Contourites

CONTOURITES

Edited by

M. Rebesco and

A. Camerlenghi

Amsterdam • Boston • Heidelberg • London • New York • Oxford Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo

Elsevier Science Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK First edition 2008 Copyright Ó 2008 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-444-52998-5 For information on all Elsevier publications visit our website at elsevierdirect.com Printed in the Great Britain 08 09 10 11 10 9 8 7 6 5 4 3 2 1

Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org

PREFACE

Contourites are sediments deposited or substantially reworked by bottom currents. The study of contourites is nowadays crucial for several fields of fundamental and applied research: 1. palaeoclimatology and palaeoceanography, since these fairly continuous and relatively high-resolution sediments hold the key for priceless information on the variability in ocean circulation patterns, current velocities, oceanographic history and basin interconnectivity; 2. hydrocarbon exploration, since accumulation of source rocks may be favoured by weak bottom currents, whereas ‘‘clean’’ deep-sea sands may be formed by robust flows; 3. slope stability, since fine-grained, low-permeability, high pore-water content contourites facilitate the formation of overpressurized gliding planes when rapidly loaded, or when their rigid biosiliceous microfabric collapses due to diagenetic alteration. Despite its significance, this group of sediments is poorly known by the majority of non-specialists. Notwithstanding the growing interest and the now intensive research in contourites, a textbook that serves as a reference book on contourites was missing until now. This book addresses all aspects of the knowledge in the field of contourites and provides a comprehensive and cross-referenced coverage of the subject. It can serve as a standard reference work for non-specialists, and in particular postgraduate students, university teachers and lecturers, researchers and professionals who are seeking an authoritative source of information about contourites. It reviews both theoretical topics and case histories, and it points to approaches that may help tackle open problems. Divided into a wide and interdisciplinary spectrum of topical sections, it provides practical advice on multidisciplinary research techniques. The ample use of illustrations, diagrams, photographs and maps, complemented by a CD-ROM including all illustrations (as many as possible in colour), provides a helpful tool for researchers in the preparation of classroom lectures and training courses, journal articles and meeting presentations. The figures presented in this book are partly new, partly adapted from previous works, and partly reproduced from previous works. Wherever appropriate, permission has been obtained for the reproduction or adaptation, but not all copyright holders could be traced. We would appreciate if such copyright holders would be so kind as to contact us. The authors invited to contribute to this book are distinguished specialists in the field. They have been invited not simply to produce a research paper on the topic, but to critically review the current state of knowledge. They have been asked to accomplish their review in a clear and concise way, including as much factual.

xvii

xviii

Preface

information as possible. They have been encouraged to submit as many tables and illustrations (diagrams, photographs and maps) as are necessary to complement the text of their chapter to present the information in the best possible way. The authors have been recommended to keep the language simple, and those whose native language is not English have been encouraged to choose co-authors whose mother tongue is English and/or to propose native English reviewers for their manuscripts. The editors and the authors are grateful to the reviewers who contributed to improve the scientific quality of each chapter and to the book as a whole: A. Arche, C. Bjerrum, A. Camerlenghi, A. Cattaneo, X.D. de Madron, G. Evans, F. Eynaud, J.-C. Fauge`res, M. Frenz, Z. Gao, M. Gardner, J. Jones. D.W. Kirkland, P. Knutz, J.S. Laberg, R. Larter, E. Llave, D. Long, J. Lo´pez-Go´mez, L. Lo¨wemark, L. Masse´, T. Mulder, H. Nelson, M. Rebesco, C. Silva, P. Talling, D. Thornalley, R. Urgeles, A.J. van Loon, M. Vanneste, A. Wiewio´ra. Michele Rebesco and Angelo Camerlenghi

LIST oF CONTRIBUTORS

A. Camerlenghi ICREA, c/o Universitat de Barcelona, Departament d’Estratigrafia, Paleontologia i Geocie`ncies Marines, GRC Geocie`ncies Marines, C/Martı´ i Franque`s, s/n, E-08028 Barcelona, Spain. [email protected] B. Chaco´n Fachbereich Geowissenschaften, Universita¨t Bremen, D-28334 Bremen, Germany. [email protected] A. Crise Istituto Nazionale di Oceanografia e Geofisica Sperimentale, B.go Grotta Gigante 42/c, I-34010 Sgonico (TS), Italy. [email protected] T. Duan Marathon Oil Company, Houston, Texas, USA. [email protected] J.-C. Fauge`res De´partement de Ge´ologie et Oce´anographie, Universite´ Bordeaux1, UMR CNRS 5805 EPOC, Avenue des Faculte´s, F-33405 Talence cedex, France. [email protected] M.A. Fregenal-Martı´nez Department of Estratigrafı´a Facultad de Ciencias Geolo´gicas, Universidad Complutense, E-28040 Madrid, Spain. [email protected] Z. Gao School of Geosciences, Yangtze University, Jingzhou, Hubei province, 434023, China. [email protected] P. Giresse Laboratoire d’E´tudes des Ge´o-Environnements Marins, Universite´ de Perpignan, 52, Avenue Paul Alduy, F-66860 Perpignan, France. [email protected] E. Gonthier De´partement de Ge´ologie et Oce´anographie, Universite´ Bordeaux1, UMR CNRS 5805 EPOC, Avenue des Faculte´s, F-33405 Talence cedex, France. . [email protected] xix

xx

List of Contributors

Y. He School of Geosciences, Yangtze University, Jingzhou, Hubei Province, 434023, China. [email protected] F.J. Herna´ndez-Molina Facultad de Ciencias del Mar, Universidad de Vigo, E-36200 Vigo, Spain. f [email protected] J.A. Howe Scottish Association for Marine Science & UHI Millennium Institute-Dunstaffnage Marine Laboratory, Oban, Argyll, Scotland, PA37 1QA, UK. [email protected] K.J. Hsu¨ Oakcombe, Marley Common Haslemere, Surrey, GU27 3PT, UK. [email protected] H. Hu¨neke Institute of Geography and Geology, University of Greifswald, D-17487 Greifswald, Germany. [email protected] S. Hunter National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, SO14 3ZH, UK. [email protected] P.C. Knutz Geological Survey of Denmark and Greenland (GEUS), Øster Voldgade 10, 1350 Copenhagen, Denmark. [email protected] A. Kuijpers Geological Survey of Denmark and Greenland (GEUS), Øster Voldgade 10, 1350 Copenhagen, Denmark. [email protected] J.S. Laberg Department of Geology, University of Tromsø, N-9037 Tromsø, Norway. [email protected] E. Llave Instituto Geolo´gico y Minero de Espan˜a, Rı´os Rosas, 23, E-28003 Madrid, Spain. [email protected] A. Maldonado Instituto Andaluz de Ciencias de la Tierra. C.S.I.C./Universidad de Granada, Campus de Fuentenueva, s/n, E-18002 Granada, Spain. [email protected]

List of Contributors

xxi

J. Martı´n-Chivelet Department of Estratigrafı´a Facultad de Ciencias Geolo´gicas, Universidad Complutense, E-28040 Madrid, Spain. [email protected] D.G. Masson National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, SO14 3ZH, UK. [email protected] I.N. McCave Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge, CB2 3EQ, UK. [email protected] T. Mulder De´partement de Ge´ologie et Oce´anographie, Universite´ Bordeaux1, UMR CNRS 5805 EPOC, Avenue des Faculte´s, F-33405 Talence cedex, France. [email protected] T. Nielsen Geological Survey of Denmark and Greenland (GEUS), Øster Voldgade 10, 1350 Copenhagen, Denmark. [email protected] M. Rebesco Istituto Nazionale di Oceanografia e Geofisica Sperimentale (OGS), Borgo Grotta Gigante 42/C, I-34010 Sgonico (TS), Italy. [email protected] S. Salon Istituto Nazionale di Oceanografia e Geofisica Sperimentale, B.go Grotta Gigante 42/c, I-34010 Sgonico (TS), Italy. [email protected] G. Shanmugam Department of Earth and Environmental Sciences, The University of Texas at Arlington, Box 19049 – Arlington, TX 76019 USA. [email protected] M.S. Stoker British Geological Survey, Murchison House, West Mains Road, Edinburgh, EH9 3LA Scotland, UK. [email protected] D.A.V. Stow National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, SO14 3ZH, UK. [email protected]

xxii

List of Contributors

F. Trincardi ISMAR-CNR, Via Gobetti 101, I-40129, Bologna, Italy. [email protected] A.J. Van Loon Geological Institute, Adam Mickiewicz University, Mako´w Polnych 16, 61-606 Poznan, Poland. [email protected] T. van Weering Royal Netherlands Institute for Sea Research (NIOZ), P.O.Box 59, Texel, 1790 AB Den Burg, and Department of Paleoclimatology and Geomorphology, Free University, de Boelelaan 1085, 1081 HV Amsterdam, The Netherlands. [email protected] G. Verdicchio ISMAR-CNR, Via Gobetti 101, I-40129, Bologna, Italy and EDISON SpA, Foro Buonaparte 31, I-20121, Milano, Italy. [email protected] A.R. Viana Petrobras, Research Center, R&D Exploration, Rio de Janeiro, Brazil. [email protected] F. Werner Institut fu¨r Geowissenschaften, Universita¨t Kiel, Olshausenstrasse 40-60, D-24118 Kiel, Germany. [email protected] A. Wetzel Geologisch-Pala¨ontologisches Institut, Universita¨t Basel, Bernoullistrasse 32, CH-4056 Basel, Switzerland. [email protected] D. Wilkinson National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, SO14 3ZH, UK. R.B.Wynn National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, SO14 3ZH, UK. [email protected] W. Zenk Leibniz Institute of Marine Sciences at the University of Kiel (IFM-GEOMAR), Du¨sternbrooker Weg 20, D-24105 Kiel Kiel Germany. [email protected]

P A R T

1

CONTOURITE RESEARCH

C H A P T E R

1

C ONTOURITE R ESEARCH : A F IELD IN F ULL D EVELOPMENT M. Rebesco1, A. Camerlenghi2 and A.J. Van Loon3 1

Istituto Nazionale di Oceanografia e Geofisica Sperimentale (OGS), Sgonico (TS), Italy ICREA, c/o Universitat de Barcelona, Departament d’Estratigrafia, Paleontologia i Geocie`ncies Marines, Barcelona, Spain 3 Geological Institute, Adam Mickiewicz University, Poznan, Poland 2

Contents 1.1. 1.2. 1.3. 1.4. 1.5.

Bottom Currents Contourites Drifts Sedimentary Structures Prospects

4 6 8 9 10

Contourites were first recognized and described only 40 years ago when photographs of the deep-sea bottom showed distinct current ripples (Heezen and Hollister, 1964; Hollister, 1967); much progress has been made since, but the current status in this field shows clearly that this relatively new research topic is still in full development: a surge of new research results (reflected in, for instance, the publication of several special volumes of scientific journals) go still hand in hand with uncertainties and with insufficient knowledge (there is, for instance, a lack of indisputable diagnostic criteria). New techniques are continuously developed and implemented (e.g., swath bathymetry and 3-D seismics). All this indicates a great potential for further research, in the same way as the now mature research on turbidites developed in the 1960s. Comparing the research on turbidites with that on contourites, it seems that the latter is presently facing a transition from adolescence to maturity. One might consider such a period of transition not to be the right moment to compile a reference book, which is aimed at providing both experts in the field a state-of-the-art overview, and the non-specialist readers an introduction that may help starting research in this field. However, we consider it just crucial that this transitional stage of contourite research is documented. It is the right moment to increase the recognition of contour currents as important transport and sedimentary phenomena that control much of the deep-sea sedimentation; in addition, this seems the right moment to bring new forces into play. The aim of this book is therefore not to present an inventory of the fossilized knowledge of a fully developed field, but rather to provide a basis for future Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00201-X

Ó 2008 Elsevier B.V. All rights reserved.

3

4

Contourite Research: A Field in Full Development

development, a platform for ample exchange of complementary expertise, ideas and experience. The insight into the role of contourites will certainly benefit from input by new researchers within the – so far – fairly restricted circle of the current specialists in this area of research. A good understanding of contourites and of the sedimentary processes behind them is often complicated by a lack of clear, unambiguous and commonly accepted diagnostic criteria, and by discussions about conflicting views about the sedimentary structures. The different views that still exist are particularly confusing for nonspecialists and students, who may easily become disoriented by such complex situations. With this book, and with this introduction in particular, we want to offer to non-specialist readers a simplified, often descriptive, overview of contourites. We are aware of the risk implied in any simplifications. For this reason, we neither hide the problems nor provide oversimplified explanations. We asked the authors of the various chapters to accomplish in a clear and concise way a critical review of the current state of knowledge regarding the subject of their chapter. This implies that the book does not contain contributions reflecting new research. Nor is it our intention to present, in this introduction, new syntheses or untested new hypotheses. In contrast, we provide a kind of glossary that may help the reader to find his way, with special reference to key terms. These terms include the ‘‘generator’’ (bottom currents), the sediments (contourites), the types of accumulation (drifts) and the types of sedimentary structures (traction, bioturbation). A comprehensive text on contourites is, in our opinion, justified also because of the incomplete coverage of the subject by reference books, to which the non-specialist readers are referred before entering the details of specific journal articles: even the Encyclopedia of Sediments and Sedimentary Rocks (Middleton et al., 2003) does not include ‘‘Contourites’’ among its entries! The Glossary of Geology (Bates and Jackson, 1987) lists 43 entries for ‘‘drift’’, of which the only one referring specifically to sedimentology lists wind or river currents as main ‘‘generators’’ of sedimentary drifts.

1.1.

BOTTOM C URRENTS

‘‘Bottom current’’ is our preferred term to refer to the water-mass flows that control the deposition of contourites. Many other terms are often used, for which reason we need to clarify the meaning. Among geologists, the most adhered to definition of contourites is that they are sediments deposited or substantially reworked by the persistent action of bottom currents (e.g., Stow et al., 2002c; Rebesco, 2005). However, the term ‘‘bottom currents’’ is not usually employed by physical oceanographers. It is obvious that the time span of the processes involved in most geological processes is quite different from that analysed by physical oceanographers. In fact, the semi-quantitative inferences derived by geologists from the impact of ocean currents on the sea-floor sediments represent an integral over a poorly defined time interval. A detailed discussion about bottom currents from the point of view of physical oceanographers is given by Zenk (2008). However, we state as a general simplification that any ‘‘persistent’’ water current near the seafloor may be called a ‘‘bottom current’’.

M. Rebesco et al.

5

Such a current has the capability to affect the sea floor by re-suspending, transporting and/or controlling the deposition of sediments. This type of current is influenced by a number of processes (tides, internal waves, barotropic waves, dynamic instabilities) that modulate the speed and the instantaneous direction. Though typically affecting the sea floor, episodic flows that are not in equilibrium conditions do not belong to this category. Bottom currents do therefore not include turbidity currents, which are – in contrast to contour currents – rapid, dense (loaded with suspended sedimentary particles) currents driven by gravity, which fail to develop equilibrium conditions, even if they perhaps have more impact on the sea floor than bottom currents. For the same reason, the specific type of thermohaline circulation (THC) called ‘‘submarine overflows’’ (e.g., Denmark Strait Overflow Water) or ‘‘cascading currents’’, that are generated by transients in water-mass distribution in response to the local effect of evaporation, cooling or freezing in the surface layer over the continental shelf (Shapiro et al., 2003; Ivanov et al., 2004), should not be strictly considered as a bottom current. These currents are produced by the sinking – in relatively confined ‘‘flow pipes’’ – of important volumes of sea water, typically with high speed and intermittent flow; they occur every few years. The flow is controlled by morphological elements such as sills, submarine canyons and narrow straits. They are included among the engines of the THC at a basin-wide scale because the intermittent flow is – in the long term – compensated by the persistence of the process over very long time intervals, often coinciding with long-term climatic cycles. Known as purely physical oceanographic processes, these currents are now known to have an effect on the sea-floor morphology and to be able to transport important quantities of sediment in suspension (Canals et al., 2006; Trincardi et al., 2007). Future work should address the sedimentological importance of these currents. Shallow-water motion produced by surface waves, storms and tides (which are intermittent and not in equilibrium conditions), does not typically develop long-lasting bottom currents, though they affect the sea floor of continental shelves. Nevertheless, bottom currents can be affected by a number of distinct forces acting at different water depths. The term ‘‘bottom current’’ should be considered as a generic term that embraces different types of current. For clarity, we list here the (sometimes coexisting) types of current included within this term: • wind-driven currents, which originate by horizontal movement of the superficial layers due to wind shear stress, and subsequent propagation of the motion through the water column down to greater depth; • thermohaline currents, driven by gravity (uneven density distributions due to variable temperature and salt content of water masses), which are the most common type of bottom current; the predominantly horizontal large-scale transport of a water mass (advection) is called ‘‘THC’’ or ‘‘meridional overturning circulation’’ (MOC) in the Atlantic Ocean and determines the so-called oceanic conveyor belt; • geostrophic currents, which are characterized by long-lasting equilibrium conditions between the horizontal pressure gradient and the Coriolis force. By definition they have zero vertical velocity, which implies that the flow is forced

6

Contourite Research: A Field in Full Development

to follow the bathymetric contours (therefore these currents are also named ‘‘contour currents’’); • contour currents, which are currents that have a net flow along-slope, sub-parallel to the topographic isobaths (contours); in spite of that, for certain parts of their path they can also flow upslope, down-slope, around and over topographic obstacles or irregularities; • boundary currents, which are currents the direction of which is controlled by sea-floor morphology (a wide canyon, continental slopes, but also flanks of major submarine mountain chains); due to the meridional distribution of continents and the Coriolis force, they are typically intensified along the western boundaries of the oceans (western boundary currents); • abyssal currents (sometimes also indicated as ‘‘ocean currents’’), which may be considered synonyms of large-scale bottom currents flowing in the deep sea (below the edge of the continental shelf) where topography and Coriolis force play a major role in determining the pathway of the current. Two more processes are important in the context of the processes related to contourite formation: • downwelling, which is the downward transport of cooled water masses from the surface; it occurs in restricted regions, mainly at polar latitudes, resulting in bottom currents; note that downwelling does not necessarily occur at the ocean margin, so it not necessarily affect the sea floor; • upwelling, which involves – in contrast – deep-water masses, and which produces bottom currents against topographic barriers and continental margins, mainly at low latitudes. Bottom currents can also be forced by deep-water tidal currents (in submarine canyons), long-wave baroclinic currents (internal waves and tides, solitary waves) and tsunami-related traction currents (Shanmugam, 2008). In conclusion, most (though not all) bottom currents start as density currents sinking to their equilibrium level (quasi-geostrophic balance). They are predominantly unidirectional subsurface currents in contact with the sea floor. They show a quasi-steady flow, though possibly affected by tides, seasonal changes and/or migrating eddies. They are controlled by topography, and when their direction largely parallels the topography, they are called ‘‘contour currents’’.

1.2.

C ONTOURITES

The term ‘‘contourite’’ was originally introduced to define the sediments deposited in the deep sea by contour-parallel thermohaline currents. Early pioneering work documented the strong influence of the deep Western Boundary Undercurrent along the continental margin of eastern North America (Heezen and Hollister, 1964; Heezen et al., 1966; Schneider et al., 1967). However, a rigorous restriction to this definition would prevent the application to ancient deposits, where both depth and direction of the currents can rarely be precisely reconstructed.

M. Rebesco et al.

7

The definition was consequently widened to embrace a larger spectrum of sediments that are affected to various extent – and in a wide range of water depths – by different types of current. According to a now widely accepted recent definition, contourites are sediments deposited or significantly affected by bottom currents (Stow et al., 2002c; Rebesco, 2005; Stow and Fauge`res, 2008). Yet, the different types of bottom current are known to influence to a greater or lesser extent many depositional environments. They affect various types of sediment, both during and after deposition. This implies the risk of an excessively wide application of the term ‘‘contourite’’, and consequently of a loss of significance. On the contrary, where the dominant action of a bottom current is ascertained, the presence of other types of sediment is not excluded. Turbidites, for example, occur frequently even where the bottom-current influence is large enough to control the overall geometry of the deposits and to generate a (contourite) sedimentary drift. It is therefore not possible to restrict the use of the term ‘‘contourites’’ to the sediments contained in a sediment drift. In contrast, it is commonly accepted that contourite sedimentary facies also include sediments, usually interbedded, that are not deposited under the influence of bottom currents. This is especially true where the persistence of bottom-current action has not been enough to determine the geometry of the deposit. Another way to restrict the use of the term ‘‘contourites’’ is to set a minimum depth above which contourites cannot reliably be distinguished from shallow-water shelf-current deposits. This water depth is suggested to be around 300 m, according to the definition by Stow et al. (2008), even though this limit should not be applied rigidly. Shallow-water contourites may reflect also other hydrodynamic factors (shelf currents, tides and waves, storms) capable of impinging the sea floor, but the influence of which is still negligible compared to a dominant (though not absolutely steady) bottom current. As discussed in detail by Verdicchio and Trincardi (2008a), the processes involved in the formation of shallow-water contourites are generally more varied and less steady than in the case of deep-water contourite deposits. Shallow-water bottom currents derived from a stable geostrophic circulation can nevertheless form contourite drifts resembling the typical deep-water contourite drifts in morphology, internal geometry and sedimentary facies. Therefore, water depth alone does not seem to be an effective criterion to restrict the use of the term ‘‘contourites’’. As is often stressed in literature, the process of contourite deposition is not a simple one: it often involves multi-phase entrainment, long-distance transport, and interaction among depositional processes induced by the various types of bottom current described above (e.g., He et al., 2008; McCave, 2008; Stow et al., 2008). In addition, contourites are generally not easy to recognize because of the lack of simple, unambiguous diagnostic criteria. A composite triple-stage approach (e.g., Fauge`res et al., 1999; Rebesco and Stow, 2001; Nielsen et al., 2008) is recommended for the identification of sediments deposited by bottom currents as can be determined beyond reasonable doubt from seismic-reflection data: the analysis must include the overall architecture of the deposit (gross geometry and large-scale depositional units), the internal architecture (structure and sub-units) and seismic attributes and facies in each sub-unit.

8

Contourite Research: A Field in Full Development

When samples are available, the problem of the lack of universally recognized unequivocal diagnostic criteria for the sedimentary characteristics of contourites must be faced. Though most workers suggest that pervasive bioturbation is the most diagnostic criterion (Stow and Fauge`res, 2008; Wetzel et al., 2008), others suggest a combination of traction structures (Martı´n-Chivelet et al., 2008; Shanmugam, 2008). Distinctive characteristics that are not diagnostic by themselves may be provided by early diagenesis (Giresse, 2008) and physical properties (Laberg and Camerlenghi, 2008). We therefore suggest that contourites as such should be identified on the basis of analysis of the characteristics of their facies and facies associations (cf. Stow and Fauge`res, 2008). Considering the above, we suggest that the well-established term ‘‘contourite’’ should be used as a generic term, in the same way as, for example, ‘‘mass-wasting deposits’’ or ‘‘gravity-flow deposits’’. These generic terms may be considered family names that simply describe a process that affects a certain kind of sediment: bottom currents in the case of contourites, down-slope mass transport of sediment and water for mass-wasting deposits. These family names include several kinds of sediment that have more specific names (e.g., turbidites, debris-flow deposits or mudflow deposits). In the case of contourites, such specific names do not exist; they should be defined in the time to come on the basis of research aimed at distinguishing between the various types of transport and depositional processes involved.

1.3.

D RIFTS

In sedimentology, ‘‘drift’’ is a general term used to describe ‘‘unconsolidated rock debris transported from one place and deposited to another’’ (Bates and Jackson, 1987). Its use is taken from one of the many definitions that can be found in a dictionary: ‘‘motion or action under external influence’’. Its use has traditionally been related to sediment transport by river currents, wind and glaciers. It is important to note that the ‘‘drifting’’ implied by the term does not refer to the deposit, but to the particles that move, transported by a flow, before settling. The first to use the term in association with bottom currents were Heezen and Johnson (1963), who stated that ‘‘scour and drift of sediments due to current activity are the most reasonable explanation of sediment moats’’ that were identified at that time next to seamounts and other striking morphological structures in the Atlantic and Pacific. The drift of sediment particles by the geostrophic circulation was later recognized as responsible not only for sediment moats, but also for sediment knolls (Heezen et al., 1966) as the drift could focus sedimentation in certain areas of the continental rise. The use of the term ‘‘drift’’ then shifted from that of a process for sediment transport by deep contour currents, to the end product of that process: the sedimentary deposit. The Blake–Bahama Outer Ridge was defined by Heezen and Hollister (1971) as a migratory sediment drift built by bottom currents. The sediments deposited by the action of bottom currents, whether or not forming a sediment drift, were termed ‘‘contourites’’ by the same authors. Because the flow responsible for the ‘‘drift’’ of particles from one place to another and the resulting sedimentary deposit is implicitly a steady one, mass-flow deposits resulting from surges of density currents were

M. Rebesco et al.

9

distinguished in principle from contourites since the very beginning (see the early definitions by Hollister and Heezen, 1971, p. 421). Bottom currents are capable of building thick and extensive accumulations of sediments not only on the continental rise, but also elsewhere along continental margins, from the abyssal floor to the outer shelf, if a significant sediment input is available (Fauge`res and Stow, 2008). These sediment bodies have received various names, including ‘‘outer ridges’’, ‘‘sedimentary ridges’’, ‘‘sedimentary mounds’’, ‘‘sediment drifts’’ and ‘‘contourite drifts’’. In our view, all these terms are synonyms, referring to accumulations of sediment deposited – or significantly affected – by bottom currents. Strictly speaking, the term ‘‘contourite drifts’’ should be specifically used for sediment accumulations deposited by currents flowing along the contours. However, we highlighted already above that different types of bottom current exist, and that even the so-called ‘‘contour currents’’ do not always follow the contours. An additional complication is provided by sedimentary accumulations for which the current direction is inferred only indirectly. We hence think that the term ‘‘contourite drifts’’ should be used for sediment accumulations deposited by bottom currents in general. Further, because of the presence of ‘‘contourite’’ in this term, we think that this term is to be preferred. In many cases, especially where mixed turbidite/contourite systems are involved, terms that not belong to the contourite terminology in a strict sense (e.g., mounds, levees, fans, lobes, channels) are used in the literature for deposits that have been significantly affected by the interaction with bottom currents. We recommend for such situations that these terms be preceded by the prefix ‘‘contourite’’ (e.g., contourite levees). Not only depositional structures are produced by bottom currents: there are a number of erosional and non-depositional structures. The classification and precise terminology of these structures that are related to bottom currents have not been fully developed yet; only some attempts have been made (Herna´ndez-Molina et al., 2008a, b).

1.4.

S EDIMENTARY STRUCTURES

What are the diagnostic sedimentary structures of contourites? The scientific community has not come yet to complete agreement. There are two viewpoints, with contrasting ways of reasoning: one is in favour of traction structures, whereas the other one is in favour of bioturbation. The initial finding in the 1950s of current ripples and other traction structures in deep-marine deposits affected by bottom currents (Hollister, 1967; Hollister and Heezen, 1967, 1972; Bouma, 1972a, b, 1973; Bouma and Hollister, 1973) suggested that contourites show lamination as a result of fluid-flow processes and depositional sorting mechanisms. Distinctly laminated deposits are still interpreted as contourites by some authors (see Hu¨neke and Stow, 2008; Martı´n-Chivelet, 2008; Shanmugam, 2008). These authors stress that traction structures are abundant on modern ocean floors that are influenced by bottom currents; they also interpret ancient coarsegrained deposits in some large outcrops as contourites. They wonder whether the general absence of such structures in the case studies based on sedimentary cores may

10

Contourite Research: A Field in Full Development

result from a bias imposed by the impossibility of larger-scale observations. They also suggest that bioturbation in contourites should not be considered a diagnostic criterion, since bioturbated mud is equally abundant in areas unaffected by bottom currents and since turbidites may be extensively bioturbated as well. According to most contourite workers (e.g., Stow and Fauge`res, 2008), crosslamination is only rarely described from modern contourites, whereas extensive bioturbation is generally dominant (Wetzel et al., 2008). Their observations are derived from a huge number of data – published over the past 15 years – from modern drift systems using conventional coring techniques. They suggest that the lack of clear lamination in contourites is due to several reasons, including the commonly low current velocity that is insufficient to produce primary lamination, the relatively low accumulation rate allowing bioturbation to destroy primary lamination, and the small sediment input that is insufficient to allow a sorting mechanism to develop lamination. The controversy regarding diagnostic criteria (traction structures versus bioturbation) for the identification of contourites is still a problem. It was not possible to provide here a solution, considering the present state of the knowledge. We therefore purposely took care that both views are adequately expressed in this volume by inviting authors belonging to the two ‘‘schools’’. In our opinion, they honestly offered their own evidence, and we hope that the comprehensive scenario provided by this scientific, non-personal confrontation will provide a sound basis for a new step forward in the ongoing struggle to clarify this controversial issue.

1.5.

P ROSPECTS

More research is obviously needed to define a universally acceptable set of diagnostic criteria for contourites. Such research should be aimed at obtaining a much deeper insight into the processes involved. Simultaneously, the existing terminology for all aspects related to contour currents (deposits, processes, morphology, etc.) needs more consistency and logic, and new terminology has to be developed for aspects for which no real terminology exists as yet. This is a prerequisite for progress, because more precise terms and a stricter application of the terminology are needed for unambiguous interpretation of descriptions, as well as for better comprehension of the numerous processes involved. Improved terminology is, as mentioned before, particularly important to distinguish between the various types of contourite, and for classification of erosional structures that are related to bottom currents. The growing attention for these sediments – because they can be hydrocarbon seal rock and reservoirs, because they form a significant part of the palaeoceanograpic record, and because they form a source for potentially hazardous submarine landsides – will hopefully give rise to an ever increasing amount of high-quality data sets that will improve the understanding of these sediments, especially on the basis of their sedimentology and their seismic characteristics. An intensified access to industrial-quality hydrocarbon-exploration data in particular may play an essential role in an increased understanding of contourites, just like was the case in the 1960s for turbidites.

C H A P T E R

2

P ERSONAL R EMINISCENCES ON THE H ISTORY OF C ONTOURITES ¨ K.J. Hsu Oakcombe, Marley Common Haslemere, Surrey, UK

Contents 2.1. 2.2. 2.3. 2.4.

Introduction Not All Deep-Sea Sands are Turbidites Turbidophiles and Turbidophobes Erosional Unconformity Misinterpreted

2.1.

11 12 14 15

INTRODUCTION

I studied metamorphic petrology at the University of California (UCLA) in the early 1950s, and my ambition was to become a teacher of petrology at an American university. That was 10 years after the repeal of the Chinese Exclusion Act and there were not yet voices in the Congress on equal employment. I had little luck in finding a job. The nearest I came was a chance to become an acting lecturer at the Oregon State University as a 1-year substitute for somebody with sabbatical leave. However, my application was rejected. Eventually I found work for Shell Research, because it was a Dutch company. No American company would employ me because it would be impossible for Orientals to make business trips to the Deep South, where hotels existed only for the white or the coloured. I was not well prepared for my new job, not having taken any course on sedimentology. I asked for some advice, when I went to say goodbye to Jerry Winterer. Winterer was a classmate, but also he taught sedimentary petrography. I did not know him very well, because we, ‘‘hardrock guys’’, had our coffee room on the fourth floor, while the ‘‘softies’’ congregated on the fifth. Winterer was very pleased to learn that I was going to work for Shell. He just read an article in the AAPG Bulletin by someone there called Nanz, on the geometry of Oligocene sandstone reservoirs of the Seeligson Field in south Texas. It was brilliant, Winterer told me. I should look him up; perhaps he might teach me a thing or two on the art and science of basin analysis. I did not have to look Nanz up: I was assigned to him as a trainee. Taking into account of my experiences in California, Nanz asked me to make a study of the Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00202-1

Ó 2008 Published by Elsevier B.V.

11

12

Personal Reminiscences on the History of Contourites

Pliocene sandstone reservoirs of the Ventura Field, California. There had been rumors that those might be turbidites. Nanz was ready to consider all scientific evidence, while most of his colleagues were downright hostile to the innovative suggestion. I went to Ventura in 1954, and the Pliocene sands are obviously turbidites, as can be deduced from their sedimentary structures. Applying Nanz’s technique of mapping the geometry of genetic units, I recognized on the electric logs a group of sand layers within a thick succession of shale. With easy correlation, I could map the distribution of this genetic unit. It had been thought, on the basis of the Grand Banks study by Heezen and Ewing (1952), that turbidites were blanket sands. I found, to the surprise of everyone, that the Ventura turbidites are shoestring bodies (Hsu¨, 1977). In hindsight, the discovery should not have been surprising. The event started from a point, and the loci of points constitute a line. Turbidites originated from a single locality, and should thus be linear sand bodies. My Shell report was internally distributed in 1954, but it was not published for more than two decades (Hsu¨, 1977), and even then only in part. Nevertheless, the impact in the oil industry was significant. After the Saticoy Field of the Ventura Basin had been discovered in 1955, the knowledge that the ‘‘turbidite reservoirs are string-sands’’ has become very helpful to production geologists. The recognition that linear sand bodies have up-dip pinch-out edges has contributed to the discovery of gas fields by the Occidental Company in the Sacramento Valley. I suspected that one of my former assistants at Ventura made the knowledge transfer; he received a generous bonus from his company after he changed employers.

2.2.

N OT ALL D EEP -SEA SANDS ARE T URBIDITES

‘‘Turbidite’’ soon became a ‘‘household word’’ in the Shell Oil Company, and the word was synonymous with ‘‘deep-sea sand’’. I did not like the word because of the genetic implication of deposition from the turbid waters of a suspension. When I was a post-doc in Switzerland, I found evidence in the Alpine flysch. The existence of groove casts as bottom markings clearly indicates that the flow depositing the laminated bottom layer (Bouma interval A) is not a turbidity current; it consists of debris in viscous motion (Hsu¨, 1959). Neither is graded bedding (also Bouma A) necessarily an evidence of settling from a suspension. The grading may have resulted from a decrease of the current-transport velocity (Kersey and Hsu¨, 1976). Finally, the cross-laminated silt (Bouma C), above the parallel laminated Bouma B interval, may or may not be a deposit from suspension. Years later my students and I did carried out experiments to show that the bedform of suspension deposits can be rippled (Hsu¨ et al., 1980). At that time, however, I was speculating on the possibility that the so-called Bouma C unit has been reworked by a bottom current. Bruce Heezen, who had acquired a fame because of his paper with Ewing on the turbidity-current deposition after the Grand Banks Event, came in 1955 to Houston to give a talk on deepsea sedimentation. I was his host. We soon found ourselves in agreement that not all deep-sea sands are turbidites. He showed me beautiful pictures of deep-sea ripples.

¨ K.J. Hsu

13

I could tell him, from my observations at Ventura, that those ripples are underlain by cross-laminated sand or silt. I had a field season, in the summer of 1955, to study the Ventura sediments. Together with Jim Valentine, who was collecting foraminifera samples for ecological studies, we climbed a steep slope covered by a dense growth of sage brush. Below a vertical cliff of a thick turbidite ledge was a mudstone deposit. Intercalated in the pelitic sediment is an ash unit which crops out everywhere in the Ventura Basin; it is the boundary ash between the Pliocene Pico Formation and the Pleistocene Santa Barbara Mudstone. The volcanic ash had settled on a deep-sea bottom, as Valentine identified the typical Uvigerina peregrina assemblage in the pelitic sediments. On a close inspection, we found that the tuff unit consists of two ash layers, with a rippled horizon between the two (Figure 2.1). Instead of a laterally continuously rippled layer, the ripples are ‘‘starved’’ micro-dune features underlain by a cross-laminated, very well-sorted coarse silt. That silt cannot be a turbidite, because the speed of a turbidity current would have disturbed or eroded away the underlying ash layer.

Figure 2.1 Deep-sea sedimentary facies. Cross-laminated silt deposited by bottom currents is found intercalated in the sediments of thin-bedded sand facies, which is found on the fringe of main turbidite sand bodies (Hsu« et al., 1980, reprinted with permission from AAPG, whose permission is required for further use).

14

Personal Reminiscences on the History of Contourites

2.3.

T URBIDOPHILES AND T URBIDOPHOBES

I wrote up the results of the Ventura Basin first as an internal report. I identified four lithofacies in the Pliocene of the Ventura Basin that represent four environments of deposition (Figure 2.2), namely: a mudstone facies deposited on the slope of the basin, a conglomerate facies in the canyons cutting across the slope, a turbidite sand facies in the deepest part of the basin trough, and a thin-bedded facies on the fringes of the turbidite sand. My interpretation of the genesis of the cross-laminated or thin-bedded sand represents little competitive value to Shell. The work was released for publication (Hsu¨, 1964) after I had presented the results at the 1963 SEPM meeting at Houston. My talk was attended by two persons, among others. Glenn Bartle was the President of the Harpur College; he came to the talk because I was being considered for a staff position at the College, or SUNY Binghamton. He did not make any comment of my interpretation, but he was very critical of my presentation. I did not hold the microphone steadily, and the variable volume of my voice was very irritating to the audience. The other person was Ken Emery, and he was furious. Emery, then teaching at the Southern California University, was a ‘‘turbidophile’’. He had convinced himself that all deep-sea sands are turbidites. He was angry because I was reinforcing the doubt of the ‘‘turbidophobes’’, and there were many of those in the oil industry. As he said, he had been working for

Tuff

Cross-laminated sands and silts

Shales

Figure 2.2 Cross-laminated silt between two ash falls.When I encountered the cross-laminated silt layer between the ash falls for the first time, I began to doubt if all deep-sea sands are turbidites (Hsu«, 1964, with permission from the Society for Sedimentary Geology).

15

¨ K.J. Hsu

(a)

(b)

Figure 2.3 Morphological similarity of cross-laminated silts (Hsu«, 1964, with permission from the Society for Sedimentary Geology). (a) Cross-laminated silt in the Pliocene, Pico Formation, California. (b) Cross-laminated silt in Holocene tidal-flat deposits of the Wadden Sea, the Netherlands.The same bedform can be present in two different environments.

more than a decade to educate the ignorant. And now I was giving them a new excuse. I was thus obstructing the progress of science, while driving a wedge between the academics and the industry. Ripples are ripples, and the same bedform can be found in different depositional environments. Deep-sea ripples are morphologically not distinguishable from ripples on tidal flats. I gave a comparison of the sedimentary structures of the deposits from the two very different environments (Figure 2.3). Emery was right, my after-dinner talk at the St Louis SEPM was a hit and gave much comfort to the ‘‘turbidophobes’’ of the oil industry. I was also right, not all deep-sea sands are turbidites. I learned then that Heezen was continuing to develop his idea on deep-sea sand deposition. Eventually, the rippled and cross-laminated sands and/or silts are called ‘‘contourites’’, because they have been deposited by deep-sea currents flowing parallel to submarine contours.

2.4.

EROSIONAL UNCONFORMITY M ISINTERPRETED

I left Shell in 1963. While I continued to be interested in sedimentology, I became almost a full-fledged geological oceanographer after I joined the JOIDES Deep Sea Drilling Project. As the Chairman of the South Atlantic Group of the Paleoenvironment Panel, I was an avid reader of reports and cruise proposals on deepsea circulations, and became acquainted with the latest investigations on the Antarctic Bottom Current (AABW) and the North Atlantic Deep Water (NADW). Particularly interesting were the results of drillings on the West African Margin. The power of the NADW caused deep erosion of the slope sediments. In places, Middle Miocene hemipelagic deposits of the South Atlantic overlie slope deposits as old as the Cretaceous.

16

Personal Reminiscences on the History of Contourites

When I was, as they called it, ‘‘on the beach’’, I taught Alpine tectonics at the Swiss Federal Institute of Technology. I encountered in the geology of the Alps a century-old puzzle that had been called ‘‘Wang Transgression’’. Unlike the common transgressive deposits of sand or gravel, the Wang is a Maastrichtian shale formation, containing a deep-sea fauna indicative of marine deposition on a Cretaceous slope. The relation was considered transgressive, because the Maastrichtian formation lies directly upon Campanian pelagic deposits in the North Helvetic palaeogeographic realm. The Wang sediments of the higher ultrahelvetic nappes were deposited farther offshore; they overlie unconformably progressive older (Santonian, Turonian, Cenomanian, Early Cretaceous, Late and Middle Jurassic) hemipelagic deposits (Figure 2.4). The orthodox interpretation postulates uplift, followed by subaerial erosion, followed by subsidence, and finally followed by a shallow-marine transgression. There is in fact no evidence of uplift, of subaerial erosion, of subsidence, or of a shallow-marine transgression. Using the modern West African Margin as an analogue, we see the obvious fact that the Wang ‘‘transgression’’ was not a transgression: it is an overlap of Maastrichtian contourites above older slope sediments. The unconformity under the Wang Formation signifies a period of very active submarine erosion by contour currents in the Helvetic realm of an Alpine Basin during the pre-Maastrichtian. This erosional progress is comparable to that on the West African Margin by the NADW during the pre-Middle Miocene. An American student, S. Diefenbach, completed a master thesis on the ‘‘Wang Transgression’’. She found clear evidence that the North Helvetic realm was a marine slope environment far south of the European continental coast. The gradient was steep enough to have caused widespread slumping of hemipelagic deposits, and the overshore slope was cut by submarine gullies or canyons in which coarse clastics were accumulated. Helvetic

Ultrahelvetic Internal Prealps Plaine-Morte Decke

Wildhorn Decke

Wang Leimer Habkern Schurfling Schuppen Wildflysch

External Prealps

Tothorn Laubhorn Decke Decke Leissigen Schlieren Flysch Flysch Habkern

Gurnigel Flysch

swell

Sea level

Priabonian

on

Luteti

SH Swell

Wang beds Amdener Schichten Turonian

n

Wang

mania

c

Malm Dogger nian rias Aale T

granite

oi oz

Lowe

o Pale beds

Lu tet ian Yp re Pa sion leo ce ne es

us

ceo r Creta

Habkern e cen

M

Ceno

n

esio

Ypr

Figure 2.4 The Wang ‘‘Transgression’’ (Hsu«, 1960, with permission from the Geological Society of America). A reconstruction of the paleogeographic relations suggests that the MaastrichtianWang Formation consists of transgressive deposits overlying sub-aerially eroded older formations. This classic interpretation is wrong. The Wang Formation consists of contourite beds.

