacement.
ESE-striking faults display a similar character to the north-striking faults (Fig. 8). Many of them clearly offset the SE-striking faults with a leftlateral sense of displacement. Some also offset the north-striking faults with a left-lateral sense of displacement. Conversely some of the ESE-striking faults are themselves offset, with a right-lateral sense of displacement, by north-striking faults (Fig. 8).
ENE-striking. A network of ENE striking faults is identifiable over all of Nias and is particularly pronounced in southeastern areas of the island (Fig. 8). Although excellent outcrops can be found on Nias within incised fiver valleys, the faults themselves are rarely exposed as the gouges are subject to rapid weathering and therefore direct kinematic indicators are generally obscured. The senses of displacement on these faults can however be determined from stratigraphic offsets and the relationships between the different faults (Fig. 8). The SE-striking faults controlled the geometry of the sub-basins during the Oligocene and Early Miocene extension/transtension when they developed as a SW-dipping synthetic set of extensional faults. The north-striking and ESE-striking faults also developed during the Oligocene and Early Miocene as oblique-slip faults. Differential throws on these faults led to the along-strike changes in sedimentation recorded in the stratigraphic record on Nias (Samuel et al. 1995; Fig. 3). Further movements occurred on these faults during the Pliocene phase of uplift and deformation as they offset some of the SE-striking faults. The apparent offsets indicate that the north-striking faults have a right-lateral sense of movement and the ESE faults have a leftlateral sense of movement although the amounts of lateral movement on individual faults are generally small (less than 5 km). The combination of the right-lateral movement on the north-striking faults and the left-lateral movement on the ESE-striking faults produces marked extension of the island parallel to the trench-forearc system. The deformation is one of pure shear, rather than simple shear. In some parts of Nias, sections of Oligocene and Lower Miocene sedimentary rocks have been rotated away from the strike of the island (Samuel et al. 1995). These rotations are confined to discrete fault bounded segments and do not exhibit any consistent rotation about vertical axes. There is no indication that they can be associated with a simple shear deformation although palaeomagnetic studies are planned to investigate this further. NW-SE extension across Nias is not only taken up by the north and ESE-striking faults but also by the ENE-striking faults which form part of an extensional fracture set (Fig. 8). Small-scale normal faults, generally with throws of less than 5 m, are exceptionally common on Nias and many of them have led to NW-SE extension.
Discussion and conclusions Although it is possible to identify a major structural feature on Nias that is in alignment with the Mentawai fault zone, there is a complete lack of
SUMATRAN FOREARC • MENTAWAI FAULT ZONE
347
Fig. 7. Detail of a 1:100 000 SAR (Synthetic Aperture Radar) image of Nias. The image has been used to map the network of faults over the island (Fig. 8).
evidence for significant strike-slip motion along it. The linear structure on Nias, previously referred to as a flexure, is the surface expression of an extensional fault that has been mildly reactivated in a contractional sense. Further synthetic Oligocene to Miocene extensional faults have been subject to varying degrees of contraction however it is clear that these faults have not been subject to any considerable strike-slip displacement. The structural studies rule out the possibility that
the Mentawai fault zone in the Nias area has a strike-slip origin. The evidence from Nias alone cannot however rule out the possibility that substantial strike-slip movements occur further to the southeast along the zone. Study of the seismic data available tends to suggest to us however that this is unlikely; although the structure of the Mentawai fault zone is variable it shows two characteristics that are most atypical of strike-slip faults zones. Firstly, sections to the northeast and southwest of
348
M.A. SAMUEL t~ N. A. HARBURY
\
\
\
L
N
0
10
20
30 i
krn %
Southeast striking faults " ~
North& east-southeast striking faults East-northeast striking faults
iiii!iiii~i~:i~i~i~: !/~ili~~ilii~¸
Fig. 8. Map of the fault network identified over Nias. The faults were traced directly from the 1:100 000 SAR image (e.g. Fig. 7). Fault movement since the Pliocene has extended the island parallel to the arc. the zone are always downthrown relative to the central regions. One of the characteristics of strikeslip fault zones is that the magnitude and senses of displacement change markedly along the length of the fault system (Underhill et al. 1988). Secondly the Mentawai fault zone is both atypically narrow (c. 25 km) and atypically straight for a major strikeslip system. Strike-slip faults with small amounts of movement (< 5 km) are identifiable on Nias but they cut across the strike of the island. Movements on these structures effectively resulted in extension of the island in a direction parallel to the arc. Furthermore
extension is accommodated on a well developed arc-perpendicular network of extensional faults. The strike-slip faults on Nias can be matched, in terms of orientation and sense of movement, to larger-scale strike-slip faults, such as the Batee and Singkel faults which cut across the present-day forearc basin (Fig. lb). Clearly the forearc is not behaving as a rigid plate but is being subjected, at least in the Nias area, to arc-parallel stretching as suggested by McCaffrey (1991). This study provides an indication of how arcparallel extension may be accommodated in forearcs. The importance of strike-slip faults such as the
SUMATRAN FOREARC • MENTAWAI FAULT ZONE Batee and Singkel faults has been highlighted by Matson & Moore (1992) who have shown that these faults controlled the development of basins within the present-day forearc basin (Fig. lb; Matson & Moore 1992). In addition to motion on relatively large strike-slip faults our work suggests that the extension that must occur along the forearc is taken up over a network of relatively small-scale strike slip and normal faults distributed across the forearc. Any vertical throws on these faults may be so small as to preclude identification on seismic sections. It was suggested by Diament et al. (1992) that the Mentawai fault zone lies on the boundary between the forearc ridge, which they believed was part of the accretionary prism, and the forearc basin which they supposed was located in a continental domain. The very straightness of the fault zone makes this unlikely. The studies on Nias have shown that, although the sub-basins developed on ophiolitic material, they evolved in an extensional environment and not as part of an accretionary prism. Basement to the west of the main basinbounding fault is ophiolitic although it is conceivable that the Mola Basement High may be composed of continental material. On a larger scale however gravity studies have failed to identify
349
any obvious oceanic-continental crustal boundary (Kieckhefer et al. 1981; Milsom 1993). Instead these studies have shown that the basement beneath the forearc basin is most likely to be extremely heterogeneous. Studies of melanges on Siberut, which lies to the southwest of the Mentawai fault zone, have identified blocks of continental basement material in the melanges on the island (Andi Mannga & Burhan 1986). The melanges have circular geometries and both include and intrude Neogene material suggesting that they have a similar origin to the diapiric melanges on Nias. Basement beneath parts of Siberut may therefore contain continental material that has been structurally covered, and brought up, in diapiric melanges. Although we suggest that the Mentawai fault zone is not a major strike slip zone defining a marked rheological boundary, this zone of deformation is an important feature whose origin needs to be understood. Onshore seismic lines across the Mola Basement High can be tied to offshore lines which run across the present-day forearc basin. These lines (e.g. Fig. 6) reveal the presence of extensional/transtensional basins actually within the present-day forearc basins and the Mola Basement High can therefore be viewed
Northeast
Southwest
Indian Ocean Trench
Basementfragment Oligoceneto recent sediments rich in ophiolitic material \ // Accretionarywedge \ r--Ni, ;---~ Forearc basin
Continental crust Sumatran fault zone & magmatic arc
/
Descending ~ocean;~c~ii~t
Crust of
\ ~Ol1~
unknown nature
Overriding continental lithosphere
Fig. 9. Cross-section from the Indian Ocean and trench across Nias to the Sumatran mainland. Nias island lies at the outer-edge of the Sumatran Forearc and the deformation seen over the island is comparable with that seen on seismic to the southeast of Nias along the Mentawai fault zone. The Mentawai fault zone represents a line of deformation at the outer edge of the Sumatran forearc.
350
M.A. SAMUEL & N. A. HARBURY
as a large horst block (Fig. 9). Gravity measurements (Milsom 1996) and seismic reflection studies (Matson & Moore 1992) reveal that the Mola Basement High has a north-south trend. Features such as the Mola Basement High have led to the variations in structural style along the Mentawai fault zone. The structure from Nias into the present-day forearc basin (Fig. 9) is readily comparable with that seen on some of the seismic sections crossing the Mentawai fault zone further to the southeast. The underlying feature is that central regions of the fault zone are structurally higher than the margins. It is striking that Diament et al. (1992) report on the identification of faulted blocks, horst and graben systems, and flexures within the Mentawai fault zone. These are all features that have been identified on Nias and need not require strike-slip motion. Structural studies on Nias, and the other forearc islands, have revealed the importance of mud diapirism in deformation along the outer-edge of the forearc (Samuel 1994). In addition diapiric activity, particularly along faults within the presentday forearc basin, has been recognized (Rose 1983; Milsom 1996). Features on several of the seismic sections crossing the Mentawai fault zone can be explained in terms of mud diapirism and these diapirs appear to produce substantial sea floor topography. In addition structures such as primaryrim synclines, which are diagnostic of diapirism (Talbot & Jackson 1987), are readily apparent. The authors suggest that features visible in seismic sections across the Mentawai fault zone can be explained by the inversion of originally extensional structures and by mud diapiric processes. This is entirely consistent with the geological data on Nias (Figs 4 & 9). The structure of the Sumatran outer-arc ridge is wholly comparable with that of the Mentawai fault zone. Diament et al. (1992) have suggested that the Mentawai fault zone may connect with faults such as the Batee Fault. Given, however, that there has not been extensive strike-slip movement along the Mentawai fault zone, where it runs onshore
southeastern Nias, linkage is not required. Structures such as the Singkel Fault and the Batee Fault appear to have had marked effects on the styles of deformation along the outer-arc ridge and may have led to the indentation into the forearc of islands such as the Banyak Group (Fig. 1). From the evidence on Nias it is clear that deformation along the flexure or Mentawai fault zone initiated during the Pliocene (Fig. 2; Karig et al. 1980; Samuel et al. 1995). This deformation may be a response to an increase in the rate of subduction (Samuel et al. 1995). The arc-parallel extension on Nias also dates from this time whilst spreading in the Andaman Sea initiated at 13 Ma (Curray et al. 1979). This implies that there may have been a time lag as the new stress regime propagated along the forearc. Karig et al. (1980) suggested that the flexure on Nias, and its associated continuation along the edge of the forearc, marked the arcward limit of accretionary prism deformation. We agree with Karig et al. (1980) that the Mentawai fault zone represents a deformation front. Our evidence indicates that the deformation, as seen on Nias, did not however occur in an accretionary prism and the Mentawai fault zone should not be viewed as a backthrust. Rather this subduction-driven deformation is manifested in the inversion of sedimentary basins which originally formed during the Oligocene to Early Miocene at the outer edge of the Sumatran Forearc. The great straightness of the Mentawai fault zone, and its position parallel to the trench, suggests that its location is controlled by parameters such as the angle and rate of subduction of the Indian Oceanic plate. We thank B. Situmorang of LEMIGAS for his support of our field programmes in Indonesia. E Banner,A. Bakri, L. Hartono, K. Bartram, A. Mitlehner and J. Ling are thanked for dating so many of our samples and M. Jones, G. Roberts and J. Milsom are thanked for their helpful comments. We are grateful to J. Malod and G. Moore for their constructive reviews. This work was sponsored by British Petroleum and the University of London SE Asia Research Group.
References ADAMS, C. G. 1984. Neogene Larger Foraminifera. Evolutionary and Geological Events in the Context of Datum Planes. In: IKEBE,N. & TSVCHI,T. (eds) Pacific Neogene Planes. University of Tokyo Press, Tokyo, 47-67. ANDIMANNGA,S. & BURHAN,G. 1986. Laporan Geologi Lembar Siberut,
Sumatra.
Sekala
1;250,000.
Geological Research and Development Centre, Bandung. BELLIER, O., St~BRIER,M. & PRAMUMUIJOYO,S. 1991. La grande faille de Sumatra: Gromrtrie, cinrmatique et quantit6 de drplacement mises en 6vidence par
l'imagerie satellitaire. Comptes Rendues de l'Academie des sciences, Serie II, 321, 12191226. BOLLI, H. M., SAUNDERS, J. B. 8/: PERCH-NIELSON, K. 1985. Plankton Stratigraphy. Cambridge University Press, Cambridge. BRADLEY, K. 1973. Geology of S.E. Nias Island, 1971 Geological Survey and relationship of the Nias Section to offshore seismic data. Union Oil
Company of Indonesia. BURROUGH,H. C. & POWER, P. E. 1968. Field Survey, Southern Part of North West Sumatra Contract
SUMATRAN FOREARC t~¢ MENTAWAI FAULT ZONE
Area. Union Oil Company of Indonesia, Report, RGE 43. CtW,RAY, J. R. 1989. The Sunda Arc: A model for oblique plate convergence. Netherlands Journal of Sea Research, 24, 131-140. - - , MOORE, D. G., LAWVER,L. A., EMMEL,E J., RAITT, R. W., ET AL. 1979. Tectonics of the Andaman Sea and Burma. In: WATKINS, J., MONTADERT, L. & WOOD-DICKERSON, P. W. (eds) Geological and geophysical investigations of continental margins. AAPG Memoir, 29, 189-198. DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S. 1990. Current plate motions. Geophysical Journal International, 101, 425-478. DIAMENT, M., HARJONO, H., KARTA, K., DEPLUS, C., DAHRIN, E, er AL. 1992. Mentawai fault zone off Sumatra: A new key to the geodynamics of western Indonesia. Geology, 20, 259-262. DOUVILLE, H. 1912. Les foraminif'eres de File de Nias. Sammlungen Geologisches Reichs-Museums, Leiden, 8, 253-278. ELBER, R. 1939. Report on a Regional Geological Survey ofNias. Permina NW Sumatra Report, 40102. F~TCH, T. J. 1972. Hate convergence, transcurrent faults and internal deformation adjacent to southeast Asia and the Western Pacific. Journal of Geophysical Research, 77, 4432-4462. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156-1167. HARaLrRY, N. A. & KALLAGHER,H. J. 1991. The Sunda Outer-Arc Ridge, North Sumatra, Indonesia. Journal of SE Asian Sciences, 6, 463-476. HARJONO, H., DIAMENT, M., DuBo~s, J. & LARUE, M. 1991. Seismicity of the Sunda Strait: Evidence for crustal extension and volcanological implications. Tectonics, 10, 17-30. HUCHON, P. & LE PICHON, X. 1984. Sunda Strait and central Sumatra fault. Geology, 12, 668-672. JARRARD, R. D. 1986. Terrane motion by strike-slip faulting of fore-arc slivers. Geology, 14, 780-783. KARIG, D. E., LAWRENCE,M. B., MOORE, G. E & CURRAY, J. R. 1980. Structural framework of the fore-arc basin, NW Sumatra. Journal of the Geological Society, London, 137, 77-91. , SUPARKA, S., MOORE, G. E & HEHANUSSA,P. E. 1979. Structure and Cenozoic evolution of the Sunda Forearc in the Central Sumatra region. In: WATKINS, J. S., MONTADERT, L. & WOODDICKERSON, P. (eds), Geological and geophysical investigations of continental margins. AAPG Memoir, 29, 223-237. KIECKHEFER, R. M., MOORE, G. E, EMMEL, E J. & SUGIARTA,W. 1981. Crustal structure of the Sunda forearc region west of central Sumatra from gravity data. Journal of Geophysical Research, 86, 7003-7012. MCCAFFREV, R. 1991. Slip vectors and stretching of the Sumatran forearc. Geology, 19, 881-884. MATSON, R. G. & MOORE, G. F. 1992. Structural Influences on Neogene subsidence in the central Sumatra fore-arc basin. In: WATKINS,J. S., FENGZHI QIANG & MCMILLEN, K. J. (eds) Geology and Geophysics of Continental Margins. AAPG Memoir, 53, 157-181.
351
MILSOM, J. S. 1993. Interpretations of gravity data from the vicinity of Nias. University of London Geological Research in SE Asia, Report 119. 1996. Basin Formation in the Nias Area of the Sumatra Fore Arc, Western Indonesia. Geological Society of Malaysia, Special Publication, in press. MINSTER, J. B. & JORDAN, T. 1978. Present-day plate motions. Journal of Geophysical Research, 83, 5331-5354. MOORE, G. E & KARIG, D. E. 1980. Structural geology of Nias Island, Indonesia: implications for subduction zone tectonics. American Journal of Science, 280, 193-223. , BILLMAN,H. G., HEHANUSSA,P. E. & KARIG, D. E. 1980a. Sedimentology and paleobathymetry of Neogene trench-slope deposits, Nias Island, Indonesia. Journal of Geology, 88, 161-180. , CURRAY, J. R., MOORE, D. G. & KARIG, D. E. 1980b. Variations in geologic structure along the Sunda Fore Arc, northeastern Indian Ocean. In: HAYES, D. E. (ed.) The tectonic and geologic evolution of southeast Asian seas and islands. American Geophysical Union Monograph, 23, 145-160. NEWCOMB, K. R & MCCANN, W. R. 1987. Seismic history and seismotectonics of the Sunda Arc. Journal of Geophysical Research, 92, 421-439. PAGE, B. G. N., BENNETT,J. D., CAMERON,N. R., BRIDGE, D. MC., JEFFREY,D. H., ETAL. 1979. A review of the main structural and magmatic features of northern Sumatra. Journal of the Geological Society, London, 136, 569-579. PAUL, E 1941. Supplementary (Commitment) Survey, Nias. Permina NW Sumatra Report, 40103. PUBELLIER, M., RANGIN, C., CADET, J-P., TJASHURI, I. BtrrrERLIN, J. & MULLER, C. 1992. L'ile de Nias, un edifice polyphase sur la bordure interne de la fosse de la sonde (Archipel de Mentawai, Indonesie). Comptes Rendues de l'Academie des sciences, Serie II, 315, 1019-1026. ROSE, R. 1983. Miocene carbonate rocks of Sibolga basin, northwest Sumatra. Proceedings Indonesian Petroleum Association, 12, 107-125. SAMUEL, M. A. 1994. The structural and stratigraphic evolution of islands at the active margin of the Sumatran Forearc, Indonesia. PhD Thesis, University of London. -& HARBURY,N. A. 1996. Basin Development and Uplift at an Oblique-slip Convergent Margin: Nias Island, Indonesia. Geological Society of Malaysia, Special Publication, in press. --, JONES, M. E. & MATTrmWS, S. J. 1995. Inversion-controlled uplift of an outer-arc ridge: Nias Island, offshore Sumatra. In: BUCHANAN,J. G. & BUCHANAN, P. G. (eds) Basin Inversion. Geological Society, London, Special Publication, 88, 473-492. TALBOT, C. J. & JACKSON,M. P. A. 1987. Salt Tectonics. Scientific American, 256, 70-79. UNDERHILL, J. R., GAYER, R. A., WOODCOCK, N. H., DONNELLY, R., JOLLEY, E. J. & STIMPSON, I. G. 1988. The Dent Fault System, northern Englandreinterpreted as a major oblique-slip fault zone. Journal of the Geological Society, London, 145, 303-316.
Tectonic evolution of the Bantimala Complex, South Sulawesi, Indonesia KOJI WAKITA 1, J A N S O P A H E L U W A K A N 2, K A Z U H I R O M I Y A Z A K I 1, I S K A N D A R Z U L K A R N A I N 2 & M U N A S R I 2' 3
1 Geological Survey of Japan, 1-1-3 Higashi, Tsukuba, Ibaraki 305, Japan 2 Research and Development Centre for Geotechnology, Jl. Cisitu, 21/154D, Bandung, 40135, Indonesia 3 Tsukuba University, 1-1-1 Tennodai, Tsukuba, Ibaraki 305, Japan Abstract: The Bantimala Complex of South Sulawesi, Indonesia is an assemblage of northeastdipping tectonically stacked slices. The slices consist mainly of high pressure metamorphicrocks, radiolarian chert, breccia, sandstone and shale, and melange. In order to understand the tectonic evolution of the Bantimala Complex, we have investigated the lithology, age, stratigraphy, structure and relationships of the components. The K-Ar ages of high P-low T metamorphicrocks suggest that an oceanic plate subducted beneath the Sundaland continent during the Late Jurassic or earliest Cretaceous. The subduction ceased during the Albian, and the high pressure schists were exhumed and eroded at the surface before and during the deposition of middle Cretaceous radiolarian chert. The exhumation of the schists was related to the collision of microcontinents derived from Gondwanaland. The Jurassic shallow marine sedimentary rocks in the Bantimala Complex are possibly remnant fragments of the collided microcontinent. Tectonic stacking of the Bantimala Complex was caused by Neogene subduction and collision of another continental fragment further to the east.
Melange complexes are widely distributed in the Indonesian and east Malaysian regions. Tectonic, sedimentary and diapiric processes have been proposed as possible mechanisms for melange formation in these regions. Asikin (1974), Hamilton (1979) and Hehuwat (1986) stressed tectonic processes occurring during interaction of continental and subducting plates in the origin of melanges of the Indonesian region. Barber et al. (1986) proposed a diapiric origin for melange of the Bobonaro Scaly Clay in Timor, Indonesia. Clennell (1991) proposed that sedimentary processes such as submarine slumps and slides were the major mode of melange formation in east Sabah, Malaysia. These three processes are not mutually exclusive, but rather are end-member driving mechanisms which can operate together or sequentially (Lash 1987; Clennell 1991). It is important to note that multiple processes occurring at different stages in their history caused the chaotic mixing of many melange complexes. In order to understand the multiple processes of chaotic disruption in the Cretaceous melange complexes in Java and South Sulawesi, the authors have tried to determine in detail the tectonic history of the melange complexes. Recently, Wakita et al. (1994a) determined the history of the accretionary process which formed the Luk-Ulo Melange Complex of the Karangsambung area,
central Java on the basis of radiolarians extracted from siliceous and argillaceous rocks. The detailed dating of sedimentary rocks suggests that subduction occurred between the Early and Late Cretaceous. Oceanic materials such as chert, limestone and pillow basalt travelled on the oceanic plate, and were accreted with terrigenous materials at the 'Karangsambung Trench'. Cretaceous melange complexes are found in Sulawesi and south Kalimantan as well as in central Java (Fig. 1). The Bantimala Complex components are very similar to those of the Luk-Ulo Melange Complex in central Java but differences of structure and lithology suggest a different tectonic evolution in each area. Described here is the geology of the melange complex of the Bantimala area along with discussion on the tectonic evolution of this area during Mesozoic and Cenozoic time.
Tectonic setting The southeastern part of pre-Cretaceous continental basement of Sundaland extends into west Kalimantan (Hutchison 1989). Cretaceous granites were intruded into the basement of central and western Kalimantan, Sumatra and the western part of the Java Sea (Hamilton 1979). Cretaceous melange complexes surround the southeastern margin of Sundaland. As they are widely covered
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 353-364.
353
354
K. WAKITAET AL.
by Cenozoic sedimentary and volcanic rocks, the complexes are exposed only in a few localities (Fig. 1) such as Bantimala (South Sulawesi), Karangsambung (central Java) and Meratus (south Kalimantan). The distribution of JurassicCretaceous granites and the direction of the continental growth by accretion suggest the oceanic plate subducted mainly towards the Sundaland continent during Cretaceous time. The Cretaceous complexes of the Meratus and Bantimala areas formed a single complex prior to Neogene foreland subsidence of the Makassar Strait (Bergman et al. 1995). Tertiary subduction complexes and obducted ophiolite are distributed to the east in central and east Sulawesi (Simandjuntak 1990; Parkinson 1991; Coffield et al. 1993; Bergman et al. 1996), and the microcontinent of the Banggai-Sula has pushed them westward (Fig. 1). Outline of the geology
The Bantimala area is located c. 40 km northeast of Ujung Pandang, South Sulawesi (Fig. 2). The detailed geology of this area has been investigated by Sukamto (1975, 1978, 1982, 1986) and Hasan (1990, 1991). The Bantimala area is underlain by the Bantimala Complex (Jurassic-Cretaceous), propylitized volcanic rocks (Palaeocene), the Malawa Formation (Eocene), the Tonasa Formation
__,,
I.
A)
(Eocene-Middle Miocene), the Camba Formation (Middle to Late Miocene), and Quaternary sedimentary cover in ascending order (Fig. 3). The Tonasa Formation rests conformably on the Malawa Formation. The other Cenozoic formations lie unconformably on the older formations. The Bantimala Complex is a tectonic assemblage of slices and blocks consisting of sandstone, shale, conglomerate, chert, siliceous shale, basalt, ultramafic rocks, schist and 'schist breccia'. The ages of components range from Jurassic to middle Cretaceous. Propylitized volcanic rocks similar to Lamasi Complex rocks to the north consist of breccia, lava and tuff mainly of andesitic and partly of basaltic and trachytic composition. K-Ar and fission track dating indicate a Palaeocene age (Sukamto 1982). The Malawa Formation is composed of sandstone, conglomerate, siltstone and marl with layers or lenses of limestone and coal or lignite. This formation is Upper Palaeocene to Lower Eocene on the basis of large foraminifers, palynology and ostracods (Sukamto 1982; Hasan 1990). The Tonasa Formation consists of limestone with subsidiary tuffaceous marl, siltstone and sandstone and contains foraminifera of Lower Eocene to Middle Miocene age. The Camba Formation is composed of marine volcanic and sedimentary rocks. The sedimentary rocks are sandstone, siltstone and claystone, intercalated with marl, limestone and coal. Volcanic
~
"i!i!!!!!i!i!i!i!i
,~
I .
N
g~G--
# nt
,~~..:.::!:i~lIMeratus
~~," e~,',',',',',',',',',',',',',','," ti al c~ e- o~ u~""s ~
/ ]-lds
/
~
n
g
s
~po
,.J .~ ~
%----T,,r:~-rv
s u b d u~ c t i o n ~ Q ~
c~
l?s
~
~ ~
1~o"
"
.i.:.~:.!..~.:.i.:.i.:..-:.~ Continental fragment
Timor / e~o -~
~ -
Fig. 1. Tectonicsetting of the Bantimala Complex (modified from Hamilton 1979)
,
BANDA SEA
~oo 4oo ,oo .
.
.
.
~oo ~ooo
,
,,.
BANTIMALA COMPLEX EVOLUTION, SULAWESI
355
Diorite (Miocene)
Bantimala Complex ~
Balangbaru Formation
D Melange . . . . . . . . . .
-Pankajene
~.v#~/~l
I
~
Schist
I B a n t J m a l a~ ) x
I~
5 I
,
10 i
15 i
20 km I
1 1914EE Fig. 2. Locality and distribution map of the Bantimala Complex. The exposures near Barru are equivalent to the Bantimala Complex (after Sukamto 1975).
rocks of basaltic and andesitic composition include tuff, breccia, volcanic conglomerate and lava. Fossils of the Camba Formation indicate it ranges from Middle to Upper Miocene. K-Ar ages of Camba Formation volcanic rocks from Buakayu, western Sulawesi range from 6-14 Ma (Bergman et al. 1996). Diorite stocks intrude the Camba Formation as well as the Bantimala Complex and the Tonasa Formation. Quaternary alluvial deposits in the western part of the Bantimala area consist of gravel, sand and clay.
Bantimala Complex The Bantimala Complex is a tectonic assemblage of various lithologies ranging in age from Jurassic to Cretaceous. The complex is unconformably overlain by, or in fault contact with, Tertiary and Quaternary formations.
Lithology The main components of the Bantimala Complex are sandstone, shale, conglomerate, chert, siliceous shale, basalt, ultramafic rocks, schist, and 'schist breccia'. The stratigraphic relationships among the
components are mostly unclear because of their fault contacts with each other, except for the following. Schist is unconformably overlain by 'schist breccia' (A. J. Barber pers. comm. 1994). The 'schist breccia' grades into sandstone that is overlain by chert. The chert is conformably overlain by siliceous shale interbedded with sandstone and is locally intercalated with conglomerate. Melange is one of the characteristic units in the Bantimala Complex, and includes rock types such as sandstone, shale, siliceous shale, chert, basalt, schist, and felsic igneous rocks within a sheared shale matrix. Jurassic sedimentary rocks are known in two tectonic slices and are quite different from Cretaceous sandstone, shale and conglomerate of the complex (see below).
Structure The tectonic slices of the complex are elongated and strike NW-SE. The dip of strata ranges from 20-80 ° towards the northeast or east (Figs 3 and 4). Palaeocene propylitized volcanic rocks unconformably cover the faults between tectonic slices of the Bantimala Complex, and the boundary faults between the Balangbaru Formation and the
K. WAKITAET AL.
356
Quaternary ooooo , oooooo ~:;.. o o o o i:.... o o o
, ,~.
o
oooo ooo o ooo oo ooo oo ooo ooo ,
~
o
"o ~.
;?;.*;,*;,~;.';.~;.';?;.';~;?;~;?;." i
o
!
oi
i
-
oooooooo
=..=.
dionte Eocene - E. Miocene
o o
o' ?;,%.*;,,~;,*;,.%,%.%,*;t;,,~;.%?;?;t; i ~ :;:~:;:~,:
alluvium
M. Miocene P l i o c e n e Camba Formation
ooooooooo
.,?;?;%tsr~rsr = o o o o o o o o .;.*;,%.*;,%t;.%.~;... o o o o o o o o o ,?;.*;?;?;?;?;?;~;.~; ' : o o o o o o o o o ~,;.';,~;.~;~;%?;?;." .=:~:=. o o o o oo ;.*;..~;,%.*;,%.*;,*' "::;:~:;.... o o
oo ooo ~
I--"]
o o o o c 4 _4 5_ _ oooooooooo
Malawa
o
& Tonasa
Formations
Paleocene - Eocene volcanic rocks
Propyfltized dacite
& diorite
Bantimala Complex '~ '=:¢~:=,:¢~:: ". . . . . . I-
"":'::
o o o,
:;:.... ::,:~'=
o .....
..¢ "
Balangbaru Formation Melange
=.:~:~:
bedded chert ....
o o o o % 0 0 0 0 o'"o'~:~.... ooo
o o o
oooo ooo
schist
.
o o
".-.=i .;%.*;?;~ .:
breccia
schist
rocks
ultramafic Jurassic
SW
NE' lOOO 500 0
_o2 _Ol~:
marine
shallow formation Fault
Fig. 3. Geological map of Bantimala area (after Sukamto 1986).
Bantimala Complex. NW-SE thrust faults cut the Miocene Camba Formation (Fig. 3) and PlioPleistocene wrench faults occur locally.
Melange The most distinct outcrop of melange occurs near the junction of the rivers Pangkajene, Pateteyang and Cempaga (Figs 4 and 5). The melange includes clasts and blocks of chert, sandstone, basalt, limestone and schist embedded within a sheared shale matrix. Major clasts are sandstone, chert and siliceous shale. Basalt and limestone are locally dominant. Fragments of metamorphic rocks are very rare. The shale matrix is usually sheared to some degree, and is severely sheared along the western part of the Pateteyang River (Fig. 5). The clasts are subrounded to subangular, and rhomboidal, spherical, blocky and irregular in shape. Clast sizes range from several millimetres to several hundred metres. Along the Pateteyang River, clasts of siliceous shale and sandstone are elongated in the highly sheared matrix. Sandstone clasts are usually less than 1 m in length, but sometimes reach several metres. These are usually poorly sorted and fine to coarse grained arkoses, containing mineral (quartz, feldspars, micas, etc.) and lithic fragments (metamorphic rocks, felsic tufts, etc.). Limestone clasts are locally dominant in the melange near the mouth of the Cempaga River
(Fig. 6a). Some include fossils such as hexacorals, foraminifers, calcareous algae and sponges. As stromatoporids are not present, the limestones are younger than Jurassic and are considered to be Cretaceous (K. Mori pers. comm. 1994). Highly sheared polymictic rocks are distributed along the Pateteyang River. Brecciated diorite and schist as well as clasts of sandstone and chert occur in severely sheared shale (Figs 5 and 6b). Diorite forms Tertiary intrusions in this area and occurs as clasts. This mixture is not typical of melange in the Bantimala Complex. Berry & Grady (1987) attributed the chaotic features to Plio-Pleistocene faulting.
Ultramafic rocks Ultramafic rocks of the Bantimala Complex are mainly distributed along the western margin of the Balangbaru Formation (Fig. 3). Very small outcrops of serpentinite also occur along some of the southwestern boundaries of the metamorphic tectonic slices. The ultramafic rocks are mostly serpentinized peridotite, with local chromite lenses. They are unconformably overlain by sandstone of the Balangbaru Formation (Hasan, 1990).
Schist The metamorphic grade of schists in the Bantimala Complex ranges mainly from greenschist to
L
, ss mx
0 , 100 , 200m
t
~
/ ch
))
~Tonasa Formation (Teariary Limestone)
Basalt
bs~ To
Sandstone
Melange m matrix MeJange atrax
ss~
x ~~ mX
I
.~ 119"41'
---F
Fault
~
sc
r~
fi "~.,)~2c'~'~-~_
"~\o/~,.e v' °"' ° ' sc"~' "~ ~l,~ .~.".bm
Breccia wilh metamorphic rock fragments
Schist
Siliceous shal
-,-~--. Thrust
bm ~
sscc ~
c--'--~~a.lone2 ~P.la ~,~/erg ka.lolae
Fig. 4. Route map along the Pangkajene River, east of the Mangilu village. The box shows the area of Fig. 5.
To
4*46'30"
I 119 ° 40'
t~ t.~ "-,3
;,o
7
©
< © t"
©
>
>
> 7~ ,.-]
358
K. WAKITA E T A L .
i
119° 40' 55"
I 119 ° 41' 00"
mx
1 119" 41' 05"
SC t~t
Pangkajene
mx
Melange matrix ( S h e a r e d s h a l e w i t h clasts)
ch
Radiolarian
ss
Sandstone
br
Breccia, w i t h basalt, l i m e s t o n e and c h e r t f r a g m e n t s
chert
~X
River 'SO 6"-
SS
b ch ch
sc
~
....
~
bs Basalt
'~', ', bs
4"46'15"-
sc
Schist
bm
Breccia, with metamorphic
x bs
1.,(t~
u
Ultramafic
fragments
rock
Fault
i"t /
Riverside
o
,brn ":*
@"2
bn
N
SC'~
t
/T/X
ch
lt~,
4°46'30"
o
5ore
:
•
•
Fig. 5. Route map along the Cempaga and Pateteyang Rivers. This is the same location as that shown in Berry & Grady (1987, fig. 5). Most of the slices and blocks are in tectonic contact with each other. Only major faults are shown on the map. Breccia (bm) includes diorite clasts as well as metamorphic fragments within sheared red shale. Eclogite and glaucophane schist occur as boulders at the base of schist in the northern part of the map.
359
BANTIMALA COMPLEX EVOLUTION, SULAWESI Table 1. K-At ages of muscovites from metamorphic rocks of the Bantimala Complex. Sample no.
Rock type
Mineral
Age (Ma)
BT-11b Mg-47 L-01B BT-08e BT-17
eclogite eclogite eclogite pelitic schist pelitic schist
muscovite muscovite muscovite muscovite muscovite
132 + 7 124 _ 6 113 _ 6 114 _+6 115 _ 6
amphibolite facies. In a tectonic slice, the metamorphic grade increases toward the northeast (tectonically towards the top). Glaucophane schist and eclogite associated with serpentinite are locally present as tectonic blocks and slices and occur along the boundary fault of one tectonic slice consisting of metamorphic rocks. Most of the metamorphic rocks of greenschist to amphibolite facies were originally eclogite and have suffered retrogressive metamorphism. Wakita et al. (1994b) reported K-Ar ages of micas from schists ranging from 113-132 Ma (Table 1). The ages of eclogite vary widely, although the ages of greenschists are concentrated c. 114-115 Ma.
Schist breccia 'Schist breccia' is one of the main lithologies of the Bantimala Complex. Most are sedimentary breccias consisting mainly of schist fragments, although it is sometimes difficult to distinguish tectonically brecciated schist from sedimentary breccia. The metamorphic grades of the schist fragments are the same as that of schists in the tectonic slices of the Bantimala Complex. The relationship between chert and 'schist breccia' of the Bantimala Complex has been described as an 'unusual unconformity' in the Paring River (Haile et al. 1979; Wakita et al. 1994b). The 'brecciated schist' grades into sandstone, which is overlain by radiolarian chert. As the chert is intercalated with thick beds of 'schist breccia' at several horizons, as well as with sandstone layers of various thicknesses, the three rock types, i.e. chert, sandstone and 'schist breccia' are considered to be contemporaneous deposits of an unstable sedimentary basin.
Chert and siliceous shale Chert layers range from 1-20 cm thick and are interbedded with thinner shale layers less than 1 cm thick. The bedded chert is mostly red or reddish brown, and sometimes pale green or grey in colour. It is composed mainly of skeletons and fragments of radiolarians, and a small amount of shale. The chert sometimes includes well preserved middle
4°mr 3.21 2.94 3.76 2.5 2.37
%40Ar
%K
93.9 90.4 96.6 92.2 94.5
5.99 5.94 8.34 5.48 4.98
Cretaceous radiolarians (upper Albian to lower Cenomanian) including Holocryptocanium barbui,
Thanarla conica, Archaeodictypomitra vulgaris and Rhopalosyringium majuroensis. The chert is underlain by 'schist breccia' and coarse grained sandstone. It is intercalated with sandstone beds and laminae in the lower part of the succession. The chert is intercalated with rhyolite tuff layers along the Pateteyang River (Fig. 6e). The rhyolite is pale green in colour, is usually fine but sometimes coarse grained. The chert locally grades into siliceous shale towards the stratigraphic top in some localities. The shale is grey or reddish-brown in colour, and composed of radiolarian skeletons, terrigenous fragments and other detrital materials. One of the authors (Munasri) extracted radiolarians including Holocryptocanium barbui, Pseudodictyomitra pseudomacrocephala and Thanarla veneta from a siliceous shale collected in the Pangkajene River.
Jurassic Paremba Sandstone Jurassic shallow marine sedimentary rocks, called the Paremba Sandstone (Sukamto & Westermann 1992), are incorporated as tectonic slices in the Bantimala Complex. The lower part of the Paremba Sandstone along the Bontolio River is composed of thin bedded sandstone and shale, intercalated with thin limestone layers. Some shallow marine sedimentary structures such as ripple and convolute laminations are recognized (Fig. 6c). The upper part of the formation is rich in conglomerate (Fig. 6d) which includes pebbles mainly of basalt and schist. Ammonites (e.g. middle Liassic Fuciniceras), gastropods and brachiopods of the Lower and Middle Jurassic are reported from the Paremba Sandstone (Sukamto & Westermann 1992).
Cretaceous Balangbaru Formation Cretaceous flysch sequences assigned to the Balangbaru Formation are widely distributed in the northeast and north in the Bantimala area. Detailed descriptions were given by Hasan (1990, 1991) and the following is based on Hasan (1990).
360
K. WAKITA ET AL.
(a)
~)
(c)-
(a)
(e)
(13
Fig. 6. Outcrop photographs of the Bantimala Complex. (a) Cretaceous limestone blocks embedded within a sheared shale matrix along the Cempaga River. (b) Highly sheared shale matrix including various kinds of rocks as clasts, along the Pateteyang River. Plio-Pleistocene faults cut the shale matrix of the Cretaceous melange. (c) Jurassic Paremba Sandstone, showing ripple marks, interbedded with shale, along the Bontorio River. (d) Jurassic Paremba Sandstone with intercalations of conglomerate, along the Bontorio River. (e) Rhyolite layers intercalated in radiolarian chert, along the Pateteyang River, near the mouth of the Sanggi River. (f) Turbidite of the Balangbaru Formation, in the upstream part of the Balangbaru River.
The formation is subdivided into three members, in ascending order: the Allup, Panggalungan and Bua Members. The Allup Member is composed of pebbly sandstone, conglomeratic breccia and interbedded sandstone with shale. The Panggalungan Member consists of interbedded sandstone
and shale and chaotic breccia deposits. The Bua Member is composed of interbedded sandstone, shale (Fig. 6f) and conglomerate. Hasan (1990) interpreted the facies associations of the Allup, Panggalungan and Bua members as inner fan, outer fan to basin plain and middle fan respectively. The
BANTIMALA COMPLEX EVOLUTION, SULAWESI mineral composition of the sandstones varies from the bottom to top of the formation. Garnet, spinel, glaucophane and chloritoid are characteristic of the lower parts of the Balangbaru Formation, while zircon, apatite and tourmaline increase in proportion towards the stratigraphic top. The mineral assemblages suggest that the lower sediments were derived principally from metamorphic rocks of the Bantimala Complex, and the upper sediments have a continental or magmatic arc provenance. Hasan (1990) reported planktonic foraminifera, such as Globotruncana helvetica, G. arca, G. foricata and Heterohelix globulosa, from the fine grained sandstone and siltstone of the Panggalungan and Bua members of the Balangbaru Formation indicating a lower Turonian to upper Maastrichtian range. One of the authors (Wakita) extracted radiolarians from shale and siliceous shale of all members of the Balangbaru Formation. Archaeodictyomitra spp., Pseudodictyomitra spp., Rhopalosyringium majuroensis, Thanarla spp., Praeconocaryomma sp., Stichomitra sp. Archaeodic~omitra and Pseudodictyomitra are most dominant genera among them. The assemblage remains very similar from the stratigraphic bottom to top of the formation, and is also very similar to that from chert of the Bantimala Complex reported by Wakita et al. (1994b). Rhopalosyringium majuroensis ranges from upper Albian to upper Turonian (Schaaf 1984) and the common occurrence of Pseudodictyomitra spp. indicates that assemblage ranges from Lower to middle Cretaceous. No species indicating an age younger than Coniacian has been found in samples of the Balangbaru Formation. The radiolaria indicate the age of the Balangbaru Formation is between upper Albian and Turonian. Sukamto (1975, 1978, 1982, 1986) and Hasan (1990) excluded the Balangbaru Formation from the Bantimala Complex. However, the new radiolarian age data suggest that the age of the Balangbaru Formation is not different from the age of chert in the B antimala Complex. The Balangbaru Formation also occurs as tectonic slices stacked with the other members of the Bantimala Complex. The authors therefore propose, on the basis of new radiolarian data, that the Balangbaru Formation is part of the Bantimala Complex. The formation is almost contemporaneous with sandstones deposited on chert of the Bantimala Complex, and on the ultramafic rocks.
Tectonic evolution The following lithologies and structures of the Bantimala Complex are critical elements in the tectonic evolution of the Bantimala area: high pressure metamorphic rocks, Jurassic shallow
361
marine sedimentary rocks, schist breccia overlain by radiolarian chert, melanges including blocks of various kinds of rocks, and tectonically stacked slices of various rocks unconformably overlain by Palaeogene volcanic rocks. To explain the l i t h o l o g i e s and structures of the Bantimala Complex, the authors propose the following hypothetical tectonic evolution: Cretaceous subduction, collision and accretion of a microcontinent, and Neogene tectonic stacking of slices caused by westward collision of another microcontinent. Figure 7 shows a possible tectonic evolution of the Bantimala area based on this hypothesis. K-Ar ages of high pressure metamorphic rocks range from 132-113 Ma, and indicate that there was subduction from Jurassic to early Cretaceous towards the 'West Kalimantan Continent' as indicated by the distribution of Jurassic to Cretaceous granites. Jurassic shallow marine sedimentary rocks are the most important piece of evidence for a microcontinent that subducted, collided and accreted in the early Cretaceous 'Bantimala trench'. Shallow marine clastic formations contemporaneous with the Paremba Sandstone of the Bantimala area are found further to the east in central and SE Sulawesi, and in the Banggai-Sula area (e.g. the Buya Formation; Surono & Sukarna 1993). On the other hand, there is no indication of unmetamorphosed Jurassic rocks to the west of the Bantimala area. Various sizes of continental fragments drifted northward and accreted along the Asian continental margin since the break-up of the Gondwanaland (Nur & Ben-Avraham 1983; Maruyama et al. 1989). The microcontinent on which the Jurassic Paremba Sandstone was deposited is thought to be one of them. Subduction of the oceanic plate caused the formation of high pressure metamorphic rocks at least from Late Jurassic to Early Cretaceous times, and brought the microcontinent with its overlying Jurassic shallow marine sedimentary rocks into the 'Bantimala trench'. After the arrival at the trench, the microcontinent was subducted, collided and accreted within the accretionary wedge. After the collision and accretion of the microcontinental block, subduction ceased at the 'Bantimala trench'. Underthrusting of the light and buoyant continental fragment caused the rapid uplift and exhumation of high pressure metamorphic rocks. Their 113-132Ma K-Ar ages indicate the time of cooling during exhumation. After metamorphic rocks of the Bantimala Complex appeared at the surface, they were eroded and provided 'schist breccia' and sandstone to a sedimentary basin in which radiolarian remains were deposited at a relatively high rate during the late Albian to early Cenomanian (Wakita et al. 1994b).
K. WAKITA ET AL.
362
late Albian
Jurassic West Kalimant an Cont inent
Shallow marine sediment at ion
schist breccia
radiolarian chert
Exhumat ion Collision & accret ion
High P/T met amorphic rocks
C e n o m a n i a n - early T u r o n i a n olist ost rome
O l i g o c e n e - Pliocene
BalanclbaruFormation sla
chert
Fig. 7. Tectonicevolution of the Bantimala area, South Sulawesi.
The radiolarian biostratigraphy and lithostratigraphic relationships reveal that the ages of 'schist breccia', chert, and the Balangbaru Formation are similar. The occurrences of slump deposits and 'schist breccia' layers within chert suggested that they were deposited in an unstable sedimentary basin. Successive exhumation of metamorphic rocks caused the uplift of the basement of the basin. The basement provided fragments of the other members of the Bantimala Complex into melanges which were originally formed as olistostromal deposits. Ultramafic rocks of the Bantimala Complex were exhumed at almost the same time as the high pressure schists. The Balangbaru Formation was deposited unconformably on the ultramafic rocks. After deposition, the tectonic slices of the Balangbaru Formation and its ultramafic basement were tectonically juxtaposed with other slices of the Bantimala Complex. The most distinct feature of the Bantimala Complex is tectonic stacking of slices. The structure is very similar to that of accretionary prisms and accretionary complexes elsewhere in the world. The stacking of structures dipping to the east are, however, opposite to that expected from a westward-dipping, oceanic plate subduction toward the Sundaland continent during Cretaceous time. Sukamto (1982) suggested that tectonic stacking and mixing in the Bantimala Complex were formed prior to Late Cretaceous, because the Balangbaru
Formation was considered to overlie the Bantimala Complex unconformably. However, as the Balangbaru Formation is found as tectonic slices in the Bantimala Complex, the stacking must have happened after its deposition. Although the geological map of Sukamto (1986) indicates that Palaeogene propylitized volcanic rocks unconformably cover the tectonic slices and their boundaries (Fig. 3), we doubt that the stacking of slices occurred in before the Palaeogene. We need to check the nature of boundaries and relationship between members of the Bantimala Complex and Palaeogene propylitized volcanic rocks. Thrust faults locally cut the Miocene Camba Formation. Some of the faults obviously moved later than Miocene time. Coffield et al. (1993) and Bergman et al. (1996) argued that stacking of the Lamasi Complex of South Sulawesi was caused by westward obduction of the ophiolite in Oligocene time and Miocene to Pliocene collision of microcontinent. A similar scenario is acceptable for the tectonic stacking of the Bantimala Complex. Obduction of an oceanic plate and subsequent collision of a microcontinent caused westward-thrusting during Oligocene to Pliocene time (Parkinson 1991; Coffield et al. 1993; Bergman et al. 1996). Tectonic stacking of the Bantimala Complex is similarly interpreted to result from ophiolite obduction followed by microcontinent collision in Tertiary time.
BANTIMALA COMPLEX EVOLUTION, SULAWESI
Conclusions The Bantimala C o m p l e x consists of tectonic slices of sandstone, shale, conglomerate, chert, siliceous shale, basalt, ultramafic rocks, schist, and 'schist breccia' The rock components of the complex range in age from Jurassic to middle Cretaceous. The c o m p l e x was u n c o n f o r m a b l y overlain by Cenozoic sedimentary and volcanic formations and was intruded by Tertiary igneous rocks. Schists in the Bantimala C o m p l e x range mainly from greenschist to amphibolite facies and locally include glaucophane schist and eclogite. The K-Ar ages of the schists range from 113-132 Ma. Schist was unconformably overlain by radiolarian chert and clastic rocks during the late Albian to early Cenomanian. Radiolarian biostratigraphy reveals that the Balangbaru Formation ranges from upper Albian to Turonian and belongs to the Bantimala Complex. Jurassic shallow marine sedimentary rocks are the most important pieces of evidence for a lost m i c r o c o n t i n e n t that subducted, collided and accreted in the early Cretaceous ' B a n t i m a l a trench'.
363
Underthrusting o f a light and buoyant continental f r a g m e n t c a u s e d the rapid uplift and exhumation of high pressure metamorphic rocks including eclogite. Tectonic stacking of slices in the Bantimala Complex was mainly caused by Oligocene ophiolite obduction and M i o c e n e - P l i o c e n e collision of a microcontinent. This paper is one of the results of the joint project between the Research and Development Centre for Geotechnology (RDCG) and the Geological Survey of Japan (GSJ) under the ITIT programme 'Research on Mineral Resources Assessment of Oceanic Plate Fragments'. The authors thank Drs A. J. Barber and C. D. Parkinson of Royal Holloway, University of London for their critical review, effective suggestions and discussion of the geology of this area. We also express thanks to Dr Ir. S. Suparka, director of RDCG for his helpful support during our geological survey. We are grateful to Dr R. Sukamto of the Geological Research and Development Centre for his kind offer of unpublished data and information on the field area. Thanks are also extended to Drs K. Mori, T. Nakamori and Y. Iryu of Tohoku University and Dr Y. Sato of the Geological Survey of Japan for the identification of fossils in limestone clasts of the melange of the Bantimala Complex.
References ASlKIN, S. 1974. The geological evolution of central Java and vicinity in the light of the new global tectonics. PhD Thesis, Bandung Institute of Technology [in Indonesian with English abstract]. BARBER, A. J., TJOKROSAPOETRO,S. & CHARLTON,T. R. 1986 Mud volcanoes, shale diapirs, wrench faults, and melanges in accretionary complexes, eastern Indonesia. AAPG Bulletin, 70, 1729-1741. BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. & GARRARD, R. J. 1995. Tertiary tectonic and magnetic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for a Miocene continent-continent collision. This volume. BERRY,R. F. & GRADY,A. E. 1987. Mesoscopic structures produced by Plio-Pleistocene wrench faulting in South Sulawesi, Indonesia. Journal of Structural Geology, 9, 563-571. CLENNELL,B. 1991. The origin and tectonic significance of melanges in Eastern Sabah, Malaysia. Journal of Southeast Asian Earth Sciences, 6, 407-429. COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A., GURITNO, N., ROBINSON, N. M. & TALBOT, J. P. 1993. Tectonic and stratigraphic evolution of the Kalosi PSC area and associated development of a Tertiary petroleum system, south Sulawesi, Indonesia. In: Proceedings Indonesian Petroleum Association 22nd Annual Convention. 678-706. HAILE, N. S., BARBER, A. J. & CARTER, D. J. 1979. Mesozoic cherts on crystalline schist in Sulawesi and Timor. Journal of the Geological Society, London, 136, 65-70. HAMILTON,W. 1979. Tectonics of the Indonesian region. US Geological Survey, Professional Paper, 1078.
HASAN, K. 1990. The Upper Cretaceous Flysch Succession of the Balangbaru Formation, Southwest Sulawesi, Indonesia. PhD Thesis, University of London. -1991. The Upper Cretaceous Flysch succession of the Balangbaru Formation, Southwest Sulawesi. In: Proceedings Indonesian Petroleum Association 20th Annual Convention. 183-208. HEHUWAT, E H. A. 1986. An overview of some Indonesian melange complexes - a contribution to the geology of melange. Memoir of Geological Society of China, 7, 283-300. HUTCHISON, C. S. 1989. Geological Evolution of SouthEast Asia. Oxford University Press, Oxford Monographs on Geology and Geophysics, 13. LASH, G. G. 1987. Diverse melanges of an ancient subduction complex. Geology, 15, 652-655. MARUYAMA,S., LIU, J. G. & SENO, T. 1989 Mesozoic and Cenozoic Evolution of Asia. In: BEN-AVRAHAM,Z. (ed.) The Evolution of the Pacific Ocean Margins. Clarendon Press, Oxford, Oxford Monographs on Geology and Geophysics, 75-99. NUR, A. & BEN-AVRAHAM,Z. 1983. Break-up and accretion tectonics. In: HASHIMOTO,M. & UYEDA, S. (eds) Accretion Tectonics in the Circum-Pacific Regions. Terrapub, Tokyo, 3-18. PARKINSON, C. D. 1991. The Petrology, Structure and Geologic History of the Metamorphic Rocks of Central Sulawesi, Indonesia. PhD Thesis, University of London. SCHAAF, A. 1984. Les radiolaires du Cr~tacd Infdrieur et Moyen: Biologie et Systematique. Sciences G6ologiques Memoire, 75.
364
K. WAKITA ET AL.
SIMANDJUNTAK,T. O. 1990. Sedimentology and Tectonics of the Collision Complex in the East Arm of Sulawesi, Indonesia. Geology of Indonesia, 13, 1-35.
SUKAMTO, R. 1975. Geologic Map of Indonesia, Ujung Pandang Sheet. 1:1,000,000. Geological Survey of Indonesia. - 1978. The structure of Sulawesi in the light of plate tectonics. In: Proceedings of Regional Conference on Geology and Mineral Resources of SE Asia. 121-141. -1982. The geology of the Pangkajene and Western part of Watampone, South Sulawesi, scale 1:250,000. Geological Research and Development Centre, Bandung. 1986. Tectonik Sulawesi Selatan dengan acuan khusus ciri-ciri himpunan batuan daerah Bantimala. Dissertation, ITB, Bandung.
8Z WESTERMANN, G. E. G. 1992. Indonesia and Papua New Guinea. In: WESTERMANN,G. E. G. (ed.) The Jurassic of the Circum-Pacific. Cambridge University Press, USA, 181-193. SURONO & SUKARNA,D. 1993. Geology of the Sanana Sheet, Maluku, 1:250,000. Geological Research and Development Centre, Bandung. WAKITA, K., MUNASRI ,~ BAMBANG, W. 1994a. Cretaceous radiolarians from the Luk-Ulo Melange Complex in the Karangsambung area, central Java, Indonesia. Journal of SE Asian Earth Sciences, 9, 29-43. --, MUNASRI,SOPAHELUWAKAN,J., ZULKAm~AaN,I. & MIYAZAKI, K. 1994b. Early Cretaceous tectonic events implied in the time-lag between the age of radiolarian chert and its metamorphic basement in Bantimala area, South Sulawesi, Indonesia. Island Arc, 3, 90-102.
The Tertiary evolution of South Sulawesi: a record in redeposited carbonates of the Tonasa Limestone Formation M O Y R A E. J. W I L S O N & D A N W. J. B O S E N C E
SE Asia Research Group, Department of Geology, Royal Holloway University of London, Egham TW20 OEX, UK Abstract: South Sulawesi, situated at the junction of three major plates and with an almost complete Tertiary sequence, is an ideal location in which to study syntectonic sedimentation. Redeposited carbonate facies of the lower/middle Eocene to middle Miocene Tonasa Limestone Formation in the Barru area prove to be reliable indicators of tectonic activity. South of the Barru area contemporaneouscarbonate sediments formed on a relatively stable shallow-waterplatform, known as the Tonasa Carbonate Platform. Redeposited carbonate facies and interbedded marls from the Barru area are described and interpreted in this study. The immaturity and provenance of clasts indicate that the redeposited facies were derived from the faulted northern margin of the Tonasa Carbonate Platform. A relay ramp between at least two major NW-SE trending faults is the inferred configuration of this margin. Three main phases of faulting are indicated by the redeposited facies: late Eocene to early Oligocene, middle Oligocene and early to middle Miocene. This is consistent with other outcrop and seismic data from the region and with the inferred plate tectonic situation during the Tertiary.
Sulawesi is located in an exceedingly complex tectonic region, where three major plates have been interacting since the Mesozoic. With reference to the hotspot frame the Pacific-Philippine plate is moving WNW, the Indo-Australian plate NNE and both are colliding with the relatively stable Eurasian plate (Hamilton 1979; Daly et al. 1987, 1991). The convergence zone of this triple junction is a composite domain of micro-continental fragments, accretionary complexes, m61ange terrains, island arcs and ophiolites. Successive accretion from the east of oceanic and microcontinental material, and the associated development of island arcs, have all controlled the stratigraphic development of Sulawesi. South Sulawesi (Fig. 1), located on the eastern margin of Eurasia, has an almost complete stratigraphic sequence representing the period between the late Cretaceous and the present day (Fig. 2; Sukamto 1975; Hamilton 1979; Van Leeuwen 1981). South Sulawesi is therefore an ideal location in which to study the effects of local or regional tectonics preserved within a sedimentary sequence. Carbonate deposits of the middle Eocene to middle Miocene Tonasa Limestone Formation comprise a major part of the Tertiary succession in the western part of South Sulawesi (Fig. 2). The aim of this paper is to document platform, slope and deep water carbonate lithologies in the vicinity of Barru (Fig. 1), which preserve evidence of local, contemporaneous tectonic activity. Sulawesi is formed of distinct north-south
trending tectonic provinces (Sukamto 1975). In the west, the north and south arms of Sulawesi are composed of thick Tertiary sedimentary and volcanic sequences overlying pre-Tertiary basement complexes (Sukamto 1975; Van Leeuwen 1981). The Barru area lies within this western arc or province (Sukamto 1975; Hamilton 1979). Central Sulawesi is composed of sheared metamorphic lithologies and in the east a highly tectonized melange complex is present (Sukamto 1975; Hamilton 1979; Parkinson 1991). The eastern periphery of this melange has been overthrust by a dismembered and imbricated ophiolite sequence (Sukamto 1975; Silver et al. 1978; Simandjuntak 1990; Parkinson 1991). Emplacement of this ophiolite and resulting formation of the melange occurred during the middle Oligocene (Parkinson 1991). The microcontinental fragments of the Buton-Tukang Besi Block and Banggai-Sula are thought to have collided with the eastern part of Sulawesi during the early-middle Miocene (Fortuin et al. 1990; Davidson 1991; Smith & Silver 1991) and late Miocene-early Pliocene respectively (Garrard et al. 1988; Smith & Silver 1991). The Tertiary stratigraphy of western Sulawesi is comparable with many of the Tertiary basins in neighbouring east Kalimantan. West Sulawesi, the East Java Sea and east Kalimantan are thought to have comprised a widespread basinal area, the formation of which commenced during the early-middle Eocene (Van de Weerd & Armin 1992).
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 365-389.
365
366
M.E.J. WILSON 8~ D. W. J. BOSENCE
Geology and stratigraphy of South Sulawesi South Sulawesi is structurally separated from the rest of the western arc of Sulawesi by a NW-SE trending depression which passes through the Sengkang Basin (Fig. 1; Van Leeuwen 1981). Geologically and geomorphologically South Sulawesi is divided by a present day N-S trending depression known as the Walanae Depression (Fig. 1). The Walanae Depression has been described as a major left-lateral strike-slip zone (Sukamto 1975; Van Leeuwen 1981). Seismic (Grainge & Davies 1983) and present-day outcrop constraints suggest significant normal displacement on basin-bounding faults occurred during the Tertiary. The Barru area of this study includes the northernmost outcrops of the Tonasa Limestone Formation and is located on the southern margin of the depression passing through the Sengkang Basin, to the west of the Walanae Depression (Fig. 1).
"
::!i:l -
7
o (
....
t
T°nasallctdil \
"1
.,
:::::
:~:~::.
.-.-.
/
.
:: :: :::.
':::~:5,-~::::~:IV~iiiiill~ i ~ .7 :I
l-".11r:: :~:3:.~:~ iiii{iiiiiiiiii7 ~~ i I:: i::: ~ '".::.~iSl!]! " '~i!!!i!!!i!i~
UpJUI~G
l
KEY
Alluvium LompobatangVolcanics Walanae Fro. Tacipi/SalayerLimestone Camba& othervolcanicFrns, Intrusives
I
I
TONASA LIMESTONE Fm.
~ ~ ~ U ~ ~
KalamisengFm. SaloKalupangFm. MalawaFro. LangiVolcanics 13alangbaruFro. BasementComplex
Fig. 1. Geological map of the South Sulawesi, showing the location of the Barru area. Modified after Sukamto (1982) and Sukamto & Supriatna (1982).
Two inliers of the pre-upper Cretaceous basement complex of Sulawesi are exposed west of the Walanae Depression (Fig. 1). They comprise tectonic slices of metamorphic, ultrabasic and sedimentary lithologies (Hamilton 1979; Sukamto 1982). Deep marine clastics of the upper Cretaceous Balangbaru and laterally equivalent Marada Formation overlie the basement complexes unconformably (Fig. 2; Van Leeuwen 1981; Sukamto 1982; Hasan 1991). Palaeocene-Eocene volcanics of the Langi Formation and marginal marine siliciclastics, shales and coals of the Eocene Malawa Formation overlie with an angular unconformity the Balangbaru Formation in the eastern and western parts of west South Sulawesi respectively (Sukamto 1982). The upper part of the Malawa Formation interdigitates with shallow marine carbonates of the middle Eocene to middle Miocene Tonasa Limestone Formation to form a transgressive sequence. Deposition of at least 400 m of shallow marine carbonates occurred in the area between Maros and Tonasa II (Fig. 1; Garrard et al. 1989; Crotty & Engelhardt 1993). These sediments of the Tonasa Limestone Formation are thought to have formed on a relatively stable, gently subsiding, large-scale platform c. 80 km across (personal observation; Garrard et al. 1989), named here as the Tonasa Carbonate Platform. An intra-formational midOligocene unconformity occurs within shallowwater carbonates of the Tonasa Limestone Formation in the eastern Biru area (Figs 1 and 2, Van Leeuwen 1981). From the early Miocene onwards in the Biru area there was a deepening of the environment, and carbonate sedimentation was strongly influenced by local tectonism (Van Leeuwen 1981). This study describes and interprets deep marine late Eocene to mid-Miocene carbonates of the Tonasa Limestone Formation, from the Barru area to the north of the Tonasa Carbonate Platform. During the middle to late Miocene carbonate production was terminated by the influx of volcaniclastic deposits of the Camba Formation. These volcaniclastics were derived from a N-S trending volcanic arc which developed in South Sulawesi (Sukamto 1982; Yuwono et al. 1985). East of the Walanae Depression, lithologies are quite distinct from those to the west and the oldest lithologies are of Eocene age (Figs 1 and 2; Sukamto 1975). The lithologies are dominated by volcanics and volcaniclastics of the Salo Kalupang, Kalamiseng and Camba Formations (Sukamto 1982; Yuwono et al. 1985). Eocene shallow-marine carbonates of the Tonasa Limestone Formation outcrop only as a fault-bounded sliver at the eastern margin of the Walanae Depression (Fig. 1; Sukamto 1982).
TERTIARY EVOLUTION OF S SULAWESI
WESTERN SOUTHSULAWESI W Formationnames. Uthologies E and thicknesses '" " ~ v%. UPPERCAMBA .%%Y, Late ~v~ FORMATION& . % ~ , Miocene 1- az~,~OTHER VOLCANICS . % ~ ,
EASTERNSOUTHSULAWESI W Formation names, lilhologies E and thicknesses ~ TACIP~IW?~.LANAE" Fro. I T .........
Midale Mk)cene _ _
%'.'.'~':" FORMATION .~:%..-:~:
Age
v~
....
I
1 1
C~te
~
~.~ - : - ~ ~=-~-~-, LOWER . ~ CAMBA Fm'
~
~"-'~ ~":-:~
~:£i~
I!
LIMESTONE Iq:;~ FORMATION~ = ~
~1=1;;;~1~~
-3oom-
~...-.-~ ~ . _ . ~ ; ; . . . ,.,_,_,
,.,~
I
____
=====================
i
Z.:.':E~>:'> > > >:+',! :.--..---~,E;.I.I.I.I.I.:.I.bN
Oligecer
...,,
Late Eocene
~
TOI LSA
~.*~
Middle
Lady Eocene Key to lithologies: ~ Shallow marine limestone Redeposited carbonate facies
PaleeocqLate Eally
C ..... Late
Basinal marls (Marginal) marina clastics '] Deep marine clastics & ~. F ORMA_TIO_N-.~ FORMATION,, r ~ volcaniclastics •~ Lavas & volcaniclastics
Middle
" ~ - ~ / 1 3 A S E M E N T ~>,',>,~
Palaeoc~
=
~
....
.~,~-~"~ D
~ ~
Deep marine clastics Metamorphic and ultrabasic
Fig. 2. Stratigraphic correlation chart of the units to the west and east of the Walanae Depression in South Sulawesi. Based on Van Leeuwen (1981), Sukamto (1982), Sukamto & Supriatna (1982) and Grainge & Davies (1983).
Structure and stratigraphy of the Tonasa Limestone Formation in the Barru area The Tonasa Limestone Formation in the Barru area is bounded to the south by pre-Tertiary metamorphic and ultrabasic basement lithologies of the Bantimala Block (Fig. 3; Berry & Grady 1987). A smaller inlier (8 km across) of metamorphic basement lithologies, known as the Barru Block (Berry & Grady 1987), is located to the southeast of Barru (Fig. 1). The overall structure of these basement complexes is of relatively rigid blocks tilted to the east and bounded on the remaining sides by eastward dipping thrusts or sinistral wrench faults (Sukamto 1982; Berry & Grady 1989). A number of generally NW-SE trending faults also cut the Tertiary sequence in the Barru area (Fig. 3). The eastern and southeastern flanks of both basement blocks are unconformably overlain by an almost continuous stratigraphic sequence from the Balangbaru Formation through to the Camba Formation. Tertiary angular unconformities occur at the base of the Malawa Formation and in some
367
localities between the Tonasa Limestone and Camba Formations. The Tertiary lithologies dip eastwards at 10-25 ° . North of the Bantimala Block the Tertiary lithologies are folded into a WSW-verging, regional-scale N N W - S S E trending anticline, named here the Rala anticline (Fig. 3). On the northeast limb of the Rala anticline there is a complete stratigraphic sequence from the Malawa Formation through to the Camba Formation, with no apparent unconformities. On the southwestern limb of this anticline Tertiary lithologies dip 20--40 ° to the WSW and the carbonate sequence is considerably condensed or absent. An angular unconformity separates the Camba Formation from the older underlying lithologies on this western limb. Igneous bodies composed of diorite-granodiorite and trachyte (Sukamto 1982) intrude the Tonasa Limestone Formation and older lithologies in a number of localities (Fig. 3). Although the age of the intrusives in the Barru area is not known, similar lithologies from other areas have been dated using K-Ar techniques as middle to upper Miocene (Sukamto 1982). The Tonasa Limestone Formation in the Barru area reaches a maximum thickness of 1100 m in the Rala section (Fig. 4). The basal few metres of the carbonate sequence are interbedded with siliciclastics of the Malawa Formation. The earliest carbonate sediments are composed of wackestones, packstones and floatstones (Fig. 4). Some beds contain a rather limited fauna of miliolids, gastropods and occasional pseudomorphs after gypsum or anhydrite. Other facies contain large alveolinids or broken branching corals as well as miliolids. These lower to middle Eocene facies (T a, T. Wonders 1993 pers. comm.) indicate a normal marine back reef-barrier environment with minor restriction. Usually the initial carbonate sediments pass upwards into thick successions of metre-scale bedded packstones and rudstones composed of abundant large Nummulites and coralline algae. The coralline algae often encrust the large benthic foraminifera to form rhodoliths, suggesting shallow marine, relatively turbulent conditions prevailed (Fig. 5a). In the Rala section only, a subsequent deepening of the environment is indicated by marls and wackestones containing abundant, monospecific, large, flat Discocyclina typical of the lower limits of the photic zone (T. Wonders 1993 pers. comm.). Most areas contain decimetre-scale planar-bedded packages of bioclastic packstones, which include planktonic foraminifera and show occasional evidence of current or wave reworking (Fig. 4). This facies shows an overall fining-upward trend and is interpreted as outer shelf or slope deposits
368
M . E . J . WILSON (~ D. W. J. BOSENCE
-?
"
÷"i'.l,'r . i Igneous intrusion
,- B a r r u /
' " ' upper volcaniclastic
"
Bantimala
. . . .
~edeposited
~
. . . . Lower marine member (CF)
.- F a c i e s " : . . 2 ' ~ ," ",
;;;;gd 552
! ecent =,uv, um
,.. Overlap of'
~.=
,-..,-,.ov~nm~reaet=o~
~/ Marls + redeposited _ _ facies (TLF) Bioclastic . . . . packstones (TLF) ~ Shalk)w-water carbonat~ tacles (TLF}
•
", ", ", ", B l o c k ' , ", ", ", ", ", ", ", ",
iiiiiiii Malawa Formation .... •% - - - - "
~
r
U
Formation
" ' Basement complex
I "., "a
CF - Camba Fm. TLF - Ton,s, Limestone I
•
~ •
~
v
~
v
v
v
~
v
v
v
~
v
v
v
~
v
v
v
~
v
v
v
v
v
v
v v v
v
- - - - - ,
v
,.,
v
v
v
v
0
,,,
-.,
v
...
v
~,
v
v
v
~
1
-.,
2
,.,
,.,
3
4
I
km v
v
~
v
~
v
v
v
v
v
v
v
~
v
v
v
v
v
v
v
v
v
v
v
v
v
v
~
v
v
v
v
v
v
v
~
v
v
v
v
v
v
v
v
~
v
~
v
v
-.
~
v
v
v
v
v
v
~
v
~
v
v
v
v
v
~
v
v
v
B. ,,v vv ,,
250
~'~++~' ~ I v v
v
~----------....~
SL
-250
SW
~
bedding, with
Fault(tickmarks
downthrown block ~
~
Anticline Unconformity
v
Z
5~
direction of
angle of dip
~ ! v
Village
Dip and strike ~- 2 5
v ~. ,, ~, v
KEY •
V" Z Line of cross A ~ ~ B section
Lineof measured
,
section Rala anticUne +
Line oL~"j secti°ln Y'Y' .
.
I
Z' ~-~'/~~v
,,-
,,,
- 1 v
v
S~¢ E o~- o oo
if
v
~. v v ~. -~
t
NE
Fig. 3. Geological map of the Barru area with cross-sections (Z-Z', Y-Y'). Palaeocurrent data from the redeposited facies are shown as arrows and rose diagrams. The cross-sections are drawn to scale. The locations of the measured sections (A-B, C-D, E-F) in Fig. 6 and the cross-section (M-M') in Fig. 9 are shown. Measured sections are composite sections along stream cuttings and tracks.
TERTIARY EVOLUTION OF S SULAWESI Lithology
Age
Facies description
- - -(3oarsev~canic/astic.s of tl~e upper Camba Fm.
---Lower-mbmber-of-the-Camba
Fm.
1105120~0[:i . . . . . . . . . . .
- d e e p marine shales interbedded _ _ _willa yotc~ielastic_s
~
1001
.....
NN2NN6 e/m Miocene
~ . . . . '~ ................. .
804
.........
.
795743- ~ 715 649 . . . . . . . . . . . . . .
4"/0 .............
4 5 0 ~ '="d"d 400 --k~-,-,-,-~, 375 ~ 251 - ~ 2 0 7 ~
Redeposited carbonate facies interbedded with marls . __^ This section shows the r~-z~, trends in the redeposited m'°hgfacies in the Rala section. The number of beds and their thickness are not drawn to scale. See Fig. 19 for a more accurate representation. P17 I.Eoc
-i
-
~l:
. . . . . . . . . . . . . . . . . . . . . . . . .
N15 Decimetre-bedded bioclastic I.Eoc packstones ..........................
z ~
Discocyclina rich marls & wackestones ........................... Decimetre-bedded bioclastic
M
C M S f rn c G P C
E u. m z O LU
packstones
- - -Num-rhulit~&fi?J-c-o~M,66~fg-al bioclastic packstones & rudstones m'-Milio/icl-& a/veo/inid . . . . . . . . "-. wackestones & packstones .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Fig. 4. Simplified composite stratigraphic section of the Tonasa Limestone Formation in the Rala measured section. Location of section C-D is shown on Fig. 3.
with a upper Eocene age (N15, T. Wonders 1993 pers. comm.). In the northern and eastern parts of the Barru area (Fig. 3) these packstones are overlain by a thick sequence of upper Eocene to middle Miocene marls interbedded with redeposited carbonate facies (Fig. 5b), which are the main focus of this paper.
Redeposited carbonate and marl facies of the Barru area Marls
This easily weathered facies is interbedded with coarser, cemented carbonate facies, and is especially abundant in northern and eastern parts of the Barru area. The percentage thickness of the marls in the Tonasa Limestone Formation decreases from the Bangabangae section (E-F; 69% thickness) moving southwards towards the Bantimala Block (Figs 3 & 6). Marls are best exposed in stream cuttings, especially in the Bangabangae section, and although poorly exposed
369
in intervening areas this facies is considered to be laterally continuous between sections. The marls are pale green-grey in colour, poorly cemented and often fissile. Planar lamination on a millimetre scale is a common feature of the marls. In some localities the marls appear more homogeneous, and water-worn surfaces reveal mottling caused by the presence of randomly orientated burrows on a millimetre to centimetre scale. A pelagic biota, including planktonic foraminifera and nannofossils is well preserved and occurs in abundance in this facies. The fine grained laminated nature of the marls indicates that this facies was deposited in a low energy environment largely from suspension. The presence of abundant, well preserved tests of pelagic organisms indicates deposition in a deep marine basinal setting above the carbonate compensation depth. Burrowing in some localities suggests an aerobic environment.
Bioclastic packstone facies
This facies is seen in the upper parts of most sections in the Barru area interbedded with marls and other coarser carbonate facies. The bioclastic packstone facies occurs most abundantly in the Rala section (13.8% thickness of total deeper water facies) and is best exposed in stream sections at this locality (Figs 3 & 6). Although only well exposed in stream cuttings, beds of this facies are thought to be laterally extensive on a scale of tens of metres. In this facies beds are planar-bedded with bed thicknesses varying from 10-90 cm. Bed contacts are usually planar and sharp. The grain size varies from fine to coarse sand grade between adjacent beds. Sedimentary structures were not observed and burrows with millimetre to centimetre diameters are common in this facies. Although delicate tests of planktonic foraminifera are frequently well preserved, more robust fossils such as coralline algae, large and small benthic foraminifera and echinoids are common as fragments. Based on 'bundling' of beds and grain composition, this facies can be subdivided into two. (a) Packages of medium bedded packstone which generally lack intervening marly units were found in only one locality in the Rala section (see below, Fig. 12). Faunal elements contained in a single bed are consistently of one age. This subfacies is very similar to the bedded packstones described above, which underlie the marls (Figs 3 & 4). (b) Single packstone beds interbedded with marls commonly contain both grains which are contemporaneous with and older than the intervening marls. This subfacies often includes
370
M. E. J. WILSON & D. W. J. BOSENCE
Fig. 5. Carbonate facies of the Tonasa Limestone Formation from the (a) lower and (b) upper parts of the Rala Section. (a) Shallow-marine Nummulites and coralline algal rudstone. The coralline algae encrusts the large benthic foraminifera forming rhodoliths. Scale is in centimetres. (b) Typical outcrop of marls interbedded with redeposited carbonate facies. The lower resistant, bioclastic packstone bed is separated from a clast-supported breccia unit by rather easily weathered green-grey marls. Vertical field of view is 1.5 m.
sand grade, angular clasts from all the formations underlying the Tonasa Limestone Formation. A redeposited origin is inferred for this facies because it is interbedded with basinal marls and contains fragmented shallow-water bioclasts. Sedimentary structures which might indicate the mode of reworking are absent. The packages of bioclastic packstones may perhaps have a similar origin to the outer shelf-slope packstones underlying the marls (Fig. 4). In terms of clast content the single beds of bioclastic packstones are comparable with the graded bioclastic packstone facies described below. Non-graded packstone beds are often documented in sequences containing abundant graded packstones and have been interpreted as grain flow or modified grain flow deposits (Lowe 1976; Cook & Mullins 1983; Gawthorpe 1986; Eberli 1987). However, criteria indicating grain flow deposits such as dish structures, inverse grading, diffuse lamination and outsized clasts (Middleton & Hampton 1976; Cook & Mullins 1983) have not been identified in this facies.
Graded bioclastic pack-grainstone facies This facies is well exposed and common throughout the upper part of most sections in the Barru area. Graded, bioclastic pack-grainstone facies
are interbedded with marls. Coarser breccia facies are almost invariably overlain by this facies (see below). The percentage thickness of this facies decreases northwards away from the northern margins of the basement blocks (Fig. 6). Beds of this facies are often laterally continuous on a scale of tens of metres, although a 10% decrease in bed thickness over 5 m has been observed. Bed thicknesses of this facies vary from 5 cm up to 110 cm (Fig. 7). The basal bed contacts are sharp and often planar. Less commonly this facies has an erosive base and exhibits rare groove or flute casts (see below for palaeocurrent data). Beds are graded both in terms of grain size and composition; fining upwards from pebble-coarse sand grade to fine sand-silt grade (Fig. 8a). Heavier gravel or sand grade schist, ultramafic or sandstone clasts tend to be concentrated in the lower parts of beds. Rare water escape structures and imbrication of planar clasts are present. Planar lamination on a millimetre scale is a common sedimentary structure in the upper part of beds (Fig. 8a). The upper bed contact may be planar or slightly undulose and sharp or gradational into marls. This facies incorporates a large variety of often fragmented bioclasts, including large and small benthic foraminifera, coralline algae, echinoderm and very rare coral debris. It contains up to 25% well preserved planktonic foraminifera, especially in the upper finer part of beds. Lithic clasts include
TERTIARY EVOLUTIONOF S SULAWESI
N
,~ Bangabangae section
F m i"~"~"~"~"~'~ 775 ~'1
10km
~
Upper coarse volcaniclastic member of the Camba Fm.
371
5km Rala section
Dm! ~,,~,,~,-~
Lower member of the Camba Fro. - Deep madne shales interbeclded with volcaniclastics
1200
['l
~4n= I.'.~
..,,,-'-'------------
S
.~ B m
Doi-doi section
520. 425 ................
.°_ . . . . - . . . . -~- . . . . -~- . . . . _---
1
~~~-~'""~-~" "~"~
488 ~ ' " - ' " ' -
Redeposited carbonate facies interbeddedwith marls of the Tonasa Limestone Fm.
~ ~
"~ ,o
~_ . . . . .
-t::::!
78 ~ ~ - - ......
E ~ MS f
m c
G PC
aliow-wa
ei
acl;soit
-;-
ona
Total thickness of made & redeposited facies Total number of redeposited beds Clast-supported breccia Graded bioclastic packstones topping the above Marl supported breccia Graded bioclastic packstone Bioclastic packstone Planktonic foraminifera wacke/packstone
Mad
C I I~ s f' m c G P C:
I"'::::::::.Marl '-
: : : : : ~
Rala section 635 rn 173 39.64 11.85 0.40 12.63 13.77 1.10 20.22 0.40
8.82 6.17 7.82 69.02 6.76
% % of limestone conglomerates, packstones and marls 100 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
........
, , , , MSfmcGPC
.........
Ban~labangae section 408 m 76 0.93 0.47
Other
,,'"
/' '
,,'A
,,
.....
Limestone Formation
207,
....
.............
Doi doi section 218 m 66 58.51 20.14 4.35 7.26 5.62 4.13
~. . . . . . . . . . . . . . . . . . . .
-_.,~.v.y.-......,..d..~~. ¢" ¢'" "'~¢"
"
r~ ~
"
0 % % Amalgamation of redeposited carbonate beds 100 5O 0 DISTAL ~ '
PROXIMAL
BASlNWARD Fig. 6. Three composite stratigraphic sections through the redeposited facies interbedded with marls showing northwards proximal to distal and basinward trends. See Fig. 3 for location of the stratigraphic sections (A-B, C-D, E-F).
metamorphic (see below) and quartzose sandstone lithologies, a wide range of limestone facies and marl clasts. Both packstone and grainstone textures occur, although in the majority of beds the matrix is marly or composed of finely fragmented bioclasts. Trace fossils on a millimetre to centimetre scale, including Taphrhelminthopsis are common at bed contacts and may be parallel or perpendicular to bedding. Silica replacement is frequent within this facies and occurs in three main forms: as chalcedony within individual bioclasts or litho-
clasts, as nodules and as irregular but apparently continuous 'beds'. Silica nodules are parallel to bedding, up to 15 cm thick and 5 0 c m across and sometimes follow centimetre diameter burrows. Because this facies is interbedded with basinal marls and contains fragmented lithoclasts and shallow marine bioclasts, it is considered to be redeposited. Sedimentary structures such as an erosive base, normal grading and parallel lamination are typical of calciturbidites (Crevello &
372
M.E.J.
W I L S O N ( ~ D, W, J, B O S E N C E
Graded bioclastic pacidgrainstone facies
Planktonic foraminifera wacke/packstone facies
6) U p p e r bed contact is planar or slightly
iiiiiiiiil l----
l:i::i::i::3 ,~~ , ~
'k 5) Millimetre scale parallel ~ laminations %.4) Rare water escape structures
T 8
5) Planar sharp or transitional contact into marls
overlyingUndul°se iand n sharp t O or m gradational a r l s
'i~i~i~i~i~i~i~ '..~.~.~,.~.~.~.~.~.~.~.~. i~i~i~i~i~i~i~'i~i~ .<~<~<~<.<~<~<~<~<~<~ -i~ ~-~ ~'~"l~'~'~"l~"l~'~ <~~<~< ~<~<~ ( ~ < ~ < . < ~ < ~ ,.~-~.~.~.~.~.-~.~.-i~.~.~.~-~ ..~ •{'~•~ •~ •~ •~. ~; ~.'.~,'-~,'-~
~ ~',,'~ imbrication .~'1~-~'1 ,~,.,~,~,~;,:,~ ,, sharp o . a n . '¢t:< . . . t . : . < . . E . . ¢ . . E . . d . . ~ . . ~ . . ( ~ E ( ~ ~ ~._ p lanar base to bed ~ |......~... ..... •-'"'""-":~ with rare groove or ............."1 f's gs (~ flute casts. .,~.~.~-~-i~.
Clast-supported breccia facies topped by a graded bioclastic pack/grainstone facies 7) A graded packstone facies '<~.~!~-~
'~<'~,.~ '~
(see above) almost invariably overlies the coarse limestone conglomerate
'i~i~
~ ~ ~%%~ ~%~s=~%~,
3) Normally graded packstone fining upward from g r a v e l / c o a r s e s a n d to • fine sand/slit,
m
6) Relatively sharp contact
?~!6!:'o!~:~9'.::~!~:'~:~:~
4) Clast s i z e varies from 4 m across down to fine sand
i~:~:i~'~:~)~),~:~ !~.~:~/:?~;~i:~!~::~ i~ :~:~fi~ grade !i~;'.~i~:~,:~i~'~<~-~ :~i~i~ii~ l 3) C o m m o n circum-clast ~~:~_:~9.i~i~I~:~~:~_:~(~. ~ stylolitesa n d rare clast i~.:.~;~':-~.~'!~.:~,'~ ~ ~l imbrication. ~.~b~£)~!6~:2:~ ,~§~'/~i6-.:~?0?~b:~!:~
!] 2) Coarsening upward at I base m a y occur. 1 ~ ~narp, planar or irregular base
M
fs
cs
G
3) Well preserved planktonic foraminifera comprise 7 0 - 9 0
% of the grain types.
~;%%%%%~ ~ ~ ' ~ ~_~.~.~-_.~ ~_~_~_~ -~-~-~ ~-~
I
2) Normal grading from m e d i u m sand ~lrade to silt grade is occas,onally seen. Other beds appear ungraded
,~=~-~;~]~ ~-~L~.I M
~ii iilil
I silt
i fs
i ms
1 ) Planar sharp or transitional basal contact i cs
Marl-supported breccia facies topped by a graded bioclastic pacWgrainstone facies
~-,,~-,,~
5) Sharply overlain by graded
~;~;~:~:~x packstone facies (see above left). ~i~2 "~!'-'!~!"!'-'!~'-3.::: =....... .d
~.--.~.-.-~"~"~ "'"~7 T"'I.~.,'[ ::'"',~.,..'L.'._"'_" _ "'~
~ ¢ 1 .... ~ ";~'~:~:~:~i~ .o. ¢1~. ~ m. . 5). Shght . ~i;~: fining upward of ~(~,~:;~;: °.'~:?.o:Z :~.~.~ angular clast-supported :;:.~.~{6.,~.,'~.~:U~~:~!~; ~ :$ breccia.
::_~DS;;(S:m;:~;~".~:~!:~
4) Grain- or matrix-supported in a marly matrix
4) Chaotic poorly 1:%1 sorted facies which
,1:.":.1
~"~/~~)!~!,~)(.~(.~t ~r~t..~.~.~(.~,l t ~ ~ /~...:~::.~
~ - - : : : : I
~ - ' - J ---7-/~ "N-..7 7" ~.-I
lacks any signs of
packst0ne channel, which includes cross
lamination and clast imbrication.
=~Sub-angular
~:'-'1 clasts up tO 1 m ~'-": I across mostly J':':l supported by a
~
1 ~) Sa~a~gr~t~tctdulOse M
fs
cs
G
C
C
Fig. 7. Schematic diagrams showing the main features of four of the redeposited facies. The bioclastic packstone facies are not shown as they are planar bedded and lack other visible sedimentary structures. Diagram not to scale.
Schlager 1980; Cook & Mullins 1983; McIlreath & James 1984). A complete Ta__e Bouma sequence (Bouma 1962) characteristic of siliciclastic turbidites is not seen in this facies. Incomplete Bouma cycles are a common feature of other modern (Schlager & Chermak 1979) and ancient redeposited carbonates interpreted as calciturbidites (Pfeil & Read 1980; Mcllreath & James 1984; Gawthorpe 1986; Eberli 1987; Braithwaite & Heath 1992; Herbig & Bender 1992). Since skeletal carbonate grains are ofter~ porous or irregular in shape, grain size distribution and sedimentary structures in calciturbidites are commonly less regular than in siliciclastic turbidites (Tucker & Wright 1990). The radial chalcedony preserving 'ghosts' of earlier bioclasts and the smooth irregular shape of nodules both indicate the secondary replacement nature of the silica in this facies. Probable sources
of the silica are sponge spicules and microfossils, present in both the marls and packstones, and lithoclasts such as quartzose sandstones present in the redeposited carbonate facies. The Tonasa Limestone Formation is similar to a number of other ancient carbonate deposits in which it is reported that redeposited facies are the only carbonate facies in which silica replacement is common (Gawthorpe 1986; Bustillo & Ruiz-Ortiz 1987; Coniglio 1987; Eberli 1987; Reijmer & Everaars 1991; Herbig & Bender 1992). Silicification is thought to be prevalent only in the redeposited facies because the rapid burial of transported siliceous tests and clasts prevents the dissolving silica passing into the overlying seawater (Bustillo & Ruiz-Ortiz 1987). Certain horizons or burrows perhaps had a higher original porosity and the silica has preferentially remobilized to these areas during diagenesis.
TERTIARY EVOLUTION OF S SULAWESI
Clast-supported breccia facies This sedimentary facies is interbedded with marls and bioclastic packstones and occurs particularly abundantly in localities close to the northern margins of the basement blocks (Fig. 3). In the Doi-doi section (A-B) this facies comprises 58.5% of the thickness of the deeper water carbonate facies (Fig. 6). Beds are laterally continuous on a scale of tens to hundreds of metres. This facies is almost invariably capped by graded bioclastic pack-grainstone facies (described above) and the contact may be sharp or transitional over about 10 cm. Usually the sequence of clast-supported breccia topped by a graded bioclastic pack-grainstone facies is between 70 cm and 5 m thick, though units up to 25 m thick do occur (Fig. 7). The basal contact of this breccia facies is usually sharp and may be planar or more irregular. In rare cases, this facies exhibits a transitional basal contact with marls, in
373
which up to 30 cm of the lowermost part of the bed contains up to 25% of the green-grey marl as a matrix. The lower part of the sequence is a coarse, clast-supported breccia with angular-sub-angular clasts usually up to 60 cm across. In thicker units the maximum clast size seen is 4 m across. The bioclastic packstone matrix of this facies is rarely seen due to circum-clast stylolites which give a fitted fabric to the breccia. Locally, this facies may coarsen upward over the lower 40 cm (Fig. 7). More usually there is a slight fining upward of clast size through the breccia facies. Rare imbrication of planar clasts does occur. Replacement or partial replacement of clasts by silica is sometimes seen within this facies (see above). A wide variety of lithic clast types is present in this facies (Fig. 8c). These include metamorphic, igneous (see below) and quartzose sandstones lithologies, and a range of shallow-water wackestone, packstone and grainstone and deeper-water marl clasts. Matrix comprises a relatively low percentage (2-8%) by
IIO !iii~iiiiiiiiiii~ii!i~ ~ ~i~ ¸¸~¸~ ...... ,i~ii!i ¸~ii~i ~...........~..............~~i!i~~ % ~
Fig. 8. Photographs of the redeposited carbonate facies. (a) Graded bioclastic packstone bed, showing fining upward and parallel lamination at the top of the bed. Darker grains are formed of older non-carbonate lithologies. Scale is in centimetres. (b) Amalgamation of clast-supported breccias topped by graded bioclastic packstone unit. The lower coarse, angular clast-supported breccia is sharply overlain by a graded bioclastic packstone unit. Not that the parallel lamination in the graded bioclastic packstone becomes closer together as the unit fines upwards. The bioclastic packstone is sharply overlain by another clast-supported breccia layer. (c) Outcrop of angular clast-supported breccia showing the variety of clast types present within the redeposited carbonate facies. The dark clasts are schists and ultrabasics and were derived from the older formations in South Sulawesi. Clasts from a range of shallow-water carbonate facies and small rip-up clasts of basinal marls occur in abundance. Scale is in centimetres. (d) Photomicrograph of a bed of planktonic foraminifera wacke-packstone. Scale bar is 1 mm across.
374
M . E . J . WILSON t~ D. W. J. BOSENCE
volume of the breccia layer and is composed of marl or finely fragmented clasts or shallow-water biota. Although this sequence of units is interbedded with marls, amalgamation of both the breccia and overlying graded packstone facies frequently occurs, especially adjacent to the northern margins of basement blocks (Fig. 8b). Sometimes a coarse breccia directly and sharply overlies another breccia and the packstone facies is missing. This facies is redeposited because it is interbedded vdith basinal marls and contains a variety of coarse lithoclasts and fragmented shallow-water biota. Clast-support, poor sorting, occasional inverse grading and clast imbrication suggest transport and deposition from a lower traction layer (Lowe 1982; Pickering et al. 1986; Tucker & Wright 1990). Circumclast dissolution has removed much of the matrix (cf. Davies 1977). Debris flows, which occasionally show inverse grading at the base, can transport coarse material with as little as 5% matrix (Rodine & Johnson 1976; Picketing et al. 1986). However, normal grading of clasts towards the top of breccia units argues against a debris flow as the mode of origin. A high density turbidity current with a lower inversely graded R 2 layer and upper normally-graded R 3 layer (cf. Lowe 1982; Tucker & Wright 1990) is inferred for the transport and depositional origin of this facies. After deposition of the coarser breccia material, the graded bioclastic packstone capping this facies is thought to have been deposited out of suspension from turbulent flow. An analogous experimental example of a high density turbidite topped by a lower density turbidite has been described by Postma et al. (1988). Many examples of two layer carbonate sediment gravity flows have been described from ancient (Cook & Taylor 1977; Hubert et al. 1977; Cook 1979; Johns et al. 1981; Kepper 1981; Hiscott & James 1985; Gawthorpe 1986; Eberli 1987; Watts 1987) and modern (Mullins & van Buren 1979; Schlager & Chermak 1979) settings. Turbidity flows are often entrained on top of coarse gravity flows and can be generated by reverse shear at the flow-water interface in the head region of the coarse flows (Hampton 1972). A genetic relationship is therefore suggested between the clast-supported breccia topped by the graded bioclastic pack-grainstone facies and the graded bioclastic pack-grainstone facies (see below). M a r l - s u p p o r t e d breccia f a c i e s
This facies occurs in only two localities in stream cuttings in the Rala section (Figs 6 & 12). This facies is laterally continuous on an outcrop scale (m). Graded bioclastic pack-grainstone facies
sharply overlie the marl-supported breccia facies in both outcrops. Bed thicknesses of this facies are up to 2 m, with sharp and undulose bed contacts (Fig. 7). In this poorly sorted facies sub-angular clasts up to 1 m in diameter are present (Fig. 7). The matrix is composed of marl, and both matrix and clastsupported fabrics occur, although the former support mechanism is more common. One bed contains a N-S orientated, flat-topped channel 60 cm deep and 2.5 m wide, composed of breccia fining upward to a grainstone. Internal sedimentary structures within this lensoid body include planar cross-lamination and a layer with abundant marl clasts, some of which show imbrication. A large variety of sub-angular limestone clasts, including wackestones, packstones and float-rudstones occur within this facies. Large rounded clasts of green-grey marls and clasts of graded bioclastic packstone and clast-supported limestone breccia occur.
The massive, chaotic, poorly sorted, nature of this facies and presence of large sub-angular clasts floating in a marly matrix are features typical of debris flows (Cook et aI. 1972; Middleton & Hampton 1976; Lowe 1979, 1982; Cook & Mullins 1983). In debris flows the cohesive strength of a fine sediment-water matrix is sufficient to support and transport large clasts on slopes as low as 1° (Cook et al. 1972; Prior & Coleman 1982; Hiscott & James 1985). The range of clast types indicates that both shallow and deeper water carbonate lithologies were incorporated into this facies. Since clasts of reworked breccia or packstones were not seen in other redeposited facies, a deeper water origin for this facies is suggested. Modern mudsupported and clast-supported debris flows occur in proximal and distal parts of the lower slope of the Bahamas respectively (Mullins et al. 1984). Graded bioclastic packstones capping debris flow deposits have been described from other modern and ancient deposits (Crevello & Schlager 1980; Mullins & Cook 1986). Channels containing graded lenses of carbonate sands within debris flows are less commonly reported (Braithwaite & Heath 1992). It has been suggested that turbulence may have played an important role in supporting grains at the top of flows (Cook et al. 1972; Hampton 1972). The channelized normally graded body within this facies perhaps indicates amalgamation of the breccia.
Planktonic foraminifera wackepackstone facies
Although this facies occurs in the upper part of all sections in the Barru area, it is best exposed and
TERTIARY EVOLUTION OF S SULAWESI occurs most commonly in the Bangabangae section (E-F) where it comprises 7.8% of the thickness of the deeper water facies (Fig. 6). This facies outcrops in stream cuttings and is only known to be laterally continuous on an outcrop scale (m). Planar beds of this facies between 20 and 60 cm thick are interbedded with marls (Fig. 7). This facies differs from the marls in that it is cemented and contains a higher percentage of planktonic foraminifera and fragmented shallow-water biota. Upper and lower bed contacts are planar and may be transitional or sharp with interbedded marls. Normal grading, fining from medium to silt-size, is occasionally seen (Fig. 7). Well preserved planktonic foraminifera comprise 70-90% of the bioclasts present in a micritic or marly matrix (Fig. 8d). Fragmented echinoderms, benthic foraminifera and coralline algae form the remaining bioclasts. Silica nodules parallel to bedding are present within this facies. The normal grading sometimes seen in this facies suggests deposition by waning current flow. Reworking could also account for the increased concentration of planktonic foraminifera and presence of fragmented shallow-water bioclasts, relative to the intervening marls. Although sedimentary structures which could differentiate between storm, contour currents or mass flow processes as the mode of reworking are absent, the location of this facies relative to other redeposited facies suggests a mass flow depositional origin (see below).
Origin of the redeposited carbonate facies Redeposited carbonate facies have been related to slope instability (Hopkins 1977; Schlager & Camber 1986), eustatic sea-level changes (Watts 1987; Reijmer et al. 1988, 1991, 1992; Burchell et al. 1990; Reijmer & Everaars 1991; Herbig & Bender 1992) or tectonics (Bosellini 1989; Fernandez-Mendiola & Garcia-Mond6jar 1989; James et al. 1989; Playford et al. 1989; Elmi 1990; Garcia-Mond6jar 1990; Kenter et al. 1990; Watts & Blome 1990; Everts 1991). In this section it is shown that the majority of the redeposited facies in the Barru area are derived from a tectonically active faulted platform margin. Downcurrent changes in the organization of the redeposited facies are also discussed below.
375
redeposited facies interbedded with marls are considerably reduced in thickness or are absent (Fig. 3). In terms of spatial distribution and basement clast content the redeposited facies can be subdivided into two groups (Fig. 3). (1) The Bantimala redeposited facies are located to the north and east of the Bantimala and Barru Blocks respectively. This facies group contains basement clasts of lithologies, such as quartz-mica schists and ultrabasics, which outcrop in the northern part of the Bantimala Block. (2) The Barru redeposited facies crop out only to the north of the Barru Block. These facies contain basement clasts of lithologies such as serpentinite and granite-granodiorite, which are present in outcrops within the northern margin of the Barru Block.
Proximal to distal changes In the Bantimala redeposited facies the total thickness of marls and redeposited facies thickens and then thins towards the north over a distance of 15 km (Fig. 6)~ When the thickness of redeposited units is expressed as a percentage of total stratigraphic thickness there is a clear decrease in redeposited carbonate units, relative to marls, towards the north (Fig. 6). Correspondingly the percentage of amalgamated redeposited carbonate beds also decreases towards the north (Fig. 6). These factors indicate proximal to distal and basinward trends towards the north. The occurrence of a clast-supported breccia topped by a graded bioclastic pack-grainstone facies suggests a genetic relationship between this sequence and single beds of the graded bioclastic pack-grainstone facies. The spatial distribution of facies (Fig. 6) might also suggest a downcurrent change from a two layer into a single layer sediment gravity flow deposit. Downcurrent or lateral changes of two component gravity flows into single layer flows have been reported from both modern and ancient carbonate deposits (Davies 1977; Cook 1979; Krause & Oldershaw 1979). The planktonic foraminifera wacke-packstone facies may be the correlative distal deposit of the graded bioclastic pack-grainstone facies (cf. Davies 1977). Outcrop constraints preclude the possibility of proving this inferred proximal-distal facies relationship.
Nature o f clasts Subdivision of the redeposited facies Redeposited carbonate facies interbedded with basinal marls are located in the upper parts of sections in the northern and eastern parts of the Barru area (Fig. 3). West of the Rala anticline
The majority of lithoclasts in the redeposited facies are angular to sub-angular, often with the clast margin cutting across the grain fabric. A continuum of clast sizes ranging from 4 m down to fine sand grade are present. A wide variety of clast types occur in both groups of redeposited facies (Fig. 8c).
376
M.E.J. WILSON & D. W. J. BOSENCE
Schist, serpentinite, ultrabasic, quartzose sandstone and shale clasts were derived from the underlying formations. A range of limestone clasts which vary both in terms of age and environment of deposition are also present in the redeposited facies. Often a redeposited bed will contain limestone clasts which are both contemporaneous with and older than the intervening marls. Rip-up clasts of marl are often found within the redeposited facies.
Tectonic versus eustatic origin The deposition of the redeposited facies may have been triggered by eustatic or tectonic events. The variety of clast types excludes other possible causes, such as gas charging, tidal or surface wave action, rapid deposition or patchy shallow marine cementation resulting in unstable slopes. A number of factors, outlined below, exclude the possibility of a eustatic mechanism as the main cause of the majority of the redeposited facies. Shallow-water carbonate lithologies in the Barru area and on the Tonasa Carbonate Platform to the south, with the exception of the lowermost beds in the carbonate sequence and some beds in the Doi-doi section, lack lithic grains. It is therefore suggested that no major areas of basement highs existed during the deposition of Eocene shallowmarine carbonates. Erosion of basement lithologies or siliciclastic grains from carbonate deposits due to sub-aerial exposure would have required the removal of between 60-470 m of shallow-water carbonate deposits. This is much greater than, or close to, the inferred maximum for, short-term sea-level falls for the Tertiary, the value of which is thought to vary from 75-100 m (Haq et al. 1987; Sarg 1988). Limestone clasts in the Bantimala redeposited facies contain no evidence for subaerial exposure until the latest Oligocene. Prior to the deposition of marls during the late Eocene, both the Barru area and the area of the Tonasa Carbonate Platform were the sites of widespread accumulation of shallow-water carbonates. During the late Eocene rapid deepening occurred in the Barru area, whilst shallow marine sedimentation with only localized evidence for Oligocene karstification continued on the Tonasa Carbonate Platform. The range and immaturity of clast types within the redeposited units suggest derivation from a tectonically active faulted margin of a carbonate platform (cf. Kepper 1981; Hurst & Surlyk 1984; Martini et al. 1986; Eberli 1987; Burchette 1988; Eaton & Robertson 1993). Carbonate slopes become unstable at angles of 30-40 ° (Kenter 1990). Derivation of the redeposited carbonates is thought to be related to oversteepening of the
platform margin caused by faulting. The underlying formations are relatively easy to erode and undercutting of the platform margin may also have contributed to the oversteepening. Provenance studies, the spatial distribution of redeposited facies and field evidence outlined below, indicate that the redeposited facies were derived from the northern faulted margins of the Bantimala and Barru blocks. A eustatic control or enhancement of a tectonic control cannot be dismissed for some of the redeposited facies. Increased shallow-water carbonate production and 'shedding' into deeper water areas, caused by increased accommodation space and more favourable conditions may be related to eustatic changes. For example, packages of decimetre bedded bioclastic packstones, which only contain fossils contemporaneous with the adjacent marls, may be related to eustatic changes (cf. Reijmer et al. 1988; 1991; 1992; Watts 1987; Burchell et al. 1990; Reijmer & Everaars 1991; Herbig & Bender 1992).
Field evidence f o r derivation from a faulted platform margin No complete sections through the Barru redeposited facies are exposed. In general, the maximum clast size decreases and the bed thickness of these redeposited facies thins towards the north (Fig. 9). These factors indicate basinward and proximal to distal trends towards the north away from the Barru Block within the Barru redeposited facies. The Barru Block is bounded to the north by a fault downthrowing and dipping steeply (60-70 °) towards the NE (Fig. 3). This fault juxtaposes basement lithologies in the footwall against coarse redeposited facies of the Tonasa Limestone Formation in the hanging wall (Fig. 9). Although this suggests normal fault movement, kinematic indicators were not seen and the fault surface itself is not exposed, so it is not clear if oblique-slip was also involved. All the basement and limestone lithologies reworked into the Barru Redeposited Facies outcrop in situ within or adjacent to the margins of the Barru Block (Fig. 3), suggesting local derivation. These facts strongly suggest that the area of the Barru Block was the site of shallowwater carbonate accumulation and the Barru redeposited facies were derived from an active syn-depositional faulted margin to the Barru Block. No similar fault is exposed juxtaposing lithologies of the Bantimala Block against redeposited facies south of Rala. In part this may be due to outcrop constraints, with the upper volcaniclastic member of the Camba Formation unconformably overlying much of the northern edge of the
377
TERTIARY EVOLUTION OF S SULAWESI SSE
NNW
~ ...... Erosion of footwall ,, %" ,, ~' ,,:~.--~--k ,,-~ %" ,~ <--~-~-~--~'i lithologies into the , ,'%',', ,', ,', t, s, s%',', , ' ~ , ' % ' ~ t % ' t ~ s % ' ~ 1 ~,~ hangingwall area ,, z...~.,~,, ,, ,, ,, ,, ,, ,,~¢...,¢_.~ ,, ,, ,,1 l i b , ............. %
%" %'~1,~
%
% % %"
'%
%
Y
+
4~ %
%
%
%
%
%
%
%
%
,',',~',','4°,%~,',g$,,',',',',~, I % %
.,.
,,-
% %
%
.,,
% %
,,,
i ~% ~ ~ %,% ~ @ @ @ i ~ ~ ~ ! %
V
i~% %"
+. ÷ Y . ". "'1~. "..Z'. ' " , , %' , , '%, , ' ,%, ' ~%, , %~ , ,%' ~ k%~~ ÷ ' ÷@' ~ ÷, 1' ,%, " , ,%~ ,%1~l,~ ' , , ~ , "% ~ %
~
".~
%
%
%
%" %
%
%
%~+
+
~
%
%
%
%
%
~
%
%
%
%
%
%
%
@
~
%
%
%
%
%
%
................................ ',KEY
:
.
:
-
.
=
.
:
....................
;
.~ ..... L : : ::~:~.?~:;-:?:.:~-¢..., .... • ...........!ili,........ . . . ,o,o,, . . . ........ ~:~E
'
~ ~ .... ~ii~!~i::~.~i~!~ --- - . . -. . .-. .-. .-. . -. . .-. .-. .= -----------"""'"" ~ ................................
II
~
................................
I
~ %
%1"...~
:2 . . ; . . ,
~
. ' . . :; :::.. ,.. i::.. , ; . . * ~
.
............... .
.
.
.
.
310/60-70 ° Main fault
~
0
.
.
l.ii i
I]'l
i"" "l. . . . . . . . . .
I ii.
Underlying lithologies not exposed
"
160
, .
/
':~:~,?i!:~i~!:~::?~i~!i?i!~.~... ~ . . . . . . . . . . . . . .
M~::ld;posited packstonesand I ~ i oE 200"1 matrix to the limestone bm~ia/~ ~ ~I 100"t Shallow marine carbonates / ~ E ~1 E o50-1 Intrusive - pre-Tertiary ' -~u'! .E_=N104 [~-~ Calcite-veined metamorphic ' , : ~ 11
,_____bas_~_m_~_n_t-_pr_~-_T_e_rtia~_. . . . . . . . .
.
.,
. ~
~
%
~
.
~
%
%
.
.::.
• ". " . ~ ". ~ [ ~ " ~
+~. ' ~ . ~ . ~ ' .
.
::
% %.
%
~ ~.%'.~ % % ~ . ~ ÷
~
~
" .. .. ':':.:'.~""%"::i:(.:.!!-:i::~:~i)!~ ::.... •............................................................................. = : . .::.. <: .:"~;4.~.~.:::: .I:.::+-. , , , , , o , . , , , , , . . . . . . . . . . . . . . . . . . . . . . . . . ,,,, .......................
.,. ,., ke -
%
|
Present day land surface
250
~
"
560
~
'
~
~
-
~
1 6 0 0 2000 4 6 0 0 8000
-~
Distance awayfromfault(metres) Fig. 9. Schematic cross-section through the northern bounding fault to the Barru Block, showing redeposited facies thinning and fining towards the north away from the fault. See Fig. 3 for location. Not to scale.
Bantimala Block (Fig. 3). Faults south of Rala with similar trends as the north-bounding fault to the Barru Block may have acted as sources for the redeposited facies. However, movements on these faults contemporaneous with deposition of the Bantimala redeposited facies cannot be proved.
Configuration of the faulted platform margin Redeposited facies are inferred to have been sourced from at least two faulted northern margins of the Barru and Bantimala blocks. The lack of channelization and facies distribution suggests that both groups of redeposited facies were derived as sheets of carbonate debris from a faulted linesource forming base-of-slope carbonate aprons (Mullins & Cook 1986). The Barru redeposited facies are inferred to have been derived from the fault which today delimits the northern edge of the Barru Block. The precise area and fault for derivation of redeposited facies from the northern edge of the Bantimala Basement Complex is less clear. Regional geological observations and palaeocurrent data can be used to define possible configurations of the platform margin and to relate the two source areas.
Derivation On the eastern limb of the Rala anticline the lower marine member of the Camba Formation conformably overlies the Tonasa Limestone Formation (Fig. 3). On the western limb of this anticline the lower marine member is absent and an angular unconformity separates a condensed sequence of the Tonasa Limestone Formation, or older formations, and the upper volcanic member of the Camba Formation. The lack of abundant limestone clasts in the volcaniclastics directly above this unconformity suggests that major erosion had occurred prior to the deposition of the Camba Formation. This angular unconformity may therefore have been generated during the Eocene to Miocene and supplied material for the Bantimala redeposited facies. Two kilometres south of Doi-doi is a fault with a trend similar to the north-bounding fault of the Barru Block (Fig. 3). This normal fault dips to the NE and juxtaposes basement lithologies against the Balangbaru Formation. Southwest of this fault a thin layer of the Malawa Formation and upper Eocene packstones of the Tonasa Limestone Formation unconformably overlie metamorphic lithologies of the Bantimala Block. The packstones are in turn separated from the upper volcanic
378
M.E.J. WILSON & D. W. J. BOSENCE
member of the Camba Formation by an angular unconformity. The eastern part of this fault, which does not penetrate the Camba Formation, is intruded by a later diorite stock. The offset of the N-S trending contact between the Malawa and Tonasa Limestone Formations north of this intrusion indicate normal or possibly strike-slip movement on the fault (Fig. 3). South of the intrusion the Balangbaru, Malawa and Tonasa Limestone formations are all present. The Tonasa Limestone Formation in this location is composed of 3 1 0 m of shallow-water Eocene-Oligocene carbonates overlain by 50 m of deeper water carbonates dated as Oligocene (E. Finch 1993 pers. comm.). This block bounding fault and/or other similar trending faults in the vicinity or under the Camba Formation are inferred to be the northern faulted margin of the Bantimala Block, which shed material forming the Bantimala redeposited facies.
Palaeocurrent data The Bantimala redeposited facies thin and fine northwards away from the Bantimala Block (Fig. 6). Rare clast imbrication and scour structures (cf. Hubert et al. 1977; Hiscott & James 1985), indicate that the transport direction of the redeposited facies had strong northerly, northnorthwesterly and easterly trends (Fig. 3). Although palaeocurrent data are scarce, the southern area shows a dominant transport direction towards the east (Fig. 3). In the northern area finer redeposited facies tended to be north-directed and are inferred to have been derived from the Bantimala Block which lay to the south. Coarser redeposited units in the north tend to be
easterly directed and are thought to have been derived from the northern margin of the Ban'u Block, an inference supported by clast provenance data (Fig. 3).
Evidence for Tertiary exposure Carbonate lithologies are poorly exposed along the margins of the basement blocks close to the northbounding faults of the Bantimala and Barru blocks. Limestone clasts reworked from these footwall blocks, however reveal evidence of karstification and subaerial exposure. Limestone clasts derived from the Barru Block during the late Eocene (NP17-19, E. Finch 1993 pers. comm.) contain features such as irregular dissolution hollows lined by blocky cement and then infilled by fine carbonate silts and micritic glaebules (Fig. 10a). The margins of clasts cut across all of these early diagenetic features, suggesting subaerial exposure of the footwall area of the Barru Block. Limestone clasts in the Bantimala redeposited facies do not show corresponding evidence for sub-aerial exposure of the Bantimala Block until the latest Oligocene.
Regional tilting A number of factors suggest that the whole of the Barru area was tilted towards the east during the deposition of the Tonasa Limestone Formation. A thinned carbonate sequence, the angular unconformity with the overlying Camba Formation and absence of the Balangbaru Formation in the western part of the Bantimala footwall block all
Fig. 10. Photomicrographs showing features of the redeposited facies. Scale bar is 1 mm across. (a) Shallow-water carbonate clast showing evidence for karstification. An irregular dissolution hollow has been coated in a blocky cement, subsequently fine sediment has infilled the void and there is a final phase of sparry calcite cement at the top of the cavity. (b) Quartz mica schist derived from the pre-late Cretaceous basement complex adjacent to an Eocene large benthic Discocyclina foraminifera. Sample from a graded bioclastic packstone bed from the Bantimala redeposited facies.
TERTIARY EVOLUTIONOF S SULAWESI suggest that erosion was more effective to the west. Only the Barru Block underwent sub-aerial exposure from the late Eocene. From the late Eocene onwards, west of the Rala anticline, redeposited facies are considerably reduced in thickness (< 45 m) or absent compared with those to the east (up to 635 m). This indicates much greater accommodation space in the east during deposition of the Bantimala redeposited facies. In certain areas, palaeoflow towards the east and the overlap of redeposited facies from the two source areas in the northern palaeocurrent grouping imply a surface sloping towards the east. If this inferred eastward palaeotilt existed, slopes would have been very low since the two northern palaeocurrent groupings indicate that palaeoflow was mainly north directed.
Morphology of the faulted margin Two configurations of the faulted northern area are feasible from the late Eocene onwards. The first is domino style faulting, with the Barru Block being downthrown relative to the Bantimala Block. The second scenario is an east-dipping relay ramp between two major parallel faults; the northbounding faults to the Barru and Bantimala blocks.
Horizontalscale 0 5 km
10 . ' ..-/""
379
An east-dipping relay ramp between two major parallel faults is the preferred configuration of the northern area for the reasons outlined below and illustrated in Fig. 11. (a) The abundance of low energy marl facies east of the north-bounding fault to the Barru Block suggests that the Barru Block was not undergoing erosion and did not have topographic relief east of present day outcrop. The general lack of coarse redeposited facies north of the easternmost outcrop of the Barru Block suggests the north-bounding fault to the Barru Block did not continue southeastwards. The Camba Formation is not displaced east of the Barru Block, which also suggests the block bounding fault dies out towards the SE. (b) Much of the palaeocurrent data from redeposited facies in the two northern palaeocurrent groupings is north-directed. This indicates that sedimentation rates were high enough to overcome the dip slope of the hanging wall block. Lower angles of dip on the hanging wall block would be expected in the relay ramp situation. In the domino fault model axial flow between the two major faults towards the east, the direction of regional tilt, would be more likely. The Bantimala redeposited facies were transported slightly further north than the northern margin of the Barru Block. This
its
/
Vedical exaggeration /
Barru Block
i ~
; ,
~"/"/ : "'"
Block
', , ~", •", ," ~.' .......::::::::::::::::::::::::::::::::::::::::::::::. --
Key
~
•
.,.
,,.
[ ] Basinal marl ~_~ Redeposited facies Shallow water platform carbonate Sandstones, shale and coal ~ . " Shales, sandstones and conglomerates I'.1 Metamorphic, volcanic and sedimentary lithologies
~
Areal extent of the redeposited facies Overlap of redeposited facies
~ Z :':':"7 - - Sea,eve,
~'-::-/~,. Fault and direction i:::'7 ~ of movement . ~ . . ~ Karstification of subaerially ~","" exposed limestone
~
directioof n
Transpod
redeposited facies
Fig. 11. Block diagram showing the preferred reconstruction of the northern margin of the Tonasa Carbonate Platform in the Barru area during the late Eocene. Two tectonically active, major NW-SE trending normal faults are separated by an east-dipping relay ramp. The area is thought to have been tilted gently towards the east.
380
M.E.J. WILSON • D. W. J. BOSENCE
situation would be most unlikely in a domino fault block model. (c) With domino style faulting and tilting, footwall exposure of both major faults might be expected. In the relay ramp situation the two main faults border different parts of the same large block. With tilting of a large single block, the northern Ban'u Block area would have been more likely to undergo emergence compared with the more southerly area of the fault bounding the Bantimala Block. Earlier subaerial exposure of the Barru Basement area did occur, suggesting a relay ramp model is more appropriate (Fig. 11). The lack of a SE continuation of the northbounding fault to the Barru Block, redeposited facies distribution, and patterns and timing of subaerial exposure of footwall areas all suggest that a relay ramp is the more appropriate configuration for the platform margin.
History of the platform margin Changes in the bed-thickness, clast-type and content of the redeposited facies through time can be used to constrain the history of the platform margin (cf. Everts 1991; Reijmer et al. 1991). Figure 12 is a plot of the bed thicknesses of the redeposited facies against the percentage of clast types and ages, based on point counting thin sections. The plots on the right hand side of Fig. 12 highlight the percentage of clasts older than the background marl sedimentation and the percentage of clasts which are lithic fragments. Ages for the background marl sedimentation are derived from nannofossil (E. Finch pers. comm. 1993) and planktonic foraminifera identification (T. Wonders pers comm. 1992; E Banner pers. comm. 1994). Provenance studies and a knowledge of the stratigraphic sequence in South Sulawesi can be used to infer an age for non-carbonate clasts. For example, metamorphic and quartzose sandstone lithologies could only have been derived from pre-upper Cretaceous and upper CretaceousEocene non-carbonate formations respectively (Fig. 10b). Although it may be possible to date limestone clasts containing large benthic foraminifera (cf. Adams 1970), many of the carbonate grains cannot be dated. The presence of Discocyclina, Pellatispira or Biplanispira indicates Eocene ages, whilst Lepidocyclina indicates an Oligocene or younger age. Unbroken delicate planktonic foraminifera or rip-up marl clasts are considered to be contemporaneous with the background basinal marl sedimentation. Clasts of planktonic foraminifera bioclastic packstones occur very rarely. Redeposited carbonate units are thickest during three periods of time: late Eocene to early
Oligocene; latest early-early late Oligocene and early to middle Miocene (Fig. 12). When the redeposited facies are thickest, thick-bedded, clastsupported breccias are the dominant facies and marls occur rarely. During these times between 4090% of grains or clasts are lithic fragments and correspondingly a significant proportion of the clasts are older than the background marl sedimentation. In comparison, when the redeposited facies beds are thinner, marls are more abundant, there is a decrease in the amount of lithic clasts, and the majority of clasts are contemporaneous with the background sedimentation (Fig. 12). It has been inferred above that the main cause of the redeposited facies was tectonic activity on a faulted carbonate platform margin. Derivation of lithic clasts and lithologies older than contemporaneous marl sedimentation would have occurred during periods of faulting. At these times the footwall block would have had relative relief leading to 'exposure' and erosion of the block margins. Earlier lithified material from these elevated areas was prone to reworking into hangingwall depocentre areas. Since karstification of the Bantimala footwall area is only inferred from the late Oligocene, it is important to note that for much of the time erosion probably occurred in a sub-marine rather than a sub-aerial realm. Submarine downslope redeposition of shallow-water carbonate material is common along the margins of many modern carbonate banks due to slope instabilities (James & Ginsburg 1979; Mullins 1983). Three main phases of tectonic activity on the platform margin are inferred. These occurred during the late Eocene to early Oligocene, latest early/early late Oligocene and the early to middle Miocene (Fig. 12). These three phases can now be compared with more regional tectonic events.
Regional comparison The Tertiary sequence in the western part of South Sulawesi is similar to sequences in neighbouring Kalimantan and the east Java Sea. It has been postulated that the whole region was part of a widespread basin; the formation of which commenced in the early-middle Eocene (Daly et al. 1987, 1991; Pieters et al. 1987; Van de Weerd & Armin 1992). The Eocene was a period of widespread plate reorganization in SE Asia during which India collided with Eurasia (Dewey 1980; Patriat & Achache 1984; Daly et al. 1987, 1991). The bend in the Hawaii-Emperor sea mount chain also developed at this time, indicating a possible change in the motion of the Pacific plate (Daly et al. 1987, 1991). The Australian-Eurasian spreading centre shifted to its present southeast Indian location, during the Eocene (Packham 1990). Although this
381
TERTIARY EVOLUTION OF S SULAWESI
~I~
1100
"~ ~
Thickness of carbonate 1, 2, 3, 4,
redeposited units (m) 5, 1, 0 1,5 2 0, 2 5,
% Clast age
0
~f--'~, NN2-NN6 early/middle I1--\ Miocene
_
Sample % Clast type 100 No. 0 % J106 t ' ~ , , ' , , ' , .'_- ,'., ,'., , ' , J
%
\
'Envelope' showing trend of increasing
" \
'Envelope' i ,. showing trend \ of increased percentage of i li~ic fragments.i
\.~-___ percentage of
\
clasts older than
\ -
100 !
background
•,
\ !:_-_-_-_:-~
marl sedimentation.
..,
•
=
1000
379
PN6-N7 middle Miocene
/"
1
"
1 J69
..
i
early Miocene
~~2,v,¢>.~ x.~.~ 900
=¢
~
~
800
~
J52 J30
J28 J22 J15 J12 J3 J2 D97
~
NP23 latest early/early late Oligocene
~,="=='==-=="=="=="=="=~-. %-, II
D85
/
D79
i
i /// L-,'::..:..:..-..-..-..., ! - -
~.~ NP23 latest early/early late Oligocene
700
/
\
\
i
D60 L-~-_, . ; , . _ . . _ . ,_..-...'...-...1 i D52 I;..==-_J~J~.J"--'3i-,,-J~: ; .- ; : ; i I D49 i . . . . . . . . . . .:I ....... i
i
. . . .
062 I~'d. , " , ; . - . - . - .'.7
i'
Early Oligocene
i" ~
D42 t--,L--=L'..;'--';'--':'-'.=!;. .
D23
J /
".=: - " -'.-'_" -~:. r-.'.-r.;~C • \.\
D2 'h=:-=:-=;-=;-=:-=;-=-~_
I
~
~ NP18-NP20 late Eocene
DR23 ~ - - . - " = - - . , ' : ' ; .
:. , :. , J i
• ".. ". • ". • ". • i
R
500
=...___)1=17 late Eocene
J an36 k-.,-_ . . . . .
F ~,e-,re?; . . . . . . . . . . . . . . . . . .
470
:, ;, I
i D28 .k-=_;=_;=_;=_;=_;,j,'_-.:-._, .-...-=! D25 I . . . . ~ ; ~ " ,." ~ 11
~ - - ' ~ / ' ~ E n v e l o p e of bed ~' thickness of the _~.~_[~ _carbonate ---redepositedfacies
_i
=lcwI
~= BP - Bioclastic packstone ~ ~ MSB- Marl supported breccia• CSB - Clast supported breccm I GBP - Graded b oc ast c packstone i~ PFW - Planktonic foraminifera wackestone -_A_rk_os_ic packst_one_ J
" ! _
600
~
~'Carf:16n-at-e ~ r a c i e s - t y l ~ s . . . . .
/
D70
~ ") D 0 nn~E] ~ o ~ Carbonate facies
1 2 3 4 5 10152025 Thickness of redeposited carbonate units (m)
I mini Pre late Cretaceous i [r~ t ~ ?Palaeocene/Eocene - - sandstone = [ ] Eocene carbonate I [ ] O h"goce n e carbo n ate i
[ ] Miocene carbonate rI.--I - I Undated carbonate ma Contemporaneous m basinal sediments I [ ] V olcanic casts
:_.:':.,"_'U-.-.,i
',',-&e-,; ~ . . . . . .
,, F~ Lithic clas '~,~ ii ~ .UlOClaSIs ... iI ~ . fraomented H r ~ _. ° . . . . I 1151oclasIs WhOle == jr.
Fig. 12. Point-counting data plotted against the bed thicknesses for the redeposited facies in the Rala measured section. Note the increased percentage of clasts derived from earlier formations or reworked from older shallow marine carbonates of the Tonasa Limestone Formation when the redeposited beds are thickest.
382
M.E.J.
WILSON & D. W. J. BOSENCE
plate reorganization is thought to have resulted in Eocene basin formation, in detail the mechanics of this basin formation are as yet unclear. Initiation and formation of different parts of the basin have variously been related to extensional back-arc formation (Hamilton 1979; Daly et al. 1987, 1991) and a foreland-fore-arc flexural origin (Williams et al. 1988; Hutchison 1989). Two major fault trends are apparent from seismic data in the Makassar Strait offshore Sulawesi (Fig. 13). Approximately NE-SW trending faults have been related to back-arc extension due to roll-back of a plate subducting eastwards under eastern Sulawesi (Daly et al. 1987, 1991; Letouzey et al. 1990). Although Eocene-Oligocene volcanics in eastern South Sulawesi are thought to delimit the plate margin, detailed analysis has not been undertaken and it is not clear if these volcanics are related to subduction (Letouzey et al. 1990; Van de Weerd & Armin 1992). In the Kalosi area upper Miocene-Pliocene igneous rocks have been related to subduction of continental crust (Bergman et al. 1996). Another possibility is that the Makassar Strait extension may represent a failed armextremity of sea floor spreading in the Celebes Sea to the northeast (Hutchison 1988; Moss 1994). The second group of faults trends NW-SE and includes the Adang and Sangkulurang faults (Fig. 13).
NW
TT-2
TT-1
1~ -~:--~
SE
I
~~ z , ~
"-~r-__-S,-_z' % t
km
~
TT-1 m 0 mid-Miocene to Recent 192
m TT-2
62 821
4
early to mid- i
.~
"
Mi°cene 1238. . . . . . .
1 5 ~
~ ,\\ ~ "~, \ ~\ \\.
~'~] rq~,~,
KEY Deepmarine shales+ marls
Redeposited :: ,,mestonetaces
Shallowmarine
Oligocene
~ 2182 to early ~i~~ Miocene ~]
', ,\ ~
i [] ,
~
elastics(Marginal). marine
~ ~ ~
clastics
Non marine Basement
...................
~ ~. ~ ~~ ~J,~i
2892. . . . . . . . . . 3055 Eocene 3238 no recovery 3560 . . . . . . . . . . ]~P'--7] Cretaceous-inferred
Fig. 14. Seismicsection and borehole data from the inverted Taku Talu fault, Makassar Straits (after Situmorang 1987). See Fig. 13 for location of boreholes.
Offshore analogue o f the northern faulted margin The late Eocene to middle Miocene redeposited carbonate facies in the Barru area are inferred to have been derived from the faulted northern margin of the Tonasa Carbonate Platform. An analogous example of syn-tectonic carbonate sedimentation is revealed from seismic and borehole data in the Makassar Straits, from the vicinity of the NE-SW trending Taku Talu fault (Fig. 14). Eocene to middle Miocene carbonate breccias containing clasts from the underlying basement were deposited as a thickened hanging wall sequence, whilst shallow-water carbonate sedimentation continued on the footwall block of the Taku Talu fault (Fig. 14). Thus the Taku Talu fault was a major active Eocene to early Miocene normal fault which underwent later inversion during the middle Miocene (Situmorang 1987). ~
Outcropping Eocene to Holocene.TT-2 Boreholelocation(seeFig. 14) [ sedimentary basin fill pre-Cretaceous continentalcrust ..-,.-j Subduct-iotick nFaUl markSzone t downthrowside
Fig. 13. Map to show the location of Tertiary basins in Kalimantan and western Sulawesi. Modified after Van de Weerd & Armin (1992).
A d a n g fault The Adang fault is a major NW-SE trending structure separating the North and South Makassar Basins (Fig. 13). This lineament is thought to extend westwards onshore forming the southern
TERTIARY EVOLUTION OF S SULAWESI boundary of the Kuta Basin and has been linked across Borneo with the Lupar Line forming a 'trans-Borneo shear' (Woods 1985; Kusuma & Darin 1989; Wain & Berod 1989; Bransden & Matthews 1992; Van de Weerd & Armin 1992). Both strike-slip (Woods 1985; Hutchison 1989; Biantoro et al. 1992) and normal displacements (Biantoro et al. 1992; Rangin et al. 1990) on the Adang fault have been inferred. It has been suggested that the Adang fault was an important Eocene to early Miocene normal fault downthrowing to the north, which subsequently underwent strike-slip reactivation during the middle-late Miocene (Kusuma & Darin 1989; Rangin et al. 1990; Biantoro et al. 1992). During the late Eocene to Miocene the Adang fault had a strong influence on sedimentation patterns. The Paternoster Platform and Barito Basin were the sites of shallow-water carbonate development, whilst deeper sedimentation occurred in the Kuta Basin to the north (Van de Weerd et al. 1987; Kusuma & Darin 1989; Wain & Berod 1989). Seismic data across the Makassar Straits (D. Coffield pers. comm. 1995) suggest there was no fault linkage between the NW-SE trending Adang Fault and the main north-bounding fault to the Tonasa Carbonate Platform. It is suggested that NW-SE trending structures influenced sedimentation patterns in Sulawesi and Kalimantan. NNESSE trending faults such as the Walanae Fault also had an effect on Tertiary sedimentation patterns in western Sulawesi (Van Leeuwen 1981; D. Coffield pers. comm. 1995). Seismic data N-S across the Sengkang Basin reveal normal block-faulting, with blocks being downthrown to the centre of the basin (A. Ngakan pers. comm. 1994). Indeed, the northbounding faults to the Tonasa Carbonate Platform and faults with a similar trend to the east may mark the southern boundary of the NW-SE trending depression, which runs through the Sengkang Basin and appears structurally to separate South Sulawesi from the rest of the western arc of Sulawesi (Fig. 1). Timing
Redeposited carbonate facies in the Barru area suggest three phases of tectonic activity in the area during the late Eocene-early Oligocene, earlylate Oligocene and early-middle Miocene. Other evidence for possible tectonic activity in the region is documented below and possible causes are discussed (Fig. 15). E a r l y - m i d Eocene. A basal angular unconformity
to many initially transgressive Tertiary sequences marks widespread basin initiation (Cater 1981; Pieters et al. 1987; Van de Weerd et al. 1987;
383
Kusuma & Darin 1989; Wain & Berod 1989; Letouzey et al. 1990; Bransden & Matthews 1992; Hutchison 1992; Van de Weerd & Armin 1992). The angular unconformity between the Malawa Formation and underlying Balangbaru Formation is one such contact. Basin initiation possibly occurred as early as the Palaeocene in some localities (Bishop 1980; Cater 1981; Wain & Berod 1989). Carbonate production, including the deposition of the Tonasa Limestone Formation, had begun in many areas by the middle to late Eocene. Active faulting and graben formation has been inferred from seismic (van de Weerd et al. 1987; Bransden & Matthews 1992) and outcrop data (Tyrrel et al. 1986; Kusuma & Darin 1989). Basin formation is thought to be a response to widespread Eocene plate reorganisation. Late E o c e n e - e a r l y Oligocene. This period corre-
sponds with the first phase of coarse redeposited facies in the Barru area, indicating the initiation (or reactivation) of faulting on the margin of the Tonasa Carbonate Platform. The geometry and timing on major faults from regional seismic data suggest that localized extension was underway by the early Eocene (P9/P10), and rifting was widespread by the late Eocene (Van de Weerd et al. 1987; Letouzey et al. 1990; Bransden & Matthews 1992). The major variation in preserved thicknesses across faults, and hence syn-tectonic deposition is observed during the late Eocene and early Oligocene (Bishop 1980; Bransden & Matthews 1992). During the late Eocene a significant tectonic event caused widespread high angle faulting and subsequent erosional truncation affecting a number of areas (Bransden & Matthews 1992). This period is thought to represent the main extensional phase in the region (Letouzey et al. 1990; Bransden & Matthews 1992). The middlelate Eocene is also the age of the oldest oceanic crust in the Celebes Sea (Weissel 1980). Middle Oligocene. A mid-Oligocene unconformity
has been reported from many offshore (Bransden & Matthews 1992) and onshore areas (Van Leeuwen 1981; Kusuma & Darin 1989; Van de Weerd & Armin 1992). This unconformity, the major shift in facies belts, localized channelling and thickness variations have been related both to the mid Oligocene eustatic lowstand (Bransden & Matthews 1992) and tectonic activity (Cater 1981; Van de Weerd et al. 1987; Wain & Berod 1989; Sailer et al. 1992; Van de Weerd & Armin 1992). Sailer et al. (1992) reported exposure and subsequent deepening prior to the middle Oligocene (29.5-30 Ma) sea-level fall of Haq et al. (1987). In Sulawesi the second phase of coarse redeposited facies in the Barru area, an unconformity in the
384
M.E.J. WILSON (~ D. W. J. BOSENCE W.Kalimantan Basins
Age
Upper Kutai Basin
SE.Kalimantan Basins
:'. L
•.
•.
E.Java Sea Kangean
-7"~-
M
" "
[ : : : : : : : ' : ' : : : : 1 % % % % % % " , , '
r.
.. • • ~--;'~.~.'.'.' "
.
.
.
.
.
• .-,-o'.-.-,-.'.','.'.','.'."
I.............................[-,-.-
.
.
.
....
.
.
.
•
. .
,
r.'.'.'.'.'.'.'.'.'.'.'.'.'.'L'.'.'.'.'.'.'.'.'.'.'.'.'.'-,
. . . . . . . . . . .:.:.:.:.:.:.:.:.:.:.:.:.:.:.r ::.- - , . :..-.-.,--~.:.:.:
,. . . . . . . . . . . . . . .
1',
"
•
"
•
"
•. •. •.
I.:.:.:.:.:.::.:.:.:.:.:.:.iil!.:.: o
r- .-- . ' . '.. -.. -.. '.. _. ' ..' _. . '.. ' .. ' ..' . ' . ' l .
'
,-..,~..
....
EI:::::...Y'~',~
I
',
~'.'.'.---.'~--':~,
,
l
m
~--~'~
I
,
.
.
•
=~=
,
• =
~ .~ ". . . . ".
"
"
I
r m
"
~ ' m
-- - I. " ' - ' ":
-.-.-
"
~
. •.
.
.~,.:,~,_. • • ., .... l l l , ~"'~""~ ~ . ". . ". " , . . . .
-
~.~.
=
m
"
i
"
m
" "
-_,. . . ."
~. -. ', .-
l
m
, i i
l
i I
l
I:~L.;~rL ~:':~"
'
'~'~.~.."
,
,
,
.
.
.
•
•
on
"
"
•
~ , ,
m
, '
,
, m
,
, l
,,~,
" "
i -. . . . . . .- ' - ~,%
,~..,
•
.
,,,,,
Sea floor
•
,
.~.~
....
"
-
~
Emplacement
m .~,..w;--- " " " I l I l V.. " " ,~. • • , , , .::~ • • ,-.:.~" " ", . . . . :&~" " , ' ~=--.--I : , ~ Y I ~l : ~ . - - .
• " • " • " • " • " • [ • " • " • " .
•.
•. •.
•. r.
•. -.
.
•. •.
.
: : . • • • • •
I l
n
m i m
m
n
n
m , n
n
n
m
, m
n
, l
Collisionof Indiawith
C.an0e,n
. '... ',.. ', .. ,' ...' ...'...'.. • ~ ; . , = ."- : -•, - -"- ~•, ~" • "• •. "• . . . . . . . . . . . . . . ~
"
~
~
J
"
" •
•. •
•. . , •
,
~
E
J
KEY
of the
EastSulawesi
Ophiolite
-
"." " . '-..'... '-.. . . . . ~
spreading S.China Sea
'I. i n t h e
. ..... > .........................
-." " . '•. ,' .. .' ...".
impinge Sulawesi
•
"
========================================= :::: :::::::
om
east
=
" ~...........
'
.'.
• ...~
-
•
"
' I'~r~.~_" m
fragments
°.
•
."'.',',',"',',",',','.',',','" " "
"
: ~ .
-..
"
"
i m ,
m
=
....
•
" ~'.~- -" . " ~ ' ~• "• "-'='~'=-" "--" -"- - _-" , ~ ' , ~ , . , ~ ' - - - " .
~,"..:t, " " .~.~,, . . , , , ~'~" | I I I l I ' ~'.~.~.~%~'~ ~
:
•
• ---_._
•
m
,
"~,~"
N .j
=
1;"~.~" m.,~."
'
•.
microcontinent
_,_,m
•
"
,
,
•
Collision of Australia with Indonesia and
,,. . . . . . .
i
~,
"
. : .
.:.
•
~"
. m
- . ., . .
•
-
"
•
~:.~. =
• " ~ t ~ , ~ .
÷:
: • : • ~• - • • • •" t " r' - r--.F' .~ : . L " ' ~ i ~
" " . ~.'#'~,
"
•
.:. . . . . .
~.z.~--'T'T'-.'--..']:~T'T'~'-.'~i~:'.'.'.','.'-[.'.'.'-:-: I'.'.'.'.'.'-"".".'-',,.;r'.. " • "1. " . " • " • " • " • " • " 4" • " i.................~.-..~.. •. •. •. •. • ~..4~..
M " .................,~.,v~..
.~?.=
, m
• : • : •
, ~.~.',1,1 I I I ~ ' ~ : : ' ~ "
•.
~ , " " ~ ~,%%'~" " " l
:.:.:.:.
'• . ~ '".
•
• • • •
.
::::::::::::::::::::::::::: : : : : : : : : : " :-: •.
"
.
• ".'.-.'.-.-.'.:.:.:.:.:.:.:, "'"'""':'"":':"':'""" ": . :- • :- . ": . :- . :" • :~. ~- ~ ~, .~~ .:.:.:.:.:....................
•
• . • •.
.........
. ~o , ; .
-.-];,:,~:_.
I'.'.'.'.'.'.'.'.'.'.'.'.'.'.'$'-'-'.'.'.'.'.'.'.'-'.'.'.'.I~Ir~ I'.'.'.'.'.'.'.'.'.'-'-'.'-'.T.'.'.'.'.'.'.'.'.'.'.'.'-'-,,
i
*
. .
-.......
"
• • . • • •. .
.
• -.-.-.-.-.-.". ". ". ". ". ".-,~, :"... "=-.. " " • " • " .• _- _._
.~.. ;~ . ; . . - - . _ ; T . . . - ' . ' _ . " "
_ - , - - . ' ~ l - T r r . - - . - . l T r r
•.'.'.'.'.'.'.'.'.'.'...'."
.
Events
'" : :" : : : : : : : : : : [ : : : : : : : : : : : : : :[vv",,"v"vvv"vvv' •
,
Southwest Sulawesi
T r 7:
..............
E
/
plate
motion
Pacific plate. Development
of of
spreading centre Indian Ocean [ 7 t
Deep marine shales + roads
r ~
Volcaniclastics
~ ~
Redeposited carbonate facies
[~
~ (Marginal) Shallow marine limestones [,_'1 clastms
marine
in
~ Non marine I',','i clastics
Fig. 15. Tertiary stratigraphic correlation chart, showing similarities between the sequences in Kalimantan, East Java Sea and Sulawesi. Data taken from Bishop (1980), Cater (1981), Van de Weerd et al. (1987), Kusuma & Darin (1989), Wain & Berod (1989), Bransden & Matthews (1992) and Van de Weerd & Armin (1992).
Biru area (Van Leeuwen 1981) and minor localized evidence for exposure on the Tonasa Carbonate Platform (personal observation) during the mid Oligocene, all suggest a tectonic cause. During the Oligocene sea floor spreading began in the South China Sea (Holloway 1982; Ru & Pigott 1986; Daly et al. 1991) and the east Sulawesi ophiolite was emplaced (Parkinson 1991). Early to middle Miocene. In the Barru area west-
verging folding affects both the Tonasa Limestone Formation and the overlying middle-late Miocene volcaniclastics. The latest Oligocene to middle Miocene phase of the Bantimala Redeposited Facies therefore records a period of faulting and uplift prior to folding (and possible inversion). Based on seismic evidence from the east Java Sea and Makassar Strait, contemporaneous faulting and syn-tectonic graben fills have been inferred until the early Miocene (Bishop 1980; Cater 1981, Letouzey et al. 1990). Inversion of many of the earlier faults with normal displacement occurred slightly later during the early to middle Miocene (Fig. 14; Letouzey et al. 1990; Bransden & Matthews 1992). This compressive regime, active
up to the present-day, is widely attributed to the collision of microcontinental fragments onto eastern Sulawesi during the middle to late Miocene (Daly et al. 1987, 1991; Van de Weerd & Armin 1992). The middle to late Miocene is also the time when a volcanic arc developed in western Sulawesi (Sukamto 1975; Yuwono et al. 1985; Coffield et al. 1993; Bergman et al. 1996). Similar to the major bounding faults of the north margin of the Tonasa Carbonate Platform, many of the faults mentioned above had major normal displacements. Inferring an additional strike-slip motion would depend on along-strike linkage of faults, a factor which is difficult to ascertain from the available seismic and outcrop information. A transtensional Palaeogene history for the region has been suggested on the basis of the length of faults, apparent reversal and observed variability of fault trends (Bishop 1980; Bransden & Matthews 1992).
C o n c l u s i o n s
Late Eocene to middle Miocene redeposited facies of the Tonasa Limestone Formation in the Barn]
TERTIARY EVOLUTION OF S SULAWESI area provide a unique example of the use of carbonate sedimentology in comparing local and regional tectonic events. In the shallow-water carbonates of the Tonasa Carbonate Platform to the south, tectonic effects are impossible to distinguish from eustatic effects. However, clast composition and facies types in the redeposited carbonates reveal that tectonic and not eustatic changes were the main controls on sedimentation. In fact, redeposited carbonates in the Barru area prove to be a remarkable natural seismograph, recording regional tectonic changes and a number of local and regional conclusions can be inferred from them. (1) The spatial distribution, b a s e m e n t clast content and p r o x i m a l - d i s t a l trends within redeposited facies suggest derivation from two main source areas. These areas were from the northern margins of the Bantimala and Barru Blocks. (2) The textural immaturity and provenance of clasts indicate that the redeposited facies were derived from the faulted margins of a carbonate platform. Redeposited facies were derived from at least two faulted line-sources and were deposited as sheet flows forming carbonate aprons at the base of the slope. (3) An east-dipping relay ramp between two main N W - S E trending faults is the preferred configuration of the northern margin of the Tonasa Carbonate Platform. (4) The redeposited facies indicate three phases of tectonic activity. These occurred during the late
385
Eocene to early Oligocene, the middle Oligocene and the early to middle Miocene. This is consistent with other similar sequences in Kalimantan and the East Java Sea and with inferred regional plate tectonic changes. (5) The northern faulted margin of the Tonasa Carbonate Platform has a similar trend to N W - S E orientated structures in the Makassar Straits and Kalimantan. These structures are inferred to have i n f l u e n c e d s e d i m e n t a t i o n patterns during the Eocene to middle Miocene. The senior author gratefully acknowledges BP Exploration, UK for their generous financial support during the course of her PhD study, of which this work forms a part. The SE Asia Research Group, Royal Holloway, University of London, especially Dr Tony Barber and Diane Cameron, are thanked for their administrative and technical support. In Indonesia, Alexander Limbong, the senior author's counterpart from GRDC, Bandung, during three gruelling 'non-stop' field seasons deserves special thanks. GRDC, Bandung, Kanwil, South Sulawesi, particularly Darwis Falah and family, BP offices in Jakarta and Ujung Pandang and LIPI all provided technical and practical support. Dr Ted Finch and Prof. Fred Banner at University College London, and Dr Toine Wonders, Consultant, UK, are thanked for their excellent biostratigraphic work. The constructive comments from referees Dr Dana Coffield and Dr Neil Harbury and those by Dr Tony Barber, Rob Bond, Nigel Deeks and Dr Steve Moss towards improving this paper were much appreciated. Keith Denyer is thanked for producing the photographic plates.
References ADAMS, C. G. 1970. Reconsideration of the East Indian Letter Classification of the Tertiary, British Museum of Natural History Geology Bulletin, 19, 87-137. BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. & GARRARD, R. J. 1996. Tertiary tectonic and magmatic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for a Miocene continentcontinent collision. This volume. BERRY,R. E & GRADY,A. E. 1987. Mesoscopic structures produced by Plio-Pleistocene wrench faulting in South Sulawesi, Indonesia. Journal of Structural Geology, 9, 563-571. BIANTORO, E., MUR1TNO, B. E & MAMUAYA, J. M. B. 1992. Inversion faults as the major structural control in the northern part of the Kutai Basin, east Kalimantan. In: Proceedings of the Indonesian Petroleum Association 21st Annual Convention. 45-89. BISHOP, W. E 1980. Structure, stratigraphy, and hydrocarbons offshore southern Kalimantan, Indonesia. AAPG Bulletin, 64, 37-58. BOSELLINI, A. 1989. Dynamics of Tethyan carbonate platforms. In: CREVELLO,P. D., WILSON,J. L., SARG, E & READ, J. E (eds) Controls on Carbonate Platform and Basin Development. Society of Economic Paleontologists and Mineralogists, Special Publication, 44, 3-13.
BOUMA, A. H. 1962. Sedimentology of Some Flysch Deposits: A Graphic Approach to Facies Interpretation. Elsevier, Amsterdam. BRAITHWAITE, C. J. R. & HEATH, R. A. 1992. Deposition and diagenesis of debris flows in Upper Ordovician limestones, Hadeland, Norway. Sedimentology, 39, 753-767. BRANSDEN, P. J. E. & MATTHEWS, S. J. 1992. Structural and stratigraphic evolution of the east Java Sea, Indonesia. In: Proceedings of the Indonesian Petroleum Association 21st Annual Convention. 417-453. BURCHELL, M. T., STEFANI, M. t~ MASETTI, D. 1990. Cyclic sedimentation in the southern Alpine Rhaetic: the importance of climate and eustasy in controlling platform-basin interactions. Sedimentology, 37, 795-815. BURCHETTE, Z. P. 1988. Tectonic control on carbonate platform facies distribution and sequence development: Miocene, Gulf of Suez. Sedimentary Geology, 59, 179-204. BUSTILLO, M. A. & RUIZ-ORTIZ,P. A. 1987. Chert occurrences in carbonate turbidites: examples from the Upper Jurassic of the Betic Mountains (southern Spain). Sedimentology, 34, 611-621. CATER, M. C. 1981. Stratigraphy of the offshore area south of Kalimantan, Indonesia. In: Proceedings of
386
M . E . J . WILSON ,~¢ D. W. J. BOSENCE
the Indonesian Petroleum Association lOth Annual Convention. 269-284• COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A., GURITNO, N., ROBINSON,N. M. & TALBOT,J. 1993. Tectonic and stratigraphic evolution of the Kalosi PSC area and associated development of a Tertiary petroleum system, South Sulawesi, Indonesia. In: Proceedings of the Indonesian Petroleum Association 22nd Annual Convention. 679-706. CONIGLIO, M. 1987. Biogenic chert in the Cow Head Group (Cambro-Ordovician), western Newfoundland. Sedimentology, 34, 813-823. COOK, H. E. 1979. Ancient continental slope sequences and their value in understanding modem slope development. In: DOYLE, L. J. & PILKEY, O. H. (eds) Geology of Continental Slopes. Society of Economic Paleontologists and Mineralogists, Special Publication, 27, 287-305. & MULLINS,n. T. 1983. Basin margin. In: SCHOLLE, E A., BEBOUT, D. G. & MOORE, C. H. (eds) Carbonate Depositional Environments. AAPG Memoir, 33, 539-612. • & TAYLOR,M. E. 1977. Comparison of continental slope and shelf environments in the Upper Cambrian and lowest Ordovician of Nevada. In: COOK, H. E. & ENOS, P. (eds) Deep-water Carbonate Environments. Society of Economic Paleontologists and Mineralogists, Special Publication, 25, 51-81 , MCDANIEL, P. N., MOUNTJOY,E. W. & PRAY,L. C. 1972. Allochthonous carbonate debris flows at Devonian bank ('reef') margins, Alberta, Canada. Bulletin of Canadian Petroleum Geology, 20, 439-497. CREVELLO, P. D. & SCHLAGER,W. 1980. Carbonate debris sheets and turbidites, Exuma Sound, Bahamas. Journal of Sedimentary Petrology, 50, 11211148. CROTTY, K. J. & ENGELHARDT, D. W. 1993. Larger foraminifera and palynomorphs of the upper Malawa and lower Tonasa Formations, southwestern Sulawesi Island, Indonesia. In: THANASUTHIPITAK, T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies & Paleontology. 31 January-5 February, Chiang Mai, Thailand. 71-82. DALY, M. C., HOOPER, B. G. D. & SMITH, D. G. 1987. Tertiary plate tectonics and basin evolution in Indonesia. Proceedings of the Indonesian Petroleum Association, 16th Annual Convention. 399-427. , COOPER, M. A., WILSON, I., SMITH, D. G. & HOOPER, B. G. D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine and Petroleum Geology, 8, 2-21. DAVIDSON, J. W. 1991. The geology and prospectivity of Buton island, S.E. Sulawesi, Indonesia. In'. Proceedings of the Indonesian Petroleum Association, 20th Annual Convention. 209-233. DAVIES, G. R. 1977. Turbidites, debris sheets, and truncation structures in upper Paleozoic deepwater carbonates of the Sverdrup basin, Arctic archipelago. In: COOK, H.E. & ENos, E (eds) Deep-water Carbonate Environments. Society of -
-
Economic Paleontologists and Mineralogists, Special Publication, 25, 221-247. DEWEY, J. E 1980. Episodicity, sequence and style at convergent plate boundaries. In: STRANGEWAY, D.W. (ed.) The Continental Crust and its Mineral Deposits. Geological Association of Canada, Special Paper, 20, 553-576. EATON, S. & ROBERTSON, A. H. E 1993. The Miocene Pakhna Formation, southern Cyprus and its relationship to the Neogene tectonic evolution of the Eastern Mediterranean. Sedimentary Geology, 86, 273-296. EBERLI, G. P. 1987. Carbonate turbidite sequences deposited in tiff-basins of the Jurassic Tethys Ocean (eastern Alps, Switzerland). Sedimentology, 34, 363-388. ELMI, S. 1990. Stages in the evolution of late Triassic and Jurassic carbonate platforms: the western margin of the Subalpine Basin (Ard~che, France). In: TUCKER, M. E., WILSON, J. L., CREVELLO, P. D., SARG, J. R. & READ, J. F. (eds) Carbonate Platforms: Facies, Sequences and Evolution. International Association of Sedimentologists, Special Publication, 9, 109-144. EVERTS, A. J. W. 1991. Interpreting compositional variations of calciturbidites in relation to platformstratigraphy: an example from the Paleogene of SE Spain. Sedimentary Geology, 71, 231-242. FERNANDEZ-MENDIOLA, P. A. & GARCIA-MONDt~JAR, J. 1989. Evolution of a Mid-Cretaceous carbonate platform, Gorbea (northern Spain). Sedimentary Geology, 64, 111-126. FORTUIN, A. R., DE SMET, M. E. M, HADIWASASTRA,S., VAN MARLE, L. J., TROELSTRA, S. R. & TJOKROSAPOETRO, S. 1990. Late Cenozoic sedimentary and tectonic history of south Buton, Indonesia. Journal of Southeast Asian Earth Sciences, 4, 107-124. GARCIA-MONDI~JAR, J. 1990. The Aptian-Albian carbonate episode of the Basque-Cantabrian Basin (northern Spain): general characteristics, control and evolution. In: TUCKER, M. E., WILSON, J. L., CREVELLO, P. D., SARG, J. R. & READ, J. E (eds) Carbonate Platforms: Facies, sequences and evolution. International Association of Sedimentologists, Special Publication, 9, 257-290. GARRARD,D., SILALAHI,D., SCHILLER,D. & MAHODIM,P. 1989. Sengkang basin, South Sulawesi. Indonesian Petroleum Association post-convention field trip, 1989. GARRARD,R. A., SUPANDJONO,J. B. & SURONO. 1988. The geology of the Banggai-Sula microcontinent, eastern Indonesia. In: Proceedings of the Indonesian Petroleum Association, 17th Annual Convention. 23-52. GAWTHORPE,R. L. 1986. Sedimentation during carbonate ramp-to-slope evolution in a tectonically active area: Bowland Basin (Dinantian), northern England. Sedimentology, 33, 185-206. GRAINGE, A. M. & DAVIES,K. G. 1983. Reef exploration in the east Sengkang basin, Sulawesi, Indonesia. In: Proceedings of the Indonesian Petroleum Association, 12th Annual Convention. 207227.
TERTIARY EVOLUTION OF S SULAWESI HAMILTON, W. 1979. Tectonics of the Indonesian region. United States Geological Survey Professional Paper, 1078. HAMPTON, M. A. 1972. The role of subaqueous debris flow in generating turbidity currents. Journal of Sedimentary Petrology, 42, 775-793. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156-1165. HASAN, K. 1991. The Upper Cretaceous flysch succession of the Balangbaru Formation, Southwest-Sulawesi. In: Proceedings of the Indonesian Petroleum Association, 20th Annual Convention. 183-208. HERBIG, H. G. & BENDER, P. 1992. A eustatically driven calciturbidite sequence from the Dinantian II of the Eastern Rheinisches Schiefergebirge. Facies, 27, 245-262. HISCOTT, R. N. & JAMES, N. P. 1985. Carbonate debris flows, Cow Head Group, western Newfoundland. Journal of Sedimentary Petrology, 55, 735-745. HOLLOWAY,N. H. 1982. North Palawan block, Philippines its relations to Asian mainland and role in evolution of the South China Sea. AAPG Bulletin, 66, 1355-1383. HOPKINS, J. C. 1977. Production of foreslope breccia by differential submarine cementation and downslope displacement of carbonate sands, Miette and ancient wall buildups, Devonian, Canada. In: COOK, H. E. & ENOS, P. (eds) Deepwater Carbonate Environments. Society of Economic Paleontologists and Mineralogists, Special Publication, 25, 155-170. HUBERT, J. E, SUCHECKI, R. K. & CALLAHAN,R. K. M. 1977. The Cow head breccia: sedimentology of the Cambro-Ordovician continental margin, Newfoundland. In: COOK, H.E. & ENOS, E (eds) Deep-water Carbonate environments. Society of Economic Paleontologists and Mineralogists, Special Publication, 25, 125-154. HURST, J. M. & SURLYK, E 1984. Tectonic control of Silurian carbonate-shelf margin morphology and facies, North Greenland. AAPG Bulletin, 68, 1-17. HUTCHISON, C. S. 1988. Stratigraphic-tectonic model for eastern Borneo. Geological Society of Malaysia Bulletin, 22, 135-151. 1989. Geological Evolution of Southeast Asia. Oxford Monograph on Geology and Geophysics, 13. Oxford Science Publications. -1992. The Eocene unconformity on southeast and east Sundaland. Geological Society of Malaysia Bulletin, 32, 69-88. JAMES, N. P. & GINSBURG, R. N. 1979. The seaward margin of the Belize barrier and atoll reefs. International Association of Sedimentologists, Special Publication, 3. , STEVENS,R. K., BARNES, C. R. & KNIGHT, I. 1989. Evolution of a lower Paleozoic continentalmargin carbonate platform, northern Canadian Appalachians. In: CREVELLO, P. D., WILSON, J. L., SARG, F. & READ, J. E (eds) Controls on Carbonate Platform and Basin Development. Society of Economic Paleontologists and Mineralogists, Special Publication, 44, 123-146.
387
JOHNS, D. R., MUTTI, E., ROSELL, J. & SI~GURET,M. 1981. Origin of a thick, redeposited carbonate bed in Eocene turbidites of the Hecho Group, south-central Pyrenees, Spain. Geology, 9, 161-164. KE~rrER, J. A. M. 1990. Carbonate platform flanks: slope angle and sediment fabric. Sedimentology, 37, 777-794. --, REYMER, J. J. G., VAN DER STRAATEN, H. C. & PEPER, T. 1990. Facies patterns and subsidence history of the Jumilla-Cieza region (southeastern Spain). Sedimentary Geology, 67, 263-280. KEPPER, J. C. 1981. Sedimentology of a middle Cambrian outer shelf margin with evidence for syndepositional faulting, eastern California and western Nevada. Journal of Sedimentary Petrology, 51, 807-821. KRAUSE, t E & OLDERSHAW, A. E. 1979. Submarine carbonate breccia beds - a depositional model for two-layer sediment gravity flows from the Sekwi Formation (Lower Cambrian), Mackenzie Mountains, Northwest Territories, Canada. Canadian Journal of Earth Sciences, 16, 189-199. KUSUMA, I. & DARtN, T. 1989. The hydrocarbon potential of the lower Tanjung Formation, Barito Basin, SE Kalimantan. In: Proceedings of the Indonesian Petroleum Association, 18th Annual Convention. 107-138. LETOUZEY, J., WERNER, P. & MARTY, A. 1990. Fault reactivation and structural inversion. Backarc and intraplate compressive deformations. Examples of the eastern Sunda shelf (Indonesia). Tectonophysics, 183, 341-362. LOWE, D. R. 1976. Grain flow and grain flow deposits. Journal of Sedimentary Petrology, 46, 188-199. 1979. Sediment gravity flows: their classification and some problems of application to natural flows and deposits. In: DOYLE, L.J. & PILKEY, O.H. (eds) Geology of Continental Slopes. Society of Economic Paleontologists and Mineralogists, Special Publication, 27, 75-82. 1982. Sedimentary Gravity Flows: II. Depositional models with special reference to the deposits of high-density turbidity currents. Journal of Sedimentary Petrology, 52, 279-297. MCILREATH, I. A. & JAMES,N. P. 1984. Carbonate Slopes. In: WALKER,R. G. (ed.) Facies Models. Geoscience Canada, 12, 245-257. MARTINI, I. P., RAU, A. & TONGIORGI, M. 1986. Syntectonic sedimentation in a middle Triassic rift, northern Apennines, Italy. Sedimentary Geology, 47, 191-219. MIDDLETON, G. V. & HAMPTON,M. A. 1976. Subaqueous sediment transport and deposition by sediment gravity flows. In: STANLEY,D. J. & SWaFT, D. J. P. (eds) Marine Sediment Transport and Environmental Management. John Wiley and Sons, New York, 245-257. Moss, S. 1994. Tertiary basins of Kalimantan - A presurvey report. University of London Southeast Asia Research Group, Report, 127. MULLINS, H. T. 1983. Modem carbonate slopes and basins of the Bahamas. In: COOK, H. E., HINE, A. C. & MULLINS, H. T. (eds) Platform Margins and Deepwater Carbonates. Society of Economic -
-
388
-
M . E . J . WILSON • D. W. J. BOSENCE
Paleontologists and Mineralogists, Short Course Notes, 12, 4.1-4.138. & COOK, H. E. 1986. Carbonate apron models: alternatives to the submarine fan model for paleoenvironmental analysis and hydrocarbon exploration. Sedimentary Geology, 48, 37-79. -& VAN BUREN, H. M. 1979. Modem modified carbonate grain flow deposit. Journal of Sedimentary Petrology, 49, 747-752. --, HEATH, K. C., VAN BUREN, H. M. & HINE, A. C. 1984. Anatomy of modem open-ocean carbonate slope: northern Little Bahama Bank. Sedimentology, 31, 141-168. PACKHAM,G. H. 1990. Plate motions and Southeast Asia: some tectonic consequences for basin development. Southeast Asia Petroleum Exploration Society Proceedings, 9, 55-68. PARKINSON, C. D. 1991. The Petrology, Structure and Geologic History of the Metamorphic rocks of Central Sulawesi, Indonesia. PhD Thesis, University of London. PATRIAT, P. & ACHACHE, J. 1984. India-Eurasia collision chronology has implications for crustal shortening and driving mechanisms of plates. Nature, 311, 615-621. PIETERS, P. E., TRAIL, D. S. & SUPRIATNA, S. 1987. Correlation of early Tertiary rocks across Kalimantan. In: Proceedings of the Indonesian Petroleum Association, 16th Annual Convention. 291-305. PFEIL, R. W. & READ, J. E 1980. Cambrian carbonate platform margin facies, Shady dolomite, southwestern Virginia, U.S.A. Journal of Sedimentary Petrology, 50, 91-116. PICKERING, K., STOW, O., WATSON, M. t~ HISCOTT, R. 1986. Deep-water facies, processes and models: A review and classification scheme for modem and ancient sediments. Earth Science Reviews, 23, 75-174. PLAYFORD, P. E., HURLEY, N. F., KERANS, C. t~ MIDDLETON, M. F. 1989. Reefal platform development, Devonian of the Canning basin, western Australia. In: CREVELLO,P. D., WILSON, J. L., SARG, F. & READ, J. E (eds) Controls on Carbonate Platform and Basin Development. Society of Economic Paleontologists and Mineralogists, Special Publication, 44, 188-202. POSTMA,G., NEMEC,W. & KLEINSPEHN,K. L. 1988. Large floating clasts in turbidites: a mechanism for their emplacement. Sedimentary Geology, 58, 47-61. PRIOR, D. B. & COLEMAN,J. M. 1982. Active slides and flows in underconsolidated marine sediments on the slopes of the Mississippi Delta. In: SAXOV, S. 8¢ NIEUWENHUIS, J. K. (eds) Marine Slides and Other Mass Movements. NATO Scientific Affairs Division, 37-79. RANGIN, C., JOLIVET,L. & PUBELLIER,M. 1990. A simple model for the tectonic evolution of Southeast Asia and Indonesian region for the past 43 Ma. Bulletin de la Societig Ggologique de France, 8, 889-905. REIJMER, J. J. G. & EVERAARS,J. S. L. 1991. Carbonate platform facies reflected in carbonate basin facies (Triassic, Northern Calcareous Alps, Austria). Facies, 25, 253-278. -
--,
SCHLAGER,W. & DROXLER, A. W. 1988. Site 632: Pliocene-Pleistocene sedimentation cycles in a Bahamian basin. Proceedings of the Ocean Drilling Program, Scientific Results, 101,213-220. --, --, BOSSCHER, H., BEETS, C. J. & MCNEILL, D. E 1992. Pliocene/Pleistocene platform facies transition recorded in calciturbidites (Exuma Sound, Bahamas). Sedimentary Geology, 78, 171-179. , TEN KATE,W. G. H. Z., SPRENGER,A. & SCHLAGER, W. 1991. Calciturbidite composition related to exposure and flooding of a carbonate platform (Triassic, Eastern Alps). Sedimentology, 38, 10591074. RODINE, J. & JOHNSON,A. M. 1976. The ability of debris, heavily freighted with coarse clastic materials, to flow on gentle slopes. Sedimentology, 23, 213-234. Ru, K. & PIGOTT, J. D. 1986. Episodic rifting and subsidence in the South China Sea. AAPG Bulletin, 70, 1136-1155. SALLER, A., ARMIN. R., ICHRAM, L. O. & GLENNSULLIVAN, C. 1992. Sequence stratigraphy of upper Eocene and Oligocene limestones, Teweh area, central Kalimantan. In: Proceedings of the Indonesian Petroleum Association, 21st Annual Convention. 69-92. SARG, J. E 1988. Carbonate sequence stratigraphy. In: WmGUS, C. K., HASTINGS, B. S., KENDALL,C. G. ST. C., POSAMENTIER, H. W., Ross, C. A. & VAN WAGONER, J. C. (eds) Sea-level Changes An Integrated Approach. Society of Economic Paleontologists and Mineralogists, Special Publication, 42, 155-181. SCHLAGER, W. t~ CAMBER, O. 1986. Submarine slope angles, drowning unconformities, and self-erosion of limestone escarpments. Geology, 14, 762-765. -• CHERMAK,A. 1979. Sediment facies of platformbasin transition, Tongue of the ocean, Bahamas. In: DOYLE, L. J. & PILKEY, O. H. (eds) Geology of Continental Slopes. Society of Economic Paleontologists and Mineralogists, Special Publication, 27, 193-208. SILVER, E. A., JOYODIWIRYO, Y. & MCCAFFREY,R. 1978. Gravity results and emplacement geometry of the Sulawesi ultramafic belt, Indonesia. Geology, 6, 537-531. S1MANDJUNTAK,Z. O. 1990. Sedimentology and tectonics of the collision complex in the East Arm of Sulawesi, Indonesia. Geologi Indonesia: Journal of the Indonesian Association of Geologists, 13, 1-35. SITUMORANG, B. 1987. Seismic stratigraphy of the Makassar basin. In: Lemigas Scientific Contribution, 3-37. SMITH, R. B. & SILVER,E. A. 1991. Geology of a Miocene collision complex, Buton, eastern Indonesia. Geological Society of America Bulletin, 103, 660-678. SUKAMTO,R. 1975. The structure of Sulawesi in the light of plate tectonics. In: Proceedings of the Regional Conference on the Geology and Mineral Resources of SE Asia, Jakarta. 1-25. -1982. Geologi lembar Pangkajene dan Watampone bagian barat, Sulawesi. Geological Research and Development Centre, Bandung.
TERTIARY EVOLUTION OF S SULAWESI & SUPRIATNA, S. 1982. Geologi lembar Ujung Pandang, Benteng dan Sinjai quadrangles, Sulawesi. Geological Research and Development Centre, Bandung. TUCKER, M. E. & WRIGHT, V. E 1990. Carbonate Sedimentology. Blackwell Scientific Publications. T W ~ L , W. W., DAVIS, R. G & McDOWELL, H. G. 1986. Miocene carbonate shelf margin, Bali-Flores Sea, Indonesia. In: Proceedings of the Indonesian Petroleum Association, 15th Annual Convention, 124-140. VAN DE WEERD, A. A. & ARMIN, R. A. 1992. Origin and evolution of the Tertiary hydrocarbon-bearing basins in Kalimantan (Borneo), Indonesia. AAPG Bulletin, 76, 1778-1803. , --, MAHADI, S. & WARE, P. L. B. 1987. Geologic setting of the Kerendan gas and condensate discovery, Tertiary sedimentation and paleogeography of the northwestern part of the Kutei basin, Kalimantan, Indonesia. In: Proceedings of the Indonesian Petroleum Association, 16th Annual Convention. 317-338. VAN LEEUWEN, T. M. 1981. The geology of southwest Sulawesi with special reference to the Biru area. In: BARBER, A. J. & WmYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication, 2, 277-304. WAIN, T. & BEROD, B. 1989. The tectonic framework and paleogeographic evolution of the upper Kutei - -
389
basin. In: Proceedings of the Indonesian Petroleum Association, 18th Annual Convention. 55-78. WATTS, K. F. 1987. Triassic carbonate submarine fans along the Arabian platform margin, Sumeini Group, Oman. Sedimentology, 34, 43-71. & BLOME, C. D. 1990. Evolution of the Arabian carbonate platform margin slope and its response to orogenic closing of a Cretaceous ocean basin, Oman. In: TUCKER,M. E., WILSON,J. L., CREVELLO, P. D., SARG, J. R. & READ, J. E (eds) Carbonate Platforms: Facies, Sequences and Evolution. International Association of Sedimentologists, Special Publication, 9, 291-323. WEISSEL, J. K. 1980. Evidence for Eocene oceanic crust in the Celebes Basin. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. American Geophysical Union, Geophysical Monograph, 23, 37-47. WILLIAMS, P. R., JOHNSTON, C. R., ALMOND, R. A. & SIMAMORA, W. H. 1988. Late Cretaceous to early Tertiary structural elements of west Kalimantan. Tectonophysics, 148, 279-297. WOODS, B. G. M. 1985. The mechanics of progressive deformation in crustal plates - A working model for Southeast Asia. Geological Society of Malaysia Bulletin, 18, 55-99. YUWONO, W. S, BELLON, H., SOERIA-ATMADJA, V. & MAURY, R. C. 1985. Neogene and Pleistocene volcanism in South Sulawesi. Proceedings Ikatan Ahli Geologi Indonesia, 14, 169-179. -
-
Tertiary Tectonic and magmatic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for a Miocene continent-continent collision STEVEN
C. B E R G M A N
1, D A N A
& RICHARD
Q. C O F F I E L D 2, J A M E S A. G A R R A R D
E T A L B O T l, 3
4
1ARCO Exploration & Production Technology, Exploration Research & Technical Services, 2300 W. Plano Pkwy., Plano TX 75075-8499, USA. e-mail
[email protected] 2 Atlantic Richfield Indonesia, The Landmark Center, Tower B, Jl. Jend Sudirman, Kav. 70A, Jakarta, Indonesia 12910. e-mail
[email protected] 3Now at: KT Geoservices, 661 N. Plano Rd, Suite 317, Richardson TX 75081, USA 4 ARCO International Oil & Gas Co., Asia-Pacific New Ventures Exploration, 2300 Plano Pkwy., Plano TX 75075-8499, USA. Now at: ARCO Alaska, 700 G Street, Anchorage AK 99701, USA Abstract: New field and laboratory data from western Sulawesi, Indonesia, integrated with available data establish its Late Cenozoic igneous framework and a new model for its tectonic evolution. Western Sulawesi contains three major Neogene N-S-trending tectonic domains (from W to E): (1) an active foldbelt, in which Pliocene and Miocene volcanogenic rocks are involved in W-vergent thrusting which extends into the Makassar Strait; (2) a central belt comprised of a deformed submarine Miocene volcanoplutonic arc built on an Oligocene-Eocene clastic and carbonate platform with Latimojong Mesozoic basement metamorphic and sedimentary rocks thrust over its eastern margin on W-vergent faults; and (3) an accreted Cretaceous-Palaeogene(?) ophiolite (Lamasi Complex) between the Latimojong basement block and Bone Bay. The Lamasi Complex ophiolite includes dioritic plutons, basaltic sheeted dykes, pillow lavas, greenstones, tufts and volcanic agglomerates with depleted (MORB-like) Sr & Nd isotope and REE characteristics of probable normal oceanic crust with possible subduction-related or back-arc affinity. New K-Ar, 4°Ar-39Ar, Rb-Sr, and Nd-Sm isotope data suggest Cretaceous to Eocene crystallization and Oligocene to Miocene obduction. Late Miocene to Pliocene extrusive and intrusive rocks form a cogenetic volcanoplutonic complex of calc-alkalic to mildly alkalic, potassic, and shoshonitic felsic and mafic magmatic rocks of bimodal composition which were erupted and intruded during a short episode of Middle Miocene to Pliocene (3-18 Ma) lithospheric melting. Based on new Rb-Sr, Nd-Sm, and U-Pb isotope, and major and trace element geochemical data, parental source rocks of the Miocene melts were Late Proterozoic to Early Palaeozoic crustal and mantle lithospheric assemblages which became heated and melted owing to a continent-continent collision in which west-vergent continental lithosphere derived from the Australian-New Guinea plate was subducted beneath eastern-most Sundaland. The timing of this magmatism and subsequent cooling and denudation history are constrained by 113 new K-Ar, 4°Ar-39Ar, and fission track ages. The new tectonic model differs significantly from previous models: 'the Makassar Strait is now interpreted as a foreland basin bound on both sides by converging Neogene thrust belts, in contrast to previous models suggesting Late Tertiary oceanic spreading or continental rifting. West-vergent obduction of a pre-Eocene oceanic, primitive arc, or back-arc crust onto western Sulawesi occurred during late Oligocene to Miocene times. The Late Miocene western Sulawesi magmatic arc is envisioned as a continent-continent collision product, in contrast to previous models involving a normal ocean--continent or ocean-ocean subduction-related magmatic arc (west or east vergent) or post-subduction rifting. The east Sulawesi ophiolite extends into western Sulawesi, suggesting that Bone Bay resulted from collapse of the overthickened Miocene orogen. The new tectonic model illustrates the central role western Sulawesi plays in unlocking the complex evolution of Indonesia as well as the temporal and magmatic details of a continent-continent collision zone.
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. t06, pp. 391-429.
391
392
s.C. BERGMAN ET AL.
The objective of this paper is to integrate new and published geological, petrological, isotopic, stratigraphic, structural and geochemical data for western Sulawesi in order to test available models and formulate a new tectonic model for the region. Sulawesi is located in a complex tectonic position, at the intersection of three major lithospheric plates: the westward-moving Pacific Plate, the northward-moving Australian Plate and the relatively stationary Eurasian Plate (Fig. 1). The complex plate tectonic position of Sulawesi is manifested in a varied Tertiary structural and stratigraphic record (Figs 1-3). The south arm of Sulawesi is dominated by Miocene and younger volcanic and plutonic rocks forming a magmatic belt that many workers have regarded as a subduction-related volcanic arc involving a westwarddipping (Sukamto 1978; Hamilton 1979), or eastdipping (Katili 1978) oceanic plate. Some recent workers interpret the magmatic arc as a postcollisional rift-related magmatic belt (Yuwono et al. 1988b; Leterrier et al. 1990; Kalavieris et al. 1992; Priadi et al. 1994). The present study presents new field and laboratory evidence based on three seasons of geological research (19901993) associated with the exploration for hydrocarbons in the Kalosi Production Sharing Contract (PSC) (see Coffield et al. 1993 for a summary of the petroleum systems framework). The results demonstrate that previous 'conventional' tectonic models for the Late Tertiary evolution of western Sulawesi and the adjacent Makassar Strait require significant modification. The authors propose that the Miocene to Recent tectonic framework of western Sulawesi and Makassar Strait was dominated by compressional continent-continent collision processes.
Geological and tectonic framework Sulawesi is located at the southeast limit of the Sunda Platform crustal domain and north and west of the Australian Continental Platform, its probably derivative fragments forming Irian Jaya, Sula, Buru, Seram and the Tukang Besi Platform. Sulawesi formed along the Oligocene-Miocene collision zone between the Eurasian Plate and micro-continental fragments derived from the Indian-Australian Plate (Hamilton 1979; Hutchison 1989a; Rangin et al. 1990; Daly et al. 1991). The four arms of Sulawesi and adjacent islands form four distinct megatectonic provinces. The northern arm is composed of late Palaeogene to Neogene subduction related volcanic arc rocks resulting from the west-dipping subduction of the Molucca Sea Plate (Jezek et al. 1981; Bellon & Rangin 1991). The east and southeast arms are
composed of Mesozoic and younger allochthonous metamorphic and ophiolitic rocks which were obducted during the Oligocene epoch (Parkinson 1991). The south arm is dominated by Miocene and younger volcanic and plutonic rocks which form a magmatic belt superimposed on the Mesozoic basement of the southeastern margin of Sundaland (Katili 1978; Silver et al. 1983a, b). The fourth megatectonic province contains Late Palaeozoic and Mesozoic Australian-derived microcontinents which have been accreted to the eastern margin of Sulawesi, comprising Banggai, Sula, Buton, Kabaena and Tukang Besi, among other islands. The present work concerns parts of the south and north arms of Sulawesi near their boundary in an area approximately bound by Mamuju, Palopo, Parepare and Barulatong (Fig: 2), which will be referred to as 'western Sulawesi' for simplicity. The present-day tectonic framework of western Sulawesi is dominated by active uplift, and compressional and strike-slip faulting. The region is seismically active and nearly all historic earthquake hypocentres are < 50 km below the surface (cf. USGS National Earthquake Information Center, Boulder, Colorado, PDE database which shows > 40 shallow events between November 1929 and July 1994 with Richter magnitudes 3-5 in the Mamasa-Parepare areas). The study area is 500-800 km north of the Sunda-Banda volcanic arc/subduction zone in which the Australian Plate is being subducted beneath Sumbawa, Flores and the Flores Sea. Miocene and younger collision of Australian continental crust has resulted in the accretion of Timor to the upper plate of Eurasia and a south-dipping thrust zone has developed north of Flores in response to the continued compression. Several major NNW trending strike-slip faults cut through the south arm; these faults have traditionally been interpreted as sinistral from field-based structural observations (Berry & Grady 1987) and LandSat imagery interpretations (M. Crawford pers. comm. 1991), although there is evidence for right-lateral offset on some faults. Our recent work has shown that some of these strike-slip faults are transpressional and possess recent compressional offset. Late Tertiary compressional structures include oblique-trending folds and shallow-angle thrust faults. The presence of several Quaternary basins within and adjacent to the south arm of Sulawesi indicate the presence o f Late Tertiary subsidence, possibly representing piggy-back basins adjacent to thrust sheets, or due to transtensional processes. Geological data relevant to the study area include published 1:1 000000 and 1:250000 scale geological maps (Djuri & Sudjatmiko 1974; Ratman 1976; Sukamto 1975, 1982; Ratman & Atmawinata 1988) and various literature information (e.g. van
Fig. 1. Present-day tectonic elements of Indonesia showing the distribution of active trenches, Quaternary subduction-related volcanic arcs, and major structures. L~
394
S.C. BERGMAN ET AL.
Bemmelen 1949; Audley-Charles 1977; Hamilton 1979; Sasajima et al. 1980; Van Leeuwen 1981; Garrard 1989; Hasan et al. 1991; Coffield et al. 1993), which are summarized in a geological sketch map in Fig. 2 and in a schematic stratigraphic section in Fig. 3. Geological studies in western Sulawesi include early Dutch work initiated during the first part of this century (Abendanon 1915; t'Hoen & Ziegler 1917; Reyzer 1920; Sax 1931a, b; Sung 1948). More recent pubfished information on geochronology and petrology of volcanic and plutonic rocks of western Sulawesi include Sasajima et al. (1980), Van Leeuwen ( 1981 ), Sukamto & Simandjuntak (1983), Leterrier et al. (1990), Kalavieris et al. (1992), Yuwono 1987, Yuwono et al. (1988), Van Leeuwen et al. (1994) and Priadi et al. (1993, 1994). The stratigraphy of western Sulawesi is dominated by Neogene rocks but also includes formations as old as Jurassic (Fig. 3). Five main lithostratigraphic sequences are recognized (Garrard et al. 1992; Coffield et al. 1993): (1) Mesozoic (mainly Late Cretaceous) metasedimentary rocks of the Latimojong, Balangbaru and Marada Formations, which include flysch deposits believed to have formed in a forearc basin setting, and ophiolites of the Lamasi Complex CretaceousPalaeogene?); (2) a Palaeogene (Middle to Late Eocene) syn-rift sequence composed of siliciclastic, coal, volcanic and carbonate sedimentary deposits of the Toraja and Malawa Formations; (3) an Eocene to Middle Miocene carbonate post-rift sequence including the Makale and Tonasa Formations; (4) a Middle Miocene to Pliocene basaltic to dacitic volcano-plutonic and epiclastic sedimentary sequence including the Enrekang and Camba Volcanic series and the Buakayu Formation, and consanguineous granitic to gabbroic intrusives; and (5) Pliocene and younger synorogenic, nonmarine to upper bathyal sedimentary deposits including the Walanae Formation, formed during widespread thrusting, regional uplift and erosion. These five sequences occur in a variety of structural positions. The older sequences are typically structurally highest in the sections and in the hinterland to the east; the younger sequences occur in structurally lower positions than in the sequences towards the west or foreland. Importantly, the western promontory of the south arm of Sulawesi (near 3°S latitude) is a relatively recent geomorphological product, resulting from uplift due to Neogene thrusting which extends for 200-300 km from the Majene foldbelt in the west to the Kalosi foldbelt in the east. Makassar Strait
The Makassar Strait is characterized by a central
deep-water trough with depths as great as 15002500 m. In contrast to most of the Makassar Strait margins which contain a broad shelf, the western Sulawesi margin north of Majene possesses a very narrow shelf (< 10-20 km) and steep slope. The tectonic evolution of the Makassar Strait has been the subject of considerable debate since Alfred Russell Wallace delineated his Australian-Asian faunal boundary along its axial trend in 1858 (for a Wallace Line review, see George 1981). Most workers have incorporated phases of Palaeogene, Miocene or Quaternary rifting which resulted from NW-SE extension along the Makassar Strait (Katili 1971, 1975, 1978; Carey 1976; Audley-Charles 1977; Hamilton 1978, 1979; Burollet & Salle 1981; Situmorang 1982a, b, 1984; Faugeres et al. 1989; Shaw & Packham 1992; Effendi 1993, among many others)i Rifting is thought to have involved oceanic (Hamilton 1979) or continental crust (Burollet & Salle 1981). Malecek et al. (1993) proposed that the Strait is underlain by trapped Cretaceous oceanic crust. Alternative models involving a foredeep to the western Sulawesi orogen have been advanced by van Bemmelen (1949), although this interpretation has largely been neglected by recent workers, until now. Most recent interpretations tend to draw long continuous normal faults down the axis of the Strait, or attempt to close up the basin by matching the Sulawesi western coastline with the eastern coastline of Borneo (Katili 1978). Whereas this procedure may work well for passive margins such as West Africa and east South America, the recently active (15 Ma to present) fold and thrust belt of western Sulawesi has produced a rapidly evolving Neogene coastline which does not lend itself to coastline matching. Seismic sections orientated NW-SE show the salient tectonostratigraphic sequences and structural styles. Whereas Eocene extensional faulting is recorded in certain deeper parts of the sections, Oligocene passive platform conditions are present, and Late Miocene to Pliocene fold and thrust belt development is apparent (Fig. 4). The leading edge of the thrust front is evident on the seismic section, as is a series of stacked thrust sheets with a series of overlying piggy-back basins. These piggy-back basins contain distinct stratigraphic packages which record the movement of the underlying thrust sheets. Contemporaneous sedimentary packages in front of the leading thrust are entirely parallel bedded and undeformed whereas those in the piggy-back basins contain diverging reflection packages which onlap the adjacent growing ramp anticlines. The Samarinda anticlinorium, a fold and thrust belt on the opposite (western) side of the northern Makassar Strait mirrors the Majene foldbelt. The Samarinda anti-
MIOCENE W SULAWESI CONTINENT
395
COLLISION
120" 00'
KALOSI
PSC.
03*
03 ° 30'
0
10
20
i .....
30 Kilt. !
II
i
II
i111 ii i ii1,1 iii
t25"
t2~
LEGEND : 04* O0'
QUATERNARYALL~
~i ~PARE "1
1
VOLCANICS
1
GRANITIC OTHER
PARE
PMOCENE WALANAE FM.
1
VOLC/LNICS& CLAt~I'ICS L EOCENE - M. MIOCENE
INTRUSIVES
0 It¢ - L EOCENETORAJA FM.
INDEX MAP llllllll,i
i PRE TERTIARYBASEIkrcNT LATIMOJOI~
I
i
I
I
i~
i
LAMASl COMPLEX OPHIOLITE
AGE ?
..........
I
,,~
I
,,i
Fig. 2. Generalized geological map of western Sulawesi.
clinorium is a Neogene east-vergent compressional trend which has developed in the Kutei Basin along strike to the north of the west-vergent Meratus Mountains orogen. The origin of these structures
is controversial and calls upon inversion of preexisting rift structures (Biantoro et al. 1992) or gravity sliding from the hinterland of the Kutei Basin (Ott 1987; van de Weerd & Armin 1992).
S.C. BERGMAN ET AL.
396
STRATIGRAPHY FORMATION
LITHOLOGY
TECTONICS
Ma
0 ,==.
40
:)CENE
~- sroc. p, ~C~.Et •
WALANAE
FM.
NEW K-Ar ages 10 ,
,
.
,
,
,
.
.
.
,
.
SYN.__q O R O G E N I C
F
PEAK
FORELAND BASIN/ FOLD & THRUST BELT INITIATION
15-
20-
OPHIOLITE OBDUCTION
"! I~,
,
,
,
>25-
POST-RIFT
IT"
< m
1-30-
rr UJ
1-35.
FM ( MALAWA )
TORAJA
40.
SYN-RIFT
4,5'
50.
55
60
LATE
PRE-RIFT ~
m
P
~L..--"~
J__--65.
VV V J VVV
~,r~
.........
VVV
+
+'1~
Fig. 3. Stratigraphic column of western Sulawesi.
Fig. 4. Interpreted E-W seismic section 201 through the Makassar Strait (after Coffield et al. ]993).
,
,
~T
'i
I
~- /
°F-
~ ,.,.~
°ILl I @
° Z _1
MIOCENE W SULAWESICONTINENT COLLISION Herein, the Makassar Strait is interpreted as a present-day foreland basin bound on both sides by converging Neogene fold and thrust belts.
Nature and distribution of CretaceousPalaeogene ? and Neogene igneous rocks Two groups of igneous rocks are distinguished on the basis of field relations, age, composition, petrology and isotope and trace element geochemistry: Cretaceous-Palaeogene(?) ophiolitic rocks of the Lamasi Complex and MiocenePliocene igneous rocks of the Camba-EnrekangMamasa Complex. Lamasi Complex ophiolitic rocks are intensely deformed, typically metamorphosed and mainly bounded by thrust faults. Associated sedimentary deposits include red cherts and mudstones. The Lamasi Complex ophiolites include a variety of intrusive and extrusive lithologies. Intrusives such as sheared diorite plutons (> 20 km diameter), amphibolite gneisses (metagabbros) and sheeted basalt dyke complexes have been recognized. Extrusives include basaltic pillow lavas, basalt to andesite lava flows and agglomerates, and andesite, dacite and rhyolitic pyroclastic deposits comprising block breccias, tuff breccias and tuffs, Neogene igneous rocks of the CambaEnrekang-Mamasa Complex and derivative volcaniclastic deposits cover more~han 75% of the surface of western Sulawesi (Fig. 2). Volcanic outcrops consist of massive lavas and pyroclastic sections with > 5 km thicknesses; plutons and laccoliths of batholithic dimensions (> 50-100 km diameter) are intimately associated with this volcanic assemblage. Volcanic and intrusive sequences are variably deformed; many volcanic sequences are steeply dipping to vertical and batholithic-scale laccoliths are detached and bounded by thrust faults. The volcanic rocks range from basalt to rhyolite in composition, but are dominated by basaltic to dacitic stratovolcano complexes, lava domes and flows, and pyroclastic sequences. Plutonic rocks range from gabbroic and dioritic stocks, sills and dykes to granitic-quartz monzonitic to monzodioritic stocks, plutons and large laccoliths. Lamprophyre dykes are common. The age of the Camba-Enrekang-Mamasa igneous sequences is mainly Middle to Late Miocene (averaging 8 _+2 Ma), although the total age range observed is 3-18 Ma (see below).
Analytical results and interpretation Petrology A summary of lithological and location data for intrusive and extrusive rocks included in the
397
present study is provided in Table 1. (For more details see table 1 of Supplementary Publication No. SUP 18099 (11 pp) available from the Society Library or the British Library Document Supply Centre, Boston Spa, Wetherby, W. Yorks LS23 7BQ, UK.) Analytical methods and details are provided in the Appendix. The locations of various types of analytical data are summarized in Figs 5-7. Modal analyses of eight selected phaneritic Miocene intrusives of the Camba-EnrekangMamasa Complex (Table 2) range from quartz monzonite and monzonite to diorite in the IUGC classification scheme. The Miocene Camba-Enrekang-Mamasa Complex comprises basaltic and gabbroic lithologies rich in augite, through andesites/diorites rich in augite and amphibole with rare enstatite, to granitic-monzonitic lithologies relatively poor in quartz and rich in K-feldspar, albite and biotite. The most primitive basaltic lithologies are phlogopitebearing lamprophyres (olivine minettes) and alkali gabbros. Volcanic rocks are dominantly subalkaline to shoshonitic, with leucite or analcime observed in many trachytes or trachyandesites. Palaeogene ophiolites of the Lamasi CoMplex include a variety of intrusive and extrusive lithologies. Intrusive rocks include basalt, diorite and gabbro of varying degrees of metamorphism (greenschist-amphibolite grade). Extrusives are dominantly basalt, andesite and dacite.
Major and trace element geochemistry Major and trace element compositions of 59 samples define specific compositional groups and trends for the Lamasi and Miocene complexes, illustrated in Harker variation diagrams and compositional plots (Figs 8-14) and REE and spider diagrams (Figs. 11-12) (selected averages are listed in Table 3; detailed analytical data are in table 2 of Supplementary Publication SUP 18099 (see above)). The Miocene igneous rocks range from basalt to rhyolite in the TAS classification scheme (Fig. 8), and are dominated by high K calc-alkalic and shoshonitic series (see Bergman 1987 for an overview of K-rich alkaline igneous rocks). On an AFM diagram, the Miocene igneous suite defines a calc-alkaline trend (Fig. 10). The presence of basaltic trachyandesites, trachyandesite, and borderline trachyte-dacite compositions indicates an alkaline affinity of the intrusive and extrusive units. Compositions are dominated by relatively alkaline basaltic and dacitic rocks, with a conspicuous lack of andesites. Miocene intrusive and extrusive lithologies are geochemically similar and define the same trends on major element cross plots (Fig. 9). Compositions of Miocene intrusive and extrusive rocks from a
lithology
phlogopite olivine basalt dyke arkosic sandstone alkali gabbro sill syenogabbro sill basalt crystal lithic lapilli tuff basalt crystal lithic lapilli tuff graded basalt crystal lithic lapilli tuff trachyandesite crystal lithic lapilli tuff monzodiorite pluton quartz monzodiorite pluton andalusite siltstone homfels basalt block in flow breccia quartz monzodiorite pluton monzodiorite stock dacite porphyry lava dacite porphyry lava coarse crystal lithic dacite tuff crystal lithic dacite lapilli tuff welded crystal lithic dacite lapilli tuff dacite porphyry lava quartz monzodiorite pluton quartz monzodiorite pluton
sample
90SUL1 90SUL2 90SUL3 90SUIAa 90SUL5 90SUL6a 90SUL6b 90SUL7 90SUL8 90SUL9 90SUL10a 90SUL 11 90SUL12 90SULI3 90SUL14 90SUL15 90SUL16a 90SUL 18a 90SUL18b 90SUL20 90SUL21 a 90SUL21 b
Table 1. Sample location summary
Awang R. Awang R. Sipe Hill Tou R. Subing R. Sadang R. Sadang R. Sadang R. Palopo pluton Palopo pluton Palopo pluton aureole Bambulu R. Palopo pluton Rantepao stock Mt. Tambakkuku Mt Tambakkuku Mt Tambakkuku Batu R. Batu R. Loca R. Mamasa Mamasa
locality Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Cretaceous Palaeog./Cret Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene
age 290 295 290 290 730 830 830 830 1070 1050 250 10 475 835 400 365 400 100 100 120 1100 1100
(m)
elev
3 3 3 3 3 3 3 3 2 2 2 2 2 2 3 3 3 3 3 3 2 2
deg 11.9 11.8 11.7 11.3 9.9 9.9 9.9 7.6 57.4 56.8 57.1 50.6 57.5 57.2 30.7 28.9 28.9 33.2 33.2 34.0 56.4 56.4
min
lat S
119 119 119 119 119 119 119 119 120 120 120 120 120 119 119 119 119 119 119 119 119 119
deg
43.9 43.8 43.6 43.7 44.7 45.8 45.8 46.9 4.2 4.9 7.5 7.2 5.7 58.9 47.0 47.1 47.6 43.5 43.5 42.3 22.7 22.7
min
long E
bp b b
bh h b bs bp
b bh bh
b b
K
A
1 1 1 1 1 1 1 l 1
i i i i i i i i
i
G
Analyses
z a
sn
P
az
snp
a
Z
sn
sn
az z
sn
a
sn snp
sn
z a
azs
sn
z sn
z
sn
a sn
sn sn sn
sn
I
~D
90SUL22a 90SUL22b 90SUL23 90SUL24 90SUL25 90SUL26a 90SUL26b 90SUL26c 90SUL26d 90SUL27 90SUL28 RAG/90/73 RAG/90/74 92SUL1 92SUL2 92SUL3 92SUIA 92SUL5 92SUL6 92SUL7 92SUL8 ND92/29 RAG92-004 RAG92-006 RAG92-008
monzonite pluton fluvial alluvium draining pluton dacite porphyry lava vitric rhyolite lava quartz monzodiorite pluton andesite crystal lithic lapilli tuff andesite crystal lithic lapilli tuff andesite block in tuff breccia andesite block in tuff breccia trachyte porphyry dome quartz monzonite gneiss clast in Walanae fine grained volcaniclastic sandstone siliciclastic sandstone vertical dacite lava diorite/gabbro xenolith in dacite lava vertical dacite lava thrust sheet horizontal dacite lava horizontal dacite lava horizontal dacite tuff dacite pyroclastic flow lapilli tuff vertical Eocene sandstones dacite lava dome tuffaceous sandstone trachyte/shoshonite tuff breccia leucite shoshonite lapilli tuff
Balakalua Balakalua Buntubuntu Sumarorong Polewali pluton Pare Pare stratavolcano Pare Pare stratavolcano Pare Pare stratavolcano Pare Pare stratavolcano Allakuang quarry Enrekang Sadang River Sadang River Biloka River Biloka River Biloka River Biloka River gorge Benteng dam Waru Balu Lambou/Kalosi Tirasa Mtn/Benteng end of New Rangas Rd end of New Rangas Rd end of New Rangas Rd
Miocene Recent Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Pliocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Miocene Eocene Miocene Miocene Miocene Miocene
1035 1035 975 850 70 60 60 60 60 100 200 300 300 189 190 185 160 30 20 55 270 425 400 5 5
59.1 59.1 0.9 10.1 24.2 55.2 55.2 55.2 55.2 59.4 33.1 12.7 12.6 34.6 33.9 34.4 38.2 41.3 34.3 42.1 19.8 37.4 37.6 37.6 37.6
119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 119 118 118 118 20.4 20.4 18.9 19.8 20.8 42.5 42.5 42.5 42.5 47.8 45.0 43.7 43.8 40.6 41.9 40.5 41.7 4O.0 34.8 37.3 49.4 37.9 49.2 49.2 49.2 z a a
sn sn
snp
s
s
b
b b
b
b
a
snp
az
a
a
Z
Z
b b b b
aZS
snp
sn
sn
bh b
b b b
400
s . c . BERGMAN ET AL.
Table 2. Modal compositions of selected Miocene intrusive rocks sample
quartz (vol%)
plagioclase
K-feldspar
4 7 11 2 11 4 5 22
61 50 55 63 49 45 38 41
17 24 14 10 26 33 27 32
90SUL8 90SUL9 90SUL12 90SUL13 90SUL21 a 90SUL22a 90SUL25 90SUL28
biotite
hornblende
augite
6 3 8 9 19
Normalized volume percentages based on c. 500 points using a 1 mm grid.
exhibit flat to LREE depleted patterns similar to N-type MORB and possess spider diagram patterns similar to MORB and supra-subduction zone rocks (SSZ; arc or back-arc) defined by slight negative Nb anomalies.
200-300 km wide area define continuous trends on these variation diagrams, suggesting a cogenetic relationship. Lamas• Complex ophiolites are compositionally distinct from Miocene intrusives and extrusives in terms of major and trace element contents. Lamas• ophiolitic rocks range from basalt to rhyolite in major element composition (Fig. 9), are marginally thole••tic on an AFM plot (Fig. 10), and are depleted in most trace elements compared with the Miocene magmatic units. Lamas• ophiolites
118 `>30' 2 ° 30'
Conventional K - A r and 4°Ar-39Ar geochronology Fifty-three conventional K-Ar ages of biotite, hornblende, plagioclase, san•dine, phlogopite, white
119 °
120 °
I
I
120 ° 30' 2 ° 30'
RAG92-0~
~
+
• 90-21a,b 9022a'0bq~Mamasa 90-2393-15• 93-18b
3°Makassar Strait
(\
90-24•
.....
~ 93RSI79A
93-13
--
251
927~2°_.09o-28
_ 501km
I0
118 ° 30'
25 mi
t
~
• 93-53•93~23~b-48a,b 90-15093"5-03~9933"~7 90-14•90-16aRAG92-018
KaJosi •
,
~
3°
93-45 93-~ O• •93-43 93-44
92-7 \
~
_
~,
93-46
]"92.6• 92.111~2.2~' En rekang • • 9U-1Ua,D ND92/29 92-4 • %.~
/ 0 •
93~193,39
RAG/90/73,74 92-9• 92-8
90"2,~1L~93"5 93-3 93"63~'~ 09~1 "~, "
~,93Rs2°1.~~e A J
90-7
9090"~O090-6a,b
•
• • 93-17 93-11
93-10#93.9
4°
90-11 J 93-3093-33 • qP93-34b 90-1390-9 90-10aJ Palopo 93-210 • 90.8~OA93.2728 , 90-12 " ~ ~ ' Bone 93.2o• -t- 93-3~ ~ 93~.~4o 93-363 II - ~,~
Pare Pare/ 9
90.2%b,0,, .
90-27
,
119 °
Fig. 5. Sample location map for all samples included in the present study.