¨ K.J. Hsu

17

The Wang is clearly a contourite formation. The interpretation was not published until my book The Geology of Switzerland was printed after my retirement (Hsu¨, 1995). Meanwhile, I committed an indiscretion when I taught the idea in my class on The Geology of Switzerland. For that, students and colleagues alike chastised me. I was told that I should not have taught Ken Hsu¨’s crazy ideas to beginning students. I often wonder, if Isaac Newton was similarly reproached when he taught gravity to his Cambridge students before the publication of the Principia. If the academic establishment of the 17th-century England had been as dogmatic as the Swiss geological community of the 20th century, Newton would probably have had to tell his students, contrary to his conviction, that the Sun went around the Earth. I recalled that one of my students had to delete the word ‘‘me´lange’’ when he intended to publish, in the early 1970s, his thesis on the Wildflysch in the Eclogae Geologicae Helvetiae. Now that a book on contourites is published, I am hopeful that the editors of that illustrious journal would permit the use of the word ‘‘contourite’’ in articles on Alpine palaeoceanography.

C H A P T E R

3

M ETHODS FOR C ONTOURITE R ESEARCH J.A. Howe Scottish Association for Marine Science & UHI Millennium Institute-Dunstaffnage Marine Laboratory, Oban, Argyll, Scotland, UK

Contents 3.1. Introduction 3.2. Oceanographic Measurements 3.3. Geophysical Methods 3.4. Sampling Strategies 3.5. Analytical Methods 3.6. Onshore Studies of Ancient Sequences 3.7. Summary of Multidisciplinary Techniques Acknowledgements

3.1.

19 20 21 27 29 31 31 33

INTRODUCTION

The physical effect of a persistent bottom current on deep-sea sediment can be studied using a wide variety of oceanographic, geophysical and sedimentological techniques. Contourite workers need to be able to study both the modern and palaeo-deep-sea floor. A definition of the term ‘‘contourite’’ is provided by Rebesco et al. (2008); this chapter aims to present a summary of the range of techniques that can be employed to investigate contourite sedimentation in the deep sea. The range of processes contributing to current-influenced sedimentation in the deep sea can be extremely diverse, and at a variety of scales from the localised to the global, such as biogenic fluxes from the water column, changes in ocean chemistry, the localised dynamics of the benthic boundary layer and the variation of thermohaline flow in response to bathymetry. Since Wu¨st (1936) first proposed the idea of thermohaline flow in the deep oceans, bottom current and perhaps more specifically, contourite researchers have needed to take a broad, multi-disciplinary approach when examining the deep sea bed for evidence of contourites. Early workers have, understandably, been restricted by the variety of techniques available to them, and used standard approaches such as locating rippled sands at abyssal depths in deep-sea photographs (Hollister and Elder, 1969; Heezen and Hollister, Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00203-3

 2008 Elsevier B.V. All rights reserved.

19

20

Methods for Contourite Research

1971) or identifying regions of enhanced sedimentation on a margin from seismicreflection profiles (Hollister et al., 1978) as evidence for persistent along-slope bottom currents. Modern contourite researchers have considerably more in their investigative arsenal, including multi-beam bathymetry, side-scan sonar, high-resolution seismicreflection profiles, as well as the ability of recovering longer, less disturbed samples from drifts as coring technologies have developed. Progress has not simply been restricted to the collection of data from deep-water basins and margins: once ashore, core samples are increasingly subjected to ever more detailed sedimentological and geochemical examination. This drive for very detailed, high-resolution core data has resulted in some spectacular discoveries that have considerably increased our knowledge of how significant global thermohaline flow is to climate and, more importantly, how changes in global thermohaline flow can influence climate change. Most, if not all, of these discoveries have arisen from detailed multidisciplinary work on regions of contourite sedimentation in the deep sea.

3.2.

O CEANOGRAPHIC MEASUREMENTS

Oceanographic measurements of modern, persistent thermohaline or winddriven flow in the deep sea are the most fundamental data needed when beginning to investigate bottom-current deposits (see also Salon et al., 2008 and Zenk, 2008). For reasons of cost, only major drilling and coring projects can undertake to obtain new, detailed current-meter data over drift sites. Ideally, two cruises are needed, one to deploy the current meters and one to recover them, usually after enough time has elapsed to measure the hydrographic conditions at the site (e.g. 6–12 months later). Typically, contourite workers have made use of pre-existing hydrographic data sets, which are not always ideally located on major drift sites and collected for other purposes. One of the earliest and certainly one of the most important multidisciplinary contourite studies showed how the deep sea can experience high-energy bottom-current events or ‘‘benthic storms’’. These storms are characterised by short-duration flow reversal and enhancement, produced in regions of high eddy kinetic energy. The HEBBLE (High Energy Benthic Boundary Layer Experiment) site (Nowell et al., 1982; McCave et al., 2002) on the Nova Scotia Rise identified these events through the combined use of current measurements, sea-floor photography as well as high-resolution seismics and coring. Studies such as the Project Mudwaves in the Argentine Basin (Manley and Flood, 1993a; Flood et al., 1993) also used direct physical measurements from current meters placed across a wave crest, flank and trough to deduce the bottomcurrent velocities and structure across the features. Measurements such as these can lead to mathematical modelling of bottom-current flow across waves (Hopfauf et al., 2001). Camerlenghi et al. (1997) used two current-meter moorings across a giant sediment drift west of the Antarctic Peninsula to ascertain present-day bottom-current conditions. In an extensive study from

J.A. Howe

21

the Campos Basin of the southwestern Brazilian Margin, Viana et al. (1998a) utilised current-meter moorings, wave measurements and infrared satellite images to investigate the origin of contouritic sand deposits on the margin. Hydrological processes considered in this study included the influence of surface currents, counter-currents, waves, tide and eddies resulting in the offshore transport of contourite sand by a combination of shelf spillover and internal waves. The direct physical measurement of bottom currents is of vital importance in relating the modern benthic conditions to ancestral bottom-current flows, inferred from core-sediment texture. Another significant activity that can be carried out in the water column above sites of bottom-current influence is the collection of sediment from the water column, including from the nepheloid layer. Such data are unusual but very helpful. It is notable from high-latitude studies where significant sediment supply can originate from the surface through iceberg rain-out. The presence of a nepheloid layer can be a useful indicator of near-sea-floor current activity. These layers of suspended sediment can be several hundred metres thick and consist of material derived from both primary productivity from the surface and in situ resuspension. These measurements can provide information on the sediment flux to a site, either by along-slope transport or directly from biogenic productivity from the overlying water masses. A simple sediment trap on a current-meter mooring can collect sediments for a prolonged period of time, typically months. The sediment record can show seasonal variations not usually detectable in core section. Some sediment traps utilise a sequential carousal of collecting bottles, permitting samples to be obtained over a seasonal cycle, such as pre-planktonic bloom, bloom and postbloom. Water samples collected during conductivity, temperature and depth (CTD) casts can reveal the nature, and – if combined with transmissometer data – the extent of suspended sediment in the nepheloid layer. Pudsey (1992) and later Gilbert et al. (1998) used transmissivity profiles collected during CTD transects in the northwestern Weddell Sea to identify an active nepheloid layer present on the rise, the product of the along-slope transport of Weddell Sea Bottom Water. In the North Atlantic, across the Bjorn and Gardar Drifts, Bianchi and McCave (2000) collected water samples to examine suspended particulate material and transmissometer readings to determine the thickness of the benthic nepheloid layer. These data were combined with pre-existing current-meter data from the region and used to calibrate the palaeo-records of bottom-current velocities from the cores, resulting in a very detailed record of the flow of Iceland–Scotland Overflow Water moving as a Deep Western Boundary Current in the Iceland Basin across this major drift complex.

3.3.

GEOPHYSICAL METHODS

In any multidisciplinary study of bottom-current-influenced sedimentation in the deep sea, geophysical – specifically acoustic – surveys play a key role (see also Nielsen et al., 2008). For simplicity, acoustic surveying is here divided into

22

Methods for Contourite Research

two main techniques: sea-floor mapping (for surface expression and morphology) and seismic-reflection profiling (for internal acoustic character, geometry and hence interpretations of the depositional environment). Some inferences can also be made on the relative ages of the sediments (seismostratigraphy). It could be argued that it is only through the use of seismic-reflection profiling and echo sounders surveys that the basin-scale extent and nature of bottomcurrent sedimentation has been recognised (Johnson and Schneider, 1969; Hollister et al., 1978). A wide range of acoustic tools are now available that can penetrate the sediment cover, depending on the frequency used. Frequencies around 12 kHz and above provide little sediment penetration and are therefore used for surface mapping, 3.5 kHz profiling systems can provide penetration up to 100 m below sea floor and are suited to detailed process-related studies, and, finally, 10–200 Hz airgun seismic sources give very deep penetration (km) and are much more suited to deep geological studies of ancient drifts and basement. Even at its most basic, sea-floor mapping purely for morphology can provide one of the most graphic illustrations of the dynamic nature of bottom-current systems available. Through detailed bathymetric surveying, it is possible to acoustically image the sea floor where sediments have become moulded into drifts and waves or sculpted into moats around obstructions. Even with basic side-scan sonar instruments, some information on sediment morphology, grain size and sea-floor roughness can be obtained (Figure 3.1). More advanced, towed side-scan systems such as TOBI (Towed Ocean Bottom Instrument) and GLORIA (Geological Long-Range Inclined Asdic) have provided stunning acoustic images of contourite systems from most of the major ocean basins. The modern development of multibeam echo sounders has revolutionised mapping of the sea floor. These systems provide wide coverage (typically in a water depth of 3000 m the ‘‘footprint’’ could be as wide as 12 km). Regional sediment processes can be identified from the morphology of the sea floor. At their best, multibeam systems can easily be used to identify drifts, moats and waves characteristic of bottom-current-influenced sedimentation (Figure 3.2a–c). Using an echo sounder can provide a wealth of information on sea-floor morphology, acoustic character and even some limited sediment thicknesses (isopach maps) in the surface sediments. Sea-floor mapping using 3.5 kHz echo sounders or, more recently, with modern parametric sub-bottom profilers (such as ‘‘TOPAS’’ and ‘‘Parasound’’) has the added advantage of being able to cover large areas of sea floor quickly and cheaply (Figure 3.3a and b). This technique has been used on numerous occasions in contourite studies. Initially pioneered by Damuth (1978, 1980), this technique classifies the echo type based on sea-floor microtopography, sub-bottom penetration and reflectors. The use of 3.5 kHz survey data has extended the range and scope of contourite study, enabling workers to occasionally observe a basin-wide, more regional picture of sedimentation. Examples of this technique can be found in Pudsey and Howe (1998), where a simple surface acoustic-character map was produced for the Central Scotia Sea using data sets from numerous cruises from the mid-1980s to the present day.

J.A. Howe

23

Water column

Ship track/ transmit pulse

First bottom return

10 m

N

Figure 3.1 Side-scan sonar image of bottom-current-reworked sands and outcropping rocks from approximately 500 m water depth southeast of the Falkland Islands, South Atlantic (Image from the British Antarctic Survey cruise 29 on RRS James Clark Ross).

Seismic-reflection profiling (Figures 3.4a, b and 3.5) has provided the basis for a number of classification schemes on the various types of drift morphology and internal configuration (McCave and Tucholke, 1986; Fauge`res et al., 1999; Rebesco and Stow, 2001). Regional seismic-reflection profiles can also be used to ‘‘backstrip’’ or map laterally continuous reflectors or mega sequences and produce models of palaeo-bathymetry (and hence palaeo-current pathways) and sediment thicknesses. Wold (1994) used this technique highly effectively producing models of North Atlantic Drift development and hence sediment fluxes during the Cenozoic. This technique is typically underused in contourite research, due to the high density of seismic profiles needed across a drift system. In addition, the reflectors within the modelled drifts need to be related to some direct age constraint, typically from a drill site or long piston core.

24

Methods for Contourite Research

(a)

Minor drifts Sediment drift

Wave field

20 km

Wave field Parasitic cones

Terraces

Concave slide scar

Moat Water depth (m)

(b)

800.00

1500.00

Sediment drift

5 km

2000.

Slide scar

Sediment waves

Parasitic cones

Moat

Terrace Slide scar

Slide Scar

(c)

63°N

20 km Sediment wave field

60°N

N

57°N

54°N

51°N

Sediment drift Multiple moat Sediment wave field

For caption see next page.

25°W

Moat –2300 m

20°W

15°W

10°W

5°W

0°W

J.A. Howe

25

Figure 3.2 Multibeam bathymetry of the Rosemary Bank Seamount, North Atlantic. (a) Shaded relief (from the northeast) and colour contoured multibeam bathymetry (Kongsberg EM120 12 kHz, 1°  1° beams, collected during cruise 99 of the RRS James Clark Ross) of the Rosemary Bank Seamount, North Atlantic Ocean. (b) Detailed view of the western spur region of Rosemary Bank showing the sediment wave-field, multiple moat and drift complex. Also indicated is a concave slide scar on the southern flank. (c) Perspective view of the seamount viewed from the northwest. Inset map shows the location of the seamount in the northern Rockall Trough (from Howe et al., 2006, with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

(a)

(b)

Figure 3.3 TOPAS sub-bottom profiles from the Fram Strait. (a) TOPAS sub-bottom profile from the western Svalbard Margin hemipelagic and contouritic sediments interrupted by debris flows and diapirs from the Storfjorden Trough Mouth Fan. (b) Small, current-controlled sediment drift developed on the Vestnesa Ridge, western Svalbard Margin, a region associated with abundant gas-escape-related pockmarks (adapted from Howe et al., 2008). Inset map shows the location of the figures on the western Svalbard Margin. (Profiles collected on the Scottish Association for Marine Science Arctic cruise 75 on RRS James Clark Ross.) A multicolour version of this figure is on the enclosed CD-ROM.

26

Methods for Contourite Research

SW

(a)

NE Northern Rockall Trough

Distal edge of Sula Sgeir Fan

Subsidiary drift

Elongate drift

Wyville-Thomson Ridge

Moat-related drift

0.8

Plio–pleistocene 1.0 Miocene 1.2 Palaeogene BGS 83/04-60

1.5 km NE Hebrides Slope

Multi-crested sediment drift

(b)

Elongate drift

SW

Northern Rockall Trough

TWT (s)

0.8 TWT (s)

Elongate drift

Elongate drift

1.0

1.2

BGS 84/06-24

Plio–pleistocene

1.4

Miocene

1.5 km

Palaeogene

Figure 3.4 Seismic-reflection profiles from the northern Rockall Trough, North Atlantic. (a) British Geological Survey single-channel airgun seismic-reflection profile 83/04 -60 illustrating the development of elongate, subsidiary and moat-related drifts adjacent to the Wyville-Thomson Ridge, northern Rockall Trough (adapted from Howe et al., 1994). (b) British Geological Survey sparker seismic-reflection profile 84/06 -24 showing a complex multicrested elongate sediment drift developed parallel to the Hebrides Slope, northern Rockall Trough (adapted from Howe et al., 2002).

TWT Northern s 0.5 NW 1.0 1.5 2.0 2.5 3.0 3.5

Rockall Trough

(b)

Rosemary Bank

Northern Rockall Trough SE

(c)

SBM SBM

(d)

SBM

(a) BGS 03/03/1

s 0.5 1.0 1.5 2.0 2.5 3.0 3.5

5 km

Figure 3.5 Major seismostratigraphic units occurring in the drift sequences surrounding the Rosemary Bank Seamount. Megasequences: RPa = Pliocene ^ Holocene; RPb = midMiocene to early Pliocene; RPc = late Eocene to early Miocene. Reflectors: C10 = early Pliocene angular unconformity; TPu = late Oligocene to mid-Miocene; C20 = late-early to early-mid-Miocene; C30 = late Eocene unconformity (after Stoker et al., 2001). TL =Top of Late Cretaceous ^ early Paleogene lavas; SBM = sea bed multiple (adapted from Howe et al., 2006). The location of the Rosemary Bank Seamount is shown in Figure 3.2d (from Howe et al., 2006, with permission from Elsevier). (a) British Geological Survey seismic-reflection profile 03/03/1 (with locations of insets b ^ d). The profile displays the distinctive moat-drift association at the base of the seamount produced by enhanced bottom-current flow.

J.A. Howe

27

(b)

Rosemary Bank Seamount

TL

Elongate drift Cut-and-fill

Onlap RPa

Drape of post-C10

RPb

Moat

0.5 s C10

?C20 RPc C30

3 km Eocene

Neogene drift

(c)

Eocene prograding wedge with Neogene veneer

Rosemary Bank Seamount

TL

0.5 s

3 km SBM

TL

? debris flow uncertain age TL

Rosemary Bank Seamount

Elongate drift RPa C10

Moat

(d)

RPb TPU

0.5 s

RPc C30 3 km TL

Eocene

Figure 3.5 (Continued) (b) Detail of drift ^ moat association showing the main mid-Miocene ^ Pliocene age construction of the drifts. (c) Eocene age prograding wedge overlain by Neogene drift sediments on the flanks of the seamount. (d) Deeply incised moat at the base of the seamount produced by sustained, enhance bottom-current flow.

3.4.

SAMPLING STRATEGIES

In most multi-disciplinary studies of contourite systems, after the oceanographic and geophysical surveys, it will usually be essential to sample the sea bed. The collection of any undisturbed contourite sediment samples from the sea bed presents no more problems than in any other deep-sea sediment facies. It is worth noting, however, the

28

Methods for Contourite Research

local variation in facies across drift complexes, not always appreciated from conventional geophysical data. To examine small-scale depositional changes, the use of sea-floor photography or video transects is very useful (Figure 3.6a–f ). The use of sea-floor photography is not a new technique: indeed, it was one of the first, and most widely used to examine the deep sea. As described earlier, workers such as Heezen and Hollister (a)

(b)

(c)

(d)

(e)

(f)

Figure 3.6 Sea-floor photographs showing contouritic sediments from the North Atlantic and Arctic Oceans (all images courtesyof David Hughes, the Scottish Association for Marine Science). (a) Sea-floor photograph of inclined glass sponges bending under the influence of Norwegian Sea Deep Water flowing west along the Faeroe ^Shetland Channel, north of the Wyville-Thomson Ridge, 1100 m. (b) Rippled, bioturbated contourite sandy muds in the northern Rockall Trough, south of theWyville-Thomson Ridge,1050 m. (c) Glaciomarine hemipelagites under seasonal sea ice on the Yermak Plateau, northern Fram Strait, 803 m. (d) Rippled contourite silty sands from 700 m on the Hebrides Slope, northern Rockall Trough. (e) Gravelly and sandy contourites from 1000 m on the Hebrides Slope, northern RockallTrough. (f ) Rippled silty/sandy contourites with emerging clasts from 1000 m on the Hebrides Slope, northern Rockall Trough. A multicolour version of this figure is on the enclosed CD-ROM.

J.A. Howe

29

(1971) and Heezen et al. (1966) made extensive collections of sea-floor photographs from the ocean basins and showed, visually, that the deep-sea floor was not a wholly tranquil place but was swept by persistent bottom currents. Photographs across drift sites were obtained by McCave (1982), Scha¨fer and Asprey (1982) and Carter and Schafer (1983) from the Orphan Knoll, Labrador Slope. These showed that rippled sands and silts are predominant where current velocities are highest (e.g. at drift margins and moats), with mud deposited marginally to the axis of the flow (such as the drift crest), in this case the Western Boundary Undercurrent. A wide variety of sampling, coring and drilling devices are available to collect sea-floor sediments, but not all of these are suited to contourite studies. Simple grabs tend to provide a homogenised sample, destroying the fabrics and structures that are invaluable to contourite interpretation. Short corers such as multiple and mega corers as well as Sholkovitch, Craib and box corers are useful in that they provide undisturbed samples from the sediment/water interface, but – being short (under 0.5 m) – can lack the temporal range needed to examine any change in the bottom-current system. Longer soft-sediment corers such as Kasten, gravity, vibrocore and piston corers are much more useful in providing longer records from drift sites, and are also heavy enough to sample some sands that may be present. Very long coring systems are now available, such as the spectacular Giant Piston Corer, the Advanced Piston Corer and the French STACOR (stationary piston corer) and Calypso corer, which can routinely collect undisturbed core samples of over 30 m. Problems associated with these corers have been highlighted by Skinner and McCave (2003) as including ‘‘over-sampling’’ (due to cable rebound) and ‘‘under-sampling’’, whereby the basal or mid-core sections are deformed or lost. These problems aside, the techniques of piston and gravity coring have provided a wealth of short-term, high-resolution palaeo-oceanographic records on bottom-current variability from drift sites across the world. Longer drilling projects have been, thus far, the domain of the original Deep Sea Drilling Project (DSDP), Ocean Drilling Program (ODP) and its present incarnation, the Integrated Ocean Drilling Program (IODP). This program has routinely drilled drift sites over the past few decades for the geological record of bottomcurrent activity with an age range of Quaternary–Cretaceous and extending from the high-latitude Arctic Ocean and Southern Ocean to the equatorial Atlantic: see Stow et al. (1998a) and Rebesco (2005) for a review.

3.5.

ANALYTICAL M ETHODS

Contourite facies collected in core samples are recognised by the criteria of structure, texture, fabric, composition and their sequence arrangement (see also Stow and Fauge`res, 2008). Extensive work has been conducted on contourite facies, notably by Stow (1979), Fauge`res et al. (1984) and Gonthier et al. (1984); it is summarised in Stow et al. (2002c). A wide variety of contourite facies have been recognised, from the initial classification of muddy, silty–muddy and sandy proposed by Stow and Holbrook (1984). Contouritic gravels, volcaniclastic, calcareous, siliceous and even manganiferous facies have now been described (Stow et al., 1998a). Identifying these facies in core samples can be difficult and workers have gone

30

Methods for Contourite Research

to considerable lengths to identify useful criteria. Meticulous core logging is necessary, with particular care needed when describing lithological contacts, grain size, structure, level of bioturbation and sequence relationships. The ability to differentiate between down-slope turbidite and along-slope contourite deposition was considered by Stow (1979) and Stow and Lovell (1979). A multi-disciplinary approach is required with a combination of morphological evidence (i.e. are there features on the sea floor indicative of a depositional pathway, such as furrows, moats and waves, visible on acoustic data?) and sedimentological information (sediment fabric and structure). Typically, contourites result from persistent or semi-persistent flows as opposed to the catastrophic down-slope movement of turbidites. Contourites therefore might posses irregular grading and sharp contacts, heavy-mineral fabric and lack the reworked shallow-water microfossils of turbidites. Interbedded turbidites and contourites as well as contourites derived from the reworking of turbidites remain difficult to distinguish, although combined sedimentological and morphological (including sub-bottom profiles) approaches should provide evidence of local scale (sedimentological) and regional scale (morphological) along-slope pathways. Flood et al. (1985) examined the magnetic fabrics of sediments from the HEBBLE site to determine along-slope fabrics in muddy contourites. Other techniques useful for contourite recognition include X-radiography of cores for internal structure, ice-rafted debris and bioturbation (Figure 3.7a–c), grain-size analysis and an examination of any microfaunal assemblages. (a)

(b)

Silty lamination

(c) D/St

Glaciogenic contourites

Silty lamination Sandy turbidites

Sandy–gravelly contourites

Glaciogenic contourites X-radiograph positive BGS core 59-08/41 Upper Hebrides Slope 470 m 0.35–0.70 m

X-radiograph negative SAMS core 067 From Strait Western Svalbard Margin Fram Strait 1226 m 1.70–2.00 m

X-radiograph positive BGS core 58-14/34 Northwestern Rockall Trough 1486 m 2.80–3.00 m

Figure 3.7 X-radiographs of contourite facies. (a) X-radiograph positive image of silty lamination in bioturbated muddy contourites from the upper Hebrides Slope (British Geological Survey core 59-08/41). (b) Negative X-radiograph image of sandy turbidites interbedded with glaciogenic contourites from the western Svalbard Margin, Fram Strait. D/St indicates dropstone clast (Scottish Association for Marine Science core 067). (c) Positive X-radiograph image of sandy/ gravellycontourites fromthe northern RockallTrough (British Geological Surveycore 58-14/38).

J.A. Howe

31

Recent developments include the use of non-destructive, pass-through core-logging systems; using these, a great deal of information can be obtained whilst at sea, such as magnetic susceptibility and other physical properties such as bulk density, acoustic velocity and natural gamma-ray emissions. More elaborate onshore logging systems use X-ray fluorescence to determine bulk geochemistry, which can be useful in examining changing sediment sources and bottom-water chemistry. The most useful sedimentological technique is grain-size analysis. The use of grain size, and its link to bottom-current flow was first demonstrated by Ellwood and Ledbetter (1977) from the Vema Channel, South Atlantic. This study demonstrated the direct relationship between subtle changes in silt grain size and bottomcurrent velocities (in this case, Antarctic Bottom Water). McCave et al. (1995a) took this technique a step further, identifying the ‘‘sortable silt’’ interval; this is the mean of the terrigenous silts between 10 and 63 mm, which are most susceptible to transport by bottom currents. Through a meticulous analysis of this sediment fraction, subtle bottom-current variations can be inferred. A number of important palaeo-oceanographic studies have been undertaken using this technique and have revealed the wider, climatic signals preserved in fine-grained contouritic sediments. Geochemical proxies are becoming more widely used, notably the use of rare-earth elements and isotopes, particularly Nd with Sr and Pb, when combined with grain size as a tracer of palaeo-bottom-current velocities (Revel et al., 1996). Other work has used stable isotopes such as 234Th, 210Pb and 210Po for particle cycling (Murray et al., 2005) and 18O and 13C to develop a detailed stratigraphy of bottom-current variations (Rasmussen et al., 2002).

3.6.

ONSHORE STUDIES OF ANCIENT SEQUENCES

Studies from outcrop as well as from boreholes of fossil contourite sequences are much rarer than the offshore, modern counterparts (see also Hu¨neke and Stow, 2008). Stow et al. (1998a, 2002c) summarise one of the main problems as being confusion with fine-grained turbidites. Many of the claimed contourite cases originate from reworked turbidite successions. The recognition of fossil contourites uses a three-stage approach: (1) small-scale studies (from field, laboratory or borehole evidence); (2) regional (drift, formation and region) trends in facies; and (3) fitting the conclusions from (1) and (2) with any independent palaeooceanographic reconstruction of the region showing a deep-water, persistent bottom-current pathway in the geological past.

3.7.

S UMMARY OF MULTIDISCIPLINARY TECHNIQUES

Contouritic sedimentation remains somewhat enigmatic, although it is becoming clear that bottom-current-influenced sedimentation is a highly significant process in the deep ocean. Outlined in this chapter are a number of approaches to investigating along-slope processes, at a number of scales, using hydrographic,

32

Methods for Contourite Research

geophysical and sedimentological and outcrop studies (Table 3.1). Developments continue, especially in the areas of geophysics and palaeo-oceanography using contourites. New high-resolution bathymetric surveys are now being routinely collected from drift and wave sites around the world and new proxies for palaeo-ocean chemistry and bottom-current pathways investigated. Deep-ocean technologies and instruments are being developed which are smaller and cheaper, and hence easier to use on research cruises. Satellite data are now more widely available, showing the distribution of surface currents and sea-surface heights, enabling regions of enhanced bottom-current activity to be better understood. Table 3.1 The recognition of contouritic sedimentation in the deep sea and investigative techniques used Scale Small-scale: field, borehole and laboratory • Non-turbiditic characteristics or origin • Mixed pelagite/hemipelagite setting with strong evidence for bottom-current sedimentation • Any cyclicity is related to bottomcurrent velocities, not to terrigenous input or biogenic productivity

Methods

• • • • • • • •

Medium-scale: drift, formation or region • Regional trends in facies, and current/ palaeo-current directions, texture, mineraological or geochemical indicators • Unconformities, condensed sequences, regional variation in thickness, drift geometry present • Shape and geometry of the sediment body, indicating an along-slope trend • Contouritic environment/palaeo-environment, including accumulation rates and in situ faunal assemblages Large-scale: system, ocean or continental margin • A bottom-current-influenced system from environmental/palaeo-environmental evidence • Modern and ancestral bottom-current system

Source: Adapted from Stow et al. (2002c).

• • • •

Detailed core logging Facies analysis Gravity, kasten or box coring Physical properties Grain-size analysis Extensive site survey data (multi-beam bathymetry, sub-bottom profiles, sea bed photography, side-scan sonar) X-radiographs hydrographic survey (current moorings, CTD) Seismic-reflection profiling Multibeam bathymetry Sea bed video/photography Onshore section logging/wireline logging

• Hydrographic surveys • Regional multi-beam bathymetric mapping

• Satellite imagery • Seismic-reflection profiling • Drilling/long-piston-coring investigations

J.A. Howe

33

As techniques for investigating contourites are continuing to be developed, this forward progress is given the added impetus of the knowledge that contourites may, in the future, provide significant discoveries in the fields of petroleum exploration, thermohaline circulation and climate studies.

ACKNOWLEDGEMENTS This review could not have been completed without the invaluable help of Peter Morris of the British Antarctic Survey for advice on the geophysics and multi-beam, and Martyn Stoker of the British Geological Survey for allowing access to the Rockall Trough seismic-reflection profiles. The text was greatly improved by the critical reviews of Rob Larter of the British Antarctic Survey and David Long of the British Geological Survey.

P A R T

2

BOTTOM CURRENTS

C H A P T E R

4

A BYSSAL AND C ONTOUR C URRENTS W. Zenk Leibniz Institute of Marine Sciences at the University of Kiel (IFM-GEOMAR), Kiel, Germany

Contents 4.1. Introduction 4.2. Abyssal Currents in the Global Thermohaline Circulation 4.3. Contour Currents 4.3.1. Physical properties and modes 4.3.2. Entrainment and modeled contour currents 4.4. Conclusion and Summary Acknowledgments

4.1.

37 40 42 42 51 55 57

INTRODUCTION

In marine geography, the large temperature difference at low latitudes of over 20°C between the sea surface and the bottom of the ocean remained a mystery for more than two centuries (cf. Schiermeier, 2006). Although in the tropics the ocean’s surface is exposed to its strongest solar irradiance, temperatures exceeding 20°C are found only in a thin top layer of some tens of meters. Roughly in the upper kilometer, temperatures decrease to about 5°C. This uppermost part of the oceanic water column is called the thermocline or, occasionally, the warm water sphere. Its well-defined upper interface is in permanent exchange with the atmospheric boundary layer, enabling free exchange of heat, mass, and energy. In contrast to the warm water sphere, the cold water sphere below is characterized by minimal vertical temperature gradients. At depths >2000 m, temperatures are usually <4°C. At moderate and lower latitudes of both hemispheres, characteristic abyssal temperatures beneath 4000 m depth lie in the range 0–2.2°C. Such a simple classification as the two-layered thermal ocean fails for ocean currents. One often associates high surface currents with vigorous wind stress upon the ocean surface or with strong tidal effects. Slow and sluggish motions are supposed to be typical for the abyss (Zenk, 2001). However, more detailed studies show a truly wide spectrum of waves and current fluctuations, virtually independent of their vertical distance from the sea surface. Generally, slow drifts in the millimeter per second range are masked by turbulence elements and Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00204-5

Ó 2008 Published by Elsevier B.V.

37

38

Abyssal and Contour Currents

Deflected organisms

Scouring deflected organisms

Lineations

Sediment tails

Current direction

Ripples

Plane bed

Anti-dunes

Bare rock with pockets of sand and gravel

100 cm s–1

Figure 4.1 Series of visual seascapes as a function of increasing current flowing from left to right (after Heezen and Hollister, 1971; with permission fromWHOI,Woods Hole, MA, USA).

eddies. Extremely high values ( >1 m s1) are rare. They may occur at the surface as well as at depth. In a series of generalized bottom sketches (Figure 4.1), Heezen and Hollister (1971) demonstrate the impact of increasing abyssal currents on the underwater seascape. It reaches from marine organisms gently deflected by a few millimeters per second current to bare rocks with pockets of gravel caused by very strong bottom currents ( >1 m s1). We note that such quantitative descriptions of the strength of ocean currents represent an integral over a rather unspecified time interval. Wind-driven currents at the surface are subject to varying weather and climatic conditions. At depth, however, the three-dimensional shape of bottom contours is one of the most prominent factors that influence the direction and the strength of internal currents. They can be forced by long-wave motion (tides, internal waves, seiches) or by horizontal density differences (gradient currents). Topographic control of oceanic currents in deep passages and straits defines the abyssal motion in substantial parts of the world ocean. Whitehead (1998) distinguishes between unidirectional currents over saddle points between neighboring deep ocean basins, depending on the source of the densest bottom water, and bidirectional currents in ocean straits with exchange of water masses between oceans and marginal seas. Marginal seas are defined as basins isolated by topography from the rest of the oceans. In Figures 4.2 and 4.A1 we show selected subsurface passages and straits that subdivide basins of the Atlantic Ocean and the Mediterranean Sea, principally allowing over- and outflows of bottom waters between them. In the long run, the deep drain of a semienclosed source basin is balanced by a refill with strongly cooled water masses from the surface. The latter phenomenon is called ‘‘convection’’ and is generally restricted to relatively small but highly effective regions of the oceans primarily located at polar latitudes. In extreme cases, observation of this process has revealed downwelling velocities of up to 0.1 m s1 (Marshall and Schott, 1999). The onset of convection is very sensitive to surface salinity. The deep cold North Atlantic limb of the thermohaline circulation (THC), occasionally called the ‘‘conveyor belt’’ (Broecker, 1991), is directed toward the equator. Farther to the south, 1

Equivalent color figures for the Indian and Pacific Oceans are given in Figure 4.A_(2b) and 4.A_(2c). Figures identified by letters are found in the enclosed CD-ROM (Compact Disk-Read Only Memory) along with the pertinent caption.

39

° De

40

o

n

ar –F

d

lan

Ice

ait ssa Str e Pa nel oe

nk

60

60

ge

rk

a nm

°

W. Zenk

a Ch

Ba

r Fa

°

40

°

Strait of Gibraltar

Discovery Gap

20°

it ra

20°

St

W ind An ward eg Pa ad s Pa a– J sage ss un ag gf e ern

Charlie Gibbs FZ

of Si c il y

Vema Gap

ara

Ce

0°

nche

Roma

0°

FZ

lain al P

yss Ab

20°

°

40

°

0°

30

°

°

60

°

90

40

60

°

South Sandwich Island Arc Gap

Sh a Pa g Ro ss ck ag s e

V Ch ema an ne l

20°

Figure 4.2 Representative selection of deep-ocean passages in the Atlantic Ocean and the Mediterranean Sea (courtesy Dr J.Whitehead,WHOI,Woods Hole, MA, USA).Water exchange across sills is called ‘‘overflow.’’ Straits show out- and inflows separated by a variable shear zone between the surface and the bottom. Light gray indicates depths <4000 m, dark regions mark basins with depth >5000 m. A multicolor version of this figure is on the enclosed CD-ROM.

the Southern Ocean with connections to all three oceans enables a global distribution of abyssal currents on a secular timescale with diffused slow upwelling into the thermocline (Figure 4.3). The belt is closed by the interoceanic surface and nearsurface current system above the abyssal layers in the tropics and subtropics. It transports water masses for new convective sinking in polar regions of both hemispheres (Rahmstorf, 2002). Two of the prime propelling engines of the THC with the upright standing ‘‘U-turns’’ (Schiermeier, 2006) are situated in the Labrador and the Nordic Seas. Figure 4.B demonstrates the complex interaction between surface and abyssal currents in the North Atlantic, coupled by deep convection (vertical arrows). The shown cartoon by V. Byfield (personal communication, 2006) strongly simplifies the deep-sea basins by showing a uniform depth and square-edged ocean margins. Because dominant parts of the Atlantic THC2 follow a meridional direction, one often refers to the displayed advection processes (large horizontal arrows in Figure 4.B) as the ‘‘meridional overturning circulation’’ (MOC). More generally, 2

In the THC scenario, the wind-driven part of the global oceanic circulation is excluded per definition.

40

Abyssal and Contour Currents

Figure 4.3 Highly simplified cartoon of the global thermohaline circulation (THC), modified from the original presentation of the oceanic ‘‘conveyor belt’’ by Broecker (1991). For further details, see the multicolor version of Figure 4.3 on the CD-ROM (courtesy Dr S. Rahmstorf, PIK, Potsdam, Germany). Surface and near-surface waters flow toward convection regions in the Labrador and Nordic Seas, and in the Weddell and Ross Seas. They recirculate as deep and abyssal currents, and participate in basin-scale slow upwelling in the interior of all three oceans. Sea-surface salinitycontrols the convection process at high latitudes decisively.

physical oceanographers define the term ‘‘advection’’ as the transport of a water mass and its properties as a current in a three-dimensional velocity field. The former describes predominantly the horizontal large-scale flow, whereas ‘‘convection’’ refers to locally induced vertical motion driven by buoyancy forces.

4.2.

A BYSSAL C URRENTS IN THE G LOBAL T HERMOHALINE C IRCULATION

Historically, the basic concept of the THC was developed by Stommel (1958) and Stommel and Arons (1960): in the oceans’ interiors, rising abyssal water from a flat bottom is laterally replaced by freshly ventilated water from a limited number of sinks in polar regions. On the northern hemisphere, the transformation from light surface waters to heavy bottom water that starts the deep convection process is restricted to late winter time in semienclosed polar seas (S1 in Stommel’s simplified diagram (Figure 4.4) such as the Greenland, Norwegian and Labrador Seas. In the south, the Weddell (S2) and Ross Seas play analogous roles in deep-water renewal. The Stommel–Arons theory postulates the existence of a confined current in a 100–200 km wide corridor in the form of the Deep Western Boundary Current (DWBC) and a remaining, much larger, uniform upwelling regime with minimal horizontal drift. Boundary currents have been observed in all oceans (cf. Imawaki et al., 2001). They are not necessarily restricted to western continental slopes, but can also be guided by flanks of major submarine mountain chains such as the Mid-Atlantic Ridge (Figure 4.C; Machı´n et al., 2006).

41

W. Zenk

S1

S2

Figure 4.4 Classic model of the depth-integrated currents of the thermohaline circulation below 2000 m depth (after Stommel, 1958). S1 and S2 symbolize convective source regions at polar latitudes. The circulation in the interior is fed and driven by boundary currents on the western sides of the ocean basins.

In contrast to the spatially limited downwelling regions, Stommel and Arons assume for the upwelling at moderate and lower latitudes fairly broad lateral and long-term scales. Calculations of the global integral upwelling speed based on steady-state downward heat-flux estimates (Hogg, 2001) result clearly in the submillimeter per second range. Experiments to directly observe the very slow upward drift in the open ocean have so far been unsuccessful. Instead, the horizontal dominance of internal waves, ubiquitous eddies, and boundary currents have been repeatedly documented in numerous trials since early current measurements in the western North Atlantic (Swallow and Worthington, 1961; Crease, 1962). The vertical velocity of the widely spread upwelling process implies vortex stretching on a basin scale. Conservation of potential vorticity on the rotating earth induces the poleward return current beneath the base of the thermocline. Up till now, dedicated observations have failed to remove all inconsistencies of the Stommel–Arons theory. Even after the end of the decade-long World Ocean Circulation Experiment (WOCE; Siedler et al., 2001) two such inconsistencies remain: 1. Observations of tracer concentrations and float trajectories, both with high spatial resolution, show a clear dominance of deep zonal current directions. The classical theory, however, favors the conventional picture of advection in the form of large gyres in each of the major ocean basins without a preference of zonal currents (Hogg, 2001). 2. In theory the diapycnal3 diffusivity, required to support the thermocline circulation, appears to be one order of magnitude too high in comparison with observations (Webb and Suginohara, 2001). 3

Direction normal to the local isopycnal surface.

42

Abyssal and Contour Currents

0 –500 –1000

Water depth (m)

–1500 –2000 –2500 –3000 –3500 –4000 –4500 –5000 –5500 –6000 –38

0.0

–36

0.1

–34

0.2

–32

0.3

0.4

–30

0.5

–28 –26 Longitude

0.6

Diffusivity

0.7 (10–4

–24

0.8

–22

0.9

2.0

–20

5.0

–18

8.0

–16

22.0

m2 s–1)

Figure 4.5 Directly observed distribution of diffusivity in the Brazil Basin of the South Atlantic (from Polzin et al., 1997; with permission from The American Association for the Advancement of Science). The zonal section runs from the continental rise off Brazil (left side) toward the western flank of the Mid-Atlantic Ridge. Note the nonlinear scale for diffusivity. High correlation between diffusivity and bottom roughness is found on the slope of the ridge on the right side. The white line marks the observed depth of the 0.8°C potential isotherm. See also the multicolor version of Figure 4.5 on the CD-ROM.

Such concerns can be tested in general circulation models. Energy–budget studies suggest that the role of tidal mixing may have been underestimated in the past (Munk and Wunsch, 1998). Also varying bottom topography and roughness have a significant impact on the spatial distribution of density and velocity microstructure even hundreds of meters above the sea floor. A confirming example for the heterogeneous distribution of diapycnal diffusivity based on velocity microstructure observations in the Brazil Basin is shown in Figure 4.5 (Polzin et al., 1997). Above the bottommost 150 m of the rough Mid-Atlantic Ridge, their observations reveal an increased diffusivity that is about two orders of magnitude higher then above the smooth abyssal plain.

4.3. 4.3.1.