4°
I 120 °
120
° 30'
119 °
118 ° 30'
120 °
I /
2 ° 30' b 7.3.+0.2 b 2.4.+0.1 sn 5 . 3 _ + 0 . ~
, /
120 ° 30'
I
2 ° 30'
/
Mamuju
/I
L ] ~
Makassar Strait ~
p120.0-.+1.0
~1
b 6.2_+0.2 h 10.7.+1.1 / Hal•DO I ,, ~+,, " ~ = b 5.4_+0.2, h 9.6_+0.9' I h I~'R°+fl~u'~I~ n4-n P 4 b N A p172.4tf,18+2.1plat" am56.9, . . . . . . . . ' ~- I' -~ ' ~ ........ ;~--' '~'-]' A m" • NA ~am 203t~one ~1_ 3 ° wr9 7_904 -/ "-p123.6+1.6, Nl~20.4tl~.2+0.2ptat " -" p1120-25 NA p1137.3tf # a y l + b 10.0-~-0.4 ' "~b 7.6-+0.20 wr 4~7_3.1 I b 11_+0.4 b 12.5_+0.5 & I • b9.6_+0.3 p1201_+9(0.10%K) [ b 12.8_+0.3 NA p1312tf; am 123.2tf / . phi 17.5_+0.4~ Kalosi• "+^ • • I • b 7.7-+0.2 . . . . ~ b 9.8_+0.3 wm ~4t_z A/A p 126.4tf; am 371tf |
" t/ ') - ' (~"
b77+03 b 72±0.3 • ' - ' b 7.0±0.3 i • " a a m asa b 7 3+0 3 • " -" • b 9.8_+0.3
, --'i
• . , , - ' - - ' - " - ' ~ h ~ l m q / ~ n+n-~ b 7.2_0.3, sn 6.7_+0.3 pl 7.8_+0.8, b 7.3_+0.3 ~A~T^~ u o. r_-lJ.,~,r~.o~-u.~ b 8.8_+0.3, pl 7.7_+0.8• b 7.1-+0.3 "~'~ } b7.6.+0.2 _-"I•.b8.1_+0.3 ii b 8.05-_0.31 b 7.4~_0.'~2b~8.1_+0.3 p162_+13 (0.02%K) sn 0
25
I
I
0
50 km i
~
/
I
25 mi
b 6.8+0.3 Pare P ~
4° 118 ° 30'
Enrekang
b 8.5-+0.3
~
7.8_+0.39b4.8_+0.2
I 119 °
I 4° 120 ° 30'
120 °
Fig. 6. K-Ar and 4°Ar-39Ar (A/A) age results location map. Age _+ 10 (Ma) follows phase dated: wm = white mica, pl = plagioclase, see Table 4 for other abbreviations.
118 ° 30'
119 °
3 ° --
120 °
.,~ ~j
2 o30'
/ . . . . J Makassar (,•
Strait
+
1 2 0 ° 30'
I
a 7.3+1.2, z 13.0.+0.6 a 7.2.+1.0 e ....... D ivlu[ll~u 7RR+t3a _mea46.+0.6 .¥ ........ z 8.8_0.5
2 o30'
a 2 1+0 5 z 6 4+0 4 s4 4+0 5 ~ _ . ...... t ...... ; / " D ~ i,-,,-,,-, z 10.8+~6 -~-ea 8.4_+4.3 . . . . ~,v + zS~0.~ ••a8.7-+8.8 -B o n_e T a 6.2_+1.1, z 6.3_+0.4 - z8.6_+0.6 | Bay
~
I I , _
•
~
,
a 20.4_+5.2 . z 4.3_+0.6 a 4.0+_2.9 a 3.8_+1.6 a 2~.6+_6.5
/ a 9.2+_1.4, z 9.8+_0.5
- 1 k % t " / /
t • a 12.9.+3.8, z 92-+10 n 7 (~+1 9 7 . . . . . a 5.8.+0.9 Kalosi • • \ ,t~ u_~,iL, LZ.U_TV.'~ • a4.1-+1.0 a 8.2+9.5 z 93-+9 \ " f-l,,Ja 2.4_+1.8, z 6.8_+0.5 • '..a 5.5_+1.6, z 67.+8 l a4.2_+06 z 5 0 + 0 3 s4.5_+0.9 ~ _~11+~~ Wjl~a6.3+~.3, z128_+12 / \ -A. . . . . . . . . . . . . . . . . . . . ~ a, ,z, ,~ a 4.5+0 9 • / "z 123_+9 \,"v f~ z67_-',~'4• a9.9.+1.6 z69-+5
?
118 ° 30' Fig. 7. Fission abbreviations.
a• 25'r.i
\ z##:,
/
119 ° track age results location
120 ° map. Central
FT age + ]~ (Ma)
fo|]ows
1200 30'
p h a s e d a t e d : see T a b l e 4 f o r
402
S . C . B E R G M A N ET AL.
Table 3. Summary of igneous, metamorphic and sedimentary rock major and trace element geochemical data averages
Lithology
number of samples
Miocene volcanic rocks
36
Miocene intrusive rocks
19
Cretaceous metamorphic rock Lamasi Ophiolite
1 13
Palaeogene clastic rocks
5
Miocene volcanic rocks
36
Miocene intrusive rocks
19
Cretaceous metamorphic rock Lamasi Ophiolite
1 13
Palaeogene clastic rocks
5
Miocene volcanic rocks
36
Miocene intrusive rocks
19
Cretaceous metamorphic rock Lamasi Ophiolite
1 13
Palaeogene clastic rocks
5
TiO 2
A1203
(wt%) Fe203
MgO
CaO
57.05 9.70 59.35 8.51 38.60 53.14 5.36 63.88 18.94
0.70 0.20 0.65 0.27 0.34 1.11 0.49 0.48 0.19
13.83 2.50 14.97 2.00 7.49 15.52 1.86 11.91 4.21
3.42 1.67 1.87 1.45 3.12 3.60 1.11 2.69 1.70
4.02 2.60 3.94 2.26 1.15 5.05 1.68 1.43 0.97
6.43 4.72 5.48 3.22 25.70 8.30 3.72 6.54 10.50
2.15 1.41 3.24 1.42 0.80 6.05 1.80 2.40 1.45
5.23 2.06 4.92 2.39 3.61 9.29 2.24 4.36 1.58
(ppm) Li 19 10 29 12 20 6 5 30 20
Be 6 3 5 1 3 4 1 4 1
Rb 169 104 179 99 21 10 13 35 22
Cs 16 19 7 4 2 2 1 6 4
Sr 610 372 553 266 581 191 182 277 261
Ba 1225 824 996 740 132 104 139 221 134
Y 24 9 19 7 12 28 15 14 11
La 56 35 41 17 13 5 4 12 5
(ppm) Hf 6.4 3.5 4.8 1.5 1.5 3.2 1.9 3.7 1.2
Nb 16.9 9.0 13.5 5.0 7.0 5.4 1.4 7.4 1.8
Ta 1.4 0.8 1.0 0.2 < 0.5 < 0.6
W 5.5 4.6 2.3 1.2 < 1 3.3 0.6 2.0 1.0
Sc 20 11 18 13 10 32 10 11 6
V 129 69 125 87 92 264 127 92 52
Cr 142 107 124 99 34 135 164 181 214
Co 18 11 20 13 9 28 13 14 9
SiO 2 mean std. dev. mean std. dev. mean std. dev. mean std. dev.
mean std. dev. mean std. dev. mean std. dev. mean std. dev.
mean std. dev. mean std. dev. mean std. dev. mean std. dev.
< 0.7
FeO
FeO*
For complete data see table 2 of Supplementary Publication SUP 18099 (see text).
mica or whole rock separates define the cooling histories of the major igneous complexes (summarized in Table 4 and illustrated on a map in Fig. 6; detailed data provided in table 3 of Supplementary Publication SUP 18099 (see above)). High temperature cooling ages (> 300°C) of 38 Miocene and Pliocene Camba-EnrekangMamasa Complex intrusive and extrusive rocks overlap and range from 2.4-17.5 _+0.2-0.4Ma (average 8 +_2.5 Ma). Where multiple phases were dated, concordant ages are generally observed with the exception of two samples which contain hornblende ages significantly older than their biotite ages. This discordance may reflect the higher blocking temperature of hornblende (500°C) than biotite (300°C), suggesting protracted cooling in the interval 500-300°C (see below). Six conventional K-At plagioclase or whole rock ages of seven samples of the Lamasi Complex range from 20-201 Ma. The rocks possessing older ages (162-201 Ma) contain 0.02-0.10 wt% K in mineral separates or whole-rock splits and are interpreted as anomalously high due to excess
radiogenic 4°Ar. In contrast, some Lamasi Complex samples with higher K contents (0.2-0.5 wt%) give younger K-Ar ages in the range 20-24 Ma and are interpreted as reset cooling ages thus dating the ophiolite emplacement. Two plagioclase separates with moderate K contents (0.10-0.14wt%) give Palaeocene or early Cretaceous ages (46_ 3 and 120 _+5 Ma) which are interpreted as a partially reset higher temperature cooling events. Ten 4°Ar-39Ar ages of plagioclase or amphibole separates from five rocks of the Lamasi Complex are summarized in Table 4 (Bergman, unpublished data). Unfortunately, the plagioclase and amphiboles possess very low K contents, yielding a variety of ambiguous ages with a bewildering and poor interpretation potential. Nevertheless, three groups of ages are distinguished: excessively old ages due to excess 4°At, high T (> 400°C) or partially reset cooling ages, and low T (< 300°C) cooling ages. The separates with among the lowest K contents (0.001-0.06 wt%) possess the oldest ages in the range 203-371 Ma. These are regarded as erroneously old due to excess 4°At contamin-
403
MIOCENE W SULAWESI CONTINENT COLLISION
MnO
Na20
K20
P205
CO2
SO 4
S
F
C1
0.13 0.20 0.10 0.05 0.16 0.18 0.04 0.18 0.14
2.78 1.37 2.98 1.04 1.40 3.55 1.30 0.73 0.97
3.18 1.56 3.81 1.66 0.90 0.35 0.56 0.92 0.62
0.40 0.32 0.29 0.21 0.07 0.17 0.13 0.07 0.02
1.04 3.92 0.94 1.89 19.00 0.05 0.04 4.45 8.15
0.44 0.49 0.05 0.05 <0.1 0.16 0.10 0.20 0.10
0.07 0.13 0.02 0.02 0.02 0.07 0.05 0.21 0.36
0.08 0.04 0.07 0.03 0.02 0.02 0.02 0.03 0.02
Ce 108 68 79 31 22 14 9 27 9
Pr 12 7 8.4 3.1 3 2.3 1.3 3.3 1.1
Nd 44.9 25.9 32.1 10.9 12.1 11.6 6.8 13.8 3.9
Sm 8.7 4.7 6.0 1.8 2.8 3.9 2.2 3.6 0.9
Eu 1.8 0.9 1.3 0.4 0.8 1.3 0.6 1.0 0.3
Gd 7.3 3.8 5.1 1.7 2.7 4.5 2.5 3.3 1.3
~ 0.9 0.3 0.7 0.2 0.4 0.8 0.4 0.5 0.3
Dy 4.9 1.3 4.0 1.3 2.3 5.0 2.8 2.8 1.7
Ni 61 52 61 45 19 30 25 47 40
Cu 47 69 33 39 30 58 42 39 46
Zn 70 13 65 12 62 85 27 70 21
As 8.2 7.2 4.0 4.0 8.1 1.3 1.4 8.0 6.4
Pb 35.0 27.2 28.6 16.5 2.0 <2
Sn 14.0 8.5 <9
T1 1.4 0.8 1.3 0.8 0.1 < 0.3
Ag 0.6 0.3 0.5 0.1 0.4 0.3 0.2 0.4 0.1
5.5 3.0
9.0 10.8 5.0 12.0 5.9
0.3 0.1
ation b e c a u s e fresh a m p h i b o l e f r o m s a m p l e 93SUL43 yield total fusion and total gas ages o f 122-123 Ma, whereas plagioclase separates from the same rock possesses a m u c h older set o f ages (294-312 Ma). The high T (> 400°C) or partially reset cooling ages fall into two groups: Early Cretaceous (121-137 Ma) and Late Cretaceous to P a l a e o c e n e ( 5 7 - 9 0 Ma) and are interpreted to indicate the t i m i n g o f Cretaceous to Palaeogene o p h i o l i t e c r y s t a l l i z a t i o n , and in s o m e cases partially reset b y subsequent e x h u m a t i o n or thermal events. A third group is interpreted as low T (< 300°C) cooling ages ranging f r o m M i o c e n e to P l i o c e n e ( 5 - 2 0 M a ) and p r e s u m a b l y date the cooling f o l l o w i n g the obduction event. Therefore, although the samples are a m o n g the m o s t difficult for u n a m b i g u o u s 4°mr-39Ar age interpretation due to variable alteration/metamorphism and very low K-contents, yielding excess 4°Ar problems, we believe that some limited useful age information h a s b e e n p r o d u c e d , i n d i c a t i n g C r e t a c e o u s to Palaeogene ophiolite crystallization f o l l o w e d by E o c e n e to O l i g o c e n e o b d u c t i o n and M i o c e n e exhumation.
H 2 0 + H20-
LOI
Total
Mg#
0.02 0.01 0.03 0.02 < 0.01 0.02 0.01 < 0.01 < 0.01
2.78 1.90 1.67 1.15 1.90 2.30 0.83 3.32 1.85
1.84 2.14 0.33 0.29 0.70 0.35 0.27 1.16 0.94
5.32 4.91 2.52 2.62 20.60 2.21 1.08 8.83 9.00
99.79 0.54 99.55 0.47 101.37 99.80 0.27 99.94 0.88
54.20 12.39 56.91 7.61 36.22 48.43 9.08 34.45 13.92
Ho 0.9 0.2 0.7 0.3 0.4 1.0 0.6 0.5 0.3
Er 2.4 0.6 2.1 0.8 1.1 3.0 1.7 1.4 1.0
Tm 0.3 0.1 0.3 0.1 0.1 0.4 0.2 0.2 0.1
Yb 2.2 0.5 1.8 0.8 0.9 2.7 1.4 1.2 0.7
Sb 0.8 0.5 0.5 0.7 0.4 0.3 0.2 0.6 0.1
B 44 25 30 21 43 24 4 56 18
G 19 3 20 2 9 20 4 15 5
Ge 14 5 11 2 14 16 5 15 2
Lu 0.3 0.1 0.3 0.1 0.1 0.4 0.2 0.2 0.1 (ppb) Au 3.1 2.4 4.1 3.4 <2 <3 <2
~ 32 24 20 10 3 1 1 5 2
U 7.8 5.0 5.4 3.0 1.4 1.3 0.9 2.2 0.6
Zr 230 127 1~ 45 85 109 62 143 31
Pd 14.4 16.3 21.7 15.7 3.0 3.3 0.7 3.6 1.5
Pt 11.5 2.9 11.0 2.9 < 10 < 10 0.6 < 18
Hg 20 13 52 97 57 43 26 82 78
W h i t e m i c a f r o m a sandy m u d s t o n e o f Oligocene d e p o s i t i o n a l age ( s a m p l e 9 3 S U L 5 3 ) y i e l d s a 114 +_ 2 M a K - A r age, indicating a M i d to Late Cretaceous p r o v e n a n c e o f detrital white mica. This 114 M a age is virtually identical to those for the P o m p a n g e o Schist C o m p l e x o f central Sulawesi and the B a n t i m a l a C o m p l e x o f S W S u l a w e s i (Parkinson 1991), which are likely candidates for the p r o v e n a n c e c o m p o n e n t s o f the E o c e n e Oligocene clastic deposits o f Sulawesi. These and other red marine or fluvial sandstones and mudstones o f O l i g o c e n e - E o c e n e age f r o m the Kalosi area contain a m i x e d h e a v y mineral assemblage including both granitic (rutile, ilmenite, zircon) and ophiolitic-accretionary provenances (chromite) ( B e r g m a n u n p u b l i s h e d data), i n d i c a t i n g the presence o f both continental granitic and oceanic ophiolitic sources in the uplands o f the O l i g o c e n e Eocene depocentres. T h e distinctive red colour is interpreted to result f r o m oxidation o f r e d u c e d iron (probably derived f r o m the ophiolitic provenance) which has been p r o m o t e d by weathering in a tropical climate.
404
S.C. BERGMAN ET AL.
~
14-
10
.
.
.
.
.
.
.
olite
.
n
K,O
/ • ......
/~kJTr~ite
6-
Trachybasalt• "
(wt%)
x~.._O~Z
Tephrite/
4-
Basanite
F•~.~ "~,,-~ '~'0
I~1[
In
•,,~
[
& ,I
~•
k
."
\ // Rhyolite ,~
I ~1~ I~
~
Dacite
•
•
\ 2-
• Basalt
'icrobasalt
0
40
45
[
Basaltic Andesite t
•
Andesite
\
~.
I
I
I
I
I
I
I
50
55
60
65
70
75
80
O Miocene Intrusive Igneous Rocks SiC 2 (wt%) • Miocene Extrusive Igneous Rocks • Paleogene? Lamas• Complex Rocks
Pumices and Shards, Electron Microprobe data Wt% data on a volatile free basis.
Fig. 8. Western Sulawesi igneous rock total alkalis v. silica (TAS) plot showing the nomenclature of common volcanic lithologies (after Cox et al. 1979).
Fission track analysis
Apatite (n = 28 samples), sphene (n = 2) and zircon (n = 23) fission track (FT) analyses of representative Mesozoic and Tertiary igneous, metamorphic and sedimentary rocks define their provenance and thermal history (summarized in Table 4, and detailed in table 4 of Supplementary Publication SUP 18099 (see above)). Zircon grains in the igneous and sedimentary rocks are euhedral to subhedral and primarily yellow-brown, but range from yellow to brown and red-brown. Sphene grains are yellow-brown and subhedral. Apatite grains from the igneous rocks are mainly euhedral, whereas those from the sedimentary rocks are euhedral and anhedral with rare rounded grains. The FT ages reflect the time elapsed since cooling through the respective closure temperatures for sphene (250 _+50°C), zircon (200_ 25°C) and apatite (120 _+25°C) (Naeser & Faul 1969; Dodson 1973; Dodson & McClelland-Brown 1985; Green et al. 1989). Some samples for which FT ages were determined on several minerals show concordant FT ages, whereas others exhibit discordance, interpreted as rapid and protracted cooling, respectively. For intrusive rocks, these FT ages represent the approximate time of cooling through 225°C (zircon) or 60-130°C (apatite) and may represent
either the time of exhumation if the intrusives were emplaced at depths > 3-4 km, or the late stage crystallization history if they were emplaced near the surface (< 1-2 km depth). For the volcanic rocks, these ages reflect the late stage crystallization history or subsequent exhumation-related thermal history if they were buried to depths > 3-4 km following deposition. Nearly concordant apatite, zircon and sphene FT age histograms and age spectra are illustrated for the Polewali pluton in Fig. 15. Extremely rapid cooling ( > 5 0 100°C Ma -1) is indicated for the interval 60-250°C for this and other samples (Figs 15-16). Apatite FT ages of Miocene intrusive and extrusive units range from 1.9-11 Ma and average 6 -+ 2 Ma. Some vitric-crystal tuffs (e.g. samples 93RS201 and 93RS279) contain identical apatite and zircon FT ages, defining the age of deposition of the host tufts and therefore the age of the host sedimentary sections. Mean track lengths of samples with statistically meaningful populations (n > 30) range from 12.3-15.5 tam with widely varying standard deviations of 1.1-3.8 pm, mainly indicating rapid cooling (> 10°C Ma -t) through the temperature interval 60-130°C. Apatite uranium contents also vary widely from 2-71 ppm U. Sphene FT ages range from 4.4_+0.5 to 4.5 _+0.9 Ma with U - 50-69 ppm. Zircon fission
MIOCENE W SULAWESICONTINENT COLLISION
i
,
405
10
,
o • •
8
ShoshoniteSeries
O M i o c e n e Intrusive Igneous R o c k s
o
M i o c e n e Extrusive Igneous R o c k s P a l a e o g e n e ? Lamasi I g n e o u s R o c k s
,
,
,
,
,
o
•"
t.=A,•.%~-"
---
•
8
t.. (')
4
~A,mi i~l.--12.~ •
Subl.alkaline or Tholeiitic Series
LI•
•
•
u l.m
m~ i 0
• l__l_
•
,o
"o 2
•
A
•
A0
m~''° mlm
0.4 2:
• _.mlI. L~,rr~ "
•
1AAA
~ • llli-,l_'U
v
o .
"lm
.
.
.
o
.
0.0
•
A
A=mo •
= ._c•
• •
0.8
A~ '
~o
A
IA
A~AI~., • -- 0
g
0
O~oO,
•
•~
I~"
•
•
~ 0_,-~ 0
•
8,~
•
l'
,~,12
g
0
4 +
•
•
mmm= No
(3
•
•
0
•
""
s-~" UA lhA ,, I ~ A •
o
o
_ ~ - ll . ....-
•
•
o
4
Oa,c~,~a~,oeSeres 0
'
iA
9
~-o
o
,
Alkaline Series O • O~o
g
1
~.-<: 16
A
0.2
All _ml l = m~ ll~ •
0
0.1 ~
"lmlO r-- -i
i)miil~ i
•
i
•
~
5
~
1.5~
0
~
•
I
--
m
•
m0
0.0 ~ 12
I
,ln'-ii_
•
~-~l~
l~ •
,
,
,
,
,
50
55
60
65
70
SiO 2 ( w t % )
r~imw--
•
ol
45
l0
• i
&li_illq)
0.0
•
~,~m-° O~ :
"~1.0
0.5
•
75
•
llcm~91
m
80 45
,
,
i
50
55
60
4~
•
0 65
70
75
80
Si0 2(wt%)
Fig. 9. Western Sulawesi igneous rock Harker variation diagrams showing various major element oxide compositions plotted as a function of silica content.
406
S . C . BERGMAN ET AL.
FeO* (FeO + 0.9 Fe203 )
~ TholeSeriesl~ ite •
y" •
....-
o
..,
,,
\
..
Calc-AlkSeri alinees
• 1~
kVI ~/
-
\
0 MioceneIntrusiveIgneousRocks • MioceneExtrusiveIgneousRocks A, Pal:eogene?L:masi ComplexRocks kv/
__~ \
~/
~./
\ _.~
k/
Na20 + K20
MgO
Fig. 10. Western Sulawesi igneous rock AFM ternary diagram (wt%) showing the trends defined by a tholeiitic series, represented by the Tonga-Marianas-S Sandwich Island suites and a calc-alkaline series, represented by rocks from Cascades, North Chile, and New Guinea (after Brown, 1982). Dashed line shows boundary between tholeiitic and calc-alkaline (and alkaline) series (after Kuno, 1968).
track ages for Miocene igneous rocks range from 4.3-10.8 Ma; U = 171-582 ppm. Apatite electron microprobe compositions vary widely in anion abundance; C1 contents in several samples approach 2-3 wt% (table 4 of Supplementary Publication SUP 18099 (see above)), indicating higher than normal apatite blocking temperatures (140 ° v. 120°C) are probably applicable (Green et al. 1989). Apatite FT track ages of five sandy non-marine or marine mudstones with Oligocene-Eocene depositional ages from the Kalosi area are in the range 6-13 Ma, indicating the Palaeogene section now exposed at the surface in the Kalosi area was buried to temperatures > 120°C and cooled through 60-120°C during the Late Miocene. Zircon FT ages of these five Oligocene-Eocene mudstones are in the range 67-93 Ma, suggesting a Late Cretaceous
provenance and that these samples had not been heated to > 225°C since deposition. Zircon FT ages of two Latimojong Complex metasedimentary rocks are 123-128 Ma, with a 6.8 Ma apatite FT age, indicating an early Cretaceous provenance and Miocene cooling event due to thrust-related exhumation. The mostly Late Miocene apatite fission track ages of igneous, metamorphic, and sedimentary rocks from western Sulawesi are interpreted to suggest the time of rapid (> 10°CMa -]) cooling through the temperature interval 130-60°C, probably due to thrust-related exhumation associated with compressional tectonism.
Rb-Sr and Nd-Sm isotope data 87Sr/86Sr and 143Nd/144Nd ratios and Rb, Sr, Nd
MIOCENE W SULAWESI CONTINENT COLLISION
and Sm concentrations were determined on 32 representative whole rock samples with the purpose of constraining the nature of crustal or mantle magma provenance and the possible presence of earlier melting events than the most recent Miocene and older Lamasi magmatism (summarized in Table 5, and detailed in table 5 of Supplementary Publication SUP 18099 (see above)). In addition, three Palaeogene and Cretaceous sedimentary samples were analyzed to determine provenance and test the possibility that near surface sedimentary units represent a possible contaminant to the igneous rocks. 87Sr/86Srttnltlal) ....... values of the Miocene intrusive and extrusive rocks overlap, both groups ranging from 0.703-0.74, and are in general positively correlated with Rb/Sr ratios (Fig. 17). The lack of strong positive correlation between 87Sr/86Sr and 1/86Sr values indicates that the correlated 87Sr/86Sr and Rb/Sr values are not the result of mixing of two unique end members (Fig. 18). Palaeogene to Cretaceous sedimentary rocks can also be eliminated as potential contaminants of the igneous rocks because the sedimentary and metamorphic samples do not form isotopic end-members of the trends defined by the Miocene igneous rocks. The similarity in Miocene intrusive and extrusive isotopic ratios and ages indicates that the two groups are members of a cogenetic suite of lithospheric melts. 87Sr/86Sr values are not correlated with major and trace element composition, such as silica and alkalis, further suggesting that mixing of two compositional end members can be eliminated in explaining the major element and isotopic compositional variation. The entire Miocene suite representing samples from a 500 × 300 km area defines a Rb-Sr pseudo-isochron (Fig. 17) with an age of 305 Ma (initial 87Sr/86Sr = 0.7094, excepting the anomalous sample 90SUL28) whereas four separate samples from the Palopo pluton define a whole-rock isochron of 506 _+21 Ma (initial 87Sr/86Sr = 0.7073). The relatively high initial 87Sr/86Sr values suggest melting of relatively old, enriched mantle and crustal lithospheric material. Because the Palopo pluton possesses Late Miocene biotite and hornblende K-Ar ages (300-500°C cooling ages = 6-11 Ma), the 506 _+21 Ma isochron indicates early Palaeozoic crustal rocks were partially melted, yet not totally homogenized, by a Miocene melting event. U-Pb zircon data (see below) substantiate this conclusion of Miocene recycling of Palaeozoic or older crustal materials. The material with the most depleted 87Sr/86Sr values and enriched 143Nd/144Nd values are from the Lamasi Complex samples and are distinct from the Miocene igneous rocks. On a Rb-Sr isochron plot, a slight positive correlation is observed and a maximum homogenization age of 50 Ma can be
407
inferred from the limited range in 87Sr/86Sr with a large range in Rb/Sr. 143Nd/144Nd(initial) ratios of Miocene igneous rocks range from 0.5119-0.5131 and Sm/Nd values are in the range 0.094-0.25, 143Nd/144Nd values exhibit a strong negative correlation with 87Sr/86Sr values and fall mainly in the field of continental lithospheric melts (Fig. 19). The western Sulawesi data form a trend ranging from the most depleted ratios of the Palaeogene or older Lamasi Complex ophiolites, which are similar to MORB and nearby Celebes Sea and Sulu Sea Eocene ocean floor basalts, to the more enriched ratios of the Mamasa granite and related rocks. The Late Miocene Buakayu alkali gabbro sill (sample 90SUL3) possesses a slightly depleted set of Sr & Nd isotopic ratios which indicate a contribution from asthenospheric mantle. Besides this gabbro and the samples of Lamasi Complex, all other samples indicate derivation from relatively enriched lithospheric parental materials with Nd-model ages of 1-2 Ga. In contrast, Lamasi Complex rocks were derived from melting of an extremely LIL-element depleted, time integrated low Rb/Sr and high Sm/Nd asthenospheric source mantle, typical of mid-ocean ridge basalts or back-arc basin basalts. 143Nd/144Nd values of Miocene igneous rocks exhibit a slight positive correlation with 147Sm/144Nd values (Fig. 20), defining a Nd pseudo-isochron age of c. 0.6 Ga. 143Nd/144Nd values are not correlated with 1/Nd, other trace element, silica or alkali contents, suggesting that mixing of two compositional end members can be eliminated in explaining the major element and Nd isotopic compositional variation. Depleted-mantle Nd model ages vary from 0.8-2.3 Ga and are not correlated with silica content or other major element composition. The 143Nd/144Nd and 147Sm/144Nd values of Lamasi Complex rocks are positively correlated and define a 178 Ma Nd pseudo-isochron (initial 143Nd/144Nd ratio = 0.51278). On a Rb-Sr isochron plot, a slight positive correlation is observed and a maximum homogenization age of 50 Ma can be inferred from the limited range in 87Sr/86Sr ratios despite a large Rb/Sr variation, possibly indicating Paleogene pertubation of the Rb-Sr system. The Sr and Nd isotopic ratios of the three Eocene to Cretaceous sedimentary and metamorphic samples are in the ranges 87Sr/86Sr = 0.705-0.707 and 143Nd/144Nd = 0.5123-0.5126, overlapping the western Sulawesi Miocene igneous rock trend. These compositions are consistent with the view that the Palaeogene to Cretaceous sedimentary and metamorphic rocks possess a mixed provenance which includes depleted mantle array-like oceanic or calc-alkaline volcanic rocks and more enriched crustal components.
location/area
Bua Kayu Bua Kayu Bua Kayu Bua Kayu Bua Kayu Palopo Palopo Palopo Palopo Palopo Palopo Enrekang Enrekang Enrekang Enrekang Enrekang Mamasa Mamasa Mamasa Mamasa Mamasa Polewali Pare Pare Pare Pare Pare Pare Pare Pare Pare Pare Enrekang Bua Kayu Bua Kayu
sample
90SUL-2 90SUL-3 90SUL-4a 90SUL-6a 90SUL-7 90SUL-8 90SUL-9 90SUL-10a 90SUL-11 90SUL-12 90SUL-13 90SUL-14 90SUL- 15 90SUL-16a 90SUL-18b 90SUL-20 90SUL-21a 90SUL-21b 90SUL-22a 90SUL-22b 90SUL-23 90SUL-25 90SUL-26a 90SUL-26b 90SUL-26c 90SUL-26d 90SUL-27 90SUL-28 RAG90-73 RAG90-74
arkosic sandstone alkali gabbro sill syenogabbro sill basalt crystal-lithic lapilli tuff trachyandesite tuff monzodiorite pluton qz monzodiorite pluton hornfelsed andalus siltstone basalt block in tuff breccia qz monzodiorite pluton monzodiorite pluton trachyandesite lava flow trachyandesite lava flow dacite tuff dacite tuff trachyandesite lava flow qz monzodiorite pluton Quat alluv/Mamasa pluton qz monzodiorite pluton diorite in pluton dacite porphyry qz monzonite pluton andesite tuff breccia crystal lithic andesite tuff andesite clast in tuff breccia andesite clast in tuff breccia trachyte flow/dome gran. orthogneiss clast volcaniclastic sandstone siliciclastic siltstone
lithology
Table 4. K-Ar, 4°Ar-39Ar and fission track age data summary
5.4_+0.2, h 9.6_+0.9 8.7_+0.9 7.1-+0.3 7.2_+0.3, sn 6.7-+0.3 7.8_+0.8, b 7.3-+0.3 8.8_+0.3, p 7.7_+0.8 8.1_+0.3 7.7_+0.3 7.2_+0.3 7.0_+0.3 7.3_+0.3 6.1_+0.3, h 5.8_+0.6 8.7-+0.9 6.8_+0.3 7.8-.+0.3 4.8___0.2 8.1-+0.3
b h b b p b b b b b b b b b b b b
b 10.0+_0.4 b 6.8_+0.3, h 6.0-+0.3 b 6.2-+0.2, h 10.7_+1.1
b 12.5___0.5 b 11__.0.4
K-Ar phase & age (Ma)
z a a z a z a z z a z a a a a
6.6_+0.5 7.3+1.2, z 13.0-+0.6 7.2_+1.0 6.6_+0.5 4.6__.0.6 6.8_+0.4 4.2__.0.6, z 5.0-+0.3, s 4.5-+0.9 6.1_+0.4 6.2_+0.4 5.9_+0.8 5.8_+0.4 6.9___1.5 1.9-+0.5 24.6_+6.5 4.0_+2.9
a 20.4_+5.2 z 4.3_+0.6 z 8.6_+0.6 a 2.1-+0.5, z 6.4_+0.4, s 4.4_+0.5 z 5.7_+0.4 a 8.4_+4.3 a 7.7_+1.2 a 6.2_+1.1, z 6.3_+0.4 z 10.8_+0.6 z 6.7-+0.4 a 4.5_+0.9
a 3.8__.1.6
FT phase & age (Ma)
Ar-Ar phase & age (Ma)
"~
.~
trachyandesite lava trachyandesite lava trachyandesite tuff trachyandesite lapilli tuff sandstone trachyte tuff leucite shoshonite lava leucite shosh, vitric crystal tuff trachyandesite lava basaltic lapilli tuff breccia equigranular granodiorite silicified mudstone lamprophyre basaltic tuff block breccia equigranular granite andesite lava andesite blocks/tuff breccia basaltic sill amphibolite/gabbro
basaltic andesite lava basaltic andesite lava massive pillow basalt lavas sheared basaltic greenstone rhyolite block in breccia sheeted gabbro dikes sheared equigranular diorite rhyolitic clasts in breccia chloritic phyllite calcareous augen/greenschist light green sandstone coarse grained sandstone sandy mudstone trachyte tuff trachyte tuff
Biloka River Biloka River Waru Balu Lambou/Kalosi Mamuju Mamuju Mamuju Tirasa/Benteng Loca divide Kanang/Polewali Kanang/Polewali Polewali Polewali Minanga Busu Pappang Lemo Lembang
Buntu Saragi Bua Bakabaka Batupapan Cilallang Bonelemo Salu Ranteballa Barulatong Langae Dea Raada Panjura Saruran Matama Limboro
7.6_+0.2 7.4_+0.2 8.0_+0.3 8.5_+0.3
wm 114+_2 b 7.7+_0.2 b 9.6_+0.3
am 203tf p 20.0-+1.0 p 23.8-+1.6 p 120-+5 wr 23.8-+1.2 p 126.4tf; am 371tf p 201_+9 (0.10%K) p 162_+13 (0.02%K)
phi 17.5_+0.4 b 12.8_+0.3 b 7.6_+0.2 b 9.8_+0.3 wr 45.7+_3.1 wr 9.7-+0.4
b 9.8-+0.3
sn 5.4-+0.2 sn 5.3-+0.2 b 2.4-+0.1 b 7.3_+0.2
b b b b
z a a a a a a
123_+9 6.3_+2.3, z 9.9__.1.6, z 5.5_+1.6, z 8.2-+2.5, z 7.0_+1.2, z 9.2-+1.4, z
a 8.7_+8.8
128_+12 69_+5 67-+8 93-+9 7.0-+0.4 9.8_+0.5
all_+l.2 a 4.1-+1.0 a 2.4-+1.8, z 6.8_+0.5 a 5.8_+0.9
a 12.9_+3.8, z 92_+10
b, biot; p, plag; h, hbl; a, apatite; z, zircon; am, amphibole; sn, sanidine; wr, whole rock; s, sphene; tf, total fusion; plat, plateau; phi, phlog.
92SUL-1 92SUL-3 92SUL-6 92SUL-7 92SUL-8 RAG92-4 RAG92-6 RAG92-9 ND92-29 93SUL3 93SUL5 93SUL6a 93SUL9 93SUL10 93SUL13 93SUL15 93SUL17 93SUL20 93SUL27 56.9tf 93SUL28 93SUL30 93SUL35 93SUL38 93SUL42 93SUL43 93SUL45 93SUL46 93SUL47 93SUL48a 93SUL49 93SUL52a 93SUL53 93RS201A 93RS279A 72.4tf,18_+2.1plat;
p 312tf; am 123.2tf,
p 15.6tf,12.2_+0.2plat p 20.4tf, 5.2_+0.2plat p 137.3tf
p
am
410
s . c . BERGMAN ET AL. I
Table 5. Summary of Rb-Sr and Nd-Sm isotope data for igneous, metamorphic and sedimentary rocks sample
90SUL3 90SUL4A 90SUL5 90SUL6A 90SUL7 90SUL8 90SUL9 90SUL10A 90SULll 90SUL 12 90SUL13 90SUL14 90SUL18B 90SUL20 90SUL21a 90SUL22a 90SUL23 90SUL24 90SUL25 90SUL26C 90SUL26D 90SUL27 90SUL28 93SUL1 93SUL9 93SUL27 93SUL28 93SUL30 93SUL43 93SUL45 93SUL49 93SUL52A
lithology code
87Sr 86Sr*i
143Nd 144Nd*i
Mi(m) Mi(m) Mv(m) Mv(m) Mv(i) Mi(i) Mi(f) Km(p) Pv(m) Mi(f) Mi(f) My(f) Mv(f) Mv(f) Mi(f) Mi(f) Mv(f) Mv(f) Mi(f) My(i) Mv(i) Mv(i) Mv(i) Mv(m) Mi(m) Pi(m) Pv(m) Pv(i) Pi(m) Pi(i) Ps(p) Ps(p)
0.70643 0.71234 0.71391 0.71235 0.7 1551 0.71449 0.71610 0.70787 0.70335 0.71094 0.71200 0.72126 0.71679 0.71217 0.71973 0.71921 0.72213 0.72210 0.71324 0.71066 0.70996 0.70717 0.73909 0.70852 0.71645 0.70300 0.70317 0.70342 0.70360 0.70316 0.70662 0.70559
0.51274 0.51212 0.51192 0.51207 0.51206 0.51205 0.51202 0.51253 0.51301 0.51220 0.51227 0.51214 0.51218 0.51229 0.51215 0.51194 0.51215 0.51216 0.51215 0.51225 0.51226 0.51251 0.51221 0.51246 0.51217 0.51304 0.51305 0.51294 0.51310 0.51262 0.51230
Age: M, Miocene; P, Palaeogene/Cretaceous; K, Cretaceous. Rock type: v, volcanic; s, sediment; i, intrusive; m, metamorphic. Composition: (i) intermediate; (m) mafic; (f) felsic; (p) pelitic. *i, initial value; whole rock samples in all cases.
I
I
I
I
I
I
I
I
I
I
I
OIB N-Type MORB E-Type MORB average REE abundances 100
~
normalized
10
Miocene
1
La
0.315
Ce 0.813
0,1
Intrusive
Is"
Igneous
Ig
l'y
Rocks
I
0.192 0.259 0.325 0.213 0.209 Nd Eu Tb He Tm Lu 0.597 0.0722 0.049 0.072 0.032 0.0323
100
chondrite normalized
10
1 La I Pr I Sm I Od I Dy I Er I Yb 0.315 0.1 0.192 0.259 0.325 0.213 0.209 Ce Nd Eu Tb He Tm Lu 0.813 0.597 0.0722 0.049 0.072 0.032 0 . 0 3 2 3
t
I
I
I
I
I
I
I
I
.......
~
100
"-,,
I
I
OIB N-Type E-Type
I
I
MORB MORB
average REE abundances chondrite normalized
Zircon U/Pb data U/Pb dating was attempted on zircon and sphene separates from six representative rocks with the purpose of testing the hypothesis that an inherited crustal component is present in the Miocene magmatic rocks (Table 6). All samples exhibit varying degrees of discordance, several (samples 90SUL12, 15, 21, 25 and 26c) less so than others (sample 90SUL28). Zircons from samples 90SUL12, 15, 21, 25 and 26C possess similar 2°6pb/238U and 2°Tpb/235U ages of 5-13 Ma, broadly similar to the K-Ar and fission track ages on the same samples. In contrast, 2°Tpb/2°rpb ages range from 88-1073 Ma for this material. Sample 90SUL28 displays Palaeozoic to Proterozoic U/Pb and Pb/Pb ages and defines a
10
Palaeogene?
1
•a I
0.315
Co 0.813
O, 1
I
Lamasi
Complex
I o' I'y
Rocks
I
I'b
0.192 0.259 0.325 0.213 0.209 Nd Eu Tb He Tm Lu 0.597 0.0722 0.049 0.072 0.032 0.0323
Fig. 11. Western Sulawesi igneous rock rare earth element (REE) profiles, normalized to chondrite compositions (Hanson 1978) showing trends observed for ocean island alkali basalts (OIB), normal mid-ocean ridge tholeiitic basalts (N-MORB) and enriched midocean ridge tholeiitic basalts (E-type MORB) (after Sun & McDonough 1989).
MIOCENE W SULAWESI CONTINENT COLLISION chord on the concordia diagram that possesses an upper intercept of 391 Ma, similar to the Rb-Sr whole rock model ages for the same sample and the entire suite of Sulawesi Miocene granitoids and andesites. Sphene U/Pb data indicate concordant 2°6pb/238U and 2°Tpb/235U ages for sample 90SUL25 (6 Ma), yet much younger U/Pb ages of 1.0-1.3 Ma compared to its biotite K-Ar age (8.1_+0.3 Ma) and similar to its apatite fission track age (1.9 _+0.5 Ma).
Discussion Petrogenesis
Mid to Late Miocene mafic, intermediate and felsic volcanic and intrusive rocks are abundant and widespread in western Sulawesi, indicating extensive lower crust and upper mantle melting during the interval 2-18 Ma (mean and mode 8 _+4 Ma). The Pliocene to Miocene intrusives and extrusives possess overlapping age and major, trace element and Sr-Nd isotopic compositions, indicating a cogenetic volcanoplutonic belt or province. These Miocene igneous rocks form a bimodal compositional suite, with basaltic and dacitic compositional subsets predominating, and generally lacking intermediate andesites. This suggests two distinct sources of magmas: a more mafic, probably peridotitic parent for the basaltic rocks and less mafic, probably basaltic parental material for the dacites and compositionally similar intrusives. The sources of these Miocene magmatic rocks, therefore, range from enriched (in LIL and other incompatible elements) ancient upper mantle peridotite lithosphere, which melted to produce the alkaline basaltic sills, dykes, and pyroclastic sequences in the region, to Palaeozoic or older lower crustal rocks of basaltic to intermediate composition, which melted to produce the rhyolite-dacite tufts and quartz monzonite to monzodiorite stocks, plugs and plutons. Because an asthenospheric mantle source does not appear to be important in the petrogenesis of the Miocene basaltic melts (except sample 90SUL3), based on the radiogenic isotopic and enriched trace element compositions, classic subduction-related melting processes were probably not involved in the formation of the Miocene basalts. The Miocene granitic rocks (quartz monzonite to monzodiorite) are relatively alkalic and silica-poor compared with granodiorites of subduction-related arcs, and the most likely mechanism f o r their formation involves lithospheric melting instead of subduction-related melting of a mantle wedge overlying a Benioff zone. The western Sulawesi igneous rocks are notably distinct in major element and isotopic composition from arc-continent collision-related
411
magmatism in the East Sunda-Banda Arc (van Bergen et al. 1993). Furthermore, andesitic compositions form a minor component of the Miocene suite of western Sulawesi. Instead, a bimodal assemblage of alkaline basaltic and dacitic compositions predominate. The few andesites (e.g. Parepare) are relatively enriched in 87Sr (87Sr/86Sri = 0.710) compared with typical continental subduction-related andesites (87Sr/86Sr = 0.705-0.707; Gill 1981; Ewart 1982) and are more compositionally allied to the group III late collisional melts of Harris et al. (1986) in terms of radiogenic 87Sr/86Sr, low Rb/Zr, and presence of mafic and felsic melts. The most likely mechanism for generation of both basaltic and granitic material in western Sulawesi involves mantle and crustal lithospheric thickening resulting from accretionary processes, which produced progressive heating of the lithosphere, and finally melting during the Miocene (see Jaupart & Provost (1985), Delisle (1986), England & Thompson (1986) and Karabinos & Ketcham (1988) for relevant details of collisional magmatism thermal modelling). It is also possible that a lithospheric delamination process (Bird 1979; Kay & Kay 1991, 1993; Davies & van Blanckenburg 1995) was involved in the melting process. Trace element geochemistry of western Sulawesi Miocene igneous rocks indicates that they are not allied directly with normal subduction-related processes. Spider diagrams suggest similarity to subduction-related rocks from volcanic arcs associated with compressional boundaries in which oceanic crust is subducted beneath continental crust (Fig. 12). Trace element discrimination diagrams proposed for granitic rocks (Fig. 13) yield ambiguous results: intermediate between syncollisional granites and volcanic arc granites. The compositions of western Sulawesi intrusives largely cluster near the granite minimum melt eutectic at 650-670°C, in equilibrium with relatively Ca-poor plagioclase (An4_10). This lends further support to the view that the granitic melts represent partial melts of continental crust. Using discrimination criteria based on mineral abundances (classification of Lameyre & Bowden 1982), the western Sulawesi granitic rocks are intermediate between normal alkaline and calcalkaline series granites (calc-alkaline granodioritic and monzonitic series). Harris et al. (1986) divided collision zone magmatism into four types based on the relative timing of magmatism and tectonism: groups I to IV for pre-, syn-, late- and post-collision intrusions. In their classification, the western Sulawesi Miocene magmatic rocks are group III late-collision calc-alkaline to alkaline intrusions, based on structural relations and compositional traits. The granitic melts are distinct from typical
412
S. C. BERGMAN ET AL.
10000
Western Sulawesi Miocene Extrusive Rocks •~
1000
.~
100
E o Z
10 I
c 1
I .
I
m
E
•=
0.1
t I
OIB E-Type MORB N-Type MORB average abundances primitive mantle normalized
0.01
1oo0oj
• High K Andesite Calc-Alkaline Andesite • Low K Andesite [] Collision Granite © Volcanic Arc Granite
I
Cs Rb Ba Th U K Ta Nb La Ce Sr Nd P Hf Zr Sm Ti Tb Y o.188o.6356.989o.085 0.021 250 o.014 0.7130.6871.775 21.1 1.354 95 0.1066 11.2 0.4441300 0.108 4.55
Miocene Intrusive Rocks
._o 1000
gc N
100
t~
, •
Z
10
.t-
O. 1
n
t
,,\
• Low K Andesite
0.01 Cs Rb Ba Th U K Ta Nb La Ce Sr Nd P Hf Zr Sm Ti Tb Y 0.1880.6356.9890.085 o.o21 250 0.o14 0.7130.6871.775 21.1 1.354 95 o.1066 11.2 0.4441300 0.108 4.55
continental crust melts associated with continental collision, such as the Himalayan belt (Dietrich & Gansser 1981), which are formed by leucogranites significantly more enriched in SiO 2 and A1203, although of similar 87Sr/86Sr, relative to the western
Sulawesi granitoids. This contrast may be a function of contrasting lower crust lithologies in the two collision zones: the Himalayan source is pelitic or granitic, whereas the western Sulawesi Miocene igneous parental rocks are intermediate to mafic
MIOCENE W SULAWESI CONTINENT COLLISION
10000 l
I
I
I
I
I
I
I
I
I
I
I
I
I
413
I
I
I
I
Palaeogene? Lamasi Complex Rocks
L F
I
1000 ",
100
QcE.~'"
,~
a'-
~',,~ t'-,
I ~-
/LV',,~., q -,
10
1
I i 0.1
I/
• ~ • [] o
'I~ORB average abundances
0.01
primitive mantle normalized I
I
I
I
I
I
I
I
I
I
High K Andesite Calc-Alkaline Andesite Low K Andesite Collision Granite Volcanic Arc Granite !
I
I
I
!
I
I:
I
F
Zr Sm Ti Tb Y 0.1880.6356.9890.085 0.021 250 0.014 0.7130.6871.775 21.1 1.354 95 0.1066 11.2 0.4441300 0.108 4.55 Cs Rb Ba Th
U
K
Ta
Nb La
Ce Sr
Nd
P
Hf
Fig. 12. Western Sulawesi igneous rock 'spider' diagrams, normalized to primitive mantle trace element abundances, showing trends observed for ocean island alkali basalts (OIB), normal mid-ocean ridge tholeiitic basalts (N-MORB), enriched mid-ocean ridge tholeiitic basalts (E-type MORB), and various representative andesite and granite compositions (after Sun & McDonough 1989; Pearce et al. 1984; Ewart 1982).
meta-igneous assemblages. Priadi et al. (1993) presented trace element and St, Pb, and Nd isotopic data on basement peridotites, granulites and amphibolites from Palu (Central Sulawesi, north of the present study area) which demonstrated an Australian lithospheric character. However, the Palu Sr and Nd isotope data fall in distinct fields and define contrasting trends compared with the Miocene magmatic rocks of the present study and the Palu basement rocks lack the isotopic and trace element compositions required to form a significant contaminant or parental contributor to the Miocene igneous rocks of western Sulawesi. The Lamasi Complex ophiolites form a distinct compositional group and should be considered separately from the Miocene igneous rocks. The Lamasi rocks are tholeiitic in composition with relatively flat REE to LREE-depleted patterns, in contrast to Miocene calc-alkaline and alkaline assemblages which possess strongly LREEenriched signatures. The geochemistry of the Lamasi ophiolites are similar to ocean floor basalts and tholeiitic subduction-related (primitive arc) and back-arc basalts in major, trace element, 87Sr/86Sr and 143Nd/144Nd composition. These Lamasi
Complex ophiolites were obducted during the Late Oligocene and Early Miocene based on new K-Ar and 40Ar-39Ar data and probably indicate that the eastern Sulawesi ophiolite (Silver et al. 1983a, b) extends into western Sulawesi. Other members of the Lamasi Complex, otherwise known as the Palaeocene-Eocene Langi Volcanic Formation or the 'Older Andesites' of van Bemmelen (1949) and the Papayato Volcanic Formation of Ratman (1976) may represent part of these CretaceousPalaeogene(?) ophiolitic assemblages. Mubroto et al. (1994) presented K-Ar data on the Balantak ophiolite of East Sulawesi which indicate Cretaceous to Eocene crystallization ages, similar to the Lamasi Complex volcanics in the present study. Nishimura et al. (1980) presented gravity data for South Sulawesi which show the existence of a gravity high east of the Bone Mountains, consistent with the area underlain by ophiolite. The Lamasi volcanic complex could represent a fragment of Indian Ocean crust, similar to that thought to be trapped in the Banda Sea (Lapouille et al. 1986), or, less likely, more proximal forearc or back-arc oceanic crust related to the Sunda subduction zone.
414
S.C.
i
1000
i
i
i lllll
i
i
i
i iiiii
[
Syn-Collision Granites • Extrusive O O Intrusive O ¢ ~1~{ 'I
-
i
i
ETAL.
BERGMAN
i l i i I i i i i I I I I I I I t I I I i i i i I I I I I I I I I I
i iiiii 1.0 --
•
0 •
WithinPlate/ Gra~
o.s-
~-,,
100
v
,,,~. ",4;.
~00_
Q.
n-
Arc
Volcanic
.$3
Granites
lO
Miocene Intrusive Igneous Rocks -Miocene Extrusive Igneous Rocks L a m a s i Complex Rocks
at P a l a e o g e n e ?
O
n Granites
i o
•
"4
",
",
,,
o •",,i
Ridge
-
"o
•
c "'.,,"',
o
"4 -05 '
'
''""1
10
'
'
'""1
1O0
'
'
-
V@~
,
'''"'1 1000
k
1000 -'J \
,
Q.
,,,,,,I
,
,,,,,,I
,
• Extrusive
,
-1,0
Granites
1
. . . . . . . .
I
~"
I
I0
OceanRidge Granites
. . . . . . . .
I
1 O0
Y (ppm)
. . . . . . . .
45
I I I I
&
I I I I I 1 I I
5/5
60
I I
IJ51
' '
70 ....
I I I I
715
80
SiO2(wt%)
IVo,cani rc . Granitesand I Syn-Collision I Granites I1~ I/
",,
,,,,,,I
Within Plate
ntrusive
o
10
,
I I I I
1 ~
~
,
--
Field of Normal
Y + Nb (ppm)
•
,
I
1000
Fig. 13. Western Sulawesi igneous rock trace element discrimination diagrams for granitic rocks, showing the compositional fields defined by granitic rocks of various tectonic associations (after Brown 1982).
Thermal history The K-Ar and fission track geochronology data place firm constraints on the thermal history of western Sulawesi. The age data are plotted on time-temperature diagrams in Fig. 16, in which the blocking temperature of each phase is used. The cooling histories of different rocks in each area are similar, and the five areas possess nearly identical cooling histories, with the exception of several sandstones from Buakayu, which possess much older fission track ages. The cooling trajectories form two distinct gradients: an earlier phase of rapid cooling (> 100°CMa -1) in the interval > 500-100°C, and a later phase of more protracted cooling (at 5-20°CMa -1) from 100-20°C. The
Fig. 14. Western Sulawesi igneous rock Cat/alkali v. S i t 2 plot showing the composition of SW Sulawesi rocks relative to typical calc-alkaline subduction-related melts (after Pearce 1983).
earlier cooling phase is probably associated with magmatic cooling, which typically occurs at > 100°C Ma -], and the latter phase with denudation, typically at rates of 5-20°C Ma -]. The thermal history of the area is one of rapid emplacement of magmatic rocks and eruption of volcanic rocks during a short term episode in the Mid to Late Miocene (3-18 Ma) followed by rapid cooling of these melts in the upper crust. Late Miocene to Pliocene cooling was more gradual (5-20°C Ma-1), although rapid enough to produce relatively long confined apatite fission track length distributions (> 14 ~tm mean track lengths), and was probably associated with thrust-induced uplift and erosion. Several sandstones from the Buakayu area possess some of the oldest apatite FT ages (11-25 Ma, relative to depositional ages) of all samples studied. The detrital apatites possess significantly older FT ages than their corresponding depositional ages, suggesting the rocks have not been heated to > 120-150°C since deposition. The best constrained crystallization, cooling and unroofing history of an individual pluton occurs at Palopo. Four samples from 475 m, 835 m, 1050 m, and 1070 m elevations (samples 90SUL12, 13, 9 and 8) exhibit a range in hornblende A1 contents (averaging 5.7, 12.9. 4.9 and 10.4 wt% A1203; Bergman unpublished data) and predicted crystallization pressures, averaging 1.7, 7.6, 1.0 and 5.7 kbar, respectively (calibration of Schmidt 1992; Hammarstrom & Zen 1986;
MIOCENE W SULAWESI CONTINENT COLLISION
ZIRCON 90SUL25 Polewali Pluton
>`6 t-
~4 O"
I-i-2.
0
>,6" c.)
i
I
5 10 15 20 Apparent Fission Track Age (Ma) SPHENE 90SUL25 Polewali Pluton
t-"
U- 2 .
0
>6
I
i
I
I
5 10 15 20 Apparent Fission Track Age (Ma) APATITE 90SUL25 m Polewali Pluton
~4 O"
0
5 10 15 20 Apparent Fission Track Age (Ma)
Fig. 15. Fission track age histograms for apatite, sphene and zircon from the Polewali pluton.
Hollister et al. 1987; Rutter et aL 1989). The presence of andalusite porphyroblasts within the Palopo pluton contact metamorphic aureole (sample 90SUL10, elevation 250m) limits the intrusion pressure to < 3.8 kbar, the aluminosilicate triple point. This indicates that the elevated pressures inferred for samples 90SUL13 and 8 probably reflect earlier phases of crystallization in the cooling history of the pluton, possibly during magma ascent through the lower and middle crust (15-23 km). The top of the pluton (sample 90SUL9 from 1050 m elevation) crystallized at pressures of 1.0 __.0.5 kbar (3-4 km subsurface depths), whereas the lower-most sample (90SUL12, 0.5 km
415
elevation) crystallized at pressures of 1.7 _ 0.5 kbar (5-6 km subsurface depths). Combining these inferences with the fission track and K-Ar geochronology data for the Palopo pluton and its contact aureole, its cooling and uplift history can be constrained. The Palopo parental granitic magma was formed in the lower crust during a Miocene melting event, ascended toward the surface, and intruded the Cretaceous deposits at depths of 3-5 km during 6-10 Ma and rapidly cooled through 500-300°C by 6-8 Ma. The pluton finally cooled through 60-120°C and ascended through 2-3 km depths by 2-3 Ma, presumably during denudationrelated uplift and erosion associated with thinskinned thrusting. Tectonic framework
The Miocene plate tectonic framework of Sulawesi has been the subject of ongoing vigorous debate (Katili 1978; Sukamto 1978; Hamilton 1979, 1989; Van Leeuwen 1981; Hutchison 1982, 1989a, b; Silver et al. 1983a, b; Wood 1985; Nishimura 1986; Leterrier et al. 1990; Letouzey et al. 1990; Rangin et al. 1990; Audley-Charles 1991; Audley-Charles & Harris 1991; Daly et al. 1991; Rangin & Silver 1991). Most workers envision a Miocene westward-dipping subduction zone to produce the widespread Miocene volcanic and plutonic rocks in western Sulawesi. The present study, however, indicates Palaeozoic to Proterozoic lithospheric parental rocks unlike any known in Sundaland (see Hutchison 1989a for details of Sundaland) were melted to produce the Miocene igneous rocks. Such Palaeozoic to Proterozoic lithospheric assemblages characterize the northern Australian Plate, such as in western Irian Jaya (Pieters et al. 1983; Pigram & Panggabean 1984) and derivative continental blocks including Sula, Buru, Seram and the Tukang Besi Platform. Silurian or Permian granitic rocks form part of an Early Palaeozoic continental assemblage and extensive Miocene collision of arc terranes and associated crustal melting are known along the northern margin of NE Australia-New Guinea (Audley-Charles 1991). This similarity to the inferred crustal parental material of the Miocene magmatic rocks of western Sulawesi suggests that a crustal fragment derived from the northern Australian plate was accreted onto SE Sundaland during the Miocene due to westward vergent plate motion of the Pacific plate or was transported along and juxtaposed against SE Sundaland by a transform margin. It is also possible that the accreted crustal block was one of the allochthonous Pacific microplates which characterize eastern Indonesia (Hutchison 1989a, b; Audley-Charles 1991; Audley-Charles & Harris 1991). Early Miocene emplacement of the East
416
S . C . BERGMAN E T A L
25 20
AGE (Ma) 15 10
20
5
E
°4e°e° ~
rein100
25, 20
...
AGE (MliI 15. 10.
20
5. ~
100
~',,.,0e <~
~
0
~
/~
,,.,,:.~400~
~" ~
0
J 25
o.Conll~ P_hy!Ute oo o,gn.~,.
+
~ooLT .."'"'~"" Ip
PALOPO PLUTON
(a
/
i
0 5
• "........... T T
i
200~ lind Turblditel
Ul
+
+
Rocks s00 o Igneous Smdztonn
M .AMAS.A BA'I:H O LI';'H PAREPARE IGNEOUS
'
a
BUA I(AYU ' ENREKANG
o
,oo
o"~" 0
," ". • I i. 50 100rni
.SO0 2s
AGE (Ms) 15 10
20
0 SarDdmon~ Igneous Rocks io
l's
1'o
AGE (Ma)
~
o
2s
~o
l's
1'o
AGE (Ma)
~
o
Fig. 16. Sulawesi igneous and sedimentary rock time-temperature plots summarizing new K-Ar and fission track age data and thermal history trajectories.
Sulawesi ophiolite marked the onset of the Sula Platform accretion event (Audley-Charles 1987; Silver et al. 1983a, b; Parkinson 1991). Lithospheric thickening following obduction of a crustal block produces melting at the base of the crust and formation of granitic magma at c. l0 Ma after initial crustal thickening, based on thermal modelling by Delisle (1986), which nicely matches the obduction and magmatism age relations observed in Sulawesi. Continued west-vergent compression led to crustal thickening beneath the south arm of Sulawesi, culminating in Late Miocene lithospheric melting, extensive volcanism, and Late Miocene to Pliocene thin-skinned thrust faulting (see Fig. 21) in a manner broadly similar
(albeit of a smaller scale) to the Eocene accretion of the Indian plate onto Eurasian plate and the subsequent development of the Himalayas. The Pliocene framework was characterized by regional deformation and uplift along a 10 000 km long belt on the northern margin of Australia and in much of Indonesia. Fold and thrust belts of AlpineHimalayan scale developed during the last 5 Ma in New Guinea, Sulawesi, the Moluccas and Banda Arc. Interestingly, the rapid uplift inferred for South Sulawesi since 5-8 Ma on the basis of apatite FF data is also suggested for two other major fold-thrust belts, the Papuan fold belt (Hill & Gleadow 1989) and the Mt Everest area of the Himalayan orogen (Bergman et al. 1993) using
MIOCENE W SULAWESICONTINENT COLLISION
0.725
0.720
I
A + O •
I
'~"~J
0.710 (0.70937 i) (0.7073i)
~
O
.... ;4,~,-
~~ 'e ' - m~;~''"
0
. . - " "" I-'"'i ..-"'"" ~(3~°~ ' c )
.-'" ~.~K.~./.ce~-" 0
0.715 -
87Sr/ 86Sr present
I
Palaeogene or Mesozoic? Lamasi Ophiolitic Rocks Miocene Palopo Intrusives Miocene Intrusive Rocks Miocene Extrusive Rocks Palaeogene and Cretaceous (~)
SedimentaryRocks
417
0.74 -
-
2
4
,
,
6
,~--',/ '~/
8
,
,
.,""
....-""'"
10 ,
~.~
oCxf77 f" f
0 HI
m
0.705 A
0.700
o7,-
I 1
2
i 3
87Rb/ 8eSr Fig. 17. Western Sulawesi igneous rock isotope plot: 87Sr/86Srv. Rb/Sr.
apatite and zircon b-rI' data. The Pliocene structural evolution of South Sulawesi is probably not composed of compressive deformation alone; Miocene collision-related nappe emplacement in Timor and Seram produced post-collisional uplift and nappe attenuation by low and high angle normal faulting (Harris 1989). The presence of the Lamasi ophiolite complex of western Sulawesi and analagous easterri arm ophiolite separated by the deep intervening Bone Bay suggests orogenic collapse may be occurring here as well. Seismic reflection data across the Bone Bay show the existence of an N-S trending, east-dipping normal fault system accommodating > 4 km of post-Miocene clastic sedimentary fill. The lithospheric melts formed in the lower crust were transported to the upper crust along well established zones of crustal weakness, probably reactivated fault zones (Fig. 21). Crystallization took place in the upper crust within 5-10 km of the surface as intrusive masses of various sizes and shapes, including narrow dykes and sills 2-50 m across, plugs and stocks 0.1-1 km in diameter, and large plutons, laccoliths, and batholiths 20-100 km in diameter. Some batholithic-scale laccoliths (e.g. Mamasa) were later detached from their site of crystallization and structurally transported westwards to the surface during the last 6-8 Ma along low angle thrust faults.
The thermal consequences of crustal accretion, magmatism and thrust faulting are variable and offsetting. These include a reduction in subsurface heat flow due to attempted subduction of relatively cool continental crust, increased heat flow due to the emplacement of large plutons and batholiths into the upper crust, and the development of inverted thermal gradients due to thrusting of hot sheets over cold sheets (see England & Thompson 1986 and references therein for more details of the latter process). Present-day heat flow values are largely lacking for Sulawesi (Thamrin 1985, 1986; Matsubayashi & Nagao 1991); the only available data are for the western coastal area of western Sulawesi and range .from 10-48 mW/m 2. The authors suggest an elevated regional thermal regime during Late Miocene and Early Pliocene magmatism and thrusting, with presumed regional heat flows in the range 50-90 mW/m 2. The large volume (> 100-1000km 3) of granitic melts emplaced in the upper crust would produce the largest thermal effects of regional importance due to conductive cooling of 650-700°C melts, accentuated by convective and advective heat transfer by fluids. The new geochronology and structural data place f'trm constraints on the Neogene tectonic evolution of the Makassar Strait. Instead of most previous models which define it as a rift-related tectonic
418
s.c. BERGMAN
ET AL.
0.725
m m 072°73 Jto'~o"l.... J' ~ .....
© 0.720 -
0
©
1•
0.715 -
•
~:
:~
0.70 1 , , , I . . . . . . . . 0.000 0.008 0.016
mm
87Sr / 86Sr initial
t,•~
0.71
00
m m
© <so>
Omm
0.710 -
O Miocene Intrusive Igneous Rocks • Miocene Extrusive Igneous Rocks • Palaeogene? Lamasi Complex Rock: Palaeogene and Cretaceous Sedimentary Rocks
m m o 0.705 -
A
A
0.700 0.000
A
t
I
I
0.002
0.004
0.006
A 0.( )08
1/Sr
Fig. 18. Sulawesi igneous rock isotope plot: 87Sr/86Srv. l/Sr.
MORBasalts I (asthenosphere) I ~ ~.2~-
I
I
I
I
I
o.o by crustal residenceage
0 •5130 -~ ,-d~ sulu- k ~ ' , , Celebes lx-I ', ^ M ',~,,.i Java _~ea ',, f-~ Mantle',, I",.l(-') Array " " ~ " t ~ Z rt~""_',,~l ',~a "~" 0.5126 -_ Bu'k_Eah_
I
O • • ~
Miocene Intrusive Igneous Rocks Miocene Extrusive Igneous Rocks Palaeogene? Lamasi Complex Rocks Palaeogene and Cretaceous Sedimentary Rocks
-lO
-5
0 -o z
nda Arc
-
"~ Z
C
-
O
-
Q. IJ.i
1.0 by crustal residenceage
C-
oj m 0.5118 0.700
t
t 0.710
-5
-10
(3 t
t 0.720 87Sr/86Sr initial
2.o by crustal residenceage I
0.730
-15
I
0.740
Fig. 19. Western Sulawesi igneous rock isotope plot: 143Nd/144Ndv. 87Sr/86Sr (fields after Whitford & Jezek 1982; Serri et al. 1991).
419
MIOCENE W SULAWESI CONTINENT COLLISION
0.5132 I
I
I
I
I
,. omptex Rocks , o ene Larnas~C . 0.5127 -
I
I
~k .j&~
•
Z
~ ~,oC%~,~so,~
T--
-o~ 0.5122Z a-
-o m 0.5117]
J
0.5112 0.00
I 0.04
O Miocene Intrusive Igneous Rocks • Miocene Extrusive Igneous Rocks • Palaeogene? Lamasi Complex Rocks Palaeogene and Cretaceous Sedimentary Rocks I
0.08
I
I
0.12 0.16 147Sm/144Nd
I
I
0.20
0.24
Fig. 20. Western Sulawesi igneous rock isotope plot: 143Nd/144Nd v. Sm/144Ndisochron plot showing two pseudoisochrons for Cretaceous-Palaeogene Lamasi ophiolites and Miocene igneous rocks.
element (e.g. Hamilton 1979), the central Makassar Strait appears to be a Neogene foreland basin, bounded to the west by the E-vergent Kutei Basin fold belt and to the east by the W-vergent western Sulawesi thrust belt (Fig. 22). In the western part of the study area, the Majene foldbelt extends offshore into the Makassar Strait where regional multichannel seismic data show frontal folds and thrusts actively developing below 2000 m of water (Fig. 4). Further west, within the deepest parts of the Makassar Strait, undeformed, horizontally stratified Late Neogene sediments extend to the easterlydirected toe thrusts associated with the Mahakam Delta. In many parts of western Indonesia, the sites of major Palaeogene sediment accumulation (many related to rifting) were inverted during Late Miocene and Pliocene times in response to the Palawan, Bangga-Sula and Australian continental plate collisions. The resulting post-collisional Neogene sedimentary sequences were deposited within more tectonically quiet platform regions preserved between the zones of active inversion. Such structural inversions have been well documented especially along the eastern Sunda Shelf at the southern entrance to the Makassar Strait (Letouzey et al. 1990). The inversion episode in western Sulawesi was particularly intense, mainly because of its location
perpendicular and close to the main centre of collision. Extensive imbrication of individual thrust sheets collectively resulted in thrust loading which depressed the adjacent lithosphere of easternmost Sundaland to produce a major foreland basin. The erosion product of the western Sulawesi uplift (termed Walanae Formation in this study or 'Celebes Molasse' by van Bemmelen 1949) is actively being transported westwards and deposited in the central Makassar Strait (North Makassar Basin). To date, this foreland basin remains only partially filled. Preliminary flexural modelling of the cross section shown in Fig. 22 (Steve Crews pers. comm. 1994) demonstrates that the lithospheric load of the western Sulawesi thrust belt and the Mahakam-Delta is sufficient to explain the general topographic framework of the North Makassar Basin and western Sulawesi. In eastern Kalimantan, a similar although easterly directed Late Neogene sediment-thrust loading event took place in association with the development of the Mahakam Delta, which includes additional compressional structures related to toe-of-slope.gravity sliding (Ott 1987; van de Weerd & Armin 1992). The dominant, present-day structural trend includes a series of tightly folded NNE-SSW trending anticlines and synclines observable onshore at the surface and
Sphene 90SUL-25 90SUL-28
Zircon 90SUL-12 90SUL-15 90SUL-21A 90SUL-25 90SUL-26C 90SUL-28 unsized+M3 ° M4 ° + M 6 °
sample #
54.69 49.95 57.11 31.61 18.37 29.33 9.68 5.78
initial
57 42
41.88 38.84 38.02 23.19 10.41 17.09 4.78 1.48
final
weight (mg)
23 22 34 27 43 42 49 76
% loss
285 63
37 40 38 37 6 646 196 30
Pb (ng)
Table 6. Zircon and sphene U-Pb isotopic and age data
241 1240
853 764 890 962 432 1279 1335 958
U (ppm)
4.9 1.5
0.9 1.1 1 1.6 0.6 38 41 29
Pb (ppm)
22 256
412 542 888 733 212 1233 534 298
2o4pb
ratio 206pb
6.0236e-3 8.5312e-3
6.5980e-3 2.2290e- 1
1.0015e-3 2.9190e-2
235U
ratio 207pb
8.4174e-4 1.2770e-3
238U
ratio 2o6pb 238U (Ma)
206pb
235U (Ma)
207pb
age
206pb (Ma)
207pb
6.1 1.27
6.19 1.15
40 -248
5.42 6.1 281 8.22 8.63 121 6.85 7.48 217 7.84 12.73 1073 4.7780e-2 6.45 6.68 88.4 5.5389e-2 186 204 428 178 194 398 178 171 -2894 391 Ma = upper intercept concordia age
5.1901e-2 4.8450e-2
2o6pb
ratio 2o7pb
t,~
Fig. 21. Schematic E - W cross-section (present-day geography) illustrating the important points of the new tectonic model for western Sulawesi (after Coffield et al. 1993). Note that the westward-verging orogen of South Sulawesi is formed over a back-thrust or retro-arc thrust system in the hanging wall of a westward-dipping subduction zone. The Miocene volcanics are derived from melting of the subducted continental lithosphere of Australian affinity. See Fig. 1 for location of sections.
t,~
422
S . C . BERGMAN ET AL.
i)( ÷ T O M I N I BAY
4:"
fl /
EAST
LEGEND ORMATION8
THEMAHAJ(AM DELTA
OP}~:)UTESOF
FOLD& THRUST BELT80Ir W.
mm ~ = ~ , qq~ I~RUSl~ '~
PLATFORM
OTHERGRANITI~R
SEDIMI~rm
r
~?
'r
KUTEO.ASl.
'r
T
NORMALFAULTS
~. ~
WRENCHFAULT8 FOLDAXEtOFTHE KtrrB 8ASlH IIA1~METRY
MAKASSA. S T . A I T
WESTE..SULAWESl
.ONS
A'
& Thrust B,Ito)
Miihelmm Delta)
~n 5
km
i° T H -S -10 l"om~v~l
SIMPLIFIED
STRUCTURAL
CROSS-SECTION
| ~ ~ !
Fig. 22. Simplified geological map and NW-SE cross-section through the Makassar Strait showing the main Neogene stratigraphic and tectonic elements.
MIOCENE W SULAWESI CONTINENT COLLISION identifiable offshore as far east as the 200 m isobath (Courteney et al. 1991). The combined crustal effect of these two converging fold and thrust belts, together with their associated erosion products, is presently manifested as the deep-water Makassar Strait. The North Makassar Basin is, therefore, not the site of a Tertiary spreading centre, but the result of NW to SE shortening in response to a Neogene continentcontinent collision that took place in western Sulawesi (Fig. 22). A unified Late Cenozoic tectonic evolution of western Sulawesi should include a phase of obduction of the Lamasi ophiolite during the Palaeogene, followed by a Late Miocene continent-continent collision of an Australian Craton-derived block with Sundaland with associated fold and thrust-belt development and extensive lithospheric melting-related magmatism.
Conclusions The integration of existing data with new geochemistry, petrology, geochronology and field data lead to the following major conclusions.
(1) Miocene calc-alkaline to mildly alkaline and shoshonitic felsic to mafic plutonic and cogenetic volcanic rocks dominate the surface and shallow subsurface geology of western Sulawesi, forming over 70% of the area. Intrusive rocks include biotite-hornblende granitic rocks ranging from monzonite and quartz monzonite, to quartz monzodiorite and gabbro. (2) Dominant lithofacies include airfall and subaqueous tufts, tuff turbidites, lapilli tufts, tuff breccias, and volcanic conglomerates. These lithotypes form stratovolcano deposit sequences and volcanic apron successions > 5 km in thickness. Domes and lavas rarely occur. Intrusive bodies form sills, dykes, stocks, plugs, plutons and small laccolithic batholiths ranging from several metres to 100 km in maximum plan dimension. (3) Intrusive rocks are temporally, geochemically, and isotopically indistinguishable from volcanic rocks in the region and both are considered members of the same cogenetic suites. These Miocene igneous rocks form a bimodal compositional suite, with basaltic and dacitic compositional subsets, with minor intermediate composition andesites. (4) Tectonic framework classifications based on trace element discrimination diagrams indicate the western Sulawesi Miocene igneous rocks are transitional between volcanic arc and
423
collisional granitoids. However, the authors urge caution in application of these classification schemes owing to lithospheric source compositional heterogeneities. (5) New K-Ar ages of mineral separates from volcanic and intrusive rocks are indistinguishable and range from 2-18 Ma, averaging 8 _+4 Ma. Apatite, zircon and sphene fission track ages exhibit the same general relationship and are slightly younger than the K-Ar ages (2-14 Ma), and average 5 _+2 Ma. A two stage cooling history consisting of an initial magmatic cooling phase at 100-400°C Ma -1, followed by a secondary cooling phase related to denudation at c. 5-20°C km -1 is indicated by the geochronology data. Assuming geothermal gradients of 30°C km -1 for the most recent denudation cooling phase, these cooling rates correspond to uplift rates of 200700 m Ma -1, rates typical of fold and thrust belts such as Taiwan and the Himalayan chain. These values are confirmed by mineral equilibria geobarometry determinations of selected plutons, which indicate relatively shallowlevel intrusion of Miocene granitic laccoliths and plutons at 3-10 km depths. (6) 87Sr/86Srinitial values of Miocene igenous rocks (range 0.7035-0.790; n = 23) exhibit a positive correlation with 87Rb/86Sr, here interpreted with age significance. Whole rock Rb-Sr isochrons for an individual pluton, such as Palopo (506 Ma) are distinct from the best fit 'errorchron' or isochron of the entire suite (305 Ma), yet both can be interpreted to indicate a Palaeozoic 'closure' or 'homogenization' event of the lithospheric source region of the magmas. Whole rock 143Nd/144Nd values range from 0.511-0.525 and are correlated with Sm/Nd, indicating a possible isochronal relationship with errorchron ages of c. 0.6 Ga. Model ages of time since separation from depleted mantle range from 0.9-2.5 Ga. Zircon U/Pb age systematics of six representative granitic intrusive and andesitic extrusive samples possess nearly concordant 2°6pb/238U and 2°7pb/235U ages in the range 5-12Ma, similar to K-Ar biotite and hornblende ages on the same rocks. In contrast, 2°7pb/2°6pb ages are grossly discordant and are in the range 88-1073 Ma, suggesting the presence of recycled ancient zircons derived from Proterozoic to Palaeozoic lithospheric source rocks in the Miocene granitic magmas. Lithosphere of this character is absent in SE Sundaland and is abundant in the northern Australian plate; the most likely source of the Miocene Sulawesi igneous rocks was Australian plate lithosphere which became
424
s . c . BERGMAN ET AL.
accreted onto Sundaland during an Oligocene to Miocene collision event. (7) In contrast to the Miocene rocks, the Lamasi C o m p l e x ophiolites possess the most depleted 87Sr/86Sr and enriched 143Nd/ln4Nd values, and trace element systematics typical of mid-ocean ridge basalts or back-arc basin basalts. (8) The fission track, K-Ar and U-Pb ages, major and trace e l e m e n t geochemistry, mineral chemistry geobarometry and Rb-Sr/Nd-Sm/ U-Pb isotope c o m p o s i t i o n s of western Sulawesi M i o c e n e m a g m a t i c rocks form c o m p l e m e n t a r y and mutually i n d e p e n d e n t datasets which constrain the following petrogenetic features: western Sulawesi M i o c e n e Pliocene extrusive and intrusive rocks form cogenetic volcanoplutonic complexes of calcalkalic to mildly alkalic and potassic, felsic to mafic magmatic rocks which were erupted and intruded during a relatively short episode (2-18 Ma) of lithospheric t h i c k e n i n g and melting, possibly involving lithosphere delamination. Parental source rocks of the M i o c e n e melts were Late Proterozoic to Palaeozoic crustal and mantle lithospheric assemblages which became heated and melted due to a continent-continent collision in which west vergent crust of the A u s t r a l i a n - N e w Guinea plate was subducted beneath easternmost Sundaland. Regional e a s t - w e s t compression continued into Pliocene to Recent times and resulted in the development of a Late Neogene fold and thrust belt with ramp-related imbrication p r o v i d i n g e n h a n c e d regional uplift rates of 200-700 m Ma -1 during the last 2-18 Ma. (9) The effects of Late Neogene thrust loading in western Sulawesi combined with sediment loading in eastern Kalimantan, associated with the d e v e l o p m e n t of the M a h a k a m Delta, resulted in crustal subsidence along much of the Makassar Strait. The North Makassar Basin is interpreted as an actively forming foreland basin located immediately in front of the Majene foldbelt. The Neogene initiation of the Makassar Strait subsidence represents an important surface manifestation of the western Sulawesi continent-continent collision episode. GRDC geologists Amarudin, Supandjono, Sukido, Trisna and their colleagues provided much assistance in the field. We thank Theo Van Leeuwen, Imans Kavalieris, Maury Cucci, Scot Krueger, Charles Hutchison and Mike Crawford for comments, discussions and insights. R. Maury and C. Rangin provided preprints of publications during an early phase of this study. We thank Jeff Corrigan and Marty Robinson for reviews of an earlier version of the manuscript. We appreciate the meticulous
analytical work provided by Shaft Kelley, Ken Foland, Fritz Hubacher, Long Liang, Tom Bills, Jenny Hopps, Paul Green, Ian Duddy and Paul Mueller. We thank Steve Crews for timely flexural modelling of the thrust belt. Christopher Parkinson, Robert McCaffrey and Martin Menzies provided most useful reviews, for which we are extremely grateful. David Nicklin, Gene Richards, Jamie Robertson, Suherman and John Duncan are thanked for their support of this project, and we thank Pertamina, ARCO and KG Kalosi for permission to publish this research.
Appendix: analytical methods Analytical work was performed at ARCO Exploration Research Laboratories in Plano, Texas and at several university and industrial laboratories. Detailed analytical data are available from the senior author upon submission of a formal written request. Sample details. Samples, weighing 2-10 kg each, were trimmed from Sulawesi surface outcrops. Representative splits were subjected to a variety of analyses and mineral separation. Modal analyses were performed on representative coarse-grained plutonic lithologies using a 0.4 mm grid and an entire 27 × 46 mm thin section (n > 500 points). Analytical techniques. Samples were crushed using a conventional jaw crusher and disk mill, and biotite, hornblende, sanidine, plagioclase, apatite, sphene and zircon were concentrated using conventional heavy liquid and magnetic methods at Krueger Enterprises, Cambridge, MA. Fission track analysis. The external detector method (Naeser 1979; Gleadow et al. 1983, 1986) was used for individual grain dating of apatite, sphene and zircon. Apatite fission track age and length measurements were made in the laboratories of Geotrack International, Melbourne, Australia (G. Laslett & D. Arne, analysts). Zircon, sphene, and some apatite FT analyses were performed by S. Kelley in the laboratories of Southern Methodist University Geology Department. Apatite grains were mounted in epoxy, polished to expose grain interiors, and etched for 25 seconds in 5 molar HNO 3 to reveal spontaneous fission tracks. Nonmagnetic, > 3.1 gm/cc splits containing lexcessive pyrite were treated with aqua regia to concentrate zircon. Zircon grains were mounted in melted FEP Teflon tape, polished and etched in a KOH-NaOH eutectic at 230°C tbr 3-6 hours. Since most samples contained multiple zircon populations (based on morphology and etching character), two mounts were made, and each etched for different times to achieve the optimum etch for the full range of grains. Sphene grains were mounted in FEP teflon tape, polished, and etched in a 5.0 M NaOH solution at 125°C for 4 hours. Apatite, sphene and zircon mounts were covered with low U muscovite detectors, apatites sandwiched between standards including Durango apatite, NBS glass SRM962 and Corning CN-6 glass, and sphenes and zircons between Fish Canyon and Myalla Road syenite zircon age standards as well as the two glasses. Both packages were
MIOCENE W SULAWESI CONTINENT COLLISION irradiated at the Texas A & M reactor (sphene and zircon) or the HIFAR reactor, Australia (apatite). Zircons and sphenes were irradiated at a flux of 6.4-7.0x 1015 neutrons/cm 2, and apatites at .7-1.4 x 1016 neutrons/cm 2. Reactor neutron fluence was calculated using accepted ages of 27.9, 172.8 and 31.4 Ma for Fish Canyon, Myalla Road and Durango standards and flux gradients were accounted for using the glass standards. Muscovite detectors were etched for 13 minutes in 48% HF to reveal fission tracks resulting from induced fission of 235U during neutron irradiation of apatite and zircon. Individual grain ages were calculated using the formulation of Price & Walker (1963); grain age uncertainty calculations are those of Hurford et al. (1984). Grain age uncertainties are quoted at the 1 sigma confidence interval. Apatite confined track length measurements were made using a Leitz microscope with a 100x air lense by Geotrack, camera-lucida tube and a digitizing tablet. Uncertainties are _.+0.2 pm. Only horizontal tracks with well developed tips in grains with prismatic faces were measured. Major and minor element contents of dated apatites from selected samples were determined to better interpret the FT age and length data using wavelength-dispersive spectroscopic analyses on a 1983 JEOL 733 Superprobe at the ARCO Advanced Analytical Services Laboratory in Piano, TX (L. Liang, analyst). Natural mineral standards were used to convert X-ray count rates to elemental concentrations; operating conditions were: 15 kV potential, 15 nA beam current, rastered beam 10-20 lain square and t0--30 second count times; an interlayered Si-W crystal was used for F determinations. Analytical uncertainties are _+ 1-3 relative % for major elements (> 10 wt%) and _ 5-10 relative % for minor elements (1-10 wt%) and > _+ 10-20 relative % for elements present in concentrations < 1 wt%. OH was calculated by difference assuming anion stoichiometry. Thermal modelling of apatite FT data was performed using the kinetic models developed by Laslett et al. (1987), Carlson (1990) and Corrigan (1991) which are capable of predicting or inverting the degree of length and age reduction as a function of time and temperature.
425
K-Ar and 4°Ar-39Ar analyses. Conventional 4°K-4°Ar dating was performed on whole rock or mineral concentrate samples in the laboratory of Geochron Inc., Cambridge, MA (Tom Bills and Jenny Hopps, analysts) using conventional techniques (flame photometry for K determinations and gas source mass spectrometry for 4°Ar measurements). Mineral concentrates were seived to -80/+200 mesh and whole rock and feldspar splits were washed in dilute HF and HNO 3 to remove alteration phases. Uncertainties were calculated using an empirical formulation dependent on K-content and age. Two or three splits were analyzed and averaged to generate the observed ages. 4°Ar-39Ar thermal release analyses were performed by Ken Foland and coworkers at the Ohio State Geochronology Laboratories, Columbus, OH, using conventional techniques. Rb-Sr and Nd-Sm isotope analyses. Conventional isotope dilution solid source mass spectroscopy was used to determine 87Sr/86Sr and 143Nd/t44Nd ratios and Rb, Sr, Nd and Sm contents in the geochronology laboratory of Ohio State Dept Geology and Mineralogy, Columbus, Ohio (F. Schubacher and K. Foland, analysts). U-Pb zircon analyses. Zircon and sphene concentrates prepared by Geochron were hand picked for purification and further separated into magnetic splits using a Frantz magnetic barrier separator. Conventional U-Pb solid source mass spectroscopy was used to determine U-Pb ratios, and U and Pb contents in the Geochronology Laboratories of Univ. Florida, Dept Geology, Gainesville, (P. Mueller, analyst). Major and trace element geochemistry. Major element contents were determined using conventional X-ray fluorescence spectroscopy, minor and trace elements using wet chemistry, ICP-MS, DCP, XRF, INAA, and other techniques in the laboratories of X-ray Assay Laboratories, Toronto. Analytical uncertainties are < 1-5 relative % for major elements (present in concentrations > 10 wt%) and < 5-20 relative % for minor and trace elements.
References ABENDANON, E. C. 1915. Geologische en Geographische Doorkruisingen van Midden-Celebes (1909-1910). E. J. Brill, Leiden, 1. AVDLEV-CHARLES, M. G. 1977. Mesozoic evolution of the margins of Tethys in Indonesia and the Philippines. In: Proceedings 5th Indonesia Petroleum Association Convention. 2, 25-52. 1987. Dispersal of Gondwanaland: relevance to the dispersal of the angiosperms. In: WHITMORE, T. C. (ed.) Biogeographical Evolution of the Malay Archipelago. Clarendon Press, Oxford, Oxford Monograph, 4, 91-102. 1991. Tectonics of the New Guinea area. Annual Reviews of Earth and Planetary Sciences, 19, 17-41. &HARRtS, R. A. 1991. Allochthonous terranes of the southwest Pacific and Indonesia. Philosophical Transactions of the Royal Society of London, A331, 115-131.
-
-
-
-
-
-
BELLON, H. & RANGIN, C. 1991. Geochemistry and isotopic dating of Cenozoic volcanic arc sequences around Celebes and Sulu Seas. In: SILVER, E. A., RANGIN, C., VON BREYMANN, M. Z., ET AL. (eds) Proceedings O c e a n Drilling Program, Scientific Results, 124, 321338. BERGMAN, S. C. 1987. Lamproites and other Potassiumrich igneous rocks: a review of their occurrence, mineralogy and geochemistry. In: FITTON, J.G. & UPTON, B. J. G. (eds) Alkaline Igneous Rocks. Geological Society, London, Special Publications 30, 103-190. --, COFFIELD, D. Q., DONELICK, R., CORRIGAN, J., TALBOT, J., ET AL. 1993. Late Cenozoic compressional and extensional cooling and exhumation of the Qomolangma (Mt. Everest) region, Nepal. Geological Society America Abstracts with Programs, 25, A243.
426
S. C, BERGMAN ET AL.
BERRY, R. E & GRADY,A. E. 1987. Mesoscopic structures produced by Plio-Pleistocene wrench faulting in South Sulawesi, Indonesia. Journal Structural Geology, 9, 563-571. BIANTORO, E., MURITNO, B. P. & MAMUAYA,J. M. B. 1992. Inversion faults as the major structural control in the northern part of the Kutei Basin, East Kalimantan. In: Proceedings 21st Indonesia Petroleum Association Convention. 1, 45-68. BIRD, P. 1979. Continental delamination and the Colorado Plateau. Journal of Geophysical Research, 84, 7561-7571. BROWN, G. C. 1982. Calc-alkaline intrusive rocks: their diversity, evolution, and relation to volcanic arcs. In: THORPE, R. S. (ed.) Andesites, orogenic andesites and related rocks. J. Wiley and Sons, New York, 437-461. BUROLLET, P. E & SALLE, C. 1981. Seismic reflection profiles in Makassar Strait. In: BARBER, A. J. & Wiryosujono, S. (eds) Geology and tectonics of the eastern Indonesia. GRDC, Special Publication, 2, 273-276. CAREY, S. W. 1976. Tectonic evolution of South East Asia. In: Proceedings 4th Indonesia Petroleum Association Convention, 2, 3-23. CARLSON,W. D. 1990. Mechanisms and kinetics of apatite fission track annealing. American Mineralogist, 75, 1120-1139. COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A., GURITNO, N., ROBINSON, N. M. & TALBOT, J. P. 1993. Tectonic and stratigraphic evolution of the Kalosi PSC area, and associated development of a Tertiary petroleum system. In: Proceedings 22nd Indonesia Petroleum Association Convention. 679-706. CORRIGAN, J. D. 1991. Inversion of apatite fission track data for thermal history information. Journal Geophysical Research, 96, 10347-10360. COURTENEY, S., COCKCROFT,P., LORENZ,R., MILLER, R., OTT, H. L., ET AL. 1991. Indonesian Oil and Gas Field Atlas, vol. V. Kalimantan, Indonesian Petroleum Association, Jakarta. Cox, K. G., BELL, J. D. & PANKHURST,R. J. 1979. The interpretation of igneous rocks. Allen & Unwin, London. DALY, M. C., COOPER,M. A., WILSON, I., SMITH,D. G. & HOOPER, B. G. D. 1991. Cenozoic plate tectonics and basin evolution in Indonesia. Marine and Petroleum Geology, 8, 2-21. DAVIES, J. H. & VON BLACKENBURG, E 1995. Slab breakoff: a model of lithosphere detachment and its test in the magmatism and deformation of collisional orogens. Earth and Planetary Science Letters, 129, 85-102. DELISLE, G. 1986. The subsurface temperature field resulting from obduction of a crustal segment. Geologische Jahrbuch, E34, 77-85. DIETRICH, V. & GANSSER, A. 1981. The leucogranites of the Bhutan Himalaya (Crustal anatexis versus mantle melting). Schweizerische Mineralogische und Petrogropische Mitteilungen, 61, 177-202. DJURI & SUDJATMIKO1974. Geologic map of the Majene and western part of Palopo quadrangles, South
Sulawesi, scale 1:250,000. Geological Research and Development Centre, Bandung. DODSON, M. H. 1973. Closure temperature in cooling geochronological and petrological systems. Contributions Mineralogy and Petrology, 40, 259-274. & MCCLELLAND-BROWN, E. 1985. Isotopic and paleomagnetic evidence for rates of cooling, uplift, and erosion. In: SNELLING,. N. J. (ed.) The chronology of the geological record. Geological Society, London, Memoir, 10, 315-325. EFFENDI, L. 1993. Selat Makasar merupukan wilayah kompleks antara perairan Bagian Barat dan Timur. In: Proceedings 22rid Annual Convention Indonesian Association Geologists, 2, 950-961 (in Bahasa). ENGLAND, P. C. & THOMPSON, A. 1986. Some thermal and tectonic models for crustal melting in continental collision zones. In: COWARD,M. P. & RIES, A. C. (eds) Collision Tectonics. Geological Society, London, Special Publications, 19, 83-94. EWART, A. J. 1982. The mineralogy and petrology of Tertiary-Recent orogenic volcanic rocks: with special reference to the andesite-basalt compositional range. In: THORPE, R. S. (ed.) Andesites, orogenic andesites and related rocks. J. Wiley and Sons, New York, 25-87. FAt;GERES, J.-C., GAYET, J. & GONTHIER, E. 1989. Microphysiographie des depots quaternaires darts le detroit de Makassar (ocean Indien). Opposition entre une marge stable (Borneo, Kalimantan) et une marge active (Celebes, Sulawesi). Bulletin Soci~t~ gdologique France, 8, 807-818. GARRARD, R. A. 1989. Sengkang Basin, South Sulawesi. Post-convention field trip guidebook, Indonesian Petroleum Association, Jakarta. --, NUSATRIYO, G. & COFFIELD, D. Q. 1992. The prospectivity of early Tertiary rift sequences in the Neogene foldbelts of South Sulawesi. Extended Abstract. The Eastern Indonesia Symposium. Petramina and Simon Petroleum Technology, Jakarta. GEORGE, W. 1981. Wallace and his line. In: WHITMORE, T. C. (ed.) Wallace's Line and plate tectonics. Clarendon Press, Oxford, 3-8. GILL, J. B. 1981. Orogenic andesites and plate tectonics. Springer-Verlag, New York. GLEADOW,A. J. W., DUDDY,I. R. & LOVERING,J. E 1983. Fission track analysis: a new tool for the evaluation of thermal histories and hydrocarbon potential. Australian Petroleum Association Journal, 23, 93-102. --, GREEN, P. F. & LOVERING, J. E 1986. Confined fission track lengths in apatite - a diagnostic tool for thermal history analysis. Contributions to Mineralogy and Petrology, 94, 405-415. GREEN, P. E, DUDDY, I. R., LASLETT, G. M., HEGARTY, K. A., GLEADOW,A. J. W. & LOVERING,J. E 1989. Thermal annealing of fission tracks in apatite: 4. Quantitative modelling techniques and extension to geological timescales. Chemical Geology (Isotope Geoscience Section), 79, 155-182.
MIOCENE W SULAWESI CONTINENT COLLISION HAMILTON, W. 1978. Tectonic map of the Indonesian region. US Geological Survey, Map 1-875-D. 1979. Tectonics of the Indonesian region. US Geological Survey, Professional Paper, 1078. 1989. Convergent-plate tectonics viewed from the Indonesian region. In: SENCOR, A. M. C. (ed.) Tectonic evolution of the Tethyan region. Kluwer Academic Publishers, Dordrecht, 655-698. HAMMARSTROM,J. M. & ZEN, E-AN 1986. Aluminum in hornblende: an empirical igneous barometer. American Mineralogist, 71, 1297-1313. HARRIS, N. B. W., PEARCE, J. A. & T1HOLLE,A. G. 1986. Geochemical characteristics of collision-zone magmatism. In: COWARD,M. P. & REIS, A. C. (eds) Collision tectonics, Geological Society, London, Special Publications, 19, 67-81. HARRIS, R. A. 1989. Processes of allochthon emplacement, with special reference to the Brooks Range ophiolite, Alaska and Timor, Indonesia. PhD Thesis, University of London. HASAN, K., GARRARD, R. & MAHODIM, P. 1991. South Sulawesi, Post-convention field trip guidebook, Indonesian Petroleum Association, Jakarta. H1LL, K. C. & GLEADOW, A. J. W. 1989. Uplift and thermal history of the Papuan fold belt, Papua New Guinea: apatite fission track analysis. Australian Journal Earth Sciences, 36, 515-539. HOLLISTER, L. S., GRISSOM, G. C., PETERS, E. K., STOWELL,H. H. & SISSON,V. B. 1987. Confirmation of the empirical correlation of A1 in hornblende with pressure of solidification of calc-alkaline plutons. American Mineralogist, 72, 231-239. HURFORD, A. J., FITCH, E J. & CLARKE, A. 1984. Resolution of the two age structure of the detrital zircon populations of the Lower Cretaceous sandstones from the Weald of England by fission track dating. Geological Magazine, 121, 269-296. HUTCHISON, C. S. 1982. Indonesia: regional distribution and characteristics. In: THORPE, R. S. (ed.) Andesites, orogenic andesites and related rocks. J. Wiley, London, 207-224. 1989a. Geological evolution of South-east Asia. Oxford University Press. 1989b. Displaced terranes of the southwest Pacific. In: BEN-AVRAHAM, Z. (ed.) The Evolution of the Pacific Ocean Margins. Oxford University Press, 161-174. JAUPART, C. & PROVOST, A. 1985. Heat focusing, granite genesis and inverted metamorphic gradients in continental collision zones. Earth and Planetary Science Letters, 73, 385-397. JEZEK, P. A., WHITFORD, D. J. & GILL, J. B. 1981. Geochemistry of recent lavas from the SangiheSulawesi arc, Indonesia. In: BARBER, A. J. & WmYOSUJONO, S. (eds) Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication, 2, 383-389. KARABtNOS, R & KETCHAM, R. 1988. Thermal structure of active thrust belts. Journal of Metamorphic Geology, 6, 559-570. KATILI, J. A. 1971. A review of the geotectonic theories and tectonic maps of Indonesia. Earth Science Reviews, 7, 143-163. -
-
-
-
-
-
427
1975. Volcanism and plate tectonics in the Indonesian islands. Tectonophysics, 26, 165-168. 1978. Past and present geotectonic position of Sulawesi, Indonesia. Tectonophysics, 45, 289-322. KAVALIERIS,I., VAN LEEUWEN,T. M. & WILSON, M. 1992. Geologic setting and styles of mineralization, north arm of Sulawesi, Indonesia. Journal SE Asia Earth Sciences, 7, 113-129. KAY, R. W. & KAY, S. M. 1991. Creation and destruction of the lower continental crust. Geologische Rundschau, 80, 259-278. & 1993. Delamination and delamination magmatism. Tectonophysics, 219, 177-189. KUNO, H. 1968. Origin of andesite and its bearing on island arc structure. Bulletin Volcanologique, 32, 141-17. LASLETT, G. M., GREEN, E E, DUDDY, I. R. & GLEADOW, A. J. W. 1987. Thermal modelling of fission tracks in apatite: 2. A quantitative analysis. Chemical Geology, 65, 1-13. LAMEYRE, J. & BOWDEN, P. 1982. Plutonic rock types series: discrimination of various granitoid series and related rocks. Journal Volcanology and Geothermal Research, 14, 169-186. LAPOUILLE, A., HARTONO, H. M. S., LAROU, M., PRAMUNIJOYO, S. & LARDY, M. 1986. Age and origin of the seafloor of the Banda Sea (Eastern Indonesia). Oceanolica Acta, 81, 379-389. LETERRIER, J., YUWONO, Y. S., SOERIA-ATMADJA, R. & MAURY, R. C. 1990. Potassic volcanism in central Java and South Sulawesi, Indonesia. Journal SE Asia Earth Sciences, 4, 171-187. LETOUZEY, J., WERNER, R & MARTY, A. 1990. Fault reactivation and structural inversion. Backarc and intraplate compressive deformations. Example of the eastern Sunda shelf (Indonesia). Tectonophysics, 183, 341-362. MALECEK, S. J., REAVES, C. M., SOERIA-ATMADJA,R. & WIDIANTARA, K. O. 1993. Seismic stratigraphy of Miocene and Pliocene age outer shelf and slope sedimentation in the Makassar PSC, offshore Kutei Basin. Proceedings 22nd Indonesia Petroleum Association Convention. 345-371. MATSUBAYASHI, O. & NAGAO, T. 1991. Compilation of heat flow data in SE Asia and its marginal seas. In: CERMAK, V. & RYBACH, L. (eds) Terrestrial heat flow and the lithosphere structure. Springer-Verlag, New York, 444-456. MUBROTO,B., BRIDEN,J. C., MCCLELLAND,E. & HALL, R. 1994. Paleomagnetism of the Balantak ophiolite, Sulawesi. Earth and Planetary Science Letters, 125, 193-209. NAESER, C. W. 1979. Fission track dating and geological annealing of fission tracks. In: JAGER, E. & HUNZIKER,J. C. (eds.) Lectures in Isotope Geology. Springer Verlag, N.Y. 154-169. & FAUL, H. 1969. Fission track annealing in apatite and sphene. Journal of Geophysical Research, 74, 705-710. NISH~URA, S. 1986. Neo-tectonics of East Indonesia. In: Geological Society of China Memoirs, 7, 107-123. --, YOKOYAMA,T. & HERRY 1980. Gravity measurements at South Sulawesi. Phys. Geol. Indonesian Island Arcs, 35-41. -
-
428
S, C. BERGMAN ET AL.