CONTOUR C URRENTS

Physical properties and modes

In physical oceanography, an abyssal current is generally defined as a flow of water masses beneath the bottom of the main thermocline or within the cold water sphere

43

W. Zenk

(cf. Zenk, 2001). In most cases, contour currents can be classified as a particular mode of abyssal currents, although their depth level is not exclusively associated with abyssal depths. Actually, most of the contour currents investigated so far occur at intermediate levels. They are controlled by topography and, as long as they have not yet reached their equilibrium level, by gravitation. Other forces acting on contour currents include pressure gradients, Coriolis and inertial forces, and – where applicable – bottom drag. Deflection against a side wall of a basin is an elementary property of all contour currents. Friction between the bottom and contour currents is part of the marine exogenic4 transformation processes. Besides the required minimum current velocity for erosion, transportation, and resuspension sedimentation depends critically on the cohesiveness of the ground and on the grain size of the available material. The simplified diagram in Figure 4.6 describes some aspects of the complex interaction of near-bottom currents versus sedimentation. Note the logarithmic scales on both axes of the experimentally determined interactions (Heezen and Hollister, 1971). A more extended version of Figure 4.6 was published later by Hollister and Heezen (1972). The term ‘‘contourite’’ was first specified by Hollister and Heezen (1972) for ‘‘contour current-deposed sediment’’ to contrast markedly with turbidite or

Current velocity (cm s–1) 0.01

0.1

1.0

10

100

Pebbles

e dim

l) E ohe rosion sive mat eria

Silt

(inc

Tr an

sp

or

ta

tio

n

Se

1.0

0.1

Grain diameter (mm)

tion

nta

Sand

(coh Erosion esive mate rial)

Granules

0.01

Clay

0.001

Figure 4.6 Simplified representation of near-bottom current velocities required for erosion, transport, and deposition (after Heezen and Hollister, 1971; with permission from WHOI, Woods Hole, MA, USA). Erosion velocities for fine particles are uncertain because they depend on the degree of sediment cohesion.

4

Originating outside the lithosphere.

44

Abyssal and Contour Currents

‘‘turbidity-current-deposed sediment’’. It is irrelevant where contour currents touch a side wall. The frictional interface between moving water layers and the sea floor may consist of the continental rise or slope, the shelf edge, or a bottom of a broad canyon. Contour currents differ from more event-controlled turbidity currents in so far that they show a quasi-steady flow, though potentially interrupted or enhanced by seasonal changes or migrating eddies. At the continental margins, contour currents run parallel to the slopes. Their counterparts, i.e. turbidity currents, follow paths perpendicular to the outer-shelf margins. Both currents are visibly involved in the morphological diversity of continental margins (Figure 4.D, after Blondel, 2003). Optical properties of contour currents, along with other physical parameters such as potential temperature, are often suitable to trace them over larger distances. As an example, the bottom part of Figure 4.7 shows a vertical section of light

Pressure (dbar)

3600

1.5

2

1.5

1

1

3800

1.5

0.5

4000

0

0.5

1

4200

0.5

4400

0

4600 Potential temperature (°C)

4800

–39.5

–39.4

Pressure (dbar)

3600

–39.3 –39.2 Longitude (°W)

–39.1

–39

Clear water minimum 5.8

3800 5.9

5.9

4000

5.8 5.8

4200 4400 5.9

4600 4800

–38.9

6

Attenuation (0.1 m–1)

–39.5

–39.4

–39.3 –39.2 Longitude (°W)

–39.1

–39

–38.9

Figure 4.7 Vertical distribution of abyssal potential temperature (top) and of red-light attenuation (bottom) in a zonal section between the Santos Plateau and theVema Channel of the South Atlantic at approximately 29°S. See also the multicolor version of Figure 4.7 on the enclosed CD-ROM (Dr A. Macrander, AWI, Bremerhaven, Germany, 2006, personal communication). The Vema Channel represents a throughflow channel between the Argentine and Brazil Basins. The graph intersects the channel extension transporting cold (<2°C) Antarctic Bottom Water (AABW) northward at a position about 210 km downstream of the Vema Sill (for location of theVema Channel, see Figure 4.2 or Figure 4.A____(2a) on the CD-ROM). Stations are indicated by open triangles (r).The x-symbols (6) denote the area of the strongest near-bottom current, which coincides with the highest attenuation.This co-occurrence indicates the active role of the abyssal current as a sediment-transporting contour current.

W. Zenk

45

attenuation between the clear water minimum (CWM) of the water column and the bottom of the Vema Channel in the South Atlantic. The main equatorward overflow of cold and dense Antarctic Bottom Water (AABW) from the Argentine into the Brazil Basin takes place at the Vema Sill (31S, 39W) at a depth of 4646 m (Zenk and Morozov, 2007). The confined bottom current acts on the sea floor and increases the number of suspended particles in the bottom boundary layer. On average, near-bottom speeds at this choke point range up to 0.3 m s1 (cf. Table 2 in Hogg et al., 1999). The core of the AABW transport is situated on the eastern side of the Vema Channel ( on the right side of the channel in both parts of Figure 4.7). In general, vertical profiles of light attenuation from open ocean sites are labeled by at least two nepheloid layers, one in the surface mixed layer (SML) occasionally reaching down to the bottom of the euphotic zone, and a second layer above the sea floor (bottom nepheloid layer, BNL). Near ocean margins, additional nepheloid levels (intermediate nepheloid layers, INL) are well documented. For example, Dickson and McCave (1986) have studied the genesis of INLs in great detail in the region of the Porcupine Bank in the eastern North Atlantic. On the western slope of the bank, at a depth of 400–600 m, several intermediate layers become detached from the sea floor and start isopycnal spreading further along-slope or even off-slope. Since the Porcupine Bank has been identified as a sector of enhanced semidiurnal tidal energy in the eastern North Atlantic (cf. Morozov, 1995), the authors propose tidal contour currents along the southern margin of the Celtic Sea as the origin of the observed INLs. The observed particle composition of INLs supports an origin as BNLs. We reproduce the specific conceptual INL/BNL model for the Porcupine Bank and the inferred general model for regions with topographic irregularities in Figure 4.8. Besides their striking optical characteristics, boundary currents are often marked by corresponding deep-water mass characteristics, which makes them easy to follow experimentally. Their pathways can be traced by conventional hydrographic signatures such as temperature, salinity, or density. The warm and salty Mediterranean outflow water (MOW) in the Gulf of Cadiz of the eastern North Atlantic represents an excellent example for the transition between a purely bottom-following current to a genuine contour current (Heezen and Johnson, 1969; see also Herna´ndez-Molina et al., 2008a). It originates from the Strait of Gibraltar with a sill depth of about 300 m. The modulated exchange of surface and deep-water masses at Gibraltar represents a master case for Whitehead’s (1998) bidirectional currents in ocean straits. Due to the high density compared to the ambient North Atlantic Central Water (NACW), the Gibraltar Water with its characteristic high-temperature/salinity stamp descends while flowing westward to the equilibrium depth of about 1200 m (Figure 4.9) (Heezen and Hollister, 1971). At the longitude of the Portima˜o Canyon, situated off the southern Portuguese continental slope (8300 W), it loses bottom contact (Ambar et al., 2002). The MOW then turns northward at Cape St Vincent toward the Lisbon Canyon (Figure 4.10) (Richardson et al., 2000), and beyond (Bower et al., 2002b).

46

Abyssal and Contour Currents

Northerly wind High productivity

West

200

East

350 400

Porcupine Bank pe

lo

s al

t en

in

nt

Co

(a)

Specific model Sinking (mucus) particles Internal waves Nepheloid layer Slope current

600

800 0

High productivity (seasonal)

helf

S Shelf edge

400 Density interface

pe

Slo

(b)

General model

Resuspension points

Depth (m)

27.4σt

Depth (m)

Upwelling

800

Figure 4.8 Intermediate nepheloid layers (INLs) of the Porcupine Bank (from Dickson and McCave, 1986; with permission from Elsevier). (a) Schematic generation of an INL off Porcupine Bank in the eastern North Atlantic. Actions of wind, the slope (contour) current, and sinking particles are indicated by different vector symbols (6 #). (b) Generalization of the Porcupine example with topographical irregularities and varying density gradients showing the generation of multiple INLs.

An equally important topic in the hydrography of the Gulf of Cadiz is the formation of eddies of MOW (meddies). These submesoscale coherent vortices or lenses transport a highly saline blend of Gibraltar Water and NACW while they drift through the eastern Atlantic. At intermediate depth, they cause a large positive salinity anomaly compared with ambient waters of equal density. Anticyclonic, i.e. clockwise (on the northern hemisphere) rotating meddies have a diameter of about 100 km. With a thickness approaching 1000 m, their core depth at about 1200 m agrees with the Mediterranean Water tongue in the eastern North Atlantic. The lifetime of meddies can exceed 2 years, unless they collide with topographic structures. When they approach a seamount or the continental slope, they cause local submarine storms with speeds of the order of 0.2–0.3 m s1 (Figure 4.11) (Richardson et al., 2000). Although their appearance resembles more turbidity currents, we must count such collision occurrences as transient contour current with tangential direction along the bottom topography. Population estimates of meddies lie in the range of 29 coexisting lenses in the whole North Atlantic. Contour currents also occur in regions with submarine overflows. Prominent examples are the overflow water types Denmark Strait Overflow Water (DSOW) and Iceland Scotland Overflow Water (ISOW) in the subpolar gyre of the North

47

W. Zenk

Distance (km)

Salinity (‰)

Velocity (cm s–1)

Rate of change of core salinity

Core salinity

Salinity of overlying wate

Depth (m) Rocky

Rolling

Sediments waves

Current swept bottom

r

Smooth

Distance (nautical miles)

Figure 4.9 Conceptual diagram of the water exchange in the Strait of Gibraltar (SoG) and the westward outflow of highly saline MediterraneanWater (after Heezen and Johnson, 1969). Figure reproduced with permission from Bibliothe¤que Muse¤e Oce¤anographique, Monaco and WHOI, Woods Hole, MA, USA. At the surface of the Gulf of Cadiz (GoC), eastward flowing less saline (lighter) overlying Atlantic water (open arrows) enters the Alboran Sea (AS). Below the sheared interface, the dense Mediterranean water spreads as a bottom current westward into the deep eastern North Atlantic (black arrows). Its pathway around the southwestern edge of the Iberian Peninsula (Cape St Vincent) toward the Lisbon Canyon has been documented by float observations (reproduced in Figure 4.10). Due to mixing with ambient waters (entrainment), the core salinity and speed (black columns from historical local observations) decrease down-slope. The Mediterranean undercurrent with its core layer (dashed line on lower left side) represents a classic contour current. Its exogenic interaction with the sea floor is documented in the different bottom types shown in the bottom profile on the lower right side.

Atlantic. Other sills with overflowing submarine cataracts are shown in Whitehead’s maps of Figure 4.A_(2a)5–(2c). At the southern edge of the Nordic seas, the convectively formed, cold dense water masses are dammed by the Greenland– Iceland–Scotland Ridge. After passing the Denmark Strait with a sill depth of about 600 m, DSOW can reach the Irminger Sea. DSOW, exiting the Denmark Strait, 5

Color versions in enclosed CD-ROM.

48

Abyssal and Contour Currents

39°

Lisbon Lisbon Canyon Setubal Canyon AM 109

Latitude (north)

38°

AM 136

Portugal

AM 127

AM 126b AM 165

Cape St Vincent

37°

00

m

Ormonde Seamount 20

AM 112 1000 m

AM 126b

2000 m

36° 11°

10°

9°

8°

Longitude (west)

Figure 4.10 Selected trajectories of the Mediterranean Outflow Current south of Portugal follow the Iberian continental slope northward around Cape St Vincent (after Richardson et al., 2000; with permission from Elsevier). On average, the shown float trajectories were collected at 1123 m depth.The AM xxx notation indicates individual RAFOS float trajectories. On its eastern flank beyond Cape St Vincent, this well-defined eastern boundary current shows properties of a contour current.

flows initially in a rather confined ‘‘pipe’’ as a distinguished current in a fluid of background water. Figure 4.12 demonstrates this overflow stream. It was experimentally determined by 18 closely spaced cross sections. Apart from some variability at the sill and early in the descent, DSOW follows a well defined path with remarkably little cross-stream scatter (Girton and Sanford, 2003). On its southwesterly way, it generates a gigantic contour current along the continental

49

W. Zenk

05/94

Coalescence with Meddy 18

04/94

06/94 08/94

38° Josephine

Latitude (north)

09/94

10/94

Lion

13

36°

Coalescence 11/94

Unicorn

Temperature (°C)

12 12/94

11 Josephine 10

9 Feb Apr Jun Aug Oct Dec Feb Apr Jun

1994

34° 16°

14°

12° Longitude (west)

10°

Figure 4.11 Coalescence of intermediate Mediterranean eddies (meddies) with a later series of collisions with seamounts of the Horseshoe Seamounts Chain in the eastern North Atlantic (after Richardson et al., 2000; with permission from Elsevier). Meddies conserve and transport concentrated Mediterranean water in their inner core. They act as roving reservoirs for salty water. Small circles identify the start of observation months in 1994 (format: mm/yy). The encounter of such a clockwise rotating warm and salty lens with the topography initiates turbulent mixing with decreasing core temperatures observed by the drifting neutrally buoyant RAFOS float. The inset shows such a temperature drop of about 3°C. When meddies approach seamounts like the Josephine Bank, they behave like local transient contour currents.

slope of East Greenland. While slowly sinking (lower inset in Figure 4.12) DSOW gets diluted with over- and underlying water through turbulent mixing. As in the case of Gibraltar Water in the Gulf of Cadiz, this entrainment process increases the initial water transport by a factor of 2 or slightly more. More to the east of the Denmark Strait, dense ISOW leaves the Norwegian Sea through the Faroe Bank Channel (sill depth almost 800 m) encountering with high speed on a zonal course the Reykjanes Ridge, part of the Mid-Atlantic Ridge south of Iceland (Lankhorst and Zenk, 2006). Like the DSOW farther to the west, the ISOW also hugs against a huge topographic impedance (Figure 4.13). The crest of the Reykjanes Ridge sinks southward. The blend of entrained ISOW spreads roughly below the isopycnal 1027.8 kg m3 ( = 27.8). Where topography permits, i.e. in the Bight (Figure 4.14) or Charlie Gibbs Fracture Zones, the

50

Abyssal and Contour Currents

Figure 4.12 Position of the core of the Denmark Strait OverflowWater (DSOW) downhill of the sill (modified after Girton and Sanford, 2003). The bars symbolize the half-width of the DSOW in each of the analyzed 16 cross sections.The left inset shows the geographical location off Greenland. The right inset demonstrates the descent of the overflow plume along the DSOW path.

powerful ISOW contour current transgresses the Reykjanes Ridge westward into intermediate levels of the Irminger Sea (Bower et al., 2002a). So far we have only considered present-day contour currents. The close relationship between these currents and contourites allows the reconstruction of palaeocurrents. For instance, Maldonado et al. (2003) studied contourite deposits in the Scotia Sea at the northeastern tip of the Antarctic Peninsula. In the confluence zone between the deep-reaching Antarctic Circumpolar Current (ACC) and the abyssal outflow of Weddell Sea Deep and Bottom Waters (WSDW, WSBW), elongate-mounded drifts developed along the left-hand margins of channelized bottom currents in the northwestern Weddell Sea and in the central Scotia Sea (Figure 4.15). These deposits record intensive bottom-current activity over geological timescales (21.3 Ma). The interaction between both currents (ACC and WSDW transport) within the Drake Passage generates an abyssal counter current beneath the ACC. It hugs against the northern slope of the South Shetland Island Arc when entering the Bellinghausen Sea. Hydrographic properties of this peculiar contour current and its biological and geological impact in the transition zone between the Atlantic and the Pacific were studied by Nowlin and Zenk (1988: Figure 4.E) and Giorgetti et al. (2003; Figure 4.16), respectively.

51

50

Latitude (°N)

60

W. Zenk

Velocity (cm s )

200

400

2 4 6 8

40

10

50

40

Longitude (°W)

20

10 (103 m2 s–1)

Figure 4.13 Streamlines inferred from subsurface float observations in the Iceland Basin at the level where the Iceland Scotland Overflow Water (ISOW) from Faroe Bank Channel in the north encounters Labrador Sea Water (LSW) entering through Charlie Gibbs Fracture Zone in the south (53°N) (from Bower et al., 2002a; with permission from Macmillan Publishers). For position, see Figure 4.2 or the multicolor version Figure 4.A____(2a) enclosed on the CD-ROM. The well-resolved contour currents at a depth between 1500 and 1750 m appear along the Reykjanes Ridge, where the streamlines lie closely together. The inset gives a measure for current speeds. Arrowheads indicate the circulation direction. A multicolor version of this figure is on the enclosed CD-ROM.

Three decades ago, the source region of this abyssal current at the northwestern exit of the Weddell Sea around the Antarctic Peninsula had already been studied by Hollister and Elder (1969) indirectly on the basis of a large number of photographic images (Figure 4.F).

4.3.2.

Entrainment and modeled contour currents

For many years, the descent and mixing of overflow and outflow waters between marginal seas and the open ocean have been modeled by numerical simulations (and in rotating tanks). The background for such efforts lies in the enormous impact that overflows from marginal seas exert on the THC at intermediate and abyssal depths of the oceans. Properties of marginal seas serve as markers or tracers which stamp the overflow plume, respectively the outflow tongue in the deep oceans. Such labeled tongues can be traced over global distances. Early numerical models considered outflows as steady currents without friction and entrainment. Soon the paramount importance of entrainment was revealed. It led

52

Abyssal and Contour Currents

(a)

2 00 −2 ug −A 21

56

6

7.8

4

28−Jan−2003 40

30 Longitude (W)

20

σθ = 2

LS W

−M ay 17

Latitude (N)

60

Potential temperature (°C)

00

5

8 (b)

−2

64

34.9

ISOW 35 Salinity

35.1

Figure 4.14 Drift of a cycling APEX float (Autonomous Profiling Explorer). The instrument drifted (a) along both sides of the Reykjanes Ridge, a subaqueous extension of Iceland (after Lankhorst and Zenk, 2006; with permission from the American Meteorological Society). Black dots indicate surfacing positions of the float. They are separated by 10-day intervals. The nominal park or drifting depth of the float is 1500 m. It was launched south of the Faroe Bank Channel extension (black dot with star). Once it was caught (60°N) by the Iceland Scotland Over Flow Water (ISOW), it reached maximal speed and paralleled the ridge as a topographyfollowing contour current or subsurface western boundary current. At Bight Fracture Zone (57°N), it escaped into the Irminger Basin on the western side of the Reykjanes Ridge where the eastern boundary current speed is clearly slower than in the Iceland Basin. The potential temperature/salinity diagram (b) contains selected samples from two vertical profiles before (O) and after (D) the float’s passage through Bight Fracture Zone. The ISOW is characterized by a density  > 27.8.

to more advanced models with a stream tube approach (Smith, 1975) and parameterization of bottom drag and entrainment (Price and Baringer, 1994). Figure 4.17 displays the schematic by Price and Baringer’s outflow model on a rotating earth (f = Coriolis parameter) and associated density () profiles.  represents the excess density due to the outflow. Outflows are relatively thin in comparison to the whole depth (D) of the water column. Their thickness (H) amounts to about 1% of their width (W). The density profile shows a stratified ocean with the underlying homogeneous outflow plume beneath H. The simple geostrophic balance between two opposing forces (Coriolis and buoyancy) is supplemented by the two parallel forces (bottom drag and entrainment) acting perpendicularly to both others. The resulting triangle of forces causes a slight downhill angle between isobaths on the slope and the spreading velocity (u). Price and Baringer (1994) were among the first who modeled a whole suite of out- and overflows in the North Atlantic and the Southern Ocean. Figure 4.G shows an example from the simulated pathway of the Mediterranean Water plume in the Gulf of Cadiz. The general course of the core layer around Cape St Vincent is quite satisfying when compared with selected RAFOS6 float trajectories shown in Figure 4.10. 6

‘‘SOund Fixing And Ranging’’ spelled backwards.

53

W. Zenk

C6A (21.3 Ma)

(a)

50°W

40°W

30°W

20°W

56°S

Present

60°S

64°S (b)

Figure 4.15 Sketch of bottom flows in the eastern Scotia Sea and northern Weddell Sea showing pathways of the Antarctic Circumpolar Current (ACC), Weddell Sea Deep Water (WSDW), and Weddell Sea Bottom Water (WSBW) (from Maldonado et al., 2003; with permission from Elsevier). A reconstruction from 21.3 Ma ago (top) can be compared with the present-day situation (bottom).The interaction of the various current systems generates a wide variety of contourites. A multicolor version of this figure is on the enclosed CD-ROM.

Further progress was achieved by the development of a hydrostatic, reduced gravity, two-dimensional primitive equation model with an application to DSOW along the slope of East Greenland (Jungclaus and Backhaus, 1994). Although models are increasingly successful in reconciling observations and simulations, the treatment of turbulent entrainment remains of primary importance

54

Abyssal and Contour Currents

Figure 4.16 The hypothetical extension of the westward counter (contour) current (see Figure 4.E on the CD-ROM) beneath the Antarctic Circumpolar Current (ACC) in Drake Passage (see inset on the lower right side) is shown as a black line. Its local existence was documented by moored current meters (D), in sediment cores (•), and by hydrographic observations (Giorgetti et al., 2003). The figure demonstrates the occasional co-existence of contour currents that parallel the continental slope and turbidity currents documented by the shown canyons or turbidity channels. Abbreviations: CTD = Conductivity, Temperature, Depth recorder; DSDP = Deep Sea Drilling Program; ODP = Ocean Drilling Program. For further details, see Rebesco et al. (1998b, 2007).

in large-scale ocean modeling. The descent and spreading of model overflow waters injected from high latitude and marginal seas, is extremely sensitive to the parameterization of the entrainment process. Mixing and acceleration of ambient waters in boundary currents strongly control transport and slow deepening of contour

55

W. Zenk

f/2

Depth

Density ρ

H

Outflow

δ+ρ

u Coriolis

u

Oceanic nt me g ain Dra r t En ttom Bo

W

δρ

H

D

Buoyancy

Figure 4.17 Schematic Mediterranean outflow in the Gulf of Cadiz (from Price and Baringer, 1994; with permission from Elsevier). The outflow plume (streamtube) leaves the Strait of Gibraltar (upper left corner of the left figure) and spreads westward on the bottom of the modeled gulf. The equilibrium of Coriolis, buoyancy, and frictional forces (bottom drag and entrainment) controls the down-slope direction of the modeled plume with velocity (u). The right figure shows the underlying density differences between the sea surface and a homogeneous outflow layer.

currents. For a better understanding of these processes in ocean models, further field observations and highly resolving numerical experiments are equally necessary (Price, 2002).

4.4.

C ONCLUSION AND SUMMARY

To the best of our knowledge, the term ‘‘contour current’’ as a particular mode of abyssal currents was first introduced in the (marine geological) literature by Heezen et al. (1966). Their original definition comprises near-bottom currents that ‘‘flow along isopycnals which are approximately parallel to the bathymetric contours’’. Due to a lack of definition in the physical oceanography literature,7 we suggest a more focused approach for ‘‘contour currents’’: Contour currents are predominantly unidirectional subsurface currents that are in contact with a sidewall. They are in quasi-geostrophic balance and are controlled by the local bottom topography. Their kinetic energy is attenuated by friction generating an exogenic stress on the sea floor. In marine sedimentology, the topographic control of contour currents gives rise to the generic name ‘‘contourite.’’ Contourites are sediments deposited or substantially reworked by

7

Even the extensive index of the comprehensive synthesis work of the World Ocean Circulation Experiment (Siedler et al., 2001) contains no entry on ‘‘contour currents.’’

56

Abyssal and Contour Currents

the action of contour and plain bottom currents, commonly in the vicinity of a continental rise. On descending slopes, contour currents start as density or gravity currents. They represent the confined advection of one fluid through another fluid (background water). While slowly sinking to their equilibrium level, their direction parallels largely the topography. They become mixing agents exchanging momentum and properties with ambient waters by turbulent entrainment. Where they are in contact with the sea floor, their bottom drag transforms kinetic energy into exogenic action on the sea floor. As a result, a large variety of contourite modes can be established. As a vicarious example, we reproduce in Figure 4.18 prime conceptual factors and processes that are engaged in contour currents and material transport along the Iberian Peninsula. The material has been compiled by H. de Haas (personal communication, 2006) for the cover of a synthesis volume of the Ocean Margin Exchange project (OMEX).8 We recognize the continental slope off Portugal with two stacked contour currents: the (seasonally) alternating slope current (Pingree et al., 1999) and the lateral northward transport of the MOW. Abyssal BNLs and INLs interact with downwelling of slope waters along the continental margin. Narrow canyons are natural obstacles in the course of bathymetryfollowing contour currents. In contrast, event-triggered turbidity currents in

La and teral re- trans sus p pen or t sio n

Win d upw -drive n elli ng

W MO

INL NL B

0

Ib

Sl

op

e

cu

e

ri

a

rre

20

00

wn Do ling l we L BN

40

00 41

°

nt m 200 0m

100

9°

n nyo

Ca

4 La 0° titu de (

N)

)

39

10°

e (W

tud ngi

Lo

°

Figure 4.18 Cartoon of the Iberian continental margin reflecting major processes and currents affecting particle transfer over the shelf edge and off the slope (from Pingree et al., 1999; with permission from Elsevier).The shown European Slope Current and the advection of Mediterranean Outflow Water (MOW) represent contour currents. Canyons act as pathways for direct transport of particles (occasionally by turbidity currents) from the shelf to the deep sea. A multicolor version of this figure is on the enclosed on the CD-ROM (Courtesy: Dr H. v. Haas, NIOZ,Texel,The Netherlands).

8

Funded between 1997 and 2000 by the European Commission under GD XII.

W. Zenk

57

canyons act as off-slope pathways for direct particle transport from the shelf right to an abyssal basin. Numerical simulations of outflow plumes as an approximation to the oceanographic problem of deep-water production and circulation show encouraging results. They also help us to better understand the dynamics of contour currents in three dimensions.

ACKNOWLEDGMENTS Many observational results shown in this chapter have been made possible by grants of the Deutsche Forschungsgemeinschaft, Bonn (Sonderforschungsbereich 460 – Dynamics of Thermohaline Circulation Variability). The German work in the Vema Channel has been funded by the Bundesministerium fu¨r Bildung und Forschung, Berlin, under CLIVAR-marine 2 (Climate Variability and Predictability).

C H A P T E R

5

D EEP - WATER B OTTOM C URRENTS AND T HEIR D EPOSITS G. Shanmugam Department of Earth and Environmental Sciences, The University of Texas at Arlington, Arlington, TX, USA

Contents 5.1. Introduction 5.1.1. Bottom currents versus turbidity currents 5.1.2. Genetic nomenclature 5.2. Thermohaline-Induced Geostrophic Bottom Currents 5.2.1. Antarctic and Artic bottom currents 5.2.2. Velocity 5.2.3. Deposits 5.2.4. Reservoir potential 5.3. Wind-Driven Bottom Currents: The Loop Current 5.3.1. Velocity 5.3.2. Deposits 5.3.3. Reservoir potential 5.4. Deep-Water Tidal Bottom Currents 5.4.1. Previous studies 5.4.2. Velocity 5.4.3. Deposits 5.4.4. Facies associations in submarine canyons 5.4.5. Reservoir potential 5.5. Internal Waves and Tides (Baroclinic Currents) 5.5.1. Nomenclature 5.5.2. Velocity 5.5.3. Deposits 5.6. Conclusions Acknowledgments

5.1.

59 59 61 63 63 63 64 66 66 67 68 70 72 72 72 75 76 76 77 77 79 80 81 81

I NTRODUCTION

The primary objective of this chapter is to discuss deep-water bottom currents and their deposits from oceanographical and sedimentological standpoints. The general term deep water is used here because bottom currents operate in deep waters of Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00205-7

Ó 2008 Published by Elsevier B.V.

59

60

Deep-water Bottom Currents and Their Deposits

both marine and lacustrine environments (Shanmugam, 2006a). The general term bottom current is used in this chapter because it covers a variety of bottom currents of different origins, flow directions, and velocities (Shanmugam et al., 1993a, p. 1242). Southard and Stanley (1976) recognized five types of bottom currents at the shelf break based on their origin. These currents are generated by (1) thermohaline differences, (2) wind forces, (3) tidal forces, (4) internal waves, and (5) surface waves. In addition, tsunami-related traction currents have been speculated to occur in bathyal waters (Yamazaki et al., 1989), but the mechanics of such currents are not yet understood (Shanmugam, 2007; 2008a). This chapter discusses four types of deepwater bottom currents, namely (1) thermohaline-induced geostrophic bottom currents (i.e., contour currents), (2) wind-driven bottom currents, (3) deep-water tidal bottom currents, and (4) internal waves and tides (baroclinic currents).

5.1.1.

Bottom currents versus turbidity currents

In discussing deep-water bottom currents, it is imperative that a clear distinction be made between bottom currents and turbidity currents. Turbidity current is a sediment flow with Newtonian rheology and turbulent state, in which sediment is supported by fluid turbulence and from which deposition occurs through suspension settling (Dott, 1963; Sanders, 1965; Middleton and Hampton, 1973; Shanmugam, 2000). Distinguishing deposits of deep-water bottom currents from those of turbidity currents has been and still is a challenge (Bouma and Hollister, 1973; Stow, 1979; Mulder et al., 2009). However, bottom currents and their deposits differ from turbidity currents and their deposits in the following respects: 1. bottom currents may occur on the shelf, slope, and in basinal environments, whereas turbidity currents are more common on the slope and basinal environments (Figure 5.1); 2. bottom currents are driven by thermohaline, wind, or tidal forces, whereas turbidity currents are driven by sediment gravity; 3. bottom currents may flow parallel to the strike of the regional slope, may flow in circular motions (gyres) unrelated to the slope, or may flow up and down submarine canyons (tidal), whereas turbidity currents always flow down-slope (Figure 5.2), though flow parallel to the strike of the regional slope may occur due to local morphology (e.g., marginal troughs); 4. bottom currents persist for long periods of time and can develop equilibrium conditions, whereas turbidity currents are episodic or surge-type events that fail to develop equilibrium conditions (Allen, 1973, 1985); 5. bottom currents can exist without the presence of entrained sediment and, for this reason, they are termed ‘‘clear water currents’’ (Bouma and Hollister, 1973, p. 82), whereas turbidity currents cannot exist without entrained sediment (Middleton and Hampton, 1973); 6. bottom currents show oscillating energy conditions, whereas turbidity currents exhibit waning energy conditions (Sanders, 1965); 7. bottom currents transport sand primarily by traction (i.e., bed load movement by sliding, rolling, and saltation; Allen, 1984), whereas turbidity currents generally transport fine-grained sand and mud in suspension;

G. Shanmugam

61

Figure 5.1 Schematic diagram showing complex deep-marine sedimentary environments occurring at water depths greater than 200 m (shelf ^ slope break). In general, shallow-marine (shelf ) environments are characterized by tides and waves, whereas deep-marine (slope and basin) environments are characterized by mass movements (i.e., slides, slumps, and debris flows), bottom currents, and pelagic/hemipelagic deposition. Turbidity currents may be common in basinal settings. Note up- and down-tidal bottom currents in submarine canyons (opposing arrows). Along-slope movement of contour-following bottom currents (contour currents) and circular motion of wind-driven bottom currents are important processes outside of the canyon. After Shanmugam (2003); with permission from Elsevier.

8. traction structures (e.g., parallel laminae, ripple laminae, and cross-beds) should be originally common in bottom-current sands (Shanmugam, 1997a; Martı´nChivelet et al., 2008) (though some authors suggest that this should be successively destroyed by pervasive bioturbation allowed by the continuous nature of bottom currents), whereas normal grading is the norm in turbidites that are deposited by relatively catastrophic episodic events of waning energy (Kuenen and Migliorini, 1950); 9. bottom-current deposits generally exhibit sharp upper contacts (Hollister, 1967), whereas turbidites show gradational upper contacts; 10. bottom currents can result in well-sorted sand with good porosity and permeability because of reworking and winnowing away of mud (Shanmugam et al., 1993a), whereas turbidites are poorly sorted, commonly mud-rich deposits with low porosity and permeability (Pettijohn, 1957; Sanders and Friedman, 1997).

62

Int erc ha nn el

Deep-water Bottom Currents and Their Deposits

Channel margin slump

Channel axis

Levee

Axial turbidity currents Overbank “turbidity” currents Contour currents

Figure 5.2 Conceptual model showing the spatial relationship between down-slope turbidity currents and along-slope bottom currents (contour currents). After Shanmugam et al. (1993a); with permission of the American Association of Petroleum Geologists.

5.1.2.

Genetic nomenclature

In science, words should have clear and consistent meanings. In geology, however, this is not always the case (Shanmugam, 2006b). The tradition of genetic nomenclature in sedimentary geology began with the introduction of the term turbidite for a deposit of a turbidity current in deep-water environments (Kuenen, 1957). Kuenen and Migliorini (1950, p. 99) and Kuenen (1967, p. 212) suggested that normal grading of a turbidite bed was a consequence of deposition from a single waning turbidity current. For a genetic term to succeed, (1) it must be based on sound fluid dynamic principles, (2) its usage must be accurate (relying on sedimentological description), precise (referring to a single process), and consistent (requiring a steady and a uniform application in time and space), and (3) it must imply a diagnostic flow behaviour. However, genetic terms of turbidites, which include (1) atypical turbidites, (2) fluxoturbidites, (3) hemiturbidites, (4) high-concentration sandy turbidites, (5) megaturbidites, (6) problematica turbidites, (7) seismoturbidites, and (8) undaturbidites, fail to reveal a clear flow behaviour (see Table 2.2 in Shanmugam, 2006a). Like turbidites, genetic terms of bottom-current-reworked sands also fail to divulge a diagnostic flow behaviour. 1. The term contourite emphasizes current orientation with respect to bathymetric contours (Hollister 1967), not the flow behaviour. 2. The term laminite represents sedimentary structure (i.e., lamina) (Lombard, 1963), not the flow behaviour. 3. The term tractionite impies traction deposition from bottom currents (Natland, 1967), but traction deposition has also been attributed to turbidity currents (Middleton and Hampton, 1973).

63

G. Shanmugam

4. The term tidalite impies alternating units of traction and suspension deposition from shallow-water tidal currents (Klein, 1971), but deep-water tidal bottom currents can also develop alternating units of traction and suspension deposition (Klein, 1975; Shanmugam, 2003). 5. The term winnowite impies winnowing action of bottom currents (Shanmugam and Moiola, 1982), but this is a reworking rather than a depositional process. 6. The term tsunamite represents tsunami-induced tractive bottom-currentreworked sediment (Yamazaki et al., 1989), but the term tsunami is not a self-defining expression of a single depositional process (Shanmugam, 2006b). In summary, genetic terms are ineffective for communicating depositional mechanics in process sedimentology (Shanmugam, 2006a). Therefore, I have adopted the general term ‘‘bottom-current-reworked sands.’’

5.2. 5.2.1.

T HERMOHALINE-I NDUCED GEOSTROPHIC BOTTOM C URRENTS

Antarctic and Artic bottom currents

Aspects of thermohaline-induced bottom currents are discussed by Zenk (2008). These deep-marine bottom currents in modern oceans became popular when Heezen et al. (1966) reported deep-water masses that flow along the ocean floor. An example of such bottom currents is the Antarctic Bottom Water (AABW). AABW was first identified by Brennecke (1921) in the northwest corner of the Weddell Sea in the Antarctic region (Figure 5.3). The origin of the AABW was attributed to the formation of ice from surface freezing over the Antarctic continental shelves. When sea ice forms, seawater experiences a concurrent increase in salinity due to salt rejection and a decrease in temperature. The increase in the density of cold saline (i.e., thermohaline) water directly beneath the ice triggers the sinking of a water mass down the continental slope and the spread to other parts of the ocean. The Western Boundary Undercurrent (WBUC or WBU), the Arctic counterpart to AABW, originates as a cold dense-water mass from the Norwegian Sea off Greenland (Worthington and Volkman, 1965). It flows along the western margin of the North Atlantic (Figure 5.3). These thermohaline currents tend to flow parallel to the slope, that is, along the slope at right angle to down-slope flowing gravity-induced currents (Figure 5.2). The WBUC is deflected in the northern hemisphere to the west as a result of the Coriolis force. Because of its tendency to flow parallel to bathymetric contours, the WBUC is known as a contour current (Heezen et al., 1966). These currents are commonly known as geostrophic contour currents, because they strike a balance between the Coriolis and the gravity forces.

5.2.2.

Velocity

Measured current velocities usually range from 1 to 20 cm s–1 (Hollister and Heezen, 1972); however, exceptionally strong, near-bottom currents with

64

Deep-water Bottom Currents and Their Deposits

80°

160°

100°

80°

0°

Arctic ocean

80°

Norwegian Sea

C

BU

40°

Ocean

W

ater

G GS

Antarctica

Indian ocean

Atl

BW AA

w

Pacific

p

40°

antic

Pacific ocean de e

0°

Weddell Sea

W

AAB

Antarctica

80°

Figure 5.3 Simplified circulation patterns of major thermohaline bottom currents (contour currents). Most contour currents originate from the Weddell Sea and from the Norwegian Sea. WBUC = Western Boundary Undercurrent. AABW =Antarctic BottomWater. GSG = Gulf Stream Gyre. Compiled from several sources (Wu«st, 1950; Stommel, 1958; Heezen and Hollister, 1971; Stow and Lovell, 1979).

maximum velocities of up to 3 m s 1 were recorded in the Strait of Gibraltar (Gonthier et al., 1984). Bottom-current velocities of 73 cm s 1 were measured at a water depth of 5 km on the lower continental rise off Nova Scotia (Richardson et al., 1981). Stow and Lovell (1979) summarized velocities of contour currents. Because of their high velocities, bottom currents in the deep sea are quite capable of erosion, transport, and redeposition of fine to coarse sand. Regional erosional unconformities in the deep sea throughout thousands of square kilometers of sea floor have been attributed to erosion by bottom currents (Berggren and Hollister, 1977; Tucholke and Embley, 1984; Shanmugam, 1988). In the Rockall Trough region, for example, bottom currents associated with the North Atlantic Deep Water (NADW) have caused an erosive area extending over 8500 km2 in water depths of 500–2000 m (Howe et al., 2001).

5.2.3.

Deposits

The following general features of bottom-current-reworked deposits have been discussed by Hubert (1964), Hollister (1967), Hollister and Heezen (1972), Bouma and Hollister (1973), Unrug (1977), Stow and Lovell (1979), Lovell and Stow (1981), Shanmugam (2000), and Ito (2002): • fine-grained sand and silt; • thin-bedded to laminated sand (usually less than 5 cm) associated with deepmarine mud;

65

G. Shanmugam

• • • • • • • • • • • •

rhythmic occurrence of sand and mud layers; sharp to gradational bottom contacts; sharp, non-erosional, upper contacts; well-sorted sand; little depositional mud matrix (clean sand); horizontal laminae; low-angle cross-laminae; ripple cross-laminae; lenticular bedding or starved ripples; inverse size grading (Figure 5.4); mud-offshoots in ripples; mud-draped ripples.

No single criterion by itself is unique to bottom-current-reworked sands. Although many of the criteria listed above can be attributed to processes other than bottom-current reworking, the association of several of the above criteria in a given deep-water example, along with the knowledge of the regional depositional setting, greatly enhances the chance of recognizing bottom-current-reworked facies. It is difficult to establish that a given sedimentary structure in the rock record was originated by contour-following thermohaline currents without establishing the paleowater circulation pattern independently. Therefore, the general term ‘‘bottom-current-reworked sands’’ is appropriate.

cm 5 4 3 2 1 0

Figure 5.4 Core photograph showing inverse grading and sharp upper contacts of sand layers (arrow), interpreted as bottom-current-reworked sands. Paleocene, North Sea.

66

Deep-water Bottom Currents and Their Deposits

A contourite model, first proposed by Gonthier et al. (1984), consists of a basal inversely graded unit overlain by a normally graded unit (Stow et al., 1998a; Stow and Fauge`res, 2008). Criticism has been addressed in detail by Shanmugam (2000, 2006a). This facies model, for example, strongly suggests that bioturbation is a diagnostic feature. This suggestion is based on the belief that active contour currents would increase the oxygen concentration of the water mass, and thereby would increase the activity by benthic organisms. Tucholke et al. (1985), however, suggested that the degree of preservation of bioturbation is a function of bottom-current intensity; strong bottom currents do not favour preservation of biogenic structures. Bioturbated mud in the deep sea is equally abundant in areas that are unaffected by contour currents. Even if bioturbation were prevalent in areas of contour currents, it would not directly reveal anything unique about current orientation (i.e., contour-following currents). The bioturbation criterion for ‘‘contourites’’ is criticized because ancient deep-water turbidites (e.g., in the Late Cretaceous Point Loma Formation near San Diego, California) are extensively bioturbated and contain the trace fossil Ophiomorpha (Nilsen and Abbott, 1979). In contrast, convincing cases of ‘‘contourites’’ without bioturbation have been documented in the rock record (Dalrymple and Narbonne, 1996). In short, bioturbation cannot substantiate the contour-following current orientation, which is the basic tenet of the ‘‘contourite’’ model (Hollister, 1967). Finally, analogous to the five divisions (Ta, Tb, Tc, Td, and Te) of the turbidite facies model (i.e., the Bouma Sequence), Stow and Fauge`res (2008, their Figure 13.9) have introduced five divisions (C1, C2, C3, C4, and C5) for their contourite facies model. In defense of their facies model, Stow and Fauge`res (2008, p. 240) have argued that process models derived from ancient strata are less reliable in comparison to their contourite model derived from modern sediments. Their argument is specious because the Bouma Sequence, which they have used as the analogy for their model, was derived strictly from a study of ancient strata. Furthermore, no one has ever documented the complete Bouma Sequence from modern sediments. Idealistic turbidite and contourite facies models, despite their popularity, are a step backward in the pragmatic science of process sedimentology (Shanmugam, 2008b).

5.2.4.

Reservoir potential

Aspects of economic relevance of bottom-current-reworked sands are discussed by Viana et al., (2007). From a reservoir point of view, thermohaline bottom currents are important because of their ability to intrude and rework sand. A thick prism of calciclastic reworked sands (middle Miocene to Pleistocene) off Little and Great Bahama Banks has been studied in detail (Mullins et al., 1980). These calciclastic reworked sands were lithified by early submarine cementation. Measured maximum porosity and permeability values were 40% and 9880 mD, respectively. High permeability values are attributed to winnowing away of mud by vigorous bottom currents. Such bottom-current-reworked sands are potential petroleum reservoirs.

67

G. Shanmugam

5.3.

W IND-D RIVEN BOTTOM C URRENTS: THE LOOP C URRENT

The Loop Current in the eastern Gulf of Mexico is a wind-driven surface current (Figure 5.5). It enters the Gulf of Mexico through the Yucatan Straight as the Yucatan Current; it then flows in a clockwise loop in the eastern Gulf as the Loop Current, and exits via the Florida Strait as the Florida Current (Neumann and Pierson, 1966; Nowlin, 1972; Mullins et al., 1987). Finally, this current merges with the Antilles Current to form the Gulf Stream. The Loop Current also propagates eddies into the north-central Gulf of Mexico, where the Ewing Bank area, a case study used in this chapter, is located (Figure 5.5).

5.3.1.

Velocity

Velocities in eddies that have detached from the Loop Current have been recorded as high as 2 m s 1 at a depth of 100 m (Cooper et al., 1990). The Loop Current and related eddies pose significant problems for deep-water drilling (Koch et al., 1991).

36° m

N

f St

200 m

An

a Current

op

Gulf of

Atlantic Ocean

Gul

a

xico Me

Lo

tille

sC

urr

en

Cu

t

rre

r id

28°

rid

Texas

Flo

Ewing Bank 826 Garden Banks

rea

United States 32°

nt

Flo

24°

Yu c

ata

nC

20°

Mexico

ur

re

m 200

16°

nt

Nicaragua

Caribbean Current Rise

Pacific Ocean 10°

102°

0

98°

Isthmus of Panama

400 km

94°

South America 90°

86°

82°

78°

74°

70°

66°

Figure 5.5 Present circulation pattern of the Loop Current in the Gulf of Mexico.This winddriven surface current is considered to be affecting the sea floor (Pequegnat, 1972). Note the detached eddies from the Loop Current in the Ewing Bank area. After Shanmugam et al. (1993a); with permission of the American Association of Petroleum Geologists.

68

Deep-water Bottom Currents and Their Deposits

For example, drilling operations in the Green Canyon 166 area were temporarily suspended in August of 1989 because of high current velocities that reached nearly 150 cm s 1 at a depth of 45 m, and 50 cm s 1 at a depth of 250 m. These intense bottom currents affect the ability of a drilling rig to hold station over a wellhead (Koch et al., 1991). Current-velocity measurements, bottom photographs, high-resolution seismic records, and GLORIA side-scan sonar records indicate that the Loop Current influences the sea floor at least periodically in the Gulf of Mexico (Pequegnat, 1972). Computed flow velocities of the Loop Current vary from nearly 1 m s 1 at the sea surface to more than 25 cm s 1 at 500 m water depth (Nowlin and Hubert, 1972). This high surface velocity suggests a wind-driven origin for these currents. Flow velocities measured using a current meter reach up to 19 cm s 1 at a depth of 3286 m (Pequegnat, 1972). Such currents are capable of reworking fine-grained sand on the sea floor. Current ripples, composed of sand at a depth of 3091 m on the sea floor, are clear evidence of deep bottom-current activity in the Gulf of Mexico today (Pequegnat, 1972). These current ripples are draped by thin layers of mud. If these mud drapes on sand ripples were preserved in the rock record, they would be termed ‘‘mud-offshoots.’’

5.3.2.

Deposits

Deposits of the Loop Current have been interpreted in the cores from the Ewing Bank 826 Field, Plio Pleistocene, Gulf of Mexico. The Ewing Bank Block 826 Field is located nearly 100 km off the Louisiana coast in the northern Gulf of Mexico (Figure 5.5). It contains hydrocarbon-producing clastic reservoir sands that have been interpreted as bottom-current-reworked sands (Shanmugam et al., 1993a, b). The Ewing Bank cores exhibit the following features: • predominantly fine-grained sand and silt; • thin-bedded to laminated sand (usually less than 5 cm) intercalated with deepwater mud (Figure 5.6); • rhythmic sand and mud layers; • numerous layers (50 or more per 1 m of core); • sharp (non-erosional) upper contacts of sand layers (Figure 5.6); • sharp to gradational bottom contacts; • internal erosional surfaces; • external truncation surfaces; • megascopic inverse size grading (Figure 5.7); • microscopic inverse size grading; • horizontal lamination and low-angle cross-lamination (Figure 5.6); • cross-bedding; • lenticular bedding or starved ripples at core scale (Figure 5.6); • current ripples with preserved crest or with eroded crest; • ripple forms with curved bases; • flaser bedding; • mud-offshoots (Figure 5.8); • well-sorted sand and little depositional matrix.

1 cm

Figure 5.6 Core photograph showing discrete thin sand layers with sharp upper contacts. Traction structures include horizontal laminae, starved ripples (arrows), and low-angle crosslaminae. Dip of cross-laminae to the right suggests current from left to right. Note rhythmic occurrence of sand and mud layers. Middle Pleistocene, Gulf of Mexico. After Shanmugam et al. (1993a); with permission of the American Association of Petroleum Geologists.

Figure 5.7 Core photograph showing megascopic inverse size grading. Arrow shows the gradational nature of basal contact from mud (dark color) at the bottom to sand (light color) at the top. Each scale division is 3 cm. Middle Pleistocene, Gulf of Mexico. After Shanmugam et al. (1993a); with permission of the American Association of Petroleum Geologists.

70

Deep-water Bottom Currents and Their Deposits

cm 5 4 3 2 1 0

Figure 5.8 Core photograph showing discrete sand units with current ripples and mud offshoots (arrow). Note sigmoidal configuration of ripples and truncated tops. Middle Pleistocene, Ewing Bank Block 826, Gulf of Mexico. After Shanmugam et al. (1993a); with permission of the American Association of Petroleum Geologists.

Most of the features listed above are interpreted as the products of deposition by traction or combined traction and suspension (Figure 5.9). Sand layers with traction structures occur in discrete units, but not as part of a vertical sequence of structures. These features are interpreted here as evidence for bottom-current reworking. Because traction structures are also observed in ‘contourite’ deposits discussed before, caution must be exercised in classifying a deposit as a ‘contourite’ based solely on traction structures without independent evidence for contour-following bottom currents in the area.

5.3.3.

Reservoir potential

Details of reservoir quality of reworked sands in the Ewing Bank area have been discussed by Shanmugam et al. (1993a). Sedimentary structures in cores reveal that the reservoir is composed of a lower unit with a turbidite channel sand, and an upper unit with a bottom-current-reworked sand. Although both the turbidite and the reworked sands show porosity values in the range of 35–42%, their permeability values are strikingly different. The basal turbidite channel sand unit shows a distinct upward decrease in permeability (with

71

G. Shanmugam

Mud offshoots

Climbing ripple cross-bedding

Suspension (mud offshoot)

Traction and suspension

5 cm

5 cm

Traction

Flaser bedding

Lenticular bedding

Suspension Traction

Suspension

5 cm

5 cm

Traction

Horizontal bedding

Rhythmic bedding Sharp upper contact

Traction

5 cm

1 cm

Traction Suspension

Cross-bedding

Sharp upper contact Sharp upper contact

Inverse grading

Erosion

10 cm

5 cm

Traction

Fine sand

Graditional Lower contact

Mud

Figure 5.9 Summary of traction features interpreted as indicative of deep-water bottomcurrent reworking. After Shanmugam et al. (1993b); with permission of the Geological Society of America.

72

Deep-water Bottom Currents and Their Deposits

maximum values of 5000 mD in the turbidites and about 500 mD in the bottom-current-reworked sands).

5.4.

D EEP-W ATER T IDAL BOTTOM C URRENTS

In this chapter, I maintain a distinction between ‘‘deep-water tidal bottom currents’’ as a general category and ‘‘baroclinic currents’’ as a special category. This is necessary because deep-water tidal bottom currents may include baroclinic currents. However, not all deep-water tidal currents are baroclinic types. In addition, our understanding of baroclinic currents associated with internal tides is still in its infancy. Controversies and complexities associated with internal tides and ocean mixing have been addressed by Garrett (2003).

5.4.1.

Previous studies

Deep-marine tidal bottom currents in submarine canyons and in their vicinity are one of the best-studied and most extensively documented modern geologic processes (e.g., Shepard et al., 1969, 1979; Shepard, 1976; Beaulieu and Baldwin, 1998; Petruncio et al., 1998; Xu et al., 2002). During the past four decades, an understanding of deep-marine tidal bottom currents has been developed by synthesizing a great wealth of published information. This information includes direct observations from deep-diving vehicles, dredging from canyon floors, underwater photographs of canyon floors, photographs and X-radiographs of box cores, seismic profiles, current-velocity measurements (Shepard and Dill, 1966; Shepard et al., 1969, 1979; Dill et al., 1975; Shepard, 1976), and from study of conventional cores and outcrops (Shanmugam, 1997b, 2002). Selected examples of studies that dealt with tidal processes and/or their deposits in modern and ancient deep-water environments were reviewed by Shanmugam (2003).

5.4.2.

Velocity

Tidal currents are significant processes in many modern submarine canyons (Shepard et al., 1979). The interaction of the canyon topography with the tidal current is particularly important. In the modern Zaire (formerly the Congo) Canyon in West Africa, the canyon head can be traced 25 km up the estuary on land (Heezen et al., 1964b; Shepard and Emery, 1973; Droz et al., 1996). The deep Zaire Canyon is simply a deep-water extension of the Zaire estuary. The width and the relief of the canyon increase seaward from the estuary reaching a maximum width of 15 km and a maximum relief of 1300 m near the shelf break (Babonneau et al. 2002). The mean tidal range in the Zaire Canyon is 1.3 m (Shepard et al., 1979). Shepard et al. (1979) documented systematically that up- and down-canyon currents closely correlated with timing of tides (Figure 5.10). Shepard et al. (1979) measured current velocities in 25 submarine canyons at water depths

73

G. Shanmugam

Hueneme Sta. 28 Lo.1 Depth 448 m 3 mAB V vs Time

Down-canyon–Up-canyon cm s–1

30

30 cm s–1

20

20 cm s–1

10

10 cm s–1 Data from tide tables

0

Tide

2m

–1

10

10 cm s

20

20 cm s–1

30

30 cm s–1

1

0

2000 h 2/12/73

12

24

36

48

60

72

84

h

Figure 5.10 Time/velocity plot of data obtained at 448 m in the Hueneme Canyon, California, showing excellent correlation between the timing of up- and down-canyon currents and the timing of tides obtained from tide tables (solid curve). 3 mAB = velocity measurements were made 3 m above sea bottom. Modified after Shepard et al. (1979); with permission of the American Association of Petroleum Geologists.

ranging from 46 to 4200 m by suspending current meters, usually 3 m above the sea bottom (Figure 5.11). These canyons include the Hydrographer, Hudson, Wilmington, and Zaire in the Atlantic Ocean; and the Monterey, Hueneme, Redondo, La Jolla/Scripps, and Hawaii canyons in the Pacific Ocean. Maximum velocities of up- and down-canyon currents commonly ranged from 25 to 50 cm s 1. Keller and Shepard (1978) reported velocities as high as 70–75 cm s 1, velocities sufficient to transport even coarse-grained sand, from the Hydographer Canyon. In the Niger Delta area of West Africa, five modern submarine canyons (Avon, Mahin, Niger, Qua Iboe, and Calabar) have been recognized (Figure 5.12). In the Calabar River, the tidal range is 2.8 m and tidal flows with reversible currents are common (Allen, 1965). In the Calabar estuary, maximum ebb-current velocities range from 60 to 280 cm s 1, and flood current velocities range from 30 to 150 cm s 1. These velocities are strong enough to transport particles of sand and gravel size. The Calabar estuary has a deep-water counterpart that cuts through sediments of the outer shelf and slope, forming the modern Calabar submarine canyon (Figure 5.12). Thus, as they do in the Zaire Canyon to the south, tidal currents are likely to operate in the Calabar and Qua Iboe canyons.

74

Deep-water Bottom Currents and Their Deposits

Submarine canyon Depth: 46–4200 m

Sea level

Ebb flood

Velocity range: 25–50 cm s–1

Current meter: 3 m above base

Figure 5.11 Conceptual diagram showing cross-section of a submarine canyon with ebb and flood tidal currents (opposing arrows). Shepard et al. (1979) measured current velocities in 25 submarine canyons at water depths ranging from 46 to 4200 m by suspending current meters commonly 3 m above the sea bottom. Measured maximum velocities commonly range from 25 to 50 cm s 1. After Shanmugam (2003); with permission from Elsevier.

Av on

Opuama Canyon

6°N

hin

Gu lf

Afam Canyon

Shelf Slope

Ma

Qua lboe Canyon

lab Ca

Ancient canyon 100 km 2°E

4°

ar

bo e Qu

al

Gu in ea

Recent submarine canyons

Niger

of

2°N 4°

6°

8°

10°E

Figure 5.12 Location of modern and ancient submarine canyons in the Gulf of Guinea,West Africa. Outside submarine canyons, the shelf ^ slope break (dashed line) is not only an important physiographic boundary between shelf and slope, but also a major controlling factor of processes on the shelf (e.g., tides and waves) and on the slope (e.g., mass transport). However, within submarine canyons (e.g., recent Calabar Canyon), the shelf ^ slope break does not control processes. Map modified after Petters (1984); with permission from Blackwell Publishing.

75

G. Shanmugam

5.4.3.

Deposits

Sedimentary features indicative of tidal processes in shallow water environments have been well established (e.g., Reineck and Wunderlich, 1968; Klein, 1970; Visser, 1980; Terwindt, 1981; Allen, 1982; Banerjee, 1989; Nio and Yang, 1991; Dalrymple, 1992; Archer, 1998; Shanmugam et al., 2000). Traction structures that develop in shallow-water estuaries also develop in deep-water canyons and channels with tidal currents (Shanmugam, 2003). General characteristics of deep-water tidal deposits are • • • • • • • • • • • • •

heterolithic facies; rhythmic alternation of sandstone/shale couplets (tidal rhythmites); thick (spring)/thin (neap) bundles; alternation of parallel and cross-laminae; double mud layers (Figure 5.13); climbing ripples; cross-beds with mud-draped foresets; bidirectional (herringbone) cross-bedding; sigmoidal cross-bedding (i.e., full-vortex structures) with mud drapes and tangential basal contacts; reactivation surfaces; crinkled laminae; elongate mudstone clasts; flaser bedding;

cm 5 4 3 2 1 0

Figure 5.13 Core photograph showing double mud layers (arrow) in Pliocene sand. Edop Field. After Shanmugam (2003); with permission from Elsevier. A multicolor version of this figure is on the enclosed CD-ROM.

76

Deep-water Bottom Currents and Their Deposits

• wavy bedding; • lenticular bedding; • alternating traction and suspension structures.

5.4.4.

Facies associations in submarine canyons

Submarine canyons are not only unique for providing a protected environment for focusing tidal energy from shallow-marine estuaries to deep-marine canyons, but also prone for generating mass movements (e.g., slides, slumps, grain flows, and debris flows) due to failure of steep canyon walls. The Edop Field is located in the ancient Qua Iboe Canyon (Figure 5.12). The Pliocene Intra Qua Iboe (IQI) reservoir in the Edop Field is a major hydrocarbonproducing siliciclastic reservoir. Based on recognition of a 3-km-wide erosional feature observed on a seismic time slice, a submarine canyon for the Edop reservoir has been documented (Shanmugam, 1997b, his Figure 21). The bulk of the cored interval is interpreted to be deposits of sandy and muddy slumps and debris flows. Some cored intervals are composed of fine to very fine sand with well-developed double mud layers (Figure 5.13), mud-draped ripples, and tidal rhythmites with thick and thin sand layers. These features have been interpreted as products of deep-water tidal currents. Such a close association of mass-flow deposits and deep-marine tidal deposits is an indication of canyon-fill facies (Figure 5.14).

5.4.5.

Reservoir potential

Recognition of tidal facies in deep-water successions has implications for reservoir potential. For example, in channel-mouth environments, down-slope turbidity currents are likely to develop depositional lobes, whereas bi-directional tidal bottom currents are likely to develop elongate bars (Shanmugam, 2003). Turbidite Double mud layers (tidal) Floating mudstone clasts and quartz granules (debris flows)

Contorted layers and floating quartz granules (slumps) Double mud layers (tidal) Mud-draped ripples (tidal)

Figure 5.14 Facies association showing interbedded occurrence of double mud layers (tidal origin), floating mudstone clasts and quartz granules (debris-flow origin), double mud layers (tidal origin), contorted layers and floating quartz granules (slump origin), and double mud layers with mud-draped ripples (tidal origin). In the rock record, such a facies association may be used as evidence for deposition within submarine canyons. Modified after Shanmugam (2003); with permission from Elsevier.

77

G. Shanmugam

lobes are aligned perpendicular to the channel axis, whereas tidal bars are aligned parallel to the channel axis. Depositional lobes are likely to be much larger than channel width, whereas tidal sand bars are thought to be much smaller than channel width. Deep-water elongate tidal bars are speculated to be analogous to tidal bar sands that develop in shallow-water estuarine environments (see Shanmugam et al., 2000). In frontier exploration areas, an incorrect use of a turbidite-lobe model (with sheet geometry) instead of a tidal bar model (with bar geometry) will result in an unrealistic overestimation of sandstone reservoirs.

5.5. 5.5.1.

I NTERNAL W AVES AND T IDES (BAROCLINIC C URRENTS)

Nomenclature

A plethora of nomenclature is in use for internal waves and tides. In particular, the term baroclinic has been used by different authors with different meanings (Wunsch, 1996). Thus it is useful to explain these terms to minimize confusion. Surface waves, caused by wind (meteorological force) blowing over the water surface (Komar, 1976), develop at the interface between water and air. Internal waves, first reported by Ekman (1904), are gravity waves that oscillate along the interface between two water layers of different densities (Figure 5.15). These waves are common phenomena in coastal seas, fjords, lakes, and the atmosphere. In shallow-water shelf environments, waters can range from well mixed to density-stratified types. Most shelf waters are vertically well mixed. In deep-water environments, however, most of the ocean is vertically stratified, with an upper low-density layer and a lower high-density layer (Figure 5.15). The interface between layers of different densities (i.e., pycnocline) can be caused either by differences in temperature (i.e., thermocline) or by salinity (i.e., halocline). Navrotsky et al. (2004) made observations of internal waves and spatial inhomogeneities of a thermohaline structure during the warm season on the continental slope and shelf of the Sea of Japan. Santek and Winguth (2005) documented internal waves, induced by the 2004 Indian Ocean tsunami, along the eastern continental slope off Sri Lanka. Deep water

Shallow water Surface waves (meteorological) Shelf

Sea level

Internal waves and tides (astronomical)

1. Low-density (warm) layer 2. High-density (cold) layer

1 2

Slo

Thermocline

pe

Basin

Figure 5.15 Schematic diagram showing meteorological surface waves in shallow-water shelf environments and astronomical internal waves in deep-water slope and basinal environments. Internal waves, occurring along a thermocline, can extend onto shelf environments. Internal waves typically have much higher amplitudes than surface waves. Diagram based on concepts of Inman et al. (1976) and Maxworthy (1979). Not to scale.

78

Deep-water Bottom Currents and Their Deposits

Figure 5.16 Satellite image of internal waves in the Sulu Sea between the Philippines (to the northeast) and Malaysia (to the southwest). Sunlight highlights delicate curving lines of internal waves moving to the north toward Palawan Island. The Sulu Sea is stratified with water layers of differing densities. Unlike surface waves, however, internal waves can stretch tens of kilometers in length and move throughout the ocean for several hours. This true-color Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) image was acquired on April 8, 2003. Image courtesy Jacques Descloitres, MODIS Land Rapid Response Team at NASA/GSFC. Uniform Resource Locator (URL): http://earthobservatory.nasa.gov/ Newsroom/NewImages/images.php3?img_id=15334 (accessed May 12, 2007). A multicolor version of this figure is on the enclosed CD-ROM.

Internal waves typically have much lower frequencies and higher amplitudes than surface waves because the density difference between two water layers is typically much less than the density difference between water and air. Unlike surface waves, internal waves can stretch tens of kilometers in length (Figure 5.16). They can move throughout the ocean for several hours. They have their greatest wave height at intermediate depths and their greatest velocities at the bottom (LaFond, 1962). Internal waves are of significance not only for maintaining ocean circulation by downward mixing of heat, but also for sustaining biological productivity by supplying nutrients. Internal waves that correspond to periods of tides are called internal tides (Shepard, 1975). It is important to distinguish the surface (barotropic) tide from the internal (baroclinic) tide. This is because sedimentological aspects of deposits of surface tides are well established (Alexander et al., 1998), whereas those of internal tides are not. The passage of tropical cyclone Bobby in 1995 over the Western Australian Shelf influenced the generation of internal tides at 300 m depth (Davidson and Holloway, 2003). The currents associated with internal tides are

G. Shanmugam

79

termed baroclinic currents or motions. These tidal currents tend to create turbulent mixing in the stratified water column as well as at the sea floor. The term near-inertial wave has been used for internal waves that correspond to periods of upwelling (Federiuk and Allen, 1996). Internal solitary waves, consisting of a single isolated wave, are common in stratified fluids. The term soliton is commonly used as a synonym for solitary waves. Although the term ‘‘soliton’’ is strictly a mathematical solution of a non-linear internal-wave theory, it has become common practice in the offshore drilling industry to use the term to describe observations of large-amplitude nonlinear internal waves in the ocean (Hyder et al., 2005). The origin of internal solitary waves has been attributed to a number of causes, including sea-floor topography (Maxworthy, 1979; Farmer and Armi, 1999). Because internal solitons are hazardous to offshore drilling operations (Fraser, 1998), a clear understanding of these internal waves has practical implications for the cost and the safety of offshore drilling operations.

5.5.2.

Velocity

Deep-water bottom currents have been attributed to internal waves in offshore California (Emery, 1956). Wunsch (1969) proposed amplification of near-bottom velocities as internal waves propagate over a shoaling bottom. Brandt et al. (2002) reported results of high-resolution velocity measurements carried out by means of a vessel-mounted acoustic Doppler current profiler on the 12 November 2000 in the equatorial Atlantic, at 44°W between 4.5°N and 6°N. The data showed the presence of three large-amplitude internal solitary waves. The pulse-like intense solitary disturbances propagated, perpendicular to the Brazilian Shelf, toward the north–northeast. These internal waves were characterized by maximum horizontal velocities of about 2 m s 1 and maximum vertical velocities of about 20 cm s 1. Shepard et al. (1979), who studied bottom currents in submarine canyons, documented that internal waves advance in both up- and down-canyon directions. Measured values of velocity reach up to 1 m s 1 in the up-canyon direction and 265 cm s 1 in the down-canyon direction. Shepard (1975) suggested that internal waves, which occur in canyon depths of up to 3500 m,were mostly tidal in origin (i.e., internal tides). In a stratified ocean, internal tides are generated commonly above an area of steep bathymetric variation, such as the shelf break. An example is the Bay of Biscay, where the internal tides are the most energetic (Baines, 1982). Internal tides travel slowly compared with surface gravity waves. Hyder et al. (2005) made observations of internal solitons that occurred between January and April 1998 at a water depth of 440 m northeast of the Andaman Islands, Bay of Bengal. Their observations indicated the occurrence of internal solitons with thermocline depressions of up to 50 m and an upper-layer current velocity of up to 120 cm s 1. These solitons only occurred during spring tides, when the tidal range exceeded 1.5 m. The advances in tidal mapping afforded by the Topex/Poseidon satellite have allowed Egbert and Ray (2000) to answer some longstanding questions about tidal energetics. Analyses of the altimeter-derived cotidal charts reveal that most tidal energy is dissipated in shallow seas, but about 25 30% of the global energy

80

Deep-water Bottom Currents and Their Deposits

dissipation occurs generally near rugged bottom topography in the open ocean. St. Laurent et al. (2003) also reported that internal tides are generated commonly at regions where the barotropic tidal current encounters variations in bottom topography. Semi-diurnal internal tidal currents are likely major factors in shaping continental slopes. Cacchione et al. (2002) have discussed the intensification of near-bottom water velocities, caused by reflection of semi-diurnal internal tides, and their role on sedimentation patterns and bottom gradients of continental slopes off northern California and New Jersey. Oceanographic studies have shown that the Makassar Strait in Indonesia is dominated by strong tidal currents, internal waves/tides, and solitons (Ray et al., 2005). The Indonesian Throughflow (ITF), which passes through the Makassar Strait, is a series of ocean currents that flow from the tropical Pacific Ocean through the Indonesian Seas into the Indian Ocean. It transports nearly 10 Sv (1 Sv or Sverdrup = 106 m3 s 1) of the Pacific Ocean water into the Indian Ocean. The ITF, a thermocline flow, is stratified along the Makassar sill depth of 680 m (Gordon, 2005). Current meter measurements from two moorings in the Labani Channel recorded velocities in excess of 50 cm s 1 at 250 m (Wajsowicz et al., 2003). At these velocities, medium-grained sand can be eroded and transported by baroclinic tidal currents.

5.5.3.

Deposits

The potential significance of shoaling internal waves for causing sediment movement on continental shelves and slopes has been discussed by Cacchione and Southard (1974). Laboratory experiments confirmed that shoaling interfacial waves could generate ripples and larger bedforms in artificial sediment (Southard and Cacchione, 1972). Stride and Tucker (1960) attributed the development of modern sand waves near the shelf edge to internal waves. Karl et al. (1986), using sparker profiles, documented sand waves in the heads of submarine canyons of the Bering Sea. In this case, a surface sediment sample (C1) was composed of 19% gravel, 76% sand, and 5% mud. The modal class of this sample was fine sand. However, no sedimentary structures were described from the cores from these sand waves. Karl et al. (1986) speculated that internal waves were responsible for the origin of sand waves. They also suggested that delivery of large volumes of fresh water and large quantities of sediment directly to the heads of submarine canyons during periods of low sea level might have enhanced the propagation of highfrequency internal waves. Gao et al. (1998) interpreted ancient strata with bidirectional cross-bedding, flaser bedding, wavy bedding, and lenticular bedding as deposits of internal tides based on associated deep-water turbidite and slump facies. The key to interpreting deposits of ‘‘internal tides’’ or baroclinic currents in the rock record is the evidence for tidal currents in a stratified deep ocean. Without that evidence for density stratification, there is no difference between a tidal deposit formed by surface (barotropic) tide in a shallow-marine shelf and a tidal deposit formed by internal (baroclinic) tide in a deep-marine slope or canyon environment. Furthermore, tidal bottom currents in submarine canyons may be unrelated to density stratification. Until we develop objective sedimentological criteria for

81

G. Shanmugam

recognizing deposits of internal tides of stratified water bodies, it is preferred to classify deep-water deposits with tidal signatures as products of ‘‘deep-water tidal bottom currents’’ rather than of ‘‘baroclinic currents’’ (i.e., internal tides). This is a topic for future research.

5.6.

C ONCLUSIONS

Four basic types of deep-water bottom currents and their deposits are discussed: (1) thermohaline-induced geostrophic bottom currents, (2) wind-driven bottom currents, (3) deep-water tidal bottom currents, and (4) internal waves and tides (baroclinic currents). • A distinctive attribute of reworked sands by thermohaline-induced and winddriven bottom currents is their traction structures. Common sedimentary features of these traction currents include cross-bedding, current ripples, horizontal lamination, sharp upper contacts, and inverse grading. These sands also exhibit internal erosional surfaces and mud-offshoots indicating oscillating energy conditions. • Deposits of deep-marine tidal bottom currents are sand/mud rhythmites, double mud layers, climbing ripples, mud-draped ripples, alternation of parallel and cross-laminae, sigmoidal cross-bedding with mud drapes, internal erosional surfaces, lenticular bedding, and flaser bedding. These features represent alternating events of traction and suspension deposition. • There are no objective sedimentological criteria to recognize deposits of internal tides in the rock record. • Sedimentary structures of bottom-current deposits occur in discrete units, not as a predictable vertical sequence.

ACKNOWLEDGMENTS I thank M. Rebesco, A. Camerlenghi, A. J. van Loon, and D. W. Kirkland for helpful comments.

C H A P T E R

6

D YNAMICS OF THE B OTTOM B OUNDARY L AYER S. Salon1, A. Crise1 and A.J. Van Loon2 1

Istituto Nazionale di Oceanografia e Geofisica Sperimentale, Sgonico (TS), Italy Geological Institute, Adam Mickiewicz University, Poznan, Poland

2

Contents 6.1. Introduction 6.2. Spectral Windows 6.3. Characteristics of the Bottom Boundary Layer 6.3.1. General characteristics 6.3.2. Influence of wind stress 6.3.3. Interactions with the sea floor and suspended particles 6.4. Analytical Approach of the BBL 6.4.1. The Ekman layer 6.4.2. The viscous sub-layer (bed layer) 6.4.3. The logarithmic layer 6.5. Conclusions Acknowledgements

6.1.

83 85 87 87 88 88 88 89 94 95 96 97

I NTRODUCTION

The water column of the ocean can be roughly divided into the following three main layers, which differ from one another because of the role played by friction: (1) the surface boundary layer, which is due to atmospheric forcing that is expressed by a frictional force (wind stress) that is felt by the ocean surface; (2) the ocean interior, which is essentially frictionless and which is characterized by the balance, commonly defined as geostrophic balance, between the horizontal pressure gradients and the Coriolis force (which forces water flows on the northern hemisphere to move to the right, and on the southern hemisphere to move to the left)1; and (3) the boundary layer near the sea floor, where friction between the water and 1

The term ‘‘geostrophy’’ comes from Greek  or  = Earth, and o = turning. The attribute ‘‘geostrophic’’ is always related to flows characterized by an equilibrium between the horizontal pressure gradient and the Coriolis force. In the ocean’s interior, currents with horizontal dimensions of over a few tens of kilometres and durations of more than a few days are in geostrophic balance. Equations describing the geostrophic balance can be used to evaluate currents at depth (see, for example, Stewart, 2005).

Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00206-9

Ó 2008 Published by Elsevier B.V.

83

84

Dynamics of the Bottom Boundary Layer

the bottom dissipates the energy from the ocean’s interior (Pedlosky, 1987; Cushman-Roisin, 1994). The ocean currents flowing in the boundary layers near the surface and the bottom meet discontinuities in the form of the air/water interface and the sea floor, respectively. The presence of such a frictional boundary affects the current velocity of the ocean’s interior and the turbulent structures. At the contact with the sea floor, the current velocity is forced to diminish to zero, so that the vertical profile of the velocity undergoes shearing (‘‘shear’’ is defined in this case as the way the horizontal component of velocity changes with distance from the boundary), particularly in the water layer close to the sea floor; the thickness of such a layer is a dynamic property of this boundary layer. The turbulent processes transfer and mix the physical properties of the near-boundary flows (momentum, energy and materials such as living organisms, atmospheric gases and sediments), thus giving rise to well-mixed boundary layers at steady state. The surface and the bottom boundary layers are therefore the pathways of communication between, respectively, the atmosphere and the sea, and the sea floor and the sea; exchange of particles, chemicals and organisms takes place in these boundary layers. The simple picture of a stratified ocean that is well-divided into three layers is valid for many regions of the open ocean, but it becomes, obviously, more complex when density gradients, a complex topography, lateral boundary currents or other features are involved. A different approach, based not on a steady description of phenomena along the vertical dimension but on their temporal variability then is required to understand the underlying processes and their morphological, sedimentary and granulometric effects. Moreover, the transport processes near the sea floor that are associated with the sediment dynamics are determined by the turbulent structures that are related to the dimensions (spatial scales) and duration (time scales) of the ocean currents (Le Couturier et al., 2000). The dynamics of the bottom boundary layer itself is driven by the ocean currents: the kinetic energy, which is mainly contained in the large-scale currents and which is ultimately derived from the solar irradiation that drives the global thermohaline circulation,2 is transferred along the so-called turbulent energy cascade down to the smallest dimensions (molecules) and, finally, is converted by molecular dissipation into heat. Turbulent phenomena are characterized by a wide range in dimensions, namely from molecular size to extremely large eddies. It should be noted in this context that eddies, identified as vortexes in fluid dynamics, are described in ocean circulation as areas of closed swirling motion. Their diameters range from a few tens of kilometres (mesoscale eddies) to a few hundreds of kilometres (synoptic eddies), and they can develop both at the surface and at great depth. Turbulence can be seen as a superposition of these eddy structures, which tend to mix the ocean water masses, thus homogenizing them with respect to 2

The thermohaline circulation is due to processes at the sea surface – such as heating and cooling, influxes of fresh water, evaporation, precipitation, river runoff, sea-ice formation and melting – that induce differences in the temperature and salinity – and thus the mass – of the oceanic waters. The thermohaline circulation extends from the upper oceanic layer to the bottom (see also Zenk, 2008, this volume).

85

S. Salon et al.

temperature, momentum, particle concentration, etc. The efficiency of mixing is related to the energy transfer by the various processes that take place at different scales: kinetic energy is supplied by large-scale processes, not affected by viscosity, whereas the kinetic energy of flows of ever smaller dimensions is eventually – as mentioned above – converted into heat by molecular (viscous) dissipation. The concept of such a ‘‘turbulent cascade’’ was clearly illustrated by Richardson (1922): Big whirls have little whirls Which feed on their velocity, And little whirls have lesser whirls, And so on to viscosity. Hence, the energy of the turbulent structures that characterize the dynamics of the bottom boundary layer, in combination with their interaction with organisms and sediments, is governed by the turbulent energy cascade. The energy transfer through the various sizes of turbulent eddies is commonly described in terms of wave number (or frequency) spectra: energy production, by means of mechanical instabilities in the fluid, occurs at small wave numbers, whereas viscous dissipation occurs at large wave numbers. After a brief discussion on the concept of spectral windows, we will deal with the description of the general characteristics of the bottom boundary layer, whose analytical investigation constitutes the core of this chapter.

6.2.

SPECTRAL W INDOWS

Spectral windows (Nihoul and Djenidi, 1987) represent typical scales for time and space relative to a certain process under investigation. This concept has been adopted to define hydrodynamic phenomena in marine systems. A schematic distribution of spectral windows in time and space for the most relevant marine dynamical processes is shown in Figure 6.1. In this case, two main spectral windows can be defined, sub-divided by the diurnal time scale: the upper part of Figure 6.1 includes physical processes such as surface and internal tides, internal and inertial waves (the latter are periodic, circular motions related to the Coriolis force, with periods depending on the latitude; at mid-latitudes the inertial period is 17 h). The lower part of Figure 6.1 encompasses processes associated with mesoscale phenomena and synoptic storms, meanderings of fronts (an oceanic front is defined as the transition zone between water masses of different density; frontal currents are driven by the strong temperature and/or salinity gradients that occur in the frontal zone), eddies and gyre circulations developed within the major currents (scales of weeks to months) and at the ocean basin scale (scales of several years). This spectral window can be extended down to the thermohaline circulation at the global scale, which may take 1000 years to overturn completely. A sub-window can be further identified for processes shorter than few hours (uppermost part of Figure 6.1). This includes, among others, some types of waves and three-dimensional turbulent structures (surface waves, Langmuir cells). The processes belonging to this sub-window are commonly considered as ‘‘stirring mechanisms’’

86

Dynamics of the Bottom Boundary Layer

1 mm 1s

1 cm

1 dm

1m

Molecular processes

10 m

100 m

1 km

10 km

100 km 1000 km

Surface waves

1 min Individual movement

Turbulent patch size

Langmuir cells

1h Inertial/internal waves 1 day

Internal tides

Surface tides

Diurnal Plankton migration

Synoptic storms

1 week Phytoplankton patch 1 month Zooplankton patch 1 year

Coastally trapped waves Fronts and eddies

Seasonal MLD and biomass cycles Gyre circulation

Mesoscale phenomena 10 years

Figure 6.1 The relevant time and length scales of several physical and biological processes in oceans (from Dickey, 1991; with permission from American Geophysical Union).

that increase the contact frequency between water parcels, and therefore increase the mixing rate of these parcels, and consequently of the water properties. As shown in Figure 6.1, spectral windows approach can be adapted also to biological phenomena, that are tightly related with physical processes. The individual movement and the turbulent patch size characterizing the micro-organisms belong to the upper spectral window, while phenomena related with plankton migration, phytoplankton and zooplankton patches have time scales longer than the diurnal one. The turbulent cascade of energy therefore proceeds from the bottom right corner of the figure to the top left corner, where molecular processes are involved in the turbulent dissipation due to viscosity. Following this approach, bottom currents variability (Shanmugam, 2008, this volume; Zenk, 2008, this volume) can have time scales of about 20 days (e.g. Camerlenghi et al., 1997) and persist for a couple of months (e.g. Giorgetti et al., 2003). Together with deep cyclones, that may have lifetimes of 6–9 weeks (e.g. Savidge and Bane, 1999), they belong to the lower part of Figure 6.1. Conversely, turbidity currents were observed to develop in few hours (e.g. Khripounoff et al., 2003; Xu et al., 2004) and therefore belong to the upper spectral window of Figure 6.1. Jankowski et al. (1996) studied the behaviour of the sediment that is brought into suspension due to deep-sea mining, with the primary objective to understand and model the mesoscale processes that are superimposed upon the stable geostrophic component. Fourier analysis of the experimental data collected

87

S. Salon et al.

from a bottom boundary layer at a depth of over 4000 m showed the following three components: (1) a slowly varying geostrophic current that could be considered constant over a duration of a week; (2) persistent long waves characterized by an inertial and a sub-inertial period (in the area of the experiment, the inertial period is 98 h); and (3) semi-diurnal tides. Spectral analysis of the velocity field responsible for tidal propagation showed that the main energy peaks were found at near- to super-inertial radial frequencies. The dynamical view of a combination of different components co-existing in the same oceanic region, whether in the surface boundary layer or in the bottom one, is therefore much more appropriate than one dividing the ocean in static layers. The latter is, however, extremely useful for an analytical description of the characteristics of each layer, as we will show below.

6.3. 6.3.1.

CHARACTERISTICS OF THE B OTTOM BOUNDARY L AYER

General characteristics

As pointed out by Kanta and Clayson (2000), the boundary layer just above the sea floor should, strictly speaking, be called ‘‘bottom boundary layer’’ in the shallowmarine environments of the continental shelf (cf. Soulsby, 1983; Grant and Madsen, 1986) and ‘‘benthic boundary layer’’ in the deep ocean (cf. Wimbush and Munk, 1970; Armi and Millard, 1976). It should be noticed that the bottom boundary layer (BBL) can extend over the entire water column on shallow coastal shelves, thus being coupled with the surface boundary layer. In such a case, the turbulent flow is very complex due to the presence of a wide variety of processes, and is beyond the scope of this chapter. In the following, we will hence describe a generic BBL, no matter whether in the deep ocean or on the continental shelf. The BBL described will be considered as separated from the surface dynamics by a well-defined geostrophic interior, with dynamics that results mainly from influences by the current in the ocean interior, by the sea-floor topography, and by sediment-transport processes. In general, a BBL is characterized by currents of a few centimetres per second (in deep water usually not over 20–40 cm s1), but at the contact with the sea floor the current velocity is zero (no-slip condition). This is because the velocity shear in the BBL is particularly strong near the sea floor, and diminishes upwards, to become zero when approaching the ocean interior, where the velocity field is approximately constant and determined by the geostrophic balance. The thickness of the BBL is of the order of 10 m in the deep ocean (depths up to 4000 m; Lueck, 2001), but under high-velocity current conditions it may reach a thickness of 40 m (Kanta and Clayson, 2000) or even involve the whole water column in shallow-water areas, where friction and currents are relatively strong compared to the deep ocean. When the dynamics of the BBL is considered over a nearly horizontal sea floor, far from mid-ocean ridges or other relief forms, a homogeneous fluid,

88

Dynamics of the Bottom Boundary Layer

vertically uniform in temperature and salinity, may safely be assumed. Kanta and Clayson (2000) have already pointed out that the benthic heat fluxes from the sea floor to the water column are so small (0.1 W m2) that they do not influence the dynamics, and therefore may be neglected. This assumption is supported by observations made over large regions of the oceans more than two decades ago (D’Asaro, 1982). Similar data have also been presented by Jankowski et al. (1996).

6.3.2.

Influence of wind stress

Grant and Madsen (1986) have discussed the characteristics of a continental-shelf BBL driven by wind stress. They deal with the shallow BBL with a spectralwindow approach, in terms of wave propagation at different frequencies, from the surface down to the bottom. They describe the currents as low-frequency wavy motions that can penetrate the geostrophic interior and then influence the BBL, whereas the higher-frequency surface waves remain confined in the surface boundary layer. Their approach to the BBL phenomenon, as representing a response to wind forcing and pressure differences may be applied also to a deep-sea BBL (on which the present contribution is focused), though generally decoupled from the surface dynamics. Surface perturbations can, in fact, also be transferred to depth through internal waves, as these propagate both horizontally and vertically (Gill, 1982; D’Asaro, 1989; D’Asaro et al., 1995).

6.3.3.