OTT, H. L. 1987. The Kutei Basin-A unique structural history. Proceedings 16th Indonesia Petroleum Association Convention, 307-316. PARKINSON, C. D. 1991. The petrology, structure, and geologic history of the metamorphic rocks of central Sulawesi. PhD Thesis, University of London. PEARCE, J. A., HARRIS, N. B. W. & TINDLE, A. G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25, 956-983. PIETERS, P. E., PIGRAM, C. J., TRIAL, D. S., Dow, D. B, RATMAN,N. & SUKAMTO,R. 1983. The stratigraphy of W Irian Jaya. Bulletin Geological Research Centre, Bandung, Indonesia, 8, 14--48. PIGRAM, C. J. & Panggabean, H. 1984. Rifting of the northern margin of the Australian continent and the origin of some microcontinents in eastern Indonesia. Tectonophysics, 107, 331-353. PRIAD1, B., POLVE, M., MAURY, R. C., SOERIA-ATMADJA, R. & BELLON, H. 1993. Geodynamic implications of Neogene potassic calc-alkaline magmatism in Central of Sulawesi: geochemical and isotopic constraints. Proceedings 22nd Annual Convention Indonesian Association Geologists, 1, 59-81. --, BELLON, H., SOERIA-ATMADJA, R., JORON, J. L. & COYrEN, J. 1994. Tertiary and Quaternary magmatism in Central Sulawesi: chronological and petrological constraints. Journal SE Asia Earth Sciences, 9, 81-93. PRICE, P. B. & WALKER, R. M. 1963. Fossil tracks of charged particles in gneiss and the age of minerals. Journal of Geophysical Research, 68, 4847-4862. RANGIN, C. & SILVER, E. A. 1991. Neogene tectonic evolution of the Celebes-Sulu basins: new insights from leg 124 drilling. In: SILVER,E. A., RANGIN,C., VON BREYMAYN, M. T., Er AL. (eds) Proceedings Ocean Drilling Program Scientific Results. 124, 51--63. , PUBELLIER,M., AZEMA, J., BRIAIS, A., ET AL. 1990. The quest for Tethys in the western Pacific. 8 paleogeodynamic maps for Cenozoic time. Bulletin de la Socigtg Gdologique de France, 6, 907-913. RATMAN, N. 1976. Geologic map of the Toli-Toli quadrangle. North Central Sulawesi, Geological Survey Indonesia. -& ATMAWINATA, S. 1988. Geological report of Mamuju Quadrangle, South Sulawesi. Geologic Research and Development Center, Bandung, Open File Report, 1:250 000. REYZER, J. 1920. Geologische aantekeningen betreffende de Zuidelijke Toraja-landen, verzameld uit de verslagen der mijnbouwkundige onderzoekingen in Midden-Celebes. In: Jaarboek van het Mijnwezen Nederlandsch Oost-lndie 1918. Weltevreden, Gov't Printing Office, 154-209. RUTTER, M. J., VAN DER LAAN, S. R. & WYLLIE, P. J. 1989. Experimental data for a proposed empirical igneous geobarometer: Aluminum in hornblende at 10 kbar pressure. Geology, 17, 897-900. SASAJIMA, S., NISHIMURA, S., HIROOKA, K., OTOFIJI, Y., VAN LEEUWEN, T. & HEHUWAT, F. 1980. Paleomagnetic studies combined with fission-track datings on the western arc of Sulawesi, East Indonesia. Tectonophysics, 64, 163-172.
SAx, H. G. J. 1931a. Geological Research South West Celebes. Confidential BPM Report, Balikpapan, Indonesia. 1931b. Temporary Report of the Geological Research of the Pare-Pare and Loewoe Permits. Confidential BPM Report, Balikpapan, Indonesia. SCHMIDT,M. W. 1992. Amphibole composition in tonalite as a function of pressure: an experimental calibration of the Al-hornblende barometer. Contributions to Mineralogy and Petrology, 110, 304-310. SERRI, G., SPADEA, P., BECCALUVA, L., CIVETTA, L., COLTORT1,M., DOSTAL,J., Er AL 1991. Petrology of igneous rocks from the Celebes Sea basement. In: SILVER, E.A., RANGIN, C., VON BREYMANN,M. T., ET AL. (eds) Proceedings Ocean Drilling Program Scientific Results. 124, 271-296. SHAW,R. D. & PACKHAM,G. H. 1992. The tectonic setting of sedimentary basins of eastern Indonesia: Implications for hydrocarbon prospectivity, The APEA Journal, 195-213. SILVER, E. A., MCCAFFREY, R. & SMITH, R. B. 1983a. Collision, rotation, and the initiation of subduction in the evolution of Sulawesi, Indonesia, Journal of Geophysical Research, 88, 9407-9418. --, JOYODrW~RYO, Y. & STEVENS, S. 1983b. Ophiolite emplacement by collision between the Sula platform and the Sulawesi island arc, Indonesia. Journal of Geophysical Research, 88, 9419-9435. SITUMORANG, B. 1982a. The formation and evolution of the Makassar Basin, Indonesia. PhD Thesis, University of London. 1982b. The formation of the Makassar Basin as determined from subsidence curves. In: Proceedings llth Indonesia Petroleum Association Convention. 383-107. 1984. Formation, evolution, and hydrocarbon prospects of the Makassar Basin, Indonesia. In: Proceedings of the 3rd Circum Pacific Energy and Minerals Research Conference. 227-231. SUKAMTO, R. 1975. Ujung Pandang, Sheet VIII, 1:1,000,000. Geologic Map of Indonesia, Bandung. 1978. The structure of Sulawesi in the light of plate tectonics. Proceedings Regional Conference Geological Mineral Resources SE Asia, 1975. Jakarta, 121-141. --. 1982. Geologic map of the Pangkajene and western part of Watampone quadrangles, South Sulawesi, scale 1:250,000. Geological Research and Development Centre, Bandung. & SIMINDJUNTAK,T. O. 1983. Tectonic relationship between geologic provinces of western Sulawesi, eastern Sulawesi, and Banggai-Sula in the light of sedimentological aspects. Bulletin Geological Research Centre, Bandung, 7, 1-12. SUN, S.-S. & MeDONOU~H, W. E 1989. Chemical and isotope systematics of oceanic basalts: implications for mantle composition and processes. In: SAUNDERS, A. D. & NORRY, M. J. (eds) Magmatism in the ocean basins. Geological Society, London, Special Publications, 42, 313-329. -
-
-
-
MIOCENE W SULAWESI CONTINENT COLLISION SUNG, G. L. 1948. Samenvatting van belangrijkere geologische gegevens over Celebes. GL.A. Report No 22575, Pertamina. THAMRIN, M. 1985. An investigation of the relationship between the geology of Indonesian sedimentary basins and heat flow density. Tectonophysics, 121, 45-62. 1986. Terrestrial heat flow map of Indonesian basins. Indonesian Petroleum Association Proceedings, 33-70. T'HOEN, C. W. & ZmGLZR, K.J. 1917. Verslag over de resultan van geologisch-mijnbouwkundige verkenningen en opsporingen in Zuidwest Celebes. Jaarb. Mijnw. Nederlandsch Oost-lndie, 1915, 44, 237-363. VAN BEMMELEN, R. W. 1949. The geology of Indonesia. The Hague, Netherlands, Govt. Printing Office. VAN BERGEN, M. J., VROON, P. Z. & HOOGEWE~'F, J. A. 1993. Geochemical and tectonic relationships in the east Indonesian arc-continent collision region: implications for the subduction of the Australian passive margin. Tectonophysics, 223, 97-116. vAN DE WEEe,D, A. A. & At~l~, R. A. 1992. Origin and evolution of the Tertiary hydrocarbon basins in Kalimantan (Borneo), Indonesia. AAPG Bulletin, 76, 1778-1803. VAN LZEUWEN, T.M. 1981. The geology of South Sulawesi with special reference to the Biru area. In: BARBER,A. J. & WtRYOSUJONO,S. (eds) Geology and Tectonics of Eastern Indonesia. Geological
429
Research Development Centre, Special Publication, 2, 277-304. , TAYLOR, R., COOTE, A. & LONSTA~E, E J. 1994. Porphyry molybdenum mineralization in a continental collision setting at Malala, northwest Sulawesi, Indonesia. Journal Geochemical Exploration, 50, 279-315. WHITFORD, D. J. & JEZEK, P. A. 1982. Isotopic constraints on the role of subducted sialic material in Indonesian island-arc magmatism. Geological Society of America Bulletin, 93, 504-513. WOOD, B. G. M. 1985. The mechanics of progressive deformation in crustal plates-a working model for SE Asia. Geological Society Malaysia Bulletin, 18, 55-99. YUWONO, Y. S. 1987. Contributions gt l'~tude du volcanisme potassique de l'Indondsie. Exemples du sud-ouest de Sulawesi et du Volcan Muria (Java). Doctorat Th~se de l'Universit6 de Bretagne Occidentale, Brest. - - . , MAURY, R. C., SOERIA-ATMADJA,R. & BELLON, H. 1988a. Tertiary and Quaternary geodynamic evolution of S Sulawesi: constraints from the study of volcanic units. Geologi Indonesia, 13, 32-48. - - . , PRIYOMARSONO,S., MAURY,R. C., RAMPNOUX,J. P., SOERIA-ATMADJA, R., ET AL. 1988. Petrology of the Cretaceous magmatic rocks from Meratus Range, southeast Kalimantan. Journal SE Asian Earth Sciences, 2, 15-22.
SE Sundaland accretion: palaeomagnetic evidence of large Piio-Pleistocene thin-skin rotations in Buton J A S O N R. A L I 1, J O H N M I L S O M 2, E D W A R D M. F I N C H 2 & B U N D A N M U B R O T O 3
1 Department of Oceanography, The University, Southampton, UK. 2 Department of Geological Sciences, University College, London, UK. 3 Geological Research and Development Centre, Bandung, Indonesia. Abstract: The Tukang Besi Platform, an Australian microcontinental fragment, began docking with Sundaland in the Pliocene, impacting on east Buton (SE Sulawesi). Fortuin et al. (1989; Journal of SE Asian Earth Sciences, 4, 107-124) postulated that south Bnton had rotated clockwise through about 60 ° relative to central/north Buton in response to the Tukang Besi collision. A palaeomagnetic investigation was carried out to test this model. Some 41 (of 72) palaeomagnetic sites from the upper Neogene Tondo and Sampolakosa Formations on Buton yielded interpretable data. Sites from south Buton record locally consistent directions, but at sampling localities < 25 km apart deflections are between 0 ° and 35 ° clockwise. There is no trend in these data with respect to the broad geotectonic setting. In central Buton declination offsets are negligible. In north Buton localized (kilometre scale) large (30-60 °) clockwise and counterclockwise declination offsets are observed. Effectively the Buton data provide spot markers on cover sequences that have been locally deformed as 'thin-skin' sheets. The underlying basement may have experienced the large relative motions proposed by Fortuin et al. (1989) but the upper Cenozoic cover has not been deformed in such a simple way. This study demonstrates that thinskin sheets associated with continental collision may undergo 30-60 ° rotations within very short intervals (< 2-3 Ma).
The Indonesian archipelago has been a zone of active convergence throughout the Tertiary. Since at least the early Miocene, the tectonic development of the area has been dominated by the interactions between the Sundaland margin of Eurasia and the Indo-Australian and Philippine Sea plates. Convergence of these plates on their triple junction in eastern Indonesia has led to obduction, subduction, fragmentation and shuffling of numerous microplates, typically covering 104 to 105 km 2, within an ever-decreasing area. The boundaries between these fragments have probably been generally short-lived (< 5 Ma) and over the next 10-20 Ma many will be incorporated into either Sundaland or Australia. This process is already well advanced in the western Banda Sea where Australian margin fragments such as the Banggai and Sula Islands, Buton and the Tukang Besi Platform have collided with the Sundaland margin represented by the large island of Sulawesi. A quantitative understanding of the development of this margin will provide valuable insights conceming the processes (and rates) involved in terrane accretion at continental margins. In this paper new palaeomagnetic data are presented from upper Neogene rocks on Buton and the results related to the Plio-Pleistocene collision processes in the region.
Buton Buton measures 150 km from north to south and approximately 60 km across at its widest point (Fig. 1). Topographically, it differs markedly from Muna Island to the west and the islands of the Tukang Besi platform to the southeast, being generally mountainous, with peaks in excess of 1000m. Geologically, the island attracted early interest because of the occurrence of deposits of asphalt which are worked commercially at a number of localities. Because of this, the geology of the island is well known, at least by comparison with most of the other islands of eastern Indonesia. Present understanding is based on the published work of Wiryosujono & Hainim (1978), Smith (1983), Fortuin et al. (1989) and Smith & Silver (1991), as well as unpublished work by M. E. M. de Smet (pers. comm. 1992). In addition, a considerable amount of land and marine seismic work has been carded out by oil companies exploring for hydrocarbons in the region (Davidson 1991). Other marine seismic, gravity and magnetic surveys have formed parts of broader scale investigations of the Banda Sea by US oceanographic institutes. There have been onshore gravity surveys on Buton, Muna and one of the Tukang Besi islands and in the extreme southeast of Sulawesi by a joint University
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 431-443.
431
432
J.R. ALI ET AL.
r-
--
41, "N
SULAWESI
l
I,
4o30'
Tanjung Batu
BUTON
R
5o30'
Traverse
5 O k m
O
I
Alluvium Quaternary o~ [ 1 ~ ] post. , . , .: WapolakaFormation (reef limestone) collision [U Pliocenet units pper Miocene~ SampolakosaFormation (chalk and marl) [Middleto Upper~ TondoFormation (fine - coarse clastics) W Miocene ~ Pre-Neogeneformations
/ (,
~
250km
Fig. 1. Geological map of Buton, SE Sulawesi (based on Smith & Silver 1991) with palaeomagnetic sample sites: 1, Bungi River; 2, Sampolakosa River (no reliable palaeomagnetic data obtained, ND); 3, Kemberu River; 4, Sampolakosa Bridge, (ND); 5, Warumbia River, Lapodi; 6, Wabiau River; 7, Wakalumbe River, (ND); 8, Waulala River; 9, Kawau River; 10, Longito River; 11, Loangkumbe River; 12, Siloi River; 13, Walue River.
SE SUNDALAND ACCRETION of London/Geological Research and Development Centre team (Milsom 1992) and these link to earlier work by members of UC Santa Cruz in SE Sulawesi (Silver et al. 1983). The style and setting of sedimentation on Buton has been described within a framework of rift-drift followed by collision and amalgamation (e.g. Fortuin et al. 1989). The pre-Neogene section records sedimentation on a micro-continental fragment which was rifted away from northern Australia in the Mesozoic. The block drifted ahead of the Australian continent and encountered the SE Sulawesi subduction zone in the early Miocene, undergoing deformation as the zone 'choked' and subduction ceased (Fig. 2). Convergence continued between Sundaland and other Australian fragments in the Banda Sea, accommodated by subduction east of Buton until the mid-Pliocene, when the Tukang Besi platform entered the trench off southern Buton, giving rise to a second deformational event (Fig. 2). Fortuin et al. (1989) suggested that continued convergence following this second event could explain the elevated Quaternary reefs in southern Buton, the pattern of
433
major faults, folds and thrusts across the island, and the 50-60°change in regional strike between its northern and southern halves. Clockwise rotation of the south with respect to the north would inevitably lead to extension seaward of central Buton, and an extensional basin does exist in approximately the expected position. The pattern of Bouguer gravity contours, which 'wrap around' southern Buton and Muna, is also persuasive evidence for differential rotation. In addition to the sedimentary rocks, outcrops of a dismembered ophiolite are scattered through western Buton, their presence being generally attributed to the first of the two collisions (e.g. Fortuin et al. 1989). Gravity surveys have failed, however, to locate any major anomalies associated with these rocks, or any sign of a high density root zone (Milsom 1992). A gravity map of the Buton-Tukang Besi region is shown in Fig. 3. These observations are evidence in favour of the interpretation of the structure and deformation of Buton in terms of a thin-skinned thrust model; such a model has also been proposed on purely geological grounds (H. Manur, pers. comm. 1993).
IwNwt EOCENE - OLIGOCENE
Sea Level ÷~
~
.......... :......................
++ + + + + ~ l ~ t ~ t ~ ~ ~ . --~,..~----,x.~..~...'.~E~~ + + + + + + ~~1~ ~ ~~- ~ ~' ' ~~~ / - - - - ~ +~ / BUTONMICRO- CONTINENT MUNA/ SE .ff'~'// SULAWESI - / CONTINENTALCRUST BUTON KAPANTOREH STRAITS OPHIOLITE BASIN .. ~
BULU BASIN
~ _ . . ~ _ _ _ _ / ~ ~ UPLIFTED • ~ M U N A / S E SULA WESI CONTINENTAL CRUST
~
"
,,
/
T U K A N G BESI MICRO - C O N T I N E N T
FOLD & T H R U S T BELT
--
I
MIOCENE
/
/
ACCRETED B UTON MICRO - C O N T I N E N T
PLIOCENE / PLEISTOCENE
ACTIVE FOLD & THRUST BELT
Fig. 2. Cartoons to illustrate the mid to late Cenozoic development of the SE Sulawesi region, based on Fortuin et al. (1989) and Davidson (1991).
434
J.R. ALI ET AL.
Formations sampled for palaeomagnetism
5"S
123°E
124°E
Fig. 3. Bouguer anomalies, Buton, Muna and the Tukang Besi Platform. Reduction density = 2.67 Mg m-3.
Muna and the Tukang Besi platform Muna, which lies to the west of Buton, from which it is separated by a narrow seaway, is an uplifted carbonate platform of generally low relief. A narrow limestone ridge which runs close to the eastern coastline is interpreted as a former barrier reef. Uplift in the extreme south has been stronger than in the remainder of the island, although less strong than in the adjacent parts of Buton, where flights of coral terraces are impressive features of the landscape. Basement rocks are exposed only in a tiny area of schist outcrop at Tanjung Batu, on the west coast. The Tukang Besi platform, east of southern Buton, is largely submerged but some moderately large islands rise a few metres above sea-level. These seem to be entirely built up of upper Neogene or Quaternary limestone. Gravity data from Wangi Wangi, the largest of the islands, and from marine surveys (Fig. 3), suggest that the crustal thickness in the area approaches that of a normal continent, supporting the view expressed by Hamilton (1979) that the plateau is another fragment of Australian-related continental crust. Seismic reflection data obtained using relatively low power sources on oceanographic cruises have imaged only very thin sedimentary sections and it seems unlikely that there are thick Neogene sediments analogous to those exposed on Buton anywhere on the plateau. The same seismic data suggest a clear structural break between the plateau and Buton itself, the intervening seaway being considerably deeper than the strait between Buton and Muna.
The aim of the project was to use palaeomagnetism to quantify the proposed differential rotation between north and south Buton (Fortuin et al. 1989). As the collision of Tukang Besi Platform with Buton is believed to have started in the mid-Pliocene, sampling was confined to two upper Neogene formations exposed across much of Buton (the stratigraphy of Buton is summarized in Fig. 1). Older formations were not sampled as they were considered to have complex pre-Miocene motion histories: isolating their record of Plio-Pleistocene deformation would be difficult, if not impossible. The upper Miocene Tondo Formation comprises terrigenous clastic rocks ranging from breccias to mudstones. The sediments were eroded following the uplift associated with the Buton-Sundaland collision. By Mio-Pliocene times the region had stabilized, and pelagic limestones and marls of the Sampolakosa Formation were being deposited. The Tondo and Sampolakosa Formations are unconformably overlain by Quaternary reef deposits of the Wapulaka Formation (Sikumbang & Sanyoto 1981). Reefal limestones are usually very weakly magnetized, and because of this the Wapulaka Formation was not sampled for this palaeomagnetic study. In addition to the palaeomagnetic studies, nannofossils were examined from 32 of the palaeomagnetic sites, thus providing information on the maximum age of the magnetic remanence. The nannofossils have been assigned to the Martini (1971) zones which in turn have been linked to the geomagnetic timescale of Cande & Kent (1992) using the magnetochron-biozone correlations of Berggren et al. (1985). Unfortunately some of the assemblages from this study lack the nominate taxon for the Martini scheme, however the total recorded assemblages are sufficient to enable assignment to a particular zone(s).
Palaeomagnetism Methods
The majority of specimens were obtained using a gasoline-powered rock-drill which was used to cut 2.5 cm diameter mini-cores. The cores were orientated to _+ 2 ° using a magnetic compassinclinometer. A palaeomagnetic site comprises 6-9 orientated mini-cores spanning 10-30cm of section within a bed. A number of orientated blocks (c. 300 cm 3) were also retrieved. Wherever possible, sites were collected with the aim of utilizing the fold and reversal tests to constrain the age of magnetization. The stability of magnetization of each specimen was assessed after using either stepwise alternating
SE SUNDALAND ACCRETION field (AF) demagnetization or continuous thermal demagnetization to isolate the various components of magnetization held within the rocks. The AF-processed specimens were analysed at Southampton, using either a 'Molspin' spinner magnetometer in tandem with a 'Molspin' demagnetizer, or a '2G Enterprises' cryogenic magnetometer which has an in-line 3-axis demagnetizing unit. Continuous thermal demagnetizations were performed on a number of specimens in the laboratory of M. Fuller at Santa Barbara, using a '2G Enterprises' cryogenic magnetometer with an in-line thermal demagnetizing unit (Dunn & Fuller 1984). In practically all cases the remanence of the Tondo and Sampolakosa Formations was found to be dominated by a single component suggesting a simple magnetization history. Characteristic components of magnetisation were identified from vector end-point plots (Zijderveld 1967), and calculated using a Kirschvink (1980) based software package. Site mean directions (Table 1) have been calculated using the statistics of Fisher (1953). The statistics of McFadden & Reid (1982) have also been used to calculate mean inclinations; in a number of cases the sites are from isolated outcrops and averaging site directions using Fisher (1953) statistics may have been incorrect. It is worth noting, however, that in all cases the difference between the Fisher (1953) and McFadden & Reid (1982) values is less than 0.5 ° (see Table 1).
Isothermal remanent magnetization Isothermal remanent magnetization (IRM) analysis was carried out on representative specimens from sites yielding reliable data to provide information on their principal remanence carriers. The IRM was generated using a 'Molspin' pulse magnetiser (up to 0.86 T) and was measured between steps using a 'Molspin' magnetometer. The shape of the IRM curve (as well as the peak IRM value) was used to evaluate the characteristic remanence carrier(s). The IRM ratio, defined by Ali (1989), is the ratio of the IRM at 0.3 T to the IRM at 0.86T. As a general guide, values above 0.9 indicate that magnetite is the dominant remanence carrier, whereas values below 0.9 indicate that the remanence may be due to other minerals. All of the specimens have IRM ratios greater than 0.95 (Table 1) suggesting that the remanence of the Tondo Formation and the basal Sampolokosa Formation is carried by magnetite.
Palaeomagnetic results: south Buton Bungi River, Bungi Four drill sites and one hand specimen were sampled from the lower part of the Sampolakosa
435
Formation at five outcrops exposed along the Bungi River, about 5 km ESE of Bungi (Fig. 1; 1). The sites were from beds which dip to the north and west at 10-15 ° . The formation is dominated by compacted marls in beds typically 30-80 cm thick. Samples from three sites were analysed for nannofossils, but the faunas were too poorly preserved to yield a reliable age; a Mio-Pliocene age is assumed. Specimens from Sites BU1, 3, 4 and 5 have NRM intensities between 0.07 and 1.4 mA m -1 (Table 1). AF demagnetization of these sites indicates that the remanence is carried by a single high-coercivity component. Site BU2.H, from the Tondo-Sampolakosa transition, revealed quite different magnetic behaviour; the remanence being carried by a relatively soft single component of magnetization. Specimens from Sites BU 1, 2 and 5 were the subject of IRM studies (Table 1). All three specimens yielded IRM ratios >0.95 suggesting that the remanence is probably carried by magnetite. The five sites had a mean in situ direction of D = 357.9 °, I = -23.5 °, a95 = 7.7 °, K = 98.4 (Table 1, Fig. 4a), and a tilt corrected direction of D = 2.7 °, I = -31.3 ° a95 = 9.6 °, K = 63.8 (Fig. 4b). A normal polarity magnetization is assumed. The remanence data can be interpreted in two ways. Assuming the individual site corrections account for all of the deformation experienced by each site, then the in situ and tilt corrected a95 and K values suggest that the remanence post-dates tilting. However, the present-day geocentric axial dipole inclination for the Bungi River area is -10.7 °, which suggests that the remanence predates deformation. The age of magnetization of the Sampolakosa Formation in the Bungi River is considered ambiguous. The declination data, whether interpreted as either pre- or post-folding, indicate negligible rotation.
Kemberu River, Masiri Numerous outcrops of the Tondo Formation were exposed in the dry bed of the Kemberu River, near Masiri (Fig. 1; 3). The formation is dominated by massively bedded conglomerates and immature sandstones; lithologies not particularly suited for palaeomagnetic investigations. The beds are folded about axes aligned NNE-SSW. Dips are typically 20-40 ° , although values of 60-70 ° were locally recorded. Eight drill and two hand sites were sampled from 10-20 cm thick mudstones and siltstones. Three sites yielded data with a mean in situ direction of D = 355.6, I = -14.2 °, a95 = 18.4 °, K = 4 5 . 8 , and a tilt corrected direction of D = 3 5 9 . 6 ° , I = - 1 4 . 8 ° a95=15.3 ° , K = 6 5 . 4 (Table 1, Fig. 4c/d). This direction is interpreted as a primary normal polarity magnetization with negligible net rotation.
a b c d e
a b c
a b c d
a a b c d d e e e f f
KEMBERU (5.67°S/122.67°E)
WARUMBIA (5.47°S/122.83°E)
WABIAU (5.35°S/123.03°E)
Exp
BUNGI (5.42°S/122.70°E)
TRAVERSE
Fm
BU30 BU31 BU32 BU33 BU34 BU35 BU36 BU37 BU38 BU39 BU40 IS (F) IS (M&R)
BU21.H BU24 BU27.H BU29 IS (F) IS (M&R) tilt (F) tilt (M&R)
T T T T T T T T T S S
T T T S
BUll T BU12 T BUI3 T IS (F) IS (M&R) tilt (F) tilt (M&R)
BU1 S BU2.H T-S BU3 S BU4 S BU5 S IS (F) IS (M&R) tilt (F) tilt (M&R)
SITE 6 4 7 6 6
9.5-8.2
8.2-3.6 8.2-3.6 9.5-8.2
9.5-5.3
NNll-15 NNll-15
NN10
NN10-11
? c. 5
indet.
NNI0
? c. 5
6 6 6 5 6 5 6 6 6 4 6
1 6 1 6
? 9.5-5.3 6 ? 9.5-5.3 6 reworked 5
? c. 5 ? c. 5
age (Ma) Np
indet.
indet. indet. NPI7-21
indet. indet.
Nanno.
6 5 5 5 6 5 5 6 6 4 6 11 11
1 6 1 6 4 4 4 4
6 5 5 3 3 3 3
5 4 7 6 6 5 5 5 5
Nc
1.5-5.0 5-10 0.2--0.5 0.1-0.25 3-6 2.5-4.8 1.3-7.0 5-12 c. 2.4 c. 0.3 0.1-0.3
29,45 2.0--4.0 91.78 30-100
10-50 22-53 0.4-1.6
0.07-0.3 c. 2.0 0.5-1.0 0.14-0.25 0.8-1.4
NRM (mA m -1)
8.4 0.7 7.7 345.5 345.2 346.5 355.7 358.1 6.0 357.3 353.8 356.8 NA
58.4 78,4 21.6 16.0 41.7 NA
1.5 342.0 3.0 355.6 NA
358.8 357.1 4.3 354.1 356.3 357.9 NA
-4.6 -15.7 -10.8 -18.7 -10.7 -11.6 -13.6 -8.6 -5.6 -17.1 -13.0 -11.9 -11.8
-27.7 -56.2 -34.8 -45.1 -43.6 -42.7
-16.3 -16.4 -9.6 -14.2 -14.1
-34.3 -14.7 -26.2 -23.0 -19.4 -23.5 -23.8
in situ DEC INC
Table 1. Summary of palaeomagnetic data at sites named on Fig. 1 with latitude and longitudes of each
075/15 SE 075/15 SE 060/12 SE 120/14 SW 065/21 SE 060/16 SE 210/10 NW 210/10 NW 200/15 W 140/08 SW 130/08 SW
062/35 S 090/27 S 080/18 S 120/12 SW
200/08 W 160/24 W 190/18 W
290/10 N 237/18 NW 210/18 NW 165/14 W 216/18 NW
Strike & dip
9.2 -1.2 -1.2 -8.5 10.0 3.8 -19.1 -13.8 -9.0 -21.8 -18.4
-29.8 -30.1
31.0 NA 8.6 0.1 6.9 347.8 345.2 346.2 358.2 359.8 7.9 355.0 351.4
-20.4 -43.6 -19.1 -33.4
-14.8 -14.7
359.6 NA 42.6 48.8 17.1 18.2
-18.7 -14.1 -11.3
-31.3 -31.7
2.7 NA 3.9 348.6 6.3
-43.5 -29.5 -32.7 -20.1 -30.1
355.7 0.8 13.9 359.5 3.0
tilt corr. DEC INC
6.2 5.9 6.0 12.4 7.1 6.4 17.7 8.1 8.1 11.5 6.8 5.4 3.0
NA 9.0 NA 6.4 28.7 20.0 21.0 18.6
12.9 9.3 14.1 18.4 9.7 15.3 9.3
7.4 6.2 5.3 9.5 5.1 7.7 9.2 9.6 10.4
a95
116.6 167.2 165.7 38.8 90.1 142.2 19.7 68.7 69.5 64.9 99.5 72.9 165.3
NA 56.9 NA 110.7 11.1 21.0 19.9 24.4
28.0 68.9 30.6 45.8 216.1 65.4 235.2
108.1 219.4 128.9 51.1 172.2 98.4 60.2 63.8 46.8
K
N N N N N N N N N N N
N N N N
N N N
N N N N N
Pol
0.99
0.99
1.00
0.98
0.96
0.99 0.99
0.99 0.96 0.95
IRM ratio
47.1
2247
377.6
8115
1521
21393 17067
28.9 348.1 57.2
peak IRM
c7~
tao
BU56 BU58 BU59 IS (F) tilt (F)
a a a
a a b c c c d
a
a
LONGITO (4.97°S/122.92°E)
LOANGKUMBE (4.70°S/123.08°E)
SILOI (4.72°S/123.00°E)
WALUE (4.53°S/122.92°E)
T
T
T T T T T T T
T T T
T T T T T T T
T
indet.
? 9.5-5.3
? 9.5-5.3
8.2-3.6
NNll-15
indet.
? 8.2-3.6
8.2-3.6
? 9.5-5.3
indet.
NNll-15
indet.
6
5
6 6 6 5 1 1 7
4 4 3
5 5 5 6 6 6 6
1
6
4
6 4 6 3 1 1 5 7 7
3 4 3
4 4 5 6 5 6 6 7 7
1
11 11
6-17
13-46
4-12 0.1-0.35 10-17 c. 7 37.68 54.23 8-59
0.1-0.2 0.1-0.2 0.1-0.3
0.2-0.35 0.16-0.5 0.4-0.9 0.9-3.6 0.5-1.3 1.5-2.5 0.4-0.9
2.93
7.0
355.1
147.1 334.1 226.9 352.4 39.5 32.3 193.6 NA
176.1 348.9 14.7 359.8
354.8 353.1 190.5 352.4 4.6 157.6 354.3 355.0
214.9
-33.7
0.9
14.9 -10.9 13.9 -38.4 -31.0 -15.9 -4.3 18.1
19.2 -12.7 -14.4 -15.7
-23.9 -18.9 33.4 -16.9 -22.5 21.1 19.6 -17.3
-4.0
SW SW SW SW SW SW SW
350/80 E
280/18 N
140/32 140/28 145/32 125/22 126/26 126/26 140/30
150/12 SW 170/34 W 030/31 SE
083/15 S 070/12 S 080/16 SE 037/37SE 130/10 SW 110/12 S 085/20 S
235/25 NW
317.6
-19.6
-16.5
9.0
NA 354.5
8.9 3.1 -17.8 -21.1 -5.0 4.1 -27.8
-9.8
1.7 153.9 337.5 226.7 0.4 39.0 32.3 188.0
13.6 -11.1 -4.7
-7.4
355.1 179.3 356.3 9.6
-8.5 -7.2 18.3 9.9 -14.2 12.1 -0.4
4.7
-6.4 ~6.6
354.7 352.5 188.0 350.8 6.5 160.0 354.3
215.1
357.0 NA
12.7
18.1
8.3 17.5 8.4 23.1 NA NA 10.3 13.0 12.9
26.5 22.0 27.8 20.4 12.7
6.9 17.3 8.7 4.2 10.1 4.6 7.6 14.6 10.0
NA
8.1 7.6
28.6
26.6
66.5 28.8 65.3 29.4 NA NA 56.0 17.1 17.2
22.6 18.3 20.7 37.3 95.2
178.4 29.0 79.0 259.5 57.8 213.2 77.9 17.8 37.2
NA
32.8 25.8
N
N
R N R R N N R M M
R N N M M
N N R N N R N M M
R
0.99
1.00
1.00
1.00
1.00
0.99
0.99
1.00
12216
56538
17756
17415
9481
106.7
154.1
1268
Abbreviations after site number are: F = Fisher (1953) statistics used to calculate mean direction; M & R = McFadden & Reid (1982) statistics used to calculate mean inclination. Other abbreviations are Fm = Formation, T = Tondo, S = Sampolakosa; Nanno. = nannofossil zones; Np = Specimens processed; Nc = Specimens used to calculate the site direction; NRM = initial intensity; in situ = site mean direction before application o f tectonic correction: DEC = declination in degrees; 1NC = inclination in degrees; NA = not applicable; Strike & Dip = bedding orientation; tilt corr. = site mean direction after tectonic correction; a95 = circle of 95% confidence about the site mean; K = precision parameter; Pol = polarity: N, normal, R, reversed; M = statistics calculated by inverting reverse polarity directions;IRM ratio = IRM at 0.3 T/IRM at 0.86 T; peak IRM expressed in m A m 2.
BU72
BU71
BU62 BU63 BU66 BU68 BU69H.a BU69H.c BU70 IS (M&R) tilt (M&R)
BU49 BU50 BU51 BU52 BU53 BU54 BU55 IS (F) tilt (F)
a a a a a a a
KAWAU (4.95°S/122.97°E)
BU46.H
a
WAULALA (5.23°S/122.95°E)
tilt (F) tilt (M&R)
4~
438
J.R. ALI E T A L . N
N
N
N
N
E c ~ d ............... i......................... t ..............
b
g
N
N
:
i
E
N
N
!Bu:4 ........e t
N
N
h
N
®® ~
i
®
- j j N
t~uov.l-la [
i BU68
........................i.......................................... i................... BU66 ~BU70
:: i
I BU62[ @ ~ BU66
i i @
BU62 ~
Fig. 4. Site mean data from (a, b) Sampolakosa Fm, Bungi River (a = in situ, IS, b = tilt corrected, TC); (c, d) Tondo Formation, Kemberu River (c = IS, d = TC); (e, f) Tondo Fm and basal Sampolakosa Fm, Warumbia River (e = IS, f = TC); (g, h) Tondo Fm and basal Sampolakosa Fm, Wabiau River (g = IS, h = TC); (i, j) Tondo Fm, Kawau River (i = IS, j = TC); (k, !) Tondo Fm, Longito River (k = IS, 1 = TC); (m, n) Tondo Fm, Loangkumbe River (m = IS, n = TC). Open and closed circles represent downward and upward dipping vectors respectively. Site directions based on mini-core samples are shown with their 95% confidence circle. Hand specimens do not have associated confidence circles.
Warumbia River, Lapodi Three hand specimens and four drill sites were collected from the Warumbia River, about 2 km north o f Lapodi (Fig. 1; 5). Drill site B U 2 9 was collected f r o m a calcilutite at the base o f the S a m p o l a k o s a Formation. The remaining sites were
from mudstones and fine sandstone in the upper part o f the Tondo Formation. Reliable data were obtained from two hand specimens, B U 2 1 . H and B U 2 7 . H , and two drill core sites, B U 2 4 and B U 2 9 . The mean in situ direction is D = 41.7 °, I = - 4 3 . 6 °, a95 = 28.7 °, K = 11.1 (Fig. 4e). F o l l o w i n g the application o f the tilt corrections the mean direction
SE SUNDALAND ACCRETION
is D = 3 1 . 0 °, I = - 2 9 . 8 ° a95=21.0°, K = 1 9 . 9 (Table 1, Fig. 4f). Although a95 and K improve with the correction for dip, the scatter of the vectors suggests that the local tectonics are more complex than simple S - S W tilting.
Wabiau River, Walumpo East Buton is close to the supposed Tukang Besi Platform-Buton Block suture. The Wabiau River (Fig. 1; 6) is adjacent to the Tondo Formation type area and contains splendid exposures of both the Tondo Formation and the basal part of the Sampolakosa Formation. The rocks are mildly deformed only with dips generally <15 ° . Nine Tondo Formation and two Sampolakosa Formation drill sites were sampled at six outcrops. The eleven sites yielded data with a mean in situ direction of D = 356.8 °, I = -11.9 °, a95 = 5.4 °, K = 72.9, and a tilt corrected direction of D = 357.0 °, I = - 6 . 4 °, a95 = 8.1 °, K = 32.8 (Table 1, Fig. 4g/h). Although the in situ directions are slightly more clustered than the tilt corrected vectors, suggesting that the remanence post-dates deformation, the magnetization of the Wabiau River Tondo-Sampolakosa Formation sequence is considered to be primary. Demagnetization isolated essentially single component low-coercivity remanences (also see Table 1 ; IRM data). Field observations suggest that the rocks have undergone nothing more dramatic than burial and slight structural upheaval. It is possible that the assumption of a tilt corrected mean direction based on simply restoring the beds to horizontal is incorrect, and that faults between each of the outcrops may have accommodated local small-scale deformation. The tilt corrected mean direction suggests negligible rotation of the site since the start of the Pliocene.
Waulala River, Lawele Superb exposures of the Tondo and Sampolakosa Formations are present in the Waulala and Loko Rivers, about 5 km south of Lawele (Fig. 1; 8). The geology is dominated by a number of major N E - S W trending faults limiting the area to which palaeomagnetic declination data obtained might be applied. Four drill sites and one hand specimen were sampled from the Tondo Formation, together with one drill site from the Sampolakosa Formation. Only site BU46.H, the hand specimen from the Tondo Formation, yielded reliable data with a reverse polarity tilt corrected direction of D = 215.1 °, I =4.7 ° (Table 1), suggesting clockwise rotation in excess of 30 ° . However, as the site is adjacent to a major fault system, it is suggested that this deflection reflects local deformation.
439
Palaeomagnetic results: central Buton Kawau River, Bubu The Tondo Formation is exposed over a kilometre of continuous outcrop in the Kawau River, 5 km inland of Baubau (Fig. 1; 9). Seven drill sites were sampled from beds that dip towards the east and south at between 10 ° and 30 °. Sites BU49, 50, 52, 53 and 55 record normal polarity magnetizations; BU51 and 54 record reverse polarity magnetizations (Table 1). The in situ mean direction is D = 355.0 °, I = -17.3 °, a95 = 14.6 °, K = 17.8 (Table 1, Fig. 4i). Application of the tilt corrections produces a direction of D = 355.1 °, I = - 7 . 4 °, a95 = 10.0 °, K = 37.2 (Fig. 4j), which suggests that the remanence pre-dates folding. The presence of both normal and reverse polarity sites in a single outcrop is highly suggestive of a primary magnetization. The mean direction indicates negligible rotation of the site since formation.
Longito River, Bubu The Kamunte River, a tributary of the Longito River, was accessed from the road linking Bubu to Pure (Fig. 1; 10). Five sites were sampled from a 600 m long exposure of the Tondo Formation in the dry river bed. The rocks, mainly mudstones and siltstones in beds 10-30 cm thick, are deformed into an open synform. The Tondo Formation samples are weakly magnetized with initial NRM intensities between 0.06 and 0.20 mA m -1. Reliable data were obtained for three sites. Scatter for specimens from these site is large with values for a95 and K typically 25 ° and 20 respectively. However, the outcrop contains both normal (BU58 and BU59) and reverse (BU56) polarity sites, which suggests that the remanence is primary. The in situ mean direction is D = 3 5 9 . 8 °, I = -15.7 °, a95 = 20.4 °, K = 37.3 (Table 1, Fig. 4k). Application of tilt corrections produces a mean direction of D = 1.7 °, I = -9.8 °, a95 = 12.7 °, K = 9 5 . 2 (Fig. 41), which suggests that the magnetization pre-dates folding, with no rotation.
Palaeomagnetic results: north Buton Loangkumbe River, Loangkumbe The Tondo and Sampolakosa Formations are well exposed along the Loangkumbe River (Fig. 1; 11). This is one of the most important areas of Buton for the collection of palaeomagnetic samples, potentially providing the Mio-Pliocene reference direction for the northern part of the island. Unfortunately the stratigraphic interval from which palaeomagnetic sites can be sampled is limited because the Loangkumbe River flows parallel to
440
J.R. ALI ET AL.
the general strike of rocks in the area. The lower reaches of the fiver expose marls from the lower part of the Sampolakosa Formation but only a single site was sampled; determining accurate dips and strikes in the formation is difficult because obvious signs of bedding have been obliterated by erosion. Eight drill sites and one hand specimen were sampled from the Tondo Formation. Seven Tondo Formation sites, from three outcrops, have yielded data (Table 1, Fig. 4rn/n). Sites BU62 and BU63 from an outcrop c. 200 m upstream of the L o a n g k u m b e - M o u s o Cabang yielded tilt corrected directions of D = 153.9 °, I = 8.9 °, a95 = 8.3 °, K = 66.5, N = 6 and D = 337.5 ° I = 3.1 ° a95 = 17.5 °, K = 17.5, N = 4. The presence of reverse and normal polarity sites in one outcrop suggests that the remanence is primary. The declinations indicate about 25 ° counter-clockwise rotation. Five sites, between 3 and 5 km upstream of Sites BU62 and 63, yielded directional data. BU66 has a tilt corrected mean direction of D = 226.7 °, I = -17.8 °, a95 = 8.4 °, K = 65.3, N = 6, indicating about 45 ° of clockwise rotation. Site BU70, about 600 m upstream, also carries reverse polarity magnetization, with D = 188.0 °, I = - 2 7 . 8 °, a95 = 10.3 °, K = 56.0, N = 5, indicating negligible rotation. Site BU68 is normally magnetized with no declination offset. From the same outcrop, hand samples BU69H.a and BU69H.c record c. 35 ° of clockwise rotation. Declination data from the Loangkumbe River suggest that rotations for this part of north Buton are significant, but appear to be local (subkilometre scale). Using the statistics of McFadden & Reid (1982) the mean tilt corrected inclination is +9.0 °, assuming a normal polarity, suggesting deposition at a latitude c. 4.5°N. However, because of the large a95 confidence circle (12.9 °) associated with this inclination, it can only be stated with confidence that this part of Buton was close to the equator during Mio-Pliocene times.
Siloi River, Ronta The Siloi River, accessed from the road which links Ronta to Maligano, exposes the Tondo Formation (Fig. 1; 12). The geology of the area is complicated by a series of major faults juxtaposing these upper Miocene rocks with a number of pre-Neogene formations. During fieldwork, exposures of the Tondo Formation were seen only in the river bed, seriously hampering the palaeomagnetic sampling. However, a site in the Tondo Formation was sampled 4 km downstream from the road bridge over the Siloi River. The site (BU71) is in a medium-grained sandstone adjacent to a N E - S W striking fault that juxtaposes the formation with the
Ogena Formation (lower Jurassic; Davidson 1991). The in situ site mean direction is D = 3 5 5 . 1 °, 1 = 0 . 9 °, a95 = 18.1 °, K = 26.6, N = 4. (Table 1). Application of the tilt correction produces a direction of D = 354.5 °, I = -16.5 °. The remanence is carried as a normal polarity, but the age of the remanence (depositional/post deformation) is not known.
Walue River, Kobakoba Exposures of the Tondo Formation in the Walue River are scarce and just one drill site, BU72, was sampled (Fig. 1; 13). The minicores were obtained from adjacent sandstone and siltstone beds which dip 80 ° to the east. The specimens from both the sandstone and siltstone had similar initial NRM intensities of about 10 mA m -1. The site has an in situ mean direction of D = 7.0 °, I = - 3 3 . 7 °, a95 = 12.7 °, K = 28.6, N = 6 (Table 1) and a tilt corrected direction of D = 317.6 °, I = -19.6 °. It is assumed that the remanence is primary because the in situ direction is much steeper than the predicted geocentric axial dipole inclination for the Walue River area (-8.9°). The steep dip of these rocks, and hence the amount of deformation that site BU72 has undergone, must limit the areal extent for which the declination offset can confidently be used to unravel rotations relative to north. However, the tilt corrected inclination angle can be used to help define the latitude of formation ( 1 0 _ 7.5 ° S, assuming that the remanence was acquired during a normal polarity geomagnetic field).
Summary of the palaeomagnetic data Latitudinal information from all of the palaeomagnetic sites yielding interpretable data is summarized in Fig. 5. They suggest negligible (c. 1o N) northward motion of Buton and, by implication, of SE Sulawesi since the late Miocene. The palaeomagnetic declination data from both the upper Miocene Tondo Formation and the Mio-Pliocene Sampolokosa Formation on Buton are summarised in Fig. 6. Sites from south Buton record locally consistent declination offsets, but at sampling localities less than 25 km apart the deflections vary between 0 ° and 35 ° clockwise. There appears to be no simple pattern to these data with respect to the regional tectonic setting. The age of magnetization of the Sampolakosa Formation sites from the Bungi River is uncertain. The five sites show marginally tighter in situ (rather than tilt corrected) clustering. However the in situ inclination o f - 2 3 . 8 ° is too steep to be the result of a recent weathering induced chemical remanence, where an inclination of around -10.7 ° would be predicted. Regardless of the age of the remanence,
SE SUNDALAND ACCRETION 10 8-
MEAN NORTHWARD DRIFT = + O ~
Z~4-© 2-
0
L
I
-18
'
I
'
I
I
'
I
'
I
'
B I
'
-12 -6 0 6 12 18 Apparent Northward Drift (c)
Fig. 5. Apparent formation latitude shift for the Tondo and Sampolakosa Formation sites, Buton.
441
the Bungi River area has experienced negligible rotation since the magnetization was acquired. In the Kemberu River, 25 km to the south of Bungi, the Tondo Formation sites also record negligible rotation. In the Warumbia River, 25 km to the east of Bungi, four upper Tondo Formation sites record clockwise rotations of c. 30 °. East Buton was the one part of the island where large rotations might have been predicted because it is closest to the Tukang Besi Platform-Buton impact point. However, the large number of sites from the Wabiau River revealed negligible declination offsets. The single reliable Tondo Formation site in the Waulala River has a large declination offset (35 ° clockwise). Large rotations might be expected here as this site is from a block located in the middle of several NE-SW striking faults. In central Buton, declinations for the Tondo Formation sites in both the Kawau and Longito Rivers indicate negligible rotation. The two sections include both normal and reverse polarity sites suggesting that the remanence is primary. Also in both sections there is a clear clustering of the site vectors following application of the tilt corrections to the in situ data. In north Buton, declination data from the Loangkumbe River suggest local (< km) largescale, clockwise and counter-clockwise rotations. A single site from the Tondo Formation in the Siloi River records negligible rotation, but the age of the magnetization is uncertain. One Tondo Formation site from the Walue River, records a counterclockwise rotation of over 40 ° . However, the site was sampled from steeply dipping beds, and the declination deflection probably reflects localized deformation.
Discussion
Fig. 6. Summary of the palaeomagnetic declination data from the Tondo and Sampolakosa Formations, Buton. The orientations of the arrows illustrate the declination offsets. Note each arrow has an error typically 10-13° (see Table 1).
The Buton region contains a record of the relatively recent successive accretion of two microcontinents to the edge of a continent, and a quantitative understanding of the processes involved provides a valuable analogue for older systems where the rock record is less complete. Prior to the present study no palaeomagnetic information was available for the area and the initial aim of the research was to test the model of Fortuin et aL (1989) which proposed large-scale (c. 60 °) Pliocene clockwise rotation of south Buton relative to central and north Buton as a result of the Tukang Besi Platform collision. Although it remains possible that the underlying basement experienced these large relative motions, the study has shown that the upper Cenozoic cover has not been deformed in such a simple way (Fig. 6). In recent publications (e.g. Davidson 1991) the geology of Buton has been interpreted in terms of
442
J.R. ALI ET AL.
thin-skinned deformation processes but no insights have been provided into the rotations which the thin sheet overthrusts might have experienced. Strictly speaking, the palaeomagnetic study has no bearing on the question of thin-skin versus thick-skin tectonics but it is perhaps easier to envisage large relative rotations of small blocks taking place in a thin-skin rather than a thick-skin context. Resolving the fine scale detail of the tectonics of Buton is beyond the scope of this or indeed any project, because of the lack of exposure, but it has been demonstrated by the palaeomagnetic results that in a collision setting thin overthrusts may well be rotated by 30-60 ° within 2-3 Ma. The palaeomagnetic study has also provided support for the two-stage model for collision, which is still controversial; if the Tukang Besi Platform were part of the Buton block prior to its collision with SE Sulawesi in the Miocene, there would be little reason for the large Pliocene to Recent rotations which are actually observed. However, the amount of thin-skinned rotation does not seem to have been a simple function of proximity to the microcontinental suture. Rotations in north Buton, which is further from Tukang Besi, are greater than rotations at sites, such as the Wabiau River, which are closer to it. The difference in deformation style between north and south is also the reverse of expectation. In the south, each individual traverse section, up to 15 km in length, shows a consistent declination deflection, suggesting the presence of coherent sheets at least 15 km across, whereas in the north rotations appear to have occurred on a much more local (c. 1 km) scale. Finally, the results presented here emphasize the importance of comprehensive sampling in any
palaeomagnetic programme in SE Asia. If just the Waulala and Warumbia sites had been sampled in southern Buton, the Fortuin et al. (1989) rotation hypothesis would probably have been regarded as confirmed, although with rotations somewhat smaller than those originally suggested. If, on the other hand, these sites had not been visited and work had been confined to the Bungi, Kemberu and Wabiau Rivers, it would have been concluded that rotations were absent in southern Buton. Neither conclusion is supported by the full dataset actually obtained.
Conclusions This palaeomagnetic study of the SE Sulawesi/ Buton/Tukang Besi accretion complex has provided a quantitative insight into the process of crustal deformation at a continent-microcontinent accretion site. In the Buton region, a significant component of the deformation has taken place as thin-skin overthrusts. The palaeomagnetic dataset suggests that the rotation of the thin-skin sheets is complex; there is no obvious relationship between the rotation of a sheet and its proximity to the microcontinent impact point. Rotations of up to 30-60 ° may be generated. Rotation rates are considerable; the Tukang Besi Platform-Buton collision began in the Pliocene, 2-3 Ma. This study was funded by the University of London Geological Research in SE Asia Consortium. Special thanks are given to Tony Barber, John Davidson, Kate Davis, Michael de Smet, Mike Fuller, Ernie Hailwood, Hendry Manur and Graham Rose for their input to this project. The constructive comments of H. Wensink, A. R. Fortuin and an anonymous referee greatly improved this manuscript.
References ALI, J. R. 1989. Magnetostratigraphy of early Palaeogene sediments from NW Europe. PhD thesis, University of Southampton. BERGGREN, W. A., KENT, D. V., FLYNN, J. J. & VAN COUVERING, J. A. 1985. Cenozoic geochronology. Bulletin of the Geological Society of America, 96, 1407-1418. CANDE,S. • KENT,D. V. 1992. A new geomagnetic time scale for the Late Cretaceous and Tertiary. Journal of Geophysical Research, 97, 13917-13951. DAVIDSON,J. W. 1991. The geology and prospectivity of Buton Island, SE Sulawesi, Indonesia. Proceedings of the Indonesian Petroleum Association 20th Annual Convention, 209-231. DUNN, J. R. & FULLER,M. 1984. Thermal demagnetization with measurements at high temperature using a SQUID magnetometer. LOS Transactions AGU, 65, 863. FISHER, R. A. 1953. Dispersion on a sphere. Proceedings of the Royal Society of London, Series A, 217, 295-305.
FORTUIN, A. R., DE SMET, M. E. M., HADIWASASTRA,S., VAN MARLE, L. J., TROELSTRA, S. R. & TJOKROSAPOETRO, S. 1989. Late Cenozoic sedimentary and tectonic history of south Buton. Journal of SE Asian Earth Sciences, 4, 107-124. HAMILTON,W. 1979. Tectonics of the Indonesian region. US Geol. Survey Professional Paper, 1078. KIRSCHVlNK,J. L. 1980. The least squares line and plane analysis of paleomagnetic data. Geophysical Journal of the Royal Astronomical Society, 62, 699-718. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: FARINACCI,A. (ed.) Proceedings of the H Planktonic Conference, Roma 1970. Edizioni Tecnoscienza, Rome, 739-85 MCFADDEN,P. L. & REID, A. B. 1982. Analysis of palaeomagnetic inclination data. Geophysical Journal of the Royal Astronomical Society, 69, 307-319. MtLSOM, J. 1992. Structure and collision history of the Buton continental fragment, eastern Indonesia (abstract). AAPG Bulletin, 76, 1117.
SE SUNDALAND ACCRETION
SIKUMBANG,N. & SANYOTO,P. 1981. Geologic map of the Buton and Muna Quadrangle, southeast Sulawesi, Scale 1:250,000. Geological Research and Development Centre, Bandung. SILVER, E. A., MCCAFFRE¥, R. & SMITH, R. B. 1983. Collision, rotation and the initiation of subduction in the evolution of Sulawesi, Indonesia. Journal of Geophysical Research, 88, 9407-9418. SMITH, R. B. 1983. Sedimentology and tectonics of a Miocene collision complex, and overlying late orogenic clastic strata, Buton Island eastern Indonesia. PhD thesis University of California Santa Cruz, USA. -& SILVER, E. A. 1991. Geology of a Miocene
443
collision complex, Buton, eastern Indonesia. Bulletin of the Geological Society of America, 103, 660-678. WIRYOSUJONO, W. & HAINIM, J. A. 1978. Cainozoic sedimentation in Buton Island. In: WIRYOSUJONO, W. & SUDRAJAT, A. (eds) Proceedings Regional Conference on the Geology and Mineral Resources of Southeast Asia, Jakarta, August 1975. Indonesian Association of Geologists, Jakarta, Indonesia, 109-119. ZUDERVELD,J. D. A. 1967. AC demagnetization of rocks: analysis of results. In: COLLINSON,D. W., CREER,K. M. & RUNCORN, S. K. (eds). Methods in Palaeomagnetism. Elsevier, New York, 254-286.
Pb and Nd isotope constraints on the provenance of tectonically dispersed continental fragments in east Indonesia P Z. V R O O N 1,2, M. J. V A N B E R G E N 1 & E. J. F O R D E 2
1 Faculty of Earth Sciences, University of Utrecht, PO Box 80.021, 3508 TA Utrecht, The Netherlands. 2 SE Asia Research Group, Department of Geology, Royal HoIloway University of London, Egham Hill, Egham, Surrey TW20 OEX, UK. Abstract: Microcontinental fragments, tectonically dispersed at the triple junction of the
converging SE Asian, Indo-Australian and Pacific plates, are a widespread phenomenon in the east Indonesian region. This paper investigates the potential of Pb-Nd isotope characteristics of igneous and (meta) sedimentary rocks from these fragments as indicators of their provenance. From the correspondence of isotopic signatures, it appears possible to identify provenance areas of four selected microcontinents, which were emplaced in different settings and regions of east Indonesia since late Mesozoic times. According to their isotopic affinities, the blocks of Ambon-Seram, Bacan, the Banda Ridges and Sumba originally formed part of southern New Guinea, north Australia, 'Pacific' New Guinea and Sundaland respectively. These conclusions are in line with geological constraints on the palaeopositions of the fragments. The results thus indicate that isotopic signatures are potentially a powerful tool in reconstructions of the tectonic history of east Indonesia.
Relatively small continental and oceanic fragments form a conspicuous feature in the late Cenozoic tectonic evolution of east Indonesia. They represent tectonically dispersed elements which are thought to be composed of: (1) Australian-New-Guinean continental crust; (2) SE Asian arcs; (3) Pacific oceanic crust; (4) Indian oceanic crust (Hamilton 1979; Bowin et al. 1980; Silver et al. 1985; Nishimura & Suparka 1986; Audley-Charles & Harris 1990; Hartono 1990). Tectonic transport of these fragments sometimes occurred over large distances (e.g. Hamilton 1979; Silver et al. 1985). Hence, their origin is difficult to assess and has been widely debated. Many of the microcontinental fragments identified so far (Fig. 1) have been considered to be related to the Mesozoic break-up of Gondwana which created the northern margins of the Australian and New Guinean continent (e.g. Pigram & Panggabean 1984). Subsequent convergent movements of the Indian-Australian and Pacific plates towards SE Asia led to the involvement of these margins in collisions, and to the accretion of the allochthonous continental fragments in recent times (e.g. Audley-Charles & Harris 1990). As an alternative mechanism, detachment of continental blocks may have occurred by strike-slip faulting induced by plate convergence. Based on this model, it has been proposed that many continental fragments (e.g. the Banda Ridges, Banggai-Sula, Obi, Bacan, eastern Sulawesi,
Fig. 1. Distribution of continental basement in eastern Indonesia (after Hutchison 1989). AustraliaNew Guinea form a large continental plate in which the northern edge is currently subducted beneath the Banda arc. In the area northwest of the Australian plate lie numerous continental fragments. The larger are: (1) western arm of Sulawesi; (2) Banggai-Sula spur; (3) Buton island; (4) Sumba; (5) Bacan; (6) Ambon-W. Seram islands; (7) Banda Ridges.
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, Geological Society Special Publication No. 106, pp. 445--453.
445
446
P.Z. VROON E T AL.
Buton, Buru, Seram) originated from northern New Guinea and moved westward to their present position along left-lateral strike slip faults such as the Sorong fault system (see Visser & Hermes 1962; Hamilton 1979; Silver et al. 1985; Hall et al. 1991). Tectonic reconstructions (Daly et al. 1987; Audley-Charles et al. 1988; Metcalfe 1988; Rangin et al. 1990) account for the large-scale evolution of continental masses, but the precise history of the small fragments remains largely obscure. As pointed out by Hartono (1990), the microcontinents probably do not all have the same origin. Some may not have drifted far and remained relatively close to a northwest Australian origin (Timor, Tanimbar, Seram), whereas others (Buton, Sula, Banda Ridges) may originate from northern New Guinea and have moved over considerable distances along the Sorong fault system (Fig. 1). In this paper, the applicability of isotopic data is tested in assessing the provenance of four microcontinents: Ambon-W. Seram, Bacan, the Banda Ridges and Sumba. The evidence is based on a comparison of Pb-Nd isotopic signatures between (meta-) sedimentary or volcanic rocks from the microcontinents and possible provenance areas.
Analytical techniques Fine-ground samples were dissolved in a HF:HNO 3 (4 : 1) mixture at 180°C for at least a week. Nd was separated using a standard ion exchange technique and measured on a variable multicollector VG354 mass spectrometer at RHUL. A quadruple jump experiment was used (Thirlwall 1991) and 21 runs of the Nd in house standard ('Aldrich') yielded 0.511422_+ 11 (2 s.d.). The average whole procedure Nd blank was 250 pg. The Pb separation procedure was similar to the one described by White & Dupr6 (1986). The Pb isotopes were measured on a variable multicollector VG354 mass spectrometer at RHUL in static mode. The results were corrected for mass fractionation (1.3%o per amu) based on replicate runs of NBS SRM-981. The average total procedure Pb blank was 300 pg.
Isotopic characteristics of continental provenance regions The microcontinents originate from the former continental margins of Australia-New Guinea or Sundaland. In terms of Pb-Nd isotopic compositions, significant variation exists within the continental masses of northern Australia and New Guinea, largely due to major variations in the age of outcropping rocks. Ancient formations of Middle Proterozoic age (2200-1400Ma) are dominant
in northern Australia, while Phanerozoic ages (<600 Ma) are characteristic of western New Guinea. Palaeozoic rocks occur in the Bird's Head and the Snow Mountains, and may be overlain by formations up to Miocene in age. The northern part of New Guinea consists mainly of island-arc volcanics and subordinate oceanic basalts which collided with southern New Guinea during the Miocene (Hamilton 1979). As a result of this arc-continent collision, the central mountains of the island were uplifted, which triggered sediment deposition in the basins of southern New Guinea. The terrigenous fractions of surface sediments present on the shelf around the Banda arc carry the isotopic signatures of the adjacent continental areas. From a geochemical survey along and across the arc, the following major domains have been distinguished (cf. Vroon 1992; Vroon et al. 1995): (1) north Australia, with very high 2°6pb/2°4pb (up to 19.57) and low J43Nd/144Nd (=0.511900.51200), corresponding to the Proterozoic complexes on the mainland. (2) western New Guinea, with low 2°6pbfl°4pb (--18.6-19.0), relatively high 143Nd/144Nd (--0.51218-0.51225), representing probably Palaeozoic/Mesozoic complexes of New Guinea. The Pb average composition of Bird's Head (Fig. 1) has a lower 2°6pb/2°4pb (18.60-18.75) than southern New Guinea (18.75-19.0), allowing a distinction between northern New Guinea (Bird's Head) and southern New Guinea provenances (Figs 2 & 3). As will be discussed below, Sundaland and the Pacific realm of New Guinea are two other domains, which should be considered as possible origins for the fragments studied here. Their isotopic characteristics are less well defined than in the case of northern Australia and west New Guinea. Sundaland cratonized in Late Triassic times and is composed of welded cratonic terranes and younger orogenic belts (Metcalfe 1988; Hutchison 1989). It is thus reasonable to assume that Pb isotopes are on average less radiogenic than in the much older Proterozoic domain of northern Australia. The isotopic signatures of sediments from the Pacific Ocean (Figs 2 & 3) can be used as reference for a 'Pacific' origin.
Isotopic affinity and origin of microcontinents The isotopic characteristics of sediments and volcanics from the continental fragments can be used to relate their provenance to one of the abovementioned domains. For terrigenous sediments
PROVENANCE OF CONTINENTAL FRAGMENTS
A
Bacan dacites
AUSTRALIA 15.8.
Ambon
g_ 15.7
Sumba
~,. ~ ' ~
~
Ind. . . .
ian
Sulawes
ts
o,1 15.6. ff Error
New BritaintlJ~ Bna da ridges meta sediments
15.5 18.0
18.5
i 190
i 19.5
20.0
206 P b / 2 0 4 p b
B
Bacan , ~ dacites
AUSTRALIA
#_ ~o
Ambon
39.5"
\/
P,o,,,oo. . . .
esian
#_
. . . . . sediments
39.0 -
sour, NEWGUINEA
.............. 38.5-
::
NORTH
Sumba i , 19.0 206 P b / 2 ° 4 p b
ZTError
New Britain Sulawesi
38.0 18.0
11~.5
p 19.5
Fig. 2. (a) 2°6pb/2°4pb vs. 207pb]2°4pb and (b) 2°6pb]2°4pb vs. 2°8pb/2°4pb showing Pb isotopic composition of Bacan, the Banda Ridges, Ambon and Sumba, and possible provenance areas (north New Guinea, south New Guinea and Australia). Data sources are given in Table 1 - New Britain volcanic arc: Woodhead & Johnson (1993), Pacific Ocean sediments and east Indonesian sediments: Vroon et al. (1993; 1995) and references cited therein.
I 0.5130
~
M
O
~
~
New Britain Arc volcanics
0.5128
IB~anda-- ridges mete sediments
~.~Z o.5126-" Z
Sul . . . . i .........~.~.
0.5124
Sumba
sediments
~ O ~ G u I NSOUTH E NORTH ~ ~ NEWGUINEA/..,,~ ~
...... ...... 0.5118 17.o
"~11
Pacific ~ Ocean ---..... ~
Ambon
i
East ,rid,. . . . ian 17'.s
1~.0
"~.
A Bacanl ,m,J ~ ~J1
~
[
sedimen~,AUST,RAL~I
185 206pb/204pb
~.o
19.5
~.o
Fig. 3. 2°6pb/2°4pbvs. 143Nd/144Ndshowing the possible provenance areas and the Bacan-Ambon crustal contaminated volcanics and Banda Ridges metasediments. I-MORB = Indian Mid-Oceanic Ridge Basalt. References as in Fig. 2. See text for discussion.
447
which were deposited prior to, or simultaneous with, the detachment of a fragment, it is assumed that their Pb-Nd isotopic signatures correspond to those of the formerly adjacent continental region. The isotopic ratios of volcanics provide an indirect means to assess the affinity of a fragment. Given the large contrast in Pb concentrations between mantle and continent-derived material, the Pb isotopes are particularly sensitive to contamination, such that ratios observed in volcanic rocks tend to be dominated by the 'continental' signature, even if the degree of contamination is small. Contamination of a subduction-related magma may occur either in the mantle source through addition of subducted sediments, or by assimilation of crustal rocks from the overriding plate during ascent. Hence, assimilation (as well as crustal remelting) yields information on the isotopic characteristics of the basement. On the other hand, source contamination provides a signature of the crustal material which entered the subduction system when volcanism was active. The volcanic Banda arc, where the Pb-Nd isotopic signatures reflect the recent input of sediments from the surrounding passive Australian margin, represents a clear example of this case (Vroon et el. 1993; Van Bergen et al. 1993). New and published isotope data for volcanic and sedimentary rocks from Ambon, Bacan, the Banda Ridges and Sumba are given in Table 1. A comparison between the respective microcontinents and potential provenance regions is presented in Figs 2 & 3. A m b o n ( a n d W. S e r a m )
The island of Ambon forms part of the extinct northern end of the volcanic Banda arc. PlioPleistocene volcanic rocks, which cover most of the island, and associated granitic intrusives have been dated at 3.4-4.5 Ma (Priem et al. 1978; Abbott & Chamalaun 1981). Exposed basement rocks are unmetamorphosed Triassic sediments but the widespread presence of polymetamorphic xenoliths in the volcanics points to the presence of a folded metamorphic deeper basement similar to formations of southern Seram (Van Bergen et al. 1989; unpublished data Utrecht). According to Pigram & Panggabean (1984), Ambon forms part of a Buru-Seram microcontinent, which was derived from the northeast sector of the former Australian continental margin and rifted away during the Middle Jurassic. However, different views exist on how much of Seram can be considered as a microcontinental block, and on whether Buru may have a structurally separate origin (see Hamilton 1979; Audley-Charles et al. 1979; Tjokrosapoetro & Budhitrisna 1982). In this paper we use the isotopic
448
p . z . VROON ET AL.
Table 1. Nd-Pb isotopic compositions of sediments and volcanics of east Indonesian microcontinents Sample
Type
143Nd/144Nd
2°6pb/2°4pb
2°Tpbfl°4pb
Ambon AM3A AM7
granite granite
0.512099 -+ 05 0.512124 -+ 05
AM26B
andesite
0.512168 + 06
AM32I AM48I
rhyolite dacite
0.512111 _+21 0.512080 ___07
AM93Al-I AM 104A 1
andesite andesite
0.512181 + 07 0.512131 -+ 05
18.893 18.874 18.878 18.871 18.861 18.865 19.001 19.017 18.886 18.842
15.695 15.681 15.683 15.676 15.663 15.676 15.709 15.725 15.678 15.676
39.172 39.114 39.119 39.098 39.053 39.094 39.387 39.435 39.128 39.083
0.512125 0.512150 0.512104 0.512161 0.51215 0.51217
_ 05 _ 05 -+ 08 -+ 06
19.934 19.865 19.911 19.907 19.88 19.94
15.797 15.807 15.795 15.805 15.80 15.83
40.200 40.165 40.172 40.193 40.17 40.33
SW Sulawesi (Balangburu Formation) CJ30.5 shale CJ68.5 shale CJ97.4 sandstone
0.512463 +_05 0.512502 +_09 0.512548 _+05
18.738 18.668 18.730
15.647 15.628 15.638
38.880 38.733 38.818
Sumba (Lasipu Formation) SL.L 1 shale SL.P1 shale SL.R3 shale
0.512476 +_05 0.512459 _+04 0.512439 _+04
18.767 18.767 18.743
15.645 15.645 15.635
38.883 38.883 38.846
Banda Ridges -
0.51290 0.51260
18.49 18.97
15.55 15.58
38.39 39.17
Bacan BE2 BE6 BE7 BE9 BC-3 BC-6
dacite dacite dacite dacite dacite dacite
metasediment metasediment
2°8pb/2°4pb
Data of this study except Bacan (BC-3 and BC-6) and the Banda Ridge samples, which are from Morris et al. (1983) and Morris et al. (1984), respectively. Errors for 2°6pb/2°npb, 2°Tpb/2°4pb and 2°8pb/2°4pb are _+12, ___12and _+40, respectively.
data of the A m b o n volcanics to assess the origin of a continental fragment represented by the basement of A m b o n and the adjacent western part of Seram where Palaeozoic m e t a m o r p h i c c o m p l e x e s are widespread (Linthout et al. 1989). Evidence exists for a significant anti-clockwise rotation of Seram during the Late Cenozoic (Haile 1981; Nishimura & Suparka 1986), but its palaeoposition is still uncertain. The acidic volcanic rocks of A m b o n display typical values of crustally contaminated rocks: 143Nd/144Nd -- 0.51208-0.51218, 87Sr/86Sr = 0.7105-0.7188 and 2°6pb/2°4pb = 18.7-19.0 (Table 1 and unpublished data RHUL). These data are consistent with the preliminary results of Morris (1984). The polymetamorphic xenoliths and the existence of a crude positive correlation between SiO 2 and 2°6pb/2°4pb-87Sr/86Sr provide evidence for assimilation of basement rocks. The most
contaminated rocks have high 87Sr/86Sr (>0.715) and 2°6pb/2°4pb (>18.80) ratios, the latter overlapping the south New Guinea fields (cf. Figs 2 & 3). Pre-Triassic metapelitic and granitic rocks from the Tehoru Formation and Kaibobo Complex of W. Seram (cf. Linthout et al. 1989) have Nd and Pb isotopic ratios which are virtually identical to the A m b o n volcanics (e.g. 143Nd/144Nd= 0.51204-0.51220 and 2°6pb/2°4pb = 18.74-18.80; Tommasini et al. 1994). These data support the supposition that A m b o n and W. Seram share the same basement and belong to one microcontinental block derived from south New Guinea.
Bacan
The island of Bacan is situated in the southern part of the Halmahera arc (Fig. 1). In contrast to the rest
PROVENANCE OF CONTINENTAL FRAGMENTS of the arc, high-grade metamorphic continental basement rocks are exposed in central Bacan (Van Bemmelen 1949) and may also underlie the southern part of the island and Obi (Hall et al. 1991). Both islands have been considered to form one microcontinent (e.g. Pigram & Panggabean 1984), but they may represent different blocks separated by a splay of the Sorong fault system. It has been suggested that the continental basement rocks of Bacan are derived from New Guinea (e.g. Hamilton 1979). Dacites of Quaternary age from Bacan have similar 87Sr/86Sr (0.7098-0.7248) and 143Nd/144Nd (0.51210-0.51217) as the Ambon volcanics, and are also considered to represent crustally contaminated rocks (cf. Morris et al. 1983). The Sr isotopes form two different groups (=0.719 and =0.725) and tend to show a positive correlation with SiO 2, which provides evidence for assimilation. Pb isotopes are much more radiogenic than in the case of Ambon, and suggest that the assimilated component has 2°6pb/2°4pb close to 19.9. However, outcropping metamorphic basement rocks of Bacan have lower 2°6pb/Z°4pb (18.64-19.40) and higher 143Nd/la4Nd (0.51227-0.51291; unpublished data, E. Forde). These signatures make these rocks unlikely assimilants. If other formations with highly radiogenic Pb isotopes are representative of (part) of the Bacan crust, it is likely that this block originates from northern Australia, unless ancient Precambrian basement is present in the basement of the northern parts of west New Guinea. Two different options for the timing of the arrival of continental crust in this area have been suggested (Hall et al. 1991). Either late Neogene westward transport of Australian continental crust along the Sorong fault system resulted in the juxtaposition with Pacific arc-ophiolitic terranes along strike-slip faults, or both types of terranes were first amalgamated by earlier (Oligocene?) arc-continent collision at the northern 'Australian' margin before they were incorporated as complex units in the Sorong fault system. However, a 'Precambrian' Australian, rather than a New Guinean origin of the Bacan fragment is difficult to reconcile with these models, as both imply a provenance in north New Guinea. An alternative may therefore be that the block had already arrived in the region before the Sorong fault system became active.
Banda Ridges
The Banda Ridges (Lucipara Ridges) are relatively shallow structural platforms in the northern part of the Banda Sea (Fig. 1). Clastic sedimentary and
449
metamorphic dredge samples, two of which were dated at 10.8 and 22.5 Ma, revealed the continental character of the ridges (Silver et al. 1985). According to these authors, the lithology and age resemble those of rocks on other microcontinents in this region, and are particularly similar to the Tamrau Formation on the Bird's Head, which contains Jurassic and Cretaceous fossils. Hence, the Banda Ridges are considered to represent fragments of the northern margin of New Guinea which were displaced along a left-slip fault system. The dredged metasedimentary rocks display the characteristics of arc volcaniclastics and have 87Sr/86Sr = 0.7047-0.7073 and 143Nd/144Nd = 0.5126-0.5129 (Morris et al. 1984). Their a°6pb/e°4pb and e°8pb/2°4pb ratios largely overlap with the New Guinea values, but the a°Tpb/a°4pb ratios are much lower than observed in any of the provenance regions defined above. Based on Pb-Nd signatures (Fig. 3), it seems more plausible that the origin of the Banda Ridges is related to the Cenozoic part of New Guinea composed of accreted Pacific Ocean sediments and island arcs (e.g. Hamilton 1979) rather than to its 'continental' Palaeozoic-Mesozoic area. The Banda Ridges plot between Pacific Ocean sediments and the island-arc volcanics (as represented by the New Britain arc (Woodhead & Johnson 1993) further to the east), thus the isotopic data favour a 'Pacific' origin outside the Palaeozoic or older domains of New Guinea and Australia (Fig. 3).
Sumba
The island of Sumba is a continental fragment in the fore-arc region between the East Sunda volcanic arc and the Java-Timor trench (Fig. 1). Based on stratigraphic and palaeomagnetic considerations, numerous studies have related the provenance of Sumba to different regions: (1) the Australian continent to the south; (2) the margin of southeast Sundaland to the north; or (3) an isolated position within the Tethys region (e.g. Hamilton 1979; Otofuji et al. 1981; Von der Borch et al. 1983; Audley-Charles 1985; Rangin et al. 1990; Wensink 1994). Marine sediments were analysed from the Late Cretaceous Lasipu Formation which are the oldest rocks on Sumba, and were deposited prior to or simultaneous with the initial stages of the island's drift history. They display limited variations in 143Nd/144Nd (0.51244-0.51248) and Pb isotopes (2°6pb/e°4pb = 18.74-18.77). These isotopic signatures do not correspond to the Australian or New Guinean continental domains, and thus favour a northern rather than a southern origin. Because of stratigraphic indications for a palaeoposition of
450
P.Z. VROON ET AL.
Sumba near SW Sulawesi (Simandjuntak 1993), Upper Cretaceous flysch sediments from the Balangburu Formation of SW Sulawesi were analysed for comparison (Table 1). They yielded 143Nd/144Nd = 0.51246-0.51255 and 2°6pb/2°4pb = 18.67-18.74, which implies a close isotopic similarity with the Lasipu Formation. In the Mesozoic prior to opening of the Makassar Strait, SW Sulawesi was probably joined with SE Kalimantan (e.g. Katili 1978; Hamilton 1979) on the border of Sundaland. The isotope results thus provide evidence for Sumba's affinity with Sundaland, and corroborate the same conclusion from recent paleomagnetic and geochemical work (Wensink 1994; Wensink & Van Bergen 1995).
Discussion and conclusions The Pb-Nd isotopic signatures of rocks from the microcontinents studied here provide new evidence on their pre-drift origin. Ambon-W. Seram has affinities with south New Guinea, Bacan with Precambrian northwest Australia, the Banda Ridges with 'Pacific' north New Guinea and Sumba with southeast Sundaland. These findings are generally compatible with previous hypotheses from geological correlations, and pose constraints on various options for palaeopositions and trajectories that have remained open so far. An important implication of the findings is that no single mechanism can be invoked to reconstruct the origin and emplacement of all of the fragments (cf. Hartono 1990). For Ambon-W. Seram only a small northwestward shift relative to western New Guinea, possibly accompanied with a significant anticlockwise rotation (cf. Nishimura & Suparka 1986; De Smet 1989; Linthout et al. 1991), is required to account for the observed south New Guinea-like isotopic signatures. The 'tectonic shaving of north New Guinea' mechanism seems less likely because it would imply a too northern origin. The interpretation presented here thus concurs with the suggestion of Pigram & Panggabean (1984), who proposed that the origin of Seram may have been along the former margin of the Australian continent somewhere east of the Joseph Bonaparte Gulf (Fig. 1). The inferred origin of Bacan is also difficult to reconcile with a dominant westward movement along the Sorong fault system. If its affinity with Precambrian domains is correct, a major northward drift of this block, starting along the NW Australian margin, seems a more plausible direction of movement. Early dislocation from the provenance area, perhaps during the Mesozoic rifting of the Australian margin, may have led to the arrival at its
northern position prior to the initiation of the Neogene Sorong fault system. Amalgamation with adjacent blocks is then a relatively young feature, related to the westward movement of north New Guinean and Pacific terranes. The provenance of the Banda Ridges from northernmost 'Pacific' New Guinea fits with the mode of emplacement suggested by Silver et al. (1985), who envisaged the ridges as fragments of New Guinea which were displaced along a left-lateral transform system. Although the isotopic composition of the andesites suggests that Australian craton material has contributed to the andesite genesis (Silver et al. 1985), it probably signals source contamination from subducted Australian margin sediments, rather than major assimilation of basement rocks. The present position of the Banda Ridges south of the Ambon-W. Seram block has consequences for the relative timing of emplacement. Irrespective of the options for Seram's origin, its emplacement must have occurred after the Banda Ridges had arrived in the Banda Sea area, perhaps some 5-10 Ma ago (cf. Silver et al. 1985; Hartono 1990). Any traces of the fault system along which the ridges moved may have been obliterated by later tectonic events associated with the closure of the Banda arc. Significant movement and rotation of Seram may still have taken place during the last 3 Ma (Linthout et al. 1991). The palaeoposition of Sumba near the border of Sundaland also has implications for the timing of the drift history of this microcontinent. Critics of this northern origin have pointed at the difficulty of how the microcontinent could have passed the Tertiary Sunda arc (e.g. Katili 1989). However, recent palaeomagnetic evidence indicates that Sumba started to drift in the Late Cretaceous and had already arrived at or near its present position in the Early Miocene (Wensink 1994; Wensink & Van Bergen 1995). Hence, its emplacement may have been contemporaneous with the initial stages of development of the adjacent easternmost volcanic sector of the Sunda arc. Alternatively, the volcanic front of the proto-East Sunda Arc in this region may have had a more southerly position (cf. Fortuin et al. 1994), and subsequently shifted northward. More detailed age dating of the Tertiary magmatic rocks is required to distinguish these alternatives. Based on present plate movements, McCaffrey & Abers (1991) presented an evolutionary scenario which illustrates the structural complexity of the east Indonesian orogenic belt as it will appear after the next 10 Ma. They emphasized the role of strikeslip faulting in juxtaposing the continental and oceanic elements. Their model can also be used to predict the future complexity of this region in terms of isotopic heterogeneity, as an example of a
PROVENANCE OF CONTINENTAL FRAGMENTS
The Banda Orogen in 10 Ma
Bacan
~iiiiiiiiiiiiiiiiiiiii~ "
•
..... :~:'"~ii'~i
.[.:.:•[.:
~
i
i i i i::i::iii::i::i::i::i::ii ! i:ii:ii i!i i i:
~ Australian continental crust [ ~ Mixture (dominant Australian) ~ Sunda Land continental crust
i ~ New Guineacontinentalcrust Pacific arc rocks ~ i !iAusiralia: :::ii}i::i:::i::: SE Asian arc rocks [5~ Ophiolites •
~
New Guinea
~ .~!~ii!~i~i~i~!~ii~i!~i~i~iii~i~iiiWi~i@~i~i~i~i~ii~i~i~!i~#ii
~ .. ~:::::::iBandai~rc SundaArc
451
. , . , . . . . . . . - .
Yiiii!iii }iEiiiiiili;iiiiii:ii!i!ii ~
Fig. 4. Hypothetical construction of the Banda arc in 10 Ma time, after McCaffrey & Abers (1991). Shown are some of the provenance areas of microcontinental fragments. Note that the orogen contains many 'exotic' terranes, which are difficult to relate to their original provenance areas. See text for discussion. AWS = Ambon-W. Seram.
collision belt containing drifted continental blocks of different provenances. Figure 4 shows the predicted configuration based on the scenario of McCaffrey & Abers (1991). The most striking feature is that relatively small isotopic enclaves will be generated: Sumba with a 'Sundaland' isotopic signature (e.g. relatively unradiogenic Pb) will become part of a dominantly Australian domain (radiogenic Pb); the Banda Ridges will be a Pacificlike element surrounded by (a) arc rocks with Pb isotope signatures which are largely determined by subducted passive-margin sediments with New Guinean and Australian affinities (cf. Vroon et al. 1993), and (b) dominant Australia-type as well as New Guinea-type domains; Bacan will yield a pocket of highly radiogenic Pb in a predominantly Pacific environment. Thus, the positioning of various terranes, together with superimposed effects of ageing and denudation, will finally yield an isotopically highly complex picture when orogenesis and future cratonization proceed. It is concluded that isotopic fingerprinting is a promising tool to delineate the origin of continental fragments in east Indonesia. In this paper, Pb and Nd isotopes of (terrigenous) sediments and crustally contaminated volcanic rocks are used to relate microcontinents with relatively large source domains. The apparent success of this attempt can be ascribed to the large isotopic differences between the continental masses involved. Further
detailed correlations may yield a greater spatial precision, particularly if well-dated stratigraphic intervals in pre-drift deposits of the fragments can be compared with analogues in continental regions of smaller size. In addition to isotopic signatures, trace-element ratios can provide extra constraints to define the geochemical characteristics of provenance areas (cf. Vroon et al. 1995). Similar techniques may also be applicable to study the origin of accumulated material in the accretionary wedges, and to distinguish between an Indian Ocean and Pacific origin of the oceanic terranes. We would like to thank Dr. H. Wensink and Dr. J. Ali for providing samples from Sumba and Sulawesi, respectively. Samples from Ambon were collected during the 1984-1985 Snellius H Expedition which was jointly organized by the SOZ (Netherlands Council of Oceanic Research) and the Indonesian Institute of Science (LIPI). Fieldwork on Ambon was done in collaboration with R. Erfan, T. Sriwana, A. D. Wirakusumah, K. Suharyono, R. E E. Poorter and J. C. Varekamp. Samples from Bacan were collected during fieldwork (1989-1992 with assistance of R. Hall, J. Mahaihollo, S. Baker and D. Agustiyanto) funded by NERC, Royal Society and the London University SE Asia Research Group. We thank Simone Tommasini for allowing us to quote isotope data from unpublished work. Discussions on the geology and tectonics of east Indonesia with A. Barber and R. Hall helped to improve some of the ideas in this paper. R. M. Ellam and M. A. Menzies provided helpful reviews.
452
P.Z. VROON ET AL. References
ABBOTT,M. J. & CHAMALAUN,E H. 1981. Geochronology of some Banda Arc volcanics. In: BARBER,A. J. & WIRYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication, 2, 253-272. AUDLEY-CHARLES, M. G. 1985. The Sumba enigma: is Sumba a diapiric fore-arc nappe in process of formation? Tectonophysics, 119, 435-449. - & HARRIS, R. A. 1990. Allochtonous terranes of the Southwest Pacific and Indonesia. Philosophical Transactions of the Royal Society of London, A331, 571-587. , BALLANTYNE,P. D. & HALL, R. 1988. MesozoicCenozoic rift-drift sequence of Asian fragments from Gondwanaland. Tectonophysics, 155, 317-330. --, CARTER, D. J., BARBER, A. J., NORVICK, M. S. & TJOKROSAPOETRO, S. 1979. Reinterpretation of the geology of Seram: implications for the Banda Arcs and Northern Australia. Journal of the Geological Society, London, 136, 547-566. BOWlN, C., PURDY, G. M., JOHNSTON, CH., SHOR, G., LAWVER, L., HARTONO,n. M. S. & JEZEK, P. 1980. Arc continent collision in Banda Sea Region. AAPG Bulletin, 64, 868-915. DALY, M. C., HOOPER, B. G. D. & SMITH, D. G. 1987. Tertiary plate tectonics and basin evolution in Indonesia. Proceedings of the Indonesian Petroleum Association. 16th Annual Convention, October 1987, 399-428. DE SMET,M. E. M. 1989. A geometrically consistent plate tectonic model for eastern Indonesia. Netherlands Journal of Sea Research, 24, 173-183. FORTUIN, A. R., ROEP,T. B. & SUMOSUSASTRO,P. A. 1994. The Neogene sediments of East Sumba, Indonesia products of a lost arc? Journal of Southeast Asian Earth Sciences, 9, 67-80. HAILE, N. S. 1981. Paleomagnetic evidence on the geotectonic history and paleography of Eastern Indonesia. In: BARBER, A. J. & WIRYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Special Publication, 2, 81-87. HALL, R., NICHOLS, G., BALLANTYNE,P., CHARLTON,T. & ALI, J. 1991. The character and significance of basement rocks of the southern Molucca Sea region. Journal of Southeast Asian Earth Sciences, 6, 249-258. HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey Profesional. Paper, 1078. HARTONO, H. M. S. 1990. Late Cenozoic tectonic development of the Southeast Asian continental margin of the Banda Sea area. Tectonophysics, 181, 267-276. HtYrcHISON, C. S. 1989. Geological evolution of Southeast Asia. Clarendon Press, Oxford. KATILI, J. 1978. Past and present geotectonic position of Sulawesi, Indonesia. Tectonophysics, 45, 289-322. 1989. Review of past and present geotectonic concepts of eastern Indonesia. Netherlands Journal of Sea Research, 24, 103-129. LINTHOUT,K., HELMERS,H. & ANDRIESSEN,P. A. M. 1991.
Dextral strike-slip in Central Seram and 3-4.5 Ma Rb/Sr ages in pre-Triassic metamorphics related to Early Pliocene counterclockwise rotation of the Buru-Seram microplate (E. Indonesia). Journal of Southeast Asian Earth Sciences, 6, 335-342. - - , SOPAHELUWAKAN,J. & SURYANILA, E. 1989. Metamorphic complexes in Buru and Seram, Northern Banda Arc. Netherlands Journal of Sea Research, 24, 345-356. M¢CAFFREV, R. & ABERS, G. A. 1991. Orogeny in arc-continent collision: the Banda arc and western New Guinea. Geology, 19, 563-566. METCALFE, I. 1988. Origin and assembly of south-east Asian continental terranes. In: AUDLEY-CHARLES, M. G. & HALLAM, A. (eds) Gondwana and Tethys. Geological Society, London, Special Publications, 37, 101-118. MORRIS, J. D. 1984. Enriched geochemical signatures in Aleutian and Indonesian arc lavas: an isotopic and trace element investigation. PhD thesis, Massachusetts Institute of Technology, Cambridge, MA. - - - , GILL, J. B., SCHWARTZ,D. & SILVER, E. A. 1984. Late Miocene to Recent Banda Sea volcanism, III: Isotopic compositions [abstract]. LOS Transactions, AGU, 65, 1135. --, JEZEK, P .A., HART, S. R. & GILL, J. B. 1983. The Halmahera island arc, Molucca Sea collision zone, Indonesia: a geochemical survey. In: HAVES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Part 2. American Geophysical Union, Geophysical Monograph Series, 27, 373-387. NISHIMURA, S. & SUPARKA, S. 1986. Tectonic development of East Indonesia. Journal of Southeast Asian Earth Sciences, 1, 45-57. OTOFUJI, g., SASAJIMA, S., NISHIMURA, S., YOKOYAMA, T., HADIWISASTRA, S. & HEHUWAT, F. 1981. Paleomagnetic evidence for the paleoposition of Sumba Island, Indonesia. Earth and Planetary Science Letters, 52, 93-100. PIGRAM, C. J. & PANGGABEAN,H. 1984. Rifting of the northern margin of the Australian continent and the origin of some microcontinents in eastern Indonesia. Tectonophysics, 107, 331-353. PRIEM, H. N. A., ANDRIESSEN, P. A. M., BOELRtJK, N. A. M., HEBEDA, E. H., HUTCHISON, C. S., VERDURMEN, E. A. TH & VERSCHURE,R. H. 1978. Isotopic evidence for a middle to late Pliocene age of the cordierite granite on Ambon, Indonesia. Geologie en Mijnbouw, 57, 441-443. RANGIN, C., JOLIVET, L., PUBELLIER, M. & THE TETHYS WORKINC GROUP. 1990. A simple model for the tectonic evolution of southeast Asia and Indonesia for the past 43 m.y. Bulletin Societe Geologique de France, 8, 889-905. SILVER, E. A., GILL, J. B., SCHWARTZ,D., PRASETYO,H. & DUNCAN, R. A. 1985. Evidence for a submerged and displaced continental borderland, north Banda Sea, Indonesia. Geology, 13, 687-691. SIMANDJUNTAK, T. O. 1993. Tectonic origin of Sumba Island. Jurnal Geologi dan Sumberdaya Mineral, 3, 10-20.
PROVENANCE OF CONTINENTAL FRAGMENTS THIRLWALL, M. E 1991. Long term reproducibility of multicollector Sr and Nd isotope analyses. Chemical Geology, 108, 85-104. TJOKROSAPOETRO, S. & BUDHITRISNA,T. 1982. Geology and tectonics of the Northern Banda Arc. Bulletin of
the Geological Research and Development Centre, 6, 1-17. TOMMASINI, S., DAVIES, G. R., STAUDIGEL,H. & MANETTI, P. 1994. Sr, Nd and Pb isotopic data of crustal melts from the island of Seram, Indonesia: evidence for disequilibrium melting during anatexis (abstract). US Geological Survey Circular, 1107, 325. VAN BEMMELEN,R. W. 1949. The Geology of Indonesia, vol. 1A. Government Printing Office, The Hague. VAN BERGEN, M. J., ERFAN, R. D., SRIWANA, T., SUHARYONO,K., POORTER,R. E E., VAREKAMP,J. C. & VROON, P. Z. 1989. Spatial geochemical variations of arc volcanism around the Banda Sea. Netherlands Journal of Sea Research, 24, 313-322. VAN BERGEN, M. J., VROON P. Z., & HOOGEWERFF,J. A. 1993. Geochemical and tectonic relationships in the East Indonesian arc-continent collision region: Implications for the subduction of the Australian passive margin. Tectonophysics, 223, 97-116. VISSER, W. A. & HERMES,J. J. 1962. Geological results of the exploration for oil in the Netherlands New Guinea. Koninklijk Nederlands Geologisch
Mijnbouwkundig Genootschap, Verhandelingen Geologische Series, 20, 1-265. WON DER BORCH, C. C., GRADY, A. E., HARDJOPRAWIRO, S., PRASETYO, H. & HADIWISASTRA, S. 1983. Mesozoic and Late Tertiary submarine fan sequences and their tectonic significance, Sumba, Indonesia. Sedimentary Geology, 37, 113- 132.
453
VROON, P. Z. 1992. Subduction of continental material in the Banda Arc, Eastern Indonesia: S r - N d - P b isotope and trace-element evidence from volcanics and sediments. PhD thesis University of Utrecht,
Geol. Ultraiectina, 90. ---,
VAN BERGEN, M. J., KLAVER, G. & WHITE, W. M. 1995. St, Nd and Pb isotopic and trace-element signatures of the East Indonesian sediments: Provenance and implications for Banda Arc magma genesis. Geochimica et Cosmochimica Acta, 59, 2573-2598. --, WHITE, W. M. & VAREKAMP, J. C. 1993. Sr-Nd-Pb isotope systematics of the Banda Arc, Indonesia: Combined subduction and assimmilation of continental material. Journal of Geophysical Research, 98, 22349-22366. WENSINK, H. 1994. Paleomagnetism of rocks from Sumba: tectonic implications since the late Cretaceous. Journal of Southeast Asian Earth Sciences, 9, 51-65. & VAN BERGEN, M. J. 1995. The tectonic emplacement of Sumba in the Sunda-Banda Arc: paleomagnetic and geochemical evidence from the early Miocene Jawila volcanics. Tectonophysics, in press. WHITE, W. M. & DUPRI~, B. 1986. Sediment subduction and magma genesis in the Lesser Antilles: isotopic and trace element constraints. Journal of Geophysical Research, 91, 5927-5941. WOODHEAD, J. D. & JOHNSON, R. W. 1993. Isotopic and trace-element profiles across the New Britain island arc, Papua New Guinea. Contributions to Mineralogy and Petrology, 113, 479-491.
4°Ar/39Ar constraints on obduction of the Seram ultramafic complex: consequences for the evolution of the southern Banda Sea KEES L I N T H O U T l, H E N K H E L M E R S 1, J A N R. W I J B R A N S 1 & J A N D I E D E R I K A. M. VAN W E E S 2
Departments o f Petrology & Isotope Geology I and Sedimentary Geology 2, Vrije Universiteit, De Boelelaan 1085, 1081HVAmsterdam, The Netherlands Abstract: On Kaibobo (W Seram), obduction of hot oceanic lithosphere produced high-grade
metamorphism and granite in overthrust continental crust. 4°Ar/39Ar plateau ages of 5.656.0 Ma and 5.4 Ma of muscovite and biotite from the sole, among the youngest so far measured on single crystals, indicate complete resetting of the Ar-isotope system. Chloritized biotite crystals gave disturbed spectra and younger ages. Single fusion on biotite from the granite gave an age of 5.5 Ma. Time limits for the obduction-exhumationhistory of the Kaibobo complex and sole are approximated by simple 1D forward P-T-t modelling, constrained by PT data and 4°Ar/39Arthermochronology(c. 400°C at 6 Ma; c. 320°C at 5.5 Ma). Best fit results indicate that the decelerating post-emplacementexhumation of ophiolite and sole began < 8 Ma ago. The rate during the first 1 Ma of this uplift exceeded 7.5 mm a-1. Undoing 8 Ma of migration in the SW Pacific and Australian regimes of motion back-tracks Kaibobo to the site where obductionemplacement ended: near the southeastern corner of the Banda Sea plate. Similar compositions and time-settings of the Kaibobo and N Timor ophiolites lead us to postulate that, during the Early Miocene, slow-rate spreading occurred in the oceanic lithosphere of the southern Banda Sea, south of the current volcanic arc.