Interactions with the sea floor and suspended particles

The abyssal circulation and, as a consequence, the dynamics of the BBL are far from constant, and this should be taken into account when the interaction between currents and sediment is considered. In particular, the HEBBLE experiment (HighEnergy Benthic Boundary Layer Experiment: Hollister and Nowell, 1991) showed a dynamic BBL that plays a significant role in the mixing of the water column and that largely affects the sedimentary processes. A peculiar feature of the near-bottom currents is their sensitivity for the seafloor roughness that is caused by the micro-topography resulting from biological activity and sediment characteristics that are due to current-induced suspension and resettling (Grant and Madsen, 1986). In particular, in the deep ocean, large amounts of sediment can be removed from the sea floor and re-suspended into the water column when the bottom stress associated to strong currents exceeds a certain threshold value that depends on the nature and density of the sediments (Kanta and Clayson, 2000). If the concentration gradient of suspended sedimentary particles surpasses a critical value, it can change the characteristics of the density field, introducing a stratification within the water column and consequently influencing the dynamics of the BBL. Sedimentary particles may remain in suspension because turbulent processes in the water column act against the gravitational force that otherwise would cause resettling of the sediments.

89

S. Salon et al.

6.4.

A NALYTICAL APPROACH OF THE BBL

The following summary describes the BBL from an analytical point of view. It is largely based on the reviews by Soulsby (1983) and Lueck (2001). The BBL is usually described to consist of three different layers (Figure 6.2): (1) a bottom Ekman layer that is due to the equilibrium between the Coriolis force, the pressure gradient and the turbulent friction; (2) a very thin (order of magnitude: millimetres) viscous layer, close to the boundary with the solid substratum, where only molecular friction and the pressure gradient are significant; and (3) a transitional ‘‘logarithmic’’ layer between the two layers just mentioned, where the turbulent friction prevails over the other forces and is in equilibrium with the pressure gradient.

6.4.1.

The Ekman layer

Scientific literature regarding atmospheric or oceanic boundary layers commonly refers to a seminal paper published in 1905 by V.W. Ekman. He analytically described the way in which the surface or bottom stress is transferred away from the boundaries towards the oceanic or atmospheric interior. The Ekman equations are an approximation of the equations of motion, and are formally based on the

Ocean interior

Z

Bottom Ekman layer

Z Y

X

Viscous layer Logarithmic layer

Z =0 U

Figure 6.2

Sketch of the vertical structure of the bottom boundary layer.

90

Dynamics of the Bottom Boundary Layer

balance between the pressure gradient that drives the flow, the Coriolis force and turbulent friction. The Ekman layer constitutes the part of the BBL where these three forces are in equilibrium. The Ekman equations are valid under steady conditions (no dependence on time) and are based on the eddy-viscosity assumption, which can be summarized as follows. In analogy with molecular viscosity, that directly relates the viscous stresses with the mean shear of a flow, eddy (or turbulent) viscosity directly relates the turbulent stresses to the mean shear (Boussinesq, 1877), thus defining the so-called first-order turbulence closure. Whereas the molecular viscosity, , is a parameter of the fluid (  106 m2 s1 for the water), the eddy viscosity, T, is a property of the flow: it may change in space and with time, and its value can be as high as four orders of magnitude larger than that of . It is very difficult, however, to conform to such an assumption under actual environmental conditions, because turbulence and stratification influence the water column to different degrees.

6.4.1.1. The Ekman solution The Ekman solution is extremely important to understand the dynamics of boundary layers, whether the boundary is the ocean’s surface (surface Ekman layer) or the ocean’s floor (bottom Ekman layer). The way in which this solution is achieved, can be summarized as follows. We consider here for the purpose a coordinate system where the positive X-axis is directed along the bed shear stress, the Z-axis is perpendicular to the bed (Z = 0) and directed to the ocean interior, and the Y-axis is perpendicular to the plane defined by X and Z in a right-hand frame. Due to the choice of the reference frame, the bed shear stress is thus defined as  0  ð X ;  Y ÞjZ ¼ 0 ¼ ð X ð0Þ; 0Þ: It may be worthwhile to note here that the shear stress,  XZ, is the rate of transfer of momentum in the X-direction across the X–Y-plane (similar definitions can be applied to the other components of the shear tensor,  XY and  YZ). It basically describes the frictional force that water at one level exerts on the level below. If the lower level is the sea floor, the shear stress exerted on it is referred as  0, and if it is higher than a certain value, it can induce movement of the bottom sediment. The shear stress ( XZ) is defined – from the equations of motion – as the sum of the viscous stresses due to the molecular friction resulting from the mean vertical shear and the Reynolds shear stresses that are due to the fluctuations in turbulent velocity:  XZ ¼ 

dU  u 0 w 0 dZ

where  ðdU=dZ Þ is the viscous stress that depends on the molecular viscosity ( ) and the mean velocity profile U = U(Z), while u 0 w 0 is the Reynolds stress. The latter is identified as the covariance of the velocity fluctuations in the X and the Z directions (equal to the average of the product of the velocity fluctuations, u 0 and w 0 ), multiplied by the fluid density, . The Reynolds stress equals the shear stress,  XZ, except in a very thin layer near the bed where viscous stresses dominate (the

91

S. Salon et al.

viscous sub-layer, see above). The eddy-viscosity assumption relates the Reynolds stress with the mean vertical shear: u 0 w 0 ¼ T ðdU=dZ Þ and therefore  XZ ¼ ð þ T Þ

dU : dZ

In the following, we will denote  XZ and  YZ by  X and  Y, respectively. At a certain distance from the sea floor, ocean currents are affected by the bottom friction, t, and depart from the geostrophic equilibrium that holds in the ocean interior. The value of this departure determines the thickness of the Ekman layer, where a balance among the Coriolis force, the pressure gradient and the friction is established. The equations of motion in the Ekman layer, written in the horizontal coordinates, read as follows: f V ¼  fU¼

1 @P 1 @ X þ  @X  @Z

1 @P 1 @ Y : þ  @Y  @Z

The left-hand sides of the above equations represent the Coriolis force (divided by the fluid density, ), while the right-hand sides are the sum of the pressure gradient and the friction, expressed as the variation of the stress in the vertical direction. The Coriolis force is formally given by the cross-product f  U, where U is the velocity field and f is the Coriolis parameter, which has only a component parallel to Z, and the amplitude of which is defined as f = 2 sin (where  is the angular velocity of the Earth, equal to 2/Trot, where Trot is the period of Earth rotation, and the latitude; at mid-latitudes, f  104 rad s1). The Coriolis force tends to deviate the flow to the right on the northern hemisphere (to the left on the southern hemisphere), and the pressure gradient acts to restore the original direction. One may safely consider the density field in the boundary layer as homogeneous; this is consistent with a pressure gradient that is independent of the height within the layer. Vertical velocities are much smaller than the horizontal current velocity and may be neglected. The boundary conditions of the Ekman layer far above the bottom (Z ! 1), i.e. in the ocean interior, constrain the flow (U, V) to match the geostrophic flow (Ug, Vg), and the bottom stress ( X,  Y) to vanish; at Z = 0, in contrast, the no-slip condition U = V = 0 is imposed. Combining the equations of motion with the geostrophic balance (that is determined by the equations of motion at Z ! 1), f Vg ¼  f Ug ¼ 

1 @P  @X

1 @P  @Y

92

Dynamics of the Bottom Boundary Layer

the equations governing the Ekman layer are thus obtained:   1 @ x f V  Vg ¼  @Z   1 @ y f U  Ug ¼ :  @Z The solution to these equations, originally proposed by Ekman (1905), is based on the eddy-viscosity assumption, i.e.  X ¼  T dU=dZ and  Y ¼  T dV =dZ, where the eddy viscosity, T, is constant and the viscous stresses are negligible. Considering, as an example, a geostrophic flow with Vg = 0, we obtain the following solution for the current in the bottom Ekman layer (Cushman-Roisin, 1994):    Z Z=D U ¼ Ug 1  e cos D   Z V ¼ Ug e Z=D sin D pffiffiffiffiffiffiffiffiffiffiffiffiffi where D ¼ 2 T =f represents the distance over which the solution approaches the geostrophic flow, i.e. the layer thickness above the sea floor for which frictional forces are much smaller than the Coriolis force, and is therefore a measure of the thickness of the Ekman layer. The vertical profile of the velocity is known as the ‘‘Ekman spiral’’ (Figure 6.3): the horizontal velocity U ¼ ðU; V Þ in the Ekman layer rotates, due to the effect of friction that acts against the current, leftward (on the northern hemisphere) with respect to the geostrophic flow, Ug, with increasing water depth, diminishing to zero at Z = 0. The eddy-viscosity assumption is, however, still far from sufficient to describe the conditions in a fully turbulent oceanic boundary layer in an adequate way. The mass transport in fluid dynamics is defined as the integral of the velocity field along the vertical dimension: in the case of the bottom Ekman layer, this

U(Z1)

U(Z2)

0.2

V

U(Z = 0) U(Z3) Ug

0 0.2

0.4

0.6

0.8

1

U

Figure 6.3 The Ekman spiral. Polar diagram of the current-velocity vector, U, in the bottom Ekman layer. The current-velocity vector is plotted at different distances from the sea floor: Z1 < Z2 < Z3. This figure is drawn for the northern hemisphere ( f > 0); a reversed plot is valid for the southern hemisphere.

S. Salon et al.

93

transport is always perpendicular to the geostrophic flow direction (to the left on the northern hemisphere, the opposite on the southern one; Cushman-Roisin, 1994). The Ekman transport generated by contour currents in a stratified ocean on a sloping bottom can influence the flow within the bottom boundary layer. The bottom Ekman transport advects the initial density gradient and creates a buoyancy force that opposes the cross-slope flow. This force gives rise (in absence of external forces) to the exponential decay of the transport from its initial value and in turn, reducing the speed-dependent bottom drag, limits the slowdown of the interior flow (see for details MacCready and Rhines, 1993; Durrieu de Madron and Weatherly, 1994). 6.4.1.2. Thickness of the Ekman boundary layer According to Lueck (2001) and numerous earlier authors (see among others Wimbush and Munk, 1970; Weatherly, 1972; Weatherly and Martin, 1978; Bird et al., 1982; Soulsby, 1983; Thorpe, 1988; Cushman-Roisin, 1994; Jankowski et al., 1996), the thickness of the turbulent Ekman boundary layer (or Ekman height, hE) is proportional pffiffiffiffiffiffiffiffiffiffi to the ratio u*/f, where u* (which is called ‘‘friction velocity’’) is equal to  0 =, and where f is the Coriolis parameter.3 The friction velocity, u*, is a characteristic of the flow and is basically determined by the turbulent fluctuations of the current velocity close to the solid boundary. To estimate the height of the Ekman layer, it is necessary to formulate the friction velocity u* (or the bed shear stress  0) in terms of variables that can be actually measured or calculated, such as the free-stream velocity, a vertically averaged velocity or a reference velocity at some fixed distance from the bottom. The simplest approach to evaluate the bed shear stress,  0, is to consider it as a quadratic function of an appropriate velocity, U:  0 ¼ CD U 2 , where CD is a drag coefficient that depends on the bottom characteristics. To estimate the order of magnitude of hE, Lueck (2001) adopted a geostrophic velocity Ug  0.1 m s1 as the reference velocity (U) and a drag coefficient CD  0.002 (values also used by Soulsby, 1983, and Thorpe, 1988), obtaining hE  u* =f ¼ 45 m (with a midlatitude value for f ), corresponding roughly to 1% of the average ocean depth. It is noteworthy that such a formulation results also in u* =Ug ¼ CD½  1=22; values between 1/20 and 1/30 have often been mentioned (Weatherly et al., 1980; D’Asaro, 1982; Bird et al., 1982; Gust and Weatherly, 1985; Jankovski et al., 1996). Weatherly and Martin (1978), Bird et al. (1982), Thorpe (1988) Cushman-Roisin (1994) and Jankovski et al. (1996) report hE ¼ u* =f for the case of neutral stratification, with the von Ka´rma´n constant, = 0.4, as a proportionality constant. For the case of stratified conditions with a specific value of the buoyancy frequency, N, Weatherly and Martin (1978) suggested the formulation hE ¼ 1:3u* =f ð1 þ N =f Þ1=4 (in a stratified fluid, the buoyancy frequency, N – also known as the Brunt–Va¨isa¨la¨ frequency – is the natural frequency of the vertical 3

As previously defined, a measure of the thickness of the Ekman layer is D. Considering that the eddy viscosity can be expressed as a product of a velocity and a distance, we can choose, respectively, the friction velocity and the Ekman depth, and express T  u*D. Therefore we obtain hE = D  u*/f.

94

Dynamics of the Bottom Boundary Layer

oscillations of fluid particles; it is expressed as N 2 ¼ ð g=Þðd=dZ Þ, where g is the acceleration due to gravity and  = (Z) is the density that depends on the vertical coordinate, i.e. the height in the atmosphere or the depth in the ocean; in the absence of stratification, N = 0). The Ekman height, hE, that is a measure of the Ekman layer, can thus be considered as the thickness of the water column where the bottom friction is important; for Z > hE, the Coriolis force dominates over friction (geostrophic balance), whereas for Z << hE the contrary is true. The latter case is detailed below.

6.4.2.

The viscous sub-layer (bed layer)

The Ekman model is not well suited to describe the dynamics of the BBL very close to the sea floor (Z << hE); a proper description should also take into account the irregularities of the sea floor, defined as roughness. We will first deal with the case of a horizontal, flat bottom. Because of the absence of normal flow at the water/sea floor interface (Z = 0), vertical turbulent fluctuations are strongly suppressed near the bottom, and the momentum transfer is completely governed by the viscous stresses, thus  X ¼  ðdU=dZ Þ. The only important quantities that play a role in such a layer, viscosity ( ) and friction velocity (u*), can be combined to define a dimensional scale, the viscous length: = /u*. At distances of the order of , the rotational effects are negligible in comparison to those depending on the viscous friction. In the viscous sub-layer, the stress is constant and equals thep value ffiffiffiffiffiffiffiffiffiat ffi the bed,  0 (Soulsby, 1983), thus ðdU=dZ Þ ¼  0 =, but since u* ¼  0 = we obtain dU=dZ ¼ u2* = . As = /u*, it is possible to solve the differential equation dU=dZ ¼ u* = with the no-slip boundary condition (U(Z = 0) = 0). A linear profile of the velocity in the Z-direction is thus obtained: UðZÞ ¼ ðu* = ÞZ. Estimates of the distance from the bottom that corresponds to the thickness of the viscous sub-layer, where the velocity remains linear with Z, were obtained by laboratory measurements and, in terms of the viscous length, are Z ¼ 5  20 (Weatherly, 1972; Soulsby, 1983; Lueck, 2001). A commonly accepted value is 12 . Linear vertical profiles of the velocity up to 1 cm were observed by Chriss and Caldwell (1984) for a smooth turbulent flow. It should be noticed in this context that the laboratory experiments of Nikuradse (1933) established the various flow regimes near the sea floor, based on the value of the roughness Reynolds number, Red ¼ u* d= ¼ d= , where d is the grain diameter of the sea-floor sediment; the various regimes are (1) smooth turbulent (Red < 3.5), (2) transitional (3.5 < Red < 68) and (3) rough turbulent (Red > 68). Measurements by Sternberg (1968, 1970) validated this approach in cases where d is defined as the height of the roughness elements (i.e. bed ripples or secondary roughness due to biological activity) if the sea floor is not flat, obtaining however 5.5 and 165 as threshold values between the three different regimes, instead of 3.5 and 68. Most of the ocean bottom has roughness elements with a characteristic scale exceeding that of the viscous sub-layer (d >> ; rough turbulent regime). The coupling between the sediment and the flow characterizes the flow dynamics near the sea floor, thus leaving the role of viscosity insignificant (see below). In

S. Salon et al.

95

particular, the thickness of the layer closest to the sea floor is mainly determined by the spacing, rather than the height, of the sedimentary roughness elements (Soulsby, 1983) that cause horizontal variations in the flow properties. The term ‘‘bed layer’’ is also used to identify this sub-layer either in case of a smooth bed where viscosity plays a major role, and in case the roughness dominates the dynamics.

6.4.3.

The logarithmic layer

The vertical profile of the velocity departs from linearity at Z >> : the Reynolds stresses then become much larger than the viscous ones and neither the viscous length nor the Ekman height can adequately describe the dimensions of the flow dynamics. The shear stress in the layer where << Z << hE still equals the value determined by the boundary condition at the bed,  X =  0, but as turbulent effects dominate over the viscosity, the proportionality with the mean shear becomes ruled by the eddy viscosity:  X ¼  T ðdU=dZ Þ. Following the classical hypothesis of Prandtl (see McComb, 1990; Perlin et al., 2005), the eddy viscosity must have the dimension of a velocity associated to the amount of the turbulent fluctuations (u*), multiplied by an appropriate length scale, l (the mixing length): T = u*l. This implies that dU=dZ ¼ u* =l. Since no other scales rule the flow in this layer, the only dimensionally correct option is that the mixing length is proportional to the distance from the solid wall: l = Z, where is the von Ka´rma´n constant. Integrating the differential equation from Z0 (that is the distance from the bottom where the velocity matches the linear profile of the velocity in the viscous sub-layer and represents a scale of the bottom roughness; in the case of rough flow regime U(Z0) = 0, as discussed in the following) to Z, it is thus found that   u* Z þ UðZ0 Þ: UðZÞ ¼ ln Z0 In the case of a smooth bottom, Z0 depends on , and it was shown by laboratory measurements that, for this condition, Z0  0.1 (Weatherly, 1972); field observations indicate proportionality factors between Z0 and ranging from 0.25 to 0.004 (Soulsby, 1983). Under the rough flow regime, the logarithmic layer overlaps the viscous sub-layer and Z0 depends on the roughness scale of the sediments, d, indicating that Z0  d/30: in this case, Z0 represents the distance above the bottom in the trough between two roughness-element peaks where the mean velocity is zero (Weatherly, 1972). In case of a rippled sea floor with ripple height H and wavelength l a quite general estimation of the order of magnitude of Z0 is Z0 = 2H (H/l)1.4 for 15 < /Z0 < 1000 (Wooding et al., 1973). The logarithmic profile (also known as the log-law) allows to estimate the bed shear stress,  0, for example by measurements of the current velocity at some reference distance from the sea floor. Soulsby (1983) refers to U100 as the velocity at 1 m above the sea floor. Using the log-law and a feasible value for the roughness length, Z0, it is 2 possible to identify a situation where  0 ¼ C100 U100 , the drag coefficient being 2 C100 ¼ ½ =ln ð100=Z0 Þ . According to Soulsby (1983), values of C100 for different types of bottom range from 0.0016 for silt/sand (Z0 = 0.005 cm) to 0.0061 for rippled sand (Z0 = 0.6 cm).

96

Dynamics of the Bottom Boundary Layer

The logarithmic profile is therefore valid for Z0 << Z << hE, and ultimately conforms the Ekman profile if the distance from the bottom is of the order of the Ekman height. Turbulent kinetic energy is produced by the Reynolds stresses in combination with the velocity shear, and in the case of a homogeneous, pure shear flow, the production rate of turbulent kinetic energy, P, is in equilibrium with the rate of dissipation of the turbulent kinetic energy, ", where P ¼ u 0 w 0 ðdU=dZ Þ. As the mean shear can be described by u*/ Z and the Reynolds stresses by u*2, a direct relationship exists, by means of the energy balance P = ", between the dissipation rate and the friction velocity: U*3 = " Z. This formulation provides an estimate of the friction velocity – and therefore of the bottom stress – as an alternative for the value found through measurements with the log-law. Observations within 5 m from the bottom (Chriss and Caldwell, 1982; Johnson et al., 1994; Sanford and Lien, 1999; Trowbridge et al., 1999) frequently show a change in the slope of the logarithmic profile: the friction velocity that characterizes the upper log-layer, which has an associated velocity gradient that is larger than predicted by the classical log-law, is almost twice as large as the friction velocity of the lower log-layer. Lueck (2001) ascribed this change to horizontal bottom features, but Perlin et al. (2005) explained it in terms of a modification of the length scale (l) that conforms the Prandtl linear assumption near the solid boundary with the sea floor, while at larger distances it conforms the suppression of the turbulent fluctuations due to either the stratification of the water column or the perturbations from the sea surface. The inner layer remains governed by the local roughness of the sea floor.

6.5.

C ONCLUSIONS

The bottom boundary layer conveys the transfer and exchange of physical, chemical and biological properties between the sea floor and the sea, and has strong implications in the dissipation of the energy produced by the large-scale ocean currents. It extends up to a few tens of metres above the sea floor, with a current velocity that is of the order of 10 cm s1. The current velocity may reach 40 cm s1, thus increasing the thickness of the bottom boundary layer. The bottom boundary layer can be sub-divided into three principal sub-layers: (1) a bottom Ekman layer, of which the dynamics is described by the balance between the Coriolis force, the pressure gradient and the friction, (2) a viscous layer mainly dominated by the viscous stresses and (3) a transitional, logarithmic layer where turbulent stresses play a dominant role. In the Ekman layer the current velocity tends to decrease exponentially and to rotate (leftward in the northern hemisphere) with respect to the current in the geostrophic ocean interior. As a consequence, the current in the logarithmic and the viscous layers differs in magnitude and direction from what is observed some tens of metres above. An analytical approach that clearly describes the characteristics of each sub-layer can help increase the insight into the main processes involved in the bottom boundary layer. Analysis of the dynamics of the bottom boundary layer is also

S. Salon et al.

97

extremely important for obtaining useful estimates of basic quantities (such as the friction velocity or the bed shear stress) that can be of great interest for the research of sedimentary processes. In particular, the friction velocity characterizes the dynamics of the bottom boundary layer. Experimental observations in the Ekman layer show that the friction velocity is between 1/20 and 1/30 of the geostrophic velocity. By exploiting the log-law, measurements of the vertical current profile at a few metres above the sea floor may provide an estimate of the friction velocity, that in the logarithmic layer can be also evaluated as function of the dissipation rate of the turbulent kinetic energy. The oceanic bed layers are mostly characterized by a rough turbulent regime, where the dimension of the roughness elements exceed the viscous length. Therefore, the coupling between sediment and flow determine the velocity and the turbulent profiles in the layer closest to the sea floor, leaving the role of viscosity unimportant.

ACKNOWLEDGEMENTS We thank Xavier Durrieu de Madron (Cefrem, Universite´ de Perpignan, France) who kindly reviewed the text and with his comments helped us to improve the clarity and quality of our manuscript.

P A R T

3

PROCESSES

C H A P T E R

7

S EDIMENT E NTRAINMENT Y. He1, T. Duan2 and Z. Gao1 1

School of Geosciences, Yangtze University, Jingzhou, Hubei Province, China Marathon Oil Company, Houston, TX, USA

2

Contents 7.1. Introduction 7.2. Benthic Storms (Transportation and Resuspension of Particles) 7.3. Gravity Flows 7.3.1. Turbidity currents 7.3.2. Debris flows 7.3.3. Hyperpycnal flows 7.3.4. Other types of gravity flow 7.4. Bioclastics-Forming Processes in Contourites 7.5. Volcaniclastics 7.6. Glacial Activity 7.7. Resuspension of Particles by Burrowing Activity of Benthic Organisms 7.8. Concluding Remarks Acknowledgements

7.1.

101 102 105 106 111 113 115 115 116 117 118 118 118

INTRODUCTION

Sediment entrainment refers to the processes by which contour currents capture and transport sediments, that is, how sediments are delivered to contourcurrent systems. Detailed sediment transport and deposition by contour currents are reviewed by McCave (2008). Because direct observations of these processes cannot be made easily, our knowledge of sediment entrainment for contourites is mainly based on theoretical interpretation, and on speculations regarding sedimentary features of modern and ancient contourites including grain size, components, textures and structures, in terms of fluid hydrodynamics of possible flows operating in oceans. Therefore, we will describe in this section the processes that are potentially capable of delivering similar sediments to contour-current systems as observed in modern and ancient contourites. Although types of contourites vary greatly in terms of their origin, they can be better classified based on their composition and grain size (Stow and Holbrook, 1984; Duan et al., 1993; Gao et al., 1998; He et al., 1998; Stow et al., 1998a). For Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00207-0

 2008 Elsevier B.V. All rights reserved.

101

102

Sediment Entrainment

example, in terms of composition, contourite types recognized currently include mainly terrigenous clastic contourites, carbonate contourites, bioclastic contourites, volcaniclastic contourites and some transitional types. In terms of grain size, contourites can include mud-size contourites, silt-size contourites, sand-size contourites and their transitional types (Stow and Fauge`res, 2008). Contourites coarser than sand are not common, but do exist in pathways or channels of some very strong contour currents (Mullins et al., 1980; Duan et al., 1990, 1993; Akhmetzhanov et al., 2007). In most contourite drifts, mud and silt contourites are dominating, representing over 80% of the total of the sediments in the systems (Stow and Holbrook, 1984; Duan et al., 1993; Stow et al., 1998a). The possible processes that can deliver similar sediments to contour-current systems can be grouped into the following categories: (1) bottom currents (benthic storms); (2) gravity flows; (3) bioclastics-forming processes; (4) volcanic activity; (5) glaciation and (6) burrowing. Bottom currents resuspend and/or rework the sediments on their pathways. Gravity flows deliver both fine and coarse sediments to the pathways of contour currents. Bioclastics-forming processes provide bio-skeletons or fragments to contourites. Volcanic and glacial processes provide significant amounts of sediment to contour currents in some special situations. Bioactivity can resuspend sediments, albeit less significant in most cases.

7.2.

BENTHIC STORMS (T RANSPORTATION AND R ESUSPENSION OF PARTICLES)

In the 1930s, G. Wust, a German marine physicist, first pointed out that deep-sea bottom currents can transport sediments. Heezen et al. (1966) and Heezen and Hollister (1971), through their detailed study of deep-sea sediment cores, reported to have found geological evidence for deep-sea sediment transport. Since that time, there has been an explosion of research on deep-water tractioncurrent deposits. Hollister and his students, and Nowell started and coordinated the High Energy Benthic Boundary Layer Experiment (HEBBLE for short) project in 1978 (Hollister and McCave, 1984; Nowell and Hollister, 1985; Hollister, 1993). The goal of this experimental research was to understand the physical and biological response of deep-sea sediments to high-energy flows, and to predict the sea-floor’s response to water-bottom high-stress events (such as high-energy pulses) by experiments and model verification (Nowell and Hollister, 1985). Ninety scientists brought forward the concept of ‘‘benthic storm’’ based on the research achievements of HEBBLE. Now benthic or abyssal storms refer to current pulses that enhance contour currents on the ocean floor to above about 20–40 cm s 1, raising a large amount of sediments. The HEBBLE project revealed that fast (20 cm s 1) near-bottom currents, high concentrations of sediments and grooved mud beds are associated with erosion and deposition from ‘‘benthic storms’’ where a fast, deep mean flow is augmented or reversed by intense intermittent currents (Hollister and

103

Y. He et al.

Table 7.1

Benthic-storm characteristics in the HEBBLE area

Duration Frequency Maximum velocities measured 10–50 m above bottom Maximum concentrations measured 1–5 m above bottom Maximum turbulent kinetic energy Direction of highest energy events Estimated sediment flux rates during storms

2–20 days (most last about 3–5 days) 8–10 storms per year 15–40 cm s 1 3500–10 000 m g l 1 5–20 g cm2 s–2 Westerly, parallel to contours 20–200 cm3 m 2 per day

Source: From Hollister (1993).

McCave, 1984). The waning phase of a storm results in the development of bedforms and the rapid deposition of a mud blanket. This mud layer is rapidly bioturbated (roughened) and thereby conditioned or destabilized to be eroded readily by the next storm. Benthic-storm activity is very intense, especially in the northwestern and southwestern Atlantic Ocean, around South Africa, in the circum-Antarctic oceans, at high northern latitudes in the Atlantic Ocean, and in the northeast tropical Pacific Ocean. Some basic characteristics of benthic storms revealed in the HEBBLE area of Canadian North Atlantic Ocean continental margin are shown in Table 7.1. The HEBBLE study indicates that benthic storms can accelerate the more constant contour currents into much stronger but short-term near-bottom currents which have the capacity to erode, suspend, transport and deposit large amounts of sediments. During benthic storms, a large volume of materials is resuspended and thus contributes to a very high-density nepheloid layer. These materials may then be transported over a long distance by the more constant contour current. These benthic storms could, because of their high energy, significantly influence the velocity and direction of contour currents and have an important impact on the geological record. Because of the velocity changes of contour currents, normal pelagites and hemipelagites accumulate during inactive periods of benthic storms and contour currents. During the active phases of benthic storms and contour currents, unconsolidated sediments can be winnowed from the sea floor and become reworked or resuspended, transported and redeposited. McCave (1986) suggests that most of the nepheloid layers in oceans are the results of the resuspension of sea-floor sediments. The grain size and amount of sediments that are resuspended and transported depend on the intensity of the benthic storms. Where bottom currents are strong, stable and persistent, the amount of resuspended sediments will be large, the grain size will be coarse, and sorting will be good. Hollister and McCave (1984), using the Lamont nephelometer, have demonstrated a clear relationship between the regions of high mean velocity in deep boundary currents and high suspended sediment content. Hollister and McCave (1984) also suggest that major resuspensions in the world generated by abyssal storms are likely to be those where surface eddy kinetic energy is high and coupled to a strong near-bottom mean flow (contour current) and sufficient muds. An extremely dense nepheloid layer would suggest benthic-storm

104

Sediment Entrainment

Concentration of suspended matter

log E/E0 2.2

Potential temperature (°C)

2.1

0.25

0.50

0.75

C11-195

10

30

50

µg l–1

C11-197 2.0 1.9

C11-194

1.8 1.7 1.6 C11-196 1.5

Figure 7.1 Light-scattering profile made by the Lamont ^ Doherty Nephelometer (left) and concentrations of suspended particulate matter measured by filtration of large-volume Niskin bottle samples (right) in the bottom nepheloid layer over the Greater Antilles Outer Ridge. Film exposure, E, in the light-scattering profile is normalized against exposure in the clearest water, E0. Profiles are labelled by cruise and station. C11-194 (etc.) is profile number (from Tucholke, 2002; with permission from the Geological Society, London).

activity, and on the contrary large sediment drifts and mud waves indicate significant long-term deposition. For example, fine-grained contourites of the Greater Antilles Outer Ridge have been inferred to have been deposited entirely from suspended load in abyssal currents, where the concentrations of suspended matter in the nepheloid layer range from <10 up to 63 mg l 1 (Figure 7.1, Tucholke, 2002). Both maximum concentrations and depth-averaged concentrations follow the contour of the light-scattering profiles and suggest a core of enhanced sediment load in Antarctic Bottom Water at about 1.6–1.7C, i.e., at depths of approximately 5050–5400 m (Tucholke, 2002). The nepheloid layer near the slope has also been inferred to have played an important role in the formation of Drift 7 at the Antarctic Peninsula Pacific Margin besides turbidity currents (Figure 7.2) (Rebesco et al., 2002). Moreover, nepheloid layers are well developed in the Baltic Sea and have had a pronounced effect on the formation of muddy contourites throughout the Holocene (Sivkov et al., 2002). The origin of benthic storms is still uncertain. Usually benthic storms may be caused by surface wind-driven circulation or ocean surface eddies (Hollister and McCave, 1984; Demidova et al., 1993). Nonetheless, benthic storms play important roles in sediment entrainment for contourites. They can erode, resuspend, transport and deposit a large volume of muds to sands in contour-current systems by intensifying contour currents and their related nepheloid layers.

105

Y. He et al.

NW

Polynya

Katabatic winds

Plumites Slope current

SE

Deforming basal sediment layer

Turbidites

~4000 m Nepheloid Hydraulic jump layer Very little bioturbation

80°W

75°W

Li pr ne of of ile 67°S Contour current

Ma ste ss wa ep sid sting e o on 0 f dr 0 40 ift

3500

Sediment conveyed to shelf edge; frequent turbidites

Tu in rbi ch dit an es ne ls nt

rre

3000 2500

u ec

Sl

op

Ice sheet grounded to shelf edge

2000 1500 0 100

Figure 7.2 Inferred sediment transport processes during glacial times in the area of Drift 7 on the Antarctic Peninsula Pacific Margin. Down-slope flow is shown as open arrow, and alongslope flow is shown as black arrow. Katabatic wind is a wind that is created by air flowing downhill; polynya is a semi-permanent area of open water in sea ice; plumites are deposits from turbid meltwater plumes (simplified from Rebesco et al., 2002; with permission from the Geological Society, London).

7.3.

GRAVITY FLOWS

A large volume of coarser sediments are carried to the continental slope, rise and abyssal basin by turbidity currents and debris flows. Subsequently, bottom currents may rework these sediments and produce new types of deposits. A lot of

106

Sediment Entrainment

contourites are often closely associated with gravity-flow deposits, and usually occur directly at the top of such deposits, or alternate with them. Sediment gravity flow is a general term for flow of mixtures of sediment and fluid in which bedding coherence is destroyed and the individual grains move in a fluid medium and sustain the movement. A detailed classification of gravity flows and a discussion of their relationship to other sediment density flows, e.g., hyperpycnal flows, can be found in Stow et al. (1996a) and Mulder et al. (2003b). Based on the mechanism of grain support (Middleton and Hampton, 1976), sediment gravity flow can be classified into four types, i.e., debris flow (clasts supported by matrix), grain flow (by grain-to-grain interactions), fluidized sediment flow (by escaping pore fluids) and turbidity current (by fluid turbulence). Among the four types of gravity flows, the turbidity current is the most important to the sediment entrainment of contour-current deposition. Hyperpycnal flow (Mulder and Alexander, 2001), is a special type of turbidity current that is directly induced by river flooding charges entering oceans, or other relatively persistent long-term processes, and will be discussed following the sections of turbidity and debris flows.

7.3.1.

Turbidity currents

In general, contour-current activity takes place mostly between the main depositional episodes of turbidity currents (Mulder et al., 2008). In some of the documented examples, the interplay of both down- and along-slope processes seems quite obvious, with top-cut thin turbidites resulting from bottom-current winnowing and erosion. In such cases, sharp or erosive contacts between turbidites and inferred overlying bottom-current deposits, coupled with distinctly different characteristics above and below the contact, such as bioturbation or anomalous grain size, can indicate reworking or removal of turbidite tops by bottom currents. Meanwhile, the composition of contourites and associated turbidites is very similar in many instances. All these characteristics indicate that contourites may result from bottom-current reworking of turbidity currents or their deposits. Contourite drifts are developed when bottom currents dominate the depositional systems, whereas gravity flows are limited or weak. In many documented examples (Gao et al., 1995; Stow et al., 2002d; Moraes et al., 2007), turbidites are interbedded or overlain directly by contourites and pelagic deposits (Figures 7.3 and 7.4). Moraes et al. (2007) well documented that the contourites in the Palaeocene–Eocene deep-water reservoirs of the Campos Basin, Brazil, were produced by bottom-current reworking and redeposition of turbiditic sands. Their interpretation was based on the very similar composition of turbidite and contourite sands and the abrupt contacts between turbidites and their overlying contourites, indicating reworking or removal of turbidite tops. Furthermore, with turbidites occurring much less, contourites are dominating in the Paleocene, whereas turbidites prevail in the Eocene, during which epoch contourites decreased dramatically, indicating that contourites are favoured when turbidites are limited. Turbidities, contourites and normal bathyal–abyssal deposits are well developed in the Late Triassic Zhuwo Formation in the Tanggor area, Zoige, Sichuan province,

107

Y. He et al.

0

144 143 142

1 141 140 139 2 138 137 3 136

4

135

137 136 Interval number in measured section

Siliceous shale

134

5m

133

Turbidite

132 131

Contourite

Figure 7.3 Measured section through part of the Pingliang contourite drift, showing the relationship between turbidite and contourite in vertical sequence. Note that turbidites are overlain by contourites in the 136^137^138 and 140^141 intervals. Siliceous shale is a shale with a substantial cherty component, whereas all other lithotypes are carbonates. A multicolour version of this figure is on the enclosed CD-ROM.

China, where turbidites are capped directly by contourites and then overlain by normal abyssal deposits (Yang et al., 1996). Grain-size analysis of thin sections showed the close relationship between turbidites and the associated contourites. The content of the matrix of the underlying turbidite sandstone is a bit high (20–25%), the mean grain size (Mz) is 2.413,  = is the unit of grain size.  = log2D (D is the diameter of grain, mm). The standard deviation (1) is 0.593, and the cumulative probability curves form a gently incline line. The grain size of the contourites ranges from 3 to 5.5, the mean grain size (Mz) is 4.36, the standard deviation (1) is 0.402 implying better sorting, the mean skewness (Sk1) is 0.006, approaching symmetrical distribution, and the kurtosis (KG) is 1.025. The value of Y, a discriminant function of Sahu (1964), is 8.832, less than the threshold value of 9.8433 for turbidites. These features indicate that the contourites show turbidite characteristics to a certain degree but still mainly demonstrate the characteristics of traction-current deposits. That is, the contourites of the Late Triassic Zhuwo Formation might have resulted from the reworking of the immediately underlying turbidites.

108

Sediment Entrainment

Lithological unit

Miocene

Depositional processes

Upper marl unit:

Contourites

Upper Lefkara

Age

Marl, chalky marl (0–200 m)

Oligocene

Chalk, marl, chert

Turbidites

Chalk chert unit:

Pelagic sedimentation

Middle Lefkara

Eocene

Lefkara Formation

Chalk unit: Chert-free chalk with occassional marl (0–200 m)

Palaeocene Cretaceous

Lower Lefkara

(0–250 m)

Lower marl unit: Pink or gery marl, chalk marl, chert (0–40 m) Base chert-free

Figure 7.4 Age, lithological units and principal depositional processes of the Lefkara Formation in Cyprus (simplified from Stow et al., 2002d; with permission from the Geological Society, London).

A number of examples studied worldwide (Stow et al., 2002f; Viana and Rebesco, 2007) attest that the sediments of contourites were supplied by turbidity currents or are closely related to turbidity currents. Rebesco et al. (2002, 2007) provided a well-documented example where contour currents are inferred to have pirated turbidity flows in forming several sediment drifts on the Antarctic Peninsula Pacific Margin. In particular, sediments initially supplied and transported by turbidity currents have been pirated by ambient bottom currents and further transported and deposited in sediment Drift 7 (Figure 7.2) (Rebesco et al., 2002, 2007). Very similar to the sediment Drift 7, contourite sediments with bedforms in the sheeted contourite drift on the Barra Fan of the northwest UK continental margin are inferred to have formed from low-density glacigenic turbidity currents, but were pirated by contour-following bottom currents on the distal part of the drift (Knutz et al., 2002a, b). Sediments of contour currents may partly result from turbidites and debrites in the Sicilian gateway between East and West Mediterranean Basins (Reeder et al., 2002). Sediment input of drifts of the Corsica Channel, northern Tyrrhenian Sea, was closely related to turbidites (Roveri, 2002). The Columbia Channel is a turbiditic channel elongated W–E on the continental rise of the south Brazilian Basin with 4200–5000 m water depth, where turbidity currents show a W–E trend parallel to the channel axis, whereas contour currents show a S–N trend parallel to the rise contours (Figure 7.5) (Fauge`res et al., 2002a), and sediment supply of contourites was closely related to turbidites.

109

Y. He et al.

35°

34°

31°

32°

33°

42 00

400

0

36°W

Vitoria

20°S

nts amou

Se

Trinda

de

0 400 2000 300

0

40

1000

21°

2000 46

00

00

420

0

4400

00

48

23°

52

00

50

00

40

00

22°

50

00

Foram- and/or mica-rich turbiditic flows derived from the Vitoria–Trindade Seamounts

Dominant sandy turbidites interbedded with muddy contourites

Quartz-rich turbiditic flows derived from the continental slope and upper continental rise

Dominant muddy contourites

Contour currents (AABW)

Figure 7.5 Synthetic map showing the contour-current and turbidity-current pathways, and the distribution of sedimentary processes, associated deposits and depositional environments on the Vitoria Trindade Seamounts, southern Brazil Basin (from Fauge' res et al., 2002c; with permission from the Geological Society, London)

The Pernambuco Contourite Fan (so-called ‘‘modified drift–turbidite systems’’) developed near the northeastern continental slope off Brazil was formed by an interplay between down-slope sediment gravity flows (mainly turbidity currents) and along-slope contour currents (Gomes and Viana, 2002). The process of combined turbidite/contourite deposition took place on the southeastern margin of the Weddell Sea, sediments were entrained by turbidity currents and Ice Shelf Water (ISW) plumes, and the contourites partly resulted from the reworking of turbidites (Michels et al., 2002). Turbidite, contourite and hemipelagic deposits are well developed on the Wilkes Land continental rise, Antarctica, and can be classified into three facies. In the succession, facies 1 (the oldest one) is dominated by turbidites, facies 2 (middle) and 3 (youngest one) are dominated by turbidites and contourites (Escutia et al., 2002). Turbidity currents redirected by the SW Pacific

110

Sediment Entrainment

Deep Western Boundary Current (DWBC) played and still play a major role in the Miocene-Recent formation of the Hikurangi Fan–Drift of Eastern New Zealand (Carter and McCave, 2002). The Neogene hemipelagite–contourite continuum in the Miura–Boso forearc basin, SE Japan, is interbedded with thick scoriaceous beds of turbiditic and pyroclastic fall origin (Stow et al., 2002e). The formation of these contourites is related to winnowing and reworking of certain turbidites. In the Plio–Pleistocene Kazusa forearc basin, Boso Peninsula, Japan, sandy contourites are commonly associated with turbidites and comprise turbidite-to-contourite continuums (Ito, 1996, 2002). The framework composition of sandy contourites is largely equivalent to that of the associated turbidites. Meanwhile, sandy contourites are better sorted than associated turbidites and have sharp or gradational basal contacts with underlying turbidites or hemipelagites, and sharp upper contacts with overlying hemipelagites (Ito, 1996, 2002). Oligocene fossil contourites in Cyprus, where sediments were mainly supplied by turbidity currents derived from the Kyrenia Range in the north, are carbonate-rich contourites (Kahler and Stow, 1998; Stow et al., 2002d). In the turbidite-to-contourite continuum in St Croix, US Virgin Islands, in the Caribbean Sea, most of these turbidites appear to have been partially reworked by bottom currents related to basin circulation and the contourites were closely related to turbidites (Stanley, 1993). The Ordovician sequence in the northeastern Lachlan Fold Belt of eastern Australia consists of three lithological units representing an overall fining-upward succession from turbidites to black shale, where the lower unit is dominated by turbidite beds, the middle unit consists mainly of sandy contourites, and the top unit is built by black shales of hemipelagic origin (Jones et al., 1993). These contourites would have resulted from the reworking of turbidite muds and sands by bottom currents. There are many other examples of contourites in the stratigraphic record found in China, such as (1) early Ordovician contourites in the Hunan province (Duan et al., 1990, 1993; Gao and Duan, 1994); (2) middle Ordovician contourites in the Gansu Province (Gao et al., 1995); (3) Cambrian contourites in the western Hunan and eastern Guizhou provinces (Liu et al., 1990); (4) Silurian–Middle Devonian contourites in the Qinzhou Basin, Guangxi Province (Yu et al., 1989); (5) Paleozoic contourites in the Mangar Sag, Tarim Basin (Wang, 1995) and (6) early Permian contourites in the Reshuitang of the Lancangjiang zone, southwestern Yunnan (Jia, 1995). To some extent, these contourites are related to gravity-flow deposits in all these ancient examples, in which turbidites are also used as indicative of deep-water environments and as a regional reference for the direction of down-slope – transportation (Figure 7.6). There is consequently a spectrum with respect to the larger or smaller contribution of turbidity currents to the formation of contourites and sediment drifts, depending on the relative strength of the turbidity and contour currents, which in turn is related to sediment supply, climate and sea-level changes (Stow et al., 2002f). If the turbidity currents dominate the depositional system, a turbidite submarine fan will be formed. If contour currents dominate and turbidity-current-related input is limited, a contourite drift will be formed, such as the fine-grained sediment drifts

111

Y. He et al.

E

Bank back

Bank margin

Slope

Contourite drift

Basin plain

Gravity flow

Contour current

Figure 7.6 Sedimentary model of the Jiuxi contourite drift, showing the relationship between contourite and gravity-flow input (modified from Duan et al., 1990; with permission from the Geological Society of China).

on the Blake Outer Ridge and the Greater Antilles Outer Ridge (Tucholke, 2002). Among these two end members are many mixed sediment drifts with still significant contributions of both turbidites and contourites. The sediment Drift 7 is dominated by deposition from contour currents that pirate turbidity currents most of the time (Rebesco et al., 2007), whereas the combined drift/turbidite systems suggest a more significant role of turbidity currents (Gomes and Viana, 2002; Michels et al., 2002). Some slopes and basin margins are of relatively low sediment input because of the lack of turbidity current deposits, so that contourites are not well developed even though the contour currents are strong. A good example is the eastern slope of the Corsica Channel (Roveri, 2002).