Ultramafic complexes on islands and peninsulas around SW Seram (Figs 1 & 2) occupy the highest Neogene thrust sheets, overlying Palaeozoic, continental, regional-metamorphic rocks (AudleyCharles et al. 1979; Linthout et al. 1989; Sopaheluwakan et al. 1992). The ultramafic rock on Kaibobo is finely banded and isoclinally folded. At five localities, in sole and ophiolite, crossing folds in the latter, ENE-directed (062°___25 °) stretching lineations were measured. Widespread serpentinization is locally associated with tremolite-talc-chlorite schist. Net-veining by very coarse-grained hornblende gabbro and sills of finegrained ophitic diabase is common. Pillow basalts, of unknown petrological affiliation, reported from Kelang (60 km west of Kaibobo) are 7.6 +_ 1.4 Ma old (Beckinsale & Nakapadungrat 1979). On the basis of geochemistry, petrography and thermobarometry, Linthout & Helmers (1994) interpreted the ultramafic rocks as remnants of weakly depleted, lherzolitic lithosphere, representing an early stage in oceanic development, about 10 Ma old at the time of obduction. Obduction-induced metamorphism in the sole reached at least 740°C, at 0.4-0.5 GPa, changing low- to medium-grade metamorphic rocks into high-grade phyllonites, mylonitic gneisses and amphibolites. Coincidence of the retrograde part
of the sole's PT-curve with the cooling path of the ophiolite indicates a joint exhumation history. Cordierite-bearing granite, intruded into the ophiolite, contains a significant metapelitic restite component and its melt formed under the PT conditions of the metamorphic peak in the sole. On these grounds Linthout & Helmers (1994) argued that its melt was generated by obduction-induced partial anatexis in that sole. Identical Nd and Pb isotope ratios in the Kaibobo granite and metapelites from the sole (Tommasini et al. 1994) confirm this conclusion. The vestige of important mantle-crust interaction on Seram should be understood within the framework of the intricate dynamics of the Neogene geotectonic setting of the Banda arc region, where three major plates interact: the slowly moving Eurasian plate, of which the Banda Sea plate currently forms the southeasternmost part; the Australian plate moving northward at c. 70 km Ma-1; and the SW Pacific plates in westward motion at c. 100 km Ma -t (DeMets et al. 1990; Smith et al. 1990). Exotic microblocks can be attributed to one or the other major plate and their translations, rotations and deformations have to be understood as a result of the interaction of the three large domains of directions of motion (cf. Rangin et al. 1990). The E-W elongated island of
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 455-464.
455
456
K. LINTHOUTET AL.
Seram, currently in the SW Pacific plates' regime of motion (Fig. 1), is a continental fragment of Australian derivation (Hamilton 1979; Pigram & Panggabean 1984; Vroon 1992). In accord with the rates and directions of the large plates, and sustained by palaeomagnetic work by Haile (1979), suggesting that since the Late Miocene Seram moved 2 ° northward and rotated 74 ° anticlockwise, Linthout & Helmers (1994), much like Johnston (1981), proposed a model in which Seram, as part of the Australian plate, initially migrated along the 'NNW transform', marking the eastern boundary of the Banda Sea plate (Fig. 1). Approaching the SW Pacific plates, Seram underwent ever stronger westerly-directed stress, exerted on the Australian plate by the strong westward motion at its northern boundary; in the Early Pliocene, during transition from the Australian plate into the regime of the SW Pacific plates, Seram was rotated (and block-
faulted, Linthout et al. 1989, 1991) into an E-W orientation, after which it migrated westward to its current location. In this scenario, the provenance of Kaibobo's ophiolite is constrained by the time of its obduction. Linthout & Helmers (1994) considered rocks generated by obduction-induced partial anatexis as time markers for obduction and subsequent cooling. To this end they used Rb/Sr ages 4.4 Ma and 3.3 Ma of cordierite-bearing dacite and granite, respectively, given by Abbott & Chamalaun (1981) and Priem et al. (1978). These young ages, however, are not reliable since, in a study comprising cordierite-bearing Kaibobo granite, featuring partly resorbed andesine cores in oligoclase phenocrysts (Linthout & Helmers 1994), evidence was given for disequilibrium melting with respect to the Sr isotopes during anatexis in the sole (Tommasini et al. 1994).
Fault zo..~4 ' /
/ C/~.ru
_...~.~.-\ o ° 0
\~'~
o
~/~ m
°
AUSTRALIAN PLATE
10 ° -
5 0 0 km !
1 4o
I
,
128o
Fig. 1. Reference map of the Banda arc region. Arrows correspond to plate motion directions. Dashed lines surround areas exposing ultramafic rock, associated basalts and high-grade metamorphic rocks on Seram (N Banda arc) and on N Timor and some smaller islands (S Banda arc). Circle indicates the restored site of obduction of the Kaibobo ophiolite, NE of Tanimbar, where the 'NNW transform' (dotted line) between the Australian and the Banda Sea plates intersects the outer Banda arc, as marked by the incision in the 1000 m depth contour.
40AR--39AR
DATING
OF
SERAM
To better constrain the Neogene thermal evolution of the rocks of Kaibobo, 4°Ar/39Ar ages of biotite and muscovite crystals were measured from the sole's high-grade phyllonites and from cordierite-bearing granite. The accurate age data prompted a model of the joint exhumation history of the ophiolite and its sole, on the basis of Linthout & Helmers' (1994) well-documented PT trajectories, in order to approximate the time and implicitly the site of obduction.
Petrography The sole samples studied, from c. 50 m below the contact with the ophiolite, are sillimanite and garnet-bearing phyllonite and gneiss with 'mica fish' crystals, typical of mylonitic deformation. Fibrolitic sillimanite occurs in microfolded and boudinaged bundles. Since the mica fish show inclusions of microfolded trails of fihrolite, their blastesis followed formation and deformation of sillimanite. As sillimanite was formed by obduction metamorphism only (Linthout et al. 1989), mica blastesis must be related to obduction as well. In BK21A, layers of lattice orientated quartz crystals contain strongly altered andalusite and feldspar clasts. These are interpreted as pre-phyllonitic relicts of the medium-grade regional metamorphism, characteristic of the Palaeozoic continental rocks of Seram (Linthout et aL 1989). Biotite crystals are typically fresh, red-brown and 2 mm long, and muscovite crystals are 0.7 mm long. Electron microprobe analysis of biotite, indicating
q
Can ~
...':-.-.=:-:-:-:-:..:::_ _::-_...:::-2..:::-:...Vv-..= -::-
OPHIOLITE
Fig. 2. Geological sketch map of Kaibobo, partly after Valk (1945); sites of the samples used for this study are indicated.
457
9.47 wt% K20 and a total dry wt% of 94.59, confirms its freshness; the annite content is 33%. In BK21D, with muscovite and biotite crystals up to 3 mm, the greenish brown biotite apparently is chloritized, having only 8.85 to 6.65 wt% K20 and dry totals of 93.46-90.75 wt%. In the mylonitic gneiss BK210 muscovite crystals are up to 7 mm and biotite is chloritized. The fine-grained granite BK18 contains medium-sized crystals of antiperthitic oligoclase, with partly resorbed cores of An55_35, and cordierite. The granular matrix of quartz, poikilitic microcline perthite and dispersed, mm-sized, fresh, brown biotite flakes, displays a gradual transition into gneissose xenoliths of spinel-sillimanite-garnet-(corundum)-cordieritebiotite gneiss and quartzite.
Analytical results and discussion Muscovite and biotite crystals of c. 250-1000 mm were separated by hand picking after gentle crushing and sieving of rock samples. Twenty crystals were loaded on a 22 mm diameter A1 tray in 3 mm diameter, 2 mm deep holes. Four positions were loaded with a mineral standard (TCR sanidine, K-Ar age 27.92 Ma). The J factor for these experiments was 0.003275 with an uncertainty of 0.4% (lcy). The remaining holes were filled with single crystals of unknowns. The experimental techniques are described more fully by Wijbrans et aL (1995). Following overnight bakeout, the measurement routine started with three replicate analyses of air argon, and system blanks were measured following every block of five runs of unknowns. Laser heating at increasing laser power levels for one minute were followed by four minutes of clean up on the Fe-V-Zr and Zr-A1 alloy getter cartridges and isotopic analysis on an MAP 215-50 mass spectrometer. All isotopes were measured using an electron multiplier collector. System blanks were typically m/e:40 4 x 10-17 moles, m/e:39 4 x 10-19 moles, m/e:38 6 × 10-19 moles, m/e:37 1 × 10-17 moles, m/e:36 1 × 10-18 moles, based on an estimated system sensitivity of 2 × 10-17 moles/mV. The experiments were carried out as single ~fusion analyses, and for the samples from the sole also as incremental heating analyses on single crystals. The latter technique permits the identification of excess 4°Ar and of contaminating phases. This was considered necessary since the Neogene obduction metamorphic sole rocks do have a Palaeozoic metamorphic history (Linthout et al. 1989), and some of the biotite samples were somewhat chloritized. As the incremental heating experiments were carried out with rather low ion beam intensities in the mass spectrometer, on single flakes of mica < 5 mg and < 6.0 Ma old, they were
~~'~i*~/~ ~ [~ obduct c;°rldai;rani ~ioi~n-ii~tenducecl ~/~;~/~.~__~ ~ rnetamorphi rockc ~"~//***~**** i i KV~AGE ~-'t O----~.thrust,2 Triassic limestone
OBDUCTION
I<. L I N T H O U T ET AL.
458
Table 1. Data age spectrum and single fusion experiments 36Ar(Atm)
Ms
37A1
BK21D age spectrum 0.000066 0.000002 0.001266 0.000102
0.000002 0.000323
0.000039
0.000002
0.000242 0.000158
0.000371 0.000002
0.000101
0.000002
38Ar
39Ar(K)
4°Ar(cor)
% 39Ar
4°Ar*/39Ar K
% 4°Ar*
Age
Sd Age
0.000436 0.013308 0.004007 0.001376 0.004322 0.003570 0.004488
0.037242 1.102730 0.346613 0.123768 0.366767 0.305908 0.377270
0.063397 1.493429 0.371159 0.133951 0.426120 0.344558 0.392930
1.4 42.9 55.9 60.5 74.3 85.8 100.0
1.175 1,015 0,984 0.989 0.967 0.974 0,962
69.2 75.0 91.9 91.4 83.2 86.4 92.4
6.93 5.99 5.80 5.83 5.70 4.74 5.68
1.71 0.07 0.18 0.51 0.17 0.21 0.17
Mean
5.90
0.14
5.0 10.4 11.9 11.3 11.8 11.5 11.9 15.7
Mean
2,55 2.67 3.16 2.95 3.11 2.88 2.85 3.54 2,91
0.18 0.17 0.21 0.25 0.28 0.15 0.48 0.23 0.21
5.0 5.4 34.1 93.2 95.6 90.6 89.8 97.9
7.68 5.63 5,84 5,78 5.70 5.46 8.39 5.99
6.08 6.64 0.73 0.15 0.16 0.40 0.71 0.19
Mean
5.83
0.26
50.7 59.7 61.4 67.8 89.1 72.5 79.0 92.2
4,59 4.23 4,02 5.69 9.53 5.81 5.02 6.29
2.05 1,62 2.72 1.88 0.84 0.96 1.68 0.82
Mean
6.64
1.22
25.1 72,5 83.6 86.1 87.5 91.8 75.4
5.68 5.65 5.35 5.39 5.35 5.50 4,38
1.38 0.42 0.65 0.43 0.36 0.37 0.83
Mean
5.41
0.48
BK21D age spectrum 0.026913 0.007246 0.006074 0.005077 0.003888 0.006815 0,001659 0.004730
Ms
Ms
0.011127 0.004526 0.003972 0.005041 0.003587 0.005940 0.001115 0.003458
BK210 age spectrum 0.000960 0.000002 0.000531 0.000578 0.000103
0.000002 0.000002 0.000002
0.000058 0.000052
0.000002 0.000023
0.000049
0.000482
0.000026
0.000002
BK21A age spectrum 0.000073 0.000245 0.000059 0.000002 0.000031 0.000654 0.000049 0.000026 0.000046 0.000980 0.000080 0.000002 0.000027 0.000002 0.000022 0.000002 BK21A age spectrum 0.000490 0.000002 0.000192
0.000086
0.000061 0.000076 0.000081 0.000050 0.000065
0.000002 0.000002 0.000002 0.000105 0.001058
0.011792 0.006230 0.005512 0.004377 0.003415 0.006352 0.001652 0.005067
0.000161 0.000111 0.001032 0.005127 0.004612 0.001928 0.001021 0.004103
0.000322 0.000434 0.000254 0.000373
0.000820 0.000724 0.000423
0.000821
0.000602 0.001939 0.001252 0.001827 0.002316 0.002201 0.000949
0.973922 0.548332 0.452996 0.382368 0.291488 0.536643 0.137284 0.432808
0.011407 0.009516 0.089212 0.425172 0.389430 0.160388 0.089340 0.350976
0.028395 0.036074 0.021331 0,031575 0.068610 0.063145 0.035239 0.071514
0.050376 0.155976 0.101022 0.152322 0,184141 0.178189 0.079223
8.372794 2.389034 2.037368 1.691693 1.302205 2.275960 0,556490 1.657361
0.298540 0.165896 0.259008 0.447379 0.393628 0. t 63667 0.141548 0.364035
0.043715 0.043225 0.023741 0.044948 0.124600 0.085900 0.038036 0.082849
0.193229 0.206207 0.109721 0.161633 0.191071 0.180763 0.078179
0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6
0.8 1.4 7.2 35.1 60.6 71.1 77.0 100.0
8.0 18.1 24.1 33.0 52.3 70.0 79.9 100.0
5.6 22.9 34.1 51.0 71.4 91.2 100.0
25.93 40.53 52.59 62.77 70.53 84.82 88.48 100
1.302 0.955 0.989 0.98 0.967 0.926 1.423 1.015
0.777 0.716 0.682 0.963 1.617 0.985 0.851 1.066
0.962 0.958 0.907 0.913 0.908 0.932 0.743
BK21A single fusion Ms
Bi
0.004680 0.006704
0.400443 0.557240
0.538321 0.673589
100.0 100.0
0.958 0.932
71.3 77.9
5.65 5.50
0.15 0.11
0.012380 0.004295
1.043467 0.357160
1.298813 4.468706
100.0 100.0
0.995 0.497
80.0
5.87
0.06
4.0
2.94
0.37
0.013080 0,013873
1.129049 1.145038
2.159830 7,992747
100.0 100.0
1.017 0.587
53.2 8,4
6.00 3.47
0.08 0.08
B K I 8 single fusion 0.007065 0.000177
0.000230
0.016010 0.007363
0.000466 0.000293
0.191014 0.429655 0.269728
4.422786 10.008335 5.477904
100.0 100.0 100.0
12,225 12,283 12.243
52.8 52.7 60.3
5.50 5.52 5.51
0.04 0.02 0.02
Mean
5.51
0.02
0,000523 0.000521
0.000002 0,000645
BK210 single fusion Ms
Bi
0.000881 0.014522
0.000002 0.004796
BK21D single fusion Ms
0.003423
0.000002
Bi
0.024772
0.010005
Bi Bi Bi
0.000498 0.000427
Intensities of Ar isotopes are given in beam intensity x 10-I 3 amps; 36Atand 39Arare corrected for contribution of a calcium-derivedcomponent, 37Arand 38Ar for radioactive decay after irradiation, and 4°Ar for the contribution of nucleogenic 4°Ar; the error is given at 1 s level; % 4oAr* represents the proportion of radiogenic 4OArin the signal. See text for J factor. For the incremental heating experiments a weighted mean age has been calculated, using the inverse of 2 of the 40 Ar* /39 ArK as the weighting factor. A standard error of the mean is calculated by dividing the standard deviation by (n) 1/2 where n the variance (l/s) is the number of steps contributing to the mean age. The ages were calculated using a decay constant of 5.543 x 10-l° a-1 for 4°K.
40AR--39AR DATING OF SERAM OPHIOLITE OBDUCTION
very sensitive to relatively small disturbances in the data used for correction of the signal such as the system blank values or the beam intensity of 36mr. However, despite some obvious outliers, which were attributed to experimental procedure rather than to sample inhomogeneity, it has been shown (Table 1, Fig. 3) that resetting of the Ar system in both muscovite and biotite was complete, and that chloritized biotite crystals contain large amounts of atmospheric Ar, precluding calculation of their precise ages (Table 1).
12 Age (Ma)
lo 8
L
6 4 2
The use of 4°Ar/39Ar data of micas for thermochronology is well established (McDougall & Harrison 1988; see Spear 1993 for recent references); it requires, however, the determination or the assumption of the argon closure temperatures (To) for muscovite and biotite. Based on laboratory and field calibrations, for conditions of slow to intermediate cooling rates, 350°C has been assumed as Tc for muscovite (Purdy & J~iger 1976; Wagner et al. 1977) and 280°C for biotite with low to intermediate annite contents (Harrison et al. 1985). Tc exceeding 400°C has been estimated for muscovite of the eastern Alps and 320°C for coexisting biotite (Blanckenburg et al. 1989). Typically, TcS for coexisting muscovite and biotite from various areas differ by 70-90°C (Blanckenburg et al. 1989; Cosca et al. 1991). The crystals used in this study were quite large and their cooling went fast. Since these factors both tend to shift TcS to higher values (Dodson 1973), we argue that the TcS for the Kaibobo micas should be at the high end of the range estimated.
m
8
/-!
4
Ms O, 5 . 8 3 + 0.26 (Vla 2
8
6
l
4
,,
I"-1 I I
|
|
Ms D, 5 . 9 0 + 0 . 1 4
The muscovite crystals yielded flat age spectra of 5.65-6.0 Ma. Muscovite BK21A gave a larger error (6.65 _+ 1.2 Ma, Fig. 3a), but is equivalent within its analytical uncertainty. One wellpreserved biotite crystal yielded an undisturbed age spectrum, indicating an age of 5.4 Ma (Fig. 3d), comparable to muscovite, but chloritized biotite crystals contained substantially more atmospheric argon and showed increased scatter in apparent ages and much lower integrated ages (Fig. 3e, Table 1). As no significant inhomogeneities were found in the age spectra, the integrated ages and the single fusion results may be considered equivalent. However, the higher beam intensities obtained for the single fusion experiments should yield the most precise ages (Table 1). Muscovite ages vary from 6-5.65 Ma and fresh biotite from the sole yielded an age of 5.5 Ma. Three replicate analyses of fresh biotite crystals from granite BK18 gave 5.51 __+0.02 Ma (weighted mean).
Closure temperatures
Ms A, 6 . 6 :I: 1 . 2 M a
0
6
459
Ma
2
8 Bt A, 5.41 _+ 0 . 4 8 Ma 6
4
2
4
Obduction exhumation history and its P - T - t modelling
i
As shown in Fig. 4, results of thermobarometry reveal distinct stages in the uplift of the Kaibobo ophiolite (Linthout & Helmers 1994). Early recrystallizations indicate isobaric cooling of the lherzolite from c. 1050 to c. 950°C at a depth of 25 km (boxes A-B); then the oceanic lithosphere was obducted, undergoing isothermal exhumation (B-E). By the end of obduction emplacement, the
B t D, 2 . 9 1 + 0 . 2 1 Ma
e o
2o
40
6o-
80
io, %39Ar
Fig. 3 a-e. Age spectra of single grains of muscovite (Ms) and biotite (Bt) from samples BK21A-O.
460
K. LINTHOUT ET AL. '
'
'
,]]:!:~!
,i[!~1
,
,
iiiiii~ii ~iiii~i~iil
5.5 ~ i
,
,
,
,
,
,
,
G
i
. . . . . . . . .
i
.
.
.
.
.
.
.
.
i
.
.
.
.
.
III
:il;i~i~,_ _
iii! ii!i!.....
7 Ma
iiiiiiii iiiiiiiiiiii
._.x O
r~
Q. (D
8 Ma
a,,,I,
8.25 Ma
8.25 Ma
(n
"O = r
iiiiii!~i i!!i!!!i~i~i~
A
ii~i!i iii~iiiiil
PC
3
Q.
~i~iii iii~iiiiii !i~iii
0
iii_>iiiiii
m!
9.25 Ma
. . . .
i ~L~:i
,
400
,
. . . . .
~
,
I
,
i
i
i
,
600 temperature
,
i
,
i
I
i
800 (oc)
i
J
L
I
I
'
'
I
I
I
t
,
,
1000
Fig. 4. Best fit ID forward P-T-t model of the Kaibobo obduction and exhumation path. For conditions and parameters see text. Shaded boxes for the ophiolite and open boxes for the sole after thermobarometry by Linthout & Helmers (1994); shaded bars mark the thermochronologybased on Ar dating of micas from the sole and obductioninduced granite. Filled and open circles mark modelled timing for ophiolite and sole, respectively.Model results indicate a 2 Ma period of post-emplacementexhumation between thermal relaxation of the ophiolite and sole and the closure temperature of muscovite for the Ar system (c. 400°C). However, cooling by granite melt displacement, serpentinizationand hydrothermal activity may allow a shorter period (c. 1.5 Ma) of exhumation.
sole and the juxtaposed hot ophiolite equilibrated to a temperature of c. 740°C (E, I-II). After burial by the obducting slab at a depth of 14 km, the'sole was lifted again, following a joint post-emplacement exhumation trajectory with the ophiolite, characterized by fast uplift at first (II-F) and by more isobaric cooling later (F-III-G). Age data for the early part of the ophiolite's history are not available but some time constraints for this period of emplacement may be inferred by considering that obduction took place in a setting of collision between the Australian and the Banda Sea plates, converging at a rate 6f 70 km Ma -1. The probable slow spreading, responsible for creating the obducted lherzolite, if still active during obduction, may have had an additional effect, but emplacement faster than 90 km Ma -1 is not likely. At this rate, considering that the width of the area covered by remains of ultramafics around SW Seram (in the ENE direction of the stretching lineations) is c. 100 km (Fig. 1), the emplacement could most probably not have been accomplished in less than about 1.25 Ma. The exhumation of about 11 km (B-E) of the ophiolite during this 1.25 Ma is similar to the highest uplift rates measured around the Banda arc by the palaeontological method
of geohistory analysis (De Smet et al. 1989). Assuming a similar thickness of sediments covering the metamorphic basement as found on nearby Australian fragments Buru and Misool, both much less disturbed by Neogene thrusting (Pigram et al. 1982; Tjokrosapoetro & Budhitrisna 1982), the sole's metapelites were at a pre-obduction depth of c. 5 km and rather cool. So, obduction replaced a sedimentary pile of c. 5 km by c. 14 km of ophiolite, effectively burying the sole rocks to a 9 km deeper level, which is of the same order as that inferred for the simultaneous exhumation of the obducted slab. Clearly, a model, searching to fit both the joint exhumation trajectory (II-F-III-G) and the accurate 4°Ar/39Ar thermochronology (c. 400°C at 6 Ma and c. 320°C at 5.5 Ma), will constrain the timing of the sole's reversal from burial to uplift, thereby closely approximating the end of obduction-emplacement. For this purpose we use a 1D forward P-T-t model, adopting simple kinematic scenarios of England & Thompson (1986) and Davy & Gillet (1986), in which obduction is modelled by instantaneous saw-tooth shaped geothermal perturbation, followed by isobaric thermal relaxation and exhumation is incorporated by
40AR-39AR DATING OF SERAM OPHIOLITE OBDUCTION transient uplift assuming instantaneous erosion. The heat conduction equation is solved using a finite difference technique (grid spacing 250 m, cf. Wees et aI. 1992). Heat conduction parameters of ultramafic and continental rocks are taken as equal (conductivity = 3.1 W m -1 °C-l; diffusivity = 8.95 x 10-1 m 2 s-l); radiogenic heat production is neglected. The initial temperature conditions of the model correspond to a linear geotherm constrained by surface conditions and 880°C at 25 km (B, D). The base of the model (viz. the base of the lithosphere, 1325°C), is characterized by a constant heat flow condition of 110 mW m -1, allowing the base of the lithosphere to migrate. The dynamic and thermal processes involved in obduction-emplacement (B-C-D-E, cool sole-I-II) are too complicated to incorporate in a 1D model. Therefore, the model was simplified by adopting three stages for the obduction-emplacement in 1.25 Ma: 1 Ma of isothermal exhumation, uplifting the oceanic lithosphere from depths of 25 to 14 km; subsequent, instantaneous, isobaric overthrusting onto the sole that was already heated to 600°C; concluded by 0.25 Ma of thermal relaxation at the contact of the juxtaposed ophiolite and sole to 740°C. Although this part of the tectonic history is subject to relatively large uncertainties in age and kinematics, possible variations will not significantly influence the estimate of timing of postemplacement exhumation - the main objective. The best fit results for the latter indicate 2 Ma of decelerating post-emplacement exhumation to 400°C (Fig. 4). The fin'st million years is marked by rapid uplift of 7.5 mm a-1 (II-F). The slower uplift of 3.2 mm a-I, from 7 to 6 Ma, is followed by an isobaric cooling of c. 80°C from 6 to 5.5 Ma, in perfect accord with the thermochronology of the micas. Longer .and shorter histories were also modelled. Longer exhumation requires a more pronounced deceleration of exhumation in order to meet the PT constraints of boxes II-F. A maximum of 2.5 Ma seemed possible, by adopting a 1 Ma phase of isothermal exhumation followed by isobaric cooling. However, the cooling rates between 6 and 5.5 Ma, predicted under these assumptions, are extremely low compared to those inferred from thermochronology. Only by incorporating renewed fast uplift in the 6-5.5 Ma interval, can the cooling rate increase; for this, however, there is no supporting geological evidence. For shorter periods of exhumation during the cooling from 740 to 400°C a minimum of c. 1.5 Ma is indicated, again constrained by boxes II-F. In this case, the predicted cooling from 6 to 5.5 Ma is extremely fast (c. 150°C) and clearly not in accord with the thermochronology. However, granite-melt displacement, metamorphic reactions, serpentinization
461
and hydrothermal activities, active in the 740400°C interval, will all tend to speed the cooling. If these effects were considered in the model, they would have reduced the predicted cooling rates in the 400-300°C interval. Consequently a 1.5 Ma exhumation period seems likely. Summarizing, this new data and the results of P-T-t modelling indicate that the post-emplacement exhumation of the Kaibobo complex and its sole started between 7.5 and 8 Ma ago. The implied rates of exhumation, prevailing around 7.5 Ma ago, at 7 . 5 - 1 0 m m a -1, are among the highest reported from the Banda arc (cf. De Smet et al. 1989).
Site of obduction of the Kaibobo ultramafic complex Undoing westward migration in the SW Pacific regime back-tracks Kaibobo in c. 4 Ma to the junction of the 'NNW transform' and the TararaAiduna fault zone, viz. the area where it was being rotated in the transition from the Australian into the SW Pacific regime (Fig. 1). The remaining 3.5 to 4 Ma spent in Australian motion brings Kaibobo back, over between 245 and 280 kin, to where the sole started its uplift 7.5 to 8 Ma ago. Thus the inferred locus of obduction is NE of the current position of Tanimbar, on the bathymetric incision in the outer Banda arc, marking the 'NNW transform' between the Banda Sea and Australian plate (Fig. 1). It is noteworthy that 600 km to the WSW, on N Timor and also on lesser islands between Timor and Tanimbar, ultramafites associated with highgrade metamorphic rocks occur (Bemmelen 1949; Pertamina-Beicip 1982). The ultramafic rocks obducted on the N coast of Timor, near Atapupu and Dili, are of particular interest since the highgrade metamorphic rocks in the sole near Dill give 4°hr/39Ar ages of 8 and 5.4 Ma for hornblende and muscovite, respectively (Berry & MeY)ougall 1986). Considering that at Atapupu cooling went slower since the overriding hot slab was c. 6 km thicker (Helmers et al. 1989) we have remarkably similar time settings of obduction for the N Timor and Kaibobo ultramafic complexes. Moreover, like in Kaibobo, the Atapupu ultramafites are lherzolitic and net-veined by coarse-grained hornblende gabbro (Helmers et al. 1989). Lherzolitic composition, association with a sole of metamorphic continental crust, rather than one of oceanic crustal descent, abundance of dykes and sills, antigorite as the serpentine mineral formed in a tectonic setting together with tremolite and talc, and absence of chromite pods, as observed in Kaibobo as well as in Atapupu, are all typical features of the lherzolite
462
K. LINTHOUT ET AL.
ophiolite type, corresponding to very slow spreading, as may be achieved by juvenile spreading in a transform environment (Boudier & Nicolas 1985). On N Timor also, there are Upper Miocene basalts. For pillow basalts of the Manamas Formation, west of and adjacent to the Atapupu complex, K/At dating suggests a minimum age of 6 Ma (Abbott & Chamalaun 1981); a palaeontological age was tentatively estimated at 7 Ma (Carter et al. 1976); K/At data provide an age of 10 Ma for Manamas basalts and dioritic dykes (H. Bellon, pers. comm. 1995). Trace element analyses of Manamas basalts indicate compositions close to MORB, not far from OIB (Vroon 1992).
Early Miocene spreading in the southern Banda Sea From the foregoing it appears that during the Early Miocene (c. 18 Ma ago, i.e. 10 Ma before obduction) one or several spreading basins were being formed in the ocean floor in the southem Banda Sea, to the south of the current volcanic arc. The origin and age of the Banda Sea have been a matter of debate. Hamilton (1979) considered the south Banda Sea an active marginal basin; according to R6hault et al. (1994a, b), 6-9 Ma old volcanic rocks, recently dredged from the north
Banda Sea and from various locations on ridges in the Banda Sea (north of the volcanic arc) have geochemical signatures suggesting a process of backarc opening. Bowin et al. (1980) and Lee & McCabe (1986), on the other hand, suggested the Banda Sea is a trapped, Cretaceous-Paleogene piece of Indian Ocean crust. Accepting the evidence in favour of an old oceanic crust, i.e. low heat flow, depth of basement and interpretation of magnetic lineations, but also considering the evidence from the obduction complexes of Seram and Timor, the authors cannot dismiss r e n e w e d spreading during the Early Miocene in the Banda Sea, south of the later developed volcanic arc. Audley-Charles et al. (1979) noted that the geology of Seram is in many respects the mirror image of the geology of Timor. The results here indicate that the apparent mirror-image relationship between both islands on opposite sides of the Banda Sea is the fortuitous outcome of a series of Late Neogene plate-tectonic translations and rotations of plate fragments caused by the interplay of three major plates. We are grateful for stimulating and constructive reviews by H. Bellon, S. Kelly and J.-P. R6hault. This is publication NSG 950409 of the Netherlands Research School of Sedimentology. Reactions to
[email protected] (fax 020-6462457) are welcome.
References ABBOTT,M. J. & CHAMALAUN,E H. 1981. Geochronology of some Banda Arc volcanics. In: BARBER,A. J. & WIRYOSUJONO,S. (eds) The Geology and Tectonics of Eastern Indonesia. Indonesian Geological Research and Development Centre, Special Publication, 2, 253-268. AUDLEY-CHARLES,M. G., CARTER,D. J., BARBER,A. J., NORVICK, M. S. & TJOKROSAPOETRO,S. 1979. Reinterpretation of the geology of Seram: implications for the Banda Arcs and northern Australia. Journal of the Geological Society, London, 136, 547-568. BECKINSALE,R. D. & NAKAPADUNGRAT,S. 1979. A Late Miocene K-Ar age for the lavas of Pulau Kelang and Seram, Indonesia. In: UYEDA,S., MURPHY,R. W. & KOBAYASHI, K. (eds) Geodynamics of the Western Pacific. Advances in Earth and Planetary Sciences, 6, 199-201. BEMMELEN,R. W. VAN 1949. The Geology of Indonesia. Government Printing Office, Nijhoff, The Hague. BERRY, R. E & McDOUGALL, I. 1986. Interpretation of 4°Ar/39Ar and K/At dating evidence from the Aileu Formation, East Timor, Indonesia. Chemical Geology, 59, 43-58. BLANCKENBURG, F. VON, VILLA, I. M., BAUR, H., MORTEANI, G. & STEIGER, R. H. 1989. Time calibration of a PT-path from the Western Tauern Window, Eastem Alps: the problem of closure temperatures. Contributions to Mineralogy and Petrology, 101, 1-11.
BOUDIER, F. & NICOLAS, A. 1985. Harzburgite and lherzolite subtypes in ophiolitic and oceanic environments. Earth and Planetary Science Letters, 76, 84-92. BOWIN, C. O., PURDY, G. M., LAWYER,L., HARTONO,H. M. S. & JEZEK, E 1980. Arc-continent collision in the Banda Sea region. APPG Bulletin, 64, 868-915. CARTER,D. J., AUDLEY-CHARLES,M. G. & BARBER,A. J. 1976. Stratigraphical analysis of island arccontinental margin collision in eastern Indonesia. Journal of the Geological Society London, 132, 179-198. COSCA, M. A., SUTTER, J. F. & ESSENE, E. J. 1991. Cooling and inferred uplift/erosion history of the Grenville Orogen, Ontario: constraints from 4°Ar/39Ar thermochronology. Tectonophysics, 10, 959-977. DAVY, PH. & PH. GILLET. 1986. The stacking of thrust slices in collision zones and its thermal consequences. Tectonics, 5, 931-929. DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S. 1990. Current plate motions, Geophysical Journal International, 101,425-478. DE SMET,M. E. M., FORTUtN,A. R., TJOKROSAPOETRO,S. & VAN HINTE, J. E. 1989. Cenozoic vertical movements of non volcanic islands in the Banda Arc area. Netherlands Journal of Sea Research, 24, 263-275.
40AR--39AR DATING OF SERAM OPHIOLITE OBDUCTION DODSON, M. H. 1973. Closure temperature in cooling geochronological and petrological systems. Contributions to Mineralogy and Petrology, 40, 259-274. ENGLAND, P. C. & THOMPSON, A. B. 1986. Pressuretemperature-time paths of regional metamorphism I. Heat transfer during the evolution of regions of thickened crust. Journal of Petrology, 25, 894-928 HAILE, N. S. 1979. Palaeomagnetic evidence for the rotation of Seram, Indonesia. In: UYEDA, S., MURPHY, R. W. & KOBAYASHI,K. (eds) Geodynamics of the Western Pacific. Advances in Earth and Planetary Sciences, 6, 191-198. HAMILTON,W. 1979. Tectonics of the Indonesian Region. United States Geological Survey Professional Paper, 1078. HARRISON, T. M., DUNCAN, I. & McDOUGALL, I. 1985. Diffusion of 4oAr in biotite: Temperature, pressure and compositional effects. Geochimica et Cosmochimica Acta, 49, 2461-2468. HELMERS, H., SOPAHELUAKAN,J., TJOKROSAPOETRO,S. & SURYA NILA, E. 1989. High-grade metamorphism related to peridotite emplacement near Atapupu, Timor, with reference to the Kaibobo peridotite mass on Seram, Indonesia. Netherlands Journal of Sea Research, 24, 357-371. JOHNSTON, C. R. 1981. A review of Timor tectonics, with implications for the development of the Banda Arc. In: BARBER, A. J. & WIRYOSUJONO, S. (eds.) The Geology and Tectonics of Eastern Indonesia. Indonesian Geological Research and Development Centre, Special Publication, 2, 199-216. LEE, C. S. & MCCABE, R. 1986. The Banda-Celebes-Sula basin: a trapped piece of Cretaceous-Eocene oceanic crust? Nature, 322, 51-54. L1NTHOUT, K., & HELMERS, H. 1994. Pliocene obducted, rotated and migrated ultramafic rocks and obduction-induced anatectic granite, SW Seram and Ambon, Eastern Indonesia. Journal of Southeast Asian Earth Sciences, 9, 95-109. ., HELMERS,H. & ANDRIESSEN,P. A. M. 1991. Dextral strike-slip in central Seram and 3-4. 5 Ma Rb/Sr ages in pre-Triassic metamorphics related to Early Pliocene counterclockwise rotation of the BuruSeram microplate (E Indonesia). Journal of Southeast Asian Earth Sciences, 6, 335-342. - - , SOPAHELUWAKAN,J. & SURYANILA, E. 1989. Metamorphic complexes in Buru and Seram, northern Banda Arc. Netherlands Journal of Sea Research, 24, 345-365. McDOUGALL,I. & HARRISON,T. M. 1988. Geochronology and Thermochronology by the 4°Ar/e9Ar Method. Oxford University Press, New York. PERTAMINA-BEICIP 1982. Geological Map of Eastern Indonesia, scale 1: 2,000,000. PmRAM, C. J. & PANGGABEAN,H. 1984. Rifting of the eastern margin of the Australian continent and the origin of some microcontinents in Indonesia. Tectonophysics, 107, 331-353. , CHALLINOR,A. B., HASSIBUAN,E, RUSMANA,E. & HARTONO, U. 1982. Lithostratigraphy of the Misool archipelago, Irian Jaya, Indonesia. Geologie en Mijnbouw, 61, 265-279.
463
PRIEM, H. N. A., ANDRIESSEN,P. A. M., BOELRIJK, N. A. I. M., HEBEDA, E. H., HUTCHISON, C. S., VERDURMEN,E. A. TH. & VERSCHURE,R. H. 1978. Isotopic evidence for a middle to late Pliocene age of the cordierite granite on Ambon, Indonesia. Geologie en Mijnbouw, 57, 441-443. PURDY, J. W. • JAGER, E. 1976. K-At ages of on rockforming minerals from the Central Alps. Institute of Geology and Mineralogy Padova University Memoirs, 30. RANGIN, C., JOLIVET, L. t~ PUBELLIER,M. 1990. A simple model for the tectonic evolution of southeast Asia and Indonesia region for the past 43 m.y. Bulletin de la Socidt( gdologique de France (8), XXVI, 889-905. Rt~HAULT, J.-e., MAURY, R. C., BELLON, H., SARMILI, L., BURHANUDDIN,S., ET AL. 1994a. La Mer de Banda Nord (Indon6sie): un bassin arribre-arc du Miocene sup~rieur. Comptes rendus de l'Acaddmie Sciences Paris, 318, s6rie II, 969-976. , VILLENEUVE, M., HONTHAAS, C., BELLON, H., MALOD, J.-A., ET AL. 1994b. New geological sampling along a transect across the South East Banda Sea Basin (Indonesia). Tectonic Evolution of SE Asia Abstracts, December 7-8 1994, The Geological Society London, 49. SMITH, D. E. & NINE OTHERS. 1990. Tectonic motion and deformation from satellite laser ranging to LAGEOS. Journal of Geophysical Research, 95, 22 013-22 041. SOPAHELUWAKAN, J., LINTHOUT, K., HELMERS, H. & PERMANA, H. 1992. Peridotite-metamorphite relation in West Seram: Constraints to the vertical movements of the North Banda Arc. Proceedings of the Indonesian Association of Geology, XXI Annual Scientific Meeting, Yogyakarta, December 7-10 1992, 599-609. SPEAR, E S. 1993. Metamorphic phase equilibria and pressure-temperature-time paths, Mineralogical Society of America, Washington D. C. TJOKROSAPOETRO, S. & BUDHITRISNA,T. 1982. Geology and tectonics of the northern Banda Arc. Bulletin Indonesian Geological Research and Development Centre, 6, 1-17. TOMMASINI, S., DAVIES, G. R., STAUDIGEL,H., HELMERS, H., MANETTI, P. & POLl, P. 1994. Sr, Nd, and Pb isotope data of crustal melts from the island of Seram, Indonesia: evidence for disequilibrium melting during anatexis. US Geological Survey Circular 1107, Abstracts of the Eighth International Conference on Geochronology, Cosmochronology and Isotope Geology, 325. VALK, W. 1945. Contributions to the geology of West Seram. PhD Thesis, University of Utrecht, De Bussy, Amsterdam. VROON, P. Z. 1992. Subduction of continental material in the Banda Arc, Eastern Indonesia. PhD Thesis, University of Utrecht, Geologia Ultraiectina, 90. WAGNER, G. A., REIMER, G. M. & J.~GER, E. 1977. Cooling ages derived by apatite fission track, mica Rb/Sr and K/Ar dating: the uplift and cooling history of the Central Alps. Institute of Geology and Mineralogy, Padova University Memoirs, 30, 1-27.
464
K. LINTHOUT ET AL.
WEES, J. D. VAN, DE JONG, K., & CLOETINGH, S. 1992. Two-dimensional P - T - t modelling and the dynamics of extension and inversion in the Betic Zone (SE Spain). Tectonophysics, 203, 305-324.
WIJBRANS, J. R., PRINGLE, M. S., KOPPERS, A. A. E & SCHEVEERS, R. 1995. Argon geochronology of small samples by laser heating. Proceedings van de Koninklijke Nederlandse Academie van Wetenschappen, 98 (2), 185-218.
Correlation of the Salawati and Tomori Basins, eastern Indonesia: a constraint on left-lateral displacements of the Sorong fault zone TIM
R. C H A R L T O N
Department of Geology, Royal Holloway University of London, Egham, Surrey TW20 OEX, UK Present address: Ridge House, 1 St. Omer Ridge, Guildford, Surrey GU1 2DD, UK Abstract: The Salawati Basin of western New Guinea and the Tomori Basin of eastern Sulawesi, Indonesia, are two sedimentary basins located either side of the main fault strands in the Sorong fault zone strike-slip system. It is suggested that prior to displacement on the Sorong system the two formed a single sedimentary basin. Movement on the Sorong system occurred largely during the latest Miocene-Quaternary, contemporaneous with deposition of a clastic sedimentary succession. An older basinal sequence, essentially Miocene in age, is composed predominantly of carbonate sediments, and this may have formed part of the foreland basin sequence related to the east Sulawesi orogenic belt. Correlation of the Salawati and Tomori Basins implies a leftlateral displacement of about 900 km on part of the Sorong fault zone.
a broad zone of regional left-lateral shear usually known as the Sorong fault zone (Figs 1 & 2). The Sorong fault zone extends between the Bird's Head region of Irian Jaya in the east and the island of Sulawesi in the west, a distance in excess of
The Salawati Basin of northwestern Irian Jaya (Indonesian western New Guinea) is the most important petroleum producing basin in eastern Indonesia. The northern edge of the basin is truncated by the Sorong fault, one fault strand in
Continental shelves:
120°E
130°E
140°E
Eurasian
PACIFIC OCEAN Australian M.S.C.Z. Molucca Sea Collision Zone
III
~...
/ .4":
Sorong
,, r~ NEW GUINEA
1BUTON
•
%
JAVA
d
"%°°s -..-....,~
-'~
'
.
°
.
°
.
"
°
°
.
.
,
°
•
_~
At
/
/.'.'.'.'.'~ /
I
8~b
. . . . . . .
INDIANOCEAN
~ .
" ,~
~ I
/
. . . . . "
I .
.
.
.
AUSTRALIA 7_
""
:
EU
:1
Fig. 1. Regional tectonic setting of the Sorong fault zone. Inset: relative convergence vectors for the Eurasian (EU), Indo-Australian (AU) and Pacific (PA) plates in the eastern Indonesia region. From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 465-481.
465
466
T.R. CHARLTON
Blank onshore: Post-orogenic basins
[-~ ~
Terranes of non-Australian origin ('allochthon')
~
0 m
200 m
m
ang ai islands ~'kj
Platform Imbricated Terranes of 'Australian' stratigraphic affinity Basement
400 km m
m
I1~
"~
• "" "
o.
S
....... ........'~::::': W T .......
~
. Inan Jaya Bird's Head')
'
iso
"'~-SO Ut''
\-Su,abesi
__,7_r ' - " " .- ~- ~_ S e r a m ~ " ' ~
~
Fig. 2. Structural setting of the Salawati and Tomori Basins and the Sorong fault zone. Bathymetry shaded at 1000 rn intervals adapted from R6hault et al. (1991), Smith & Silver (1991), Milson et al. (1992) and Hamilton (1979).
1500 km. The total offset on this fault system is poorly constrained, but most estimates are in the range of many hundreds to several thousands of kilometres (e.g. Visser & Hermes 1962; Tjia 1973; Pigott e t al. 1982; Pigram & Panggabean 1984). This poor constraint results primarily from the absence of unambiguous displacement markers either side of the fault zone. The main aim of this paper is to suggest that prior to the development of the Sorong fault system the Tomori Basin, located off the east coast of Sulawesi, formed the northern half of a single sedimentary basin with the Salawati Basin. This implies a displacement on the Sorong system of approximately 900 km. This interpretation has important implications for regional tectonic evolution and hydrocarbon exploration in eastern Indonesia, both of which will be considered in this paper.
Regional tectonic setting Eastern Indonesia is situated in the zone of interaction between three of the Earth's major tectonic plates (Fig. 1): the Eurasian, Indo-Australian and Pacific plates (the Philippine Sea and Caroline plates, which separate the Pacific plate proper from
the eastern Indonesia region, have only small motion vectors relative to the Pacific in this region, and for the purposes of the present paper can be treated as sub-plates of the main Pacific plate). Relative to the mantle, Eurasia is nearly static, whilst the Indo-Australian plate is moving towards the NNE, and the Pacific plate is approaching Eurasia in a WNW direction (Fig. 1). The Sorong Fault Zone is a transcurrent boundary between the Pacific and Indo-Australian plates, with a WSW-ENE orientation sub-parallel to the relative movement vector between these two plates. One important effect of the Sorong transcurrent system has been to strip continental terranes from the northern margin of the Australian continental block and transfer them westward as elements of the Pacific plate. Subsequently these allochthonous terranes have collided with and been accreted into the western margin of Eurasia. Such displaced continental terranes with Australian stratigraphic affinity include Buton, southeast and east Sulawesi, the Banggai-Sula block and southwest Obi island (Figs 1 & 2; e.g. Pigram & Panggabean 1984). The Tomori Basin, one of the primary focuses of the present paper, is a successor basin developed on the Banggai-Sula continental fragment.
SALAWATI& TOMORI BASIN CORRELATION The geology of the Sorong fault zone is relatively poorly known (Hall et al. 1991), and various lineaments have been proposed as the main fault strands in this system. The fault pattern preferred by the present author is shown in Fig. 2. In the east, the Sorong fault sensu strictu forms a well-defined east-west fault zone some 10-20 km broad through the Bird's Head. West of the Bird's Head, the author interprets the main fault strands to be the direct westward extension of the Sorong fault south of Obi island to immediately east of Sulabesi island (here named the Sorong-Sulabesi fault), and the South Sula fault (the western half of the South Sula-Sorong fault of Hamilton 1979) west of Sulabesi island. Present-day seismicity in the Sorong fault zone is primarily associated with the Sorong-Sulabesi fault (Kertapati et al. 1992). At its western end, this seismically active belt connects via Sulabesi island southeastwards to a NW-SE trending belt of shallow level seismicity through Seram island, possibly the right-lateral fault system identified by Linthout et al. (1991); and northward into seismicity associated with the eastern margin of the Molucca Sea collision zone (Fig. 1). The island of Sulabesi has a structure controlled by N-S trending normal faults, and marks a right step between the presently active and inactive strands of the main Sorong fault system. Sulabesi has probably been rotated some 90 ° anticlockwise relative to the main Sula islands to the north as a result of transtension. Other fault strands, such as the North Sula-Sorong fault and the Molucca-Sorong fault proposed by Hamilton (1979) and others may or may not have had significant displacements in the past, but are not active at the present day. These faults, however, lie north of the Banggai-Sula displaced terrane which is the primary concern of the present paper, and any movement on these faults would be additional to the strike-slip offsets interpreted here. A number of estimates have been made as to when the Sorong fault zone became an active feature. These include the Oligocene (Pigott et al. 1982), Early Miocene c. 25 Ma (Hermes 1968), Early Miocene (Tjia 1973), Early-Mid Miocene (Hamilton 1979), post Mid Miocene (Visser & Hermes 1962), Early Pliocene (Dow & Sukamto 1984) and mid Pliocene (Froidevaux 1977). As will be discussed in more detail subsequently, it is suggested that the fault system has been active since the Late Miocene (c. 6-8 Ma). In the following sections, the geology of the Salawati and Tomori basins will be outlined, and then the points of similarity that lead to the conclusion of a common origin will be discussed. This will be followed by discussion of the regional tectonic and hydrocarbon exploration implications of this interpretation.
467
The Salawati Basin The Salawati Basin, located between the western margin of the Bird's Head and the island of Misool (Figs 2 & 3), is essentially a Neogene feature, although the older basinal succession is concordant with an underlying Palaeogene shelf succession. Two distinct phases of basinal development are apparent: a Miocene phase in which carbonate sedimentary environments predominated, and a Pliocene-Recent phase dominated by clastic sedimentation. Structural setting
Total Tertiary isopachs for the Salawati Basin define a NW-SE trending trough which is truncated in the north by the Sorong-Sulabesi fault (Fig. 3). The eastern margin of the basin is found in the island of Salawati and the northwest corner of the Bird's Head, whilst the western margin is marked by the island of Misool. As mentioned above, the basin is readily divisible into Miocene and Plio-Quaternary sub-basins. The PlioQuaternary basinal depocentre is situated midway between Salawati and Misool islands, whilst the Miocene depocentre was located beneath the present-day Sele Straits separating Salawati island from the Bird's Head (Fig. 4). The Miocene carbonate basin beneath Salawati island and the NW corner of the Bird's Head had a semi-enclosed horseshoe shape open to the NW, with deep-water environments surrounded to the NE, east and south by shallower carbonate shelf (e.g. Vincelette & Soepardjadi 1976; GibsonRobinson et al. 1990; see Fig. 10). No significant faulting is recognized in association with this earlier basinal phase, which seems to have formed by essentially passive downwarping after deposition of the shallow marine Oligocene Sirga Formation (described subsequently). The main fault-related structural elements of the presentday basin developed during accumulation of the younger, clastic basinal sequence. According to Cockcroft et al. (1984), regional tilting of the basin occurred after deposition of the Miocene Kais and Klasafet formations, and this was followed during the mid Pliocene-Pleistocene by development of N-S trending normal faults. E-W trending folds, found in particular near the Sorong fault, are also of Pliocene age (Gibson-Robinson et al. 1990). Pre- Tertiary stratigraphy
Pre-Tertiary stratigraphy, ranging in age from Palaeozoic to Upper Cretaceous, is exposed on the flanks of the Salawati Basin in the northern Bird's Head of Irian Jaya and in Misool island (Fig. 5).
T.R. CHARLTON
468
130°E 5O km
0 I
131°E
Batanta,
I
I
I
1os
Kofiau
rian Salawati
. . . . . .
*..
SALAWATi\ BASIN \ ,
~ Misool
~ -
Plio-Quaternary Miocene
~
Mesozoic-ealaeogene
I-~
Tertiaryvolcanics on continental crust Non-continentalelements
11--~Contni ent basaelment
Fig. 3. The Salawati Basin. Tertiary isopach contours are in km, interpolated from Imperial (feet) contouring by Redmond & Koesoemadinata (1976).
In the northern Bird's Head, continental basement comprises Siluro-Devonian shales and turbidites (the Kemum Formation) which were deformed, metamorphosed and intruded by granitic rocks (e.g. the Melaiurna Granite) during the Late DevonianEarly Carboniferous (Visser & Hermes 1962; Pieters et al. 1983). Lower Carboniferous(?) synorogenic sediments of the Aisajur Formation are folded but not metamorphosed. The mid Palaeozoic basement is succeeded unconformably by a platform sequence commencing with the Upper Carboniferous-Permian Aifam Group. Further granitic intrusion into the Kemum basement occurred during the PermoTriassic (the Anggi Granite Suite), contemporaneous with the deposition of the Aifam Group and the succeeding Tipuma Formation, which is a red bed sequence of Triassic-Lower Jurassic age. In Irian Jaya generally the Tipuma Formation is succeeded by the Kembelangan Group of upper Lower Jurassic-Upper Cretaceous age. In the northern Bird's Head the only element of the Kembelangan Group recognized at outcrop is the Cretaceous Jass Formation (Pigram & Sukanta 1982). This comprises calcareous and micaceous mudstone with minor limestone and sandstone.
A third phase of igneous intrusion also occurred during the Late Cretaceous, with granitic core samples from the base of the Salawati K-IX well yielding Campanian-Maastrichtian K-Ar radiometric ages (71 _+ 1.5 Ma from biotite and 79.3_ 1.1 Ma from amphibole: Lunt & Djaafar 1991). In Misool, metaturbidites apparently similar to the Kemum Formation of the Bird's Head form the oldest recognized stratigraphic unit (the Ligu Metamorphics: Pigram et al. 1982). Their age, however, is uncertain, and Siluro-Devonian (Pigram et al. 1982), Permian (Froidevaux 1974) and Triassic (Simbolon et al. 1984) ages have been suggested for the sedimentary protolith. The oldest unmetamorphosed sequences in Misool are Triassic turbidites of the Keskain Formation, and partly contemporaneous Upper Triassic reefal limestones of the Bogal Limestone Formation. The Bogal Limestone is overlain unconformably by LowerMiddle Jurassic shales (Yefbie Shale), and these are in turn succeeded by Upper Jurassic shelf carbonates of the Ligu Formation and contemporaneous(?) shales of the Lelinta Shale Formation. These pass up conformably into Cretaceous bathyal limestones of the Facet Limestone Group (Gamta
u
Une
Misool
~
/
,
TBO-lX
I
~
~
..........
-'
' :
:-":
i
,
-.,
!
-.-
J e f Lio-1
-
~:~..........
~
i
~'.~ ,~,~
" ""
~P,iot - . - . = .-] Quaternary
::::::P:P:P:]:::::::::::::::::::::::::::::" ~°~'.,.~" r-.'~ ~ ° ~ ' "
iiiiiiiiiii:;......... ........
•
I Klamonol Klawilis
Klamogun Klamumuk
Bird's Head
I Miocenebasina' facies
......................................... " " n
......
in Salawati-N-1X
~
-...-
Miocene deepwater limestone facies (Klamogun Fm.)
"-
....................
Sele Straits
,'7. , ,"7 ,'7",72. :'7,,~, ~ :'7,,'7",",,, ,'~, 7":, :'7,,,, ,~ i " ~..:.::::i:iiii::!iiii~!~:i::::: .....
K-lX
Salawati island
NE'W
Change of direction
Fig. 4. Cross-section through the Salawati Basin showing Tertiary geology and pre-Tertiary subcrop (eastern part after Vincelette & Soepardjadi 1976). The Miocene Salawati Basin had a depocentre beneath the present-day Sele Straits, distinct from the main Plio-Quatemary depocentre between Salawati and Misool islands.
Miocene she,f carbonate with pinnacle reef
i
=u Km
q""~-~..-.-... - . - - -
. . . . . . . . . . . . . . . . . . . .
Vertical exaggeration x5
,
volcanics I
""
(TBF-lX)
lililiiiiiiiiiiiii!]Pa,aeogen e [ ~ ~
4 km
3
2
SW
4~
0 Z
©
:Z
E/3
;>
©
©
R~
>
470
T.R. CHARLTON
Salawati Basin
SYSTEM/ SERIES Pliocene >_ n" ~ F-rr
Miocene
•
~__
Oligocene
Eocene
North
Bird's Head
Misool
' • ' , - " - S e e Fiqure 6 ' ' - ' I
I
b
Ooe~taEat. I ~ ~ , h-..--~
O W O < I-
•
~
, , Limestone,
I~aag
~
~
~
~
i
i
I--:-Fafanlap Formation--=~'_:-~ F_or.m_at.io_n:'~. ~...,- I . . .i. I I-i-I- -II o •-. --. - - - - :..- - - . ~,1
3 .5 Lower
~
<
Middle Lower
Z
,~ •
Upper
i~ n
I
I
~
- --Yefbie S h a l e - ' - - - - • • . .'I'~R:~'I'~':~'I'~':~R"~'-~"
.
.
.
.
.
.
.
.
.
.
i
i
i
- ~
i
- -
.
" . . . . . . . . . . .
i
,
•
"=
,
•
"-
• " •
•
- _A!nlm • : - - E o r m a ~• ~ "
'e~ . . . . . . . . .
.
.
.
.
~
~
Upper
.<_"-_'-Aimau Formation--.:_.
.~.._?~_~1Aisajur
......
'
Formation - !
. . . . . . . . . . . .
-I+.B.uo
~ T o k ~ a l"a q"~ " ___.
+ i
(Relationship with Ligu M e t a m o r p h i c s uncertain)
- --='-R . ."' "--='-:-R:-: . . . "-" '-" R: -- - .
Lower
/Nanaka
Volcanics
'
4-
-Tetambahu/Nambo'Fms,I I I I " .............. I ........
- - -%--
(Age of m e t a m o r p h i c b a s e m e n t protolith unknown)
4-
:~.~;2;~;,~,. Formation 4-
4-
( B a s e m e n t not seen)
4-
Shelf limestone
R i
Red beds Coal
D e e p w a t e r limestone
-•
Chert
o o
Marl Conglomerate
V
Acid volcanics
""
Shale
T~ <
.
.':Bobor~g Formatior~.; ': ".'~,'-'R~" " "
: "•- T.p. .u.m a F o. .r .m• .a t i o. .n. . . . ::' - ---•-=-'-: ~,? .. I."• '6" "-' '. ' " ".'• "" "-':. ~" s I o. .n. .e.-. -. :. -. .' -. -. .-.-.-.~. .- : - ' : . .~ N o f a n ~ .r' ", .'"...'... - - , ..•LLime, ,~ • " ' " ' " " "fl" " " ' "" 4- "J"~"" K e ~ k ~ i f f Formation. "',
"-""-~t-'-r""-'-r"'=~"
J " , d& " l' -I~ormation~.~l~l & -' r ' Lk , &u~ =O k -t"
,13emu F m , . . ~ " ~ e ~ n ~ S-h~-e: ---.=Buya F o r m a t i o n ' . . " ' .... .-..r" . . . . . . ~ . . . .
+ ~ - - A i f a t FormationL9 . . . . . . . . . --E. . . . . . . . . .
O
I
~,
Lower
iiUJ
' I '
I
,--.-:-.-.- -:--I
"
0
I
i"'~="~=~'=~'~'~
F°rmati°n-I-
Ill
"'I --
n-"
-r'Tanamu
E
.
Upper
i
--~
Middle
o
i
M t__r0Lno=:
o
Lower
i
. . . . . . . . .
o Upper
.••
~ ' • . • . ' ' . . .-.. ~ S a l o d i k L / 1 ~ ~ ~F~rmatiorl ; ~ ; ; ~ S a l o d i k G r o u - p Z
~
[~~'-'---'~--~.-~-.-'-:]-IUpper
East Sulawesi .'•~" ". S u l a w e s i ' G r o u p
I
- - r - e t c . (see Figure 6 ) - - - r / I . I / '
Banggai-Sula
I Atkari Formation
I Neff~-in'ea I 'l "r I L i m e s t o n e G r o u p I .
Palaeocene ~
Tomori Basin
Siltstone
O
Sandstone/shale
SlLURO ............................. -•_ K e m u m F o r m a t i o n : . DEVONIAN
Sandstone
r~
Periods of acid igneous intrusion
Fig. 5. Simplified stratigraphy of basement terranes beneath the Salawati and Tomori Basins• Primary data sources for the Northern Bird's Head: Visser & Hermes (1962), Pieters et al. (1983) and Pigram & Sukanta (1989); Misool: Pigram et al. (1982) and Simbolon et al. (1984); Banggai-Sula: Garrard et al. (1988), Surono & Sukarna (1993) and Supandjono & Haryono (1993); East Sulawesi: Simandjuntak (1990), Corn6e et al. (1994).
Limestone and Waaf Formation). The uppermost Cretaceous is represented by shales passing upwards into shelf limestones (the Fafanlap Formation), which are succeeded conformably by Lower Palaeogene sandstones (the Daram Sandstone Formation). A shelf limestone sequence equivalent to the New Guinea Limestone Group of the Bird's Head follows conformably; this will be described in more detail in the following section.
A number of petroleum exploration wells located between the Bird's Head and Misool have encountered the pre-basinal succession beneath the Salawati Basin. Broadly comparable sequences to those found in the Bird's Head and Misool were encountered, but significant differences include the presence of Jurassic section (post-Toarcian) in equivalents of the Kembelangan Group, and the occurrence of volcanic breccias, volcaniclastic sediments and tuffs in the Upper Cretaceous.
SALAWATI & TOMORI BASIN CORRELATION
comprises a thin sequence of basal clastics succeeded by shelf carbonates. The overlying Oligocene Sirga Formation is predominantly a clastic sequence consisting of sandstone and shale with minor limestone. The Sirga Formation was deposited under neritic to epineritic conditions, and is up to about 200 m thick. The oldest part of the truly basinal succession comprises a series of contemporaneously developed Lower-Middle Miocene carbonate lithofacies, corresponding to a SE-NW transition from shelf to deep-water environments (Figs 5 & 6). The shelf carbonates are assigned to the Kais Formation, and deep-water time equivalents form the Klamogun Formation. The latter comprises interbedded marly and flaggy limestones with abundant planktonic foraminifera. The Kais Formation persisted into the Upper Miocene, but the Klamogun Formation was progressively replaced by a shallowing-upward sequence of
Tertiary stratigraphy
The older Tertiary succession of the Bird's HeadMisool region, which overlies Mesozoic and older rocks unconformably, is primarily developed in limestone facies. In the Bird's Head this sequence forms the New Guinea Limestone Group (Visser & Hermes 1962; Pieters et al. 1983). The Tertiary of this region is best known from the NW Bird's Head and Salawati island, where it has been intensively studied during hydocarbon exploration. However, the Tertiary section of Misool also forms part of the marginal Salawati Basin succession. The Tertiary succession in the western Bird's Head and Salawati island (Fig. 6) commences with the Faumai Formation, dated as upper LowerUpper Eocene (Gibson-Robinson et al. 1990), although lithological equivalents extend up into the Oligocene (Fig. 6; Pieters et al. 1983). The Faumai Formation, which is up to about 250 m thick,
~,1~ ~
~0 I ~ ) ~uaternal~
,2
~ "Jppe,
TOMORI NORTH
K~lomba Fro. o ' L giaF-m.o
1
•4 ~ .o-~," " " ^" " " = . w, - . . . . . . . . . . .
"
a.
~ ]~U~'
•8
~-
Mantawa Mbr.
•10 •~2
I
I
' ,
t
~ "O -° ~ ~ .'7 ~ .... ® ~ ~i-.Matindok • 16 ._~ ~ ~ ~ a~
-3° L24
~
'
,
'
I
[
I
I
I
I
'
,
'
'
t'''"
I
l
l
Formation
.
~
i
I Midd!eTom.ori
I
.
I Lr(Undiff.),_
~11~14244462, .~ "~
....
"
o o
o
.-'..~.-.~'.f
- .....
i ....
"'" .....
"
--
~
r
~
• J4~i~L-.
....
Formation:.-.-.I T . . . . . . . . .'11 "J;; - - - ~ - - - ' - - " ~ ' ' ' "
a
....
"''"
, x ' L ~ ."" " - - l ~ r - ~ ._ • .
s t-ormat
J
on.
J
l
I
~
' , r,K l a m o ~ u n , , : : ~ ~" ', ,,-ormation, I i " I ', ,', : , : , ,',,',,',,~1~
..........
~o ~
o Sele Con Iomerate
~[.Klas.a.fet. :q-LF~.~__~ I . Formation/ ~ ~ .t, r •.. "
I
, • • . , q ~,40 ~ -Upper Tomon -I r i Formation-'-_rFormation i
SE
m~_O_..~
I,m" . . . . . . .
I I I
I I
I I I
I I
_ ..............
.... f . . . . . . 34' 38, ~!
,
F o r m a t on
'14
•18 2C
BASIN
NW
• - ._. ... i:-.--.-.Klasaman ~.Kintorn t-ormadonl-- . . . . . . . Ubr.
I ,
l
SOUTH
m _ ,LuwukFormafiort
~
vlnahak ~
SALAWATI
I
,. . . . . . . . . . . . . . . . .
" "" "" " - - ' @ K o l 0
"
BASIN
/CENTRAL
471
;;:i
j_
, L wer--rl TTomqr .For~na.tiorq
i:: '1
: : : :; ~1
............
:
52 54
]
I Ig n e o u s r o c k s - IM i c {2241~9Ma} ~ts(~fiis{
:
.......................... r- . . . . . . . . . . . . . . . . . . . .
~ Pre-Tertiary (Silurian-U. Cretaceous)
Fig. 6. Stratigraphic comparison of the Salawati and Tomori Basins (Salawati Basin after Gibson-Robinson et al. 1990; Tomori Basin after Davies 1990, Handiwiria 1990 and Abimanyu 1990). The simpler, more laterally continuous appearance of the Tomori Basin is probably an artefact of less intense exploration in that basin.
472
T.R. CHARLTON
marl and siltstone (the Klasafet Formation). On the basinal flanks the shelfal Kais Formation is typically 350 m thick, whilst the basinal Klamogun Formation is 1159 m thick and the Klasafet Formation about 1925 m thick in their respective type sections (Visser & Hermes 1962). Pinnacle reefs developed locally near the shelf-slope break during the Middle-Upper Miocene, and these extend the typical thickness of the Kais Formation by up to 533 m (Gibson-Robinson et al. 1990). The second, clastic, basinal sequence is represented primarily by the Pliocene Klasaman Formation. This consists of interbedded sandy calcareous mudstone and argillaceous sandstone with minor conglomerate and lignite. This sequence has a maximum thickness of about 4500 m (Visser & Hermes 1962; Pieters et aI. 1983). It is succeeded unconformably by the Sele Conglomerate of Quaternary age. In Misool (Fig. 5) the Upper PalaeoceneOligocene(?) Zaag Limestone, which is approximately the equivalent of the Faumai Formation in the Bird's Head, succeeds the Daram Sandstone
......
Syn- to postorogenic sediment East Sulawesi ~ Ophiolite
conformably (Pigram et al. 1982; Simbolon et al. 1984). No equivalent of the Oligocene Sirga Formation is recognized, this time period being represented by an unconformity possibly associated with minor folding. As in the Bird's Head, the Lower Miocene is represented by contemporaneous shelf carbonates (the Openta Limestone) and a somewhat deeper water sequence (the Kasim Marl) which consists of well bedded marly and sandy limestones containing planktonic foraminifera. The Openta Limestone extends up into the Middle Miocene, but the Upper Miocene is absent. The Pliocene-Quaternary Atkari Formation overlies the Openta Limestone unconformably. In contrast to time equivalents in the Bird's Head, the Atkari Formation is predominantly a shelf carbonate sequence.
The Tomori Basin The Tomori Basin is located off the east coast of Sulawesi island (Fig. 7). Unlike the intensively explored Salawati Basin, the Tomori Basin is a
an0 a .n0:
Continental cover Continental basement
,.." Subsurface limit of oO" East Sulawesi °° Ophiolite
Fig. 7. The Tomori Basin, east Sulawesi. Structure of Tolo Gulf after Davies (1990). Basin isopachs in km adapted from Hamilton (1979) taking into account later drilling data. Onshore geology simplified from GRDC mapping.
SALAWATI t~ TOMORI BASIN CORRELATION region of frontier hydrocarbon exploration with only ten wells drilled to date. These are all located along the western margin of the basin, either on the coastal plains of east Sulawesi or immediately offshore.
Figure 8 shows two cross-sections through the western margin of the Tomori Basin; one in the north through the Minahaki well, the other in the south through the Tiaka field (Fig. 7). The northern and southern halves of the basin have very different structural styles, and also significantly different timing of deformation. The southern part of the basin is characterized by thrust faults, whilst the northern part is characterized by normal and wrench faulting (Davies 1990). There is also a marked change in the strike of thrust structures, from SW-NE in the east Sulawesi orogenic belt west of the northern basin, to nearly N-S in the south, offshore in the Tolo Gulf (Fig. 7). In the southern foldbelt, section as young as the Lower Pliocene Kintom Formation (described later) is affected by thrusting (Fig. 8b), whilst in the north there is no clear evidence for active thrusting during deposition of this unit (Fig. 8a). In the north the dip of sediments in the Kintom and Bia Formations away from the orogenic front suggests post-orogenic molasse-type deposition. The only faulting recognized on the northern cross-section is normal faulting downthrowing west toward the
Structural setting Figure 7 shows a Tertiary isopach map for the Tomori Basin based on the contouring of Hamilton (1979), but modified in the light of the more recent well data. The basin has a triangular or T-shaped map plan, probably reflecting two stages in the basinal evolution. In the north, the N E - S W elongation of the basin corresponds closely to the thrust front of the east Sulawesi fold-and-thrust belt, and it has been suggested (e.g. Davies 1990; Handiwiria 1990) that this portion of the Tomori Basin originated as the foreland basin to this orogenic belt. The NW-SE prolongation of the basin does not have such an obvious origin. It will be suggested later that this trend marks the former connection between the Tomori and Salawati Basins.
NW
o,
,
,
,
,
5km ,
473
Minahaki-1
.O.
a!
SE
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
0
.
m
' +
..I-
N~^I "'
I
~
+
0
5kin
i
i
,
+
,
+
,
+
i
...i-
,
..i-
~,,,.._~.
,
+
,
-I-
Tiaka-3
Tiaka-1
.
_.t_., ...........
~
,
+
•
+
-
..1..
¢.,t-
o._
0
b
~:~Kaiomba~n... " : ~
I
--2. ,
. . . . .
.~
0
Fig. 8. Sketch cross-sections through the westem margin of the Tomori Basin based on seismic sections figured by Davies (1990), Abimanyu (1990) and Handiwiria (1990). (a) Section through the Minahaki-1 well in the north of the basin. (b) Through the Tiaka field in the south of the basin. Note different horizontal and vertical scales. Horizontal scales were not included in the original figures and are therefore approximate. See Fig. 7 for locations.
474
T.R. CHARLTON
mountain front. This only affects rocks as young as the Miocene (and older) Salodik Group, and could be interpreted as downfaulting in the foreland associated with development of the east Sulawesi orogenic belt. This is consistent with the usual dating of orogenesis in this region as approximately Middle Miocene (e.g. Davies 1990). However, the younger, Pliocene, deformation seen in the south does not seem to have significantly affected the northern region. The timing and significance of this deformation history will be discussed in more detail later. As in the Salawati Basin, most of the variation in total stratigraphic thickness of the Tomori Basin is accounted for by the Plio-Quaternary, predominantly clastic, sequences. Thickness variations in the Miocene succession are less marked, but there is a general southward thickening from about 600-650 m in the northern wells to more than 1000 m in the south (where the foldbelt deformation makes precise estimation of true stratigraphic thickness from well sections more difficult). PreMiocene sedimentary section is absent from the north of the basin, and is only 60-90 m thick in the central basin. The pre-Miocene has not been penetrated by drilling in the south of the basin, although it is recognized in the subsurface from seismic data (R.A. Garrard, pers. comm.). Thus, as in the Salawati Basin, basinal subsidence commenced at the beginning of the Lower Miocene, at least in the northern and central portions of the Tomori Basin. Pre-basinal stratigraphy
Pre-basinal sequences forming the 'basement' of the Tomori Basin are exposed on the Banggai and Sula islands to the east, and are imbricated in the east Sulawesi fold-and-thrust belt to the west. In Banggai and Sula (Fig. 5) the oldest rocks recognized include unnamed mica schist and gneiss, lower grade metasediments (the Menanga Formation) and granitic intrusives (the Banggai Granites). The age of the sedimentary protolith is unknown, but schists on Peleng island have yielded Carboniferous radiometric ages (305 -- 6 Ma: Sukamto 1975), interpreted as the age of metamorphism. The granites have yielded Late Permian-Triassic radiometric ages (Pigram et al. 1985), and are probably cogenetic with acid volcanics (the Mangole Volcanics) of probable Triassic age (radiometric ages of 210 __ 25 Ma and 330 _ 90 Ma: Sukamto 1975). The oldest unmetamorphosed sedimentary rocks in the Banggai and Sula islands are probably reefal limestones of the Nofanini Formation. These were interpreted by Garrard et al. (1988) as Upper Triassic, although they have not been dated
accurately (Garrard, pers. comm.). The present author has also examined outcrops of the Nofanini Limestone, and they appear directly comparable to reef limestones also provisionally dated as Upper Triassic from the Tanimbar islands. The Nofanini Formation is succeeded by or is partly contemporaneous with red beds of the Bobong Formation. This is dated as Lower-Middle Jurassic by Surono & Sukarna (1993) and Supandjono & Haryono (1993), but it may range as old as Upper Triassic (Garrard et al. 1988). Marine transgression occurred during the Toarcian (e.g. Sato et al. 1978), and subsequently during the rest of the Jurassic and the early part of the Cretaceous (at least to the Hauterivian) a sequence of claystone and marl with minor limestone accumulated (the Buya Formation). Above this, apparently after a mid Cretaceous hiatus, are deep-water chalky limestones of the Tanamu Formation. This ranges in age from at least the Santonian until the Upper Palaeocene (Garrard et al. 1988). Succeeding this in the Banggai islands are Eocene-Middle Miocene shelf carbonates of the Salodik Formation. Further east, in the Sula islands, the Palaeogene is absent according to the mapping of Surono & Sukarna (1993), although Garrard (pers. comm.) has obtained Palaeogene ages from samples collected in Mangole island. Shelf limestones equivalent to the upper part of the Salodik Formation were deposited in this area during the Lower-Middle Miocene. Dolerite dykes have been recorded locally intruding the Banggai Granite and the Jurassic Bobong Formation. Their age has not as yet been better constrained than post-dating deposition of the Bobong Formation (Surono & Sukarna 1993; Supandjono & Haryono 1993). In the east Sulawesi fold-and-thrust belt, two distinct stratigraphic sequences are imbricated together: a parautochthonous sequence of continental origin with affinities to the Banggai-Sula sequence described above; and an allochthonous sequence related to the obducted east Sulawesi ophiolite. Only the parautochthonous sequence is described here (Fig. 5) as the allochthonous sequence has only limited relevance to the Tomori Basin. In the parautochthonous sequence basement has not been recognized. The oldest sedimentary rocks are of Triassic age, and comprise contemporaneous hemipelagic carbonates (the Tokala Formation) and turbiditic sandstone-shale sequences (the Bunta Formation) (Katili 1978; Rusmana et al. 1986). Recently Upper Triassic reefal facies have also been identified (Corn6e et al. 1994). These are succeeded by Lower Jurassic conglomeratic sandstones with shale and coal (the Nanaka Formation) and Middle-Upper Jurassic limestones and marls
SALAWATI• TOMORI BASIN CORRELATION (the Nambo and Tetambahu formations) (Surono et aI. 1987; Simandjuntak 1990).
In the Cretaceous two formations (the Luok and Matano Formations) have been recognized through regional geological mapping. Both consist of relatively deep-water carbonate-shale-chert sequences, and the two have been interpreted as more proximal and more distal facies within a low energy sedimentary environment (T. Simandjuntak, pers. comm.). The Luok Formation is dated as Turonian-Maastrichtian in age (Rusmana et al. 1984). It is succeeded unconformably by shelf carbonates of Lower Eocene-Lower Miocene age which, as in the Banggai islands, are assigned to the Salodik Formation (Simandjuntak 1990). In the Eocene and Oligocene, deeper water marl with limestone intercalations has been mapped as the Poh Formation (Rusmana et al. 1984; Surono et al. 1987). The Salodik and Poh formations have been placed within a re-defined Salodik Group by Handiwiria (1990). The Salodik Group is succeeded unconformably by molassic sediments of the Sulawesi Group (Abimanyu 1990). The Salodik and Sulawesi groups as recognized in the offshore Tomori Basin will be described in more detail in the following section. Pre-basinal basement has been penetrated in several exploration wells drilled through the Tomori Basin (Davies 1990; Handiwiria 1990). Three wells in the north of the basin (Matindok-1, Minahaki-1, Mantawa-1) encountered intrusive igneous rocks, whilst the Tiaka-2 well in the central portion of the basin penetrated quartz-mica schist, dated radiometrically as Middle-Late Triassic (224 _+9 Ma: Handiwiria 1990). Wells in the south of the basin did not reach basement except for Dongkala-1 which penetrated an ophiolitic sequence. This is believed to be part of the overthrust and allochthonous east Sulawesi ophiolite rather than pre-basinal basement related to the Banggai-Sula microcontinent (Davies 1990). No unmetamorphosed pre-Tertiary sedimentary rocks have yet been encountered beneath the Tomori Basin. Basinal stratigraphy
The oldest sedimentary rocks recognized through exploration drilling in the Tomori Basin are assigned to the lower member of the Tomori Formation (Fig. 6). This consists of a thin basal clastic sequence succeeded by shelf carbonates, dated as Upper Eocene-Lower Oligocene in age (Handiwiria 1990). The unconformably succeeding middle and upper members of the Tomori Formation are essentially Lower Miocene in age, although they include Upper Oligocene sediments in one well (Minahaki-1). In the north and centre
475
of the basin this sequence comprises shelfal argillaceous limestones passing up into dolomitized limestones with coal interbeds. In the south the undifferentiated Tomori Formation comprises deeper water carbonates lacking coal interbeds. The Middle Miocene Matindok Formation, comprising claystone and sandstone with minor limestone and coal, overlies the Tomori Formation conformably (Handiwiria 1990). This is in turn succeeded conformably by the essentially Upper Miocene Minahaki Formation which marks a return to shelf carbonate deposition. Pinnacle reefs occur in the north of the basin (the Mantawa Member) and claystone interbeds in the south, again suggesting deepening of the palaeoenvironment to the south. The Pliocene-Recent succession is composed predominantly of clastics assigned to the Sulawesi Group (Abimanyu 1990). The Lower Pliocene Kintom Formation, which apparently succeeds the Minahaki Formation without significant hiatus, consists of sandstone and claystone grading up into conglomerate and sandstone. This is succeeded conformably by the Upper Pliocene Bia (or Biak) Formation, composed of conglomerate and sandstone. The Quaternary Kalomba Formation, which is also composed largely of conglomerate, follows unconformably. This developed contemporaneously with the Luwuk Formation which is a reef limestone sequence.
Correlation of the Tomori and Salawati Basins The following points of correlation can be recognized between the Salawati and Tomori Basins: • Both basins are underlain by continental basement consisting of Palaeozoic(?) metasediments intruded by Carboniferous-Triassic granites. • Comparable unmetamorphosed Mesozoic sedimentary sequences subcrop the basins or outcrop nearby. Common stratigraphic features include the occurrence of reef limestones of probable Upper Triassic age in east Sulawesi, Sula and Misool; Upper Triassic turbidites in east Sulawesi and Misool; Triassic-Lower Jurassic red beds in Banggai-Sula and the Bird's Head; Toarcian transgression followed by MiddleUpper Jurassic shales in Misool and BanggaiSula; Upper Jurassic limestones in east Sulawesi and Misool; mid Cretaceous hiatuses in BanggaiSula, the Bird's Head and possibly east Sulawesi; and deep-water Cretaceous limestones in Misool, Banggai-Sula and east Sulawesi. In addition, it is possible that the Late Cretaceous granite of the Bird's Head and volcanic/volcaniclasticrocks in
476
•
•
•
•
T . R . CHARLTON
Misool and offshore to the east might correlate with the as yet undated but post-Middle Jurassic dolerites of Banggai-Sula. Tertiary sedimentation commenced in the Tomori Basin and the eastern Salawati Basin (Salawati island and the western Bird's Head) during the Eocene with thin transgressive elastics succeeded by shelf carbonates. In both basins, however, the base of the Tertiary is strongly diachronous, ranging from upper Lower Eocene-Lower Miocene in the eastern Salawati Basin, and at least Upper Eocene-Lower Miocene in the Tomori Basin (Fig. 6). In Misool the basal elastics (the Daram Sandstone) are thicker (about 50 m: Pigram et al. 1982) and as old as Palaeocene. The succeeding shelf carbonates (Zaag Limestone) may also be locally as old as Palaeocene. Older Tertiary section may underlie the southern part of the Tomori Basin, but this has not yet been penetrated by exploration drilling. A break in predominant Tertiary carbonate sedimentation occurred in both basins in the mid Oligocene. In the eastern Salawati Basin this was marked by deposition of the elastic Sirga Formation, whilst to the west in Misool there is an unconformity possibly associated with minor folding. In the Tomori Basin a mid Oligocene unconformity separates the lower and middle members of the Tomori Formation. Prior to the mid Oligocene break in carbonate sedimentation both the Salawati and Tomori regions were accumulating relatively thin shelf sequences: the Faumai Formation is about 250 m thick in the eastern Salawati Basin (Visser & Hermes 1962; Pieters et al. 1983), and the lower member of the Tomori Formation is up to 100 m thick (Handiwiria 1990). Truly basinal successions did not begin to accumulate in both areas until the Early Miocene. In the eastern Salawati Basin the Lower-Middle Miocene consists of contemporaneous shelf (Kais Formation) and basinal (Klamogun Formation) limestones, with basinal deepening to the NW, towards the Sorong fault. In the Tomori Basin, shelf carbonates are found in the north, with more basinal limestones in the south, towards the South Sula fault. The Miocene section in the northwestern part of the eastern Salawati Basin is more than 2000 m thick (e.g. Vincelette & Soepardjadi 1976), and is more than 1000 m thick in the southern Tomori Basin (Handiwiria 1990). In the Middle Miocene fine-grained siliciclastics began to accumulate in both basins. In the eastern Salawati Basin the Middle-Upper Miocene partly clastic Klasafet Formation progressively replaces the deep-water carbonate Klamogun Formation. The proportion of siltstone in the
•
•
•
•
Klasafet Formation decreases southward away from the Sorong fault (Visser & Hermes 1962; Pieters et al. 1983), suggesting a elastic sedimentary source to the north or west. In the Tomori Basin the Middle Miocene Matindok Formation is recognized in all parts of the basin so far drilled, but it thickens considerably and the palaeoenvironment deepens to the south towards the South Sula fault. During the Late Miocene there was a return to shelf carbonate sedimentation in the Tomori Basin. On the southern flank of the eastern Salawati Basin carbonate shelf sedimentation had continued without significant interruption from the Early-Middle Miocene. Isolated patch and pinnacle reefs are recorded in both basins at this time. Carbonate sedimentation and growth of the economically important patch and pinnacle reefs became area!ly restricted in both the Tomori and Salawati basins near the Miocene-Pliocene boundary. The Pliocene Klasaman Formation of the eastern Salawati Basin corresponds approximately with the Kintom and Bia formations of the Tomori Basin. Both sequences show an overall coarsening-upward, from predominant claystone at the base to more sandy and conglomeratic facies at the top. Shelf carbonate facies persisted into the Plio-Quaternary of Misool (the Atkari Formation) and the Luwuk Formation of the southern Tomori Basin. A strongly erosional base Pleistocene unconformity is recognized in both basins. A succeeding conglomeratic unit is also recognised in both areas: the Sele Formation in the Salawati Basin, and the Kalomba Formation in the Tomori Basin. In addition to stratigraphic similarities, the Salawati and Tomori Basins share common hydrocarbon kerogen characteristics (SPT/ Pertamina 1992). In both cases an anoxic, restricted marine carbonate or calcareous shale sequence is the suspected source rock (Robinson 1987). In the Salawati Basin this is either the Lower Miocene Klamogun Formation (Robinson 1987) or the Upper Miocene Klasafet Formation (Livingstone 1992). In the Tomori Basin the source sequence is probably the Lower Miocene middle member of the Tomori Formation (Davies ! 990; Handiwiria 1990).
Regional tectonic implications Figure 9 shows a possible reconstruction of the eastern Sulawesi-western New Guinea (Bird's Head) region prior to development of the Sorong fault zone. The primary control on this reconstruction is the alignment of the Salawati and
SALAWATI (~; TOMOR! BASIN CORRELATION
•
/ Banggai r Islands .L • ~
, / ,- .
Sula islands
477
_
"-:~ WATt, =2:::::i:i~i:~ii::iii~BASlNili!i~i:
,soo,
] :!!i!i!i!iii!i!iii!i!i!i!i!:
"-- 3 ~
Tertiaryisopach (km) ~
Australiancover .......... I ~ AIIochthonoUSbasement I
~
Basin >2km sediment ~
Australian basement
i
I
Fig. 9. Reconstruction of the Bird's Head-Sula-Banggai-East Sulawesi region prior to development of the Sorong fault. The Bird's Heads block is drawn in its present-day orientation with respect to north. In order to keep SW Obi as an identifiable body, the inferred wrench faults bounding this terrane have not beeen restored.
Tomori basinal depocentres. However, as the Salawati Basin has a double depocentre, there is some ambiguity as to which depocentres should be aligned• Both alignments give a reasonable correlation of the basin isopachs, but placing the Tomori Basin north of the eastern (Sele Straits) depocentre of the Salawati Basin (Fig. 9) gives a more plausible basemap for palaeogeographic reconstructions (such as that for the Miocene described below - see Fig. 10). The net offset on the Soron~ fault zone based on this reconstruction is about 900 km. The Salawati and Tomori Basins show strong stratigraphic similarities up until fairly recent times, and this probably indicates that offset of the two half-basins occurred relatively recently. The restored composite basin in Fig. 9 is roughly elliptical in shape, with an oblique orientation suggestive of a strain ellipse resulting from leftlateral shear parallel to the Sorong fault zone. This probably suggests that the basin was formed largely as a result of transtension on the Sorong system. As most of the Tertiary stratigraphic thickness that gives rise to the elliptical shape is accounted for by the Plio-Quaternary clastic sequences, and additionally as the Salawati Basin
in particular shows little evidence of fault-related deformation prior to the Early Pliocene, it is likely that movement on the Sorong fault zone occurred primarily during the Pliocene and Quaternary periods. The precise age of the onset of clastic sedimentation in the Salawati Basin has not been stated in the published literature, but a stratigraphic column illustrated by Livingstone (1992) indicates the earliest deposition of the clastic Klasafet Formation in the Upper Miocene. Livingstone (1992) also showed an idealized burial plot for the Waipili-1 well in the northern part of Salawati island which indicates that the top of the Kais platform sequence began to subside rapidly between 6-7 Ma ago. This might indicate the time at which the Sorong fault system began to develop. A further constraint on the age of the Sorong fault zone comes from regional plate tectonics. It was suggested earlier that the orientation of the fault system parallel to the relative convergence vector between the Pacific and Indo-Australian plates indicates that the present-day Sorong system acts as a transcurrent boundary between these two plates (Fig. 1). The motion of Australia relative to the Pacific in this region is at a rate of 128 km Ma -1
478
x.R. CHARLTON
on a bearing of 068 ° (using the Euler poles of Minster & Jordan 1978). At this rate the 900 km displacement would be achieved in a period of 7 Ma. Assuming that plate motion has been constant through this period, this provides a minimum age for the Sorong system. This estimate is very similar to the commencement of subsidence in the Salawati Basin indicated by Livingstone (1992). The absence of significant pre-Pliocene faulting, together with the largely non-clastic nature of the Miocene basin fill, suggests that the Miocene carbonate basin pre-dated (and is therefore genetically unrelated to) the development of the Sorong fault system. Figure 10 shows a possible interpretation of the Miocene basinal palaeogeography based on extrapolating the well established palaeoenvironments of the Salawati Basin into the Tomori Basin, and taking into account the limited geological data from Banggai-Sula and east Sulawesi. It would seem reasonable from Fig. 10 that the combined Salawati-Tomori Basin originated at least in part as the foreland basin to the east Sulawesi orogenic belt. If this interpretation is correct, then the clear evidence from the Salawati Basin for strong basinal subsidence in the Lower Miocene suggests that orogenesis was underway in east Sulawesi by this time. This is somewhat
Fig. 10. Miocene palaeogeography of the eastern Salawati Basin (after Gibson-Robinson et al. 1990) extrapolated into the repositioned Tomori Basin.
earlier than the usually accepted Middle Miocene age (e.g. Davies 1990). There is, however, a discrepancy between the age of the east Sulawesi orogen as inferred from onshore geology (which suggests important orogenesis during the Middle Miocene or earlier) and the offshore seismic and well data which indicates fold-and-thrust belt development in the Tolo Gulf during the Pliocene (Davies 1990). As already described, the strike of the Tolo Gulf foldand-thrust belt is markedly oblique to that of the orogenic belt onshore in east Sulawesi. It appears that there were two distinct phases of development in the east Sulawesi orogen: an earlier pre-Middle Miocene phase which gave rise to a foreland basin on its eastern flank (the Miocene Salawati-Tomori Basin), and a later phase of fold-and-thrust belt development during the Pliocene. The earlier phase of deformation arose from arc-continent collision (Davies 1990), whilst the Pliocene deformation is probably related to the westward translation of eastern Sulawesi in the Sorong fault system. The Southeast Arm of Sulawesi shows clear structural and stratigraphic similarities with the East Arm, and probably formed a direct continuation of the the east Sulawesi orogenic belt prior to the Pliocene. The Tolo Gulf foldbelt probably developed contemporaneously with oroclinal bending of eastern Sulawesi, with the Southeast Arm rotating anticlockwise with respect to the East Arm. In addition to the Bird's Head and eastern Sulawesi, three displaced terranes are provisionally repositioned in Fig. 9. These terranes are relocated to give the tightest geographic fit consistent with the regional geology. Banggai and most of the Sula islands are treated as a single terrane which is fixed relative to eastem Sulawesi and the Tomori Basin. The positioning of Banggai-Sula is thus largely defined by the re-positioning of the Tomori Basin north of the Salawati Basin. However, some freedom remains as to body rotation, and in Fig. 9 Banggai-Sula (and eastern Sulawesi) have been rotated about 15 ° anticlockwise relative to the Bird's Head in order to achieve the tightest geographic fit. If, as is likely, the Bird's HeadMisool block has itself undergone body rotations relative to Australia during the Neogene, the inferred 15 ° rotation of Banggai-Sula would be additional to that in the Bird's Head. The second displaced terrane in Fig. 9 comprises Sulabesi island, eastern Mangole and the southwestern part of Obi island. The boundaries of this terrane may be a simplification: Mangole island is cut obliquely by a number of ENE-WSW trending steep faults which may be left-lateral faults (as suggested in the restoration) or could be normal faults. In Fig. 9 Sulabesi is restored by reversing
SALAWATI • TOMORI BASIN CORRELATION an assumed anticlockwise rotation of about 90 ° relative to Banggai-Sula and also relative to eastern Mangole. Sulabesi and eastern Mangole are then translated about 175 km eastward relative to Banggai-Sula (cf. 180 km displacement without rotation estimated by Garrard et al. 1988, based on the similarity of structural style and the nature of Jurassic sediments in Sulabesi and the Taliabu Shelf immediately west of the Sula islands). The third displaced terrane in Fig. 9 is the Tamrau Terrane of the northern Bird's Head. Most authors (e.g. Pigram & Davies 1987) have interpreted very large displacements between the Tamrau Terrane and the Bird's Head, in part based on the apparent necessity of accommodating the large strike-slip offsets of the Sorong fault zone to the west on the single fault strand of the Sorong fault. However, the geology of the Tamrau Terrane is not greatly different from that of the Bird's Head (cf. Visser & Hermes 1962), and it is suggested that the Tamrau Terrane can be repositioned by a relatively small eastward translation of about 60 km. This would place the Netoni Igneous Complex (Pieters et al. 1981) north of the Anggi Granites of the eastern Bird's Head, and re-aligns the eastem end of the Tamrau Terrane with the eastern edge of the continental Bird's Head block. (It has been suggested by several authors that the Netoni Block is itself a distinct 'mini terrane' within the Sorong fault; however, the overlap of the Cretaceous Amiri Sandstone across the supposed Tamrau-Netoni terrane boundary (Hartono et al. 1989) does not support this distinction). A consequence of this relatively limited displacement on the Sorong fault through the Bird's Head is that most of the movement on the Sorong system must be taken up north of the Tamrau Terrane. In Fig. 9 most of the inferred left-lateral movement is taken up along the northern edge of the Tamrau Terrane such that Sulabesi island is re-positioned immediately northwest of the Tamrau Terrane. This aligns N-S trending normal faults in the Sula islands with similar faulting on the eastern flank of the Salawati Basin, and also juxtaposes the Jurassic-Cretaceous shale sequences of the Sula islands with comparable sequences in the Tamrau Terrane.
479
Implications for hydrocarbon exploration The reconstructions in Figs 9 & 10 may also have some implications for future oil exploration in this region. In the Salawati Basin, nearly all oil production is from pinnacle reefs which developed near the shelf-slope break around a semi-enclosed Miocene basin. If this palaeogeography can be extrapolated into the repositioned Tomori Basin as suggested in Fig. 10, then the belt of pinnacle reefs should follow the eastern margin of the basin to the south and west of the Banggai island group. In addition, it now seems widely agreed that the primary source rocks for the Salawati oils are Miocene deep marine and poorly oxygenated calcareous mudstones and/or marly limestones which were deposited either contemporaneously with or immediately above the Kais pinnacle reefs. It is probably more than coincidence that these restricted basinal sediments were deposited in the semi-enclosed Miocene Salawati embayment (Fig. 10). The palaeogeographic map also indicates that a similar embayment would be expected at the northern end of the Tomori Basin, and such a restricted basin might well be the source for oil and gas discoveries along the western flank of the Tomori Basin. However, the predominant structure of the northern Tomori Basin is that of a foreland basin, and any oil generated in this region would be more likely to migrate updip towards the foreland; that is towards the postulated pinnacle reef trend on the eastern flank of the Tomori Basin. Unfortunately, most of the extrapolated pinnacle reef trend lies under water depths of 10002000 m, and is therefore unlikely to be a commercial prospect with present-day economics. This work was initiated as part of the London University study of the Sorong fault zone under the leadership of Prof. Robert Hall. Fieldwork in this area was sponsored by the Royal Society, NERC grant GR3/7149 and the London University Consortium for Geological Research in Southeast Asia. Thanks to Chris Gibson-Robinson and Peter Lunt (Petromer Trend Corp.) and Andy Livsey (Simon Petroleum Technology, Jakarta) for useful discussions, and to Tony Barber (Royal Holloway University of London) and Dick Garrard (ARCO) for useful comments in review.
References AB1MANYU,R. 1990. The stratigraphy of the Sulawesi Group in the Tomori PSC, East Arm of Sulawesi. Paper presented at the 19th Annual Convention of the Indonesian Association of Geologists. COCKCROFT,P. J., GAMBER,D. A. & HERMAWAN,H. M. 1984. Fracture detection in the Salawati Basin of Irian Jaya. Proceedings of the Indonesian Petroleum Association, 13, 125-133. CORNICE,J.-J., VILLENEUVE,M., MARTINI,R., ZANINETFI,
L., VACHARD, D., E~r AL. 1994. Une plate-forme carbonatEe d'~ge rhEtien au centre-est de Sulawesi (rEgion de Kolonodale, CEl~bes, IndonEsie).Comtes Rendus Acad~mie des Sciences de Paris, 318(II), 809-814. DAVIES,I. C. 1990. Geological and exploration review of the Tomori PSC, eastern Indonesia. Proceedings of the Indonesian Petroleum Association, 19, 41-67.
480
T.R. CHARLTON
Dow, D. B. & SUKAMTO,R. 1984. Western Irian Jaya: the end-product of oblique plate convergence in the late Tertiary. Tectonophysics, 106, 109-139. FROIDEVAUX,C. M. 1974. Geology of Misool island (Man Jaya). Proceedings of the Indonesian Petroleum Association, 3, 189-196. • 1977. Tertiary tectonic history of the Salawati area, lrian Jaya, Indonesia. Proceedings of the Indonesian Petroleum Association, 6, 199-220. GARRARD, R. A., SUPANDJONO, J. B. & SURONO. 1988. The geology of the Banggai-Sula microcontinent, eastern Indonesia. Proceedings of the Indonesian Petroleum Association, 17, 23-52. GIBSON-ROBINSON, C., HENRY, N. M., THOMPSON, S. J. & HARYONO TRI RAHARIO 1990. Kasim and Walio fields - Indonesia. In: BEAUMONT,E. A. (compiler) Stratigraphic Traps, I, AAPG Treatise of Petroleum Geology, 257-295. HALL, R., NICHOLS, G. J., BALLANTYNE,P., CHARLTON,T. & ALl, J. 1991. The character and significance of basement rocks of the southern Molucca Sea region. Journal of SE Asian Earth Sciences, 6, 249 258. HAMILTON, W. 1979. Tectonics of the Indonesian Region. US Geological Survey Professional Papers, 1078. HANDIWmlA, Y. E. 1990. The stratigraphy and hydrocarbon occurrences of the Salodik Group, Tomori PSC area, East Arm of Sulawesi. Paper presented at the 19th Annual Convention of the Indonesian Association of Geologists. HARTONO, U., AMRI, C. & PIETERS, P. E. 1989. Geological
map of the Mar Sheet, Irian Jaya. 1:250,000. Geological Research & Development Centre, Indonesia. HERMES, J. J. 1968. The Papuan geosyncline and the concept of geosynclines. Geologie en Mijnbouw, 4 7 , 81-97. KATILI, J. A. 1978. Past and present geotectonic position of Sulawesi, Indonesia. Tectonophysics, 45, 289-322. KERTAPATI, E. K., SOEHAIMI, A. & DJUHANDA, A. 1992. Seismotectonic map of Indonesia. Geological Research & Development Centre, Indonesia. LINTHOUT, K, HELMERS,H & ANDRIESSEN,P. A. M. 1991. Dextral strike-slip in Central Seram and 3-4.5 Ma Rb/Sr ages in pre-Triassic metamorphics related to Early Pliocene clockwise rotation of the BuruSeram microplate (E. Indonesia). Journal of SE Asian Earth Sciences, 6, 335-342. LIVINGSTONE, H. J. 1992. Hydrocarbon source and migration, Salawati Basin, Irian Jaya. Eastern Indonesia Exploration Symposium, Simon Petroleum Technology/Pertamina. LUNT, P. & DJAAFAR,R. 1991. Aspects of the stratigraphy of western Irian Jaya and implications for the development of sandy facies. Proceedings of the Indonesian Petroleum Association, 20, 107-124. MILSOM, J., MASSON, D. 8,: NICHOLS, G. 1992. Three trench endings in eastern Indonesia. Marine Geology, 104, 227-241. MINSTER, J. B. & JORDAN, T. 1978. Present-day plate motions. Journal of Geophysical Research, 83, 5331-5354. PIETERS, P. E., HARTONO, U. & AMRI, C. 1981.