7.3.2.

Debris flows

In some cases, debris-flow deposits can also be reworked into contourites by contour currents, similar to turbidite reworking. The difference is that debrisflow deposits are usually coarser and more cohesive, and cannot be easily reworked by contour currents. Only in cases when the contour current is very strong, can debris-flow deposits be reworked. For example, contourites of the Nova Scotia continental rise may partially result from the reworking of turbidites and debris lobes (McCave et al., 2002). Contourites in the Sicilian gateway between East and West Mediterranean Basins may come partially from reworking of debrites in addition to the reworking of turbidites (Reeder et al., 2002). Another interesting example is the contourites of Wulalike Formation of the middle Ordovician in the Zhouzishan Mountains, northwestern China. The lenses of contourites consisting of echinoderm bioclastic limestones in the lower part of Wulalike Formation are mostly 2–5 m thick, with a maximum thickness of 7 m. The overlying and underlying layers are dark graptolite shales, and the lenses of contourites grade laterally into echinoderm-clast-rich breccias (Figure 7.7a). The

112

Sediment Entrainment

(a)

(c)

(d)

(b)

0.4 mm

(e)

0.4 mm

Figure 7.7 Lenses of contourites consisting of echinoderm clastic limestones in the lower part of the Wulalike Formation of the middle Ordovician in the Zhouzishan Mountains, northern China. (a) Occurrence of echinoderm clastic limestones. (b) Series of beds with cross-bedding in echinoderm clastic limestones. (c) Large-scale cross-bedding in echinoderm clastic limestones (lamina dipping to left of the photo, though slightly). (d) and (e) Secondary enlargement of encrinite in echinoderm clastic limestones and lime mud infill (in (e)). A multicolour version of this figure is on the enclosed CD-ROM.

lenses are composed of several individual beds with thicknesses of 20–50 cm (Figure 7.7b), and each bed contains large-scale cross-bedding (Figure 7.7c). Foreset laminae in the cross-beds dip towards the northwest (left of the photo), which is nearly perpendicular to the palaeocurrent direction indicated by flute casts and ripples in the turbidites of the overlying Lashizhong Formation. Bioclasts in echinoderm-rich clastic limestones are mainly crinoids, which comprise up to 65% of the lenses. Other components include carbonate sands and gravels composed of micritic limestone and bioclast-bearing micritic limestone, and bryozoan

113

Y. He et al.

fragments. Diagenesis (secondary enlargement) of crinoid fragments is intense (Figure 7.7d), so that grains are cemented by secondary calcite with local infill of voids by lime mud (Figure 7.7e). The echinoderm-rich clastic limestone lenses are interpreted as the product of relatively strong contour currents that reworked the echinoderm-clast-rich breccias. The character of texture, structure, palaeocurrent direction and palaeogeographic analysis indicate clear differences from turbiditycurrent deposits. Furthermore, there is a remainder of debris-flow deposits surrounding the contourite (Gao et al., 1995, 1998).

7.3.3.

Hyperpycnal flows

Unlike turbidity currents generated by sediment collapse discussed in previous sections, river discharge-generated hyperpycnal flows are usually long-duration phenomena, lasting days or weeks, and they can be considered as a special type of turbidity currents. In contrast, hyperpycnal flows are one type of the so-called density flows. This type of density flow, described and defined first by Bates (1953), usually forms at sea or in a lake by continuous river flooding discharge. Storm weather or other mechanisms can also generate days-lasting hyperpycnal flows, transporting sedimentary particles off carbonate platforms, or continental shelves. When sediment-laden river water enters oceans one of three types of flows will form, depending on its density relative to the density of the sea water of the receiving ocean basin. The three possible situations are shown in Figure 7.8. The flow is hypopycnal if the density of the river plume is less than that of the receiving sea water. In this case, the sediments carried by the flow will form a spreading plume at the sea surface and will gradually settle to the sea floor as a hemipelagite.

Homopycnal flow

ρf = ρw

ρf ρw (a)

ρf ρw (b) Sediment-laden channelized flow

Hypopycnal flow ρf < ρw Hyperpycnal flow ρf>ρw ρ f

Flow expanding into receiving water body

Figure 7.8 Types of density flow (from Mulder and Alexander, 2001; with permission from Blackwell Publishing). f = density of river plume; w = density of ambient fluid (sea or lake water). Hyperpycnal flow: f > w ; homopycnal flow: f = w ; hypopycnal flow: f < w.

114

Sediment Entrainment

However, if these usually fine-grained sediments reach a contour current, they will be further transported and deposited as a contourite when the strength of the contour current decreases. A large proportion of the sediments in muddy contourites may have been delivered in this way. The flow is described as homopycnal if its density is equal to that of the receiving sea water. In this case, the plume will mix vertically with the sea water and drop its load forming a mouth bar. It is possible that the surface part of the plume, after being diluted by dropping its coarser particles, will form a secondary hypopycnal flow, to deliver its fine material to a contour current. If the plume density is higher than that of the receiving sea water, the flow is termed hyperpycnal. In this case, the plume will sink to the sea floor and may form a long-lasting turbidity current. This type of flow has recently attracted much attention in sedimentological community because of its potential importance for the generation of a special type of turbidite, and because of its possible role in shaping meandering submarine channels in deep-sea depositional systems (Mulder and Alexander, 2001; Mulder et al., 2003b; Mutti et al., 2003). Mulder et al. (2001) correlated flood-generated turbidity currents in the Var Canyon, southeastern France, with turbidite sequences cored at 2 km water depth in the Mediterranean. It is shown that, over the past 100 years, 13–14 hyperpycnal turbidity currents were generated. However, only two or three Bouma sequences were recognized in the same interval. Their study also shows that, over 100 years in the Var Canyon, hyperpycnal flows contributed a greater volume of sediment to the ocean than submarine landslides. Mulder et al. (2001) calculated that 68–71% of the sedimentary column was deposited by hyperpycnal turbidity currents; 3–6% resulted from classical, slide-induced turbidity currents; and 26% consisted of hemipelagites. Furthermore, hyperpycnal currents usually transport only particles finer than medium sand and can travel a very long distance, which is determined by their origin (Mulder et al., 2003b). Therefore, another important role that hyperpycnal flows may play is to deliver a significant proportion of sediments from mud to fine sand for contour-current entrainment when they interact. From the composition and grain size of contourites (Stow and Holbrook, 1984; Stow et al., 2002c), it is reasonable to speculate that a significant fraction of sediments in contour-current deposits (muds, silts and fine sands) are delivered probably by this type of flow to the pathway of contour currents. Wilson and Roberts (1992, 1995) propose the sediment-charged hyperpycnal flows generated on carbonate platforms during winter cold fronts, or summer intense heating and evaporation as a mechanism for rapid off-bank and vertical transport of shallow-water fine sediments in carbonate-periplatform sedimentation. This type of density flow with estimated velocities of 0.5–2.5 m s 1 has been confirmed by direct submersible observations and sediment-trap experiments along the western margin of Great Bahamas Bank, and is inferred to be able to transport fine sands on the top of carbonate platforms (Wilber et al., 1993; Wilson and Roberts, 1995). This mechanism should be particularly important in delivering sediments to carbonate contourite drifts such as in the Straits of Florida (Mullins et al., 1980), in the Cretaceous Talme Yalfe Formation (Bein and Weiler, 1976) and in the Ordovician Jiuxi carbonate drift (Duan et al., 1990, 1993).

115

Y. He et al.

It should be pointed out that, besides their role in delivering a large volume of fine sediments into deep-sea depositional systems such as submarine fans and contourite drifts, hyperpycnal flows deposit a special type of turbidite, hyperpycnalites, which may be confused with contourites under criteria as described by Gonthier et al. (1984) and Duan et al. (1993) in sense of variation of vertical textures and structures in a typical succession. As Mulder et al. (2003b) discussed, the successions of contourites and hyperpycnalites may show very similar characteristics. This is especially true if the transition between the coarsening-up lower facies and the fining-up upper one in a typical hyperpycnalite succession is very gradational, and intensely bioturbated. This type of hyperpycnalite succession can be formed by a hyperpycnal flow with a low-magnitude flood, but with a sufficient discharge to create enough density.

7.3.4.

Other types of gravity flow

Owning to their rare occurrence in any given depositional system, the importance of grain flows and fluidized sediment flows is limited, but we cannot exclude the possibility that they deliver minor sediments for contourites.

7.4.

BIOCLASTICS-F ORMING PROCESSES IN C ONTOURITES

Bioclasts and biogenic calcareous or siliceous materials are among the most common constituents of contourites, and may be dominant in some settings (Figure 7.9). For example, there are 20–50% biogenic carbonates (planktonic and benthonic foraminifers, bivalves, ostracods) in the sediments of the Faro–Albufeira contourite

Biogenics

Feni, Snoni and Hatton Drifts

Gloria and Bjorn Drifts

Volcanogenic (+smectites)

Feni and Faro Drifts and Blake–Bahama Outer Ridge

Terrigenous Vogel and Nashville Seamount Drifts

Figure 7.9 Composition of North Atlantic contour-current deposits (from Stow and Holbrook, 1984; with permission from the Geological Society, London).

116

Sediment Entrainment

drift, northern Gulf of Cadiz (Stow et al., 2002b), where primary biogenic productivity yielding mainly microfossil remains is an important aspect of an abundant sediment supply of contourites. The bioclasts are mostly disarticulated and broken, with only a small minority of bioclasts being completely preserved. In the central Scotia Sea, the Holocene sandy contourites consist predominantly of large diatoms. All cores exhibit a biogenic/terrigenous cyclicity, with biogenenic sediments at the core top passing down into a more terrigenous unit and then into another biogenic unit (Pudsey and Howe, 2002). The upper biogenic unit (0.5–2 m thick in most cores) consists of dark greyish brown forminifera-bearing diatom-rich silty mud overlying olive grey to greenish grey diatom mud (in which diatoms may dominate the mud); the terrigenous unit (0.4–2 m) consists of diatom-bearing mud without biogenic carbonate; and the lower biogenic unit (at least 1–1.5 m thick) consists of grey to dark grey muddy diatom ooze and diatom mud. Foraminifer-rich sandy contourites on mid-Pacific seamounts are mainly composed of foraminifers (Viana et al., 1998a). Deep-water sandy contourites on the Carnegie Ridge, equatorial E. Pacific, are mostly composed of moderately sorted biogenic sands, i.e., broken fragments and tests of Quaternary foraminifers (Viana et al., 1998a). Moreover, most of drifts in North Atlantic Ocean contain biogenic materials. Feni, Snoni and Hatton Drifts contain 35–70% foraminiferal sand in sandy contourites. Gardar Drift, Bjorn Drift, Eirik Drift, Gloria Drift, Faro Drift, Bermuda Rise, Corner Rise, West Reykjames Ridge, Newfoundland Ridge and Blake–Bahamas Out Ridge contain 10–20% foraminiferal sand. Gilliss Seamount, Great Meteor Seamount and Gibbs Fracture Zone contain relatively less biogenic materials (Stow and Holbrook, 1984). Calcareous biogenic contourites are also found in the Davis Drift of the Mozambique Channel, on the Mozambique Ridge in the Indian Ocean, on the Ontong Java Plateau of the Southwest Pacific Ocean and in the Marion Drift of the Pacific Ocean (Stow et al., 1998a). The most common components of organisms in modern contour-current deposits are identifiable to benthonic and planktonic foraminifers, ostracods, coccoliths and diatoms. Rarely are bioclasts locally concentrated enough to form bioclast-rich contour-current deposits. Biogenic siliceous materials are located primarily in abyssal basins, and include diatom ooze, radiolarian ooze and siliceous sponge-spicules ooze, which are transported by contour currents during the process of deposition, or are resuspended and then deposited further down the flows. Benthic and swimming or planktonic trilobite shells or fragments are important components of contourites in the Early Ordovician Jiuxi Drift (Duan et al., 1993).

7.5.

V OLCANICLASTICS

Volcaniclastics are also an important constituent of contourites in many cases (Figure 7.9). In general, volcanic materials in contour-current deposits consist largely of volcanic ashes. For example, airborne volcanic tephra as well as Saharan dust are dispersed towards the east in the Sicilian gateway, Central Mediterranean region. These volcanic ashes occur as graded layers of muds, sands and gravels and are

117

Y. He et al.

preserved as primary air-fall tephra layers and as secondary, mixed bioclastic– volcaniclastic turbidites or contourites on the deep-basin floors (Reeder et al., 2002). Neogene volcaniclastic contourites are also found in the Central Pacific Ocean, and consist mostly of reworked volcanic tuffs (Stow et al., 1998a). The provenance of the early Permian contourites in Reshuitang, Lancangjiang zone, southwestern Yunnan, China, is the eastern continental volcanic island arc (Jia, 1995). The content of volcanic ash in contour-current deposits varies greatly, and depends primarily on the eruptive level, the eruptive frequency and the number of volcanoes.

7.6.

GLACIAL ACTIVITY

A detailed description of the relationship between glaciation and contourites is presented by Mulder et al. (2008) and Van Weering et al. (2008). We focus therefore here on how glacial processes deliver sediments to contourcurrent systems. Terrigenous clastics are one of the uppermost constituents of contourites which are transported by gravity flows, and also by glacial activity, especially in high-latitude areas, or during a glaciation. The sediments of contourites are derived locally or partially from ice-rafted components (Stoker et al., 1998a). For example, glacigenic contourites are developed in both the northern Rockall Trough and, especially, the Faroe–Shetland Channel, North Atlantic Ocean (Duan et al., 1994; Howe et al., 2002). These sediments show many of the features of the parent glaciomarine deposits, including dropstones, extremely poor sorting and a grain size with both coarse and fine fractions that are obviously supplied through ice rafting (Howe et al., 2002). Terrigenous clastic material derived from glacial melting is very common in contour-current deposits in the oceans around Antarctica. The Barra Fan drift development was facilitated by fluxes of glaciomarine sediments from the Hebrides Shelf Margin, NW UK continental margin (Knutz et al., 2002a, b). Although Holocene contourites in the central Scotia Sea consist predominantly of alternating biogenic and terrigenous silts, sediment cores exhibit a terrigenous/ biogenic cyclicity corresponding to glacial/interglacial cycles (Pudsey and Howe, 2002), which indicates that glacial activity played certain role in the sediment supply of contourites. The occurrence of sedimentary and lithified clasts within units III and IV of the Lofoten Drift in the Norwegian Sea reveals that there was an input of ice-rafted material in a glaciomarine environment (Laberg et al., 2002). There are some other contourite examples related to or partially related to glaciation: (1) muddy contourites with glacial patterns of the Eirik Drift and the Gloria Drift in Irminger Basin, North Atlantic Ocean (Stow et al., 1998a); (2) turbidite, contourite and hemipelagic deposition on the Wilkes Land Continental rise, Antarctica (Escutia et al., 2002) and (3) Plio–Pleistocene interbedded silty and muddy contourites (glacial muddy contourites) in the Bellingshausen Basin, South East Pacific Ocean (Pudsey and Camerlenghi, 1998). Glaciations deliver sediments to contour-current systems by both ice rafting and sediment density flows originated from the melting of debris-laden glaciers.

118

Sediment Entrainment

7.7.

RESUSPENSION OF PARTICLES BY B URROWING ACTIVITY OF B ENTHIC O RGANISMS

Organism activity causing disturbance of the sea floor and hydrodynamic winnowing promote the resuspension of unconsolidated sediments. Resuspension of particles by the burrowing activity of benthic organisms is one of the processes feeding the nepheloid layer and hence bottom-current transport. The presence of high-density nepheloid layers is an important condition for active deposition from bottom currents in oceans. For example, the supply of sand in the Holocene contourite sand sheet on the Barra Fan slope, NW Hebrides Margin comes partially from in situ winnowing (Stow et al., 2002c). The presence of various assemblages of fragmented benthonic and planktonic tests, coupled with an overall upward increase in grain size through the sand sheet, indicates vertical supply by along-slope transport and reworking of the surface sediments by vigorous bottom currents (Stow et al., 2002c). The Lofoten Drift in the Norwegian Sea probably originates from the deposition of suspended particles derived from winnowing of the shelf and upper slope (Laberg et al., 2002).

7.8.

CONCLUDING R EMARKS

Based on sedimentary textures, structures and components of various contourites, as well as on their spatial variation, it is believed that the most important mechanisms for sediment entrainment in contourites are (a) the resuspension from bottom currents/benthic storms, (b) sediment density flows, including turbiditic and hyperpycnal flowsand (c) micro-bioclastic supply by pelagic processes. Of course, other mechanisms such as volcanic and glacial activity are also important in certain environments. Obviously, there are often probably two or more types of sediment-entrainment mechanisms involved in the formation of a certain contourite. For instance, glacial activity, turbidity currents and contour currents all have played important roles in the formation of the contourite drifts in the northern North Atlantic Ocean and along Antarctica, such as the Rockall Trough and Faroe–Shetland Channel area (Duan et al., 1994; Stoker et al., 1998a; Akhurst et al., 2002; Howe et al., 2002) and the sediment drifts on the Antarctic Peninsula Pacific Margin (Pudsey and Camerlenghi, 1998; Rebesco et al., 2002).

ACKNOWLEDGEMENTS Thanks are due to Dr Michele Rebesco, Dr Angelo Camerlenghi and Dr Tom van Loon for their kind support, reviews and constructive suggestions, and to Dr Michael Gardner and Dr Dengliang Gao for their reviews and

Y. He et al.

119

comments on an earlier version of the manuscript. This work was supported by the National Natural Science Foundation of China (40272060 and 40672071 to He and Gao), the Teaching and Research Award Program for Outstanding Young Teachers in Higher Education Institutions of MOE, P.R.C. (to He) and the CNPC Innovation Foundation (2002F70102 to He). Duan also thanks Marathon Oil Corporation for its encouraging publishing this article.

C H A P T E R

8

S IZE S ORTING D URING T RANSPORT AND D EPOSITION OF F INE S EDIMENTS : S ORTABLE S ILT AND F LOW S PEED I.N. McCave Department of Earth Sciences, University of Cambridge, Cambridge, UK

Contents 8.1. Introduction 8.2. Size Analysis of Fine Sediments 8.2.1. Measurement of fine grain sizes 8.2.2. Size distributions of components 8.3. Unsorted Delivery to Deep Current Systems: Pelagic Flux and Down-Slope Transport 8.4. Controlling Factors for Input and Transport 8.4.1. Settling velocity of grains and aggregates, ws 8.4.2. Boundary-layer flow structure and turbulence 8.5. Sorting in Suspension Transport 8.6. Processes of Deposition from Turbulent Boundary Layers 8.6.1. Rate of deposition 8.6.2. Limiting shear stress for deposition td 8.6.3. Sorting by selective deposition: cohesive versus sortable silt 8.6.4. Repeated sorting events under deep-sea storm conditions 8.7. Deposits from Currents 8.7.1. Spatial variability of sediment size over bedforms and drifts 8.8. Some Examples of Palaeoflow Inferred from Sortable-Silt records 8.8.1. Gardar Drift and Bermuda Rise at the penultimate glacial termination 8.8.2. Flow on the Iberian Margin over the last glacial cycle Acknowledgements

8.1.

121 123 123 125 126 127 127 130 131 132 132 132 134 135 137 137 140 140 141 142

I NTRODUCTION

The transport and deposition of contourites, involving many cycles of erosion and deposition under intermittently strong deep-sea currents, produces fine sediments that show some sorting. Palaeocurrent reconstruction with high temporal Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00208-2

 2008 Elsevier B.V. All rights reserved.

121

122

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

resolution requires sensitive parameters from rapidly accumulating and continuously deposited fine sediments. As most structures are severely degraded by biological disturbance, grain-size parameters are needed and the mean size of medium to very coarse silt has proved to be a useful measure of the speed of the depositing flow. Because these sediments are cohesive, they pose several analytical problems not encountered in the study of sands and gravels. It is important to remember that for the finest sediments (but not for sands) the disaggregated state in which the samples are analysed is not the state in which they were deposited, because that involved varying degrees of particle aggregation. Therefore, dynamical inferences may not be made from the properties of the whole disaggregated size distribution. However, as particles increase in size, they become less prone to aggregation, and aggregates are more easily broken up by turbulent stresses. It was this fact that led McCave et al. (1995b) to propose the use of the 10–63 mm silt fraction, ‘‘Sortable silt’’ (mean size denoted by SS), as a flowspeed indicator, because the grains were more likely to have been deposited individually in response to fluid stresses. Because the particle size is not sensitive to the flow direction, but responds to the scalar flow speed, the latter term is used here rather than ‘‘f low velocity’’, which has vector connotations. Geologists have long related particle size to the speed of the depositing or eroding flow. In the great majority of cases, this has been for non-cohesive sands and gravels via grain size and sedimentary-structure analysis. The advent of the electrical-resistance pulse counter (Coulter Counter) in marine research (Sheldon and Parsons, 1967), especially the 16-channel Model T in 1970, provided new

5.40

23.7 U = 183.4–30 ϕ 21.8

5.64

20.1

5.76

18.5

5.88

17.0

6.00

0

5

10 15 20 Current speed (cm s–1)

Size (μm)

Mean size (ϕ)

r=

0.9

0

5.52

15.6 25

Figure 8.1 Ledbetter’s (1986) relation of fine-particle size to changes in current strength measured inVema Channel, South Atlantic; with permission from Elsevier.

123

I.N. McCave

high-resolution particle-size data for sizes down to 0.5 mm. In particular the work of Ledbetter (Ledbetter and Johnson, 1976), using a related instrument (Elzone counter), led the way for the use of fine-particle size as a flow-speed indicator. Palaeocurrent-speed studies have primarily focussed on contourite drifts, because they occur under important currents and are regions of high sediment-accumulation rate, giving good temporal resolution. For fine sediments, few attempts have been made to calibrate grain size to current strength, the only one of significance being Ledbetter’s (1986) (Figure 8.1). This indicates the path that must be followed to better understand the dynamical significance of contourites and unlock the messages they carry for the vigour of past deep-current systems.

8.2. 8.2.1.

S IZE ANALYSIS OF FINE SEDIMENTS

Measurement of fine grain sizes

Virtually all work on fine marine sediments starts from sample disaggregation and wet sieving to separate sand from mud at 63 mm. The sand is then dried and weighed prior to counting and analysis of foraminifers, lithic grains, volcanic ash, etc., and grain-size analysis of the fine fraction. Particularly comprehensive flow charts for sediment processing were given by Robinson and McCave (1994) and McCave et al. (1995b), modified as Figure 8.2 here. The need to process large numbers of fine-grained samples and the paucity of material often encountered in the SS range has promoted the development of new analytical instrumentation and techniques. This includes X-ray scanning settling tubes, laser diffraction size analysers and electrical sensing zone particle counters. Broadly, methods of size measurement may be divided into those which yield information on the whole size distribution, and those for which the information comes from a size window with upper and lower limits. Wholesale misunderstanding has arisen from attempts to attribute significance to differences in parameters obtained for the same sediment by different methods, for example when the mean or sorting from a Coulter Counter analysis with a 2–40 mm window is compared with a Sedigraph analysis over a 0–63 mm window. The Coulter Counter in this case does not ‘‘see’’ the clay, which may comprise a significant part of the size spectrum, and thus gives a coarser mean size. Several comparative studies of these instruments have been undertaken (Singer et al., 1988; Syvitski et al., 1991; Konert and Vandenberghe, 1997; Bianchi et al., 1999). They have fundamentally different theoretical underpinnings, so each analytical device has its disadvantages, and one instrument may be better suited for a specific application than another (McCave and Syvitski, 1991). Presently, the Sedigraph is the instrument of choice for the study of deep-sea sediments as proxies for current intensity, because it is based on the settling principle (Stokes’ Law) and therefore measures a ‘‘dynamical’’ grain size – a velocity-equivalent spherical diameter (ESD) distribution – which can be best related to transport and depositional processes. Bianchi et al. (1999) showed how the Coulter Counter – measuring volume-ESD via electrical-conductivity fluctuations of a suspension – gives similar results and constitutes a viable alternative

124

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

125

I.N. McCave

to the Sedigraph, especially when the relative abundance of material in the SS range does not exceed 5%. Laser sizers are based on the diffraction of coherent light caused by particles passing through a laser beam (Agrawal et al., 1991; Xu, 2000). This method is the fastest (if sample preparation time is ignored) and most precise for coarse (silty/ sandy) and equant particles, but has long been known to overestimate the size of clay particles compared to the settling method (see among others McCave et al., 1986). Konert and Vandenberghe (1997) showed that fine-particle platy shapes dominate size records and quantified this important problem, demonstrating that clays and silts with settling diameters of 2 and 16 mm are equivalent to 8 and 22 mm sizes, respectively, when measured by a laser particle sizer – a result that is consistent with earlier work based on the comparative study of marine sediments analysed by the Sedigraph 5000 and the Malvern 2600 E laser (Weber et al., 1991).

8.2.2.

Size distributions of components

The size distribution of compositionally distinct parts of the sediment can be obtained by determining the size distribution of the sediment before and after removal of the component and determination of its amount (Figure 8.2). The most obvious one is carbonate. Simply, the total fine fraction is size-analysed, carbonate is removed gently if clays are to be analysed later (e.g. by sodium acetate or EDTA), and the proportion of carbonate, C (where 1 > C > 0), is determined on a sub-sample. The size distribution of the residue is then determined. The size distribution of the carbonate is given by the difference between the size distribution of the total minus (1–C) times the size distribution of the residue (Figure 8.2). Formally, in unit mass of sediment, defining the size-distribution function f(dp) as f(dp) = dm/ddp, where dm is the mass of particles between size dp and (dp þ ddp), we have f ðdp ÞC ¼ f ðdp ÞT  ð1  CÞ f ðdp ÞR

Figure 8.2 Analytical flow chart for size analysis and composition slicing. This is a very complete scheme for sedimentological measurements and preparation of samples for almost all geochemical and isotopic data. Critical actions are in shaded boxes. There are four areas: (I) analysis of the bulk sample for carbonate and opaline silica content and, by difference, terrigenous matter content; (II) determination of mineral magnetic properties; (III) processing of coarse fraction (‘‘cf ’’) (sand) to yield samples for microscopic, isotopic and geochemical analysis; and (IV) processing of the f ine fraction (‘‘ff ’’) to yield size distributions of the terrigenous and biogenic components, as well as the composition of the f ine fraction in terms of terrigenous and biogenic (carbonate and opal) components. The principal operations are numbered and are shown in shaded boxes.The term ‘‘calcimetry’’ refers to determination of calcium carbonate content, which in this scheme is implicitly by a method which involves dissolution and removal of the carbonate, for example, by coulometry, or measurement of CO2 gas released in the Chittick apparatus. If carbonate is determined by the CHN analyser, a modification to this scheme is necessary. After Robinson and McCave (1994) and McCave et al. (1995b). A multicolour version of this figure is on the enclosed CD-ROM.

126

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

where the subscripts C, T and R denote the fractions of carbonate, total-sediment and non-carbonate residue; thus T = (C þ R) = 1 (McCave et al., 1995b). Figure 8.2 gives the analytical flow chart for the operations required to generate these data. When the sediment also contains a significant amount of biosiliceous debris (diatoms, radiolarians), further removal steps are necessary, rendering the whole procedure very involved. This technique was originally used by Paull et al. (1988) and Wang and McCave (1990). More recently, Trentesaux et al. (2001) and Frenz et al. (2005) have inferred carbonate grain-size distributions and foraminifer and coccolith components in hemipelagic sediments.

8.3.

U NSORTED D ELIVERY TO D EEP C URRENT SYSTEMS : P ELAGIC FLUX AND D OWN-SLOPE T RANSPORT

Sediment gets to the deep sea either by sinking from the surface (pelagic sediment), or by gravity-driven flow down-slope in turbidity currents and debris flows. This material is then resuspended and transported to sites of deposition to form contourite drifts. Typical sedimentation rates on drifts are 5–20 cm ka1, while pelagic rates are generally <2 cm ka1. Clearly, drifts represent much sediment focussing. The pelagic average rate results from the fact that the oceans are full of particles that are sinking, some slowly, some fast. The origin of most of this material is from biological production in the upper ocean, comprising organic matter, calcium carbonate and opal. It has been shown theoretically and by use of sediment traps that most of this material would be re-mineralised and not reach the sea floor, were it not for the process of aggregation which transforms its settling velocity spectrum. The aggregates are faecal pellets and ‘‘marine snow’’, loose aggregates bound together by organic mucous and gelatinous structures made by zooplankton. This material sinks at 100 m per day, a huge increase over the 2 m per day of a 5 mm coccolith. Radiochemical measurements also show that short half-life particlereactive elements (e.g. 234Th, t½ = 24 days) are being rapidly taken down to the sea floor (DeMaster et al., 1985). This probably involves scavenging by colloidal particles, initial Brownian aggregation and collection of the small aggregates by large rapidly sinking ones (McCave, 1984; Honeyman and Santschi, 1989; Hill and Nowell, 1990). The result is that the particle population delivered to the sea floor is not appreciably altered or sorted by the processes involved in vertical flux. A terrigenous mixture delivered to the sea surface (e.g. ice-rafted detritus (IRD) or aeolian dust) should arrive at the sea floor with most of its fine particles. The vertical flux should deliver records of biological productivity, carbonate and silica skeletal mineralisation, IRD flux, volcanic eruption magnitude, and wind strength. The sizes of carbonate and silica materials are likely to be distorted in transit by dissolution. Only if there is very slow flow in the bottom boundary layer will these aggregates plummet down directly onto the bed and remain there. This may be true for some of the ocean in mid-gyre regions some of the time, but contourites occur

127

I.N. McCave

under relatively fast currents which can resuspend sediment, break up aggregates and control deposition. Close to continental margins, sediment is carried down-slope in copious quantities by turbidity currents and debris flows that are responsible for sediment accumulation to form much of the continental rise. The mass failures yielding large turbidity currents and debris flows involve failure of ten to several tens of metres of sediment on the slope (Embley and Jacobi, 1977), thereby mixing the products of several glacial/interglacial cycles (Weaver and Thomson, 1993). Basal erosion by turbidity currents will also produce a mixture of sediment ages and properties. Gravity flows do not, therefore, deliver sediment bearing a pure signature characteristic of conditions in the source area at the time they were triggered. Strong deep-sea currents resuspend this sediment, forming nepheloid layers, and move it to areas of spatial decrease in flow speed, causing deposition. These turbid layers are found all over the sea bottom, some more concentrated than others (McCave, 2001). Most fine sediment deposition involves transport in and removal from nepheloid layers. Continental margins thus have much material that is moved in bottom nepheloid layers and rained out of suspensions having varying origins. Processes acting in them are responsible for sorting fine sediments.

8.4. 8.4.1.

C ONTROLLING FACTORS FOR I NPUT AND TRANSPORT

Settling velocity of grains and aggregates, ws

The still-water settling velocity (ws) of spheres follows Stokes’ Law, ws = Dgd2/ 18, as long as Rep < 0.5 or d < 100 mm in water (Reynold’s number, Rep = wsd/, where  is the molecular viscosity, and kinematic viscosity  = /, d is diameter, D = (s – ) is sediment minus fluid density) (see the list of fluid flow variables and symbols in Table 8.1). However, particles are often not spheres, so Lerman et al. (1974) assembled the relationships for the settling velocity of nonspherical particles in the viscous range from which it is seen that deviations of up to a factor of 3 slower occur, depending on shape. The molecular viscosity of sea water ranges mainly from 0.97 to 1.88  103 Pa s (25–0C), giving nearly a factor of two variation in ws (Winegard, 1970). Density exerts a strong control, particularly for aggregates made up of quartz–carbonate density solids of 2500–2900 kg m3, plus neutrally buoyant organic matter, yielding particles of much lower saturated bulk density, often <1100 kg m3 (McCave, 1984). Bulk density variations of particles that are not solid – aggregates, hollow particles and grains containing gas bubbles – are very large. The most important classes of hollow particles are foraminifers, diatoms and radiolarians. Foraminifers on the bed are often partially sediment filled, resulting in saturated bulk densities (that is the mean density of foraminiferal shell plus sediment and water in the cavities) of 1150–1550 kg m3. This gives a large range of D, from 100 to 500 (for simplicity we use a deep-sea water density of 1050 kg m3), which translates into a settling velocity of 125–500 m per day for 200 mm foraminifers. Aggregates typically have settling velocities of 1–2 mm s1 or 85–170 m per day. With variable size,

128

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

Table 8.1 Fluid-flow variables: symbols, units, dimensions, typical deep-sea values Parameter

Symbol

Unit

Dimension

Flow depth Turbulent kinetic energy dissipation rate Kolmogorov microscale (/G)1/2 Sediment diffusivity Shear rate ("/)½ Acceleration due to gravity Fluid density Sediment density Density difference Probability Rate of deposition Time Shear stress Shear velocity Grain size Dynamic viscosity Kinematic viscosity Concentration (by mass/vol) Space coordinates Velocity components (w, vertical) Stokes settling speed ‘‘Viscous sub-layer thickness ‘‘Sortable silt’’ mean size Dimensionless numbers Reynolds number

D E

m m2 s–3

L L2T3

lk

m

L

"s G g

m2 s1 s1 m s2

L2T1 T1 LT2

 s D = s   p Rd t t u (=(t/)½) d, dp m  (=/) C

kg m3 kg m3 kg m3  kg m2 s1 s Pa (=N m2) m s1 m, mm, mm Pa s m2 s1 kg m3

ML3 ML3 ML3  ML2T1 T ML1T2 LT1 L ML1T1 L2T ML3

x, y, z u, v, w

m m s1

L LT1

ws = Dgd2/18 v  10/u

m s1 m

LT1 L

SS

mm

L

Re

wsd/, ud/, etc. to/Dgd Dgd3/ 2 ws/ u ( = "s/ "m  1)

Shields number Yalin number Rouse number

  

von Karman’s constant



Value in water (SI) at 2C

9.8 1050 2650 1600

1.75  103 1.67  106

0.41

density and viscosity, particle sinking rates can be from 50 to 1000 m per day. This gets material from surface to bottom in a matter of days to a few weeks. Aggregates may be formed by physical (often referred to as flocculation, sometimes coagulation) or biological processes, frequently involving feeding and production of faeces or mucous. The finest particles (d below 1 mm) are brought into

129

I.N. McCave

contact by molecular buffeting known as Brownian motion. This is favoured by high particle concentration and high temperature. Bigger particles are brought together either by turbulent shear or by larger, fast-sinking particles sweeping up finer ones, like rain falling through mist (McCave, 1984). In both cases, larger particles grow more rapidly. A key feature of much aggregation is the presence of organic mucous, which acts as ‘‘glue’’, allowing particles which get close to stick. Many organisms produce mucopolysaccharides, notably bacteria which sit on particles. The importance of aggregation is shown in Figure 8.3, where the ‘‘flocculation factor’’ (the ratio of the settling velocity of an aggregate to the settling velocity of the primary particles of which it is made) can be up to 105. Aggregates are not stable entities. The organic membrane covering faecal pellets decays and the mucous that holds aggregates together also degrades, so some particles fall apart while sinking (Honjo and Roman, 1978; Lampitt et al., 1990) Aggregates assembled by moderate levels of turbulence in the outer part of the boundary layer may be broken up by more energetic turbulent eddies close to the boundary. The relationship between the aggregate size, df, and the shear stress, t, is very poorly known. If the floc diameter df / tn, then n ranges from 0.25 to 1. The largest aggregates, which can be up to 5 mm in diameter, are found in the

105

Flocculation factor, F = ws/wsd

104

C = 10 kg m–3 S = 30 Still water

103

102

10

1

C = 1.2–11 kg m–3 S=0 Flowing water

0.0

1

10

100

ds

Figure 8.3 Flocculation factor F (ratio of floc settling velocity to settling velocity of the primary particles of which it is made) versus diameter of the primary particles.The main curve is for still-water settling, the short one is with flow. Note that F is very small in flowing water for d50 > 10 mm. After Migniot (1968) and Mehta (1986).

130

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

moderately turbulent, high-concentration mid-water region of estuaries. Closer to the bed, a high shear breaks these large sloppy aggregates into smaller pieces. The density of flocs (f) decreases as they increase in size. A simple expression based on field data is Df = 4.9d0.61 , where the floc excess density, Df = (f – ), is in f kg m3 and the floc diameter, df, is in mm. This yields floc excess densities down to 10 kg m3, but nevertheless – because the settling velocity goes up as the square of the diameter – large aggregates can settle in the open ocean at the same speed as fine quartz–sand grains, i.e. over 200 m per day.

8.4.2.

Boundary-layer flow structure and turbulence

A fluid flowing over a surface exerts a drag force on it. The drag at the boundary slows the fluid down, but some distance out, known as the boundary-layer thickness, the average flow speed no longer changes much with distance. In the deep sea, the boundary layer extends several tens of metres above the surface. Near the bed, boundary layers are intensely turbulent and the drag force, t0, exerted on the bed is related to that intensity because the stress is transmitted by eddies. In the vertical plane, t0 ¼ u 0 w 0 , where u0 is the turbulent component in the flow direction and w0 is the vertical component. This expression is important because u0 and w0 are related, so that t0 / w 0 2. This vertical turbulent velocity is responsible for keeping particles in suspension, and the turbulent stress, u 0 w 0 , either causes aggregation or, at higher values, disaggregates fine particles. The term (t0/) has the dimensions of a velocity squared and that velocity is called the shear or friction velocity, u = (t0/)1/2. From the above, it can be deduced that u / w 0 : In a turbulent flow, the speed decreases logarithmically towards the bed because of the drag, so very close to the bed the flow is slow and becomes laminar, or at least dominated by viscosity, in a layer known as the viscous sub-layer of the turbulent boundary layer. This layer is very thin (thickness v). In water, for a flow that just moves very fine sand (u = 0.01 m s1,  = 106 m2 s1), v = 10/u is just 1 mm thick. However, this is ten times the diameter of very fine sand. The shear across this layer is large; for u = 0.01 m s1, the speed goes from 0 to 0.1 m s1 in just 1 mm. Weak aggregates cannot survive this shear and break up (Hunt, 1986). Above this sub-layer, there is an intensely turbulent transition (‘‘buffer layer’’) to a region in which the flow speed varies as the logarithm of height above the bed (Figure 8.4). As the flow speed decreases, u decreases and v increases, so that in deposition most particles arriving at the bed have passed through the viscousdominated layer of 2–10 mm thick. Although viscous-dominated, this layer is not actually laminar. Spatially it has a structure of high- and low-speed streaks, and temporally very high-speed bursts of fluid out of the layer and sweeps of fluid into it from outside. These are associated with stresses typically up to ten times the average (and extremes of thirty times), so even strongly bound particles may be ripped apart on approaching the bed. Equally, weaker aggregates that have made it to the bed and are partly stuck to it, may also disintegrate with their components either ejected into the main flow or redeposited. These are powerful size-sorting mechanisms.

131

I.N. McCave

103 Logarithm

Linear

Wake, non-log

Uz

Logarithmic layer

u* z/v

102

Turbulence dominant

Buffer layer

10

Viscous sublayer 5

Viscosity dominant

10 15 20 Uz /U

*

Uz

Figure 8.4 Regions of the turbulent boundary layer for a flow at 1^10 m deep. In the centre, the linear representation of flow speed versus height cannot resolve the viscous sub-layer, but speed versus logarithm of height (z, expressed as a Reynolds number) shows it very well. From McCave (2005), with permission from Elsevier.

8.5.