Preliminary geologic map of the Mar Quadrangle,
lrian Jaya. 1:250,000. Geological Research & --,
Development Centre, Indonesia. PIGRAM, C. J., TRAIL, D. S., DOW, D. B., RATMAN, N. & SUKAMTO, R. 1983. The stratigraphy of western Irian Jaya. Bulletin of the Geological
Research & Development Centre, Indonesia, 8, 14--48. PIGOTT, J. D., TRUMBLY, N. I. & O'NEAL, M. V. 1982. Northern New Guinea wrench fault system: a manifestation of late Cenozoic interactions between Australian and Pacific plates. In: Watson, B. (ed.)
Transactions of the Third Circum-Pacific Energy and Mineral Resources Conference, 613--620. PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and accretion history of the New Guinea orogen. BMR Journal of Australian Geology & Geophysics, 10, 193-211. - & PANGGABEAN,H. 1984. Rifting of the northern margin of the Australian continent and the origin of some microcontinents in eastern Indonesia. Tectonophysics, 107, 331-353. -& SUKANTA,U. 1982. Geological data record of the Taminabuan 1:250,000 sheet area, Irian Jaya. Open file report, Geological Research & Development Centre, Indonesia. -& -1989. Geology of the Taminabuan Sheet area, lrian Jaya. 1:250,000. Geological Research & Development Centre, Indonesia. , CHALLINOR,A. B., HASIBUAN, E, RUSMANA, E. & HARTONO, U. 1982. Lithostratigraphy of the Misool Archipelago, Irian Jaya, Indonesia. Geologie en Mijnbouw, 61,245-279. , SURONO & SUPANDJONO, J. B. 1985. Origin of the Sula Platform, eastern Indonesia. Geology, 13, 246-248. REDMOND,J. L. & KOESOEMADINATA,R. P. 1976. Walio oil field and the Miocene carbonates of Salawati Basin, lrian Jaya. Proceedings of the Indonesian Petroleum Association, 5, 41-57. RI~HAULT,J. P., MALOD, J. A., LARUE,M., BURHANUDDINN, S. ~ SARMILI,L. 1991. A new sketch of the central North Banda Sea, Eastern Indonesia. Journal of SE Asian Earth Sciences, 6, 329-334. ROBINSON, K. M. 1987. An overview of source rocks and oils in Indonesia. Proceedings of the Indonesian Petroleum Association, 16, 97-122. RUSMANA,E., KOSWARA,A. & SIMANDJUNTAK,T. O. 1984.
[Report on the geology of the Luwuk Quadrangle, 1:250,000]. Open file report, Geological Research --,
& Development Centre, Indonesia• [In Indonesian]. SUKIDO, SUKARNA, D., HARYONO, E. & SIMANDJUNTAK, T. O. 1986. Preliminary geological
map of the Lasusua-Kendari Quadrangle, Sulawesi. 1:250,000. Geological Research & Development Centre, Indonesia. SATO, T., WESTERMANN, G. E. G., SKWARKO, S. K. & HASIBUAN, E 1978. Jurassic biostratigraphy of the Sula islands, Indonesia. Bulletin of the Geological
Research & Development Centre, Indonesia, 4, 1-28. SIMANDJUNTAK,T. O. 1990. Sedimentology and tectonics of the collision complex in the East Arm of Sulawesi, Indonesia. Geologi Indonesia, 13(1), 1-35.
SALAWATI & TOMORI BASIN CORRELATION SIMBOLON, B., MARTODJOJO, S. & GUNAWAN,R. 1984. Geology and hydrocarbon prospects of the preTertiary system of Misool. Proceedings of the Indonesian Petroleum Association, 13, 317-340. SMITH, R. B. & SILVER,E. A. 1991. Geology of a Miocene collision complex, Buton, eastern Indonesia. Bulletin of the Geological Society of America, 103, 660-678. SPT (SIMONPETROLEUMTECHNOLOGY)/PERTAMINA.1992. Eastern Indonesia: Biostratigraphy, Geochemistry and Petroleum Geology. Unpublished non-exclusive consultancy study. SUKAMTO,R. 1975. Geologic map of Indonesia, sheet 8: Ujung Pandang. 1:1,000,000. Geological Survey of Indonesia. SUPANDJONO,J. B. & HARYONO,E. 1993. Geology of the
Banggai Sheet, Sulawesi-Maluku.
1:250,000.
Geological Research & Development Centre, Indonesia.
481
SURONO & SUKARNA, D. 1993. Geology of the Sanana Sheet, Maluku. 1:250,000. Geological Research & Development Centre, Indonesia. --, SIMANDJUNTAK, T. O., SITUMORANG, R. L. & HADIWIJOYO, S. 1987. Geologic map of the Batui Quadrangle, Sulawesi. 1:250,000. Geological Research & Development Centre, Indonesia. TJIA, H. D. 1973. Irian fault zone and Sorong melange. Sains Malaysiana, 2(1), 13-30. VINCELETTE,R. R. & SOEPARJADI,R. A. 1976. Oil-bearing reefs in Salawati Basin of Irian Jaya, Indonesia. AAPG Bulletin, 60, 1448-1462. VISSER, W. A. & HERMES,J. J. 1962. Geological results of the exploration for oil in Netherlands New Guinea.
Koninklijk Nederlands Geologie en. Mijnbouw Genootschaap Verhandlingen, Geologische Serie, 20.
The geology and tectonic evolution of the Bacan region, east Indonesia JEFFREY
E A. M A L A I H O L L O
& ROBERT
HALL
SE Asia Research Group, Department of Geological Sciences, University College London, Gower Street, London WCIE 6BT, UK
Abstract: Bacan is located in the zone of convergence between the Eurasian, Philippine Sea and Australian plates. The oldest rocks in Bacan belong to the Sibela Continental Suite and are probably of Precambrian age. The complex includes continental phyllites, schists and gneisses of upper amphibolite facies. Isotopic dating yielded extremely young ages due to interaction with hydrothermal fluids. Juxtaposed against the continental rocks is the mostly unmetamorphosed, arc-related Sibela ophiolite, probably derived from the Philippine Sea plate. Isotopic dating yielded a Cretaceous age with an Oligocene-Miocene overprint. In north Bacan, the oldest formation is the Upper Eocene Bacan Formation which comprises interbedded arc volcanic and turbiditic volcaniclastic rocks, metamorphosed under conditions between the prehnitepumpellyite and greenschist facies. A similar Lower Miocene sequence, assigned to the same formation, is exposed in south Bacan. The Oligocene Tawali Formation on Kasiruta, NW of Bacan, consists of arc basalts and volcaniclastic turbidites, metamorphosed to zeolite facies. The Bacan and Tawali Formations represent different parts of an arc, active from Late Eocene until Early Miocene, resulting from northward subduction of the Australian plate under the Philippine Sea plate. There is a major Lower Miocene unconformity, representing collision of the Australian continent with the Philippine Sea plate, above which shallow marine limestones of the LowerMiddle Miocene Ruta Formation were deposited. This deposition was interrupted by sudden influxes of volcaniclastic sands, forming the Amasing Formation. The Upper MiocenePleistocene Kaputusan Formation, rests locally unconformably on older rocks, and includes three members. The Goro-goro Member consists of arc andesites, originating from four eruption centres, which erupted from Late Miocene to Pleistocene, with the oldest in south and the youngest in north Bacan. The Pacitak Member consists of shallow marine pyroclastic rocks. The Mandioli Member formed fringing coastal reef limestones. The volcanic rocks of this formation were produced by eastward subduction of the Molucca Sea plate. Quaternary basalts are related to movement along the Sorong fault. It is concluded that most of the Bacan region has been part of the Philippine Sea plate since the Cretaceous; volcanic rocks of different ages all have an arc character and chemistry, and lithological variations reflect different positions within the volcanic arc; and there is evidence for continental crust of Australian origin in the Bacan area by the Early Miocene.
Three major plates, the Eurasian, Philippine Sea and Australian plates, converge in the Bacan region (Fig. 1). The Eurasian plate has a complex eastern margin which includes continental and volcanic arc crusts accreted during collision with the Philippine Sea plate (e.g. Rangin 1991). The Philippine Sea plate has a long arc history (Hall et al. 1995a), extending back at least to the Cretaceous, and is currently converging with Eurasia in the Philippines, and overriding the Molucca Sea plate in the south. The Australian plate has been moving northwards since the Cretaceous. During much of the Palaeogene, oceanic crust of the Australian plate was subducted under the southern edge of the Philippine Sea plate (Hall et al. 1995b). A strike-slip fault system, the Sorong fault, subsequently developed at this plate boundary. Bacan contains rocks of continental, ophiolitic and arc affinities (Fig. 2), and therefore an under-
standing of the tectonic evolution of Bacan should greatly enhance our knowledge of the development of this area of complex plate boundaries. This paper briefly describes and summarizes the geology of each of the formations in the Bacan region, discusses them in the light of regional tectonic events, and proposes a synthesis of the tectonic evolution of this region. A detailed description of the lithostratigraphy of Bacan (Fig. 3), with chemical and palaeoenvironmental +analyses can be found in Malaihollo (1993).
Geology and tectonic evolution of Bacan Sibela continental suite The oldest rocks in the Bacan region form part of the Sibela Continental Suite. The complex includes continental phyllites, schists and gneisses of upper
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 483-497.
483
484
J.F.A. MALAIHOLLO 8z R. HALL
PHILIPPINE SEA
500 krn
-4°N
I
CELEBESSEA
PACIFIC OCEAN
f~ 2
Morotai
~ GULF OF TOMINI
o~..
MOLUCCA SEA
I I BACAN
~ano'kwari 1"to,.
LMAHERA
-.,-... o, Gag ~
Waigeo
~
"
~$orang Fau, "~0.~(
~
Bird's Head
IRIAN JAYA %
BANDA SEA
FLORES SEA
ARAFURA SEA D
o
Fig. 1. Present-day tectonic setting of east Indonesia based on Hamilton (1979).
amphibolite facies (c. 600°C, c. 5 kbar), with strong penetrative fabrics indicative of a polyphase deformation and recrystallization history, typical of Barrovian type dynamo-thermal metamorphism (Brouwer 1923; Hall et al. 1988). There is also evidence for retrograde metamorphism associated with post peak-metamorphism deformation. The protoliths were pelites with minor amounts of sandy and carbonate/marly horizons. Whole rock chemistry of the pelites suggests that they were derived from a cratonic area and were deposited on an active margin underlain by continental crust. Based on regional stratigraphical arguments (Hamilton 1979) and the Pb isotopic signatures of Quatemary volcanic rocks (Vroon 1992), these metamorphic rocks are postulated to be of Precambrian or Palaeozoic age. This suggestion has not been confirmed isotopically, as K-Ar and Ar-Ar step heating results yielded extremely young ages (< 0.21 Ma; Malaihollo 1993), interpreted to be a result of interaction with a hydrothermal fluid carrying fractionated argon which were subsequently concentrated between the sheet silicate layers, causing an apparent young age. This mechanism could account for other unusually young ages from high-grade metamorphic rocks in the region (cf. Linthout et al. 1991). There are highly foliated metasedimentary rocks
in the south Saleh Islands and the SE comer of north Bacan. These rocks have a similar character to the Sibela Continental Suite, particularly in having brown metamorphic biotite which throughout the whole region is found only in the Sibela Continental Suite, but are of lower metamorphic grade. This paper interprets the Sibela Continental Suite to be derived from the Australian plate, where these rocks formed either part of the Mesozoic-Tertiary passive margin of north Australia or fragments rifted from Australia during Gondwana break-up. Many authors have suggested or implied that the continental rocks in Bacan were introduced into the region via strike-slip motion of the Sorong fault system from the Bird's Head region of Irian Jaya (e.g. Hamilton 1979; Silver et al. 1985) or central Papua New Guinea (Pigram & Panggabean 1984). 87Sr/86Sr isotopic compositions of volcanic rocks (E. Forde, pers.comm. 1993) suggest that the Upper Miocene Kaputusan Formation on south Bacan was erupted through continental crust, indicating the presence of continental material under the Bacan region by the Late Miocene. The presence of a Lower Miocene post-collisional intrusion (the Nusa Babi Monzodiorite) in Bacan suggests the presence of continental crust by the Early Miocene (as discussed below).
7
"•
.I~"
.
, ~ "
a
k
~
:.Sibela Mts::
~
v.'
~,,%-
Goro-goro
Isles~,
Saleh
1127"45'E
~ ~
®
L)
Continental Suite
Sibela Ophiolite
I'7~_-~'~ Bacan Fm
Tawali Fm
Intrusions
Ruta Fm
NusaBobi
~
Volcanics Kaputusan Fm Mandioli Mb Kaputusan Fm Pacitak Mb Kaputusan Fm Goro-Goro Mb Saleh Intrusions Amasing Fm
~
~
~
~
o~ o o
8
i
-
~_ ®
o~
(~
•
Alluvium& Limestones
LEGEND
[--1
Fig. 2. New simplified geological map of the Bacan region, based on aerial photographic interpretation and field mapping (Malaihollo 1993).
20
t
i
km
i
I ~27°3°'E
Nusa Babi
30
10
~
K,apuTusan ~,
, ~ =~-~.
!
I
,,'
I
i
-
J127°15'E
0
Kasiruta ~e--(TawaliBesar)
,,s
t
s'
0"45'S
m
0,30,'~
0"~5"~
4~ Oo C~h
>
Z 0
z
©
('1
© ©
486
J.F.A. MALAIHOLLO 8Z R. HALL
Ma 0-
EPOCH
Kasiruta Alluvium
Reef
Bacan ~
Pleistocene Pliocene
5.2
Possible Tectonic Events Quaternary volcanism related to movements on Sarong Fault.Upliftof Sibela Block. Thrustingin Halmahera, movement on splaysof Sarong Fault.
aputusan m
Kapulusan FB
Goro-goro Mbr
Goro-goro Mbr
Pacitak Mbr
Pacltak Mbr
Mandioli Mbr
Mandioli Mbr
10,4
Associated subaerial and shallow marine pyroclasticdeposits,and fringingreefs, Four eruptive centres which migrated northwards with time. Start of volcanism in Bacan (-7 Ma), Halmahera and Obi (- 12 Ma).
Initiationof eastward subducfion of Molucca Sea Plate under Halmahera,
Miocene
Ruta Formation
16.3 m
Shallow marine carbonate platform [open platform, patch reef,tidalbar and foreslope facies). Local shallow marine volcaniclasticand littoraldeposit,
Collisionof Australiaand arc on the PhilippineSea plate.
23.3 Tawali Fm Marikapal Mbr
Oligocene
Tawali Fm Jojok Mbr
Bacon Formation
Arc volcanic and turbidites, Repeated, high concentration, proximal volcaniclasticturbidites, i Metamorphosed to PrehniteInner and middle fan facies, Pumpellylte to Lower Greenschist facies. I Volcanism related to the northward movement of Pillowlavas with volcanic arc chemistry. Australiaunder the Zeolitefaciesmetamorphism. PhilippineSea plate, I
35.4 Eocene
I Similarsequence in Bird's Head and southern Philippines,
Fig. 3. Cenozoic stratigraphy of the Bacan region with related local tectonic events. Timescale after Harland et al. (1990).
Sibela ophiolite
Juxtaposed against the continental suite is the Sibela ophiolite (Hall et al. 1988). There are two types of material recognized in this suite: ophiolitic rocks, most of which are of lower crustal plutonic and mantle origin (microgabbros, gabbros and serpentinised harzburgites) with few volcanic rocks, and spatially-related rocks including unusual amphibole cumulates and many rocks with magmatic-tectonic fabrics. This dismembered, incomplete ophiolitic complex is mostly unmetamorphosed, although locally there are mylonitic rocks which have been affected by upper amphibolite-lower granulite facies metamorphism (c. 1000oc, c. 5 kbar) related to ductile deformation of hot rocks at or near their place of formation, and local recrystallization in shear zones (Malaihollo 1993). Chrome spinel compositions suggest that the peridotites may have formed in an arc setting. This is supported by geochemical evidence from cogenetic metagabbros. Petrographic study reveals that most of the cumulate rocks consist mainly of cumulus amphibole with minor pyroxene and intercumulate plagioclase, indicating crystallization
under hydrous conditions, implying an arc-related setting. The cumulate rocks in the Sibela ophiolite are very different from those of the Halmahera ophiolite, which consist of olivine, orthopyroxene and clinopyroxene (Ballantyne 1992). The Halmahera cumulates are genetically related to the ophiolite which is interpreted to have formed in a supra-subduction zone, normally associated with the initiation of a subduction zone, from the intraPacific subduction (Ballantyne & Hall 1990; Ballantyne 1991). Preliminary Nd-Sm ages on the Halmahera cumulates are Jurassic (M. E Thirlwall, pers. comm. 1992). Ar-Ar and K-Ar dating of the Sibela cumulates and metagabbro (Malaihollo 1993) yielded Cretaceous (97-94 Ma) and Oligocene-Miocene (37-25 Ma) ages, although amphiboles analysed contain excess argon and therefore plateaux ages will be slightly older than closure ages (Lanphere & Dalrymple 1976). The Cretaceous age may be linked to volcanic activity recorded in the east Halmahera ophiolite (9480Ma; Ballantyne 1992), suggesting a link between the Sibela ophiolite and the east Halmahera ophiolite. The Oligocene-Miocene ages are related to the Oligocene volcanism (discussed below).
GEOLOGY • TECTONICS OF BACAN, E INDONESIA There are two possible interpretations for the association of ophiolitic and cumulate rocks. The first possibility is that the ophiolitic rocks represent older crust of possibly Jurassic age (cf. east Halmahera-Gag ophiolite) intruded by Cretaceous arc cumulates related to intra-oceanic subduction. This interpretation is supported by chemical similarities between the cumulates and zoned ultramafic complexes, normally associated with arc magmatism. Alternatively, both the ophiolitic and arc cumulate rocks may be related to the same intra-oceanic subduction during the Cretaceous. The age and nature of the original crust upon which the arc was built is unknown.
Juxtaposition of continental and ophiolitic rocks The continental and ophiolitic rocks are juxtaposed in the Sibela Mountains. Differences in metamorphic character indicate that metamorphism occurred before amalgamation of continental and ophiolitic rocks. The continental suite is interpreted to be part of the Australian continental basement, whereas the ophiolitic rocks represent the Philippine Sea plate. Regional mapping shows that since the Early Miocene, rocks from these two plates have the same geological history, indicating an Early Miocene collision (Hall et al. 1995a). There are three possibilities for the mode of juxtaposition of the continental and ophiolitic rocks: (1) the Sibela Mountains represent a suture zone resulting from collision between the Australian and Philippine Sea plates; (2) the continental rocks collided with the Philippine Sea plate and were subsequently translated by strike-slip faulting as part of a composite fragment to their present position; (3) the continental rocks remained part of the north Australian Margin until the late Neogene and were then translated by strike-slip faulting on the Strong fault into the region. These possibilities are discussed below.
487
erupted in an arc setting. Turbiditic rocks were deposited by dilute, low concentration, possibly distal or low energy currents. The Bacan Formation was affected by a thermal event at c. 15 Ma. In south Bacan the oldest rocks dated are a very similar sequence of Early Miocene age. These interbedded volcanic and volcaniclastic turbidite rocks were metamorphosed to prehnitepumpellyite facies (c. 240-330°C, c. 2 kbar) due to burial metamorphism, with local hydrothermal alteration. Whole rock and mineral chemistry indicates an arc origin for the volcanic rocks. Graded volcaniclastic arenites indicate deposition by turbidity currents which dissipated their energy as they travelled down slope. Isotopic dating shows that this formation was affected by a thermal event atc. 8Ma. Undated metabasites in the Saleh Islands were metamorphosed at conditions between upper prehnite-pumpellyite and lower greenschist facies (c. 250-360°C, c. 4 kbar). The metamorphic character of these rocks suggests static, regional metamorphism. Geochemical and lithological evidence from the metabasites indicates that they represent an arc-related calc-alkaline sequence, possibly formed in a backarc, and they have similar chemical characteristics to the Upper Eocene and Lower Miocene sequences. The Upper Eocene, Lower Miocene and the Saleh Island sequences are similar in all aspects, except age. More dating is needed, although this may prove difficult, since despite examination of > 200 thin sections, only four samples were found to be suitable for biostratigraphic or isotopic dating. This is attributed to lack of fossils in the original sediments, metamorphism which destroyed fossils, and alteration by thermal overprints of the volcanic rocks. The rocks assigned to the Bacan Formation could represent either different parts of a single, continuously active arc, or separate arcs active at different times between the Late Eocene and Early Miocene.
The Tawali Formation The Bacan Formation In north Bacan the oldest rocks exposed belong to the Upper Eocene Bacan Formation. This comprises interbedded basic-intermediate volcanic and volcaniclastic turbiditic rocks. The Bacan Formation was folded and subsequently metamorphosed under conditions transitional between the prehnite-pumpellyite and lower greenschist facies (250-330°C, c. 2 kbar upwards), with characteristics of burial metamorphism. Hydrothermal alteration may be related to the Neogene Kaputusan volcanism. Mineral and whole rock geochemistry indicates that the volcanic rocks were
The oldest rocks on Kasiruta belong to the newly defined Tawali Formation, named after the island of Kasiruta (also known as Tawali Besar). Rocks forming part of this formation were originally assigned to the Bacan Formation (Yasin 1980), but here they are assigned to a new formation because of differences in lithology and the character of metamorphism and deformation. The lower part of the Tawali Formation consists of basaltic pillow lavas and interbedded sediments (Jojok Member), and it is overlain by fossiliferous volcaniclastic turbidites (Marikapal Member). This formation is affected by burial metamorphism transitional
488
J.F.A. MALAIHOLLO & R. HALL could be a function of different structural positions during collision with the Australian continent or different thermal environments due to different positions in the pre-collision arc. Thus, the simplest explanation for all these Upper Eocene, Oligocene and Lower Miocene volcanic arc formations is that they represent arc volcanism at the edge of the Philippine Sea plate due to the subduction of the Indo-Australian plate under the Philippine Sea plate (Fig.4). This is supported by the fact that all of these formations record similar palaeomagnetic declinations and inclinations which are characteristic of the Philippine Sea plate (Hall et al. 1995a; Ali & Hall 1995).
between low and high temperature zeolite facies (c. 180°C, < 2 kbar). Whole rock chemistry indicates that the Lower Oligocene basalts are highly differentiated arc lavas, erupted in a deep, open marine environment above the CCD. The Upper Oligocene volcaniclastic rocks are products of repeated, high concentration, proximal turbidity currents with associated slumped deposits. The Tawali Formation is interpreted as an equivalent of the Bacan Formation although there are several differences between them. There is an apparent lack of Oligocene rocks on Bacan, although as noted above, only a few samples of the Bacan Formation have proved dateable. There are also differences in lithology and metamorphic character (particularly the lower grade of metamorphism suffered by the Tawali Formation relative to the Bacan Formation), and trace element differences between the Tawali and the Bacan Formations. Rocks of similar lithology and age to the Tawali Formation have now been identified throughout the Halmahera-Waigeo region (the Tawali Formation in Halmahera, the Anggai River Formation in Obi and the Rumai Formation in Waigeo; Hall et al. 1991). The Oha Formation of Halmahera is probably not of Upper CretaceousEocene age as previously suggested (Hakim & Hall 1991), but is interpreted as part of the same arc as the Bacan Formation. Despite differences, all these formations range in age from Upper Eocene to Upper Miocene and they all pre-date the c. 22 Ma unconformity (discussed below) interpreted to be related to the arrival of the continental crust in the region. Differences in lithology could be explained if the different formations represent different parts of the arc (e.g. arc slopes, backarc basin, forearc slopes). Differences in the grades of metamorphism
~
TawaliFm,,MarikapalMb.& RumaiFm.(turbidites) ~
Early Miocene unconformity Above the Bacan and Tawali Formations, there is a major regional unconformity. This is characterized by a major change in lithological, metamorphic and structural character. Rocks below the unconformity are folded arc volcanic rocks and turbidites, metamorphosed up to greenschist facies, whereas the rocks immediately above it are mostly unfolded, unmetamorphosed carbonates. The age of the unconformity is estimated to be c. 22 Ma (Early Miocene) throughout the Halmahera region (Hall et al. 1995a). Nusa Babi intrusive rocks The Nusa Babi Monzodiorite (NBM) includes quartz-monzodiorite dykes and plugs with associated aplite dykes of monzogranite composition. These intrude the Bacan Formation and have been reported to intrude the Sibela Complex (Yasin
BacanFm.arc BacanFm.&AnggaiRiverFm, (forearcturbidites
(back-arcpillowbasalts)
SibelaContinentalSuite .
Philippine Sea plate
X~4
×
×X
XX
,
Australian plate No Scale Intended
Fig. 4. Different positions, in the pre-collisional stack, of the volcanic formations in the Bacan-Halmahera-Waigeo region which were part of the Late Eocene-Early Miocene arc.
GEOLOGY & TECTONICS OF BACAN, E INDONESIA
1980), but do not intrude the Ruta and Amasing Formations. The NBM is locally altered, possibly due to auto-metasmorphism. Magma evolution was primarily controlled by fractionation and the addition of water. Whole rock geochemistry indicates that the NBM is of plutonic arc or collisional magmatism origin. K-Ar dating yields an Early Miocene age (Baker & Malaihollo 1996). The geochemistry, isotopic age and the fact that the NBM intrudes the Bacan Formation indicate a postcollisional origin, probably related to the arrival of the Sibela Continental Suite in the region.
489
The Saleh Diorite
The Saleh Diorite intrudes the Sibela Continental Suite in the Saleh Islands and is juxtaposed against it in central Bacan. This intrusion consists of amphibole-bearing diorite and micro-diorite. Locally these rocks have suffered low temperature alteration, probably due to auto-metamorphism, and subsequent hydrothermal alteration. Whole rock geochemistry indicates that the Saleh Diorite is different from the NBM and is of arc or postcollisional origin. K-Ar dates suggest the c. 15 Ma age is related to the initiation of the Halmahera arc.
The Ruta and A m a s i n g Formations
The Lower-Middle Miocene Ruta Limestone lies unconformably above the Bacan and Tawali Formations. There are four microfacies recognized in the Ruta Limestone: skeletal wackestonespackstones, algal boundstones, foraminiferal packstones and bioclastic-lithoclastic packstones. These represent deposition on an open platform, a platform margin build-up (patch reef), a tidal bar downslope from a patch reef (open platform) and foreslope talus respectively, all forming part of a shallow marine carbonate platform. The deposition of the Ruta Formation began in the Early Miocene, predominantly as open platform, platform margin build-up and foreslope talus facies• Between the Early~ Miocene and Middle Miocene, deposition was locally interrupted by sudden and high influxes of volcaniclastic material forming sandstones of the Amasing Formation. Three facies in the Amasing Formation have been recognized: shallow marine with storm horizons, shoal or estuarine, and beach deposits. These facies represent shallowing upwards from an open carbonate platform to a beach environment. Deposition of the Ruta Formation continued until the Middle Miocene and its upper part is dominated by the tidal bar facies. There was widespread development of carbonate platforms at this time throughout the Halmahera-Waigeo region (Hall et al. 1995a), in north Irian Jaya (Pigram & Davies 1987; Pigram et al. 1990) and in the Philippines (Mitchell et al. 1986). Widespread carbonate platform development suggests that most of the Bacan-Halmahera-Waigeo region was in a quiet tectonic setting, in an equatorial position and was part of an extensive shallow marine area. In north Irian Jaya, the Middle Miocene Moon Volcanics (Pieters et al. 1989) may be linked with the Maramuni arc of Papua New Guinea which Dow (1977) suggested represented post-collision subduction reversal. However, the character and distribution of this volcanism is still poorly known and if an arc existed, it may have been limited to the eastern part of New Guinea (Ali & Hall 1995).
The Kaputusan Formation
The Kaputusan Formation was deposited above the Ruta and Amasing Formations. It consists of three members: the Goro-goro Volcanic, the Pacitak Volcaniclastic, and the Mandioli Limestone Members• Field mapping and aerial photographic interpretation indicates that the Kaputusan Formation rests directly upon the Bacan Formation in some areas implying an unconformable contact at its base. The Goro-goro Member consists of twopyroxene andesites (TPAN), hornblende-pyroxene andesites (HPAN), hornblende andesites (HBAN), and hornblende-biotite andesites (HBIAN). Associated with the volcanic rocks are subaqueous pyroclastic flows with related base surge deposits. The petrography, mineral chemistry, whole rock major and trace element chemistry of these rocks are typical of island-arc volcanic rocks. They are mostly fresh, akhough locally they are metamorphosed to zeolite facies, interpreted as due to hydrothermal activity. Magma diversification of the Goro-goro Member was achieved mainly by fractionation of plagioclase, pyroxene, Fe-Ti oxide, amphibole and biotite, with evidence for replenishment, immiscibility, resorption and assimilation. Magmatic conditions were < 8 kbar and 2-10 wt.% H20 with PH2'-' < Ptotal" The Goro-goro Member U includes at least four eruption centres (as defined by bulk trace element ratios): in south Bacan (mostly HBIAN), the Goro-goro area (mostly TPAN), and the north Mandioli and the Kaputusan areas (largely HPAN and HBAN). The Mandioli group are shoshonitic rocks, which may indicate eruption in an extensional setting within a dominantly convergent zone, such as a strike-slip zone where similar rocks commonly occur (e.g. Ellam et al. 1988). K-Ar ages indicate that these centres ~ were erupting from the Late Miocene (c. 7.5 Ma) to the Pliocene (c. 2.2 Ma). Each of these centres appears to have been active for c. 2 Ma, with a history of multiple eruptions. The Upper MioceneUpper Pliocene Pacitak Member consists of •
490
J.F.A. MALAIHOLLO ,~ R. HALL
Discussion
reworked pyroclastic and volcaniclastic material from the Goro-goro Member, deposited in an oxygen-rich, nearshore, shallow marine environment. The Upper Miocene-Lower Pliocene Mandioli Member consists of wackestones and packstones which formed fringing coastal reefs. The Kaputusan Formation is interpreted to be the product of the eastward subduction of the Molucca Sea plate under Halmahera, and can be correlated with the Weda Group of Halmahera and the Woi Formation of Obi (Hall et al. 1995a; A l i & Hall 1995).
The history and motion of the Philippine Sea plate has been complex, although recent palaeomagnetic results are contributing to a clearer picture. The work of Hall et al. (1995a, b) suggests that between c. 50-40 Ma the plate experienced a c. 50 ° clockwise rotation with a southward translation; no rotation between 40 to 25 Ma, and from 25 to 0 Ma the plate rotated clockwise 35 ° with northward translation. The first rotation may be related to the change of motion in the Pacific plate (Clague & Jarrard 1973) at about 42 Ma. Australia was moving north throughout the Tertiary and we interpret the regional unconformity at c. 22 Ma as the result of collision of Australia with a volcanic arc at the southern edge of the Philippine Sea plate (discussed below). This may have caused or contributed to renewed clockwise rotation of the Philippine Sea plate.
Q u a t e r n a r y deposits Although currently there are no active volcanoes on Bacan, Quaternary volcanic rocks are present. These are fresh olivine-, pyroxene- and plagioclase-phyric basalts. Mineral and whole rock geochemistry indicate that these are arc basalts which erupted through a quartz-rich basement (Sibela Continental Suite). There is evidence of a withinplate basalt component from whole rock and mineral chemistry, possibly indicating magmatism related to movements along the sinistral Sorong fault. This is supported by the linear distribution of the volcanic centres, and in accordance with the view of Silitonga et al. (1981). Detritus from the Sibela Continental Suite can be found solely in the Quaternary deposits, indicating that it has become available for erosion only recently, suggesting extremely fast rates of uplift (c.
Pre-Tertiary and early Tertiary evolution of Bacan There are two types of pre-Tertiary rocks in Bacan: continental crust and ophiolitic/arc material. The continental crust is presumably of Australian margin or micro-continental origin; north of this margin was oceanic crust. Ophiolite/arc material in Bacan records Cretaceous and Oligo-Miocene ages, both of which can be correlated with ages recorded in Halmahera which now forms part of the
1 mm a-l).
I / / I / I / I / / / / / / / I ~
,~/, ,~//I Philippine [~ /, // /, /, /, 1 [Se 0 .Plgte b / / / / / ] /I/I/////~
i/i/ill/i/i/////
// JPhilippi ne I ii i,i i,i ii i i i , / e a Plate / / / / / / /
4 / / ~ P h i l i p p i n e i~, / 4 Sea Plate I / / L./ / / /
eurasia I = , i # / ~
///
/ / / /
IIIIIIIIIII/I//
~ J l l l l l ~
i i i / / / / / / 1 1 1 / / i
-
IIIIIIIIIIlilll .~111111~-~1 ~ ~ / / / :::.::::. ::::::'*::::,.
i:iSi:i:!:!:!:!..
Late Eocene
, ~
i
'Iregion [ . . . . . i
"
I
II
i
/ I
.
::'.::::
Oligocene
Early Miocene
Fig. 5. Simplified tectonic setting of the Bacan region during the Late Palaeogene-Early Miocene. The oceanic crust of the Australian plate was being subducted under the Philippine Sea plate. The Bacan region moved from arc-forearc (Late Eocene) to a backarc (Oligocene) setting and returned to an arc-forearc position (Early Miocene).
GEOLOGY ~ TECTONICS OF BACAN, E INDONESIA
Philippine Sea plate. This implies that the ophiolitic rocks of Bacan were part of the Philippine Sea plate since at least the Cretaceous. Plate tectonic reconstructions show that almost all the oceanic floor of the Mesozoic west Pacific has long since been subducted. The ophiolites of Halmahera and Waigeo, which represent the oldest parts of the Philippine Sea plate, were formed in an intraoceanic arc setting (Ballantyne 1991, 1992) and the Cretaceous Halmahera volcanic arc was situated at sub-equatorial latitudes (Hall et al. 1995a). There is no lithostratigraphic evidence on Bacan for the history of the region between the Cretaceous and Late Eocene. 45-22 Ma
Between the Late Eocene and Early Miocene there was extensive arc activity throughout the BacanHalmahera region. Sukamto et al. (1981) suggested that Oligocene volcanism in the Bacan-Halmahera region was the result of west-dipping subduction from the Pacific side, implying that the region lay on the Eurasian plate. Recent palaeomagnetic evidence (Hall et al. 1995a) from the Tawali Formation shows that the region was part of the Philippine Sea plate. Oceanic crust of the Australian plate was subducted northwards, forming an arc at the southern margin of the Philippine Sea plate and fragments of this arc extend from northern New Guinea, through the Bacan-Halmahera-Waigeo region, into the east Philippines (Rangin et al. 1990; Hall 1995; Hall et al. 1995a). The Oligocene Tawali Formation is interpreted as part of this volcanic arc. The Upper Eocene to Lower Miocene Bacan Formation is genetically related to the Tawali Formation and represents temporal and spatial variations of the same arc (Fig. 5). There are two possible correlatable volcanic arcs in the east Indonesia-west Pacific area: the OligoceneMiocene volcanic rocks of Irian Jaya and Papua New Guinea (e.g. Pigram & Davies 1987) or the Eocene-Oligocene Palau-Kyushu remnant arc (Karig 1975; Sutter & Snee 1980). The PalauKyushu and the West Mariana ridges are associated with opening of the Parece Vela Basin (at 3017 Ma) and these remnant arcs are attributed to subduction of Pacific ocean (e.g. Uyeda & BenAvraham 1972; Seno & Maruyama 1984) and are therefore different from the arc represented by the Tawali Formation. Correlation of volcanic rocks in Bacan and the Bird's Head was suggested by Van Bemmelen (1949) and Verstappen (1960). The arc volcanic rocks of north Irian Jaya, for example the Arfak (Ratman & Robinson 1981; Pieters et al. 1982), Batanta (Sanyoto et al. 1985), Yapen (Atmawinata et al. 1989) and Mandi Formations
491
(Pieters et al. 1989) and Papua New Guinea, for example the Bismarck Volcanic Province (Dow 1977), are interpreted to be related to subduction between the Australian and Pacific plates (Dow 1977; Pigram & Davies 1987) and may be a continuation of the Bacan-Tawali Formation volcanic arc. 22 Ma unconformity
A collision between the Australian margin and an island-arc has been implied or suggested by many authors (e.g. Dow 1977; Jaques & Robinson 1977; Pieters et al. 1983; Pigram & Davies 1987), although suggestions of the age of this collision vary. Pigram & Symonds (1991) review estimates of its age ranging from Eocene to Late Miocene; for east New Guinea they argue for a collision unconformity at c. 30Ma. Charlton et al. (1991) tentatively suggested a mid-Oligocene age for collision on Waigeo, based on a poorly defined Upper Oligocene age of the Mayalibit Formation. However, new isotopic dating of volcanic rocks from this formation indicates a narrow age range, of uppermost Oligocene-lowermost Miocene (our unpublished results). At the south end of Mayalibit Bay on Waigeo these volcanic rocks dip at up to 40 ° beneath Miocene limestones, suggesting they are below the unconformity. The presence of the Australian continental rocks on Bacan could be due to: (1) Early Miocene collision; (2) Early Miocene collision followed by Neogene strike-slip translation; or (3) Pliocene or younger pure strike-slip translation. Although there is no indisputable evidence, the proximity of the Sibela Continental Suite to the arc, interpreted to be the result of subduction of the Indo-Australian plate under the Philippine Sea plate, and the juxtaposition of the continental rocks with the ophiolite interpreted to represent Philippine Sea plate material favours a collisional interpretation (1 or 2). The character of the Nusa Babi Monzodiorite, suggesting post-collisional melting of a continental source, and the evidence of a continental crustal contribution to Late Miocene volcanic rocks also suggest the presence of a continental basement beneath central and south Bacan by the early Neogene. On Bacan there is a regional unconformity of Early Miocene age which is interpreted as resulting from the collision of Australian and Philippine Sea plates (Fig. 6). There is a change from folded, metamorphosed arc rocks to unmetamorphosed carbonates and a possible stitching intrusion of Philippine Sea plate rocks (the Bacan and Tawali Formations) and Australian plate rocks (Sibela Continental Suite) by the Nusa Babi Monzodiorite which is of Early Miocene age. This interpretation
492
J.F.A. MALAIHOLLO • R. HALL //////////////////// / ///~/ /~/~/ /~l ~P~h'i'l' ip i e ] p n ~ / / / / /, , // / // / // /// /// /// /// /// /~Sea //// Platel/////
,/
~-.i~/Iregion
," !
.:.:....:.:.:.:.::..:. + 1~ : ! : ! $t . - '.-.mi!~!ii!!~iiiiiiiiir,~+ : i~: ! : ! : +! : ! : ! :i -: ,...................
,:-:-.*.:o:-.'.;..'-,'..'.."
~r ................
':':':':':':':':':':':';'. .-'.:.:.:.;-'.:.-'.'..'.'..:.:.:.:.;..
+
===================== ,',.'.
x
"1";'.": +
;.;.:.:.,.;
Fig. 6. Simplified tectonic setting of the Bacan region during the Early Miocene. Continuation of subduction culminated in collision and thrusting of the continental crust, under the arc-forearc region of the Philippine Sea plate. Following collision, the Philippine Sea plate rotated westward.
is supported by a similar lithostratigraphy throughout the Bacan-Halmahera region (Hall et al. 1988 1991), where most of the Lower to Middle Miocene is represented by carbonate rocks deposited above Upper Eocene-Lower Miocene arc rocks of Philippine Sea plate affinity and rocks of Australian continental affinity. It is therefore concluded that the collision in east Indonesia occurred in the Early Miocene and the authors agree with Ali & Hall (1995) who suggest that much of the evidence from east Indonesia and New Guinea can be understood in terms of pre-Miocene intra-oceanic arc tectonics, an arc-continent collision at about 22 Ma, and Neogene strike-slip tectonics. 19-15 Ma
After the Early Miocene collision, the region was uplifted and deposition of shallow marine carbonates followed. A tectonically quiescent period is envisaged at this time. Localized uplift contributed influxes of volcaniclastic material onto the carbonate platform (Fig. 7), probably derived from the pre-Miocene formations. Initiation of Molucca Sea subduction and s u b s e q u e n t arc v o l c a n i s m
Northward movement of Australia during the Neogene occurred without subduction at the boundary of the Australian and Philippine Sea plates, which was the left-lateral strike-slip Sorong fault system. However, arc volcanism did result from the eastward subduction of the Molucca Sea
plate under Halmahera (the Philippine Sea plate) at the Halmahera trench (Fig. 8). On Bacan there is a local unconformity of Late Miocene age, and in places the Upper Miocene Kaputusan Formation rests directly on the Bacan Formation. The Kaputusan Formation is the equivalent of other Upper Miocene arc sequences on Halmahera and Obi. Although the oldest isotopic age obtained from the Kaputusan Formation is c. 7.5 Ma, there was a thermal event at c. 15 Ma recorded by widespread resetting of isotopic ages in the Bacan Formation and rocks on Obi. The Saleh Diorite, interpreted as a precursor to the Kaputusan volcanics, was intruded at c. 15 Ma. The oldest isotopic ages obtained from volcanic sequences on Obi are c. 12 Ma (Baker & Malaihollo 1996). Thus, initiation of subduction possibly started at c. 15 Ma and the first volcanic products were erupted at c. 12Ma. Late Neogene-Recent
At c. 3 Ma there was more than 60 km shortening between east and west Halmahera, attributed to movement along the Sorong fault (Nichols & Hall 1991). Splays of the Sorong fault running through the Bacan region may have contributed to the ending of Kaputusan volcanism. Quaternary volcanism was later reactivated along the fault splays (Fig. 9). These faults are also responsible for the shaping of Bacan coastlines and the creation of deep basins surrounding Bacan. The culmination of collision processes in the Molucca Sea will eventually result in the transfer of the Bacan-
GEOLOGY & TECTONICS OF BACAN, E INDONESIA
493
~ / / / / / / / / / / / I;~:::::::::::::::~////l Philippine 1/ / / / / ] S e a Plate[ / I-.----------4////) ...... / 1::74=. . . . . ~ , . , I : = ~ / / / / " I I . Z _ ~ / / /
#,! r~
~ ~"q,',',1~,,~ I
~
I
I
I
I~
¢-
~
~
~
" ~
~.
~
~ }
~, , ...... 2,',',',',','," t
I
I
I
I
~'~I.~L~ ° " ~
-
I
1
i
i
"'""-41 -,,a,~
1 : ~ ~ I I / / / / / / / / / / / / / l---~--'---J/..,~..///////// i
I • i, ,
volcaniclastic
I ,
i
LOCQI ~
Ie
I ,
i
~
i
I ,
~
i
I ~'~-~.# ~
,
~ ,_.t
I
I
~
,
,
.%1
I ,
~
I ,
i
~
i
I ,
~
i
[',
~
~' {
I
'
~
_
t
~
+
~
% +~J~l-~,~_,_%~_~o?
I ~ " ~ " ~-",ae,-".,. ~':':.:.:, i : ~ . . , a k T d ] ~ i ' % ~ : ; : ; : ; :
..if
.%:.:
~
~
I
ioput - "
~
~
~
+
~
+
~
'
!
~
i
::iiii
Fig. 7. Simplified tectonic setting of the Bacan region during the Middle Miocene. Northward movement of Australia was accommodated by a transform boundary, which moved north with Australia (Hall e t al. 1995a, b). The Philippine Sea plate was rotating clockwise due to subduction on all sides. There was widespread development of platform carbonates in the Bacan region, away from the plate boundary.
H a l m a h e r a region f r o m the edge of the Philippine Sea plate to the Eurasian margin (Hall & Nichols 1990).
Implications for regional tectonics This study has p r o v i d e d n e w geological data and elucidated the tectonic d e v e l o p m e n t of the Bacan
region. In a regional context, it has contributed to an u n d e r s t a n d i n g o f the tectonic history o f the convergent zone b e t w e e n the Australian, Philippine Sea and Eurasian plates in east Indonesia. S o m e o f the m o s t important implications of this w o r k are: (1) M o s t of Bacan is interpreted as part o f the Philippine Sea plate. Cretaceous links are indicated by the presence of the Sibela ophiolite,
'~ ./ ! ~lZl
~. ~,
,&:~!~F~i~."i!~!##!~7 "~;~i~i;ii;;i;!!;!!!;/'
I " VolcQniccentres I~ ~g,%%o~,oMb
irn~in
~
F / ]Philippine I I¢ " 4seQ PIQtel E / / "" " "" " F
/
/
/
/
/
/
/
/
A:~:i:i&. . . . . . . .
.:
!!i!i.:i!i.:i::"..i!i!i!i!i!i!i!i!i!:==================================================== !:!:!:~:!:..':!:!:..':!:!:..':!:!:!:!:!: ~iiii~iiiiiiiiiiiiiiiii~-
iiiiiii i!ii!!!iiiiiii
::;:::::;St::::::;:::::;::::::::::, :,:,:,:,:,:,;,:°:,:,:°:,:,:,:,:,:°:,:,,
+ : :::::::::::::::::"
~ i:'.i~i!i!iiiii~ii~i~ii~i~i:.. ~!~!~!~!i
+ +
~x?-~~' +' 3~
Fig. 8. Simplified tectonic setting of the Bacan region during the Late Miocene. Westward subduction of the Molucca Sea plate under the Philippine Sea plate started at c. 15 Ma, probably from the south and moved northward. Volcanicity started in Bacan at c. 7.5 Ma. Shallow marine sequences and fringing reefs formed around volcanic centres.
494
J.F.A. MALAIHOLLO ~; R. HALL
Fig. 9. Simplified tectonic setting of the Bacan region during the Late Pliocene-Quaternary. The Sorong fault system transform boundary cut through the Bacan region, contributing to the cessation of Miocene-Pliocene volcanism. Quaternary volcanism erupted through continental crust along strands of the Sorong fault. Faults controlled coast lines and basins developed between strands of the fault.
(2)
(3)
(4)
(5)
and younger connections are indicated by the similarity of Eocene-Oligocene arc sequences which also record a similar palaeomagnetic history (Hall et al. 1995a). The Bacan and Tawali Formations are most likely to be arc-related sequences produced as a result of subduction of oceanic crust of the Australian plate, probably directly linked to the Indian Ocean, under the Philippine Sea plate. The Sibela Continental Suite arrived in the Bacan region after collision of Australia or a rifted fragment of Australia with the Philippine Sea plate. The collision of Australia with the Philippine Sea plate can be dated at c. 22 Ma in the Bacan region. Younger Neogene-Recent movements on strands of the Sorong fault have modified the geology of the region and permitted emplacement of younger volcanic rocks.
Several different models for the tectonic development of NE Indonesia have been proposed, e.g. the indentor model of Charlton (1986); the terrane model of Silver et al. (1985); and the marginal basin model of Ben-Avraham (1978), and all were based upon limited geological data. The BenAvraham model considers Bacan-Halmahera as part of the Australian continent, which is wrong as almost all of the region has been part of the Philippine Sea plate since Cretaceous. Both the Charlton and Silver et al. models imply that the Sibela continental block was translated by postMiocene strike-slip motion to the Bacan region,
which has been disputed by results of this study. Ali & Hall (1995) and Hall (1995) offer alternative interpretations of the region which incorporate the results of this and other recent geological and geophysical studies. In a wider context, if the terrane definition of Howell et al. (1985) is followed, the Sibela Continental Suite, the Sibela ophiolite, the Bacan, Tawali and Kaputusan Formations could be considered as five different tectono-stratigraphic terranes. This may lead to interpretations such as those of Struckmeyer et al. 1993, who apparently assign separate histories to each of these 'terranes'. Karig et al. (1986) argued that rocks in the north Philippines, which have a similar character to the Sibela Continental Suite, Sibela ophiolite, Bacan and Kaputusan Formations, were separate allochthonous terranes juxtaposed by strike-slip faulting. Whilst not arguing against their interpretation, this study has shown that the juxtaposition of rocks of different character does not necessitate the notion of allochthonous terranes. In the Bacan region, the only fragment that may genuinely be an allochthonous terrane is represented by the Sibela Continental Suite. For the most part, the region has always been at the edge of the Philippine Sea plate, and different 'tectono-stratigraphic terranes' reflect different tectonic regimes at the edge of the plate. The use of stratigraphical similarities is often used in the literature to correlate tectono-stratigraphic terranes and ultimately tectonic history (e.g. Hamilton 1979; Pigram & Panggabean 1984). Using this technique, the Bacan region could be
GEOLOGY • TECTONICS OF BACAN, E INDONESIA correlated with parts of Papua N e w G u i n e a (Oligocene pillow lavas against continental basement; Dow 1977), the Zamboanga Peninsula of Mindanao (high grade continental rocks juxtaposed against meta-ophiolite unconformably overlain by Middle Miocene volcanic rocks; Rangin 1991) and north Philippines (see above; Karig et al. 1986). This type of correlation may lead to erroneous interpretations since each of these regions has different tectonic affinities (Bacan was and is still part of the Philippine Sea plate; North Papua New Guinea was part of the Philippine Sea plate, but now is part of the Australian plate; Z a m b o a n g a and north Philippines were and are still part of the Eurasian plate). Similarities in the stratigraphy are consequences of similar tectonic histories (areas recording the collision of continental with oceanic plate), and one should be cautious in correlation and interpretation of m o v e m e n t of terranes along strike-slip fault. All too often tectonic reconstruc-
495
tion of a region is based on meagre geological data, such as in the Bacan region, resulting in a diversity of interpretations. Although this study has provided the most comprehensive geological dataset from the Bacan region, the history of the region in the Early M i o c e n e and before the E o c e n e is still unclear. Bearing this in mind, one should be cautious in interpreting the tectonic evolution of other, similarly complex, older regions. This work was supported by NERC award GR3/7149, grants from the Royal Society and the University of London SE Asia Geological Research Group, and financial assistance from Amoco Production Company. We thank S. J. Baker, P. D. Ballantyne, E T. Banner, T. R. Charlton, E. M. Finch, G. J. Nichols and S. J. Roberts for their contributions and discussion, and C. C. Rundle and D. C. Rex for guidance and help with isotopic dating. Logistical assistance was provided by GRDC, Bandung and the Director, R. Sukamto with field support by D. A. Agustiyanto, A. Haryono, and S. Pandjaitan.
References
ALl, J. R. & HALL, R. 1995. Evolution of the boundary between the Philippine Sea Hate and Australia: palaeomagnetic evidence from eastern Indonesia. Tectonophysics, 251, 251-275. ATMAWINATA, S., RATMAN, N. & PIETERS, P. E. 1989. Geological map of the Yapen Sheet, lrian Jaya. Geological Research and Development Centre, Bandung, Indonesia. BAKER, S. J. & MALAIHOLLO,J. E A. 1996. Dating of Neogene igneous rocks in the Halmahera region: arc initiation and development. This volume. BALLANTYNE, P. D. 1991. Petrological constraints upon the provenance and genesis of the East Halmahera Ophiolite. Journal of Southeast Asian Earth Sciences, 6, 259-269. - 1992. Petrology and geochemistry of the plutonic rocks of Halmahera ophiolite, eastern Indonesia, an analogue of modern oceanic forearcs. In: PARSON, L. M., MURTON, B. J. & BROWNING, E (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publication, 60, 179-202. - & H A L L , R. 1990. The Petrology of the Halmahera Ophiolite, Indonesia: an early Tertiary forearc. In: MALPAS, J., MOORES, E. M., PANAYIOTOU,A. & XENOPHONTOS,C. (eds) Ophiolites: Oceanic Crustal Analogue. Proceedings of the Symposium Troodos 1987. The Geological Survey Department: Ministry of Agriculture and Natural Resources, Nicosia, Cyprus, 461-476. BEN-AVRAHAM,Z. 1978. The evolution of marginal basins and adjacent shelves in east and Southeast Asia. Tectonophysics, 45, 269-288. BROUWER, H. A. 1923. Bijdrage tot de geologie van het eiland Batjan. Jaarboek Mijnwesen Nederlandsch Oost-lndie 1921, 50, 73-105. CHARLTON, T. R. 1986. A plate tectonic model of the eastern Indonesia collision zone. Nature, 319, 394-396.
--,
HALL, R. & PARTOYO, E. 1991. The geology and tectonic evolution of Waigeo Island, NE Indonesia. Journal of SE Asian Earth Sciences, 6, 289-297. CLAGUE, D. A. • JARRARD, R. I. 1973. Tertiary Pacific plate motion deduced from Hawaiian-Emperor chain. Geological Society of America Bulletin, 84, 1135-1154. Dow, D. B. 1977. Geological synthesis of Papua New Guinea. Bureau of Mineral Resources Australia Bulletin, 201. ELLAM, R. M., MENZIES, M. A., HAWKESWORTH,C. J., LEEMAN, W. P., RosI, M. & SERRI, G. 1988. The transition from calc-alkaline to potassic orogenic magmatism in the Aeolian Islands, Southern Italy. Bulletin of Volcanology, 50, 386-398. HAKIM, A. S. & HALL, R. 1991. Tertiary Volcanic rocks from the Halmahera Arc, Eastern Indonesia. Journal of SE Asian Earth Sciences, 6, 271-287. HALL, R. 1996. Reconstructing Cenozoic SE Asia. This volume. -& NICHOLS, G. J. 1990. Terrane Amalgamation in the Philippine Sea Margin. Tectonophysics, 181, 207-222. - - , ALI, J. R., ANDERSON,C. D. & BAKER, S. J. 1995a. Origin and motion history of the Philippine Sea Plate. Tectonophysics, 251, 229-250. --, AUDLEY-CHARLES,M. G., BANNER,F. T., HIDAYAT, S. & TOBING, S. L. 1988. The basement rocks of the Halmahera region, east Indonesia: a Late Cretaceous-Early Tertiary forearc. Journal of the Geological Society, London, 145, 65-84. , FULLER,M., ALl, J. R. & ANDERSON,C. D. 1995b. The Philippine Sea Plate: Magnetism and Reconstructions. In: TAYLOR,B. & NATLAND,J. H. (eds) Active Margins and Marginal Basins: A Synthesis of Western Pacific Drilling Results. American Geophysical Union Monograph, 88, 371-404.
496 --,
J . F . A . MALAIHOLLO & R. HALL
NICHOLS, G. J., BALLANTYNE,P. D., CHARLTON,T. & ALl, J. 1991. The character and significance of basement rocks of the southern Molucca Sea region. Journal of SE Asian Earth Sciences, 6, 249-258. HAMILTON, W. 1979. Tectonics of the Indonesian Region. US Geological Survey Professional Paper, 1078. HARLAND, W. B., ARMSTRONG,R,. L., COX, A. V., CRAIG, L. E., SMITH, A. G. & SMITH, D. G. 1990. A geologic time scale 1989. Cambridge University Press. HOWELL, D. G., JONES, D. L. & SCHERMER,E. R. 1985. Tectonostratigraphic Terranes of the Circum-Pacific Region. In: HOWELL, D. G. (ed.) Tectonostratigraphic Terranes of the Circum-Pacific Region. Circum-Pacific Council for Energy and Mineral Resources Earth Science Series, 1, 3-30. JAQUES, A. L. & ROBINSON, G. P. 1977. The continent/ island-arc collision in northern Papua New Guinea. BMR Journal of Australian Geology & Geophysics, 2, 289-303. KARIG, D. E. 1975. Basin genesis in the Philippine Sea. Initial Reports of the Deep Sea Drilling Program, 31, 857-879. , SAREWITZ, D. R. & HAECK, G. D. 1986. Role of strike-slip faulting in the evolution of allochtonous terranes in the Philippines. Geology, 14, 852-855. LANPHERE, M. A. & DALRYMPLE, G. B. 1976. Identification of excess 4°Ar by the 4°Ar/39Ar age spectrum technique. Earth and Planetary Science Letters, 32, 141-148. LfNTHOUT, K., HELMERS, H. & ANDRIESSEN, P. A. 1991. Dextral strike-slip in Central Seram and 3-4. 5 Ma Rb/Sr ages in pre-Tertiary metamorphic related to Early Pliocene counterclockwise rotation of the Buru-Seram microplate (E. Indonesia). Journal of SE Asian Earth Sciences, 6, 477-493. MALAIHOLLO,J. F. A. 1993. The Geology and Tectonics of the Bacan Region, Eastern Indonesia. PhD thesis, University of London. MITCHELL, A. H. G., HERNANDEZ,E & DELA CRUZ, A. P. 1986. Cenozoic evolution of the Philippine Archipelago. Journal of SE Asian Earth Sciences, l, 3-22. NICHOLS, G. J. & HALL, R. 1991. Basin formation and Neogene sedimentation in a back arc setting, Halmahera, eastern Indonesia. Marine and Petroleum Geology, 8, 50-61. PIETERS, P. E., HAKIM, A. S. & ATMAWlNATA, S. 1982. Preliminary Geological Map of the Ransiki Quadrangle, lrian Jaya. Geological Research and Development Centre, Bandung, Indonesia. , HARTONO,U. & AMRI, CH. 1989. Geological Map of the Mar Sheet, lrian Jaya. Geological Research and Development Centre, Bandung, Indonesia. .... , PIGRAM, C. J., TRAm, D. S., Dow, D. B., RATMAN,N. & SUKAMTO, R. 1983. The stratigraphy of western Irian Jaya. Geological Research and Development Centre Bulletin, 8, 14-48. PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and the accretion history of the New Guinea orogen. BMR Journal of Australian Geology & Geophysics, 10, 193-211. & PANGGABEAN,H. 1984. Rifting of the northern
-
-
margin of the Australian Continent and the origin of some microcontinents in Eastern Indonesia. Tectonophysics, 107, 331-353. & SYMONDS, P. A. 1991. A review of the timing of the major tectonic events in the New Guinea Orogen. Journal of SE Asian Earth Sciences, 6, 307-318. --, DAVIES, P. J., FREARY, D. A., SYMONDS, P. A. & CHAPRONIERE, G. C. U. 1990. Controls on Tertiary Carbonate Platform evolution in the Papuan Basin: new play concepts. In: CARMAN, G. J. & Z. (eds). Petroleum Exploration in Papua New Guinea. Proceedings of the First PNG Petroleum Convention, 185-195. RANGIN, C. 1991. The Philippine Mobile Belt: a complex plate boundary. Journal of SE Asian Earth Sciences, 6, 209-220. --, JOLIVET,L. & PUBELLIER,M. 1990. A simple model for the tectonic evolution of southeast Asia and Indonesia region for the past 43 m.y. Bulletin de la Socidtd ggologique de France, 8 VI, 889-905. RATMAN, N. & ROBINSON,G. P. 1981. Geological Map of the Manokwari Quadrangle, lrian Jaya. Geological Research and Development Centre, Bandung, Indonesia. SANYOTO, P., PIETERS, P. E., AMRI, CH., SIMANDJUNTAK, W. & SUPRIATNA,S. 1985. Preliminary Geological Map of Parts of the Sorong, Kasim, west Waigeo and Misool Quadrangles, lrian Jaya. Geological Research and Development Centre, Bandung, Indonesia. SENO, T. & MARUYAMA, S. 1984. Paleogeographic reconstruction and origin of the Philippine Sea. Tectonophysics, 102, 53-84. SILITONGA, P. H., PUDJOWALUJO,H. & MOLLAT, H. 1981. Geological reconnaissance and mineral prospecting on Bacan island (Moluccas, Indonesia). In: BARBER, A. J. & WmYOSUYONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Indonesia, Special Publication, 2, 373-81. SILVER, E. A., GILL, J. B., SCHWARTZ,H., PRASETYO,H. & DUNCAN, R. A. 1985. Evidence for a submerged and displaced continental borderland, north Banda Sea, Indonesia. Geology, 13, 687-691. STRUCKMEYER,H. I. M, YEUNG, M. & PIGRAM,C. J. 1993. Mesozoic to Cainozoic plate tectonic and palaeogeographic evolution of the New Guinea Region. in: CARMAN, G. J. & CARMAN, Z. (eds) Petroleum Exploration and Development in Papua New Guinea. Proceedings of the Second PNG Petroleum Convention, 261-290. SUKAMTO, R., APANDI, T., SUPR1ATNA, S. & YASIN, A. 1981. The geology and tectonics of Halmahera Island and surrounding areas. In: BARBER, A. J. & WIRYOSUYONO,S. (eds.) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Bandung, Indonesia, Special Publication, 2, 349-62. SUTTER, J. F. & SNEE, L. W. 1980. K/Ar and 4°Ar/39Ar Dating of Basaltic Rocks from Deep Sea Drilling Project Leg 59. Initial Reports of the Deep Sea Drilling Project, 59, 729-734.
-
-
GEOLOGY •
TECTONICS OF BACAN, E INDONESIA
UYEDA, S. & BEN-AVRAHAM, Z. 1972. Origin and development of the Philippine Sea. Nature Physical Sciences, 240, 176-178. VAN BEMMELEN, R. W. 1949. The Geology of Indonesia. Volume Ia, General Geology. Government Printing Office, The Hague. VERSTAPPEN, H. TH. 1960. Geomorphological observa-
497
tions on the north Moluccan - northern Vogelkop Island Arcs. Nova Guinea Geology, 3, 11-38. VROON, P. Z. 1992. Subduction of Continental Material in the Banda Arc, Eastern Indonesia. PhD thesis, University of Utrecht. YASIN, A. 1980. Geologic Map of the Bacan Quadrangle, North Maluku. Geological Research and Development Centre, Bandung, Indonesia.
Dating of Neogene igneous rocks in the Halmahera region: arc initiation and development SIMON
BAKER
& JEFFREY
MALAIHOLLO
SE Asia Research Group, Department of Geological Sciences, University College, London WC1E 6BT, UK Abstract: Potassium-argon ages of Neogene to Recent igneous rocks from the Halmahera region
record a history of intra-oceanic arc development since the late Middle Miocene following an earlier phase of collisional plutonism. Arc formation from the Middle Miocene onwards was due to the east-directed subduction of the Molucca Sea plate beneath the Philippine Sea plate as it arrived at the Eurasian margin. The distribution of ages within the Neogene arc indicates a northward migration of volcanic activity during the Late Miocene to Pliocene. Results of the dating work show that after collision with the Australian margin at c. 22 Ma there was a period of volcanic quiescence and limestone deposition before a new arc formed. This arc began erupting at around 11 Ma on Obi as a result of subduction of the Molucca Sea plate. Initiation of subduction is thought to have occurred around 15-17 Ma and may have been responsible for disturbing potassium-argon ages of pre-Neogene rocks. Dates from fresh rocks show that the volcanic front migrated northwards through Bacan and Halmahera throughout the Late Miocene to Early Pliocene. Limestone deposition was curtailed as arc activity migrated north while volcanism died out from the south. No Neogene volcanism younger than 8 Ma is observed in the Obi area while on Bacan subduction-related volcanism ceased at c. 2 Ma. Late Pliocene crustal deformation caused a 30-40 km westward shift of the volcanic front. Quaternary volcanic rocks exposed in Bacan and the extreme south of Halmahera are not direct products of subduction but, rather, display geochemical characteristics of both subduction and fault-related magmatism. These volcanic rocks are distributed along splays of the Sorong fault system. The formation and propagation of the Halmahera arc is a consequence of the clockwise rotation of the Philippine Sea plate as the southern edge moved across the northern Australian margin and impinged on the east Eurasian margin. The ages of initiation of volcanism and subduction track the developing plate boundary as subduction propagated northwards.
The principal islands of the Halmahera group (Halmahera, Bacan and Obi) lie in northeastern Indonesia in the province of Maluku, straddling the equator between 127°E and 129°E. These islands lie at the junction of three major plates (the Philippine Sea plate, the Australian plate and the Eurasian plate) where the A l p i n e - H i m a l a y a n and the C i r c u m - P a c i f i c orogenic belts meet. The present-day tectonics are therefore complex (Fig. 1). These belts meet in the region of the Sorong fault system which is responsible for transferring crustal fragments of Philippine Sea and Australian origin into the complex of island arcs and small ocean basins that forms the Eurasian margin. Present-day tectonics are the result of the northward m o v e m e n t of continental Australia (Australia and New Guinea) into the Pacific region throughout the Tertiary. Clockwise rotation of the Philippine Sea plate since c. 25 Ma led to the development of the Sorong Fault Zone as it collided with the northern Australian margin (Hall et al. 1995a, b). Current motion between these two plates is taken up by sinistral m o v e m e n t on the Sorong fault system in northern New Guinea.
The currently active H a l m a h e r a arc, at the eastern edge of the Molucca Sea (Hamilton 1979; Moore & Silver 1983), formed as a consequence of eastward subduction of the Molucca Sea plate beneath the Philippine Sea plate as the latter rotated clockwise. Geophysical evidence from earthquake data indicates a seismic zone dipping at c. 45 ° to the east to depths of about 200 k m (Cardwell et al. 1980). To the south the subduction zone appears to be terminated by a strand of the Sorong fault system just north of the eastern tip of Mangole (Sula Platform). Opposing the Halmahera arc is the Sangihe-north Sulawesi arc approximately 250 k m to the west which is the product of westward subduction o f the Molucca Sea plate; beneath the Sangihe arc the slab dips at 55-65 ° and reaches a depth of 600 k m (Cardwell et al. 1980). The oldest rocks known from the Sangihe arc are of early Middle Miocene age and particularly voluminous arc activity occurred between 5 and 14 Ma (Hamilton 1979). The present Halmahera arc lies north of the equator and is built upon a Neogene arc that extended from Obi northwards through Bacan to
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 499-509.
499
500
S. BAKER & J. MALAIHOLLO
I[ " .+.~. " ~+.! Euraeian Margnn +
.
~_. • . + • +
¥. #
t
7Z'l/'TZ' ,/\
,...
/ ;
-,- MOROTAI
tN
A
. + ~_."
• ~
•
;,-
#
O 0
,'-',',-',', HALMAHEF, A.-I -' ;'."
+/:# ~ ~#" Molucca Sea Collision
N.E
.~ I A
. . . . . . . . .
~
~= % /
<
\-~ "-; -~-~-;
O
Phnlnnnnne -',',r r .: . ' - Sea Rate-; :'\-; :-i :w~,J~o.o°:_
-
Complex -,
"-
z -
-.
"- " ~ -
\
\
~.÷~'~"~9~~b~C~UINP_A
~ustra -1,50 km Fig. 1. Locationmap showingthe principaltectonic elements and islands of the study area.
north Halmahera (Hall et al. 1988b; Hall & Nichols 1990). This paper describes the Neogene to Pliocene development of this arc based on potassium-argon and petrographic studies of the volcanic rocks. Other Neogene rocks dated by the potassium-argon method are described and placed in a plate tectonic context.
Stratigraphy The Tertiary stratigraphy of the region is summarized in Fig. 2. The pre-Neogene geology of the region consists of a basement of ophiolitic and arcrelated rocks (Hall et al. 1988a; 1991; Ballantyne 1990) overlain by and imbricated with, arc volcanic, plutonic and sedimentary rocks of late Mesozoic to Eocene age (Hall et al. 1988b; Hakim 1989; Ballantyne 1990; Hakim & Hall 1991). Following an Eocene unconformity (possibly linked with the Pacific plate reorganization at this time) a Late Eocene arc developed with associated basaltic pillow lavas and volcaniclastic turbidites (Hall et al. 1991; Malaihollo 1993); this arc ceased activity in the earliest Miocene and its termination is marked by a regional unconformity. Middle Miocene limestones were deposited throughout
the region during a period of stability before the development of the Neogene arc and associated sedimentary rocks. Deformation in the Pliocene is marked by another unconformity. Continental metamorphic rocks representing Australian crust and possibly of Palaeozoic or younger age (Hamilton 1979; Malaihollo 1993) are juxtaposed against ophiolitic rocks in the southern part of the region. Detailed stratigraphic descriptions of preNeogene and Neogene rocks are given in Hall et al. (1988a, b, 1991), Hakim & Hall (1991) and Roberts (1993).
Analytical methods and data presentation Samples were selected for potassium-argon dating primarily on the basis of their freshness and the presence of suitable potassium-bearing phases. Mineral separates were used; amphiboles and micas were the preferred minerals. Whole rock analyses on fresh volcanic rocks provided good quality results (confirmed by biostratigraphic dating in many areas) with small errors. Biostratigraphic dating gives confidence that young, fresh rocks with a potassium rich, glassy residuum provide reliable isotopic ages. Sample locations are given
ARC DEVELOPMENT IN THE HALMAHERA REGION
501
HALMAHERA
(& KAYOA) i¢.aya(~a
BACAN
OBI
vvvvvvvvvvvvvvvv
Ma 0
c;. ~ . ~ . ~.,c:~..
Form,l;,o° ~
Formal;ion
~
d
2
Quaternary
Wcd~ Group
(rn=intyW~e~)
KaFul;u~a~ Format..ion
For'm=l:ion
q .L Vo[can'mla~t.ic~' z- "1 ,4 Ar~ V.olc,,~nic~ ,4 . / " I
Oha Volcanic. Formal;Ion (lh'¢~l;)
Pliocene
~
, volc~nlcla~r.lc~
mI
5
Miocene Rut,a
LIm~B1;one Formal;ion
Tawali Forma'cion
~
Woi 6, Gu~t.l ]< Forrnat.ione
(E=n~)
~ubaim Lim¢~t,onc Formation (Eal~t.)
Format.ion
Tawali Forma¢ion
Bacan Format.ion
Wayamli Formal;Ion (Ea~.)
Fluk
~
F.......
~
Fo~,l;~o~ ~
LIm¢~'f.on~e
23 Oligocene
Air I~,,a~i
35 !lIHI)'//.l'it~-~l~ !...... ,!lh
•
~ri~
?
?
~J I
E~SM8
J
?
56 I~leocene
~5 Fig. 2. A simplified Tertiary stratigraphy for the three principal islands of the region studied.
in Fig. 3 whilst analytical results are presented in Table 1. Samples for potassium-argon analysis were crushed and sieved to yield a fraction between 300 l.tm or 250 l.tm and 125 ~m (if the sample was small the sieve fraction 500-125 pm was used); this was then washed in deionized water. The washed fraction was either used directly for potassium-argon analysis or further processed to produce mineral separates by a combination of heavy liquid and magnetic separation techniques. The samples were then split for separate potassium and argon analyses. All potassium-argon analyses were carried out at the NERC Isotope Geoscience Laboratory (NIGL) at the BGS site in Keyworth, Nottinghamshire. Potassium analyses were conducted in duplicate using flame photometry against standard potassium solutions and a pure water blank after sample digestion in excess hydrofluoric and perchloric acids. Argon analyses on a separate aliquot were performed by complete sample fusion in a gas extraction line; the sample gases were mixed with an enriched 38Ar spike (isotope dilution method) and cleaned up using a titanium getter and a liquid nitrogen cold trap. The extraction line was coupled to a gas source Micromass 1200 mass spectrometer with a magnetic deflection of 60 ° and sensitive to 3 x 104 Amps/Torr; calculations and a direct printout of the results were provided by VG software using the constants recommended by Steiger & J~tger (1977). The mass spectrometer and spike volume were regularly calibrated by comparison with the International Standard G1-O glauconite (reproducibility results available from the authors on request).
Error calculation and data reduction methods All isotopic ages are quoted with errors at the 95% confidence level (2o). The major factor affecting the error is connected with the correction for atmospheric argon. If the atmospheric component (40Aratm) is small, uncertainties in its measurement will be minor compared to the total 4°At. However, if 4°Aratm is the major component then the same degree of uncertainty will cause large errors in the smaller radiogenic component (40Arrad). For samples with low atmospheric contamination a minimum error of 1% is assumed based on the uncertainty in the calculation of the spike volume. Potassium errors were calculated from the average of at least two analyses and 10 error calculated from the deviation of the average from the actual values obtained, l o error in % was used for age calculation and then converted to 2o error for reporting purposes (see Table 1). For some samples K error is relatively high. This is due to a number of factors namely: matrix effects affecting the flame photometer reading for some samples; undigested sample (opaques) and sample heterogeneity exacerbated by the small size of aliquots that must be used for samples with high K content. All efforts have been made to minimize these problems including triplicate analyses if necessary. For some very young samples 40Arrad is below the limits of detection for the spectrometer so the VG software at NIGL allows an maximum age calculation for such samples by assuming there is less than or equal to 1% radiogenic argon present (the limit of precision for argon analysis). These
502
s. BAKER & J. MALAIHOLLO
Kahatola HB16, A
HB93 HB94 HB95
Ternate~aAt147fA mid0re~lklH2H~m!260 M~aki~~H
~
probable gaps in the spectrum of Neogene ages; iterative cluster analysis of the data and their associated uncertainties was used. First, a density trace was obtained by computing the fraction of data points that fall into a certain interval or window (0.7 Ma was chosen in this case) using a distance weighted kernel estimator (to assess the contribution of each data point in the current window) and assuming a Gaussian distribution for the errors about the data points. Following this the probability of gaps in the suite of ages was calculated by sampling one point from equally spaced locations in the range 0-12 Ma (including points from uncertainty ranges) to produce a trace of 'gap ages' (see below). All age data discussed here are presented in Table 1; sample localities are shown in Fig. 3; results from volcanic rocks are plotted in geographical order in Fig. 4. Ages are described by area, working from south to north. The time scale of Harland et al. (1990) is used.
Pre-arc Neogene ages Bacan
BACAN ~ ~
HR209
0R303
~°~isa ~ ~D42~OJD140
BI ,~J58 Fig. 3. Sample location map. Black triangles represent currently active arc volcanoes.
ages are of limited use and in some cases the use of a larger sample for argon analysis (or of a more K-rich fraction) yields sufficient radiogenic argon for analysis. Isochron plots provided weighted averages for suites of ages in some cases. Two types of isochron plots were used, 4°Ar vs. 4°K and 4°Ar/36Ar vs. 4°K/36Ar, and ages determined from the slope of the line. The intercept of the line with the y-axis qualitatively indicates loss/gain of argon and initial 4°Ar/36Ar ratio respectively. Where potassiumargon ages from a single formation are within error of one another a site or formation mean is quoted along with the standard error on the mean (Table 1) and in these cases this is accepted as the best estimate of the age. Potassium-argon data were also tested statistically to try to identify periods of volcanism and
The oldest Neogene rocks from the region are quartz-monzodiorites of the Nusa Babi intrusions from Bacan. These plugs are distributed throughout eastern Bacan and petrographically are characterized by plagioclase with lesser orthoclase, green amphibole, biotite, quartz and opaques; some samples show chlorite and/or epidote alteration. Trace and major element tectonic discrimination diagrams suggest either an arc or a post-collisional magmatic origin (Malaihollo 1993). Dating of biotite from the type locality (BR80) gave an age of 19.8 _+ 1.6 Ma interpreted to be a crystallization age which is supported by the field observation that these intrusions cut only rocks older than Middle Miocene. Samples from other localities show the effects of later alteration and yield younger potassium-argon ages. Rocks belonging to the Saleh intrusions from Pulau Saleh Kecil (SR6) and south Bacan (B66) are amphibole-rich diorites and chemically distinct from the Nusa Babi intrusions. Petrographic studies indicate an order of crystallization typical of arc rocks; tectonic discrimination diagrams are ambiguous, suggesting either a post-collisional or arc origin. These samples yield Early to Middle Miocene ages (9.3_ 1.5 Ma and 15.1 _+ 1.6 Ma respectively) on amphibole separates interpreted as crystallization ages. Halmahera
In NE Halmahera a minor syeno-gabbro sill (c. 2 m thick, lateral extent probably less than 100 m) was
SR6 B66
P. Saleh Kecil Sibela Mtns.
E Nusa Babi
N Bacan Nr. Labuha Ra R. cent. Bacan P. Mamalaya SW Bacan Bibinoi R. P. Mandioli S Bacan S W Bacan NE Bacan NE coast Bacan N coast Bacan SE Bacan
Diorite Diorite
Monzonite
Hb A n d Hb A n d Hb And Bi px A n d Hb A n d Welded tuff Hb A n d Hb A n d Px And Px And Px And Hb And
Hb A n d Hb A n d
Px And (f) Bi px And And (f) Crystal tuff Px And (f)
S-gabbro S-gabbro S-gabbro
2px A n d Hb px A n d 2px A n d (f) Hb A n d (f) Px A n d Hb A n d (f) Hb A n d
2px A n d (f) Welded tuff Hb A n d (f) Mdiorite
Bi tuff
Rock type
A B
C
A A B B B A A A A A A A
C C
B A A B B
A A A
A A A B B A A
B A A O
A
F
Amp Wr
Bi
Wr Wr Wr Wr Hb, px Wr Hb, px Hb Wr Wr Wr Wr
Wr Wr
Wr, Bi, Px Wr Bi Wr
W Bi Hb
Wr Wr Wr Hb Wr Wr Hb
Wr Wr Hb Wr
Bi
Material
125-250 125-250
125-250
125-250 250-500 250-425 250425 125-250 250-500 125-250 125-250 125-250 250-425 250-425 250-500
500-125 500-125
250-125 250-125 300-125 250-125 300-125
250-125 250-125 250-125
500-125 500-125 500-125 500-125 300-125 500-125 500-125
500-125 250-125 500-125 250-125
250-125
Grain size (~tm)
± 0,13 _+ 0,02 _+ 0.11 _+ 0.01 ± 0.02 ± 0.03 ± 0.01
± 0,02 _+ 0,05 ± 0,02 ± 0.02
± 0.04 ± 0.02 _+0,3 ± 0.26 _+0.3
_+ 0.04 ± 0.02 _+ 0.09 ± 0.04 ± 0.09 _+ 0.03 _+0.06 ± 0,05 _+ 0,03 ± 0.16 _+ 0,01 ± 0.04
0.36 ± 0.01 0.65 -+ 0.01
4.29 _+ 0.09
1.36 1.09 1.54 1.38 1.89 1.63 1.10 2.26 0.97 1.50 1.16 2.10
1.63 + 0,05 2,01 ± 0,03
1.79 0.83 1,64 4,76 1.50
2.77 _+ 0.06 3.42 ± 0.07 0.88 ± 0.02
1,34 1.12 1.60 0.30 1.07 1.12 0.66
0.91 2.68 0.94 0.94
3.79 ± 0,08
% K (2(y error)
1.0611 2.110
0.1024
1.0238 1.2044 2.4686 2.2700 1.0420 0.3493 0.6240 1.0276 0.3237 1.0052 2.2149 0.0584
1.0283 1.232
0.733 1.0418 1.353 0,1403 0.9353
0.5295 0.4168 0.4074
1.1727 1.3133 1.3356 1.5009 1.1572 1.2911 1.0282
1.0018 2.0039 1.3406 0.519
0.1484
Wt for Ar (g)
89.20 75.85
68.74
66.66 87.57 84.74 81.91 92.59 72.16 77.58 81.35 86.48 80.11 58.28 78.64
86.41 78.96
75.1 65.33 35.46 67.48 42.39
72.58 56.22 89.73
55.38 79.97 61.31 82.84 86.42 83.11 86.14
76.42 93.26 91.18 84.34
98.76
± ± ± ± ± ± ±
1.67 4,16 1,94 5,09 6,54 5.10 6.38
_+ 373 ± 15.13 ± 10,57 ± 5.98
1.297 + 2.88 1.723 _+ 1,74 0.359 ± 8.93 Site Mean for Alkaline Sills
0.220 0.203 0.301 0.065 0.239 0.273 0.196
0.076 0,228 0.085 0.109
4°At* (nl g-t) (1~ error)
± 2.42 _+ 7.23 ± 5.74 _ 4.69 ± 3,13 ± 2,98 _+ 3.73 ± 4.57 ± 6,75 +_ 4.19 ± 1.78 ± 6.42
0.132 _+ 8.49 0.131 ± 3,33
3.325 ± 3,87
0.118 0.107 0.153 0.117 0.311 0.304 0.212 0.445 0.207 0.341 0.304 0.612
0.744 ± 3.2l 0.347 _+ 2.16 0.708 +_ 1.16 2.100 ± 2.35 0,669 _+ 1,27 Formation Mean for Woi and Guyuti Fms 0.557 ± 6,63 0.748 ± 3,95 Site Mean for Bisa Volcanic Rocks
4°Ar atm (hi g q ) (%)
± 0.7 ± 0.5 +__0.3 ± 1.3 + 0.5 _+ 0.2
± 0.7 _+ 0.5 ± 1.9 ± 0.7
± 0.5 ± 0.4 _+ 0.4 ± 0.6 _+ 0.8 ± 0.7 _+ 1.0
0.2 0.7 0.5 0.4
<0.4 ± ± ± ±
± 0.1 _+ 0.4 ± 0.3 _+ 0.2 ± 1.1 ± 0.3 ± 0.5 ± 0.5 ± 0.8 +_ 0.8 ± 0.3 ± 1.0
9,3 ± 1.6 15.1 ± 1.6
19.8 ± 1.6
2.2 2.5 2.5 2.7 4.2 4.8 5.0 5.1 5.5 5,8 6.7 7.5
8.8 ± 1.2 9.5 ± 0.8 9.2 ± 0.4
10.7 10.7 11.1 11.3 11,5 11,1
12.0 12.9 10.5 11.8
4.2 4.7 4.9 5.7 5.8 6.3 7.7
2.1 2.2 2.3 3.0
Age Ma (2(~ error)
c = clast from conglomerate or similar, if neither then in situ. Freshness F: A = fresh (no secondary minerals, glass fresh); B = mainly fresh (< 5% modal secondary minerals); C = slightly altered (5-10% modal secondary minerals); O = opacitic alteration of mafic minerals (usually hornblende). Material: Wr = whole rock; Bi = biotite; Musc = muscovite; Hb = hornblende; Px = pyroxene. Others: L O D = limit of detection.
Abbreviations. Sample: dp = duplicate; tp = triplicate. Rock type: A n d = andesite; Bas = basalt; Dac = dacite; Tuff = tuff; Bas and = basaltic andesite; Mdiorite = microdiorite; S-gabbro = syeno-gabbro; f = float,
BR80 dp
Saleh Diorite
BR82 dp BP30 dp BM131dp BM453dp BR203 BM256 BM26 BR174 BM361 BR16 tp BR58 dp BR49
P. Bisa P. Bisa
OR303 OR306dp
Nusa Babi Intrusion
Baean Kaputusan
Bisa
Central Obi NE Obi Central Obi NE Obi N Obi
OJ58 OJ40 dp OD194 OD15 OD42V
NE Halmahera NE Halmahera NE Halmahera
HB93 HB94 HB97
Alkaline Sills
Obi Woi and Guyuti
Central Halmahera Central Halmahera Central Halmahera Central Halmahera Kayoa Island Central Halmahera Central Halmahera
HR705 HR703 HA147A H256 KT5b dp H275 HR711
Central Halmahera N W Halmahera Central Halmahera N W Halmahera
SW Halmahera
Weda Group
HR209
H260 HB16 H258 HR717
Halmahera Quaternary
Location (for localities see Fig. 3)
Kayasa
Sample
Table 1. K-Ar Dates from Halmahera, Obi and Bacan
Formation
504
s. BAKER & J. M A L A I H O L L O
Ma M.Miocerle
NORTH
50UTH
~)
[_
T
I I kA; . . . . .
' ......................
............
~
W
j
~
~
,~v . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~
,
i
.
"
~
~
~
~
................................. ~ . . . . . . . . . . . . . .
o
. . . . . Q u a t e r n a r y Art; V'ol~ania~ . . . . I;;;;; ............................. ;;;:
~
u
g
,, . . . .
u
e
a
I¢~ u'r,u e a n F
n
vvvvvvv,
vJz . . . .
I
VolGani6~ot;k~
I;m. ',T,~:I~
...............
V o l ~ a n i ~ acr, ivit,y a l o n o f a u l t , e l a y ~ P rc~
Halmahera
15a6an
Obi
Fig. 4. Primary potassium-argon ages (and their errors) for Neogene volcanic rocks plotted in geographical order
from north to south. Shaded areas illustrate periods of volcanism within the named formations derived from the statistical treatment described in the text and shown graphically in Fig. 5. Note the younging trend from south to north is particularly marked from Obi, via Bisa, to Bacan. Quaternary volcanic activity is also shown.
found intruding redeposited limestones of Eocene age. The sill is fresh and has two distinct grain sizes, coarse and medium. Analcite occurs as a primary constituent of the groundmass together with alkali and plagioclase feldspar; titanaugite, sodic amphibole and biotite are all intimately intergrown along with titanomagnetite. Dating of three samples (whole rock, biotite and amphibole) yield similar ages which can be reduced to a site mean of 11.8 _ 0.7 Ma interpreted as primary crystallization ages. Arc-related
Neogene
ages
Obi
The Woi Formation in central and northwest Obi consists predominantly of volcanic rocks (lava breccias and flows with pyroclastic rocks) with interbedded volcaniclastic sandstones and breccioconglomerates. Chemically the volcanic rocks are sub-alkalic calc-alkaline arc-related rocks and the majority are andesitic although there is a significant spread into the basaltic andesite and dacite fields (Miyashiro 1978; Cox et al. 1979). Petrographically they are typically plagioclase- and clinopyroxene-phyric, minor phenocryst phases are biotite and rarely amphibole (both commonly show opacitic rims). Feldspar and clinopyroxene are
often associated as glomerocrysts. Phenocrysts are set in a fine-grained groundmass of small clinopyroxenes, opaque grains and plagioclase microlites showing felty, intersertal and, occasionally, trachytic textures. Variable amounts of glass occur in the groundmass, amygdales occur in some areas and are infilled with opaline silica and a variety of fibrous minerals. The lateral deep-water equivalent of the Woi Formation is the Guyuti Formation which consists of debris flow deposits containing andesite and limestone clasts representing a deep marine slope forearc environment (R. Hall & D. Atmo, pets. comm., 1994). Potassium-argon results on rocks from the Woi and Guyuti Formations in central Obi are similar, with small errors. The formation mean for Woi and Guyuti Formation samples is 11.1 _+0.2 Ma. This age is supported by nannofossil and foraminiferal ages from sedimentary rocks within the Woi Formation as well as from limestone clasts within the Guyuti Formation (foraminiferal and nannoplankton dates indicate a late Middle to Late Miocene age range). Neogene volcanic rocks also occur just north west of Obi island on Pulau Bisa. These are predominantly fine-grained aphyric volcanic rocks (although small opacitic brown hornblendes may form a minor phenocryst phase) and contain xenoliths of ultramafic material probably derived from
ARC DEVELOPMENT IN THE HALMAHERAREGION the Tapas ophiolite. Two samples from this island, OR303 and OR306, yield ages of 9.5_ 0.8 and 8.8_ 1.2 Ma respectively, giving a site mean of 9.2 ___0.4 Ma - notably younger than ages from the Woi Formation.
0.~50 0.25 0.20
"~
o.18
I~
0.10
Bacan
0.05
The Kaputusan Formation is widely distributed throughout Bacan and consists of volcaniclastic sedimentary rocks, lavas and associated pyroclastic rocks deposited in a subaerial or shallow marine environment. The Goro-Goro Member is the volcanic member of the Kaputusan Formation and is divided petrographically into: two pyroxene andesites, hornblende andesites, hornblende pyroxene andesites and hornblende biotite andesites. All these rocks display typical andesitic features such as strongly zoned plagioclase and mafic phenocrysts, glomerocrysts of early crystallising minerals, abundant opaques (magnetite, titanomagnetite), microcrystalline to glassy groundmasses and occasional trachytic textures with the addition of the named phenocryst phases. Chemically the volcanic rocks are typical calc-alkaline arc rocks, predominantly basaltic andesites to andesites with minor basalts and dacites; some samples plot in the high-K to shoshonitic fields. Their trace element ratios are also typical of arc rocks (Malaihollo 1993). Twelve andesite samples from the Goro-Goro Member of the Kaputusan Formation yielded good quality potassium-argon results. Biostratigraphic dates (nannofossils) from sedimentary rocks of the Pacitak Member of the Kaputusan Formation indicate a Pliocene age ranging from N19 to N21. Isotopic ages obtained from rocks in the Goro-Goro Member fall between 7.5 Ma and 2.2 Ma but within this range there appear to be two concentrations of ages. This apparent division may be an artefact of incomplete sampling. The ages form a broad continuum (Figs 4 & 5, Bacan area), but there does appear to be a gap at 3.5 Ma. This gap corresponds to an overall lack of volcanism at this time in both Halmahera and Bacan. The two groups of rocks show minor chemical differences, based on trace element data (Malaihollo 1993), and a broad difference in geographical distribution; the older rocks are found in south Bacan.
0.0
Halmahera and Kayoa Neogene volcanic rocks on Halmahera and Pulau Kayoa belong to the Volcanic Member of the Weda Group and the younger Kayasa Formation. These rocks occur in thick sequences that show abundant evidence of sub-aerial eruption from large volcanic edifices in central and northwest Halmahera whilst
505
vvvvvvv ¸
o
~ 2
vvvvvvvwvvv~v
vvw
vvv,
~vvvvvvv,
~ Age Ma "/~5
8:.&
10.5 11.E
e,,, O.20_
0,150.10 i 0.06. o.o o
t
4
5 ~ AgeMa
7
b
9
io
h
Fig. 5. Gaussian kernel smoothed density estimates for observed Neogene ages (using a window of 0.7 Ma). For description of the technique see text. (a) Densities of ages at certain times suggests that volcanic activity is pulsed rather than continuous over the lifetime of the arc. (b) Calculated probability of gaps in the volcanic record implying lack of volcanic activity at intervals of high probability.
in the southwest arm volcanic rocks are found in redeposited volcaniclastic sequences. These were built on an early to mid Tertiary volcanic arc that deposited the Oha Volcanic Formation on Halmahera and the Bacan Formation on Bacan. Petrographically the rocks are predominantly porphyritic pyroxene and/or hornblende andesites with strongly zoned plagioclase in a microcrystalline to glassy matrix; hornblendes frequently show opacitic or highly included rims. Alteration is rare and the vast majority of the samples are fresh. Chemically the rocks are basaltic andesites to dacites and are medium to high-K rocks of the calc-alkaline series with typical arc volcanic REE patterns (Hakim 1989; Hakim & Hall 1991) that are directly comparable to volcanic rocks from the Sangihe Arc (Morris et al. 1983; Morrice & Gill 1986). Samples analysed for potassium-argon were collected from Pulau Kayoa, midway between Bacan and Makian, and from central Halmahera; most are float samples as they are invariably the least weathered. Potassium-argon ages for Weda Group volcanic rocks on Halmahera and Kayoa range between 4.2-7 Ma whilst those from the Kayasa Formation
506
s. BAKER •
are restricted to around 2 to 3 Ma. All these samples were included in the computation of the density trace and 'gap ages' along with those from Bacan. The results of this statistical test suggest one fairly long period of continuous volcanism (7.8 to 4 M a ) a n d a second period from 3 to 1.8 Ma on both Bacan and Halmahera.
Quaternary volcanic activity Current volcanic activity in the region is restricted to an area north of, and including, Makian Island as far as Galela town in northwest Halmahera. Evidence of Quaternary volcanism on Bacan is witnessed by the presence of (now extinct) volcanic edifices that retain much of their original form; eruptive rocks from these are olivine and pyroxene phyric basalts. These have not been dated but chemically they show characteristics of both arc and within-plate volcanism (Malaihollo 1993). Off the extreme southern tip of Halmahera is the small island of Pulau Doworalama; a biotite separate from a biotite tuff collected here did not yield sufficient radiogenic argon for analysis (below the limit of detection) indicating that it is very young. Calculation, assuming < 1% radiogenic argon, yields an age of less than 0.5 Ma. A number of other small islands lie off the south of Halmahera but none were visited by the University of London group. Morris et al. (1983) visited the islands of Woka, Kekik and Pisang and sampled the Quaternary volcanic rocks exposed there. They found them to have some shoshonitic characteristics with higher alkali contents than are found in their 'oceanic segment' of the arc, with dramatic enrichments of Ba and St.
Neogene reset ages from pre-Neogene rocks Potassium-argon data relating to thermal overprints on older rocks are notoriously difficult to interpret and meaningless 'mixed' ages can be the result. The problems of incomplete resetting under the localized low grade metamorphic and fluid-rich conditions obtaining within island arcs are encountered in the Halmahera region. This is observed in areas where rocks with disturbed potassium-argon ages are found within a few metres of rocks with unaffected, primary ages (also reflected in the petrography of the samples). A number of demonstrably pre-Neogene rocks (based on stratigraphic and biostratigraphic evidence) yield Neogene ages, probably as a result of enhanced heat flow during initiation and development of the Neogene arc. These reset ages appear to show a systematic variation, generally younging from south to north.
J. MALAIHOLLO
Discussion Post-collisional Neogene igneous activity The oldest Neogene age from the region is associated with intrusive rocks of the Nusa Babi intrusions (BR80) whose age is 19.8 _+ 1.6 Ma. As noted earlier, tectonic discrimination diagrams suggest that the Nusa Babi Intrusive may have a collisional or an arc magmatic origin. The regional unconformity of Early Miocene age (c. 22 Ma) is interpreted by Hall et al. (1995a) to be the result of Australia-Philippine Sea plate collision. The Nusa Babi Intrusive rocks post-date collision by several million years and are therefore considered to be the result of magmatic activity associated with heating and stacking of the young orogenic belt in the Bacan area (Malaihollo 1993). The Nusa Babi monzodiorites are the most acidic plutons in the region and this supports the possibility that continental crust was involved in their formation. The sills found in NE Halmahera are dated c. 12 Ma;. The geographical position of these rocks suggests they are due to isolated, shallow magmatic activity in a region well behind the developing arc front. Their La/Nb ratios are less than 1 (0.86 for HB93 and 0.72 for HB97, unpublished ICP data) indicating a true alkaline chemistry (M. E Thirlwall, pers. comm. 1995) that strongly suggests they are unrelated to the early arc activity occurring in Obi at this time.
Neogene arc initiation and development Arc precursors: The two diorite samples dated from Bacan (Saleh intrusions, SR6, B66) display mineral chemical and geochemical characteristics that indicate an arc origin. The interpretation of their isotopic ages as primary crystallization ages implies a period of plutonic activity preceding the development of an eruptive arc in the Bacan region, perhaps in response to subduction initiation (no volcanic rocks of this age range have been found in Bacan).
Age of arc inception: The earliest primary ages from arc volcanic rocks are from samples collected from the Woi and Guyuti Formations in central Obi. None of these ages deviate by more than half a million years from 11 Ma, indicating a short but prolific eruption of andesites at this time. These samples were collected from a wide range of geographical and stratigraphic locations and are thought to be representative of the earliest volcanic episode on Obi; no younger dates have yet been obtained from this region. These rocks are interpreted as the first effusive products resulting from subduction of the Molucca Sea plate and indicate that the slab must have reached depths
ARC DEVELOPMENT IN THE HALMAHERAREGION sufficient to have caused melting (100-110kin, Tatsumi 1986).
Age variation within the arc Ages obtained from acid lavas on Pulau Bisa, just north of Obi, indicate volcanic activity between 9.5-8.8 Ma. The gap in ages between Obi and Bisa volcanic rocks suggests either a hiatus in volcanic activity between 11 and 9.5 Ma or erosion causing removal of younger arc products on Obj. Significant erosion of the arc is unlikely since ages of volcanic clasts from Neogene sedimentary formations on Obi (notably the Guyuti Formation) are also within the age range of Woi Formation volcanism. The rocks from Bisa may represent late stage products of the waning arc in the Obi region as arc activity moved northwards. The oldest volcanic rocks found in Bacan are of early Late Miocene age (7 to 8 Ma) and activity appears to have been fairly continuous until the Late Pliocene-Pleistocene (c. 2.5 Ma). Within this time span there are greater densities of ages at 7-8, c. 5, c. 2.5 and 0.5Ma. These groupings are reflected in concordant chemical and fractionation trends indicating each group evolved from a separate magma chamber (thus representing different volcanic centres). Within the Bacan group of islands there is no clear northward younging of ages but rather a vague suggestion of general younging to the north and west. Neogene and probable pre-Neogene rocks outcrop on Pulau Kayoa (c. 30 km north of Bacan) and KT5B has an age of c. 5.5 Ma. Ages of Halmahera volcanic rocks lie in a spectrum from c. 8.7 Ma to c. 2 Ma. This is divided into two groups corresponding to the two divisions on Bacan (c. 2 to 3 Ma and 4 to 7.8 Ma) but it has not been possible to relate these to discrete volcanic centres as on Bacan. Volcanic rocks from Halmahera, like Bacan, do not display a clear younging of ages within this area and the effect is only seen on a regional scale. A coarse northward younging trend is illustrated in Fig. 4 where the ages of samples are plotted against their locations from north to south. The present-day active arc that stretches from Makian through Tidore, Ternate and into the northwest arm of Halmahera is located approximately 5 0 k m west of the Neogene arc in central Halmahera. This shift must have occurred at or very near the time of the Pliocene shortening event that thrust together the eastern and western halves of Halmahera and overthrust arc rocks of the Woi Formation onto the forearc Guyuti Formation on Obi. It also coincides with a probable gap in the age spectrum indicating no volcanic activity around 3.6 Ma (Fig. 5).
507
Significance of reset ages As indicated earlier, apparent Neogene potassiumargon ages have been obtained from pre-Neogene rocks. They demonstrate a systematic northward younging whilst also consistently pre-dating the ages of Neogene volcanic rocks typically by 3 Ma within a given area.
Fault-related volcanic activity The very young volcanic rocks found in the Bacan and southern Halmahera area of the region (Pulau Doworalama, Pulau Pisang, Pulau Woka and Pulau Kekik) are thought to be related to volcanism occurring along splays of the Strong fault system. Chemical analyses of these rocks show them to have an arc-like character but with a significant within-plate component (Morris et al. 1983) that supports the notion that they are fault-related. Quaternary volcanoes in Bacan lie along a lineation interpreted to be a fault strand whilst the southern islands lie away from the line of the arc in an area thought to contain several fault strands. It is possible that these faults have provided lines of weakness through which small partial melts, related to subduction processes, could pass. If these fault strands do provide conduits for magma migration this indicates that they must be deep crustal discontinuities.
Conclusions Using a conservative estimate of the rate of convergence between the Molucca Sea plate and the Philippine Sea plate of c. 20 km Ma -1 (R. Hall, pers. comm. 1994) in the Neogene and assuming a depth of subducted slab of 100 km beneath the Obi sector of the arc at the time of the first volcanic products, the time of initiation of subduction can be estimated. It is suggested here that the Molucca Sea slab began to sink beneath the Philippine Sea plate at around 15 to 17 Ma (corresponding to reset ages of older rocks from the Obi region). This resetting reflects locally high geothermal gradients as a result of crustal heating and magma rise prior to volcanic activity. Models for the thermal structure of island arcs indicate a higher heat flow and raised isotherms over the volcanic portion of the arc (Oxburgh & Turcotte 1970; Toksoz et al. 1971) especially in very young arcs. Oxburgh & Turcotte (1970) suggest that heat transferred high in the crust by rising magmas could produce nearsurface vertical gradients of c. 100°Ckm -1. Convective hydrothermal cooling through fractures and permeable rocks could lead to even higher gradients (Parmentier & Scheidl 1981). The patchiness of the alteration and resetting of these
508
S. BAKER • J. MALAIHOLLO
rocks suggests that the m e c h a n i s m for these changes is predominantly fluid based rather than a regional raising of isotherm levels. The very localized nature suggests that fluid migration occurs along fractures rather than permeating the rock pile and there is abundant evidence of faulting and fracturing with associated mineralized veins within pre-Neogene formations, notably the Bacan and Oha Formations. Arc activity in the Halmahera region began on Obi at 11.5 Ma causing uplift and cessation of limestone deposition. By 9 Ma volcanic activity had shifted north to Pulau Bisa. Meanwhile, intrusion of diorites was occurring in Bacan prior to eruptive v o l c a n i s m at c. 7 Ma. Arc d e v e l o p m e n t on Halmahera was more or less synchronous with that on Bacan and limestone deposition in these areas ceased (except in the backarc region of NE Halmahera where it continued into the Pliocene). Pliocene deformation caused a westward shift of the arc and at c. 2 Ma arc volcanism ceased south of the equator (Pliocene to Recent volcanic activity in this region was restricted along strands of the Sorong fault system). A large volume of volcanic products was erupted in the first 1 Ma on Obi with little evidence of subsequent volcanic activity. These oldest lavas are calc-alkaline and there is no indication of unusual boninitic or tholeiitic melts as observed in the early stages o f some intra-oceanic systems, particularly the I z u - B o n i n - M a r i a n a System (Hawkins et al. 1984; Stern & Bloomer 1992). Subsequent volcanic
rocks from other parts of the arc are also normal calc-alkaline types. Any unusual chemical and isotopic c o m p o s i t i o n s are related to crustal contamination of subduction melts or fault-related volcanism. On Bacan and Halmahera, volcanism also appears to have occurred in pulses of 1 to 2 M a duration followed by similar periods of relative volcanic quiescence. Subduction-related volcanic activity did not occur along the whole length of arc simultaneously but migrated northwards with time probably as a c o n s e q u e n c e of the c l o c k w i s e rotation of the Philippine Sea plate (Hall et al. 1995b). Using the oldest age from Obi ( l l . 5 Ma) and that from central Halmahera (7.7 Ma), volcanic activity in the arc has m o v e d north at c. 65 k m Ma -~. This work was supported by NERC award GR3/7149, and grants from the Royal Society and the University of London SE Asia Geological Research Group to R. Hall, and financial assistance from Amoco Production Company. We thank J. R. Aii, C. D. Anderson, E D. Ballantyne, E T. Banner, T. R. Charlton, E. M. Finch, E. Forde, R. Hall, G. J. Nichols and S. J. Roberts for their help and discussion, C. C. Rundle of NIGL for guidance and help with isotopic dating and R. Howarth for suggestions and help with statistical analysis. Logistical assistance was provided by GRDC, Bandung and the Director, R. Sukamto with field support by D. A. Agustiyanto, S. Atmawinata, A. Haryono, S. Pandjaitan and Sardjono. We also thank reviewers M. E Thirlwall and T. R. Charlton for their helpful comments and suggestions in preparation of this paper.
References
BALLANTYNE,P. D. 1990. The Petrology of the Ophiolitic Rocks of Eastern Halmahera, Indonesia. PhD thesis, University of London. Cox, K. G., BELL, J. D. & PANKHURST,R. J. 1979. The Interpretation of Igneous Rocks. Allen and Unwin, London CARDWELL,R. K., [SACKS,B. L. & KARIG,D. E. 1980. The spatial distribution of earthquakes, focal mechanism solutions, and subducted lithosphere in the Philippine and northeastern Indonesian islands. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. American Geophysical Union Monograph, 23, 1-35. HAKIM, A. S. 1989. Tertiary volcanic rocks from the Halmahera Arc, Indonesia. MPhil Thesis, University of London. & HALL, R. 1991. Tertiary volcanic rocks from the Halmahera Arc, eastern Indonesia. Journal of Southeast Asian Earth Sciences, 6, 271-287. HALL, R. & NICHOLS,G. J. 1990. Terrane Amalgamation in the Philippine Sea Margin. Tectonophysics, 181, 207-222. , FULLER,M., ALl, J. R. & ANDERSON,C. D. 1995b. The Philippine Sea Plate: Magnetism and
Reconstructions. In: TAYLOR,B. & NATLAND,J. H. (eds) Active Margins and Marginal Basins: A Synthesis of Western Pacific Drilling Results. American Geophysical Union Monograph, 88, 371-404. , AUDLEY-CHARLES, M. G., BANNER, F. T., HtDAY•r, S. & TOBING, S. L. 1988a. The basement rocks of the Halmahera region, east Indonesia: a Late Cretaceous - Early Tertiary forearc. Journal of the Geological Society, London, 145, 65-84. &--. 1988b. Late Paleogene, , , - Quaternary Geology of Halmahera, Eastern Indonesia: initiation of a volcanic island arc. Journal ~" the Geological Society of London 145, 577-590. - - . , ALI, J. R., ANDERSON,C. D. & BAKER,S. J. 1995a. Origin and motion history of the Philippine Sea Plate, Tectonophysics, 251,229-250. , NICHOLS, G . J., BALLANTYNE, P. D., CHARLTON, Z. &
ALI, J. 1991. The character and significance of basement rocks of the southern Molucca Sea region. Journal of Southeast Asian Earth Sciences, 6, 249-258. HAMILTON, W. 1979. Tectonics of the Indonesian region. US Geological Survey Professional Paper, 1078.
ARC DEVELOPMENT IN THE HALMAHERA REGION
HARLAND,W. B., ARMSTRONG,R. L., Cox, A. V., CRAIG, L. E., SMITH, A. G. & SMITH, D. G. 1990. A Geologic Time Scale 1989. Cambridge University Press. HAWKINS, J. W., BLOOMER, S. H., EVANS, C. A. & MELCHIOR, J. T. 1984. Evolution of intra-oceanic arc-trench systems. Tectonophysics, 102, 175-205 MALAIHOLLO,J. E A. 1993. The Geology and Tectonics of the Bacan region , East Indonesia. PhD thesis University of London. MIYASHIRO, A. 1978. Nature of alkalic volcanic rock series. Contributions to Mineralogy and Petrology, 66, 91-104 MOORE, G. F. & SILVER, E. A. 1983. Collision processes in the Northern Molucca Sea. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Part 2. American Geophysical Union Monograph, 27, 360-372. MORRICE, M. G. & GILL, J. B. 1986. Spatial patterns in the mineralogy of island arc magma series: Sangihe Arc, Indonesia. Journal of Volcanology and Geothermal Research, 29, 311-353. MORRIS, J. D., JEZEK, P. A., HART, S. R. & GILL, J. B. 1983. The Halmahera island arc, Molucca Sea collision zone, Indonesia: a geochemical survey. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Part
509
2. American Geophysical Union Monograph, 23, 373-387. OXBURGH, E. R. & TURCOTTE, D. L. 1970. Thermal structure of island arcs. Geological Society of America Bulletin, 81, 1665-1688. PARMENTIER, R. D. & SCHEIDL, E 198t. Thermal aureoles of igneous intrusions: some possible indications of hydrothermal convective cooling. Journal of Geology, 89, 1-22. ROBERTS, S. J. 1993. The Neogene Stratigraphy of the Halmahera Region, East Indonesia. Ph.D. Thesis University of London. STEIGER, R. H. • J.~GER, E. 1977. Subcommission on Geochronology: Convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters, 36, 359-362. STERN, R. J. & BLOOMER, S. H. 1992. Subduction zone infancy: Examples from the Eocene Izu-BoninMariana and Jurassic California arc. Geological Society of America Bulletin, 104, 1621-1636. TATSUMI, Y. 1986. Formation of the volcanic front in subduction zones. Geophysical Research Letters, 8, 717-720 TOKSOZ, M. N., MmEAR, J. W. & JULIAn, B. R. 1971. Temperature field and geophysical effects of a downgoing slab. Journal of Geophysical Research, 76, 1113-1138.
Docking and post-docking escape tectonics in the southern Philippines M. P U B E L L I E R 1, R. Q U E B R A L 2, M. A U R E L I O 2 & C. R A N G I N 1 1 C N R S URA 1759, D d p a r t e m e n t de G~otectonique, Universit~ P & M Curie, 75252 Paris, F r a n c e 2 M i n e s a n d G e o s c i e n c e s Bureau, Manila, P h i l i p p i n e s Abstract: The structure of the Philippine archipelago results from the juxtaposition, between the Late Miocene and the present, of a volcanic belt against fragments of the Eurasian margin and associated marginal basins. The southern Philippines offers the opportunity of studying the mechanics of the deformation from active contraction to a more complex post-docking setting. The docking period is characterized by a compression which began during the early Late Miocene in the central Philippines and has been recently studied in the island of Mindanao. There, deformation initiated in the Late Miocene-Early Pliocene on a NW-trending wrench zone, and continued until the Late Pliocene with thrusting on west-verging fiats and ramps within the arc and east-verging thrusts within the Sangihe forearc. This deformation is still active to the south in the Molucca Sea. The post-docking period began in the Early Pleistocene in northern Mindanao and is represented by a new geodynamic framework with a paired subduction zone and strike-slip fault. However, convergence is still active in the Manila and the Negros trenches, although the Philippine fault is not offset. Large wrench faults, which reactivate the ramp faults of the collision stage, transfer strain from the Philippine fault to the Manila and Negros trenches. These observations imply active intra-arc extension and fragmentation within the Philippine mobile belt.
The Philippine archipelago is composed of fragments of the Eurasian margin, which have been rifted away from mainland Eurasia (Fig. 1), and a large volcanic belt, referred to hereafter as the Philippine arc, whose history is linked to that of the Philippine plate. The juxtaposition of the Philippine arc against the margin occurred during the late Neogene but in detail varies from north to south. The tectonic features of the collision include strikeslip faults and thrust tectonics. The post-collision setting involves two subduction zones of opposite polarity which define beween them a varied assemblage of continental, older volcanic and oceanic crust fragments upon which the active volcanic arcs are built (Cardwell et al. 1980; Lewis & Hayes 1984; Fig. 1). The intra-oceanic belt is built upon an EoceneOligocene volcanic arc and is in tectonic contact with the Eurasian margin by means of subduction zones in front of the marginal basins and by means of collision zones in front of the continental fragments. The subduction zones are arched toward the marginal basins and anchored at the collision zones (Fig. 1). The collision zones are from north to south: the Taiwan area (Davis et al. 1983; Barrier 1985; Pelletier et al. 1985), the Mindoro-Panay collision zone (Rangin et al. 1985; Marchadier &
Rangin 1990), and the Mindanao collision zone (Moore & Silver 1983; Hawkins et al. 1985; Mitchell et al. 1986; Pubellier et al. 1991, 1994). The re-entry zones within the belt imply active convergence at the Manila, Negros, and Sulu trenches. However, the Philippine fault does not respect such geometry. The elongated Philippine mobile belt is traversed by the 1200 km long leftlateral Philippine fault which roughly parallels the Philippine trench (Willis 1937; Allen 1962; Aurelio et al. 1991; Barrier et al. 1991). It has been demonstrated that the onset of m o v e m e n t on the Philippine fault post-dates the arc-continent collision in the northern Philippines (Pinet & Stephan 1990; Ringenbach 1992), in the central Philippines (Aurelio et al. 1991; Aurelio 1992), and in Mindanao (Pubellier et al. 1991, 1994; Quebral et al. 1995). In addition, active extensional tectonics have been documented in various parts of the Philippines (Fig. 1). Several basins have been described in the island of Luzon, some of which are not directly linked with motion along the the Philippine fault. Some extensional lineaments, such as the Macolod Corridor (Fig. 1), are still controversial. Extension also occurs in the Visayas (central Philippines) controlled by faults which parallel the fold axes.
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 511-523.
511
512
M. PUBELLIER ET AL.
The docking phase The Neogene collisional setting involves large strike-slip fault zones (Karig et al. 1986; Sarewitz & Karig 1986), and local collision zones (Holloway 1982; Bachman et al. 1983; Haeck & Karig 1985; Hawkins et al. 1985; Rangin et al. 1985; Marchadier & Rangin 1990; Pinet & Stephan 1990; Pubellier et al. 1991), depending on whether the Philippine arc was in contact with a continental fragment or oceanic crust. In Mindanao, wrenching began at the Late Miocene-Early Pliocene boundary as a NW trending wrench zone, and extended untill the Late Pliocene with movement on west-verging flats and ramps within the arc and east-verging thrusts within the Sangihe forearc.
Mindanao: wrenching and thrust tectonics
Fig. 1. Geodynamic setting of the plate boundary between the Philippine arc and the Eurasian margin, and major extensional zones of the Philippine Mobile Belt. Map distinguishes kinds of faults (thick lines for subduction and strike-slip, thin lines for normal faults) and nature of lithosphere (see key). Lithosphere is classified as continental margin (cross-hatching), intraoceanic volcanic arc (fine stipple) or marginal basin floored by oceanic crust (blank). Labelled tectonic units are: West Mindanao continental block (WM), Palawan continental block (PB), Palawan-Sulu trenches (PST), Sulu arc continental block (SA), SVPF; Sibuyan-Verde Passage fault (SVPF), Daguma Cotabato fault (DCF), Saranggani Peninsula block (SP), Zambales Massif block of oceanic crust material (Z). South China Sea (SCS), Philippine Sea Basin (PSB), Siargao fault (SF), Lianga fault (LF), Bislig fault (BF), Macolod Corridor (MC), Visayan basins (VB), Bohol Sea (BS), Lanao Lake (LL).
A recent marine survey also indicated an active basin floored with oceanic crust SE of Marinduque (Sarewitz & Lewis 1991). In this paper, the deformation associated with the docking of the Philippine arc is distinguished from that associated with the post-docking setting. It is shown that most of this late deformation is extensional on the basis of field and an escape model is proposed for the deformation of the mobile belt of the Philippines (Fig. 2).
No detailed investigation of collision tectonics has previously been conducted in Mindanao because the recognition and the dating of the tectonic elements is complex and the access is difficult. In addition, Miocene compressional features (Fig. 2) were overprinted by Pleistocene to recent extension and wrenching. A synthesis of the results of the recent tectonics is compiled on the neotectonic map of Mindanao (Pubellier et al. 1993; simplified as Fig. 2). Deformation began by the latest Miocene in western Mindanao and is marked by wrench faulting and folding (Fig. 3, top). The westernmost part of Mindanao (west Cotabato Basin and Illana Bay) underwent deformation on parallel NWtrending fault-bounded elongated pressure ridges. These ridges are known from seismic lines (Fig. 3, bottom) and their surface geometry was deduced from drainage anomaly mapping (Deffontaines et al. 1993; Pubellier et al. 1994). Some lie along the eastern flank of the Daguma Ridge near Lake Sebu or emerge from the Quaternary sediments of the basin (Roxas Ranges). Synthetic cross-sections drawn from a compilation of unpublished seismic lines by Letouzey et al. (1987) show that Pliocene reflectors unconformably overlie these structures, which are associated with west-dipping low angle faults. Their 'en echelon' geometry also suggests a strike-slip environment which affected Upper Miocene to Lower Pliocene strata (Pubellier et al. 1991). The largest of these ridges is buried and known from drilling and seismic lines only (unpublished PNOC report 1986). Seismic lines show NW-SE faults with jogs resulting from a combination of reverse and strike-slip faults trending NW. Correlation of seismic lines show these structures to be at 90°to the basement reverse faults, and to connect with the Flecha Peninsula of northern Zamboanga. Reflection seismic data (Fig. 3)
DOCKING 8Z ESCAPE TECTONICS, S PHILIPPINES
513
Bohol iiiiiiiiiiilsilandiiiiiii
Sea
Cagayan
....
~ ~ ~ ~ i ~
Cotabato Basin t
:iiig~ i:iii
I
I Phil. Arc basement
~ -'---
Talaud• ridge + Mindanao oe,~"nic series - - -
~iiilt
Sang,hewedge inferred Eurasian Margin basement
--,,,x
-
.
....:ii ii!ili~:i!::~,• , ........
Pre ple,stocene rocks
,
':iliiilii!!:
~ ii::! iiiil; :::::: :i~;/~I],, ,, I'
~ii::M:: ~:::~'.~! ::~:.~:~:!! ........ .~
~.:::~:~::..~.: i............ i!i:i:i"i:i .~:|~:~:~:~"::::::::::::::| ~:::~:~:~!iiiiiiii~:
. . . . . "W,~r.,~fl / ,. -. t ~ ~" ! •
~
~.:
6
!:!:!:~::.?.::..
o
Fig. 2. Structural sketch map of the Neogene compressional features of Mindanao island prior to the Pleistocene.
indicate that the deformation is sealed by thick Pliocene and possibly uppermost Miocene sediments. In the field, the Opol Formation, whose base is dated as uppermost Miocene ( N N l l ) by nannoplankton (C. Muller, pers. comm. 1994), is unconformable on the basement, although substantially uplifted and affected by more recent (Late Pliocene) open folds. By the end of the Pliocene, deformation was more intense in central Mindanao. It is marked by folds axes and thrusts whose orientation varies from N-S to ENE (080°), which die out alongstrike as they branch into NW-trending faults (Figs 2 & 4). South and west of Cagayan de Oro city, overturned folds and large thrusts have brought the metamorphics and the ultramafic complex with its overlying arc series onto upper Oligocene or Lower Miocene foraminifera-bearing limestones. The
Opol Formation, dated at its base as uppermost Miocene (NN11, Pubellier et al. 1991), is unconformable on these structures. The northern part of the Cotabato Basin also includes two narrow ranges (north of Pigkawayan) composed of peridotites, pillows and radiolarites where similar structural observations were made (Figs 2 & 4). The folds are sealed by a sequence of rocks including columnar basalt flows and thick tufts, which have yielded K - A r ages of 2 Ma (Sajona et al. 1993). This papwe therefore separates a N W - S E strip (Fig. 4) with very few outcrops sandwiched between the Philippine arc and the continental margin on which the Sangihe forearc rests. This central strip is interpreted as a zone of wrench deformation along-strike from the Molucca Sea collision complex, deduced from the bathymetry observed during the MODEC cruise (Rangin et al. 1995), and
514
M. PUBELLIER E T AL.
Sindan~/t~r'~urorari~'dge~a.~kjl~,A . . . . . / ~"i~ariyl~iiocenetopre:seni~
~b~~l~
=.// ~
,pil~
Interpreted sectionfromvar~us datain centralCotab~y
Illana l N ~
Cotabato
,
/
,.-i 12~Ramn~e M~arbel~~ 6N
SSW
Flechalanticline ~.....~_ .
' ''=~2...... "~ "
Auroraridge ~ ~ ~
.
~lt~pA
~,'
I SSW / FlechaAnticline _ I
1sac
.,
IllanaBay
/ IllanaAnticline
"" ' ~"
NNEj
IllanaBay
-: :'-.-~-'~~'~-.'~" ~ ~'~~ '~" ~ "'~L~~ . . . . . . . . . .
NN~_EL ~
1sec
-. . . .
~---~"4sec
Fig. 3. Structural map of westem Mindanao and seismic lines within the Illana Bay. Similar features in the Cotabato basin are deduced from Letouzey et al. (1987) and drainage anomalies (Pubellier et al. (1994). Upper map also shows the Aurora ridge (dark pattern), c 3 values from fault slip analyses in Pleistocene sediments and focal mechanisms from shallow earthquakes.
with rare outcrops of ophiolites and melanges in central Mindanao (Fig. 2). Ramp faults are associated with this event, the largest of which is the 50 km long left-lateral Tagoloan fault (Figs. 2 & 4), SE of Cagayan de Oro, but several others appear on Landsat images. The Tagoloan fault affects affects the Upper Miocene to at least Upper Pliocene clastic Opol Formation overlying serpentinized peridotites. The
fault is capped by 50 m of columnar basalts. The flows were not dated in this area but samples from the huge volcanic field of central Mindanao yield ages of 0.4 Ma and younger (Sajona et al. 1993) and the volcanic ridge of northern Zamboanga is intercalated with lower Pleistocene (NN19) marine sediments. It is suggested that the wrench tectonic phase was coeval with the thrusting in northern Mindanao and is of late Pliocene age. The most
DOCKING & ESCAPE TECTONICS, S PHILIPPINES
LF
•
,..
lind
'~
IIIIIIIIII
.....
--~2=,"i
Jll
Eurasian basement ~ 1 -- ( ' ; ; ] ] ~ F-!7;!~Collision complex of the "~'~.._ :~..... ~,.J,~;; ! L ~ Molucca Sea (Talaud ridge ~ . ~ . . . . . . . . ~" ) ",~=:~... TI I ltl & wedge and cent. Mindanao JLWPhilippine Arc ~-----1,~
Fig. 4. Schematic map of the structural units of the transfer zone during the Pliocene. Folds and thrusts are east-verging in the Sangihe forearc and the margin (SW), and are west-verging in the Philippine arc (horizontal pattern) and the central strip (dark pattern). The central strip is interpreted as the extension of the Molucca Sea collision complex deduced from the bathymetry of the MODEC cruise (Rangin et al. 1995) and with scarce outcrops of ophiolites and melanges in central Mindanao.
accurate dating of this Late Pliocene compressional event comes from shallow marine sediments of the Agusan-Davao basin in eastern Mindanao. There, tight 050 ° to 070°-trending folds affect Middle Pliocene (NN15) to Upper Pliocene (NN16) clays and greywackes. These folds are sealed by Upper Pleistocene (NN20-21) sediments (Quebral et al. 1995). To the west, the basin rests on the eastern side of the Central Cordillera, which in an assymetric west-verging thrust-anticline. In the southern part of eastern Mindanao, serial seismic lines onland and offshore of Davao Gulf (Quebral et al. 1995; Rangin et al. 1995) show the ridges to be very young asymmetrical anticlines, most of them being bounded on the eastern side by reverse faults. Unconformable reflectors overlying the folds are rare except for locally seismically transparent detrital fans eroded by channels. These anticlinal structures are very tight east and north of Samal island which is formed by the emergence of such a ridge. Only the southwestern part of Mindanao shows a difference in tectonic style because duplexes are thrust to the east (Fig. 4). The most remarkable structure is the Saranggani Peninsula which is an anticline cored with middle Miocene limestone underlying a thick unit of volcanic and volcaniclastic upper Miocene rocks (whole rock K-Ar
515
dating, Pubellier et al. 1991 ). This series is overlain by a sandy and marly unit whose base is of lower Pliocene age and whose upper section includes upper Pleistocene sediments. This again indicates that the beginning of the deformation began by the Late Miocene-Early Pliocene boundary. According to Quebral et al. (1995), the Pujada Peninsula at the southernmost end of the Pacific Cordillera, is one of these east-verging thrusts. These interpretations were later confirmed by the MODEC cruise (Rangin et al. 1995).
Luzon: M i d d l e M i o c e n e c o m p r e s s i o n
Coarse upper Oligocene-lower Miocene clastic sediments (Klondyke Formation) are interpreted to have come from mass wasting along the arc and are covered by upper Lower Miocene to lower Middle Miocene platform limestones. This is interpreted to record a tectonic event related to the early accretion of the Philippine arc against the Eurasian margin with sedimentary detritus carried as far as Ta'iwan (Florendo 1994; Pelletier et al. 1985). The tectonic features of this event are folds, thrusts, and poorly understood strike-slip faults in the Philippine arc as in Catanduanes island (Geary & Kay 1989) or in the Zambales massif (westernmost Luzon; Hawkins & Evans 1983; Karig 1983). Stratigraphic evidence of strike-slip motion is found within upper Middle Miocene rocks (Maleterre et al. 1988; Pinet & Stephan 1990). However, these faults reactivated previous structures that controlled sedimentation during the Oligocene. This contraction, which involved wrenching, was closely followed by a renewal of volcanic activity attributed to the present Manila trench (Defant et al. 1990). The second arc started activity by the late Middle Miocene and its products are clearly recorded in sedimentary basins (Ilocos and Cagayan basins) on both sides of the Cordillera (Pinet & Stephan 1990).
N E Visayas a n d central Visayas: Late M i o c e n e - E a r l y P l i o c e n e accretion
The Visayan archipelago in the central Philippines (Fig. 1) is composed of a Cretaceous to Palaeogene ultramafic, metasedimentary and Eocene to middle Oligocene volcanic and plutonic basement (Irving 1950; Gervasio 1966). Most important is the presence of an unconformity at the middle and upper Miocene boundary (9 Ma; Rangin et al. 1991). This unconformity is typical of the central Philippines; it is known in Palawan (Fricaud 1984), and Panay-Mindoro (Rangin et al. 1985; Marchadier & Rangin 1990). It was also found in oil wells in the Bantayan graben NW of Cebu. The
516
M. PUBELLIER ET AL.
upper Miocene-Pliocene calcareous Carcar Formation is unconformable on older rocks, although moderately folded. Fault set analyses in Bondoc, Masbate and Leyte indicate that stress tensors correspondingly shifted from a collision-related direction to a strikeslip-related one between the latest Miocene to the Pliocene (Aurelio et al. 1991; Aurelio 1992). This shift in tectonic regime is best manifested in Bondoc Peninsula, where collisional ~1 orientations are generally N-S (incompatible with leftlateral motion along the Philippine Fault) while the Philippine Fault sl orientations are between 050 and 080 ° . It should however be noted that traces of older strike-slip faults can be found in the region, the most conspicuous of which are in Mariduque island (Fig. 5). Here, a NNW-SSE trending leftlateral fault affects Middle Miocene sedimentary formations but is sealed by Plio-Pleistocene formations. Furthermore, a Quaternary volcano built on the southern tip of the island does not appear to be truncated by the strike-slip fault (Aurelio 1992).
an
azpiqon 3asin ,rgao strait ianga Bay 3islig Bay ateel Fault
The post-docking phase: Late Piiocene to Present strike-slip, extension and escape In contrast, and with the exception of the active thrusts described in Zamboanga and in NW Mindoro, most of the active deformation of the Philippine Mobile Belt is presently extensional and strike-slip (Fig. 1). Rather than a classical postorogenic collapse in the sense of Dewey (1988), the extensional features observed seem closely linked to the tectonic setting of oblique convergence and the geometry of the suture zone. In addition to the pull-apart basins associated with the Philippine fault, two fabrics are observed in the the active grabens on either side of the Philippine fault. A set of extensional basins situated within the strip between the Philippine fault and the Philippine trench is attributed to a process referred to here-
•Mati Fault Fig. 6. Sketch diagram showing the dispersion of the tectonic sliver between the Philippine fault and the Philippine trench, as a result of the southward migration of the trench-fault pair,
after as a post-docking dispersion since it removes fragments of the Philippine arc (Figs 6 & 7). On the other hand, grabens located in the middle of the Philippine arc seem to be controlled by the subduction zones in front of the marginal basins to the west, which generate differential displacements inside the belt, a process referred to hereafter as lithospheric escape.
Post-docking dispersion
Fig. 5. Structural map of Marinduque island showing the Late Miocene left-lateral wrench fault zone (modified from Aurelio 1992).
Dispersion between the Philippine fault and the Philippine trench (Figs 6 & 7) results from the oblique convergence along the trench and the southward propagating trench-fault system. Strikeslip motion along the Philippine fault separated and defined a strip of volcanic material between the the Philippine fault and the trench (Barrier et al. 1991). Field data show that the fault is older in the northern Philippines (Pinet & Stephan 1990) than it is in the central (Aurelio et al. 1991; Aurelio 1992) and the southern Philippines (Quebral et al. 1995). This implies that the total offset of the fault increases from south to north in the manner
DOCKING ~ ESCAPE TECTONICS, S PHILIPPINES
Fragmentation Dispersion
517
towards the bay; whereas its southern coastline is controlled by a major NNE-SSW escarpment which exposes the Oligocene volcanic basement of the cordillera. Between the Lianga fault and Bislig Bay is a carbonate platform whose morphology is interrupted by normal faults expressed as northfacing escarpments with NNE-SSW and NNWSSE segments which affect rocks as young as Late Pleistocene.
Northern Philippines
Fig. 7. Simplified map of the major post-early Pleistocene extensional faults of Mindanao and the two main strike-slip fault zones, with labels. Light pattern for the Eurasian margin and dark pattern for the accreted Philippine arc undergoing dispersion and fragmentation (see text). Visayan basins (VB), Tandag fault (TF), Lianga fault (LF), Bislig fault (BF), Cateel fault (CF), Daguma fault (DF), Cotabato fault (DCF), Valencia (East Cotabato fault, VF).
described by McCaffrey (1992) for the forearc of Sumatra. Differential displacement of discrete blocks due to a southward decrease in the total displacement along the Philippine fault accounts for extension in the region (Fig. 6).
Lianga and Bislig Bays in Mindanao The northern Pacific Cordillera (Figs 1, 2 & 7) is considered to be a large ramp anticline guided along its southern limit by the NNW-SSE Lianga fault. During the early deformation which affected upper Pliocene (NN15-16) rocks, the Lianga fault acted as a left-lateral strike-slip fault. However, truncation of the northern Pacific Cordillera by the Philippine fault and its subsequent northward strike-slip displacement resulted in the reactivation of the Lianga fault as a normal fault (Fig. 7). The youngest slickensides observed in upper Pleistocene (NN20-21) marls therefore correspond to normal faulting. A NNE-SSW structural fabric, observed throughout the Pacific Cordillera, strongly influences the coastline of eastern Mindanao (Bislig, Cateel Bays, Fig. 7). Bislig Bay, for example, is interpreted as a half-graben. Sediments along its northern coast dip gently
Several other areas in the Philippines display features which are compatible with this mechanism. The Baler-Casiguran basin (Fig. 8) is located along the southern coast of the Sierra Madre of Luzon, and was studied on the basis of seismic lines between the San Ildefonso Ridge and the Casiguran Coast by Ringenbach et al. (1993). The elongated graben is considered to have undergone minor right-lateral slip according to Ringenbach et al. (1993) and therefore to be the result of oblique rifting. However, the overall geometry on the seismic lines is consistent with a basically extentional motion. The active faulting reactivated pre-existing 050°-trending basement faults. The Legaspi Lineament (Fig. 8) is a left-lateral strike-slip fault trending WNW-ESE in the southeastern Luzon area. It intersects both the Philippine Fault in the Ragay Gulf and the Philippine trench east of Catanduanes Island. Aside from displacing the Philippine trench left-laterally by about 40 km, its activity is demonstrated by the occurrence of earthquakes on its southeastern segment. Onshore, the Legaspi Lineament is responsible for the formation of an elongated depression represented by Lake Bato, probably situated in a local pull-apart setting in a restricted segment of the lineament. It also appears to partly control the activity of Mayon Volcano, the most active of all the Philippine volcanoes. Into the Ragay Gulf, the lineament occurs as a steeply SW-dipping fault cutting through the entirety of the sedimentary fill (seismic profile). Simultaneous activity along the Legaspi Lineament and the Philippine fault gives rise to the elongated geometry of the Ragay Gulf depression (Fig. 8).
Post-docking escape Extensional transfer faults The term transfer faults is used for large wrench faults which connect the collision zone or the Philippine fault to the intraplate deformation front (e.g. the Negros and Manila trenches, or the Mindoro-Panay thrusts). Geologic data and seismic
M. PUBELLIERET AL.
518
-0
\ \ \
S W
-0
\
-' ®
"
NE
\ \ y d . Slbuyan Sea Fault \
-0
Fig. 8. Structure of the Legazpi Lineament and location of seismic lines at the termination of the Sibuyan-Verde Passage fault with major focal mechanisms.
lines across these faults reveal that an extensional component is important (Figs 1 & 7).
Cotabato fault and Aurora volcanic ridge This ridge is located fight on the suture zone and is presently an elongated bayonet-shaped horst which obliquely separates northern Illana Bay from the southern Dipolog basins. The left-lateral motion on the bordering NW faults is indicated by several focal mechanisms determined on shallow earthquakes (Fig. 3). Microtectonic studies along the Sindangan-Cotabato-Daguma lineament clearly indicate E-W compression generating left-lateral faulting. Microtectonic measurements on Pleistocene sediments and overlying basalts indicate N-S extension. E-W normal faults accommodate the strike-slip motion and have created pull-apart basins west of Aurora in northern Zamboanga.
Daguma fault Both edges of the Cotabato Basin are faultbounded with tilted blocks (Fig. 7). On the eastern side, a N-S diffuse fault zone exists in central Mindanao. Tilted blocks and small volcanic cones, some of them dated between 1.15 and 0.25 Ma (K-Ar), occur there and underline normal faults which are thought to reactivate the short limb of the Central Cordillera ramp anticline. On the western side, Pliocene thrust-cored folds orientated NW (N130 °) along the Daguma range have been reactivated as normal faults with considerable offset southeast of Cotabato (Daguma or Tiruray fault). Some of these normal faults cross-cut the Parker volcano in southern Kudarat Province and cut the Quaternary reefal terrace south of General Santos. Although this extensional pattern can be related to backarc extension behind the Cotabato
DOCKING ~; ESCAPE TECTONICS, S PHILIPPINES trench, the authors favour a wrench-related interpretation following the accretion.
519
the Pantabangan-Carranglan basins which are marked by the deposition of middle Miocene sequences and presently undergo transtension as well as extension with subsidence. Presently, the western-central Philippine area is characterized by the presence of NW-trending strike-slip faults such as the Sibuyan Sea Verde Passage faults (Fig. 8). To the south, near Masbate Island, the Sibuyan Sea fault trends WNW-ESE and cuts across the southeastern extension of the Marinduque Basin (Sibuyan Sea) filled with recent sedimentary deposits. Seismic reflection profiles clearly indicate extensional structures best exemplified by tilted blocks and graben structures. Recent activity along this fault is manifested both in terms of minor earthquakes (magnitude < 5 Mb) and variations in stress orientations in Masbate where the fault intersects the Philippine fault (Aurelio 1992). To the NW, the Verde Passage fault trends in the same direction as the Sibuyan Sea fault as it passes between Mindoro Island and Luzon towards the Manila trencE Its termination into the trench is,
Other occurrences in the central and northern Philippines Gabaldon Basin (Ringenbach 1992; Ringenbach et al. 1993) is a Pliocene basin controlled by the Cascades fault zone which is presently associated with the Philippine fault in central Luzon (Fig. 9). The basin in its present stage is represented by fluvial deposits of Pliocene age which post-date the folded lower Miocene sediments. The present configuration is a N W - S E elongated graben bounded by listric faults as indicated by tilted terrace deposits on the basin's southern rim. This basin has been interpreted to be the result of movement at a releasing bend of the Philippine fault (Ringenbach et al. 1993), but it also fits well the model of an escape of the Baguio-Zambales block from a fixed Philippine fault (Fig. 10). Similar observations were made by Ringenbach (1992) in
t
Cretaceous
~to
Eocene Basement
~
Miocene
Philippine
Arc
Basement
150Kin _
SW
Cascades FZ
Northern Sierra Madre + ~
_
1
+ •
NE #
Philippine FZ
\~"
So rn +/Madre Sierra
Baler Casiguran Basin San IldefonsoRidgc
W r'--~'~--~-- =
-:--~"
- - - - - -. . . .
--':-~
"
-- . . . . . .
~ _ ~ _ ~
Jl~
Fig. 9. Location and structure of the Baler Casiguran Basin of Luzon and the Mocalod corridor modified from Ringenbach (1992) and Wolfe & Self (1983). Manila fault (MnF), Marikina fault (MkF).
520
M. PUBELLIERET AL.
Extension resulting from subduction at the Negros and Manila trenches L a n a o extensional field, B o h o l Sea a n d Visayan basins
FRAGMENTATIO~
~
Sol,~niab.
L ~ l l l V l ~ " ~ i i ! i : Gabaldon Jr" ii~Bl~.~:~i~i!::~ f Pantapangan South [ ~ i i i : : : , Carranglanb. China ~' ~ k ~ ~ i i i i : : : i : : i ! i i . S e a ~ ~ ' ~ d Coridor
~u,u
Sea
_ ~~
~
~1 ~I!:i:i:!:i3::: nao ,=~1
field
'''n
......
V
Buluan ;:!i!ii!!!!:i:ii oasm
~
Sea
=
Fig. 10. Sketch diagram and map showing fragmentation and escape west of the Philippine fault due to the trench pull in front of the marginal basins.
however, unclear. It appears to die out into the accretionary prism which is well developed at this latitude as observed during the POP2 cruise (Rangin et al. t988). Seismically, minor to moderate earthquakes are known to have occurred along this structure for the past 100 years (Aurelio 1992). Its connection or non-connection with the Sibuyan Sea fault, however, still remains to be understood. Bathymetrically, these two strike-slip faults are in fact linked by the Sibuyan SeaMarinduque Basin.
Extension almost parallel to the convergence direction was first observed by Rangin et al. (1989) in the central Philippines where a radial set of graben post-date the Late Miocene-Early Pliocene docking phase and tend to parallel the Negros trench (Figs 1 & 7). The basins were explored for oil (Glocke 1980) and some of them are filled with up to 6 km of Miocene to Recent sediments. Normal faults of the grabens reactivate the short limbs of the folds and affect the Pliocene to Pleistocene Carcar Formation (Rangin et al. 1989). The southernmost basin is the Bohol (or Mindanao) Sea which is a poorly known basin of 2.5 s reflection seismic floor with 0.5 s mean sediment thickness (Hamilton 1979). In central Mindanao, the 200 km long Lanao volcanic field is a huge plateau situated around Lanao lake. Nb-enriched calc-alkaline and potassic lava samples have been dated between 2.31 and 0.29 Ma (Sajona et al. 1993) in this area. Extension is marked by 20 to 50 km long arcuate normal faults (Ranneft et al. 1960) which appear on Landsat images and control the geometry of the lake. Although the faults do affect the volcanic aprons, the active or dormant volcanic cones are aligned in the same direction, as pointed out by Hamilton (1979). The normal faults connect with the active transtensile Cotabato fault and Aurora ridge as part of a single tectonic system with ~3 striking N-S (Pubellier et al. 1994). Symmetrically, on the opposite side of the Cotabato fault, an extensional zone affects the Quaternary volcanics and the Quaternary reefal terraces of Mt Blick in the northern Daguma Range. The 050°-trending normal fault provide the structural control for the tiny Bongo island. M a c o l o d corridor
The Macolod Corridor (Figs 1 & 8) is a complex graben structure which has been interpreted as a pull-apart basin characterized by the occurrence of intense volcanism (Defant et al. 1990). However, the bounding faults are not clearly identified and the mechanism debatable (F"rster et al. 1988). Wolfe & Self (1983) suspected that this feature may be a link between the northwestern shelf of Palawan and the Philippine fault of central Luzon. Regardless, the distribution and abundance of volcanoes suggest lithospheric fractures striking mainly NE-SW but also NW-SE. The large Taal volcanic centre is located at the intersection of these directions. Clearly the Macolod corridor
DOCKING 8Z ESCAPE TECTONICS, S PHILIPPINES accommodates a differential displacement between the central and northern parts of Luzon, in response to forces induced between the Manila trench and colliding Benham Plateau, and the encroachment of the southern part of Luzon on the Palawan Block (Fig. 1). Marinduque Basin
Recently, incipient sea floor spreading has been documented in the central Philippines (Sarewitz & Lewis 1991), in connection with a wrench environment. The Marinduque Basin, located east of Marinduque Island and Bondoc Peninsula, has a NW-SE bathymetric axis. Offshore data, however, suggest the presence of an E-W trending bathymetric ridge which, according to Sarewitz & Lewis (1991), might correspond to a N-S spreading axis, based upon the observation that the ridge is made up of volcanic rocks whose magnetic signature is comparable to that of rocks produced during oceanic accretion. In this model, the system, now inactive, could have operated along N-S trending strike-slip zones. Such structures are presently observed as thrust faults lining up the western edges of the Bondoc Peninsula compressive ridge, associated with the Philippine fault. In view of its location between the Sibuyan Sea fault and the Verde Passage fault, it is tempting to argue that the Marinduque Basin is an extensional relay zone between strike-slip faults.
Interpretation and discussion One of the best recorded Tertiary tectonic events in the Philippines is the collision of the continental block of north Palawan with island arc units in the central Philippines. Manifestations of this collisional event are readily observable in the islands of northern Palawan, Mindoro, Panay and other neighboring islands (McCabc et al. 1983; Rangin et al. 1985; Mitchell et aL 1986), and recently mapped in Mindanao (Pubellier et al. 1993, 1994; Quebral et al. 1995). Folds and thrusts produced by this event are associated with lateral ramps striking NW-SE. It is believed that the lateral ramps are part of a large transfer zone with follows the Eurasian margin geometry and truncates the Molucca Sea (Rangin et al. 1995). In Mindanao, this wrench tectonic phase began at the Late Miocene-Early Pliocene boundary and lasted
521
until the end of the Pliocene resulting in deformation of the Agusan-Davao Basin. In the northern Philippines this event is dated as Middle Miocene. The post-docking period is now represented by a new geodynamic framework with paired subduction zones and strike-slip faults, due to continued convergence at the Manila and Negros trenches. Large transfer wrench faults (Fig. 4), which reactivate the lateral ramps of the collision stage, transfer strain from the Philippine fault to the Manila and Negros trenches. If the Philippine Mobile Belt east of the Philippine Fault is considered to be fixed on the basis that the 1200 km long Philippine fault is not offset nor disrupted by the transfer faults, it can be inferred that the western half is undergoing extension driven by slab-pull at the Manila and Negros trenches. In response to the extension, this part of the belt is fragmented and the fragments tend to escape toward the marginal basins. The escape concept, which refers to the motion of a package of rocks towards a free edge within a compressional environment, has been much described as a byproduct of frontal collision (e.g. the Indochina block: Tapponnier et al. 1982; Davy and Cobbold 1988; or the Alps: Ratschbacher et al. 1991), or from the combination of movement of the Anatolian block triggered by Arabian-Eurasian convergence and extensional forces due to slabpull at the Aegean trench (McKenzie 1972; Sengtr et al. 1985). This paper proposes that oblique convergence along an irregular continental buttress is also a likely setting for escape of continental fragments. The escape motion affects the continental fragments dragged by the northward motion of the Philippine plate to a smaller extent. This paper is a result of five years of co-operation between the Mines and Geosciences Bureau of the Philippines, and the Universitl~E et M. Curie in Paris in the southern Philippines. We are very grateful to the Philippine National Oil Company (PNOC) and the Bureau of Energy for allowing access to unpublished data. This work has been funded by the MinistEre des Affaires EtrangEres. Laboratory facilities were supplied by the Universit6 P. et M. Curie in Paris, and some palaeontological dating were provided by C. Muller (Institut Francais du Pttrole) and J. Butterlin (Inst. Phys. Globe de Paris). Two of the authors (M.P and C.R.) belong to the Centre National de la Recherche Scientifique, and two others (R.Q. and M.A.) are from the Mines and Geosciences Bureau of the Philippines.
References ALLEN• C. R. 1962. Circum-Pacific faulting in the Philippine-Taiwan region. Journal of Geophysical Research, 67, 4795-4812. AURELIO, M. 1992. Tectonics of the central segment of the Philippine Fault: structures, kinematics and
geodynamic evolution. Th~se de Doctorat de l'Universite Paris 6, France. • BARRmR,E., RANOrN,C. & Mt3LLER,C. 1991. The Philippine Fault in the late Cenozoic evolution of the Bondoc-Masbate-N. Leyte area, Central
522
M. PUBELLIER ET AL.
Philippines. Journal of SE Asian Earth Sciences, 6, 221-238. BACHMAN, S. B., LEWIS, S. D. & SCHWELLER,W. J. 1983. Evolution of a forearc basin, Luzon Central Valley, Philippines. AAPG Bulletin, 67, 1143-1162. BARRmR, E. 1985. Tectonique d'une chaine de collision active: Taiwan. Thbse de Doctorat d'Etat, 85-29, Paris, France. BARRIER, E., HUCHON, P. & ACP,EUO, M. 1991. The Philippine fault: A key for Philippines kinematics. Geology, 19, 32-35. CARDWELL,R. K., ISACKS,B. L. & KARIG, D. E. 1980. The spatial distribution of earthquakes, focal mechanism solutions and subducted lithosphere in the Philippine and northeastern Indonesian islands. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. American Geophysical Union Monograph, 23, 1-35. DAWS, D., SUPPE, J. & DAHLEN, F. 1983. Mechanics of fold and thrust belts and accretionary wedges. Journal of Geophysical Research, 88, 1153-1162. DAVY, P. & COBBOLD, P. R. 1988. Indentation tectonics in nature and experiment. Bulletin of the Geoligical Institute of Uppsala, N. S. 14, 129-141. DEFANT, M. J., MAURY,R. C., JORON,J. L., FEIGENSON,M. D., LETERRIER, J., ET AL. 1990. The geochemistry and tectonic setting of the northern section of the Luzon (the Philippines and Ta'iwan). Tectonophysics, 183, 187-205. DEFEONTAINES, B., PUBELLIER, M., RANGIN, C. & QUEBRAL, R. 1993. Discovery of an intra-arc transform zone in Mindanao (Philippines) using morphotectonic data. Zeitschrift fiir Geomorphologie, 94, 261-273. DEWEY, J. E 1988. Extensional collapse of orogens. Tectonics, 7, 1123-1139. FLORENDO, E E 1994. Tertiary arc rifting in northern Luzon, Philippines. Tectonics, 13, 623-640. FORSTER, H., OLES, D. & DEFANT, M. J. 1988. The Macolod Corridor: a rift crossing the Philippine island arc. International Symposium on the Geodynamic Evolution of Eastern Eurasian Margins, Abstracts. Paris, 53. FRICAUD,L. 1984. Etude g~ologique et structurale de la marge Ouest Palawan (Mer de Chine m~ridionale). Thesis, Universit6 Paris. GEARY, E. E. & KAY, R. W. 1989. Identification of an Early Cretaceous ophiolite in the Camarines NorteCalaguas islands basement complex, Eastern Luzon, Philippines. Tectonophysics, 168, 109-126. GERVASIO,E C. 1966. Age and nature of orogenesis of the Philippines. Tectonophysics, 4, 379-402. GLOCKE, A. 1980. Geology and hydrocarbon prospects of the Visayan Basin. Appendix V: Results of seismic interpretation. 77, BED unpublished report. HAECK, G. D. & KARIG, D. E. 1985. Strike-slip genesis of the Ilocos Norte melange. EOS, 64, 871. HAMILTON,W. 1979. Tectonics o f t he Indonesian Region. U. S. Geological Survey Professional Paper, 1078. HAWKINS, J. W. & EVANS, C. 1983. Geology of the Zambales range, Philippine islands: ophiolites derived from an island arc-back arc basin pair. In: HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of South-east Asian Seas and Islands. Part 2. American Geophysical Union Monograph, 27, 96-123 - - , MOORE, G. E, VILLAMOR,R., EVANS,C. & WRIGHT, E. 1985. Geology of the composite terrane of East and Central Mindanao. In: HOWELL, D. (ed.) Tectonostratigraphic Terranes of the CircumPacific Region. Circum-Pacific Council for Energy and Mineral Resources, Earth Sciences Series, 1, 437-463. HOLLOWA¥, N. H. 1982. The stratigraphic and tectonic evolution of Reed Bank, North Palawan and Mindoro to the Asian mainland and its significance in the evolution of the South China Sea. AAPG Bulletin, 66, 1357-1383. IRVING, E. M. 1950. Review of Philippine basement and its problems. Philippine Journal Science, 79, 267-307. KAR~G, D. E. 1983. Accreted terranes in the northern part of the Philippine archipelago. Tectonics, 2, 211-232. --, SAREWrrz, D. R. & HAECK, G. D. 1986. Role of strike-slip faulting in the evolution of allochtonous terranes in the Philippines. Geology, 14, 852-855. LETOUZEY,J. E, SAGE, MULLER, C. 1987. Geological and structuctural maps of Eastern Asia. Institut Francais du Petrole, Paris. LEWIS, S. D. & HAYES, D. E. 1984. A geophysical study of the Manila Trench, Luzon, Philippines; forearc basin structural and stratigraphic evolution. Journal of Geophysical Research, 89, 9196-9214. MALETERRE,P., STEPHAN,J. E, ANDREIEFF,P., BELLON,H., CHOROWICZ,J., BOmAT,J. M. & BALCE,G. R. 1988. The southern Central Cordillera of Luzon: a multistage Upper Eocene to Pleistocene arc deformed on the northern end of the Philippine strike-slip fault. International Symposium on the Geodynamic Evolution of Eastern Eurasian Margin (Abstracts), Paris 13-20 September 1988. MARCHADIER,Y. & RANG~N,C. 1990. Polyphase tectonics at the southern tip of the Manila trench: MindoroTablas islands, Philippines. Tectonophysics, 183, 273-288. MCCABE, R., ALMASCO,J. & DIEGOR, W. 1983. Geologic and paleomagnetic evidence for a possible Miocene collision in Western Panay, central Philippines. Geology, 10, 325-329. MCCAFFREY, R. 1992, Slip vectors and streching of the Sumatran fore arc. Geology, 19, 881-884. MCKENZIE, D. E 1972. Active tectonics of the Mediterranean region. Journal of Geophysical Research, 30, 109-185. MITCHELL, A. n. G., HERNANDEZ, E & DE LA CRUZ, A. E 1986, Cenozoic evolution of the Philippine archipelago. Journal of SE Asian Earth Sciences, 1, 3-22. MOORE, G. E & SILVER,E. A. 1983. Collision Processes in the northern Molucca Sea In: HAYES,D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. Part 2. American Geophysical Union Monograph, 27, 360-372 PELLETIER, B., STEPHAN,J. E, BLANCHET,R., MULLER,C. & HU, H. N. 1985. L'rmergence d'une zone de collision active tl la pointe sud de Taiwan (Prninsule
DOCKING & ESCAPE TECTONICS, S PHILIPPINES d'Hengshun): tectoniques superpos6es et mise en 6vidence d'une tectonique miocene moyen. Bulletin de la SocMtg gdologique de France, 8, I, 161-171. PIm~T, N. & STEPHAN,J. E 1990. The Philippine wrench fault system in the Ilocos Foothills, Northwestern Luzon, Philippines. Tectonophysics, 183, 207-224. PUBELLIER, M., DEFFONTAINES, B., QUEBRAL, R. & RANG~N, C. 1994. Drainage network analysis and tectonics of Mindanao, Southern Philippines. Geomorphology, 9, 325-342. , QUEBRAL, R., DEFFONTAINES, B. & RANGIN, C. 1993. Neotectonic map of Mindanao (Philippines). 1:800,000 with explanatory notes. Asia Geodyne Co. Manila. , RANGIN, C., DEFFONTAINES,B., MULLER, C., BUTTERLIN,J. & MANZANO,J. 1991. The Mindanao, collision zone, a soft collision event within a continuous strike-slip setting. Journal of Southeast Asian Earth Sciences, 6, 239-248. QUEBRAL, R., PUBELLIER,M. & RANGIN, C. 1995. Eastern Mindanao, Philippines: transition zone from a collision to strike-slip environment. Tectonics, in press. RANGIN, C., DAHRIN, D., QUEBRAL, R. & THE MODEC SCIEYrIFIC PARTY 1995. Collision and strike-slip faulting in the northern Molucca Sea, (Philippines and Indonesia): preliminary results of a morphotectonic study. This volume. RANGIN, C., MULLER, C. & PORTH, H. 1989. Neogene Geodynamic Evolution of the Visayan Region. In: PORTH, H. & VON DANmLS, C. H. (eds) On the geology and hydrocarbon prospects of the Visayan Basin, Philippines. Geologisches Jahrbuch, B 70, 7-28. , STEPHAN, J. E, BLANCHEr, R., BALADAD, D., BOUYSSE, P., ET AL. 1988. Sea-Beam survey at the southern end of the Manila trench. Transition between subduction and collision processes, offshore Mindoro Island, Philippines. Tectonophysics, 146, 261-278. ., BUTTERLIN, J., BELLON, H., MOLLER, C., CHOROWICZ, J. (~ BALADAD, D. 1991. Collision n6og6ne d'arcs volcaniques dans le centre des Philippines: stratigraphie et structure de la chaine d'Antique 01e de Panay). Bulletin de la Soci6t6 g6ologique de France, 162, 465--477. , & Muller, C. 1985. Middle Oligocene oceanic crust of the South China Sea. jammed into Mindoro Collision Zone, Philippines. Geology, 13, 425-428.
523
RANNEFT, T. S. M., HOPKINS, R. M. J., FROELICH,A. J. & GWlNN, J. W. 1960. Reconnaissance geology and oil possibilities of Mindanao. AAPG Bulletin, 44, 529-568. RATSCHBACHER,J., FRISCH, W., LINZER, H. G. & MERLE, O. 1991. Lateral extrusion in the Eastern Alps. Tectonics, 10, 245-271. RINGENBACH, J. C. 1992. La Faille Philippine e t l e s cha~nes en ddcrochement assocides (Centre et Nord de Luzon). These de Doctorat et Travaux de l'Institut de G6odynamique de Nice, 16. ~, STEPHAN, J. E, MALETERRE, PH. • BELLON, H. 1993. Structure and geological history of the Lepanto-Cervantes releasing bend on the Abra River Fault, Luzon Central Cordillera, Philippines. Tectonophysics, 183, 225-242. SAJONA, F., MAURY, R., BELLON, H., COTTEN, J., DEFANT, i . & PUBELLIER,M. 1993. Initiation of subduction and the generation of slab melts in Western and Eastern Mindanao, Philippines. Geology, 21, 10071010. SAREWITZ, D. & KARIG, D. E. 1986. Processes of allochthonous terrane evolution in Mindoro Island, Philippines. TECTONICS,5, 525--552. & LEWIS, S. D. 1991. The Marinduque intraarc basin, Philippines: basin genesis and in situ ophiolite development in a strike-slip setting. Geological Society of America Bulletin, 103, 597-614. SENGOR, A. M. C., GORUR, N. & SAROGLU, E 1985. Strike-slip faulting and related basin formation in zones of tectonic escape; Turkey as a case study. In: BIDDLE, K. T. & CHRISTIE-BLICK, N. (eds) Strike-Slip Deformation, Basin Formation and Sedimentation. Society of Econonomic Paleontologists and Mineralogists Special Publication, 37, 227-264. TAPPONN/ER, P., PELTZER, G., LEDAIN, A., ARMIJO, R. & COBBOLD, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611-616. WILLIS, B. 1937. Geologic observations in the Philippine islands. Natural Resources Council of the Philippines Bulletin, 13. WOLFE, J. A. & SELF, S. 1983. Structural lineaments and Neogene volcanism in southwestern Luzon. Sea. In: HAYES, D. E. (ed.) The Tectonic and Geologic Evolution of South-east Asian Seas and Islands. Part 2. American Geophysical Union Monograph, 27, 157-172.
Thermochronological and geochemical constraints on the tectonic evolution of northern Papua New Guinea R V. C R O W H U R S T , K. C. HILL, D. A. F O S T E R & A. R B E N N E T T
Victorian Institute of Earth and Planetary Sciences, School of Earth Sciences, La Trobe University, Melbourne, Victoria, 3083, Australia
Abstract: The Bewani-Torricelli-Prince Alexander Mountains, along the northern margin of Papua New Guinea, probably formed as a tholeiitic island arc in the Late Eocene-Early Oligocene. The arc is interpreted to have accreted to the margin by the Late Oligocene, but may have formed on a ribbon of extended continental crust along the New Guinea margin. Inferred roll-back of the subducting slab beneath New Guinea in the Early Miocene placed the margin into extension, creating starved graben in northern New Guinea and causing regional subsidence. Near the graben, metamorphiccore complexes resulted as the upper crust was pulled off the lower crust along low angle detachments such that lower crustal rocks cooled rapidly from temperatures >500°C. The two inferred core complexes that have been dated show rapid cooling from 27-23 Ma and 20-18 Ma. Continued subduction beneath New Guinea resulted in formation of the Maramuni arc in the Middle Miocene and the end of extension. In the Late Miocene, collision of the Melanesian arc caused regional uplift of all basement of northern Papua New Guinea, mainly from 8-5 Ma, causing at least 3-4 km of denudation. The compressional deformation propagated south causing uplift, denudation and cooling in the Papuan Fold Belt at c. 4 Ma, but is continuing at the present.
New Guinea is a prime example of arc-continent collisions, one of which continues today. The recent and ongoing events make New Guinea an important area to study in order to understand arc-continent collision, terrane accretion, the consumption of microplates and the role of extension in the collision process. The collision zone in northern Papua New Guinea is the New Guinea Mobile Belt (Fig. 1), variously interpreted to be underlain by Palaeozoic continental crust or accreted Tertiary oceanic microplates. The Mobile Belt is seismically active and the neotectonics and Mio-Pliocene collision of the Melanesian arc with the Australian continent are moderately well understood. However, several models have been proposed for previous collision, accretion or docking events of arcs/terranes, particularly in the Eocene and the Late Oligocene-Early Miocene. Pigram & Davies (1987) and Struckmeyer et al. (1993) proposed docking of the Sepik or North New Guinea terrane in the Late Oligocene causing a compressional orogeny resulting in exposure of Lower Cretaceous medium to high grade metamorphic rocks in the Mobile Belt, with the Sepik-Ramu Basins as foreland basins (Fig. 2). In contrast, Hill et al. (1993) suggested that the Late OligoceneEarly Miocene was a time of extension in northern New Guinea, creating the SepikRamu Basins as rift basins. They proposed that the exposure of the metamorphic rocks was
due to tectonic delamination of the crust creating a metamorphic core complex. Critical areas to resolve these conflicting models are the Bewani-Torricelli Mountains separating the Sepik and Aitape Basins, and the metamorphic provinces immediately south of the Sepik Basin (Fig. 2). Although along-strike and largely basic in nature, the Bewani-Torricelli Mountains may not be akin to the accreted Palaeogene Melanesian arc in the Adelbert and Finisterre Ranges (Fig. 2) as they include Upper Cretaceous igneous and metamorphic rocks. In order to test the models and determine the tectonic history, this paper presents the results of field mapping, apatite fission track thermochronology and geochemical analyses of basement rocks in the collision zone, from around the Sepik Basin (Fig. 1). Field mapping defines the present structure and stratigraphy, fission track analyses delineate the Neogene events and geochemical analyses help to constrain the Palaeogene tectonogenesis.
Stratigraphy and structure of northern Papua New Guinea The study area around the Sepik Basin can be divided into four main structural-stratigraphic units: the Mobile Belt metamorphic and ophiolitic
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 525-537.
525
526
r'. v. CROWHURST
I
140° I Caroline Plate I"ml~v, . ~ ~ "
500 km A_AdelbertArc Arc
~ h ~ "~ J ~/ ~ ' - ' - " ~ ~ ~ i t a ~r"~pe Basin ~ ......
F - Finisterre
SSP- Solomon Sea Plate ~ ~ 120°~
140°~
-'0,~
~, : -
j
/ ~ 160°
/
/
~
° t 20°. ~ /
racitic
~ ~
.~,~"~.
0 -B"i s m a r c k "
P O
~
°~ite
;
-'~ l~,~ f
-ea
Plate
~
.
L
.7 ~
r/ .
r
PapuanUltramafica
'%127 " \
NG- NewGuinea SP- Solomon SeaPlate BP - Bismark Plate O - Melanesian Arc
~
~a~in~s~. . . . . .
0°_
(
ET AL.
3
,
! q e ~ ~"-~ .
'
~
..
Indo-/1
A
Australian Plate
I 150 °
i
Fly Platform A
Komewu 1& 2
I
I Darai Plateau
Papuan Fold Belt Gobe
New Guinea Mobile Belt Kubor Anticline
Strike-slip faults? ®1× OlX
Melanesian Arc
Marum Ophiolite
Adelbert-FinisterreHuon Block A l
SW
NE V.E. approx, x 2
\ Continental Basement i
Approx. 100 km
i
Ramu Basin Oceanic Basement (Early Miocene)
Fig. 1. Plate tectonic setting of New Guinea showing the major structural zones, after Dow (1977) and the area of this study, around the Sepik-Ramu and Aitape Basins and the Bewani-Torricelli Mountains. The inset shows the present plate convergence vector and the location of the Melanesian arc. The schematic and stylized section illustrates the complex structure of folding and thrusting which has formed in response to the Late Miocene to Recent collision of the Papua New Guinea margin with the Melanesian arc. rocks; the Sepik Basin, the Bewani-TorricelliPrince Alexander Mountains and the Aitape Basin (Fig. 2).
Mobile Belt metamorphic and ophiolitic rocks Between the Papuan Fold Belt, formed in the Pliocene (Hill 1991), and the Sepik Basin is a highly faulted zone of Cretaceous and Palaeogene metasediments and medium-high pressure metamorphic schists, gneisses and amphibolites with Early Miocene cooling ages. These are interpreted to have been thrust faulted against Middle Miocene volcanics and to be overlain by an uppermost thrust sheet, preserved in mountain tops, comprising the April ophiolite, probably of Palaeogene age (Fig. 2; Rogerson et al. 1987).
The Sepik Basin Francis (1990) describes the Sepik Basin as a Pliocene to Pleistocene successor basin which
unconformably overlies three Miocene infrabasins. Reflection seismic data suggest that these are rift-graben or pull-apart basins bordered by Lower Miocene carbonates, but inferred to have a condensed Lower Miocene basinal sequence (Doust 1990; Hill et al. 1993). The basins are filled with mainly bathyal, but shallowing upwards, Middle to Upper Miocene volcaniclastic rocks. The basins are probably underlain by Cretaceous to Lower Tertiary basic igneous and metamorphic rocks of oceanic origin (Doust 1990), but may include areas of crystalline Palaeozoic continental crust (Rogerson et at. 1987).
Bewani-Torricelli-Prince Alexander Mountains These mountains expose sediments of the Sepik and Aitape Basins on the flanks, including deformed Pliocene sediments, but the mountain core comprises Palaeocene to Lower Miocene Bliri Volcanics and the Cretaceous-Palaeogene
TECTONIC EVOLUTION OF NORTHERN PAPUANEW GUINEA
527
P l ~ c e ~ - Q u t ernary
Be~la~o'~T~ee/l i uuas
~
50,
,
100kml
[
I depeat~ (sed, 'volt,) F.ocene-Mi~.ene arc n ~ t e r i ~ (sed, vok)
Eocene,-Mioeene sediments, intru~ives and e x t r u v ~
7+1
~1:1
Meltable-Miocene metamorphica mid ledimenla of the Mobile Belt
b~. ,17.6bt-19.3hbMa
Basin 6t4
Radiometric dates axe by : ~ff"~¢.~ Potassium-Argon analyses on... bt -"biotite Rubidium-Strontium analyses on... hb - amphibole In - muscovite Rb - whole rock ph - pheugite Rbm- muscovite gl - glauco~aue ZFTA - Zircon Fission Track Age ~ Apatite Fission Track Ages BTFZ - Bewa~i-Tomce]ti Fault Zone 6~1 (dosure temp. - ll0+l-10oc) Ultramafics south of Sepik Basin are termed April Ophiolites (Rogerson et al, 1987)
Fig. 2. The general geology of northern Papua New Guinea, showing the Mobile Belt to the south, the Bewani-Torricelli-Prince Alexander Mountains to the north and the Finisterre-Adelbert (Melanesian) arcs to the east. Also shown are K-Ar and Rb-Sr metamorphic cooling ages from Rogerson et al. (1987), Hutchison & Norvick (1980) and Page (1976), which indicate that cooling from high temperatures occurred during the Cretaceous and in the Late Oligocene to Early Miocene. Thirty-four apatite fission track ages overlay the geology, indicating that regional cooling from >110-<60°C occurred during the Late Miocene-Early Pliocene. The single zircon fission track age indicates an earlier cooling period through c. 200--250°C. The ultramafic rocks south of the Sepik basin are termed the April ophiolite after Rogerson et al. (1987). Most intrusive and extrusive rocks in the Mobile Belt are from the Miocene Maramuni arc (Rogerson et al. 1987).
Torricelli Intrusive Complex (Hutchison & Norvick 1980; Rogerson et al. 1987). The Torricelli Intrnsives consist of dolerites, gabbros and diorites, whereas the Bliri Volcanics comprise basalts, andesites and volcaniclastic rocks. The Prince Alexander Mountains in the east include amphibolite and gneiss with Mesozoic and Early Miocene metamorphic cooling ages, together with igneous intrusions in metapelites, the latter extending beneath the Sepik Basin (Hutchison & Norvick 1980; Rogerson et al. 1987; Doust 1990). Aitape Basin
Kugler (1990) described the Aitape Basin as a mildly deformed half-graben which deepens rapidly to the south towards the Bewani-Torricelli fault zone (Fig. 2). Using reflection seismic data, he
showed preserved normal faults within the basin and suggested that the Aitape Basin was connected to the Sepik Basin prior to c. 10 km of uplift of the Bewani-Torricelli Mountains. Recent detailed fieldwork (Hilyard et al. 1994) has better defined the main stratigraphic units of the southern margin of the Aitape Basin (Fig. 3). Restricted marine deposition is inferred before the Early Miocene followed by deposition of the coarse volcaniclastic Wogamush Beds, correlated with subduction beneath PNG initiating the continental margin Maramuni arc (Rogerson et al. 1987; Hill et al. 1993). This was followed in the Late MiocenePliocene by deep-water turbidite deposition of the Barida Beds, synchronous with thrust-related uplift of the Bewani-Torricelli Mountains, from which most of the shoaling sediments of the Aitape Basin were derived (Fig. 3).
528 B S
P.V. CROWHURST Apatitefission track age(Ma) recording coolingfrom >110°C to <60°C
]
...._
(
,- ... -- "" ~ - - "- ..
-'""
----
/11
.-V'" /
,,,,'"~
•
older basement
0
~lt
f[ ~+~ 5 ~/11_+6
ET AL.
.
.
BewamMountmns
k volcanics
',,/
."
4
~' 500o
Bewani-TorriceUi Fault Zone
/
\ J
I \ ~
""
... ~ - -
t
40o0
,,
17'
./y
/Z'"--''"/'--'',
,,,,,,,,,,,',',',',"
[.;.~.,~.q
2 kilometres
3 ~000.
O~-~
~..~ ~
.....
,~
, ,.,.,',',.'..',',L..',',.",',',','..',
Fig. 3. An interpretative structural cross-section trending N - S from the southern margin of the Aitape Basin, through the Bewani Mountains and into the Sepik Basin (see Fig. 2 for location). The boundary between the Bewani Mountains and the Aitape Basin, known as the Bewani-Torricelli fault zone, is dominated by an imbricate fan structure with south over north folding and thrusting, probably with a component of sinistral strike-slip movement. The fission track data from the Bewani Mountains indicate that the peak of denudation of the basement, and hence the major deformation, probably occurred during the Late Miocene.
Structure of the north Bewani Mountains-Aitape Basin The southern boundary of the Aitape Basin is marked by the east-west trending BewaniTorricelli fault zone. Stream traverse mapping of the northern Bewani and Torricelli Mountains demonstrated basement rocks uplifted to the south across the fault zone, locally overturning the sedimentary section. Faults and brecciated zones were observed to be generally steep-dipping, where seen in the mountains, but no estimation of slip direction was possible due to poor outcrop. Multiple faults are inferred within the Bewani-Torricelli fault zone, but it is difficult to correlate these and the intervening thrust sheets between widely spaced stream sections. Therefore, the structural style is illustrated on a north-south section across the Bewani Mountains (Fig. 3). The mountains are interpreted to have been upthrust to the north on a series of steep-dipping imbricate thrusts which may coalesce downwards into a shallow-dipping detachment or a steep-dipping strike-slip fault. The tightly to isoclinally folded nature of both the basement and the cover suggests that folding may have occurred before the steep thrust faults broke through. However, the lack of any cleavage in the sediments and the absence of any metamorphism (except hydrothermal alteration) indicates that the rocks were relatively cool when deformed.
The pattern of steep-dipping imbricate thrust faults shown in Fig. 3 resembles that of a positive flower structure formed in a transpressional regime, as proposed by Doust (1990) for the BewaniTorricelli-Prince Alexander Mountains. This is consistent with the narrow, ribbon-like nature of the mountains and their obliquity to the direction of Indo-Australian and Pacific plate convergence (Figs 1 & 2). It is also consistent with the first motion solutions from earthquakes in the region which record both overthrust and transcurrent movement (Hutchison & Norvick 1980). The faulting shown in Fig. 3 clearly deforms the Neogene section. Therefore, the BewaniTorricelli-Prince Alexander Mountains are interpreted as positive flower structures due to leftlateral transpressive motion along the BewaniTorricelli fault zone during the Late MiocenePliocene. This is consistent with the time of accretion of the Adelbert and Finisterre arcs, c. 500 km along-strike to the east (Figs 1 & 2). The estimate of the time and amount of uplift and denudation can be tested by apatite fission track analysis.
Apatite fission track analysis A total of 51 surface samples were processed from intrusive and metamorphic complexes in northern Papua New Guinea and 34 yielded apatite for fission track analysis (Table 1). Zircon yields were
141-09-30 141-11-45 141-25-30 141-14-05 141-11-20 141-11-20 142-46-35 142-46-30 142-47-10 142-46-45 142-47-30 142-47-30 143-00-10 143-00-10 142-47-10 142-09-30 142-08-45 142-08-45 142-09-00 142-54-30 144-49-20 141-10-00 141-11-45 141-09-30 141-14-05 141-11-20 143-13-10 143-09-10 142-09-20 141-09-45 141-09-45 142-09-00 142-25-00 145-18-30 145-43-45
17 8# 127 13t 23? 24~ 28t 29t 32t 34t 35t 36? 42t 43~ 45? 52? 60t 617 69? 79? 081at 2# 4# 011a# 14# 25# 37# 39# 053+54# 58# 59# 68# 102# K8-22# K8-11#
3-20-05 4-18-30 4-17-45 3-36-40 3-14-30 3-14-30 4-26-15 4-26-20 4-25-20 4-25-40 4-25-15 4-25-15 4-15-45 4-15-45 4-47-15 3-20-50 3-25-35 3-25-35 3-24-10 3-26-30 5-21-15 3-20-50 4-18-25 4-16-20 3-36-40 3-14-30 4-39-30 4-39-30 3-21-30 3-21-40 3-21-40 3-24-20 5-00-30 5-45-30 6-15-00
Latitude
320* 250* 320 360 600 605 180 150 ------200* 310 780 780 1180 300* 150 -230 200* 360 600 --460 570 650 1140 280 1700" 600
Elevation (m)
diorite gneiss diorite diorite granite granite granodiorite diorite meta-felsic granite diorite granodiorite diorite diorite andesite granite ultramafic ultramafic meta-sed.? gneiss volc. sandstone diorite amphibolite garnet gneiss diorite granite gneiss meta-felsic granite granite granite meta-mafic greywacke diorite gneiss-granite
Lithology
>38 >38 42.6-20.1 >38 97.5-14.4 97.5-14.4 >38 >38 >38 >38 >38 >38 19.7-22.8 19.7-22.8 >38 97.5-14.4 97.5-14.4 97.5-14.4 97.5-14.4 97.5-38.0 ->38 >38 >38 >38 97.5-14.4 >38 >38 97.5-14.4 97.5-14.4 97.5-14.4 97.5-14.4 -c. 60 c. 150-190
Chronostrat, age (Ma)$
2 20 20 20 11 12 6 13 7 20 20 21 6 5 16 6 8 2 27 26 7 5 20 21 20 16 20 20 35 20 20 30 30 30 10
Number of grains
2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 2.943(3973) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1.397(6285) 1,397(6285) 1.212(5488) 0.784(3723)
Standard track density (×106 cm -2) 0.185(2) 0.032(4) 1.841(85) 2.288(84) 0.101(6) 0.055(2) 1.107(27) 1.683(133) 1.518(47) 1.815(82) 0.4817(62) 0.383(28) 1.975(40) 0.638(12) 0.107(15) 0.7136(7) 1.671(63) 1.389(10) 1.069(82) 0.521(74) 0.373(8) 0.025(4) 0.017(21) 0.138(212) 0.020(11) 0.042(21) 0.005(3) 0.002(1) 0.005(5) 0.005(3) 0.012(6) 0.018(40) 0.015(19) 0.015(14) 14.02(2184)
Fossil track density (xl05 cm -2) 0.861(93) 2.780(343) 132.6(6121) 158.9(5835) 10.30(609) 12.91(467) 82.98(2024) 118.1 (9329) 61.89(1916) 135.3(6113) 40.18(5171) 35.28(2578) 68.84(1394) 21.69(408) 18.18(2536) 53.41(524) 103.2(3890) 69.17(498) 70.87(5434) 35.82(5.87) 37.58(805) 0.575(91) 1.239(1504) 4.606(7068) 0.924(498) 1.480(738) 0.208(119) 0.132(87) 0.434(399) 0.303(196) 0.400(193) 0.304(672) 0.437(553) 0.638(615) 37.90(5906)
Induced track density (×106 cm -2)
t counted by Peter Crowhurst (Zeta value = 343.5 ± 5), # counted by Kevin Hill (Zeta value = 350.0±5) ; * Denotes estimation of elevation $ Chronostraigraphic ages are estimated from Rogerson et al. (1987), Hutchinson & Norvick (1980) and Page (1976); Fission track age is central age after Galbraith & Laslett (1993); note K8-11 is a zircon fission track analysis; Track densities in ( ) are the number of tracks counted.
Longitude
Sample Number 92PNG-
Table 1. Apatite fission track analytical results and details from northern Papua New Guinea
68.7 95.4 100 99.8 87.2 99,1 77.4 99,7 97.9 100 99.7 1t30 76.3 98.8 99.7 96.1 100 4.9 99.8 98.3 60.4 15.8 76.4 35.4 8.9 5.1 19.5 99.4 99.1 62.7 66 4 20.2 9.9 0
(%)
Chi square probability
10.9 5.9 7.0 7.3 5.0 2.2 6.7 7.2 12.4 6.8 6.1 5.5 14.5 12.5 2.5 5.7 6.9 8.6 6.4 6.2 4.2 10.8 3.4 7.4 5.4 7.2 6.3 2.8 3.1 3.8 7.6 15.0 8.4 5.1 17.5
+ ± ± ± + ± ± ± ± + ± ± ± + ± ± ± ± ± ± ± ± ± + ± ± ± ± ± ± ± + + ± ±
7.8 3.0 0.8 0.8 2.0 1.5 1.3 0.6 1.8 0.8 0.8 1.0 2.3 3.7 0.7 2.2 0.9 2.7 0.7 0.7 1.5 5.5 0.8 0.5 1.9 2.0 3.7 2.8 1.4 2.2 3.2 3.0 2.0 1.5 1.3
Fission track age~[ (Ma)
Ix#
>
>
L
= ~=~
Z O
©
~-©
c~ O Z c~ r~ <
P.v. CROWHURST ET AL.
530
generally poor, and only one zircon fission track analysis is reported. Methods followed those recently reviewed by Foster & Gleadow (1992). Fission track ages were calculated using the zeta calibration method and standard fission track age equation (Hurford & Green 1982). Errors were calculated using the techniques of Green (1981). The zircon sample was analysed similarly, following the methods of Gleadow & Lovering (1978).
Sample results The 51 rocks processed comprised 29 igneous, 19 metamorphic and 3 sedimentary samples. Apatite yields for the 34 samples counted were variable from poor (1-10grains) to very good recovery (>20 datable grains): Due to the consistency of young fission track ages and low uranium concentrations of many samples very few confined tracks were present and none were measured. Similarly, the number of tracks counted was usually low, resulting in low precision on single grain ages, such that statistical tests were used with caution. The age calculations of all but two samples pass a Chi2 test at the 95% confidence level, indicating that the single grain ages for each sample can be treated as a statistically valid single population. The pooled age is presented for those samples. Two samples failed the Chi 2 test suggesting that the dated grains represent multiple populations. For these samples the central fission track age is presented (Galbraith & Laslett 1993)
Interpretation The 34 samples listed in Table 1 show a spectrum of ages from 2.2_+1.5Ma to 16.9_+3.6Ma.
Twenty-one of the samples lie within the range 5-8 Ma with a weighted mean age of 6.0 _ 0.3 Ma which indicates a major cooling event at this time. The distribution of ages is shown on the map of northern Papua New Guinea in Fig. 2 showing no apparent trend to the data. Indeed the ages are consistent throughout a very large area (c. 500 km E-W and c. 200 km N-S). Eight of the samples analysed were collected from a north-south traverse across the Torricelli Mountains over a wide spread of elevations. The data are shown on a smoothed topographic profile in Fig. 4. Most notable on this traverse is the single older sample at a higher elevation. Elsewhere in the Bewani-Torricelli Mountains samples with Late Miocene fission track ages are present at similar elevations suggesting relative fault displacement of hundreds of metres, uplifting the rocks with young ages (Fig. 4). This is consistent with the 'flower' or 'pop up' structural interpretation presented above. In addition, the 5-8 Ma apatite fission track ages throughout the mountains are consistent with uplift in the Late Miocene-Pliocene, dividing the previously continuous Sepik and Aitape Basins. Fission track ages are plotted against elevation in Fig. 5 in which the concentration of ages between 5-8 Ma is apparent, indicating a period of rapid cooling during the Late Miocene-Early Pliocene. As the samples are currently at surface, the authors interpret this rapid cooling event to be related to regional uplift and denudation of northern Papua New Guinea. Samples from the Marum ophiolite and Adelbert accreted arc record ages of 5 +_ 1 and 8 _ 2 Ma respectively, suggesting that both were denuded in the Late Miocene (cross-section of Fig. 1 and Fig. 2). Although largely constrained by only one sample, the elevation vs. age plot in Fig. 5 shows a possible slight trend of increasing fission
Ct -
-
2000 1500
~
-
-1000
-
~
~
r
Sample # from Table I from right to left (prefix 92PNG): 052, 053+054, 0 5 8 , 0 5 9 , 0 6 9 . 0 6 8 , 0 6 1 , 0 6 0 . t I I I ~1 I I 1 2 3 4 5 6 7 142o08'30" DISTANCE (kilometres) 3026,00 ,,
+tl~....~ 6+2 3-.~ ~-...~
, , ' 5 8
9 142009'30 '' 3°21 '00"
F i g . 4 . A north-south topographic profile across the Torricelli Mountains with fission track ages and interpreted thrust faulting. See Fig. 2 for location. The bulk of the ages are under 10 Ma and reiterate the Late Miocene-Early Pliocene regional cooling event. The single anomalous sample recording an age of 17 Ma could represent an earlier stage of denudation in the area and is now closely juxtaposed by a younger cooling age due to later faulting. This interpretation is in general agreement with earlier discussion of thrusting and 'pop-up'/flower structures (Fig. 3).
TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA 1200 :
"9
Miocene. This collision caused deformation throughout the Mobile Belt and the Papuan Fold Belt (cross-section of Fig. 1). Thus the thermal history of northern Papua New Guinea is entirely consistent with that of the Papuan Fold Belt to the south, where Hill & Gleadow (1989) recorded Pliocene uplift and denudation from apatite fission track analyses. No evidence for older orogenesis is preserved in the apatite fission track data, so the higher temperature thermal history must be assessed by studying the cooling of igneous and metamorphic rocks.
!
1000
800
-
--
I
;
,,"m
I
[]
600 I
D
531
I
Thermochronology
m 400
Mobile Belt I,
[]
200
0 0
'
'
5
10
's 1
20
Fission Track Age (Ma)
Fig. 5. Fission track age versus elevation plot of all
samples with true recorded elevation, excluding float, all showing 1 ~ error. The bulk of the samples lie within 2-10 Ma with a concentration around 5-8 Ma, which suggests this was the time of maximum cooling and denudation, due to regional uplift. It is possible that a slight increase in age occurs with elevation, which is typical of upthrusted rocks; i.e. samples at the top cool first thus recording the oldest age which gradually decreases with depth.
track age with elevation typical of fission track data sets from mountain ranges. Also, the similarity of ages (c. 5-8 Ma) from 150-1200 m supports the hypothesis of rapid cooling at that time. The one sample from high elevation was probably slightly below the temperature of total track annealing and records a mixed age with fission tracks preserved from both before and after the Late Miocene cooling event. Rocks at lower elevations were more deeply buried and hence hotter, with no fission tracks present prior to uplift and cooling, so record the age of cooling. In general, apatite fission track analyses of 34 basement samples from northern Papua New Guinea demonstrate regional rapid cooling from temperatures of >100°C in the Late MiocenePliocene. The cooling was due to denudation caused by regional uplift which resulted from continent-arc collision starting in the Late
Page (1976) and Rogerson et al. (1987) presented K-At and Rb-Sr ages from metamorphic rocks in northern Papua New Guinea. In the Mobile Belt, K-At apparent ages of hornblende and biotite are clustered around 27 and 23 Ma respectively. The exposed basement around Amanab has Permian K-Ar biotite ages in strong contrast to the Oligo-Miocene ages of the Mobile Belt. The hornblende and biotite data from the Mobile Belt (Landslip Range, Fig. 2) indicate relatively rapid cooling through the temperature interval of c. 500°C to 300°C (see McDougall & Harrison 1988 for a review of closure temperatures for argon in these phases) at a rate of_>50°C Ma -1 between 27 and 23 Ma. A single zircon fission track age from the Karmantina Gneiss, 400 km along-strike, is consistent with such cooling (Fig. 2 and Table 1). The zircon fission track age of 18 _+ 1 Ma records the time of cooling below c. 200-250°C. These data have been combined with the apatite fission track data to construct a cooling history for the metamorphic rocks of the New Guinea Mobile Belt (Fig. 6). The graph shows rapid cooling from >500°C to c. 150°C in the Early Miocene and a second cooling to surface temperatures in the Late Miocene to Pliocene. Regional subsidence of the Papua New Guinea margin occurred during the Middle Miocene with deposition of 1-2 km of limestone and/or thick volcaniclastic rocks (Home et al. 1990; Hill et al. 1993). Thus the geology indicates stable or increasing temperatures in the Middle Miocene rather than ongoing uplift or cooling (Fig. 6). Bewani-Torricelli-Prince Alexander Mountains
Hutchison & Norvick (1980) reported 36 K-Ar dates of rocks from the Torricelli Intrusive Complex. Twelve K-At hornblende analyses on dioritic-granodioritic rocks yielded Early
532
P . V . C R O W H U R S T ET AL.
Time 35
D 100
L) O
¢D
30
25
(Ma) 15
20
10
5
0
ewani-Torricelli Mountains (Apatite Track Fission Age) ~ a Si[iii~!!!ii i~:~i? M ~
New Guinea Mobile Belt
~ 200
( Zircon Fission Track Age) 20
300
Rapid cooling possibly due to extensional --" tectonics and denudation-" / ~ 5 : : • of the metamorphxc core
9
:!B,q~.__
apid cooling due to erosional denudation caused by arc collision
. . . . . . . . . 151via [r~-e~r age - olOtlte) Ma (K-Ar age - biotite)
?
,®
9
/ ~ 'A~N~ ~ " ~ C o o l i n g history of metamorphic rocks / [ , .,~'~ I from the Prince Alexander Mountains 500 -.--,/~ [Mid-crustal k'K~N,~X7 m rocks ~ .~p.~ 7'----~l>15kmdepth S e . , ~ . . . . . . . . ? 20Ma(K-Arage-hornblende)
fJ
37Ma(K-Ar age - hornblende) (Crystallization Age ?)
27Ma (K-Ar age - hornblende)
Fig. 6. Cooling histories of the mid-crustal rocks now exposed in the Mobile Belt and of the intrusive and metamorphic rocks from the Bewani-Torricelli-Prince Alexander Mountains constrained by fission track and K-At dating. K-Ar ages are from Rogerson et al. (1987), Hutchison & Norvick (1980) and Page (1976) as shown in Fig. 2. Closure temperatures used to generate the curves are: K-Ar hornblende indicates a closure temperature of c. 500°C; K-Ar biotite indicates a closure temperature of c. 300°C; zircon fission track ages indicate a closure temperature of c. 200-250°C and apatite fission track ages indicate a closure temperature of c. 120°C.
Oligocene ages of 30-41 Ma, with a mean of 37 Ma. Two other analyses yielded Late Cretaceous ages. Five K-Ar biotite analyses from these rocks yielded Early Miocene ages (16-23 Ma). In contrast, six K-Ar hornblende analyses of amphibolites from the Prince Alexander Mountains yielded Early Cretaceous ages (Fig. 2; Hutchison & Norvick 1980), but six metamorphic rocks from the same area dated by Rogerson e t al. (1987) yielded K-Ar biotite ages of 18 Ma and two K-Ar hornblende ages of 20 Ma. Thus the Torricelli Intrusive Complex appears to comprise mainly Oligocene intrusions into Cretaceous or older rocks, but there may have been a later minor phase of Early Miocene intrusions and local contact metamorphism. Alternatively, the Oligocene intrusions may have resided at temperatures of c. 400°C until the Early Miocene, below the blocking temperature of hornblende but above that of biotite. In the Early Miocene they then cooled below c. 300°C, giving an Early Miocene K-Ar biotite cooling age consistent with cooling of the metamorphic rocks.
These data have been combined with the apatite fission track analyses to construct cooling histories for the metamorphic and intrusive rocks of the Bewani-Torricelli Mountains (Fig. 6). The graph shows Oligocene crystallization and cooling then further rapid cooling in the Early Miocene to temperatures of c. 150°C. The rocks then underwent cooling to surface temperatures in the latest Miocene.
Geochemistry Sixteen surface samples collected from the Bewani-Torricelli Mountains, consisting of tonalites, diorites and gabbro were analysed for major and trace elements. The details of procedures, petrogenesis, figures, analytical and petrographic tables are available as Supplementary Publication No. SUP 18100 (11 pp) from the Society Library or the British Library Document Supply Centre, Boston Spa, Wetherby, N. Yorks, LS23 7BQ, UK. Interpretation of the results follows Pearce et al. (1984) and is outlined below.
TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA The results of the geochemical analyses can be summarized as follows: (1) the major elements (alkalis) suggest a sub-alkalic magma; (2) the chemistry of clinopyroxenes from mafic samples suggest a tholeiitic magma from an active margin setting; (3) Nb depletion in the tonalites suggests volcanic arc and collision-generated granites; (4) trace element plots for tonalite and mafic sampies are most akin to those from a tholeiitic to calc-alkaline volcanic arc; (5) MORB normalized plots suggest a subductionrelated setting, possibly with partial melting of subducted sediment and fractionation of clinopyroxenes; (6) although the precise sample ages are currently unknown, the geochemical trends are consistent with a simple genetic relationship between the samples.
Summary of main results 1. The geochemical analyses presented show a tholeiitic island-arc signature for the BewaniTorricelli samples and indicate a simple genetic relationship between the samples. Diorites and granodiorites dated by Hutchison & Norvick (1980) suggest a Late Eocene to Early Oligocene crystallization age. 2. The Bewani-Torricelli-Prince Alexander Mountains are interpreted as positive flower structures that owe their present relief to left-lateral transpressive motion along the Bewani-Torricelli fault zone, during the Late Miocene-Pliocene. This is consistent with the time of accretion of the Melanesian arc. 3. Apatite fission track analysis shows regional cooling of northern Papua New Guinea from temperatures of > 100°C during the Late MiocenePliocene. This indicates regional denudation contemporaneous with compressional deformation and Melanesian arc collision. 4. K-Ar thermochronology suggests rapid cooling of metamorphic rocks in the Early Miocene, from >500°C to c. 150°C. The cooling occurred from 25-20 Ma in the Mobile Belt and from 20-18 Ma in the Prince Alexander Mountains. In both cases these medium to high grade metamorphic rocks are in close proximity to similar rocks with Cretaceous to Permian K-At cooling ages.
Discussion The interpreted tholeiitic island-arc of the Bewani-Torricelli Ranges is probably of Early
533
Oligocene age, similar to the Adelbert and Finisterre (Melanesian) arcs to the east (Jaques 1976). However, the arc rocks appear to be overlain by the Miocene-Pliocene Sepik and Aitape Basins which received 'Wogamush' volcaniclastic rocks from the Maramuni arc within the Mobile Belt (Fig. 3). Therefore, unlike the Melanesian arc, the Bewani-Torricelli arc probably accreted prior to the Miocene or was formed on the northern margin of the Australian plate with subduction to the south. The structural and fission track data show that the present Bewani-Torricelli Mountains were upthrust in the Late Miocene-Pliocene, such that the Sepik and Aitape Basins were probably continuous in the Early Miocene, implying that the Bewani-Torficelli arc was then largely submerged. Reflection seismic data from both basins have been interpreted to show extensional faulting of basement (Kugler 1990; Hill et al. 1993). The cooling histories in Fig. 6 show two regional cooling events in northern Papua New Guinea occurring between 27-18 Ma and also at c. 6 Ma. The tectonic history of northern Papua New Guinea must include these two cooling events and account for the nature and amount of denudation involved. The rapid cooling of >300°C in the Early Miocene implies c. 10km of denudation assuming normal temperature gradients. The Early Miocene cooling ages have been interpreted to result from accretion/collision, causing subsidence of the New Guinea margin and creation of the Sepik-Ramu basins as foreland Basins (Pigram & Davies 1987; Struckmeyer et al. 1993). However, such rapid cooling is unusual in compressional orogenies and, as the metamorphic rocks cover thousands of square kilometres, the metamorphic detritus would have inundated adjacent basins. To the contrary, the Sepik Basin was a deep and probably starved basin with marginal carbonates in the Early Miocene (e.g. Doust 1990), despite having Early Miocene metamorphic terrains on both sides. Regionally, in the Mobile Belt and northern Fold Belt the inundation of sediment was in the Middle Miocene and comprised mainly volcanogenic material from the Maramuni arc (Hill et al. 1990, 1993). Therefore, it is inferred that the rapid Early Miocene cooling resulted from tectonic denudation during regional extension, which also created the Sepik graben and regional subsidence of the Papua New Guinea margin (Hill et al. 1993). During extension the upper crust was pulled off the lower crust, such that hot lower crustal rocks rose rapidly towards the surface as metamorphic core complexes. However, these lower crustal rocks did not reach the sur-face, but remained covered by up to 5 km veneer of extended upper crustal rocks, with low relief, causing little denudation. The
534
r'. v. CROWHURSTET AL.
Cretaceous to Permian K-Ar cooling ages for adjacent metamorphic rocks around Amanab and in the Prince Alexander Mountains suggest that these areas remained part of the upper crust during extension, and therefore show no Tertiary cooling. Following arc-continent collision in the Late Miocene, the rocks were thrusted, denuded and cooled giving the 6 _+2 Ma apatite fission track cooling ages. The metamorphic rocks of the Mobile Belt are in thrust contact with the unmetamorphosed April ophiolite (Fig. 2) which Rogerson et al. (1987) interpret as the uppermost thrust sheet cropping out in mountain tops. The ophiolite is probably Palaeogene in age (Rogerson et al. 1987) and outcrops occur over thousands of square kilometres, indicating that it was previously a significant area of oceanic crust. Like Rogerson et al. (1987) it is inferred that the ophiolite was thrust over the metamorphic rocks in the Late Miocene-Pliocene compressional deformation, consistent with the regional apatite fission track cooling ages and particularly the 5 Ma age for the Marum ophiolite. Therefore, prior to the Late Miocene there must have been a significant area of oceanic crust adjacent to the Mobile Belt and separating it from the Sepik Basin and BewaniTorricelli arc. Such a marginal basin must be included in any tectonic model. Note that an alternative model under consideration is that denudation of the metamorphic rocks could be due to sinistral transpression along wrench faults, similar to the Alpine fault of New Zealand. The detritus could have been deposited on the marginal basin and then been eroded during Late Miocene thrusting. However, an extensional model is preferred due to the rapid cooling rates, the apparently extensional graben and the regional subsidence of the entire margin.
Tectonic model As shown in this paper and elsewhere, the factual geologic data for northern Papua New Guinea are limited, so that any model is necessarily speculative. However, these new data further
constrain tectonic models, so we present a model that is consistent with the present dataset, illustrated in Fig. 7. Oligocene
In the Early Oligocene the New Guinea margin comprised continental crust as far north as the Mobile Belt, then a narrow marginal basin of Palaeogene oceanic crust bound by a ribbon of extended Cretaceous continental and oceanic crust. The marginal basin probably opened in the Late Cretaceous-Palaeocene, contemporaneous with Coral Sea spreading, as suggested by Pigram & Symonds (1991). Oceanic crust north of New Guinea was subducted to the north beneath an Oligocene tholeiitic volcanic arc, but slow subduction may also have occurred beneath New Guinea (Fig. 7; Hill et al. 1993) as suggested by Upper Oligocene-Lower Miocene intrusive rocks and Wogamush Volcanics in the Sepik Headwaters region (Rogerson et al. 1987). By the Late Oligocene, the arc had accreted to the ribbon of possible continental crust along the margin of New Guinea and had been partially subducted such that temperatures remained high. A new south-dipping subducting slab developed, with the trench to the north of the accreted arc. Early Miocene
Extension occurred in the New Guinea margin above the newly developed slab, probably due to roll-back of the slab hinge. The extension was focused in the relatively thicker and weaker continental crust and occurred along low angle faults through the crust (Fig. 7). As the upper crust was pulled off, the lower crust rose to within a few kilometres of the surface and cooled rapidly as a metamorphic core complex. This occurred in the Mobile Belt from c. 27-20 Ma and in the accreted arc (Prince Alexander Mountains) from 20-18 Ma. Adjacent rocks in the upper crust retained their Mesozoic or older age. As part of the regional extension, the ribbon of possible continental crust was extended to create
Fig. 7. Sequential cross-sections illustrating the Early Miocene extensional model for the evolution of northern Papua New Guinea (refer to D-D' on Fig. 1 for location). The Late Eocene passive margin of Papua New Guinea was disrupted by the onset of rapid convergence between the Australian and Pacific plates. The resultant subduction remote from Papua New Guinea, created the Caroline backarc basins and the Finisterre-Adelbert arcs. In the Late Oligocene accretion of arc material jammed the subducting plate, thus producing another southerly dipping subduction zone further to the north. Extension occurred in the continental crust above the newly formed slab during the Early Miocene, due to either roll-back of the slab hinge or slowing of convergence following collision of the Indo-Australian plate with Asia to build the Himalaya. During the Middle Miocene, partial melting of the subducting slab created the Maramuni arc and saw the onset of obduction of oceanic crust. The Pliocene saw the onset of major compression which produced large-scale thrusting and deformation in the Mobile Belt and caused the upthrust of the Bewani-Torricelli Mountains, which now separate the Sepik and Aitape Basins.
535
TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA
D
D Early Oligocene Late Eocene I
Pacific Plate
P a p u a N e w Guinea I
Island arc
Palaeogene marginal basin (Sontlnental crust . . . . . . . . . . . . . . . . . . . . . .
.... Mantle .... y . . . . . . . . . . . . . . . .
I
diorites granodiorite
,
i?
Late Oligocene
ccreted arc
Palaeogene marginal basin
.-?
Early Miocene
Sepik Graben 20-18 Ma
27-20 Ma
Rollback
Middle Miocene Incipient / obduction
Volcaniclastics /-\
Pliocene
N e w Guinea Papuan M o b i l e Belt Fold Obducted ophiolite Belt
_/
Sepik Basin
\
Prince Alexander 'metamorDhic rocks
Caroline Plate
Incipient low-angle / detached normal fault Metamorphic core complex
Maramuni Arc
N e w subduction zone forming
Bewani-Torricelli Mountains _ / Aitape Basin
!
536
P.V. CROWHURST ET AL.
the Sepik and Aitape Basins, which were continuous at that time. There was little uplift and erosion, so the basins were starved of sediment and carbonates were deposited around the margins (Wilson et al. 1993). The regional extension also resulted in Miocene subsidence of the whole New Guinea margin leading to deposition of thick Miocene carbonates.
Middle Miocene
Partial melting occurred above the south-dipping subducting slab resulting in extensive intrusions and volcanism to create the Maramuni arc. Abundant volcanogenic sediment was shed into the Sepik and Aitape Basins and into the deep basins abutting the Miocene platform carbonates to the south (Hill et al. 1990).
Late Miocene-Pliocene
Accretion of the Melanesian arc in eastern New Guinea and attempted subduction of the young and buoyant Caroline plate (Oligocene; Hegarty & Weissel 1988) placed the New Guinea margin into
compression. In the Sepik area the compression was initially taken up by obduction of a sliver of oceanic crust from the Palaeogene marginal basin, thrust over the Lower Miocene metamorphic rocks of the Mobile Belt. Continued compression caused more intense thrusting throughout northem and central New Guinea, resulting in regional uplift, denudation and cooling from 8-5 MR. The deformation propagated south into the Papuan Fold Belt in the Pliocene and the Sepik Basin and particularly the Ramu Basin underwent regional subsidence as foreland basins to the Melanesian arc. The oblique plate convergence was partitioned into transpression along the northern margin causing uplift, denudation and cooling of the Bewani-Torricelli-Prince Alexander mountains and N E - S W compression in the Mobile Belt and Fold Belt to the south (Abets & McCaffrey 1988). This work has been sponsored by LL&E Sepik Pty Limited, Mobil New Exploration Ventures Company, Mobil Exploration Niugini and Esso Australia Ltd. Fission track neutron irradiations are supported by a grant from the Australian Institute of Nuclear Science and Engineering. The views presented here are entirely those of the authors and do not necessarily correspond to those of any company.