S ORTING IN S USPENSION TRANSPORT

Once the material is moved out of the near-bed region, it is held in suspension by the action of fluid turbulence. Because the vertical turbulent component of velocity is about the same as the shear velocity u, the normal suspension criterion is that ws/u £ 1. Particles in steady transport are diffused up from the source at the bed and sink back down under gravity, with a balance in steady state. This is expressed as dC Cws þ "s ¼0 dz where the first term is gravity settling and the second is upward diffusion ("s is sediment diffusivity, the same value as fluid eddy diffusivity for fine particles in dilute suspension). The result of this is that, for a given value of u, the fastersettling grains are found closer to the bed and the finer, slow-settling particles are more uniformly distributed over flow depth. In the bottom of a deep flow, the concentration at height z in the flow Cz = Ca(a/z) , where Ca is a reference concentration at height a, and  = ws/ u, in which is von Karman’s constant, 0.41. This means that with the suspension criterion, ws £ u, we would not expect much material in suspension for  > 2.5. Relatively fine material (with  < 0.125) is distributed through the whole flow. For sediment in slow deep-ocean flows where u may be in the range 0.1–0.5 cm s1, this value of  gives values for ws of 0.005–0.025 cm s1, equivalent to solid-particle sizes of 1021 mm. Closer to the bed, the relatively coarser sediment is concentrated (coarse to very coarse silt). This travels more slowly than the finer particles of low settling velocity that are higher up in the flow, so sorting is produced by the flow speed differential with the finer material travelling far downstream. This gives a large-scale (500 km) fining along

132

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

sediment drifts. Clearly, if the flow slows down, the coarser material will be more rapidly deposited because it is both closer to the deposition threshold and has very little distance to go to reach the bed.

8.6. 8.6.1.

P ROCESSES OF D EPOSITION FROM T URBULENT B OUNDARY LAYERS

Rate of deposition

‘‘Deposition’’ of bed load is rather straightforward: it stops moving. This occurs in water at a shear stress only slightly lower than the critical erosion stress. In fact, we cannot measure the erosion stress precisely enough to distinguish between erosion and deposition stresses, so they are effectively the same, though one would expect a difference analogous to the coefficients of static and sliding friction. For suspended sediment, once the stress has decreased below the suspension threshold, material will sink into the near-bed region, thereby increasing the concentration and causing some material to be deposited. Experiments in flowing water by Einstein and Krone (1962) showed that – for initial concentration C0 < 300 g m3 – below a certain shear stress td, the concentration in suspension, Ct, decreases exponentially with time (t):  w pt  s Ct ¼ C0 exp D where D is the depth of flow (or thickness of the boundary layer) and p is the probability of deposition, the probability being given by p = (1 – t0/td).

8.6.2.

Limiting shear stress for deposition td

In the expression given above, td is the limiting shear stress for deposition, the stress below which all the sediment will eventually deposit. This yields Rd ¼ Cb ws ð1  t0 =td Þ for the rate of deposition, Rd (kg m2 s1). Here, Cb is the concentration near the bed. If there is no flow, the deposition rate reduces simply to the settling flux, Cbws. The critical deposition stress (td) is the stress below which particles of a given settling velocity will deposit, while those of smaller settling velocity will be ejected from the viscous sub-layer. The value of td is not well known. McCave and Swift (1976) suggested that it is probably related to the diameter, or, more properly, to the settling velocity of the particles, whether aggregates or single grains. They assumed that it is given by the critical erosion stress (te) for non-cohesive grains because below that value movement ceases and any grain reaching the bed would stick. This is shown in the critical-erosion diagram on non-dimensional axes,  = t0/Dgd, and

133

I.N. McCave

100 Suspension Su

Bagnold

sp

en

sio

n

θ

Ero

sion

No movement

10–2 10–1 8.5 0.066

100 39 1.4

ld

sho

Thre

10–1

Bedload

thre

sho

ld

101 182 18.7

102 846 116

103 3930 445

X½ d (mm) ws (mm s–1)

Figure 8.5 A critical erosion diagram on non-dimensional axes with critical suspension lines added which divide the diagram into regions of suspension, bedload and no movement. Below the suspension threshold, material falls out and deposition will ensue.Two suspension lines are shown,‘‘suspension threshold’’ results from the view that at critical movement conditions the turbulent intensity can hold particles up, so it depends whether they are ejected from the viscous sub-layer (calculated from the data of Nin‹o et al., 2003), while an alternative view ‘‘Bagnold’’ is that sub-layer ejections are always fast enough, so it depends whether the vertical turbulent velocity can hold the particles up after injection into the flow (after Bagnold, 1966). Axes are  = t0/Dsgd, 1/2 = (Dsgd3/ 2)1/2.The three rows of values on the x-axis are for 1/2, d (grain size) and ws (settling velocity). From McCave (2005), with permission from Elsevier.

 = Dgd3/ 2, with the variables defined previously (Figure 8.5). An alternative, based on measurements in a laminar flow cell, is td = 0.048Ds gd (Self et al., 1989). This expression is equivalent to td  103d (in SI units), thus deposition of 10 mm particles occurs at stresses <0.010 Pa (shear velocity u <0.32 cm s1 , geostrophic flow speed Ug < 7–10 cm s1). These values are rather lower than the critical deposition stresses from McCave and Swift (1976), which yield td  0.045 Pa (u = 0.67 cm s1, Ug = 15–20 cm s1) for 10 mm silt, or via the analytical expression of Dade et al. (1992) giving 0.015–0.03 Pa (Ug = 8–23 cm s1). The trend of these curves, however, is the same. This range of shear stresses (0.01–0.045 Pa) is a little lower than that indicated by Hunt’s (1986) experiments for the breakup of montmorillonite and illite flocs due to shear, namely 0.04–0.16 Pa. The implication is that most aggregates less than about 10 mm in diameter will survive during deposition from currents. However, above that size aggregates are increasingly likely to be broken up in the ‘‘buffer layer’’ located just above the viscous sub-layer where large turbulent shear fluctuations occur (Figure 8.4). Hydrodynamic processes of sorting in the viscous sub-layer will thus tend to act on primary particles for sizes greater than 10 m. Consequently, under stronger flow this material will be size-sorted according to its primary grain size. This is the basis for designating the 10–63 mm fraction as

134

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

‘‘sortable silt (SS)’’, because it is sorted by its primary particle size, whereas finer silt is not because it occurs in aggregates. The SS component is a useful index of the flow speed of the depositing current.

8.6.3.

Sorting by selective deposition: cohesive versus sortable silt

Sorting takes place through differing rates of sediment transport, so that an originally unsorted mixture is converted downstream into narrower distributions. During deposition, particle populations are sorted by some grains and aggregates being trapped in the viscous sub-layer, while those of lower settling velocity are not, and are transported further downcurrent. There is a substantial region of overlap between the trapped and rejected populations. The controlling variables are the critical erosion stress (te), the critical suspension stress (ts) – which may be greater than the erosion stress for fine non-cohesive silts according to Dade et al. (1992) – and the critical deposition stress (td). In general, td < te < ts, giving a bedload region for fine sediment between te and ts (Dade et al., 1992) in which silt ripples are found (Mantz, 1978). Sorting of muds thus arises mainly from selective deposition. Selective erosion (winnowing) is not such an important sorting process for cohesive muds, as the cohesion means that there is little selective removal of the particles which are most prone to stick together (Winterwerp and Van Kesteren, 2004). It can make the deposit somewhat coarser overall by producing intermittent erosion horizons marked by coarse silty and sandy lags, but at the cost of decreasing accumulation rate (McCave and Hall, 2006). Fine-sediment distributions may have a moderately well sorted silt mode but overall are poorly sorted. The fact that fine sediment distributions do not become well sorted is due to the effect of aggregation, which causes smaller particles to be deposited with larger individual grains where strong flocs and grains have similar settling velocities. Mehta and Lott (1987) argued for selective deposition on the basis of the presumed relationship between the critical deposition stress and settling velocity, and the likelihood of aggregate break-up close to the bed. The presumed floc break-up near the bed appears to be borne out by some recent measurements in shallow water, where particle size in suspension decreases towards the bed in the bottom 1.0 m of the flow (Figure 8.6) (Fugate and Friedrichs, 2003). This is still far above the buffer layer, but is much closer to the bed than the regions of flow in which large flocs are usually measured (Dyer et al., 1996). These particles, conditioned by high near-bed turbulence are deposited, whereas larger, weaker flocs formed higher up in the flow are transported but then broken up as they get close to the bed. Sorting thus occurs by selective deposition of particles selected on the basis of settling velocity. In regions where the currents frequently exceed about 20 cm s1, the fine components are removed, leaving a lag of sand. At equatorial to mid-latitudes, this is most frequently foraminiferal sand with a particle density of 1.1–1.5 g cm3 and a critical erosion shear stress of 0.067–0.085 Pa for 200–300 mm sand (Miller and Komar, 1977) (Ug = 24–27 cm s1). Although medium foram sand has the erodibility of very coarse silt, it has the settling velocity of very fine sand and is not transported far, being largely confined to bedload. To move quartz sand of this

135

I.N. McCave

Median particle size (μm)

300

100 d50 = 0.55λ k

100

0.3 mab 1 mab

300 Kolmogorov microscale (μm)

1000

Median particle size (μm)

(b)

(a)

300

100

10

30 100 Total suspended solids (gm–3)

300

Figure 8.6 Data on suspended sediment size and concentration in a boundary layer near the bed. (a) Data from a laser in situ sizer (LISST) within 1 m of the bed (1 mab) showing a dependance of particle aggregate size on Kolmogorov microscale length which decreases as shear increases closer to the bed (microscale lk = (/)1/2, where  is the kinematic viscosity and  is the shear rate ("/)1/2 in which " is the turbulent kinetic energy dissipation rate). (b) Particle size decreasing with increasing concentration low in the boundary layer because of shear break-up of aggregates (whereas increasing size would normally be expected as collisions are more frequent at higher concentration). From Fugate and Friedrichs (2003), with permission from Elsevier.

size would take currents of Ug = 35–37 cm s1. The speeds required to move foraminiferal sand are generally not exceeded frequently enough to yield a wellsorted sand, though sand dunes have been detected on the lower flanks of Hatton Drift (McCave et al., 1980). The significance of this is that the deposits of lateral flux driven by deep-sea currents are in the silt/clay size range and are sorted to a greater or lesser extent depending on currents, whereas those resulting from vertical flux comprise both fines, foraminiferal sand and, during glacials, quartz sand and gravel. Only the sandand gravel-sized material may be unambiguously related to the properties of the overlying water column.

8.6.4.

Repeated sorting events under deep-sea storm conditions

Size sorting on beaches is achieved by repeated sorting events under waves. It is actually controlled by the settling velocity of the particles determining whether they enter the deposit or move offshore in suspension. The deep-sea setting provides an analogous situation because of the unsteadiness of deep-sea currents resulting in so-called ‘‘deep-sea storms’’, periodic resuspension–transport–deposition events (Hollister and McCave, 1984; Gross and Nowell, 1990; Gross and Williams, 1991). As seen in Figure 8.7, the response of the suspended sediment concentration field to forcing is complex, with substantial periods of time with flow speed below 10 cm s1 during which deposition occurs (Gross and Nowell, 1990; Gross and Williams, 1991). Regions of deep storms were identified by Hollister and McCave (1984) as areas of high abyssal eddy kinetic energy,

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

Concentration

U

Concentration

U

136

Figure 8.7 Part of the 1985^1986 HEBBLE record of (a) flow speed, (b) turbidity and (c) shear stress (as t0/ = u2) in the bottom boundary layer (speed at 4.9 m above bottom (mab), turbidity at 2.2 mab) on the Nova Scotian Rise (4830 m water depth) (from Gross and Williams, 1991; with permission from Elsevier).The response of the suspended sediment concentration field to forcing is complex with evidence of (A) local erosional supply, (B) advection of turbid pulses created by erosion elsewhere, (C) supply limitation, possibly controlled by the surface reservoir of biologically conditioned suspendable material running out, (D) concentration decrease due to dilution in a thickening boundary layer and (E) depositional periods with flow speed below 10 cm s1. Lines have been put on the speed at 15 cm s1, about critical erosion conditions. (Avalue of t0/ of 0.5 cm2 s2 (u = 0.71 cm s1) is equivalent to Ug  13 cm s1.)

occurring particularly in western ocean basins and the Antarctic Circumpolar Current. Evidence of strong current variability has been found on several sediment drifts by current and/or turbidity records. It is a key feature of the transport/deposition system because of the profound asymmetry of erosion and deposition (a few minutes of erosion can remove deposits formed during several weeks). Putting the material into suspension takes a short time in storms, and the intervening time interval is then occupied by sorting through selective deposition.

137

I.N. McCave

8.7. 8.7.1.

D EPOSITS FROM C URRENTS

Spatial variability of sediment size over bedforms and drifts

In drift settings, sediment focussing implies that with increasing current speed there must first be an increase in accumulation rate, followed by a decrease as intermittent erosion events start to dominate (Figure 8.8) (McCave and Hall, 2006). Eventually, above some relatively high value of flow speed, probably 20–25 cm s1, i.e. critical for movement of foraminifers (Miller and Komar, 1977), net removal Interstitial

C + FS 50 SS

Framework

Grain-size components percent

100

Sand

0 Sorting by selective deposition

Selective deposition & winnowing

Accumulation rate (cm ka–1)

20

C + FS 10

SS 2 0 0

5

10

15

20 –1

Flow speed (cm s )

25 Sand

Figure 8.8 Hypothetical variation of sedimentation rate with increasing current speed shown as deposition flux (below) and component percentage (top). C þ FS = clay and fine silt (<10 mm), SS = sortable silt (10^63 mm). At zero speed, a pelagic rate of 2 cm ka1 is assumed, and erosional winnowing is assumed to occur above 20 cm s1. Above 20 cm s1, mud will be mainly interstitial in sand, and above 30 cm s1 (not shown) the sand will be sufficiently mobile to contain little mud at all and be forming ripples and sand waves.The curve is shown as peaking between 10 and 15 cm s1 but this is not at all well known and may well be dependent on the magnitude ^ frequency structure of deposition and erosion events. A peak between 5 and 10 cm s1 would be entirely feasible as some records suggest the onset of surface erosion above 10^12 cm s1. From McCave and Hall (2006). A multicolour version of this figure is on the enclosed CD-ROM.

138

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

of fines by winnowing occurs, leaving a sandy residue. One expects to see this type of pattern expressed spatially on short and large length scales (mud waves to drifts), expressing the spatial variability of flow. A simple theoretical justification of the form of the curve in Figure 8.8 is as follows. The rate of sediment deposition is given by R = Cbws (1 – t0/td). Shear stress t / U2g, flow speed squared. Comparing cases with the same concentration and settling velocity, then R / (1 – U2g/U2d). If we normalise Ug by setting U0 = Ug/Ud (so if Ug is 10% of Ud, which in reality might be 1.5 cm s1 for Ud = 15 cm s1, U0 = 0.1), then R / (1 – U0 2). Representing the advection of sediment to a site by CUg, but, as before, comparing cases with the same concentration, then sediment advection is proportional to Ug, which we also normalise by Ud. Now, if Ug = 0, no new sediment is advected so the deposition rate is 0 (after an initial clear-out of the water column). If Ug = Ud, then similarly there is no deposition. Taking the deposition rate to be proportional to the product U0 (1 – U0 2), we have simply R / (U0 – U0 3). That function is plotted in Figure 8.9 and it can be seen that it bears a close similarity to the form of Figure 8.8. Sediment properties vary over mud waves with wavelengths of 1–5 km (Ledbetter, 1993; Manley and Flood, 1993a) because flow near the bed is strongly controlled by local topography. Mud waves can be either like dunes migrating in the direction of flow, or, more commonly, like anti-dunes responding to in-phase lee-wave disturbances in the stratified water column (Flood, 1988). In the latter case, the flow slows down on the upstream face, yielding a maximum deposition rate and speeds up over the downstream face, resulting in slower deposition or even erosion and coarser silt. The theories of Flood (1988) and Blumsack and Weatherly (1989) for the lee-wave case are corroborated by the grain-size work of Ledbetter (1993) showing larger sizes on the upstream 0.4

U′ – U′3

0.3

0.2

0.1

0 0

0.2

0.4

0.6

0.8

1

U′

Figure 8.9 The deposition rate function (U0 ^ U0 3) combining advection and sub-layer effects as a function of increasing flow speed up to the depositional limit U0 = 1. (U0 = Ug /Ud, the ratio of geostrophic flow speed to critical deposition flow speed). A multicolour version of this figure is on the enclosed CD-ROM.

139

I.N. McCave

East

West

3.5 kHz

10 kHz

Figure 8.10 Acoustic profiles across a sediment drift. Upper figure: 3.5 kHz WNW-ESE profile across northern Gardar Drift at 60°N, 23°W, showing a reduced sedimentation rate on the eastern slope of the drift by closer spaced reflectors. Lower figure: 10 kHz echo sounder record along the same track showing higher amplitude reflection (redder) of the ‘‘harder’’ sea floor that is more strongly affected by currents, corresponding to coarser size and slower net accumulation rate (from McCave, 1994). The coarser sediments on the eastern side of the drift are shown by Bianchi and McCave (2000, their Figure 19) to be finer than on the western side. A multicolor version of this figure is on the enclosed CD-ROM.

face. The flow is also topographically steered by large sediment ridges (drifts). This results in the flow on one side being stronger than on the other, due to Coriolis effects. The two sides of Gardar Drift show clearly the reduced sedimentation rate and increased particle size on the eastern slope of the drift (Figure 8.10). This results in coarser sediments on the eastern side of the drift compared with the western side (Bianchi and McCave, 2000). There is thus a potential for aliasing of inferred flow speeds downcore. In other words: a change in grain size observed downcore may not represent a temporal change in flow speed but a spatial change in speed from one side of a bedform to the other. This is not very likely for the movement of drifts, which are too large to move much even on a 1 Ma scale. Mud-wave migration on the 0.1–1 Ma timescale, however, could result in the stoss and lee sides of waves being sampled successively, with downcore size variations giving the impression of a time-varying flow, even if the flow were constant. Such complications may not be confined to grain size but also extend to other proxies.

140

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

8.8. 8.8.1.

S OME EXAMPLES OF PALAEOFLOW INFERRED FROM SORTABLE -SILT RECORDS

Gardar Drift and Bermuda Rise at the penultimate glacial termination

Sortable silt mean size (SS) records from two cores from different North Atlantic basins – but both under the influence of North Atlantic Deep Water (NADW): one close to its origin and the other downstream in its evolution – are illustrated in Figure 8.11. The first is from (core NEAP-18K, 52460 N, 3020.70 W, 3275 m water depth) at the southern extremity of Gardar Drift. This was deposited from Iceland– Scotland Overflow Water (ISOW) in the Iceland Basin (Bianchi and McCave, 2000). Core MD95-2036 was taken from eastern Bermuda Rise (3341.40 N, 5734.60 W, 4462 m water depth), where the sediments are also deposited under

Figure 8.11 Sortable silt mean size, SS (circles), in core NEAP-18K (Gardar Drift) and (open squares) in core MD95-2036 (Bermuda Rise). Letters A to D distinguish events of acceleration (a) and deceleration (d) which, on the basis of independent age models for the two sites, match very closely in time, demonstrating coordinated flow changes across the North Atlantic (from Hall et al., 1998; with permission from Elsevier). The SPECMAP mean benthic isotope curve of Martinson et al. (1987), used for age and correlation, is shown for reference.

141

I.N. McCave

current influence (McCave et al., 1982; Laine et al., 1994). This core is also sited under NADW at present, with Southern Source Water (SSW, originally Antarctic Bottom Water) at greater depths over the adjacent Sohm Abyssal Plain (5500 m depth). Because bottom currents exert primary control on sediment deposition and there is negligible interference by direct fall-out from icebergs at either site during the interglacial, they are optimal locations to record bottom-water current changes. The two cores show several synchronous events (labelled A to D) in the SS record, interpreted as deceleration (letter d) followed by acceleration (letter a) of the deep current (Figure 8.11). Core NEAP-18K shows an abrupt and large decrease in the SS (from about 16 to 11 mm (Figure 8.11, event Ad), indicative of slowing flow speeds, from 13l–130 ka, just before full stage 5e. Inferred flow speeds at both sites increase rapidly into full interglacial conditions (event Aa), suggesting a strengthening NADW flow. The most notable feature in both SS records is the sharp decrease in the flow speed at about 120 ka, reaching a minimum value at 119–118 ka (event Bd). During that event between 120 and 118 ka, the abundance of the warm species Neogloboquadrina pachyderma (dex.) (now called N. incompta) in NEAP-18K decreases from 35% to <5% and planktonic 18O values increase by 0.65%, which, with negligible ice-volume effect, suggest a cooling of the northward-flowing surface water temperature of 2–3C. These data indicate that the flow slowed down in an event that marked the end of the last interglacial, and that was not clearly marked at this time in any chemical proxy. The flow-speed evidence here was the key to seeing the role of the circulation in this major climate shift.

8.8.2.

Flow on the Iberian Margin over the last glacial cycle

The sortable-silt current-speed proxy, SS shows a consistent temporal pattern of slow flow in cold, and faster in warm intervals in core MD95-2040 at 2465 m on the northern Portuguese Margin (Figure 8.12) (Hall and McCave, 2000), with the exception of stage 2. Another core from this margin (OMII-9K) indicates slow flow in stage 2, as do others further north, suggesting that the MD95-2040 record is anomalous at this point. Stage 2 mass accumulation rates in the latter core are high ( > 20 g cm–2 ka1), suggesting that the material may in part have been supplied down-slope at such a rate that the weak current could not sort it and develop an SS signal appropriate to the flow speed. This view is supported by the fact that Heinrich layers with similar accumulation rates also show high values of SS. The record is quite consistent with warm=fast and cold=slow, and thus either stage 2 is anomalous in core MD95-2040, or stages 4, 5b, 5d and 6 are. If there is an unsorted source effect on SS in stage 2, then there is not in the other stages. It coincides with the lowest stand of sea level. Elsewhere on this margin, this time is marked by down-slope transport with distinctive sedimentary structures (Baas et al., 1997). This illustrates both the utility of the method and the pitfalls associated with drifts on continental margins. Far the best are detached drifts such as Gardar and Blake Outer Ridge in the North Atlantic.

142

Size Sorting During Transport and Deposition of Fine Sediments: Sortable Silt and Flow Speed

Figure 8.12 Sedimentological parameters for core MD95-2040 on the Iberian Margin plotted against age (from Hall and McCave, 2000; with permission from Elsevier). (a) SPECMAP stack of benthic 18O values (Martinson et al., 1987). (b) Terrigenous silt/clay ratio (wt% 2^63 mm/wt% <2 mm). C: SS (mm); the line under the data in stage 2 indicates the part of the record believed to be unreliable as a current indicator because of possible down-slope contamination. The vertical dashed lines indicate the boundaries between Marine Isotope Stages (MIS) achieved by correlating the benthic isotopic record for the core to SPECMAP. Note the clear relationship of slower flow in cold/cool periods from stages 6 to 3 with a lag of a few thousand years relative to the SPECMAP isotope curve.

These examples, and many others reviewed by McCave and Hall (2006), demonstrate the use to which records of fine sediment size and sorting in drifts can be put in the service of palaeoceanography and its relationship to climate change.

ACKNOWLEDGEMENTS I am most grateful to David Thornalley for reading this chapter, and for the attention of the series editor, which allowed me to clarify this chapter at several points. The work summarised here has been supported by funding agencies in the UK (NERC) and USA (ONR) as well as by Cambridge University, to whom I am most grateful.

P A R T

4

SEDIMENTS

C H A P T E R

1 0

T RACTION S TRUCTURES IN C ONTOURITES J. Martı´n-Chivelet1, M.A. Fregenal-Martı´nez1 and B. Chaco´n2 1

Department of Estratigrafı´a Facultad de Ciencias Geolo´gicas, Universidad Complutense, Madrid, Spain 2 Fachbereich Geowissenschaften, Universita¨t Bremen, Bremen, Germany

Contents 10.1. 10.2. 10.3. 10.4.

Introduction The Traction-Structures Controversy Setting the Stage for Sedimentary Structures in Contourites Sedimentary Structures 10.4.1. Current ripples and cross-lamination 10.4.2. Lenticular and flaser bedding 10.4.3. Horizontal, subhorizontal, and sinusoidal lamination 10.4.4. Large-scale cross-bedding 10.4.5. Erosional structures 10.4.6. Grading 10.4.7. Less common structures 10.5. Sedimentary Structures in Facies Models 10.5.1. The Faro Drift facies model 10.5.2. The Gulf of Mexico facies model 10.5.3. The Caravaca facies model 10.6. Conclusions Acknowledgments

10.1.

159 160 161 162 162 170 171 172 174 175 176 177 177 179 181 181 182

INTRODUCTION

Even after 40 years of study of contourites, their sedimentary structures continue to be a matter of debate, and probably they are the most controversial topic in contourite research. There is no general agreement yet regarding such basic aspects as (1) what kind of structures can be generated in contour-current settings, (2) which of these structures can be preserved in the stratigraphic record, and (3) which of these structures can be used as diagnostic features to distinguish contourites from other deep-sea deposits such as fine-grained turbidites. In this chapter, an extensive review is presented of the structures that are formed in deep marine systems under the influence of oceanic bottom currents. Some general and theoretical constraints on the range of likely processes forming primary Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00210-0

Ó 2008 Elsevier B.V. All rights reserved.

159

160

Traction Structures in Contourites

sedimentary structures in contourites are established. Each sedimentary structure is documented and analyzed under those constraints, and the main controversial points are discussed. Particular attention is paid to small- and medium-scale traction structures (larger bedforms and sediment waves are described in Wynn and Masson, 2008).

10.2.

T HE T RACTION -STRUCTURES C ONTROVERSY

The discovery in the 1950s of current ripples and other traction structures in deep marine deposits accumulated or affected by bottom currents permitted the recognition and establishment of the contourite facies (Hollister, 1967; Hollister and Heezen, 1967, 1972; Bouma, 1972a, b, 1973; Bouma and Hollister, 1973). These works provided the first set of reliable criteria for the recognition of contourites in both ancient and modern deposits. Good sorting, thin bedding, normal and reverse grading, absence of massive bedding, cross- and horizontal lamination, and grain orientation and inclination were proposed as diagnostic of such deposits. The seed of the controversy appeared soon after, with the discovery and characterization of another type of facies: fine-grained turbidites (Piper, 1972, 1976, 1978; Piper and Brisco, 1975). Such turbidites may show similar sedimentary structures (good sorting, thin bedding, and horizontal or cross-lamination) and can be difficult to distinguish from contourites. This finding marked the need of looking for new diagnostic criteria. In this context, Stow and Lovell (1979) introduced new insights into contourite research: (1) the acknowledgment of widespread burrowing and bioturbation as a common feature of contourites and, hence, the low likelihood of preservation of primary sedimentary structures; (2) a clear distinction of contourites from fine-grained turbidites by the lack of a vertical sequence of structures in the former; and (3) a proposal that palaeocurrents are a good criterion to separate down-slope (i.e., turbidite) from along-slope (i.e., contourite) systems. In the 1980s, works in the same line of reasoning abounded, and with these, the controversy on the generation and preservation of traction structures in contourites rapidly grew (Lovell and Stow, 1981; Stow, 1982; Gonthier et al., 1984; Stow and Piper, 1984). These authors criticized the pioneer works of Hollister, Heezen, and Bouma, and pointed out that traction structures cannot be used as standard criteria for contourite recognition, neither in sediment cores, nor in outcrops. This conception was supported by the idea that the intense biological activity that develops on the deep-sea floor – favored by the introduction of oxygen by bottom currents (Chough and Hesse, 1985) – could induce sufficiently strong bioturbation to destroy any previously generated traction structure. Hence, traction structures were replaced by thorough bioturbation as diagnostic criteria of contourites in deep marine sediments. This concept allowed the proposition of distinctively different depositional models for contourites – without significant current structures – and fine-grained turbidites – rich in small-scale traction structures – (e.g., Stow and Shanmugam, 1980; Stow and Piper, 1984; Piper and Stow, 1991). On that basis,

J. Martı´n-Chivelet et al.

161

beds and successions from various localities and of various ages that had originally been interpreted as contour-current deposits were reinterpreted as fine-grained turbidites (e.g., Stow et al., 1998a). Those facies models and the related criteria have been defended by some authors up to the present day (Stow and Fauge`res, 1993, 1998; Stow et al., 1996b, 2002c; Stow and Mayall, 2000; Stow, 2005). In contrast, other workers kept adhering to the validity of the original ideas and criticized any contourite facies model that excludes traction-current structures (e.g., Shanmugam et al., 1993a, b, 1995; Shanmugam, 2000). They emphasized the importance of traction processes on the ocean floor and re-stated that primary sedimentary structures should be the basic criteria for the recognition of contourites and the presence of bottom currents, instead of bioturbation and palaeocurrent directions, which vary strongly with the oceanographic setting.

10.3.

SETTING THE STAGE FOR SEDIMENTARY STRUCTURES IN C ONTOURITES

Because of the existing controversy, we emphasize the following generally accepted views and facts regarding the deposition of contourites before dealing with the description and discussion of sedimentary structures in detail. 1. Contourite deposition arises in immediate or close association with the transport of sedimentary material by water flowing over the ocean floor. During sediment accumulation, traction structures induced by currents are often generated. However, in the time interval between deposition and significant lithification, burrowing can be sufficiently intense to destroy previous traction structures partially or totally. 2. Traction structures are abundant on recent ocean floors that are influenced by bottom currents. Deep marine environments have been intensively explored in the second half of the 20th century. The earliest explorations came in the form of photographs and cores from hundreds of deep oceanic sites all over the world (Heezen and Hollister, 1964; Hollister, 1993). By the end of the 1950s, the photographs of ripple marks and scour marks at abyssal depths (Me´nard, 1952; Heezen et al., 1959) eliminated any remaining doubt on the occurrence and importance of bottom currents in the oceans. 3. Contourites are generated in diverse settings, defined by a variety of sedimentological, hydrochemical, hydrodynamical and ecological conditions. The different environmental conditions under which contourites can form result in a wide variety of depositional facies and successions. Despite the noticeable compositional variability (siliciclastic, volcanoclastic, biogenic, etc.; see, among others, Stow et al., 1996b), two main types of contourite deposits are usually accepted, namely: ‘‘muddy contourites’’ (<15% of sand) and ‘‘sandy contourites’’ (>15% of sand) (Stow and Lovell, 1979). If it is assumed that the sand content of such deposits is strongly dependent on the current power and its winnowing effect, traction structures will be more abundant and easily preserved in sandy contourites.

162

Traction Structures in Contourites

4. ‘‘Actualism’’ cannot be strictly applied to ancient contourites. Changes that have taken place in the ocean throughout the geological history could have produced different styles of contourite deposition. The present-day circulation pattern, which is mainly controlled by the formation and sinking of dense cold waters in the subpolar regions, seems to have prevailed (although with noticeable variations) since the middle Cenozoic (e.g., Lawver et al., 1992). However, thermohaline circulation in older, greenhouse times, which was probably driven by active sinking of saline waters in intertropical seas, is still poorly understood (e.g., Hay, 1996). 5. Structures in contourites: a matter of scale. Acquisition of data on sedimentary structures and bedforms has always been limited by the available and affordable technology. Because most case studies come from cores collected in boreholes or from small outcrops in ancient rocks, some interpretations might be biased by the impossibility of direct observation of bedform geometry, internal architecture and lateral arrangement. 6. Deep bottom currents and turbidity currents are essentially different. Although from a rheological perspective both are Newtonian, turbulent flows (e.g., Shanmugam, 2000), the main difference is that, whereas turbidity currents are sediment gravity flows (Middleton, 1993), contour currents are water flows, controlled essentially by thermohaline properties and geostrophic motion. Turbidity currents are rapid, single events, whereas bottom currents are much slower, usually acting over long time periods.

10.4.

S EDIMENTARY STRUCTURES

The available documentation of sedimentary structures in both ancient and modern contourites is still relatively scarce and disperse. In this section, this information is summarized, reviewed, and integrated with that from experimental models (Figure 10.1). Given the variety of deep-sea settings in which contourites can be found, not all the following structures will be present in every sedimentary system. Also, as each sedimentary structure is the result of a process and not of an environmental setting, the following structures are typical but not exclusive of contourite deposits.

10.4.1.

Current ripples and cross-lamination

Present-day examples of current-rippled surfaces generated by bottom currents in deep marine environments are abundant (Figures 10.2 and 10.3). Bottom-current velocities from 0.1 to 0.4 m s 1 have been measured at various locations in abyssal environments (Lonsdale and Spiess, 1977). Ripples can have straight, sinuous or linguoid crests. Recent work by Wynn et al. (2002a), who analysed migrating ripples at the surface of large submarine dunes in the Faroe–Shetland Channel, demonstrated a clear and direct relationship between the current energy and the complexity of the crest lines (Figure 10.2). This result is in agreement with the classic experimental works performed in flumes in the 1960s (Allen, 1984). More

163

J. Martı´n-Chivelet et al.

Sedimentary structures

Scale 1 cm

Grain size (dominant)

Environmental implications

Horizontal or sinusoidal lamination; stripped, fine-grained deposits; “wispy” lamination

Fine sand, silt and mud

Low current strength, predominance of deposition from suspension

Lenticular bedding starved ripples

Fine sand, silt, mud

Alternating flow conditions; low to moderate current strength, winnowing.

Wavy bedding, flaggy chalks

Fine sand, silt, mud

Alternating flow conditions, low to moderate current strength

Flaser bedding, mud offshoots

Fine sand to silt

Alternating flow conditions, current speed: 0.1–0.4 m s–1

Climbing ripples (subcritical to supercritical)

Very fine to medium sands

Current speed: 0.1–0.4 m s–1 high suspension load

Large-scale cross-bedding, megarriples,dunes, sandwaves

Medium sands

Current speed: 0.4–2 m s–1; barchan dunes usually form at 0.4–0.8 m s–1

Parallel lamination (upper stage plane beds), presence of primary current lineation

Very fine to medium sands

Current speed: 0.6–2 m s–1

Minor erosive surfaces, mud rip-up clasts, upper sharp contacts

Sand, silt, mud

Alternating flow conditions, low to moderate current strength

Sole marks: flutes, obstacle scours, and longitudinal scours; cut and fill structures

Sand, silt, mud

Flow speed peaks

Longitudinal ripples

Coarse sandy muds (20% sand)

Low current strength (2–5 cm s–1), winnowing

Bioturbation (strongly variable)

Sand, silt, mud

Low current speed, strong paleoecological control, low to moderate accumulation rates

Normal and reverse gradding at different scales and within different types of deposits

From coarse sand to mud; usually fine sand, silt and mud.

Gradual changes in flow strength

Pebble lags, furrows

Coarse sand, microconglomerate

Current speed over 2 m s–1

1 cm

1 cm

1–5 cm

1–5 cm

10–50 cm

1 cm

1 cm

1–5 cm

5 cm

1–10 cm

3–20 cm

0.1–2 cm Abundance in the fossil record:

very common

common

rare

not yet described

Figure 10.1 Main types of primary sedimentary structures that may be found in contourite deposits. The relative abundance in the fossil record of contourites is shown in the right-hand column.

164

Traction Structures in Contourites

Current direction

1 Gently rippled seafloor

(a)

(straight-crested ripples)

Sinuous/ linguoid ripples

(b) 50 cm

1

Straight-crested ripples on flat floor Approximate edge of barchan dune

Linguoid ripples

Not to scale

2

6 Smooth seafloor Sinuous/ linguoid ripples

Sinuous ripples

Linguoid

Linguoid ripples on barchan surface

50 cm

2

Accumulation of pale sediments

5 General flow direction

Linguoid ripples

3 ripples

3

Linguoid ripples near dune crest

7

Gently ripples seafloor with gravel patches

50 cm

5

4

Sinuous ripples on lower barchan slopes Ripple stoss face (dark sediment) Ripple lee face (pale sediment) Organic detritus in ripple troughs

Linguoid ripples on barchan surface

50 cm

7

Pale sediment along inner edge of dune

Barchan slip face Pale sediment at base of slip face Zone of smooth seafloor with faint lineations Zone of small fan-shaped ripples

50 cm

Smooth seafloor with faint lineations

Closely spaced ripples arranged in a fan shape

Short-crested linguoid ripples

50 cm

without ripples

Straight-crested ripples on flat sea floor Gravel streaks

Transverse ripples showing bifurcation and change in orientation near barchan edge

50 cm

6

Smooth seafloor with gravel streaks Transverse ripples

50 cm

8

Faint ripples (on flat sea floor?) Sea pen bent by active bottom current

Approximate edge of barchan dune Sinuous and linguoid ripples on barchan surface

General flow direction

Figure 10.2 Patterns of distribution of sedimentary structures in recent active contourite systems (from Wynn et al., 2002a; with permission from Elsevier). (a) Pattern of ripple types across a barchan dune of the barchan field of the Faroe ^Shetland Channel, and interpreted current flow over dune surface. Interpretations based on WASP sea-floor video and photography. (b) (1^8): Details of sea-floor structures featured in (a).

recent works based on flume experiments suggest, however, that straight and sinuous ripples are metastable, always evolving into linguoid ripples if the experiment is given sufficient time (Baas et al., 1993; Baas, 1994). Ripples are also abundant in ancient deposits interpreted as contourites (Figures 10.4–10.7). Measured ripple wavelengths (l) are in the range of 5–30 cm, and their height (h) usually ranges between 1 and 4 cm. The ripple

165

1060

1040

Track of photographic traverse

1020

J. Martı´n-Chivelet et al.

500 m

Channel floor Escarpment Erosional groove gravel field Ripples

2 3

1060

1

50 m 4 2

3 2 1

1

3 2 Large, short-crested ripples (2)

1 ~2 m Symmetrical and asymmetrical parallel, long-crested ripples (1)

4 Erosional grooves (3) and gravel fields (4)

Figure 10.3 Pattern of distribution and sequences of different sandy bedforms across the bottom of a ‘‘free-standing’’ bottom-current channel in the Gulf of Cadiz (after Habgood et al., 2003; with permission from Blackwell Publishing).

index (or vertical-form index), defined by l/h, usually varies from 10 to 40, a value which is consistent with those obtained in flume studies. These experiments do not give a conclusive relationship between ripple size and current velocity, although some reveal that, under flow conditions near the upper limit of ripple stability, both

166

Traction Structures in Contourites

(a)

90° Section A Section B

5 cm

5 cm

(b) Hemipelagite

Erosional top

5 cm

Figure 10.4 Examples of horizontal lamination with subordinated small-scale cross lamination in mid-depth calcareous contourites (from Martı´ n-Chivelet et al., 2003; with permission from Elsevier). Quipar-Jorquera Fm., Late Cretaceous (Maastrichtian) from the Subbetic of Caravaca (Spain). (a) Two polished sections of a contourite bed with sigmoidal cross-lamination, draped lamination and a reactivation erosional surface. The change in the dip of foresets located under and above the reactivation surface reflects a change in the direction of ripple migration. (b) Polished section showing sinusoidal to horizontal lamination and intercalated cross-lamination. Note the sharp, erosional contact at the top of the contourite bed, overlain by a white muddy hemipelagite.

167

J. Martı´n-Chivelet et al.

(a)

(b)

5 cm

Figure 10.5 Outcrop examples of ripple lamination and related small-scale structures in ancient contourite deposits. (a) Cross-laminated calcareous contourite from the Ordovician of the Lachlan Fold Belt, eastern Australia ( Jones et al., 1993). Climbing ripples, flaser bedding, and erosional reactivation surfaces are present (courtesy B.G. Jones). (b) Subhorizontal to sinusoidal lamination and ripple lamination in bottom-current-reworked fine sands. Note the internal erosional surfaces. Jurassic, Neuquen Basin, Central Andes, Argentina (Palma, Martı´ n-Chivelet and Lo¤pez-Go¤mez, unpublished data). A multicolor version of this figure is on the enclosed CD-ROM.

the ripple height and its wavelength increase rapidly with a small increase in shear stress. However, a clearer relationship exists between l and grain size. According to Yalin (1964, 1972), l = 1000d, where d is the mean diameter (in mm) of the bed material (e.g., sands of 0.1 mm diameter will generate l of about 100 mm = 10 cm). Flume experiments show that ripples do not usually form in sands coarser than about 0.7 mm (e.g., Harms et al., 1975; Southard and Bouguchwal, 1990). In contourites most ripples are preserved in very fine to fine sands that can be easily transported both in suspension and as bed load. Vertical sections through rippled contouritic deposits commonly reveal sets of cross-lamination defining climbing-ripple cross-lamination (Figures 10.4, 10.5a and 10.6). The relationship between the stoss-side slope, x, and the angle of climb, z, defines two modes of climbing ripples (Allen, 1984). When x > z, sets are subcritical, that is, separated by sharp, erosional boundaries inclined upward relative to the bases of the parent bed features. In contrast, x < z determines supercritical sets, characterized by gradational boundaries and preservation of stoss-side laminae. Generation of climbing-ripple successions in flumes has shown how the final type of climbing ripples strongly depends on the relationship between suspended and bed load (Ashley et al., 1982). Supercritical sets form when large amounts of suspended material are supplied, whereas the subcritical ones form when the bedload is dominant. Supercritical sets seem to be a common feature in contourite deposits (e.g., Shanmugam et al., 1995; Martı´n-Chivelet et al., 2003). The time required for a climbing-ripple succession to form can be very variable. Pioneer work by Kuenen (1967) estimated that single climbing-ripple successions are usually deposited in a few to some tens of hours. In contrast, Ashley et al. (1982) showed that ripples in fine sand can migrate extremely slowly, at equilibrium with the current, at rates as low as centimeters per day. Under such conditions, deposition of a single set of climbing ripples, even with steep angles of climb, would

168

Traction Structures in Contourites

(a)

2 cm

(c) 2 cm

2cm

(b)

Figure 10.6 Core of sandy material from the upper Pliocene and Pleistocene of the Gulf of Mexico showing different traction structures (from Shanmugam et al., 1993a; with permission from the American Association of Petroleum Geologists). (a) Erosional truncation (arrow) separating a cross-laminated sand unit and overlying mud unit. (b) Lamination. (c) Several discrete units of cross-laminated sands with current ripples showing variable dip directions that could indicate multiple current directions. Preserved (lower arrow) and eroded (upper arrow) tops of ripples indicate variable energy conditions of the current.

require 10 days or even more. For ripples to remain active during deposition at such slow rates, the flow strength needs to remain very stable, within a very narrow velocity range. Such conditions are rare in most sedimentary environments, including that of turbidity currents, but can exist on a sea floor that is affected by thermohaline currents. Although detailed palaeocurrent analyses based on fossil contourite ripples are scarce, some studies show that the dip of foresets in these forms may be unidirectional ( Jones et al., 1993; Stanley, 1993), bi-directional (Zhenzhong and Eriksson,

169

J. Martı´n-Chivelet et al.

(a)

(b)

(c)

Figure 10.7 Thin sections of calcareous, fine-grained contourites from the uppermost Cretaceous of SE Spain, showing well-defined cross-lamination (from Martı´ n-Chivelet et al., 2003; with permission from Elsevier). Lamination is delineated by the alignment of biogenetic very fine sand and silt. Photo (a) shows a small vertical burrow that cuts lamination (scale bar 1 mm).