References ABERS, G. & MCCAFFREY,R. 1988. Active deformation in the New Guinea fold-and thrust-belt: seismological evidence for strike-slip faulting and basement involved thrusting. Journal of Geophysical Research, 93, 13 332-13 354. DOUST, H. 1990. Geology of the Sepik Basin, Papua New Guinea. In: CARMAN,G. J & CARMAN,Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 461-478. Dow, D. B. 1977. A geological synthesis of Papua New Guinea. Bureau of Mineral Resources Australia Bulletin, 201. FOSTER, D. A. & GLEADOW,A. J. W. 1992. The morphotectonic evolution of rift-margin mountains in central Kenya: Constraints from apatite fissiontrack thermochronology. Earth and Planetary Science Letters, ll3, 157-171. FRANOS, G. 1990. The North New Guinea Basin and associated infra-basins. In: CARMAN, G. J & CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 445-460. GALBRA1TH, R. E ~: LASLETT, G. M. 1993. Statistical models for mixed fission track ages. Nuclear Tracks and Radiation Measurements, 21, 459-470. GLEADOW, A. J. W. & LOVERING,J. E 1978. Thermal history of granitic rocks from western Victoria: a
fission track dating study. Journal of the Geological Society of Australia, 25, 323-340. GREEN, P. E 1981. A new look at statistics in fission track dating. Nuclear Tracks, 5, 77-86. HEGARTY, K. A. & WEISSEL, J. K. 1988. Complexities in the development of the Caroline Hate region, western equatorial Pacific. In: NAIRN, A. E. M., STEHLI, E G. & UYEDA, S. (eds) The Ocean Basins and Margins 7B, The Pacific Ocean. Plenum Press, New York and London, 277-301. HILL, K. C. 1991. Structure of the Papuan Fold Belt, Papua New Guinea. AAPG Bulletin, 75, 857-872. & GLEADOW, A. J. W. 1989. Uplift and thermal history of the Papuan Fold Belt, Papua New Guinea: Apatite Fission Track Analysis. Australian Journal of Earth Sciences, 36, 515-539. - - , GREYA., FOSTER,D. A. 8z BARRETTR. A. 1993. An alternative model for the Oligo-Miocene evolution of northern PNG and the Sepik-Ramu Basins. In: CARMAN, G. J & CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 241-259. , MEDD, D. & DARVALL,P. 1990. Structure, stratigraphy, geochemistry and hydrocarbons in the Kagua-Kubor area, Papua New Guinea. In: CARMAN, G. J & CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG
TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA Chamber of Mines and Petroleum, Port Moresby, 351-366. HtLYARD, D. B., FORD, C. & MCDONALD S. 1994. Aitape Basin, PPL 150, Papua New Guinea: A report for LL&E Sepik Pty Ltd. Ford Geoconsultancy Pty Ltd. HOME, P. C., DALTON, D. G. AND BRANNAN, J. 1990. Geological evolution of Western Papuan Basin. In: CARMAN, G. J & CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 107-118. HURFORD, A. J. & GREEN, P. E 1982. A users' guide to fission track dating calibration. Earth and Planetary Science Letters, 59, 343-354. HtrrCHISON, D. S. & NORVICKM. S. 1980. Geology of the North Sepik region, Papua New Guinea. BMR Record 1980/24. JAQUES, A. L. 1976. High K20 island-arc volcanic rocks from the Adelbert Ranges and Finisterre Ranges, northern Papua New Guinea. GEOLOGICALSOCIETY OF AMERICABULLETrN,87, 861--867. KUGLER, A. JR. 1990. Geology and petroleum plays of the Aitape Basin, New Guinea. In: CARMAN, G. J & CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 479-490. McDOUGALL,I. & HARRISON,T. M. 1988. Geochronology and thermochronology by the 4°Ar/39Ar method. Oxford University Press. New York. PAGE, R. W. 1976. Geochronology of igneous and metamorphic rocks in the New Guinea Highlands. Australian Bureau of Mineral Resources, Bulletin, 162.
537
PEARCE, J. A., HARRIS, N. B. W. & TINDLE, A. G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25,956-983. PIGRAM C. J. & DAVIES H. L. 1987. Terranes and the accretion history of the New Guinea orogen. BMR Journal of Australian Geology and Geophysics, 10, 193-211. & Symonds, P .A. 1991. A review of the timing of the major tectonic events in the New Guinea Orogen. Journal of Southeast Asian Sciences, 6, 307-318. ROGERSON, R. J., HILYARD, D. B., FINLAYSON, E. J., HOLLAND, D. J., NION, S. Z. S. ET AL. 1987. The geology and mineral resources of the Sepik Headwaters region, Papua New Guinea. Geological Survey of Papua New Guinea, Memoir, 12. STRUCKMEYER,n. I. M., YEUNG, M. & PIGRAM, C. J. 1993. Mesozoic to Cainozoic plate tectonic and palaeogeographic evolution of the New Guinea Region. In: CARMAN, G. J • CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 261-290. WILSON, A. C., BARRETT, R., HOWE, R. & LEU, L. K. 1993. Occurrence and character of outcropping limestones in the Sepik Basin: implications for hydrocarbon exploration. In: CARMAN, G. J & CARMAN, Z. (eds) Petroleum Exploration in PNG: Proceedings of the First PNG Petroleum Convention, Port Moresby. PNG Chamber of Mines and Petroleum, Port Moresby, 111-124. -
-
Gondwana origin of the Baoshan and Tengchong terranes of west Yunnan HELMUT WOPFNER
Geologisches Institut, Universitdt zu Kgln Ziilpicherstrasse 49, D-50674 K61n, Germany
Abstract: Upper Palaeozoic glaciomarine deposits combined with the presence of cold-water faunas and Glossopterisindicate a Gondwana provenance of the Baoshan and Tengchong Blocks in western Yunnan. The glacial origin of these Permo-Carboniferous deposits is verified by a three-fold subdivision of the basal sequence comprising, in ascending order: diamictites with faceted and striated clasts, laminites with dropstones, and a black anaerobic de-glaciation facies. The succeeding carbonates of the upper Permian and Lower Triassic originated during a postglacial phase of a much warmer climate. Although the two blocks are adjacent to each other their Upper Palaeozoic sequences show significant differences. The glacigene successions of the Baoshan Block are comparatively thin and are overlain by thick basalts and red beds. On the Tengchong Block the glacigene marine deposits exceed 1000 m and are followed by thick reefal limestones of Lower Permian age. An active volcanic rift setting is suggested for the Baoshan Block and a proximal passive margin environment at a somewhat lower latitude for the Tengchong Block. The succession of the latter is comparable with Permo-Carboniferous sections in NW Australia, especially that of the Bonaparte Gulf basin, and it is assigned to the Sibumasu tectono-stratigraphic unit. The Baoshan sequence shows similarities to sequences of Tibet and NE India and is therefore assigned to the Tibetan realm. Both terranes separated from Gondwana in the late Early Permian. Docking commenced in the Late Triassic concomitant with the closure of the Changning-Menglian Belt. Lateral displacements in the course of the Himalayan orogeny moved the Tengchong Block north, bringing it into juxtaposition with the western margin of the Baoshan Block. This tectonic contact is now the Nujiang Line.
In southwest China the east-west tectonic trends of the Himalaya turn sharply into a north-south direction. This direction is maintained essentially throughout Burma, Laos and Thailand, up to the southern tip of the Malay Peninsula. In western Yunnan the latitude-parallel structural elements control the paths of the three great rivers, the Yangtze, Mekong (Lancangjiang) and Salween (Nujiang). South of latitude 26°N some southeasttrending tectonic elements branch off the north-south alignment. This is reflected by the course of the Red River (Yuanjiang) and the middle and lower reaches of the Mekong. Both north-south and southeast-striking tectonic fabrics resulted from the collision of the Indian plate with Laurasia in Tertiary times, thereby masking a number of structural features of earlier tectonic events, especially the original boundary between detached G o n d w a n a fragments and Laurasia (Cathaysia). Within the area of the Three River region of SW China and the adjoining Golden Triangle in Burma and Laos, a number of sutures have been suggested which supposedly form the boundary between Gondwana and Laurasia. These concepts, which have been summarized by Jin Xiaochi (1994), were developed either from palaeontological
and/or palaeobotanical evidence or from field observations, as for instance the presence of rocks like serpentinites or ophiolites, or diagnostic facies associations such as banded cherts. An important, but hitherto not unequivocally accepted piece of evidence to determine the position of the boundary is the presence of Upper Palaeozoic deposits of presumed glaciomarine origin, generally referred to as pebbly mudstones, as found in two structural units in western Yunnan. Recent investigations in that region carried out by the author and by Jin Xiaochi (1994) have substantially enhanced the probability that these pebbly mudstones are indeed of glacial origin and hence originated from part of Gondwana (Wopfner & Jin 1993; Jin Xiaochi 1994).
Regional overview Southwestern Yunnan consists of six major tectonic provinces which are shown in Fig. 1. These comprise the Yangtze Platform which, in middle to late Palaeozoic times was clearly part of Cathaysia, although its Upper Proterozoic and Cambrian successions show certain affinities to Gondwana. To the west and southwest of the Yangtze Platform are the L a n p i n g - S i m a o Fold System and the
From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, Geological Society Special Publication No. 106, pp. 539-547.
539
540
H. WOPFNER 26"
gtze Platform
/ S o u t h China g
Fold System
BURMA \
VIETNAM 22-
o
50
I
I
,} L A O S
100
'"
I
kilometres
~8=
loo,
/
~ , .,,I,-,.L.i.~ 'lo21
104 J
Fig. 1. Schematic map of the major tectonic provinces of southwestern Yunnan.
Changning-Menglian Belt. Both comprise folded sedimentary and some volcanic rocks ranging in age from Silurian to Lower Triassic. Within the Lanping-Simao Fold Belt these older rocks are covered by a thick succession of little-deformed red beds of Lower Jurassic to Eocene age. In contrast, the Palaeozoic succession of the ChangningMenglian Belt is fairly well exposed, including thick successions of banded cherts and basic volcanics (Liu Benpei et al. 1991). The Ailaoshan Belt which is wedged between the Yangtze Platform and the Lanping-Simao Fold System consists mainly of migmatites and other high grade metamorphic rocks and some Palaeozoic and Mesozoic cover. Mylonites attest to intense sinistral shearing in mid-Tertiary times (Tapponnier et al. 1990). The westernmost parts of Yunnan are formed by the Baoshan Block and the Tengchong Block. These two blocks host Upper Palaeozoic marine successions with diamictites and pebbly mudstones which are interpreted to be of glacigenic origin. The Baoshan Block is built up substantially by sedimentary and some volcanic rocks, ranging in age from (?) Upper Proterozoic to Cretaceous. An angular unconformity exists between Visean carbonates of the Pumenqian Formation and older rocks and the Upper Carboniferous/Permian
Dingjiazhai Formation, the basal unit of the Pangaea sequence (Fig. 2). Its base contains diamictites and other coarse-grained deposits of presumed glaciomarine origin. The Dingjiazhai Formation is succeeded by basalts, red beds and thick carbonates which extend into the Ladinian. The presence of bauxites above the Ladinian carbonates indicates at least a temporary interval of subaerial weathering. Mildly deformed middle Jurassic red beds unconformably overlie the Pangaea sequence in some places. These red beds represent the molasse stage of the Indosinian movements. The Tengchong Block adjoins the Baoshan Block in the west and is separated from it by the Nujiang tectonic line. By far the largest portion of the block is built up by the Gaoligongshan Metamorphics which also form the mountain range of the Gaoligongshan, immediately west of the valley of the Nujiang. The metamorphic rocks, which form a large synform, consist mainly of gneisses, granulites, mylonites and some marbles. They are of Middle Proterozoic age. The sedimentary cover comprises some Lower Devonian siliciclastic rocks and dolomites succeeded by the Pangaea sequence, commencing with the Upper Carboniferous-Permian Menghong Group. This sequence comprises the thick sandstone-siltstone
541
BAOSHAN AND TENGCHONG TERRANES, W Y U N N A N
BONAPARTEGULF BASIN NW AUSTRALIA
TENGCHONG
'
~
Yanzipo Limestone intrusive
3
I ~ ~ ~
2- ~
Illl:lll'I II L~I ./ // ¢ ~ ~. ~\//
i.It~.i.~~.i,~i[ ~~ 1- - - - - ~
V V v V~
v W ~ v v v v v v v?
andesite ] bauxite
2
Hewanjie Formation
TENGCHONG TERRANE
Lote Permian
i.'.'.:.'i" 'J K--~yting Formotion Artinskion
Siguaping Formation
., ,..i ii I
I~!~!~-.-~'~)!~
~ ~ l
~ ~ ~ - "
Dodongchong
~
Formotion
j
o
Luogengdi Formation
ShazipoFm. [ ~ I YongdeFm. 1 Woniusi Formation
T RI A S S I C
granite
Damuchang Formation BAOSHAN
---~
~
Tr--~ocheryShole __--__ ---Assetion/Sokm. ~~ ~ ~ ~,~J Assetion '~.'-.'.'- i
Bangdu Formation
~ ~
block lutite Kongshuhe Formotion
Stephonion
-,~Eorly NamurionOml
Zishi Formotion
Dingjiazhai
0
Fomation
PumenqianFm. ( Visean)
0
no base exposed
Fig. 2. Lithostratigraphy of the Upper Palaeozoic successions (Pangaea Sequence) of the Baoshan and Tengchong Blocks.
No bose Fig. 3. The Pangaea sequence of the northern Tengchong Block and that of the distal Bonaparte Gulf basin in northwestern Australia.
unit of the Bangdu Formation at the base, succeeded by glaciomarine deposits. Within the southern Tengchong Block the latter are termed, in ascending order, Luogengdi, Siguangping and Damuchang Formations (Fig. 2). In the northern part, near the Burmese border, the basal unit termed the Zishi Formation consists entirely of thick sandstones, and the glaciomarine succession is referred to as the Kongshuhe Formation. The succession of the northern Tengchong Block is shown in Fig. 3. The glaciomarine beds are overlain by thick coralline limestones of Permian age. The whole region has been intruded by Indosinian granites, some of which are tin-bearing. They form large stocks and make up about 15% of the outcrop area. The youngest rocks are Miocene or Holocene basaltic lava flows, pyroclastic deposits and perfectly preserved volcanic cones (Bureau of Geology Yunnan 1990).
as Gondwana-derived terranes. The question is, just how strong is the evidence for a glacial origin of these rocks and can it withstand scrutiny? In a marine environment it is much more difficult to prove sediment transport by glaciers than on land, where lodgement tills, and glaciated pavements with multi-directional striations and pressure gouges can provide unequivocal proof for the action of glaciers. The glacial load shed into a marine realm is much more subject to reworking by currents, slurry transport and other massmovements which will obscure the original glacial signature. Thus single indicators on their own may not be conclusive. Only a combination of features viewed in context of their interrelationship will be able to provide convincing evidence. A glacigene sequence is characterized by a general three-fold subdivision reflecting, in ascending order, the final glacial high-stand, the periglacial stage and finally the deglaciation stage (Wopfner & Diekmann 1992; Diekmann 1993). The rock units which correspond to these stages
Glacial deposits or mud flows?
are:
The Pangaea sequence of both the Baoshan and the Tengchong Blocks contains within its basal rock units diamictites, pebbly mudstones and other ill-sorted, rudaceous deposits which have been interpreted to be of glaciomarine origin. A glacial character of these rocks would identify both blocks
(1) diamictites and other rudaceous deposits both in autochthonous position and as slumped and slurried re-deposited masses (Dowdeswell & Scourse 1990); (2) graded laminates with dropstones and inter-calated, usually sandy, turbidites and
542
I-I. WOPVN~g
dark-coloured siltstones, containing a typical cold-water assemblage of bryozoa and mollusca (bry-mol facies), as observed not only in many Lower Permian glaciomarine successions but also in Pleistocene/Holocene arctic and subarctic environments (Henrich et al. 1992); (3) black, pyritic lutites with high organic contents, carbonate concretions and sulphates (Wopfner 1994). Well-developed diamictites are present in the lower parts of the Dingjiazhai Formation as well as in the Kongshuhe Formation and its southern equivalents. They are generally massive diamictites with angular to subangular clasts, supported in a dark grey to dark greenish-grey lutite matrix. The composition of the latter reflects the lithology of
the source area scraped by the glaciers. Despite its fine-grained nature it contains few primary clay minerals. Within the Baoshan Block, where the glaciers traversed Lower Carboniferous limestones, the matrix is calcareous, whereas in the diamictites of the Tengchong Block, which were derived mainly from metamorphic rocks, the matrix resembles a ground-up arkose (Jin 1994). The clast sizes vary from an average of about 20 cm in the lower Kongshuhe Formation to about 6-10 cm in the Dingjiazhai Formation (Fig. 4a). Generally there is a dominance of trapezoid or pentagonal-shaped, subangular to subrounded clasts, some facetted or soled due to grinding during intra- or sub-glacial transport. Finer grained diamictites with an average clast size of 1-1.5 cm and chaotic texture are interpreted as distal slump masses. They contain sporadic clasts up to 20 cm,
j.
.¢ +
;
:
>,
....
•
7 7 Fig. 4. Features indicating a glacial origin of the Dingjiazhai Formation (Baoshan Block) and Kongshuhe Formation (Tengchong Block): (a) massive diamictite from section along Burma Road, about 20 km SW of Baoshan, western Baoshan Block; (b) slumped mass of fine-grained diamictite with faceted and faintly striated rain-out clasts, Qingshuigou Section, central Baoshan Block; (c) graded laminates with cluster of drop stones, Liuku Section, northern Baoshan Block; (d) siderite concretions in black, pyritic lutites, Kongshuhe Section, northern Tengchong Block. Scales, including those on hammer, in centimetres; diameter of coin is 21 ram.
BAOSHAN AND TENGCHONG TERRANES, W YUNNAN some of them well faceted and with faint striations. They represent rain-out clasts, material which was ice-rafted into more distal basin positions and released upon melting, which accounts for the preservation of some of the glacial signatures (Fig. 4b). The heavy mineral fraction of the diamictites and associated deposits contains large amounts of garnet and other unstable minerals, making up 50-75% of the heavy mineral spectrum in some diamictites of the Kongshuhe Formation (Fig. 5). Garnet is resistant to mechanical abrasion but susceptible to chemical attack, thus its dominance within the heavy mineral spectrum is a typical feature of glacial deposits (Hamilton & Krinsley 1967; Gravenor 1979; Diekmann 1993). Other features indicative of a glacial or periglacial origin are rhythmically bedded darkgrey siltstones with lonestones, measuring up to 25 cm, and graded laminates with dropstones as depicted in Fig. 4c. These laminates consist of regular couplets and may represent seasonal deposits. Locally, the periglacial siltstones contain an abundance of bryozoa and brachiopoda, a fossil assemblage regarded as typical of a cold-water marine environment (cf. Shi & Archbold 1993). The black lutite facies forms the third member of a glacigene sequence. It reflects depositional and climatic conditions during the final deglaciation.
KSH4 D
Garnet = 75.0
[]
Tourmaline = 11.0
[]
Zircon = 5.0
[]
Opaque = 8.0
•
Others = 1.0
543
The rocks consist of black, kaolinitic and pyritic, generally massive or faintly bedded shales or siltstones. They are anaerobic sediments which characteristically show intermediate to high contents of organic carbon, generally with a pronounced participation of algal kerogen. Commonly, large carbonate concretions are present in the lower parts of this facies (Fig. 4d). In the case of Yunnan, the concretions consist of siderite (Jin 1994). Gypsum and baryte may also be present. This facies has a very wide distribution. Besides Yunnan it has been recognized in South and East Africa, Oman, northern India and northwestern Australia (Wopfner 1994). Table 1 shows a compilation of the most common features typical of glacial deposits. From the evidence presented here the author is very confident that the successions of both the Dingjiazhai Formation and of the Kongshuhe Formation and equivalents represent glacial, periglacial and postglacial deposits. During the Late Carboniferous/Early Permian, the Baoshan and Tengchong Blocks must have been in a position where they could be traversed by glaciers of the Gondwana glaciation. Thus they are terranes which were separated from Gondwana after the glaciation.
Morphotectonics Although the Baoshan and Tengchong Blocks are in direct contact with each other today, their Permo-Carboniferous successions are quite different, which is apparent from a comparison of the two stratigraphic columns of Fig. 2. The glacial beds of the Baoshan Block are relatively thin and, especially the periglacial units, are quite fossiliferous. The glacigene succession is overlain by thick basalts of the Woniusi Formation,
Table 1, Features of glacial deposits which have been observed within the basal Pangaea Sequence of Baoshan and Tengchong Blocks in western Yunnan are indicated by crosses
KSH []
Garnet = 49.0
[]
Tourmaline = 7.9
[]
Zircon = 6.3
[]
Opaque = 36.4
•
Epidote = 0.5
Fig. 5. Heavy-mineral spectrum of the basal diamictite (KSH 4) and a diamictite from the lower third (KHS) of the type section of the Kongshuhe Formation, showing very high garnet concentrations (from Jin Xiaochi 1994).
Baoshan Tengchong diamictites with source matrix glaciated pavements pentagonal clasts striated clasts faceted clasts graded laminates dropstones slurried diamictite slump masses with rain-out clasts high content of garnet bry/mol biofacies deglaciation lutites
× 0 × × x x x x x x x x
x 0 × 0 x x 0 0 x x x x
544
H. WOPFNER
followed by red beds of the Yongde Formation and Artinskian/Kungurian carbonates of the Shazipo Formation. Geochemical data indicate that the basalts of the Woniusi Formation are of intracratonic origin (Liu Benpei, pers. comm. 1992). The glacigene succession of the Tengchong Block (Fig. 3) on the other hand, may exceed 1000 m, indicating a much higher sediment accumulation rate. The black lutites of the deglaciation phase exhibit a well-developed, fossiliferous transition zone to the overlying Lower Permian coralline carbonates. Volcanic rocks are completely absent and there is no trace of red beds, suggesting a more distal environment than that of the Baoshan Block. From the evidence available so far, the author suggests that the Pangaea sequence of the Baoshan Block originated in a volcanic rift setting which, during the Late Permian and Early Triassic, was transformed by crustal foundering to a restricted carbonate platform, indicated by the Lower Triassic dolomites and bituminous limestones of the Hewanjie Formation (Fig. 2). For the succession of the Tengchong Block a passive margin setting is suggested, probably distal shelf to upper slope, which subsequently developed into a carbonate platform.
Origin of Yunnan terranes In view of the environmental interpretation of the Pangaea sequences of the Baoshan and Tengchong Blocks, their original position within Gondwana may be deduced by identifying comparable morphotectonic settings with similar lithostratigraphy at the margin of known Gondwana fragments. Permian volcanic rift structures have been identified by Gaetani & Garzanti (1991 ) in Kashmir and Zanskar, at the northwestern margin of the Indian Gondwana fragment, and a similar morphotectonic situation was also suggested by Acharyya (1987). Figure 6 shows a comparison between the succession of the Baoshan Block with that of Kashmir. There, a comparatively thin glacigene sequence of Asselian age rests unconformably on the mid-Carboniferous Fenestella Shale. The unconformity reflects disturbances of the Hercynian orogenic phase, also recognized in the Lahul-Spiti district (Kanwar & Ahluwalia 1979). The Agglomeratic Slate which conformably overlies the glacigene sequence contains an Artinskian faunal assemblage. It is followed by the Nishatbag Formation with Gangamopteris and Glossopteris. The thick basaltic to andesitic rocks of the Panjal Traps which form the major part of the Pangaea sequence of Kashmir, are of Kazanian to Lower Tatarian age (Murgabian-Midian; Stampfli
KASHMIRNW-INDIA Daonelta ShaLe
BAOSHAN TERRANE
-Khunamuh Formation _ _ ~ ZewanFro. Ar~ VvVvV v vv VvVvVv~
ShazipoFm. YonggdeFro.
Panjar Trops
500~
Sokm.~ Assel ~'~'~
Aggtomera[ Beds
Fenestetta
Shore
s°°l ilv~-~v~ Assel om
vis. ~
v v v
-:'
Woniusi Formation Dingjiazhai Formation ian Formation i!ikPumenq
Fig. 6. Comparison between the sedimentary sequence of the Baoshan Block with the Pangaea succession of Kashmir (after Gaetani & Garzanti 1991).
et al. 1991). They are thus younger than the basalts of the Woniusi Formation and apparently reflect the separation stage of the Gondwana terranes from the supercontinent. The basalts of the Woniusi Formation, on the other hand, were associated with the inception of rifting. Gondwana successions in NE India and Bhutan, as for instance in the Rangit Valley tectonic window in Sikkim (Sinha Roy 1973) show many similarities with the Baoshan succession. The three-fold subdivision of diamictites-laminitesblack lutites is quite clearly developed, and at least the upper part of the succession was deposited in a marine environment (Singh 1987). Amygdaloidal basalts of up to 3 m thickness occur together with the 'pebble-slate' (Sinha Roy 1973). The glacial sequence is overlain by sandstones with interbedded coal seams, but in comparison with the coal fields of peninsular India, seam thickness is much reduced, indicating a position closer to the shores of the Tethys. Palaeontological evidence presented by Singh (1987) and Singh & Archbold (1993) shows that the postglacial marine beds of the eastern Himalaya are of Lower Permian age and that they correlate with similar postglacial beds in Tibet (Lhasa etc.) and these in turn may be equivalents of the postglacial siltstones of the Baoshan Block. This correlation is strengthened by the fact that the glacigene sequence of the Lhasa region as well as that of the Baoshan Block are overlain by thick basalt accumulations (Jin 1994). There are also many similarities between the Lower Permian brachiopod faunas of the eastern Himalayas (Arunachal Pradesh etc.) with those of Western Australia (Singh & Archbold 1993). It is suggested,
BAOSHAN AND TENGCHONG TERRANES, W YUNNAN therefore, that the Baoshan Block occupied a position adjacent to northeastern India, which is consistent with the available evidence. The Pangaea sequence of the Tengchong Block lacks volcanic rocks and is characterized by great thicknesses of the glacigene succession. Furthermore, the latter is preceded by a thick sandstone or siltstone unit (Fig. 2). Such thick glaciomarine successions occur in distal parts of Permian basins of western and northwestern Australia, as for instance in the Carnarvon, Canning and Bonaparte Gulf basins. The Pangaea sequence of these basins is floored in places by Lower Palaeozoic rocks, and in others by Proterozoic metasediments or highergrade metamorphic rocks. Figure 3 shows a comparison between the Pangaea sequence of the northern Tengchong Block and that of the Bonaparte Gulf basin in northwestern Australia. In both cases the diamictites are underlain by Upper Carboniferous sandstones which apparently reflect a periglacial facies during the early stage of glacier advance. The glaciomarine beds are unusually thick and contain both rain-out tills and mass-transported material, and
o ~5
545
both sections are capped by an anaerobic, black lutite deglaciation facies. New palynological data obtained by Yang Wei-ping (1994) from the black lutites of the Kongshuhe section demonstrate equivalence with the Treachery Shale of the Bonaparte Gulf section. According to Foster (1986) the Treachery Shale belongs to the Granulatisporites confluens Oppel Zone (Foster & Waterhouse 1988) of Upper Asselian to Lower Sakmarian age. The black lutites of the Kongshuhe Formation thus fit in perfectly with the Early Permian deglaciation event, as suggested by Wopfner (1994). The succession of the Tengchong Block also compares well with Permo/ Carboniferous sequences in Thailand, western Malaya and Sumatra, thus connecting it with the Sibumasu Block in the sense of Metcalfe (1993). The wide distribution of tin-bearing granites of Indosinian age adds another similarity. Conclusions
The evidence presented in this paper suggests that the Baoshan Block was the easternmost part of the
J P
-..7 :.'" . "......."
\F
'i};
. "..:
. .:
:" ...
z' ...
'"~..y""
,...
? .... ....
(....; ""-. '.
~.. ..
•.' ............ ( .............. ::
'/ ,....'"'.".~¢t
•
.13:' ~. . ..... ~ ""..
),...,,
(". !
%
.:~:.:, --'.:..........~
('K- q~
Fig. 7. Reconstruction of Gondwana showing occurrences of Permo-Carboniferousglacial deposits (dashed outlines) and the limit of glaciation in Africa and Arabia after Wopfner (1991b). Latitudes are drawn relative to an Early Permian South Pole (star) similar to that given by Levellet al. (1988). The fragments of Gondwana to the north of India and northwest of Australia are schematic. The position of the glacial deposits, indicated by triangles, is well within the latitude of active glaciation.
546
H. WOPFNER
Lhasa Block. It originated in a volcanic rift setting which extended from Kashmir to a position northeast of Arunachal Pradesh. The Tengchong Block was part of the Sibumasu Block and its provenance was a non-volcanic rift off the northwest coast of Australia. The two rift systems connected at a triple junction with the Westralian trough (WA trough in Fig. 7). This triple junction was situated northwest of the present northwest coast of the Australian continent (Veevers 1988). The inception of rift movements along the western and northwestern coast of Australia during the Late Palaeozoic glaciation is well documented by pronounced differences in thickness of the glacial deposits across active faults (Veevers 1988; Wopfner 1991a) and comprehensive geophysical data (O'Brien 1993). The volcanic rift along the Tethyan margin is well established in Kashmir and Zanskar (Acharyya 1987; Gaetani & Garzanti 1991) where substantial parts of the rift system remained attached to the Indian plate. Further east, most of the rift basin was detached and drifted off as the Lhasa Block and the Baoshan Block. The difference may be explained if removal was effected by detachment in the case of Kashmir/Zanskar but caused by delamination and subcrustal shear in the case of Lhasa and Baoshan Blocks (cf. O'Brien 1993). The timing of initial separation of the terranes from Gondwana may be deduced from the ages when deposition of terrigenous detritus ceased. In the case of the Tengchong Block and presumably also of the remaining Sibumasu Block this stage
was reached not later than Late Sakmarian although there are still some disparities in the dating of the limestones succeeding the glacigene sequence. In the case of the Baoshan Block terrigenous intake was interrupted after deposition of the red beds of the Yongde Formation in about Artinskian times. This corresponds closely with the models suggested by Acharyya (1987), Gaetani & Garzanti (1991) and Liu Benpei et al. (1991), whereas Liu Guanghua & Einsele (1994) suggest a somewhat younger date. Neo-Tethys was formed between the new margin of Gondwana and the departing fragments. In Late Triassic to Early Jurassic times the drifting terranes and micro-continents collided with elements of Cathaysia, thereby closing an oceanic realm generally referred to as Palaeo-Tethys (Acharyya 1987). In Yunnan, the result of this collision was the Changning-Menglian Fold Belt (Liu Benpei et al. 1991). At this time, structural elements of Yunnan were still more or less latitudeparallel, as indicated by structural analyses and palaeomagnetic data (Chen Haihong 1992) and the Baoshan and Tengchong Blocks were probably not in contact with each other. The present juxtaposition of the two blocks is due to dextral strikeslip movements along the Nujiang dislocation zone during the Himalayan orogeny in mid-Tertiary times. The author records his gratitude to Deutsche Forschungsgemeinschaft (DFG) and the National Science Foundation of China (NSFC) for financial support of this project.
References ACHARYYA, S. K. 1987. Limits of Greater India Plate
Early Permian selected samples from Kulshill Nos. 1 & 2, Leseur No. 1, Lacrosse No. 1 Pelican Island No. 1 and Berkley No. 1 of the Bonaparte Basin.
during Gondwana time. The Palaeobotanist, 36, 290-301.
Australian Geological Survey Organisation, Open File Report.
BUREAU OF GEOLOGY AND MINERAL RESOURCES OF
YUNNAN PROVINCE. 1990. Regional Geology of Yunnan Province. Geological Memoirs, Series 1, 21, Geological Publishing House, Beijing (in Chinese with English summary). CHEN HAIHONG. 1992. Cenozoic rotations of East Asia and a possible dynamic genesis. Memoirs Lithospheric Research, 1992, 52-54, Beijing. DIEKMANN, B. 1993. Pal~ioklima und glazigene KarooSedimente des sp~ten PaI~iozoikums in SWTansania. SonderverOffentlichungen Geologisches Institut UniversitiT"t KOln, 90. DOWDESWEEE, J. A. & SCOURSE, J. D. 1990. On the description and modelling of glacimarine sediments and sedimentation. In: DOWDESWELL J. A. & SCOURSE J. D. (eds) Glacimarine Environments, Processes and Sediments. Geological Society, London, Special Publications, 53, 1-13. FOSTER, C. B. 1986. A review of Earl3, Carboniferous-
--
& WATERHOUSE, J. B. 1988. The
Granulatisporites
confluens Oppel-zone and Early Permian marine
faunas from the Grant Formation on the Barbwire Terrace, Canning Basin, Western Australia. Australian Journal of Earth Sciences, 35, 135-157. GAETANI, M. & GARZANTI, E. 1991. Multicyclic history of
the northern India continental margin (northwestern Himalaya). AAPG Bulletin, 75, 1427-1446. GRAVENOR,C. P. 1979. The nature of the Late Paleozoic glaciation in Gondwana as determined from an analysis of garnets and other heavy minerals. Canadian
Journal
of
Earth
Sciences,
16,
1137-1153. HAMILTON, W. & KRINSLEY,D. 1967. Upper Paleozoic deposits of South Africa and Southern Australia. Geological Society of America Bulletin, 78, 783-800.
BAOSHAN AND TENGCHONG TERRANES, W YUNNAN HENRICH, R., HARTMANN, M., REITNER, J., SCH.~.FER, P., FREIWALD, A., ET AL. 1992. Facies belts and communities of the Arctic Vesterisbanken Seamount (Central Greenland Sea). Facies, 27, 71-104, Erlangen. JIN XIAOCHI. 1994. Sedimentary and palaeogeographic significance of Permo-Carboniferous sequences in western Yunnan, China. SonderverOffentlichungen Geologisches Institut Universitiit K61n, 99. KANWAR, S. S. & AHLUWALIA,A. D. 1979. Lithostratigraphy of Upper Palaeozoic Tethyan Sequence in Chandra Valley near Bara Lacha La, District Lahul and Spiti, Himachal Pradesh, India. In: GUPTAV. J. (ed.) Upper Palaeozoics of the Himalayas. Hindustan Publishing Corp., Delhi, 147-153. LEVELL, B. K., BRAAKMAN,J. H. & RUTTEN, K. W. 1988. Oil-bearing sediments of Gondwana glaciation in Oman. AAPG Bulletin, 72, 775-796. L ~ BENPEI, FENG, QINGLAI & FANG, NIANQIAO. 1991, Tectonic evolution of the Paleo-Tethys in Changning-Menglian Belt and adjacent regions, westem Yunnan. Journal China Univ. Geosciences, 2 (1), 18-28. L ~ GUANGHUa& E~SELE, G. 1994. Sedimentary history of the Tethyan basin in the Tibetan Himalayas. Geologische Rundschau, 83, 32-61. METCALFE, I. 1993. Southeast Asian terranes, Gondwanaland origins and evolution. In: FINI)LA¥, R. H., UNRUG, R., BANKS, R. M. & VEEVERS,J. J. (eds) Gondwana Eight. Balkema, Rotterdam, 181-200. O'BRIEN, G. W. 1993. Some ideas on the rifting history of the Timor Sea from the integration of deep crustal seismic and other data. Petroleum Exploration Society of Australia Journal, 21, 95-113. SHI, G. R. & ARCrmOLD,N. W. 1993. Permian brachiopod faunal sequence of the Shah Thai terrane, biostratigraphy, palaeobiogeographical affinities and plate tectonic implications. Abstracts of 3rd International Symposium IGCP 321: Gondwana Dispersion and Asian Accretion, Kuala Lumpur, 1993, 46. SINGH, T. 1987. Permian biogeography of the Indian supercontinent with special reference to the marine fauna. 7th Gondwana Symposium. American Geophysical Union, Geophysical Monograph, 41, 239-249. -& ARCHBOLD, N. W. 1993. Brachiopoda from the
547
Early Permian of the Eastern Himalaya. Alcheringa, 17, 55-75. SrNHA ROY, S. 1973. Gondwana pebble-slate in the Rangit Valley tectonic window, Darjeeling Himalayas and its significance. Journal of the Geological Society of India, 14, 31-39. STAMPFLI, G., MARCOUX, J. & BAUD, A. 1991. Tethyan margins in space and time. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 373-409. TAPPOr,mmR, P., LACASSrN,R., LELOUP,P. H., SCHA_RER,U., ZHONG DALAI ET AL 1990. The Ailao Shan/Red River metamorphic belt, Tertiary left-lateral shear between Indochina and South China. Nature, 343, 431-437. VEEVERS, J. J. 1988. Morphotectonics of Australia's northwestern margin - a review. In: PURCELL,V. G. & PURCELL, R. R. (eds) The Northwest Shelf Australia. Proceedings of Petroleum Exploration Society of Australia Symposium, Perth, 19-27. WOPFNER, H. 1991a. Permo-Triassic sedimentary basins in Australia and East Africa and their relationship to Gondwanic stress pattern. In: ULBRICH H. ROCHA-CAMPOS A. C. (eds) Gondwana Seven. Proceedings of the 7th International Gondwana Symposium, Sao Paolo, 1988, 133-146. 1991b. Extent and timing of the Late Palaeozoic glaciation in Africa. Sonderverb'ffentlichungen Geologisches Institut Universitiit KiJln, 82, 447-453. 1994. Early Permian deglaciation sequences reflect end of "ice house" conditions in northern Gondwana. Permophiles, 24, 22-23. & DJ.EKMANN, B. 1992. Neue Ergebnisse aus der spatpal~iozischen Abfolge in der Karoo Tansanias. Zentralblatt fiir Geologie und Paliiontologie, Teil 1, 1991, 2689-2701, Stuttgart. & JIN, XIAOCHI. 1993. Baoshan and Tengchong Blocks of Western Yunnan (China) in the Late Palaeozoic mosaic of the eastem Tethys. Abstracts of the 3rd International Symposium IGCP 321Gondwana Dispersion and Asian Accretion, Kuala Lumpur, 1993, 47-49. YANG, WEI-PING. 1994. Lower Permian palynological studies of the Tengchong Block in western Yunnan, China. Abstracts of the International Symposium on Permian Stratigraphy, Environment & Resources, Guiyang, China, 1994, 44.
-
-
-
-
-
-
-
-
Early Mesozoic orogeny in Fujian, southeast China Z U Y I Z H O U 1, Q I U Y U A N L A O 1, H U A N J I A N G C H E N l, SIJIANG D I N G 2 & ZHONGTING LIAO 1
1Department of Marine Geology & Geophysics, Tongji University, Shanghai 200092, China 2Hainan Geological Survey, 24 Nanhang Road, Haikou, Hainan 570005, China
Abstract: The Mesozoic sedimentary successions of west Fujian evolved from Lower Triassic deep marine fine turbidites to Upper Triassic and Jurassic coal-bearing molasse deposits, and represent a typical foreland basin sedimentary sequence. This foreland basin was generated by an early Mesozoic continental collision to the south, between the South China block and the South China Sea block. This collision led to the formation of Permo-Triassic S-type granites, and a first stage of thrusting in a NNW-SSE direction. Post-collisional convergence resulted in the accumulation of thick molasse deposits. Two main unconformities, within the molasse and separating the molasse from the underlying flysch, necessitate a re-evaluation of the traditional concepts of 'Indosinian' and 'Yanshanian' movements in SE China. The tectonic and stratigraphic evolution of the west Fujian foreland basin, together with the structural features in the basin, show that the Early Mesozoic orogeny in the region was a mild and continuous process.
South China is geologically a complicated region. It is traditionally considered as being composed of two main tectonic elements, the so-called 'Yangtze para-plafform' and 'the South China Caledonian orogen' (e.g. Ren et al. 1980; Guo et aI. 1983; Ai et al. 1985; Wang 1986). The deformed Upper Palaeozoic and Mesozoic cover of south China is supposedly typical of a para-platform (Ren et al. 1980). Intense Mesozoic igneous activity in SE China is related to a postulated Mesozoic subduction of a Pacific plate under east China and the adjacent regions (e.g. Guo et al. 1983). These assumptions, however, have been challenged as oversimplifications because they failed to explain some new geological data which have emerged during the past years (Hsti et al. 1987). In order to explain these new data, Hsti et al. (1987, 1990) postulated that the Palaeozoic and Mesozoic rocks of south China are components of a Mesozoic orogen called the Huanan (south China) Alps and that the deformation was caused by continental collisions. Much of the geological terminology that dominated Chinese geological literature before the 1980s was also used, naturally enough, for the geology of SE China. Thus, on one hand there are erroneous extrapolations of European features such as 'Ca/edonian' or 'Hercynian' movements and fold belts, while on the other hand, new orogenic events like 'Indosinian' and 'Yanshanian' and more were invented, based mainly on the practice of relating unconformities to orogenic movements. These terms often hid ignorance of regional tectonics and other geological puzzles. They have all caused great confusion when attempting to
distinguish one orogenic event from another in regions where there are several unconformities. The SE Chinese province of Fujian consists of at least two major tectonic entities, i.e. the east Fujian block and the west Fujian block, with boundaries which are still controversial (Fig. 1). Widely varying views exist as to the tectonic nature and evolutionary patterns of the region (Guo et al. 1983; Ai et al. 1985; Wang 1986; Ren 1986; Zhou 1989; Bian et al. 1993). New discoveries have convinced some that it is a collage of several continental blocks (Guo et al. 1984; Zhou 1989; Bian et al. 1993). Complex interactions among the blocks and present Eurasia and Pacific plates and their precursors should therefore be recorded by structural, petrological and geophysical characteristics of the region, although intense magmatic activity in the Cretaceous has rendered these characteristics obscure. In this paper, more emphasis will be directed towards the geological characteristics of the west Fujian foreland basin. From these characteristics and other available geophysical and regional geological data from this region and the northern shelf of the South China Sea, the Early Mesozoic orogeny and its regional significance will be discussed.
West Fujian foreland basin General setting It was traditionally believed that the NNE trending 'Zhen'he-Da'pu fault' (Fig. 1) is the boundary between two different geological units, i.e. the west
From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, Geological Society Special Publication No. 106, pp. 549-556.
549
550
ZUYI ZHOU ET AL.
Yangtze Block
/
Shangha~l
I
Fuzhou
/
SouthChina Fujian~ I#~ I / Block I J~ h'~iwan / Nan'ning II / N aI,~ n ' a.... o ~I''/( .J'Hong Kong Reiqiong-~ / J Strait ~ South China an Sea Block Synclines " ~ Anticlines " ~
Profiles I ~ Fig. 4 II
EastFujianTectonicZone is./.] [--// CretaceousVolcanicRocks ~ Permo-Triassic Granites Palaeozoic-Mesozoic Rocks~ .... ! Pre-CambrianRocks ~f~ Fu'an-Nan]inFault
F-N Fault
Zhen'he-Da'pu Fault Z-D Fault Fig. 1, Tectonic map of Fujian, SE China and its regional location. Dashed line marks the boundary of the west Fujian foreland basin.
Fujian block and the east Fujian block. This belief was based largely on the restriction to the west Fujian block of the present surface distribution of pre-Sinian metamorphic rocks and Upper Palaeozoic strata. However, in the past decade, similar strata have been discovered in the area to the east of the 'Zhen'he-Da'pu fault' and to the west of the 'Fu'an-Nan'jin fault' which is also NNE trending. Furthermore, the Bouguer anomaly map of the region also shows evidence of the 'Fu'an-Nan'jin fault' (Fig. 2), but not of the 'Zhen'he-Da'pu fault'. It has therefore been suggested by Chen et al. (1982) that the boundary between the west and east Fujian should be placed along the 'Fu'an-Nan'jin fault'. Neither of the two faults, however, can be traced northwards into Zhejian province to form a conspicuous long fracture in SE China using geological or geophysical evidence. In addition, due to the cover of thick Upper Mesozoic igneous rocks, the existence of 'Fu'an-Nan'jin fault' remains geologically unconvincing. Nevertheless, evidence of the large geological contrast between coastal east Fujian and west Fujian can be easily identified, with coastal east Fujian having been discussed fairly recently by many authors (Zhou & Lao 1990; Li et al. 1993; Lu et al. 1993). While many geologists in recent years (including those who identified the socalled 'Zhen'he-Da'pu fault' and its northern
equivalent in Zhejian - the 'Zhen'hai-Li'shui fault') have discarded the concepts surrounding the definition of these faults ( Bian, pers. comm. 1989; Zhou 1989), it is still quite common in the English literature to invoke this NNE trending fault in the southeast corner of south China as a plate margin (e.g. Hsti et al. 1990). The northern, eastern and southern boundaries of the west Fujian foreland basin as shown in Fig. 1 are much more clearly defined than the two NNE trending faults in Fujian and the basin continues southwestwards into northeastern Guangdong. The region that is covered by the basin is usually called the Yongmei depression and is believed to be either a 'Hercynian fold belt' (Bian & Gao 1982; Wang 1986) or a 'Hercynian-Indosinian fold belt' (Wang, E. K. et al. 1982). Zhou (1992a, 1992b) indicated that the region had evolved from a Late Palaeozoic passive continental margin to an Early Mesozoic foreland basin, which was related to an Early Mesozoic collision and the subsequent compression between the SE China and the South China Sea continental blocks. Stratigraphy Upper D e v o n i a n - U p p e r Permian. The Upper Devonian succession in west Fujian consists mainly of a series of slightly metamorphosed quartz
551
EARLY MESOZOIC OROGENY IN FUJIAN, SE CHINA I 118"E
.... . . . . . .
;> 38.8o\ •
+~o '-.•.
• "___._.L. . . . .
~
"",
~
. .-35.."
~
.
. .'"
37.60 //
/:/~
•
•
~@ ."~'
24ON •
Fig. 2. Bouguer anomaly contours (dashed lines, in mgal) after processing to remove the effect of the upper 15 km of the crust, and isopachs (solid lines, in kilometres) of the crust of Fujian and environs (original data from Chen et al. 1983).
sandstones and sandy conglomerates, which unconformably overlie pre-Devonian metamorphosed rocks. The Lower Carboniferous sandstones have been interpreted as fluvial and mudflow deposits by Li (1989). The Middle Carboniferous deposits in some parts of the basin contain small amounts of volcaniclastic rocks whose S i t 2 contents are bimodal, i.e. either <52% or >65%. The REE patterns of the basalts interbedded with these volcaniclastic rocks are similar to those formed in a continental rift setting (Wang, D. P. et al. 1982). As the thickness of the Upper Devonian to Upper Carboniferous sedimentary rocks gradually decreases northwards, the sediment grain size increases. The lower part of the Permian sequence in the basin consists of coalbeds interbedded with sandstones and sandy siltstones, while the upper part comprises siltstones, calcareous
siltstones and radiolarian-bearing sedimentary rocks.
siliceous
Triassic. The lower Triassic Xikou Formation consists of sandstones, calcipelites, mud-siltstones and siltstones. Deep-water turbidite sedimentary structures are well developed in the Xikou Formation, among which are Bouma sequences, convolute laminations, ball-and-pillow structures, scour marks and flute marks. The palaeocurrent direction determined from flute casts and climbing-ripple beddings is around 300 ° in the central area of the basin, i.e. Da'tian and Yong'an. In the same area, contourites and carbonate gravity flow deposits were identified (Yang et al. 1993). The basin was asymmetric with water depth decreasing towards the northwestern part of the basin. Syndepositional slumps and brecciation of Lower
552
ZtWI ZHOU ET AL.
strata are also well developed. The grain size frequency diagram and C - M diagram of 35 samples indicate suspension and saltation transportation (Zhou 1992a). The rate at which sediments of Xikou Formation were accumulated (0.136->0.460 m per 100 years) exceeds that of the Palaeozoic deposits (0.01-0.17 m per 100 years) in the region and is within the range of that of the flysch deposits in other foreland basins in the world (0.15-1.927 m per 100 years, Schwab 1986). The major element compositions of the Xikou Formation rocks exhibit low Fe203 + MgO (3-5%), TiO 2 (0.6-0.95%), A1203/SiO 2 (0.2-0.3%) and high K20/Na20. Discriminant function analysis (Zhou 1992a) of sandstones from the Xikou Formation using 11 major element oxides as variables and adopting the values of unstandardized discriminant coefficients for sandstone suites of eastern Australia (Bhatia 1983) suggest that the provenance of the Xikou Forrmation was an active continental margin (Zhou 1992a). Another prominent feature of the Xikou Formation is its well-developed structures. While some of these structures are of syndepositional origin, the majority were caused by the postdepositional compression evident from the closely spaced mesoscopic scale isoclinal folds, chevron folds, duplexes and associated cleavages (Zhou 1992b; Yang et al. 1993). The Middle Triassic calcareous sandstones were deposited in small-scale shallow marine depressions. They are overlain unconformably by the thick Upper Triassic conglomerates, pebble-bearing sandstones, sandy conglomerates and coal-beating sandstones. The Lower to Middle Jurassic quartz conglomerates, coarse and fine sandstones and siltstones interbedded with marine deposits are in turn overlain unconformably by thick tuffaceous sandy conglomerates, sandstones and siltstones of the Upper Jurassic. The accumulation rates of these sediments range between 0.078 m and 0.490 m per 100 years, close to those of the molasse deposits in other foreland basins (0.10-0.40m per 100 years, Schwab 1986). In summary, the Mesozoic sedimentary successions of west Fujian evolved from Early Triassic flysch deposits to coal-bearing Upper Triassic to Upper Jurassic molasse deposits, forming a typical foreland basin sedimentary sequence (Fig. 3). This foreland basin trended ENE and increased in size during its later evolution. In the Early Triassic, the foredeep was in Da'tian, central Fujian, where deep-water turbidites were deposited. Subsequently, as the size of the basin increased, the water depth decreased. This process culminated in Late Triassic and Late Jurassic time when the
Age & Rock Units
Tectonstratigraphy
Lithology
156 Ma . V . . V . . V , , -..V,
. . V
ChanglinFm.:ill J3c
.
.
.
.
.
.
iiiiiii:
....?.....?....?. 163 Ma =~"~°~~ Zhangping ~
:
Fm
dZz 187Ma
Lishan Fm J1
:-.~.--~.:-~. Molasse ".~::~-:.~: ~:.~: ~.:' Deposits .'~.'.' .~.': .~. in .............. Foreland Basin . . . . . . . . . . . . . . ..............
.............. . . . . . . . . . . . . . .
208 Ma
..............
Wenbingshan . . . . . . . . . . . . . . Fm T3w
230 Ma
.............. ~ . . c,>.. ~ . . ~ : ~ . ~ : ~ . ~ : .~
~ m n Fm
T2n 240 Ma
" - : , . = . .x .~x :' . . . . . .
Xikou Fm Tlx 245 Ma p
Middle Triassic-Upper Jurassic.
250 Ma
=
.......
360 Ma
'
[11
Foreland
..... • ..... ...r, . . . . . . ~ .... ' ' '
Basin
•
..............
Shelf
..............
Deposits
•
-
;
j
~
r - - '...T - - - ' .
C
Flysch
Deposits
, "--:.-
,
I
- : , . - • . = • .' ., ? , - 1 -
I
..~'
1
" ~ ' -
l
C~--
-
Tianwadong . . . . . . . . . . . . . . Fm >.-o.-c.. D3t ..............
Fluvial Deposits
' ,
~..C3,..6>.. •,,
376 Ma
,~.,.
--,..
~..
~,..
~,m
i
r-,- -
Tuffaceous
Sandstone
Coal-bearing
sandstone
Conglomerate Fine
Sandstone Calcpelite
~:~
Mudstone Sandstone Mud-Siltstone
Siliciclastic C a r b o n a t e Carbonate Sandy
[ ~
Conglomerate
3. Generalized Upper Palaeozoic-Mesozoic stratigraphic column of west Fujian.
Fig.
EARLY MESOZOIC OROGENY IN FUJIAN, SE CHINA extremely thick (maximum thickness of 3400 m in Da'tian) molasse sequence was rapidly deposited and the width of the basin reached a maximum.
553
Permian
SouthChina SeaBlock
SouthChinaBlock
Foreland fold-and-thrust belt Widely developed folds and low-angle thrusts in west Fujian have been recognized by regional geological surveys and drill cores that penetrated into both the sedimentary cover and the basement. The foreland fold and thrust belt of west Fujian is characterized by multistage and multilevel thrusting of the cover and basement decoupling. The most prominent folds were developed in the Lower Permian coal-bearing sediments, and the folding in these rocks is intense and complex. This coal-bearing Lower Permian sequence also acted as the detachment zone which enabled the development of imbricate thrusts (Fig. 4) and folds in the overlying sedimentary cover. Other detachment zones in the basin were in the Lower and Upper Carboniferous strata. Low angle thrusts in the basin usually occur in groups withvery similar dips (15-20 °) and strikes (ENE-WSW). These occur not only as cover thrusts related to different layers of detachment, but also as steeper thrusts in the basement of the basin. Kinematic studies of the thrusts reveal two main stages of thrusting. The earlier thrusting episode has been related to Early Mesozoic north-south convergence between two continental blocks (Fig. 5) and was overprinted by Early Cretaceous thrusting from west to east (e.g. Tao 1987). The northwest limbs of the anticlines are shorter and steeper, and at some localities these anticlines are even overturned. Microfabric studies show that the direction of the earlier stage of thrusting was from
I
II
-~SE
Early Triassic
Late Jurassic
~
Continental Crust
1
Oceanic Crust
Granites ~
Attenuated Crust
Fig. 5. Schematic cross-section to show the evolution of west Fujian foreland basin. Symbols are the same as those in Fig. 3.
SSE to NNW, with the root of the thrust belt, therefore, in the south of the basin.
Permian-Triassic granites Permian to Triassic S-type granites in the west Fujian foreland basin generally cut across the fractures and the anticlinal axes. They are alkaline granites with high 87sr]S6Sr ratios and low Fe3+/ (FEZ++ Fe 3+) ratios. In addition, they are also LREE enriched (Wang & Liu 1986). A discriminant diagram of Rb against Y + Nb (Fig. 6) suggests a syncollisional and volcanic arc origin. Most of these granites postdate the closure of the previous ocean basin and may be related to the partial melting of the partly subducted oceanic crust or to voluminous sediments accumulated between the two continents before the final closure of the intervening ocean basin.
Early Mesozoic continental collision in SE China Fig. 4. Structural profiles to show the thrusting activity in west Fujian. See Fig. 1 (I and II ) for the locations of the profiles.
Studies of the geology of SE Asia and of south China in the last decade or so have culminated in the recognition of several continental blocks
554
zuYi ZHOU E T AL.
1000 100
Rb ppm S y n - C O L ? /
~
',~
WPG
10 Y+Nb ppm
I'O
16o
ic;oo
Fig. 6. Discriminant diagram of Rb against Y+Nb for Permo-Triassic granites from west Fujian. The determination of tectonic settings is based on Pearce et al. (1984). ORG = ocean ridge granites; Syn-COLG = syn-collisional granites; VAG = volcanic arc granites; WPG = within plate granites. The original data are from Fujian Bureau of Geology and Mineral Resources (1985). that were assembled mainly during the Late Palaeozoic-Early Mesozoic (e.g. Seng0r 1990). Most of the blocks originated from Gondwana. The Middle-Late Triassic collision between the South China block and the Indochina block along the Black River (or Song Da) suture and the Red River (or Song Ma) suture is well documented, which is in marked contrast to the little attention that has been paid to the possible suture zones further east in the northern shelf of the South China Sea and the south China margin. According to Zhu (1987), there existed a pre-Cretaceous active margin along the northern shelf of the South China Sea, which passes through the southern part of Hainan Island and may be connected to the Red River suture. The similarity in Cambrian sedimentary successions, ore deposits, trilobites and brachiopods between Hainan Island and Australia, together with the discovery of Upper Palaeozoic glacio-marine
~,/
,-/
~
) ~"
t
--
/
.
h II
deposits of similar origin to those in Tibet, west Yunnan and SE Asia (Yu 1989) suggest that Hainan Island is a fragment that rifted from Gondwana at or after the end of the Palaeozoic. The location of the suture zone in Hainan Island and its vicinity remains a problem to be solved. Wang (1986) proposed that the suture zone extends along the Reiqiong Strait that separates Hainan Island from mainland China. Others (Zhen 1989) have argued that it is located in northern Hainan (Fig. 7). Further east, Hsti et al. (1990) postulated the existence of a Mesozoic (Triassic or Jurassic) collision between the Huanan and the Donnanya Blocks. However, it is worth mentioning here that the suture zones to the south and to the east of Fujian in SE China are of different origins and ages (e.g., Zhou 1992b). The deep-water flysch deposits of the Lower Triassic Xikou Formation, the intense deformation of these deposits, the formation of S-type granites and the first episode of thrusting in west Fujian fold-and-thrust belt have all been interpreted above to be genetically related to a continent-continent collision which took place to the south of the west Fujian foreland basin in the Early Triassic. This collision marked the beginning of the Early Mesozoic orogeny in the region. However, deformation did not stop as collision ceased. Continuing post-collisional convergence made the orogeny a long process and led to the uplift of colliding continental blocks, resulting in the deposition of thick molasse in the adjacent west Fujian foreland basin (Figs 3 & 5). The difficulty in identifying the suture zone has been due mainly to the massive magmatic and deformation activity related to the subduction process to the east of Fujian in the Taiwan Strait in the Cretaceous (Zhou e t al. 1992b), and to submergence of the area beneath the South China Sea. However, geophysical data from the region record some evidence of the existence of an ENE trending tectonic zone. The Bouguer anomaly trend in east Fujian swings from NE to ENE at
//Pearl River /" Basin /' _--
. . . . . .
........
//
,oo
Fig. 7. Postulation of the positions of Early Mesozoic suture zones on south China margin. Lines 1, 2, 3 are suture zones proposed by HsiJ et al. (1990), Zhen (1989) and Wang (1986), respectively.The area marked 4 is the possible location of the Early Triassic suture between the South China and South China Sea blocks suggested in this paper.
EARLY MESOZOIC OROGENY IN FUJIAN, SE CHINA the southern border of the west Fujian foreland basin. Folding and thrusting have resulted in the thickening of the crust in the basin (Fig. 2). Along the coast from Nan'ao to Hong Kong is an ENE tending tectonic zone characterized by a relatively low aeromagnetic anomaly and gravity gradient, reflecting proposed ultrabasic igneous rocks at a depth of 8-12 km (unpublished report of No. 909 Aeromagnetic Survey Brigade, Ministry of Geology and Mineral Resources of China). In the Pearl River Mouth Basin, three ENE trending aeromagnetic lows are also attributed to deep basic intrusive rocks (Guong et al. 1989). In order to unravel the basement structures beneath the Tertiary basins of the northern South China Sea shelf, an integrated geological-geophysical approach has been used to interpret several N-S trending profiles for which there are gravitational, magnetic, seismic and borehole data supplemented by onshore data. Forward and inversion iterative modelling was applied to the profiles. These data enabled the postulation of Palaeozoic metamorphic rocks and Upper Mesozoic sedimentary and metamorphic rocks in the basement. Fault mapping revealed an ENE fault system cross-cut by younger NW trending faults. All these phenomena are consistent with a collisional event. A schematic evolution for west Fujian is proposed in Fig. 5, though at this juncture we can only conjecture as to the position of the suture zone (Fig. 7).
Discussion According to Seng/3r (1985), the Cimmerian Continent rifted from the northern margin of Gondwana during the latest Palaeozoic to the earliest Mesozoic. This continent disintegrated as it moved through the Tethyan domain. Palaeomagnetic and palaeontological data suggest a possible Gondwana origin for the Yangtze, south China and Hainan Island blocks. The evidence of Early Mesozoic continental collision in SE China suggests that, before the Cimmerian Continent united with Eurasia during the Late TriassicMiddle Jurassic (Sengtir 1985), some of the continental blocks (e.g. South China block and South China Sea block in this paper) had already collided with each other during their northward drifting. The collisions of continental blocks that constitute the Cimmerides therefore took place diachronously.
555
The so-called 'Indosinian' movement in Indochina and south China is the regional geological phenomenon that resulted from this diachronous collision process. As mentioned above, the Early Mesozoic orogeny in SE China was not a short event, because N-S post-collisional convergence continued until the end of the Jurassic. The two regional unconformities within the syn-orogenic molasse and separating the molasse and the underlying flysch (Fig. 3), reflect two main stages of tectonic relaxation (or tectonic quiescent periods of Blair & Bilodeau 1988) in the collision zone. The coarse sediment above the unconformities actually represents flexural rebound of the thrust belt as the uplifted area was eroded, and the finer sediment above the two coarse units might represent renewed uplifts that were related to the continuing N-S compression. The traditional practice of relating these two unconformities to, respectively, 'Indosinian' and 'Yanshanian' orogenies should therefore be reassessed. The sedimentary sequence and evolution of the west Fujian foreland basin are comparable with those of typical foreland basins (e.g. the Appalachian foreland basin, Tankard 1986). Thrusting and drcollement took place along ramps at different levels, and the detached pieces could not displace freely on a perfect 'sole' of the southern Appalachian type (Zhu 1989). Finally, the Lower Triassic deep marine fine turbidites in the foredeep of the west Fujian foreland basin conformably overlies the Upper Permian deposits (Fig. 3), which shows that the transition from passive continental margin to active continental margin was gradual. The fact that neither mud nor shale was deposited over the fine turbidites also reveals that the compression was a slow and continuous process. In addition, the lack of evidence of metamorphism shows that the continental collision was a mild collision. This paper was prepared at the University of Wales, Aberystwyth in early 1994, while the senior author was a recipient of a Royal Society Visiting Fellowship. R. J. Whittington, W. R. Fitches, T. Horscroft, J. Charvet, M. Allen and B. C. Burchfiel are thanked for their constructive comments and contributions to the manuscript. Robert Hall is thanked for improving the figures and manuscript. This research was supported by the National Science Foundation of China (project NSFC49202034).
References At, C. X., CHEN, B. W. & HUANG,H. 1985. Preliminary study of iron ore-control tectonics of south Fujian and east Guangdong. Bulletin of the Institute of Geology of Chinese Academy of Geological Sciences, 13, 53-66 (in Chinese) BHATIA, M. R. 1983. Plate tectonics and geochemical
composition of sandstones. Journal of Geology, 91, 611-627. BIAN,X. Z. & GAO, T. J. 1982. Discussion on the types of iron ore deposits and their exploration in 'Yongmei depression'. Geology of Fujian, 1 (2), 53-67 (in Chinese).
556 --,
zuYI ZHOU ET AL.
Cue, Z. X. & ZHOU, W. D. 1993. Framework of Palaeozoic-Mesozoic tectonic evolution of Fujian Province. Geology of Fujian, 12 (4), 280-291 (in Chinese). BLAIR, T. C. & BILODEAU, W. L. 1988. Development of tectonic cyclothems in rift, pull-apart, and foreland basins: sedimentary response to episodic tectonism. Geology, 16, 517-520. CHEN, Y. A. & WANG, P. Z. 1982. A study on the regional geophysical and geochemical characteristics of Fujian. Geology of Fujian, 1 (2): 69-82 (in Chinese). FUJIAN BUREAU OF GEOLOGY AND MINERAL RESOURCES. 1985. Memoirs of the Geology of Fujian. Geology Press, Beijing, (in Chinese). Guo, L. Z., SHI, Y. S. & MA, R. S. 1983. Formations and evolution of Meso-Cenozoic active continental margin and island arcs of west Pacific. Acta Geologica, 57 (1): 11-18 (in Chinese). --, YE, S. F. & Lu, H. E 1984. Tectonostratigraphic terranes of Southeast China. Journal of Nanjing University (Natural Sciences Edition), 20 (4), 732-739 (in Chinese). GUONG, Z. S., JIN, Q. J., WAND, S. S. & MENG, J. M. 1989. Geology, tectonics and evolution of the Pearl River Mouth Basin. In: ZHU X. (ed.) Chinese Sedimentary Basins. Elsevier, Amsterdam, 181-196. Hst3, K. J., SUN, S. & LI, J. L. 1987. Huanan Alps, not South China Platform. Scientia Sinica, B series, 1107-1115. , LI, J. L., CHEN, H. H., WANG, Q. C., SUN, S. & SENGOR,A. i . C. 1990. Tectonics of South China: Key to understanding the West Pacific Geology. Tectonophysics, 183, 9-39. LI, J. L., HE, H. Q., YANG, M. E, WANG,Z. M. & L1N, D. Y. 1993. Structural features of the Mesozoic ophiolitic melange in coastal Fujian. In: LI, J. L. (ed.) Lithosphere structure and geological evolution of SE China. Metallurgy Industry Press, Beijing, 199-206 (in Chinese). LI, Y. W. 1989. Sedimentary and structural analysis of Lower Palaeozoic Yongmei depression. M Sc thesis, Tongji University (in Chinese). Lu, H. F., J1A, D., Guo, L. Z., SHI, Y. S., ZHANG, Q. L. & WANG, Z. H. 1993. The formation and colliding history of Fujian-Taiwan microcontinent. In: LI, J. L. (ed.) Lithosphere structure and geological evolution of SE China. Metallurgy Industry Press, Beijing, 12-26 (in Chinese). PEARCE, J. A., HARRIS, N. B. W. & TINDLE, A. G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25, 956-983. REN, J. S. 1986. Some problems about the geotectonics of South China. Scientia Bulletin, 49-51 (in Chinese). , JIANG, C. E, ZHANG, Z. K. & Q1N, D. Y. 1980. Geotectonics of China and its evolution. Science Press, Beijing (in Chinese). SCHWAB,E L. 1986. Sedimentary 'signatures' of foreland basin assemblages: real or counterfeit? In: ALLEN,P. A. & HOMEWOOD, P. (eds) Foreland Basins. Blackwell Scientific Publications, 359-410. SENGOR, A. i . C. 1985. The story of Tethys: how many wives did Okeanos have? Episodes, 8 (1): 3-12. 1990. Plate tectonics and orogenic research after 25 -
-
years: a Tethyan perspective. Earth Science Review, 27, 1-201. TANKARD, A. J. 1986. On the depositional response to thrusting and lithospheric flexure: examples from the Appalachian and Rocky Mountain basins. In: ALLEN, P.A. & HOMEWOOD, P. (eds) Foreland Basins. Blackwell Scientific Publications, 369-392. TAO, J. H. 1987. Characteristics and mechanics of thrust tectonics in southwest Fujian. Geology of Fujian, 7 (3), 196-204 (in Chinese). WANG, D. P., LIu, Z. J. • WANG, D. Q. 1982. Palaeogeography and structures of Makeng-type iron ore deposits, Fujian. Journal of Changchun College of Geology, 3, 43-58 (in Chinese). WANG, D. Z. & LIU, C. S. 1986. Distribution regularities and genetic series of granites of HercynianIndosinian cycle in Southeast China. Acta Petrologica Sinica, 2 (4), 1-13 (in Chinese). WANG, E. K., ZHANG, J. Z. t~ OUYANG, Z. 1982. The discovery of fusulina in metamorphic rocks in Datian, Fujian and its significance. Journal of Nanjing University (Natural Science Edition), 18 (2), 25-32 (in Chinese). WANG, H. Z. 1986. Geotectonic development of China. In: ZUNYI YANG, YUQI CHEN & HONGZHENWANG (eds) The Geology of China. Clarendon Press, Oxford, 235-276. YANG, M. E, LI, J. L., HAO, J., Hou, Q. L. & HE, H. Q. 1993. The discovery of Lower Triassic contourites and carbonate gravity flow deposits in southwest Fujian. In: El, J. L. (ed.) Lithosphere structure and geological evolution of SE China. Metallurgy Industry Press, Beijing, 213-218 (in Chinese). Yu, Z. Y. 1989. The determination of the early Permian glaciomarine deposit in Hainan Island and its tectonic significance. Journal of Nanjing University (Natural Science Edition), 25 (1), 108-119 (in Chinese). ZHEN, J. Z. 1989. Tectonics of west Hainan Island. MSc thesis, Tongji University (in Chinese). ZHOU, Z. Y. 1989. The study on the basement of SE China: a review. Geology of Fujian, 8 (1), 46-53 (in Chinese). 1992a. The depositional environment and tectonic setting of Xikou Formation in western Fujian. Experimental Petroleum Geology, 14 (2), 135-142 (in Chinese). -1992b. Eastern Fujian tectonic zone. In: Lru, G. D. (ed.) Geological-Geophysical Features of China Seas and Adjacent Regions. Science Press, Beijing, 320-327 (in Chinese). & LAO, Q. Y. 1990. Mesozoic evolution of eastern Fujian and its adjacent areas. In: WILEY, T. J. & HOWELL, D. J. (eds) Terrane Analysis of China and the Pacific Rim. Circum-Pacific Council for Energy and Mineral Resources, Earth Science Series, 13, 185-187. Znu, X. 1987. On the evolution of Chinese continental margins and basins. Marine Geology and Quaternary Geology, 7 (3), ll5-119 (in Chinese). 1989. Remarks on Chinese Meso-Cenozoic sedimentary basins. In: ZHU X. (ed.) Chinese Sedimentary basins. Elsevier, Amsterdam, 1-5. -
-
-
-
Index
References in italics are to Tables or Figures accretionary prisms and complexes Banda arc 53, 57 Borneo 252, 253 Songpan Ganzi 99 Sumatra 23-4 Timor trough 81-2 Visayas 515-16 Aceh sliver microplate 22-3 Adang fault 382-3 Aibi-Xingxing suture 99 Aifam Group 468 Ailaoshan belt 540 Ailaoshan suture 99, 106 Aileu Formation 57 Air Bangis granite 331,333 Aitape Basin 527 Ala Shan terrane 99 Altyn Tagh fault 124, 125 Amasing Formation 489 Ambon 445 crustal isotopic signature 447-8 amphibolite, Darvel Bay 265 Andaman Sea 160 Anggi granite 468 apatite fission track analysis Papua New Guinea 528-31 Thailand 244-5 western Sulawesi 404-6, 424 4°mr/39Arages, Kaibobo complex 457-9 arc-continent collision 525 Arip Volcanics 253 Aru trough 86-7 deformation front 92-4 gravity data 89-92 seismic reflection data 88-9 Asem-Asem Basin 252 Ashmore platform 62, 63 asphalt 431 ATLAS model 154 Aurora volcanic ridge 518 Australian plate collision with Philippine Sea plate 137 continental shelf characters crustal thickness 63--6 gravity survey 66-8 marginal slope features 69-81 reflection profile 49-51 structure 63 isotope characteristics 446 plate motion 32 Ayu trough 168
Bacan 445 crustal isotopic signature 448-9 K/Ar dating 502, 503, 505 plate setting 483 Sibela continental suite 483-4 Sibela ophiolite 486-7 tectonic evolution 490-3 Tertiary stratigraphy 487-90, 501 Bacan Formation 487, 491 Bakit Mersing line 248, 252 Balangbaru Formation 359-61,366, 367 Ban Ang Formation 236 Ban Thalat Formation 236 Banda arc 11, 47, 64, 451 backarc region 56, 68-70 forearc structure 68 indentor model 70-2 orogen evolution 57-8 reflection profiles methods 48-9 results 49-56 Banda arc (east) see Aru trough Banda orogen 194-5, 197, 451 Banda Ridges 445 crustal isotopic signature 449 Banda Sea Miocene history 462 Miocene setting 144-5 tectonic blocks 140, 141 tectonic setting 175-7 Bangdu Formation 541 Banggai granite 474 Banggai Island stratigraphy 474 Banggai-Sula continental fragment 160, 466 Banggai-Sula Spur 445 Banggong suture 99, 107 Bantimala melange complex 354-5 lithology 355 outcrop distribution 356-61 structure 355-6 tectonic evolution 361-2 Banyak thrust 25 Baoshan block 101,540 diamictite 541-4 terrane motion 544-6 Barail Formation 134 Barisan Mts 321-2 Barisan orogen 189-91,197 Barito Basin 252 Batu fault 21 Bawang Dacite 253 557
558 Belaga Formation 248, 250, 251,253 Bentong-Raub line 204 Bia Formation 475 Biak Formation 475 biostratigraphy, Pak Lay sediments 229-30 Bird's Head 139, 142-3, 146 Pb isotope ratios 446 stratigraphy 470, 471 tectonic reconstruction 477 tectonic setting 165, 174 Bisa, K/Ar dating 503 Bislig Bay 517 Black River suture see Song Da Bliri volcanics 527 Bobang Formation 474 Bogal Limestone Formation 468 Bonaparte basin 62, 63 Borneo (Kalimantan) 252 Schwaner Mts 252-3 tectonic setting 157-9, 168-70 see also under Kalimantan Bouguer gravity measurements Aru trough 89-92 Australian continental shelf 66 Boyan Melange 252 Boyan suture 99, 110-11 Browse basin 62, 63 Buan Formation 251 Bunga basalt 253 Bunta Formation 474 Burma see Myanmar Buru, tectonic setting 168 Buton 160, 167, 431-3,445 displaced terrane 466 palaeomagnetism study methods 434-5 results 435-4 1 results discussed 441-2 Buya Formation 474
Cagayan ridge 164 Camba Formation 354-5, 367 Camba-Enrikang-Mamasa Complex 397 Cambro-Ordovician terrane distribution 111 Canning basin 62, 63 Carboniferous studies Fujian province 551 rifting 104 Salawati stratigraphy 468 terrane distribution 112 Tomori stratigraphy 474 Carcar Formation 520 Caroline plate 5-6, 141, 145-7, 466 tectonic setting 168, 177 Cascades fault zone 519 Cathaysia block 100 Cathaysialand 113
INDEX Celebes Sea 140, 147, 164, 171 Cenozoic tectonic setting for SE Asia 124-5 Champa Formation 236 Changning-Menglian Belt 540, 546 Changning-Menglian suture 99, 107-8 China, South 539, 549 collision tectonics 525 Asia-India 129, 133-5 Australia-Philippine Sea plate 137 Fujian evidence 553-5 Indochina 225, 230 Philippines 512-16 Sulawesi 171-4 continental fragments 445 Cotabato fault 518 Cotabato trench 39-40 Cretaceous studies Khorat Plateau 237-8 palaeomagnetism 207 Salawati stratigraphy 468 Sumatra plutonism 324, 325, 330 terrane distribution 114 Tomori stratigraphy 475 Crocker Formation 256-8 crustal provenance analysis Pb-Nd isotope analysis methods 446 results 447-50 results discussed 450-1
Daguma fault 518-19 DAMAR line 48, 49-56, 64, 76, 78-82 Damuchang Formation 541 Dangerous Grounds Block 308 Daram Sandstone Formation 470 Darvel Bay Complex 263-4 metamorphic geology 264-73 origins 273-7 ophiolite emplacement 277 declination 435,436, 437, 439 Devonian studies Fujian province 550-1 rifting 104 Salawati stratigraphy 468 terrane distribution 111 diamict, Yunnan 541-3 Dingjiazhai Formation 540, 541,542 displaced terranes 466 Dongnanya block 100 Dungun Graben 295-6
East Java Sea 384 East Malaya terrane 99, 102 East Natuna Basin 299 East New Guinea composite terrane 140, 141 East Papua composite terrane 140
~NDEX East Talaud bank 40-2 Elat Formation 87 Embaluh Group 251,253, 254 Eocene studies Asia-India collision 129 plate settings 380-2 Sulawesi tectonics 383
Facet Limestone Group 468 Fafanlap Formation 470 Faumai Formation 470 Finnisterre terrane 140, 141, 145, 148 Flores thrust 11 forearc environments, examples of see Sumatra margin 'Fu'an-Nan'jin fault 550 Fujian province 549 foreland basin 549-50 collision tectonics 553-5 fold and thrust belt 553 granites 553 stratigraphy 550-3
Gabaldon basin 519 Gamta Limestone Formation 468 Gaoligongshan metamorphics 540 Garba pluton 331 garnet amphibolite, Darvel Bay 272-3 geochemistry Darvel Bay 267-8, 269 Papua New Guinea 532-3 Sumatra granitoids 325-8 western Sulawesi 397-400, 425 geochronology s e e 4°mr/39Ar;K]mr; Rb/Sr; Sm/Nd; U/Pb glacigenic sediments, Yunnan 541-3 Gondwana terrane rifting 103-5, 539, 544-5 gravity see Bouguer Gujuti Formation 504 Gunung Api 68
Hainan Island 99, 103, 554 Halmahera 29, 32, 499-500 K/Ar dating methods 500-2 results 502-6 results discussed 506-8 stratigraphy 500 tectonic setting 165,491 Hatapang granite 331 Himalaya trench 8 Himalayan front 9-11 hornblende gneiss, Darvel Bay 265 Huai Hin Lat Formation 206, 209, 236 Huanan Alps orogen 549
559
Huon-Finisterre arc terrane 13 hydrocarbon potential Lao PDR 246 Malay Basin 281,285 Penyu Basin 281 Salawati Basin 479 Thailand 244 Tomori Basin 479
inclination 435,436, 437, 439 indentor model, Banda arc 70-2 India-Asia collision kinematic model 125-8 forward modelling 135-7 model constraints 133-5 rotation evidence 131-3 Tertiary development 129-31 Indian Ocean, seafloor spreading 127 Indoburman flysch 134, 135 Indochina block/terrane 99, 102 tectonic history 204-6 tectonic setting 157 Triassic collisional tectonics 225, 230 Indonesia orogenies 188-96 tectonic development 186-8 tectonic history 204-6 tectonic setting 186 Indosinian orogeny 230, 549, 555 Indus Yarlung Zangbo suture 99 Irian Jaya see Bird's Head; Salawati Basin isothermal remanent magnetism measurement methods 435 results 435, 436, 437 Izu trench 8 Izu-Mariana trench 16
Jass Formation 468 Java, tectonic setting 160 Java trench 7, 8 Java-Sumatra margin, collision evidence 133-4 Javan Forearc 337 Jinshajiang suture 99, 106 Jurassic studies Fujian province 552 palaeomagnetism 206-7 rifting 104 Salawati stratigraphy 468 Sumatra plutonism 324, 325, 330 terrane distribution 114 Tomori stratigraphy 475
K/Ar dating
Bacan 502, 503, 505
560 K/Ar dating (contd) Bisa 503 Halmahera arc 506-8 Khorat granites 244 Obi 503, 504 Papua New Guinea 531 Pak Lay 229 Rajang Group 253 Sumatra granitoids 328-9, 333 western Sulawesi 400-3 Kai Islands 87-8 Kaibobo granite 456 Kaibobo ultramafic complex 4°Ar/39Ardating 457-9 closure temperature 459 implications for Banda Sea 462 implications for obduction 461-2 P-T-t modelling 459-61 petrography 457 Kais Formation 471 Kalaw red beds 208, 210-11,212-13, 216 Kalimantan (Borneo) stratigraphy 384 Rajang Group 251-2, 253, 254-6 see also under Borneo Kalomba Formation 475 Kaputusan Formation 489-90, 505 Kasiruta stratigraphy 487-8 Kayasa Formation 505-6 Kayoa 505 Kazakhstan terrane 99 Kelabit Formation 256 Kelalan Formation 250, 256 Kembelangan Group 468 Kemum Formation 468 Kemum terrane 143 Kerabai volcanics 253 Keskain Formation 468 Ketapang batholith 252 Ketungau basin 254 Khlong Marui fault 159 Khok Kruat Formation 207, 209 Khorat Group 225, 230, 231,233 Khorat Plateau 233-5 hydrocarbon potential 244, 246 inversion history 237-45 palaeomagnetic studies 206-7 stratigraphy 235-7 tectonic development 245 Kimberley block 62, 63 Kintom Formation 475 Kisar Island 68 Klamogun Formation 471 Klasafet Formation 472 Klasaman Formation 472 Klondyke Formation 515 Kluet Formation 324 Kongshuhi Formation 541,542 Kontum massif 206
INDEX Kuantan Formation 324 Kulapis Formation 313, 316 Kun Lun fault 125 Kunlun suture 99 Kunlun terrane 99 Kurogegawa terrane 99, 103 Kutei Basin 384
Labang Formation 312, 316 Lamasi Complex ophiolite 397, 400 Lampang Group 208, 212, 214 Lancangjiang suture 99, 107 Langi Formation 366 Lanping-Simao fold system 539, 540 Lao PDR 539 Khorat Plateau 233-5 hydrocarbon potential 244, 246 inversion history 237-45 palaeomagnetic studies 206-7 stratigraphy 235-7 tectonic development 245 Pak Lay Fold Belt history of research 225-8 recent studies 229-31 summary history 231 Lassi pluton 331 Laurasia-Gondwana marginal terranes 539, 544-5 leaky transform fault 70 Legaspi Lineament 517 Lelinta Shale Formation 468 Lenggurur thrust belt 146 Lhasa terrane 99 Lianga fault 517 Ligu Formation 468 Long Bawang Formation 256 Longmen Thrust 124, 125 Lubok Antu Melange 248 Luogengdi Formation 541 Luok Formation 475 Lupar fault 254 Lupar Formation 248, 250, 251,253 Lupar Line 248, 252 Lupar Line ophiolite 248 Lurah Formation 256 Luwuk Formation 475 Luzon, Miocene tectonics 515
Macolod Corridor 511,520-1 Makassar Strait 394-7 tectonic setting 162 Malacca microplate 323, 324 Malawa Formation 354, 366 Malay Basin 281,292-3 basement 283-4
INDEX heat flow 282 hydrocarbons 281,285 stratigraphy 283, 293 structure 284-7, 293-7 Malay blocks, tectonic setting 159-60 Malay Peninsula palaeomagnetic studies 207-10 tectonic setting 539 Maleta basin 62, 63 Manday suture 99, 110 Manunggal batholith 331 Mariana trench 8, 15-16 Marinduque Basin 521 Matano Formation 475 Matindok Formation 475 Mekong basin 282, 299-300 Melanesian orogeny 195-6, 197 melanges Java 353 Nias Island 342 Sulawesi 356 Melawi Basin 254 Menanga Formation 474 Mengliong Group 540 Mentarang Formation 256 Mentawai fault 21, 24 Mentawai fault zone 338, 345-6, 346-50 Mentawai sliver microplate 21-2 Meratus Mts 252 Meratus suture 99, 110 Mergui microplate 323, 324 metachert, Darvel Bay 271-2 metamorphic studies Darvel Bay 264-73 Gaoligongshan 540 Sulawesi 140-2 metatuff, Darvel Bay 270-1 Miangas ridge 35 Mindanao collision zone 511, 512-15 Mindoro-Panay collision zone 511 mineralogy (heavy), Yunnan glacigenics 543 Miocene studies Asia-India collision 130-1 Banda Sea 462 Malay and Penyu Basins 283 Papua New Guinea 533,534-6 Sulawesi tectonics 384 Sulu Sea Basin 311-16, 317 Misool Island, stratigraphy 468, 470, 472, 571 Molucca Sea 29, 147 bathymetric transect features 32-43 geodynamic framework 32 lithospheric boundaries 43-5 subduction 492 tectonic setting 174-5 Molucca-Sorong fault 467 Moon volcanics 489 Mukus assemblage 323, 324
561
Muna 434 Mussau trench 6 Myanmar (Burma) collision evidence 134-5 palaeomagnetic studies methods 212 results 212-13 stratigraphy 210-11 tectonic setting 539
Nagan granodiorite 331 Nalang Formation 236 Nam Con Son Basin 300 Nam Set Formation 236 Nam Thom Formation 236 Nam Xoi Formation 236 Nam-Phong Formation 206, 209 Nambo Formation 474 Nan-Uttaradit ophiolite zone 225 Nan-Uttaradit suture 99, 107 Nanaka Formation 474 Nasa Formation 236 natural thermal remanent magnetism 435,436, 437 143Nd/144Nd and provenance methods of analysis 406-7, 425, 446 results 447-50 results discussed 450-1 Neogene studies Halmahera arc stratigraphy 500 orogenies 188-96 Philippine tectonics 511-12 docking 512-16 extension 520-1 post docking 516-20 Sundaland rotations 217-18 see also Miocene; Pliocene New Guinea convergence evidence 145-6 indentor 43 Pb isotope ratios 446 plate motion 12-13 see also Papua New Guinea New Guinea Limestone Group 470 New Guinea Mobile Belt 525, 526 New Guinea orogen 147 New Hebrides trench 8, 13 Nias basin 21 Nias Island 338-41 deformation studies 341-6 fault pattems 346 significance of structures 346-50 Nofanini Formation 474 North China terrane 99 North Sula-Sorong fault 467 Northeast China terrane 99 Nusa Babi monzodiorite 488-9, 503
562 NW Borneo-Palawan Trough 308 NW Sabah Platform 308 Nyaan Volcanics 253 Nyalau Formation 251
Obi displaced terrane 466 K/Ar dating 503, 504 Tertiary stratigraphy 501 oblique subduction/convergence 3, 19 Oligocene studies Asia-India collision 129-20 Malay and Penyu Basins 283 New Guinea orogen 147 Papua New Guinea 534 South China Sea extension 248 Sulawesi tectonics 383-4 Ombilin granite 324, 329 ophiolites Darvel Bay 263-4, 277 Lamasi Complex 397, 400 Lupar Line 248 Nan-Uttaradit 225 Sibela 486-7 Opol Formation 513 Ordovician studies 111 orogenic events Banda 57-8, 194-5, 197, 451 Baresan 189-91,197 Huanan Alps 549 Indonesia 188-96, 230, 549, 555 Melanesian 195-6, 197 New Guinea 147 Pacific 225 Sabah 258 Sarawak 256 South China Caledonian 549 Sulawesi 191-4, 197 Sunda 188-9, 197 Talaud 191,192, 197 Yanshanian 549, 555
Pacific orogeny 225 Pacific Plate, Mesozoic subduction 549 Pak Lay Fold Belt history of research 225-8 recent studies 229-31 summary history 231 palaeogeography, terrane distribution 111-14 palaeomagnetic studies methods 212, 434-5 regional results Asia rotation evidence 131-3 Banda Sea 142 Buton 435-42
INDEX Khorat Plateau 206-7 Malay Peninsula 207-10 Myanmar 212-13 relation to plate setting 154-6, 217-19 Pangaea 539, 544-5 Papua New Guinea apatite fission track analysis methods 528-30 results 530 results discussed 530-1 geochemistry 532-3 plate setting 525 stratigraphy and structure 525-7 tectonic summary 533-6 thermochronology 531-2 Pasir Basin 252 Pb isotope ratios methods of analysis 446 results 447-50 results discussed 450-1 Pedawan Formation 250 Penyu Basin 281,297-8 basement 283-4 stratigraphy 283 structure 287-8 Permian studies Fujian province 551,553 Khorat Plateau 235 palaeomagnetism 206 rifting 104 Salawati stratigraphy 468 Sumatra magmatism 323-4, 329 terrane distribution 112 Tomori stratigraphy 475 petroleum see hydrocarbon potential Phalat Formation 236 Phetchbun Fold belt 233, 234 Philippine arc 511 Philippine archipelago Neogene tectonics 137-9, 164-5, 511-12 docking 512-16 extension 520-1 post docking 516-20 Philippine fault 15, 511 Philippine Sea plate 5, 32, 466, 491 boundary evidence 147 coupling with Australia 137, 145-6 rotation estimates 138-9 tectonic setting 162 Philippine trench 15, 32, 44 bathymetry 33-5 Phon Hong Group 235, 236, 244 Phu Phanang Formation 236 Phulekphey Formation 236 Piring granodiorite 253 Piyabung Volcanics 253 plate motions 4, 5-6 boundary convergence 3
INDEX boundary zone deformation 6-7 Plateau Limestone 208, 212 Pliocene studies 536 Poh Formation 475 Prince Alexander Mts 527 provenance analysis see under crustal Pujada ridge 35-9 Pumenqian Formation 540
Qaidam terrane 99 Qamdo-Simao terrane 99 Qiangtang terrane 99 Qinling-Dabie suture 99, 109-10 Quaternary studies Bacan stratigraphy 490 Halmahera arc 506
radiometric dating s e e 4°Ar/39Ar, K/Ar; Rb/Sr; Sm/Nd; U/Pb Rajang Group 253 Kalimantan 251-2 Sarawak 248-51 Ratburi Limestone 208, 212, 214 Raub-Bentong suture 99, 108-9 Rb/Sr dating Sumatra granitoids 324, 328-9 western Sulawesi 406-7, 425 Red River fault 124, 125, 157, 204 Red River suture see Song Ma rifting events 103-5 Ruta Formation 489 Ryukyu trench 8, 15
Sabah 307-8 Central basin 311-16 Darvel Bay Complex 263-4 metamorphic geology 264-73 origins 273-7 ophiolite emplacement 277 NW region deep regional unconformity (DRU) 308-10 structure 311 Rajang Group 254-6 Sabah orogeny 258 Sagaing suture 99, 110 Sahul platform 62, 63 Sainabouli Province see Pak Lay Fold Belt Salawati Basin correlation to Tomori Basin 475-6 hydrocarbon potential 479 stratigraphy 467-72 structure 467 tectonic significance 476-9 Saleh diorite 489, 503
563
Saleh Island stratigraphy 487, 489 Salodik Formation 474, 475 Sampolakosa Formation 435,438, 439, 440 Sandakan Formation 314, 316 Sangihe arc 29, 32, 39 Sangihe basin/trough 39, 40 Sat Khua Formation 207, 209 Sapulut Formation 256 Sarangani-Davao depression 39 Sarawak Ketungau Basin stratigraphy 254 Rajang Group stratigraphy 248-51 Sarawak orogeny 256 Saysomboun Formation 236 Schwaner Mts 252-3 SE Asia, tectonic setting in Cenozoic 155, 158, 161,163, 166, 169, 170, 172, 173, 176, 178 SE Asia (SEA) plate 5 Sebuku Formation 256 seismic sections Sandakan basin 314, 315 Vientiane Plain 239-243 Selangkai Formation 251,253 Semitau ridge 254 Semitau terrane 99, 103 Sepauk tonalite 252, 253 Sepik Basin 526 Seram 445, 455-7 crustal isotopic signature 448 tectonic setting 167-8 ultramafic complex study 4°Ar/39Ar values 457-9 closure temperature 459 implications for Banda Sea 462 implications for obduction 461-2 P-T-t modelling 459-61 petrography 457 Seram trough 86 Serantak Volcanics 253 Setul Limestone 208, 212, 214 Shan boundary 99, 110 Shan Plateau palaeomagnetic studies methods 212 results 212-13 stratigraphy 210-11 Shan-Thai block 126, 132 see also Sibumasu block Shazipo Formation 541,544 shear faults 3 Sibela continental suite 483-4 Sibela ophiolite 486-7 Sibolga granite 324, 329 Sibumasu block/terrane 99, 100-2 tectonic history 204 tectonostratigraphy 205 Triassic collision studies 225, 230
564 Sibuyan sea fault 519-20 Siguangping Formation 541 Sikuleh granite 331 Silantek Formation 248, 254 Silurian studies 468 Simao block/terrane 99, 101 Sintang intrusives 254 Sirga Formation 470 slip partitioning 3, 6-16, 19 slip vectors 19 sliver microplates 19, 21-3 Sm/Nd isochron, western Sulawesi 407, 425 Snellius ridge 42-3 Solomon trench 13 Song Da (Black River) suture 105-6, 204, 554 Song Ma (Red River) suture 99, 105-6, 204, 554 Songpan Ganzi accretionary complex 99 Sorol trough 6 Sorong fault 12-13,467 Sorong fault system 465-6 Sorong-Sulabesi fault 467 South China block/terrane 98-100 Cenozoic motion 124, 125 South China Caledonian orogen 549 South China Sea models of tectonic evolution 247-8 spreading mechanism 133 tectonic setting 157 see also Malay Basin; Penyu Basin South Sula fault 467 South West Borneo terrane 99, 103 Sturt block 62, 63 Sukadana granite 252, 253 Sula Islands 474 Sula platform 167 Sulabesi Island 467 Sulan pluton 331 Sulawesi Bantimala melange complex 354-5 lithology 355 outcrop distribution 356-61 structure 355-6 tectonic evolution 361-2 collisional tectonics 171-4 displaced terrane 466 metamorphic belt 140-2 tectonic setting 160-2 Tertiary magmatic evolution 392-4 distribution of igneous rocks 397 geochemistry 397-400, 425 geochronology 400-11,424-5 petrogenesis 411-13 petrology 397 thermal history 141-15 Tertiary stratigraphy 365-9, 394 depositional environments 375-7 facies description 369-75 Tertiary tectonics 383-4, 415-23
INDEX Sulawesi Group 475 Sulawesi orogeny 191-4, 197 Sulu Sea 164 basin tectonics 311-16, 317 Sumatra plutonism 323-5 geochemistry 325-8 geochronology 328-9 tectonic setting 160, 322-3 Sumatra fault 21, 24, 160 Sumatra fault system (SFS) 322 Sumatra margin accretionary prism transfers 23-4 plate setting 20 sliver plates 21-3 subduction rates 20 tectonic model 24-7 Sumatra trench 8 Sumatran Forearc 7, 337-8 see also Nias Island Sumba 445 crustal isotopic signature 449-50 metamorphic belt 140-2 Sunda orogeny 188-9, 197 Sundaland 250, 281 collision evidence 133-5 deftned 291 marginal features 353,354 northern shelf basin studies see Malay; Penyu; West Natuna; East Natuna; Mekong; Nam Con Son palaeomagnetic studies history of study 206-10 methods 212 results 212-17 results discussed 217-19 setting for sites 210-12 significance of results 219-22 tectonic setting 171,204 tectonic summary 300-5 see also Sumatra suturing events 105-11
Tagoloan fault 514 Taiwan collision zone 511 Talaud orogeny 191,192, 197 Talaud ridge 40 Tamangil Formation 87 Tanamu Formation 474 Tang Ting fault 125 Tanimbar trough 86 Tanjong Formation 313 Tarim block/terrane 99, 124, 125 Tatau formation 250, 251 Tawali Besar, stratigraphy 487-8 Tawali Formation 487-8, 491
INDEX tectonic overview of SE Asia 155 Temburong Formation 256 Tengchong Block 540 diamictite 541-4 terrane motion 544-6 terrane evolution origins 98-103 palaeogeography 111-14 Tertiary studies Bacan stratigraphy 487 Bacan tectonics 490-3 Salawati stratigraphy 471 Sumatra plutonism 331-3 Sunda Shelf kinematics 301 Tomori stratigraphy 475 see also Eocene; Miocene; Oligocene; Neogene; Pliocene Tetambahu Formation 474 Tethys Ocean, phases of 103-5,546 Thai-Malay Peninsula 159-60 Thailand hydrocarbon potential 244 palaeomagnetic studies 207-10 rotation 217-18 tectonic setting 539 Thangon Formation 236 thermochronology, Papua New Guinea 531-2 Three Pagodas fault 159 thrust faults 3 Tibet block, Cenozoic motion 124, t25 Timor role in Banda orogen 57 structural model 47 TIMOR line 48, 49-56, 64, 76, 78-82 Timor trough 11, 51-3, 86 evolution 75-6 seismic reflection profile 81 Tipuma Formation 468 Tokala Formation 474 Tomori Basin 472-3 correlation to Salawati Basin 475-6 hydrocarbon potential 479 stratigraphy 470, 471,474-5 structure 473-4 tectonic significance 476-9 Tomori Formation 475 Tonasa Limestone Formation 354, 365, 366 depositional environment 377-80 facies 368-75 origins 375-7 stratigraphy 367-9 tectonic environment 383-4 Tondo Formation 435, 438, 439, 440 Tonga trench 8, 13-15 Torricelli Intrusive Complex 527 Torricelli terrane 147 Triassic studies collisional tectonics 225
Fujian province 551-2, 553 palaeomagnetism 206 rifting 104 Salawati stratigraphy 468 Sumatra plutonism 324, 329-30 terrane distribution 112 Tomori stratigraphy 475 Trusmadi Formation 256 Tukang Besi platform 160, 167, 434
U/Pb dating, western Sulawesi 410-11,425 Ujung Kulon fracture zone 21 Ulai intrusion 331
Verde Passage fault 519-20 Visayas, Neogene accretion 515-16 volcanic arc, role in Banda orogen 55-6, 57 Vulcan basin 62, 63
Waaf Formation 470 Waigeo, tectonic setting 491 Walanae Depression 366 Wallace Line 394 Wang Chao fault 159 Weber basin 87 Weda Group 503, 505 Weduar Formation 87 Weryahan Formation 87 West Andaman fault 22, 23 West Burma Block 99, 102 West Kutei basin 253 West Malaysia see Malay Peninsula West Natuna Basin 282, 298-9 stratigraphy 283 West New Guinea composite terrane 140, 141 West Philippine basin 32 Wetar thrust 11, 61 Wharton basin 19 Woi Formation 504 Woniusi Formation 541,543-4, 544 Woyla suture 99, 110 Woyla terranes 99, 103
Xianggui block 100 Xikou Formation 551-2, 554 Xilao-He suture 99
Yangtze block 100 Yangtze para-platform 549
565
566 Yangtze Platform 539, 540 Yanshanian orogenies 549, 555 Yap trench 6 Yefbie Shale 468 Yongde Formation 541, 544 Yunnan (SW) 539-41
INDEX glacigenic sediments 541-3 origins 544-6
'Zhen'he-Da'pu fracture 549-50 Zishi Formation 541