1991; Fauge`res et al., 1993), or even multidirectional (Shanmugam et al., 1993a). This variability has been attributed to the different environmental settings in which contourites can form. Permanent and stable bottom currents generate dominantly unidirectional cross-lamination. Bi-directional cross-lamination has been interpreted as the result of tide-related bottom currents, such as those generated in submarine canyons (e.g., Shepard et al., 1979), or as a consequence of the interference of bottom currents and overbank turbidity currents. Multidirectional cross-laminations have been attributed to the natural seasonal variation of bottom currents or, as is more likely, to benthic storms (McCave, 2008). As stated above, the diagnostic character of ripples for the recognition of contourite deposits in deep-sea sandy deposits is still a matter of debate. Obviously, the mere occurrence of ripples does not constitute undisputable evidence for deposition by a contour current. Ripples can also be generated by traction and

170

Traction Structures in Contourites

deposition of sand from the ‘‘diluted tail’’ of turbidity currents (e.g., Kuenen, 1964; Walker, 1965). The mechanism of ripple formation is somehow different in turbidites and contourites, and the resultant deposits should reflect the differences. Ripples related to turbidites are generated over a very short time span, as they are the product of rapid flow deceleration. This results in the occasional formation of rippled intervals, which can range from quite homogeneous climbing-ripple sets to thin horizons of fading ripples. The action of bottom currents on the sea floor is usually more prolonged, and the formation of rippled sand beds usually occurs over longer periods. This gives way to more continuous and complex rippled intervals, often showing features which indicate multiple episodes of sedimentation such as reactivation surfaces, erosional features in the ripples and lenticular bedding, which can be hardly explained in the framework of a normal turbidity current. Another diagnostic character is the recognition of bi-directional or multidirectional patterns of ripple orientations (i.e., palaeocurrents), which are common in contourites but very rare in turbidites. The orientation and the regional slope of the sea floor, as defined by the direction of the current deduced from ripples have also been proposed as a tool to distinguish contourite from turbidite ripples. This criterion has, however, two main objections: (1) not all ripples generated by dilute turbidity currents migrate down-slope; and (2) many bottom currents do not follow trajectories that parallel the continental slopes. To discern between the two types of ripples is never a simple task. The study of the associated structures and facies should help. However, the topic may become even more complex as some turbidite sequences show reworking by bottom currents, and this results in mixed facies associations (Stanley, 1988a, 1993; see also Mulder et al., 2008).

10.4.2.

Lenticular and flaser bedding

Occasional changes in the flow strength on the sea floor induce deposition of heterolithic facies. These can be represented by centimeter-scale alternations of sand, silt and mud to produce sedimentary structures such as lenticular and flaser bedding (Figures 10.5a and 10.6). Lenticular bedding is characterized by the dominance of muddy sediment enclosing isolated, sandy ripples which are usually less than 1 cm thick. These ‘‘starved’’ ripples are formed by relatively mud-free ‘‘clear-water’’ bottom currents. Wavy bedding is characterized by a higher presence of sandy or silty material. Thus, the small lenses of sand or silt, which often outline fading ripples, are interconnected and overlap one another within the muddy material. In flaser bedding, the muddy sediment is subordinate, and occurs as thin (less than 1–2 mm) and discontinuous laminae which partially drape ripple forms. These drapes, named ‘‘mud offshoots’’ by Shanmugam et al. (1993a), are generated by the settling of fine-grained sediment from suspension during short periods of weak currents, when the ripples are inactive. If the current strength on the sea floor diminishes gradually, the deposition of fine sediment can increase, and thicker mud drapes can form, and eventually entirely bury the ripples. This process gives way to

171

J. Martı´n-Chivelet et al.

sinusoidal bedding that shows a progressively upward flatter shape (Martı´n-Chivelet et al., 2003). These heterolithic structures were considered by Shanmugam (2000) as diagnostic features for reworking by a bottom current. Shanmugam argues that they are the result of alternating high- and low-energy conditions, which are unlikely in turbidity currents, where the fluid turbulence would prevent formation of features such as ‘‘mud offshoots.’’ Other authors, however, argued that these heterolithic facies can also be generated by turbidity currents, where fading, starved ripples, and even mud offshoots can be the result of a relatively rapid fallout of sediment from fine-grained turbidity currents (Stow et al., 1998a).

10.4.3.

Horizontal, subhorizontal, and sinusoidal lamination

Almost all studied examples of fossil contourites show some kind of fine, usually heterolithic lamination, which can be horizontal, subhorizontal or sinusoidal (Figures 10.4 and 10.8a), consistent with deposition under the influence of weak to moderate, fluctuating bottom currents. Main types of lamination can be defined as (a) horizontal lamination in medium sand to silt, produced by lower plane-bed traction; (b) very fine sand – mud lamination induced by deposition from suspension and shear; (c) wispy lamination (i.e., flat, very thin, discontinuous lenses of fine sand to silt within a silty or muddy bed), produced by alternating suspension and tenuous traction; and (d) wavy lamination (defined by interlayered thin, sinusoidal beds of fine sand, silt and/or mud) that result from deposition from traction under changing flow strength. The four types are common in contourites (e.g., Pequegnat, 1972; Piper and Brisco, 1975; Bein and Weiler, 1976; Lovell and Stow, 1981; Stanley, 1988a, 1993; Duan et al., 1993; Fauge`res et al., 1993; Shanmugam et al., 1993a, b; Dalrymple and Narbonne, 1996; Kahler and Stow, 1998; Ito, 2002; Luo et al., 2002; Martı´n-Chivelet et al., 2003). (a)

(b)

5 cm

Figure 10.8 Medium- to large-scale traction structures in calcareous contourite deposits (Maastrichtian, Subbetic Zone near Caravaca, Spain). (a) Planar horizontal lamination in fine calcarenites from a contourite bed. Note the intercalated beds with current ripples. (b) Crossbedding in calcareous contourites. Note the sharp bottom contact of the laminated contourite bed, which overlies hemipelagic carbonate. A multicolor version of this figure is on the enclosed CD-ROM.

172

Traction Structures in Contourites

Associated with these laminations are usually other structures such as microcross-lamination, starved ripples, and contorted lamination. Syn-sedimentary deformation, although very typical of turbidites, can be sometimes present in contourites.

10.4.4.

Large-scale cross-bedding

Large volumes of sand and silt accumulated by bottom currents can produce bedforms such as dunes and sand waves, which have been recorded from many recent mid-depth and deep contourite deposits (Wynn and Masson, 2008). They may show a wide range of sizes and wavelengths, and often form vast fields that resemble dune fields of subaerial deserts remarkably well (e.g., Kenyon and Stride, 1967; Lonsdale and Malfait, 1974; Lonsdale and Spiess, 1977; Kenyon, 1986; Wynn et al., 2002a). The height of the described examples of these bedforms varies from a few decimeters to more than 60 m, and the wavelenghts are usually within the range of tens of meters, although in some cases they can be much larger, reaching more than 1500 m. The external shape of dunes and sand waves can also vary notably, mainly influenced by the intensity of the currents and the availability of transportable sand. They occur as transverse forms, with straight or sinuous crests, as barchan sand waves, or even as longitudinal sand ribbons (see Wynn and Masson, 2008). Published information on the sedimentary material and structures that constitute these bedforms is scarce. Lonsdale and Malfait (1974) described moderately sorted sand, composed dominantly of bioclasts (fragmented tests of planktonic foraminifers) in the dunes on the floor of the Carnegie Ridge in the Central Pacific. Internal sedimentary structures of deep-sea barchan dunes are probably comparable to those described from their subaerial equivalents and flume experiments. In these, dune migration gives rise to large-scale planar or trough cross-stratification, where individual cross-laminae usually show a tangential contact with the base of the set. This morphology results from the relatively weak lee separation eddies and from a high fallout rate of particles from suspension at the leeside of the dune (Leeder, 1999). In the sedimentary record, several cases of contourites with large-scale crossbedding have been documented: Villars (1991) described nice examples produced by migration of calcarenite megaripples and larger forms from the Upper Cretaceous of the Alps; similar facies have been described from the Campanian–Maastrichtian of the Subbetic (Figure 10.8b) (Martı´n-Chivelet et al., 2003). Stanley (1993) found cross-bedding with a height reaching 30 cm in the Cretaceous sandy contourites of the Virgin Islands. Duan et al. (1993) described 80–180-cm-thick beds with a calcilutitic contourite facies with s-shape progradational bedding and truncation surfaces from the Lower Ordovician of southern China (Figure 10.9), and interpreted them as formed by down-current migration of mud waves (10–20 m wavelength, 1–2 m amplitude); they also described erosional furrows of 2–10 m deep with a width of 0.25–3 m. A good example of possible large, mid-depth marine dunes exposed on land is the Weka Pass Limestone in New Zealand (Figure 10.10; Carter et al., 1996; Carter, 2007). This is composed of pelagic, fine-grained calcarenites deposited at

173

J. Martı´n-Chivelet et al.

Figure 10.9 Large-scale cross-stratified calcareous contourites from the Lower Ordovician of Jiuxi (Hunan Province, China). These fine-grained deposits are believed to result from the development of mud waves and erosional furrows on the sea floor under the influence of a semipermanent bottom-current regime (Duan et al., 1993; Photo courtesy T. Duan). A multicolor version of this figure is on the enclosed CD-ROM.

(b)

(a)

(c)

Figure 10.10 Cross-bedded, glauconitic, calcarenite sediment drifts of Late Oligocene age; Weka Pass Limestone, Waihao Forks, South Island, New Zealand (Carter et al., 2004). The drifts represent resumed sedimentation above a regional unconformity, present in strata deposited between shallow shelf and abyssal water depths, and are inferred to have been cut by vigorous bottom currents associated with the onset of the predecessor Antarctic Circumpolar Current. (Photos courtesy R.M. Carter). A multicolor version of this figure is on the enclosed CD-ROM.

174

Traction Structures in Contourites

bathyal depths under the influence of the Antarctic Circumpolar Current (ACC)–DWBC (Deep Western Boundary Current) current system, and shows spectacular large-scale cross-bedding, with sets up to 5 m thick and many tens of meters long (Carter et al., 1996). These dunes resulted from a dominant, semipermanent eastward-flowing current, interrupted by intermittent, northward-flowing storm surges that created contemporaneous scour channels (Ward and Lewis, 1975).

10.4.5.

Erosional structures

Erosional surfaces in contourite successions reflect pulses of increased bottomcurrent flow strength. These pulses can be induced by changes in the energy of a stationary current, and/or by shifts in the path of a current. Erosional surfaces on the present-day deep sea-floor form large furrows and grooves which develop under bottom currents that can exceed velocities of 100 cm s 1 (Flood, 1983). In the sedimentary record, the following types of medium- and small-scale erosional structures are usually found in contourites (Figures 10.5b, 10.6, 10.9–10.11). (b)

(a)

10 cm (c)

(d)

2 cm

2 cm

Figure 10.11 Small-scale erosional structures and winnowing deposits. (a) Scour marks (including flute casts), reflecting erosional processes on a cohesive substratum consisting of hemipelagic carbonate mud. Maastrichtian, Caravaca, SE Spain. (b) and (c) Small-scale erosional surfaces. In both cases, erosion precludes de sedimentation of sandy contourite facies. In (b), the erosive surface cuts down into a laminated contourite, whereas it affects pelagic carbonate in (c) Maastrichtian, Caravaca, SE Spain. (d) Small remobilization level with several shell fragments of large hemipelagic bivalves (inoceramids). Upper Cretaceous, Xixona (Alicante, SE Spain). A multicolor version of this figure is on the enclosed CD-ROM.

J. Martı´n-Chivelet et al.

175

1. Microerosional surfaces. Bottom currents are often too weak to generate thick contourite beds. However, several microstructures can be observed at millimeter scale; these indicate minor erosion, winnowing, and reworking. Good examples are the small cut-and-fill structures, often accompanied by winnowed foraminifer debris lags and small muddy rip-up clasts described by Hayward (1984) from hemipelagic chalks of Miocene age in Turkey, and the centimeter-scale beds with accumulation of mudstone microclasts derived from erosion of muddy facies by bottom currents described by Oaie (1998) from the Upper Proterozoic of Romania. 2. Reactivation surfaces in cross-laminated units. These are discontinuities cutting across foresets of ripples or dunes. They indicate erosion by a change in flow strength before the resumption of the forward migration of the bedform (Figure 10.8). This implies successive periods of sedimentation and erosion on the deep-sea floor, which are typical of bottom currents. Although they have been also described from turbidite series, there is no reliable explanation for them apart from intervening of bottom currents. 3. Sharp contacts of sandy layers. Centimeter- to decimeter-scale sandy beds are usually bounded by sharp contacts in contourite successions (Figures 10.6 and 10.10). In particular, sharp upper contacts have been considered as diagnostic of contourite settings, i.e., as the result of moderate erosion by bottom currents (Shanmugam, 2000). 4. Sole marks. These are usually preserved as casts at the base of sandy beds (Figure 10.10a). They are cut by turbulent flows capable of eroding the soft, cohesive fine-grained sediment which goes directly into suspension instead of moving as bedload-generating bedforms. Sole marks are much less abundant in contourites than in turbidites, probably because the changes in flow strength of bottom currents are much more gradual and not as strong as in turbidity currents. In addition, their potential of preservation is usually low. Sole marks in contourites include small flutes, obstacle scours, and longitudinal scours. Intrabed erosional structures are probably the most reliable feature indicating bottom traction currents. Fine-grained turbidites are capable of some auto-erosion, but the presence of numerous erosional features, reactivation surfaces, pebble or bioclast lags in sediments reflect repeated variations in bottom-current strength.

10.4.6.

Grading

Deposits generated under the influence of bottom currents usually show a grading of grain size that represents the sedimentary response to temporal changes in flow strength. Both normal and inverse grading can develop at very different scales in muddy and sandy contourites. Examples exist in both the recent and the ancient record. According to Stow and Holbrook (1984), the alternation of normal and inverse grading is characteristic of contourite sequences, whereas Shanmugam et al. (1993a) consider inverse grading in deep marine deposits as a reliable indicator of bottom-current reworking, which helps to differentiate such deposits from those of gravitational flows. Inverse grading rarely occurs in deposits produced by turbidity

176

Traction Structures in Contourites

currents, although it can be generated from sandy debris flows (Lowe, 1982; Kneller, 1995; Shanmugam, 1997a).

10.4.7.

Less common structures

10.4.7.1. Upper-phase plane beds (parallel lamination) Upper-phase plane beds can be generated in contourite settings if the flow strength increases enough to erase the preexisting bedforms. Intense sediment transport then takes place over a nearly featureless surface (the upper-stage plane bed), generating very small-amplitude/long-wavelength waves. According to Best and Bridge (1992), these bedforms range from 0.75 to 11 mm in thickness, from 0.7 to 1.0 m in wavelength, and propagate with velocities of up to 10 mm s 1. When net accumulation occurs, the migration of these thin sediment waves over a smoothed surface generates parallel lamination. Several flow-parallel ridges and hollows are superimposed upon these flat forms to produce primary current lineation (parting lineation). This type of high-regime parallel lamination is abundant in turbidites (division B of the Bouma sequence). Shanmugam et al. (1993a) suggested, however, that the generalized assumption of a turbiditic origin for deep marine parallel-laminated sands might have arisen because of a mistaken interpretation of laminated contourites as turbidites, and stressed the need for additional supporting evidence of turbidite deposition, e.g., by the presence of a normally graded deposit beneath the laminated interval. 10.4.7.2. Wave ripples Wave ripples, which are a common structure of shallow-water sands deposited under the influence of waves, can also form in deeper water, probably as a result of internal-waves motion within the bottom waters. There are only a few descriptions of wave ripples in contourites (Karl et al., 1986; Shanmugam et al., 1993a,b). Because some cross sections of current ripples belonging to a succession with multiple and minor erosive surfaces often resemble wave-induced structures – sometimes even resembling hummocky cross-stratification – it is important to examine ‘‘wave-like’’ ripples carefully in three dimensions. 10.4.7.3. Longitudinal triangular ripples Longitudinal triangular ripples have been reported from the deep ocean since the 1960s (cf. Heezen and Hollister, 1964). They are intermediate-scale bedforms elongated parallel to the current direction. They are symmetrical, with a characteristic triangular shape in transversal section. Their crests are straight or slightly sinuous (Flood, 1981; Tucholke, 1982; McCave et al., 1984). Individual longitudinal triangular ripples are up to 20 cm high and 1–10 m long. Their width reaches up to 1 m and this varies approximately in proportion to their height. They are composed of fine-grained cohesive sediment, essentially sandy mud (20% sand), with a poorly defined internal structure made of variably burrowed, thin horizontal layers (up to several millimeters thick) of calcareous bioclastic sand, consisting of

J. Martı´n-Chivelet et al.

177

foraminifer fragments (McCave et al., 1984). The bases of the longitudinal ripples have a sharp contact with the underlying bioturbated ooze. Two models have been proposed for the origin of these ripples: (1) the bedforms are deposited rapidly following a brief, high-velocity current event which stimulates secondary, helical flow in the bottom boundary layer (Flood, 1981; McCave et al., 1984); (2) these ripples are initially ‘‘tail-deposits’’ deposited in the lee of sea-floor obstacles (e.g., biological mounds), which are subsequently extended downstream by oblique currents (Tucholke, 1982, 1986). Their preservation potential in the sedimentary record is low, and they cannot be detected in cores. They have not been described yet from the stratigraphic record.

10.5.

SEDIMENTARY S TRUCTURES IN F ACIES MODELS

The wide spectrum of sedimentary structures in contourites reflects the numerous factors that control their formation, which include current intensity, steadiness of flow, supply of sediment, hydrochemistry, and texture and composition of available sediment. Burrowing also plays an important role in contourite generation, by means of (1) delaying consolidation and decreasing sediment shear strength, i.e., making them more easily winnowed by bottom currents, and (2) homogenizing and disturbing the sediments, and hence destroying the primary structures. With this variety of facies and controlling factors, the scarce number of facies models proposed for contourites seems strange. In this section, we briefly describe these models with emphasis on the role of the primary sedimentary structures.

10.5.1.

The Faro Drift facies model

This model is based on the contourites of the Faro Drift in the Gulf of Cadiz, south of Portugal. The Faro Drift is composed essentially of mud (95%) and is generated in a setting of relatively shallow water (600 m depth) under the influence of the Mediterranean Outflow Water (MOW), a relatively warm, nutrient-rich and moderately oxygenated water mass that flows through the area with velocity of 0.1–0.3 m s 1 (Fauge`res et al., 1984; Gonthier et al., 1984). This study of the Faro Drift provided the basis for an ideal composite facies sequence (Gonthier et al., 1984; Stow and Holbrook, 1984; Stow and Piper, 1984), which essentially consists of mud, silt and very fine sand, and is characterized by the almost complete absence of primary sedimentary structures (Figure 10.12). The just-mentioned authors acknowledged their potential formation, but considered that these structures had been systematically obliterated by bioturbation. They also mentioned the absence of a regular sequence of facies, except successive units of normal and reversed grading, which can form couplets representing a gradual increase, a maximum, and then gradual decrease in the average current velocity at a given site.

Sedimentary structures in contourite facies models

Bioturbated Mud Silt

Mottled silt and mud

cm

Sandy silt

Sandy contourites

Structure: homogeneous; bedding poor or absent; rare primary silt laminae buoturbated and burrowed Texture: dominantly silt-mud; 0–15% sand-sized; poorly sorted Composition: combination of biogenic and terrigenous (i.e. hemipelagic) part may be far-travelled organic carbon (av. 0.3–1.0%) carbonate (commonly high) absence of shallow water biogenics

Shanmugam et al., 1993a, 1995 Gulf of Mexico Pliocene – Pleistocene Shanmugam, 2000 Horizontal bedding

Climbing ripples

Mud offshoots

Faro Drift Recent

Lithology grain size and structure 4 8 16 32 64 µm

cm

Muddy contourites

Structure: thin irregular layers (lag concentrations); more rarely, horizontal and cross lamination preserved; commonly bioturbated Texture: silt-sand size, rarely gravel lag; poorly sorted, muddy or well-sorted, clean; slight negative skewness Composition: concentration of coarse fraction at sediment surface; commonly local origin; biogenic + terrigenous sands mixed; iron-stained + broken biogenic debris

Stow and Piper, 1984a Stow et al., 1986, 1999

Mottled silt and mud

Laminated Cross-laminated, bioturbated Silt mottles and lenses, bioturbated

Massive, irregular sandy pockets, bioturbated, contacts sharp to gradational Silt mottles, lenses and irregular layers; bioturbated 10

Mud

cm

Bioturbated

0

Martín-Chivelet et al., 2003

Caravaca Late Cretaceous

Facies sequence Hemipelagite (marly limestones)

5 cm

5 cm

Sharp top 5 cm

Normal grading

North Atlantic Drifts Tertiary to Recent

Inverse grading

Stow and Holbrook, 1984

Some burrowing

0.25 m 0

Parallel lamination Flaser bedding

Cross-bedding

Lenticular bedding

some small ripples parting lineation

Sigmoidal cross-lamination

5 cm

5 cm

5 cm

Sinusoidal lamination

Rhythmic bedding

Mud offshoots Erosional surfaces

Sharp upper contact

Low angle cross-lamination

10 cm

5 cm

Inverse grading

Fine sand

Gradational Mud lower contact

Erosional base scour casts Hemipelagite (marly limestones)

Figure 10.12 Sedimentary structures in facies models and sequence types. See text for explanation. Figures reproduced with permission from: The Geological Society, London (Stow and Holbrook, 1984; Stow and Piper, 1984), the American Association of Petroleum Geologists (Shanmugam, 1993a) and Elsevier (Martı´ n-Chivelet et al., 2003).

J. Martı´n-Chivelet et al.

179

Fauge`res and Stow (1993) took one further step toward the facies model and ventured an estimation of the time span needed for the formation of a complete contourite sequence of the order of tens of thousands of years. Their interpretation of the ideal composite sequence in terms of processes developed during a definite time span gave it a genetic sense that automatically transformed it into a ‘‘true’’ type sequence. In fact, from this moment on, it became the general facies model for contourites, and was presented as such in handbooks of sedimentology and successive reviews (e.g., Stow, 1994, 2005; Stow et al., 1996b, 1998a, 2002d; Stow and Fauge`res, 1998; Stow and Mayall, 2000). The usage of the Faro Drift sequence as a general facies model has, however, received strong criticism (Shanmugam et al., 1993a,b, 1995; Shanmugam, 2000). In these works, the facies model is not considered as representative of most contourite deposits, essentially because (1) the scarce evidence of current influence shown in the model, (2) the absence of true sands and associated sedimentary structures derived from the typical hydrodynamical behavior of this grain size, and (3) the use of pervasive bioturbation as a diagnostic feature of contourites, while bioturbation is widespread in many deep water environments. Recent work on both modern and ancient cases demonstrated that the strong, pervasive bioturbation reported from the Faro Drift is not a ubiquitous feature of contourite deposits, and that the generation and/or preservation of bioturbation can be highly variable (e.g., Dalrymple and Narbonne, 1996; Ito, 1996; Viana et al., 1998a). Bottom currents can play a fundamental role in the activity of benthic organisms by controlling the amount of oxygen and nutrient advection to the deep-sea floor (Chough and Hesse, 1985), but – if they are sufficiently strong – also in reworking of the sediment and in the destruction of any evidence of burrowing (Tucholke et al., 1985). With these premises, the Faro Drift model could be considered as the end member of a spectrum of situations that control bioturbation in sediments swept by bottom currents rather than as a general model for both sandy and muddy contourites.

10.5.2.

The Gulf of Mexico facies model

Based on a study of cores from the Pliocene–Pleistocene of the Gulf of Mexico, Shanmugam et al. (1993a,b) proposed a series of contourite facies and diagnostic criteria for the definition and recognition of contourites; these were essentially based on the presence of traction structures: thin-bedded or laminated sand, silt and mud layers, inverse grading, sharp contacts with sand–silt layers, internal erosional surfaces, lamination, cross-lamination, lenticular and flaser bedding, and mud offshoots (Figure 10.12 and Table 10.1). The proposal by Shanmugam c.s. does not include any typical vertical succession of facies. These facies and diagnostic features are radically different from those of the Faro Drift contourite model, indicating conditions of deposition in which the sandy material is much more abundant and the activity of bottom currents prevails over the burrowing activity of benthos in the resulting facies successions. The diagnostic criteria and facies of Shanmugam c.s. have been criticized severely by Stow et al. (1998a), who considered them as a step backward, toward the earlier contention that bottom-current deposits can be recognized solely on the basis of sedimentary structures.

180

Traction Structures in Contourites

Table 10.1 Comparison of the sedimentary structures that have been considered as diagnostic for contourite recognition by Stow c.s. and by Shanmugam c.s. (based on several works from both schools, see text for references), and the sedimentary structures of contourite deposits considered in this chapter Sedimentary structure

Stow’s contourite sequence type (facies model)

Shanmugam’s diagnostic criteria for bottom-current deposits

This review: traction structures in contourites (Martı´ n-Chivelet, Fregenal and Chaco¤n)

Burrowing

Relevant in most cases; main diagnostic feature Subordinate

Not a diagnostic feature; rare in most cases Common

Strongly variable, depending on environmental conditions; rare in many cases Very common

Subordinate

Diagnostic feature

Very common

Not diagnostic; common in fine-grained turbidites

Diagnostic feature

Very common in nonbioturbated contourites

Rare

Common

Common

Common and diagnostic

Common and diagnostic

Common and diagnostic

Not diagnostic; common in fine-grained turbidites Predominance of gradational contacts

Diagnostic criteria

Common

Diagnostic criteria

Sharp contacts are common; erosional bottom contacts are frequent.

Not common

Diagnostic criteria

Very common

– Not common; not diagnostic

– Diagnostic criteria

Rare but diagnostic Very common; diagnostic

Low-energy parallel lamination; thin-bedded or laminated sand in deep water mud Small-scale cross-lamination; climbing ripples Flaser bedding, lenticular bedding, ‘‘mud offshoots,’’ starved ripples Large-scale cross-bedding Coexistence of reverse and normal grading in vertical sequences at different scales ‘‘Rhythmic’’ occurrence of sand and mud layers Sharp (nonerosional) upper contacts and sharp to gradational bottom contacts Internal erosional surfaces Longitudinal ripples Lag deposits, winnowing of bioclasts

(Continued)

181

J. Martı´n-Chivelet et al.

Table 10.1

(Continued )

Sedimentary structure

Stow’s contourite sequence type (facies model)

Shanmugam’s diagnostic criteria for bottom-current deposits

This review: traction structures in contourites (Martı´ n-Chivelet, Fregenal and Chaco¤n)

Reactivation surfaces in ripples and larger bedforms Upper-phase plane beds, parting lineation

Not diagnostic

Diagnostic criteria

Diagnostic criteria

Not diagnostic; common in turbidites

Diagnostic criteria if not a part of a bouma sequence, with a basal graded unit

Very common

10.5.3.

The Caravaca facies model

Martı´n-Chivelet et al. (2003) provided a facies model for Maastrichtian foraminifer-rich calcarenitic contourites in the Betic Cordillera of southern Spain (Figure 10.12). The identification of the deposits as mid-depth contourites did not result from application of the ‘‘standard facies model’’ of Stow et al. (1996b), and in fact the results do not resemble that standard in any way. The authors used independent palaeogeographical and sedimentological evidence, and analyzed and discarded other options (e.g., turbidites) to explain the origin of these sediments before concluding that bottom currents were the most plausible agent. Sedimentary structures were an important key in this work since they permitted the determination of the type of current and its hydrodynamic characterization, showing common characteristics with typical bottom-currents features. An interesting point is that, in this case study, a characteristic vertical succession of facies and structures was recognized, reflecting a gradual increase, a maximum and, finally, a gradual decrease in current strength. This model is not intended to be generally applicable, but it indicates the need for opening the field of contourite facies models to more possibilities than the ‘‘standard facies model,’’ since the possible scenarios forming contourites in recent and past oceans might be much more diverse than initially thought.

10.6.

CONCLUSIONS

Traction structures are inherent to sediments generated or influenced by water flows. Contouritic deposits, developed in close association with bottom currents, are not an exception, and show a noticeable variety of bedforms and structures; their study is essential for understanding the patterns of oceanic

182

Traction Structures in Contourites

palaeocirculation. Sedimentary structures of ancient contourites are ‘‘diagnostic indicators’’ of the currents from which they were derived, but their interpretation should always be restricted to the processes rather than to the type of sedimentary event or environment until its full context (particularly, the genetic association with other structures and the palaeogeographic framework) is understood (Table 10.1). If we accept the high variability in the typology of deep-sea sediments that are affected by bottom currents, the controversy on whether primary traction structures or bioturbation should be the basic diagnostic criteria for the recognition of contourite deposits becomes, obviously, much less important. It clearly demonstrates the irrelevance of concentrating the discussion on the diagnostic use of only sedimentary structures, or, on the contrary, of completely ignoring these in facies models and denying their preservation potential. Such a complex system cannot be simply summarized in just a single or a few facies models. This analysis implies the need for a notable increase in our knowledge of contourites, and a more profound analysis of recent and ancient study cases with an open mind. It will not be possible to propose new facies models if all examples that do not fit with the existing ones are simply discarded. For that purpose, it is important to invest new efforts in: (1) finding out the relationships between sedimentary structures and facies of contourites in the framework of large-scale bedforms; and (2) comparing well-exposed ancient examples with recent cases. The future of contourite research has a great potential in hydrocarbon exploration, palaeoclimatology and palaeoceanography, but it must gain a much more complete knowledge of the processes that are responsible for contour currents and the resulting deposits.

ACKNOWLEDGMENTS We are grateful to M. Rebesco, A.J. van Loon, and A. Camerlenghi for their critical reviews and editing of the manuscript. Graham Evans, Alfredo Arche, and Jose´ Lo´pez-Go´mez are thanked for their constructive comments and suggestions. The work is a contribution to project CGL2005-06636-C02-02/BTE, and to the ‘‘Basin Analysis’’ and ‘‘Palaeoclimatology and Global Change’’ Research Groups (UCM-CAM). The research by B.Ch. was supported by a postdoctoral grant of the Spanish Ministry of Education and Science.

C H A P T E R

9

T HE N ATURE OF C ONTOURITE D EPOSITION D.A.V. Stow1, S. Hunter1, D. Wilkinson1 and F.J. Herna´ndez-Molina2 1

National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, UK Facultad de Ciencias del Mar, Universidad de Vigo, Vigo, Spain

2

Contents 9.1. 9.2. 9.3. 9.4. 9.5. 9.6.

Introduction Bottom-Current Characteristics Sedimentation rates and budget Drift Deposition and Erosion Bottom-Current Bedforms Contourite Facies and Features 9.6.1. Sedimentary structures 9.6.2. Sedimentary texture and fabric 9.6.3. Contourite composition 9.7. Contourite Cyclicity 9.8. Summary Acknowledgements

9.1.

143 144 146 147 149 151 151 152 153 154 155 155

I NTRODUCTION

Contourites are sediments that have been deposited by or significantly affected by bottom (contour) currents (Stow et al., 2002b; Rebesco, 2005; Stow and Fauge`res, 2008). They are a group of closely related, essentially deep-water facies, typically deposited below about 300 m water depth under the influence of semipermanent current action, and are commonly referred to as along-slope deposits resulting from semi-continuous depositional processes. This distinguishes them from other deep-water facies that have been deposited either by episodic down-slope processes or events (turbidites, debrites, slides and hyperpycnites), or from continuous vertical settling – the so-called background processes (pelagites and hemipelagites). Contourites are found covering large areas of the present-day sea floor beneath modern bottom-current systems, in some regions building up gigantic contourite mounds or drifts through semi-continuous deposition over a period of millions of years, often closely associated with and adjacent to regions of erosion (Rebesco and Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00209-4

Ó 2008 Published by Elsevier B.V.

143

144

The Nature of Contourite Deposition

Stow, 2001; Fauge`res and Stow, 2008). This deposition represents a significant degree of sediment focussing compared with the much lower rates and thinner accumulation of normal background sediments. Contourites also occur closely interbedded with the other deep-water facies. They range from very fine-grained (mud and silt) to relatively coarse-grained (sand and gravel) deposits, and include siliciclastic, bioclastic, volcaniclastic and chemogenic compositional varieties. Based on the large amount of information gleaned from modern contourite systems, it is possible to construct a fairly accurate picture of just how, where and when contourite deposition occurs. An associated picture emerges of when and where deposition gives way to non-deposition and erosion by bottom currents, and how long-term accumulation of contourites can result from the alternation of deposition and erosion in time and space. Following a brief overview of the bottom-current process, this chapter aims to review the depositional mechanisms of contourites at the large, intermediate and small scales. We therefore consider, in turn, the information that can be derived from drift construction and erosion, sea-floor bedform development, and the detail of sediment facies characteristics.

9.2.

BOTTOM -C URRENT C HARACTERISTICS

Bottom (contour) currents are those currents that operate as part of either the normal thermohaline circulation or wind-driven circulation systems, and are generally semi-permanent features in the ocean basins, often long-lived through geological time. In general, therefore, they are acting continuously to affect the pattern of sedimentation in the areas where they occur. The principal characteristics of bottom currents that most affect contourite deposition have been derived from numerous sources (Nowell and Hollister, 1985; McCave et al., 1988; Gross and Nowell, 1990; Gross and Williams, 1991; Stow et al., 1996b) and can be summarised as follows (Figure 9.1). 1. They have a net flow along-slope, but can also flow up-slope, down-slope, around and over topographic obstacles or irregularities. The level within the water column at which maximum flow occurs is dependent on the density (determined by salinity, temperature, and suspended load) of the water mass involved, and the major effects on sedimentation are felt where this flow impinges on the sea floor. 2. They typically act as a broad sluggish movement of water (mean velocity <10 cm s 1) over low gradient slopes and in ocean basins, as more constricted intermediate velocity flows (10–30 cm s 1) over steeper slopes and around topographic obstacles, and as highly constricted high velocity flows ( >30 cm s 1) through narrow gateways, passages and over shallow sills. The lower velocity flows can only transport and deposit clay and fine silt-sized material, intermediate velocities can transport up to fine sand-sized material in suspension as well as move coarser sediment as bedload, whereas the higher velocity flows can affect still larger grain sizes.

145

D.A.V. Stow et al.

s

w m

and

– w m

w

Figure 9.1 Bottom-current characteristics: a schematic summary of principal features. A multicolour version of this figure is on the enclosed CD-ROM.

3. Bottom currents are highly variable in location, direction and velocity over relatively short timescales (from hours to months). Velocity increase, decrease and flow reversal occur as a result of deep tidal effects (e.g. Shanmugam, 2008). Seasonal changes can result from variation in properties of the water masses generated in the source regions. The mean flow velocity decreases from the core to the margins of the current, where large eddies peel off and move at high angles or in a reverse direction to the main flow. The flow velocity is directly affected by changes in slope gradient and other topographic irregularities along its course; and also by current meandering and subdivision into two or more strands around obstacles. This kind of flow variability leads to many cycles of deposition, non-deposition and erosion during the course of contourite accumulation. 4. Eddy kinetic energy, sea-surface topographic variations and surface current instabilities can all be transmitted through the water column and so result in marked variation in kinetic energy at the sea floor. In places, this leads to an alternation of short (days to weeks) episodes of higher velocity benthic storms, and longer periods (weeks to months) of lower velocity. Benthic storms can result in further erosion and resuspension of large volumes of sediment, its incorporation into the bottom nepheloid layer (i.e. suspended sediment load) and transport downstream. Deposition occurs during the quieter low-velocity periods. 5. Bottom currents also show longer period variability (from decadal to millennial). Some of this can be directly related to climate and sea-level change, for example at the scale of Milankovitch cyclicity, which in turn influences the temperature–salinity

146

The Nature of Contourite Deposition

properties of the deep-water masses generated in the source regions as well as the volume of deepwater generated and, in some cases, the amount of water that escapes through oceanic gateways to feed thermohaline circulation. Controls on other period changes are less well understood at present, although their effects on cyclic deposition can be observed in the sediment record of contourites.

9.3.

SEDIMENTATION RATES AND BUDGET

The average sedimentation rate of contourite deposits varies significantly depending on the location with respect to long-term nature and velocity of the bottom-current system and volume and source of sediment supply (McCave, 2008). Regions of long-term erosion and non-deposition will record zero rates of accumulation, whereas typical sedimentation rates on sheeted drifts range from 3 to 10 cm ka 1, and on mounded drifts from 5–30 cm ka 1. Rates as high as 65 cm ka 1 have been recorded on some drifts (Howe et al., 1994, 2002). These values compare with pelagic rates that are generally <2 cm ka 1 and hemipelagic rates of 5–15 cm ka 1 (Stow and Tabrez, 1998). Sediment is transported within, and ultimately passes through, the bottom nepheloid layer on its way to form a contourite deposit. Although typical sediment concentrations are relatively low in nepheloid layers associated with bottom currents (0.01–0.1 ppm, or 0.02–0.2 mg l 1) (McCave, 1981), they are episodically increased up to tenfold as a result of benthic storm erosion and resuspension. This increase presumably also applies to other temporal increases in sediment supply, e.g. from localised fine-grained turbidity current input. Sediment supply to the nepheloid layer (see also He et al., 2008) comes from a range of sources: (a) vertical flux from windblown particles, river suspension plumes, glaciomarine suspension and volcanic dust delivered to the sea surface; (b) vertical flux from sea surface primary productivity, including organic material, calcareous and siliceous bioclastic debris; (c) vertical to slow horizontal advection by a combination of hemipelagic processes, including suspension cascading; (d) direct downslope flux from low-density turbidity currents and hyperpycnal plumes; (e) intermittent downslope flux via spillover processes, including bioturbational and shelf-edge current resuspension and (f ) erosion of the sea floor and resuspension by bottom currents immediately adjacent to and upstream from the site of deposition. The relative importance of sediment flux from these different sources will vary considerably between drifts of different regions. As a schematic example, we can consider the total sediment budget for the Eirik Drift in the NW Atlantic off the southern tip of Greenland (Hunter et al., 2007a, b). Currently, the Eirik Drift measures some 300 km in length, has an average width of 70 km and is up to 0.7 km thick. The total flow of the Deepwater Boundary Current (DWBC) into the northern end of the drift is measured at approximately 6 Sv (6  106 m3 s 1). If the mean flow concentration (i.e. nepheloid layer) is 0.1 mg l 1, then the mass sediment flux is around 600 kg s 1, yielding an annual sediment flux of around 2  1010 kg (or 2  107 tonnes). We show the possible distribution of this flux between the different inputs in Figure 9.2.

147

D.A.V. Stow et al.

Outflux in bottom current 200–250

Influx for drift construction Turbidity currents 30–50

Pelagic settling 30–50

Slope spillover 30–50

Pelagic settling 100–150

Deposition on drift 500–600

Local erosion by bottom current 30–50

Upstream influx to bottom current

Bottom-current influx 600

Slope spillover Turbidity 50–150 currents 100–150

Bottom-current erosion 200–300

Figure 9.2 Estimated sediment flux from different inputs into the Deep Water Boundary Current that flows over and feeds sediment to the Eirik Drift. A multicolour version of this figure is on the enclosed CD-ROM. Arrows indicate sediment flux, not current position.

9.4.

D RIFT D EPOSITION AND E ROSION

At the large scale, contourite deposition is focused in contourite drifts, which range in scale from around 50 to >106 km2, and in larger contourite depositional systems that comprise several related drifts and associated erosional elements (Fauge`res et al., 1993; Stow et al., 2002b; Rebesco, 2005). The larger drifts clearly demonstrate long-term continuity of deposition over several millions of years, which allows the accumulation of several hundreds of metres of contourite sediment. The associated erosional elements represent regions of marked erosion and/or non-deposition by bottom currents that may also persist for up to millions of years. In general, these areas display erosive winnowing of the sea floor and accumulation of coarser-grained (sand and gravel) contourites. The principal types of contourite drift that have been identified include sheeted, mounded-elongate, channel-related, confined, infill and mixed systems (McCave and Tucholke, 1986; Fauge`res et al., 1993; Rebesco and Stow, 2001; Rebesco, 2005; Fauge`res and Stow, 2008). A fundamental difference exists between sheeted drifts that accumulate with very low relief over a broad area, and mounded drifts that develop an elevated relief of thicker accumulation over a narrower, elongate region. Both sheeted and mounded forms can occur as generally smaller deposits within channel settings (e.g. oceanic gateways), structurally confined basins, and as the infill of isolated topographic lows (e.g. slump scars). Mixed drift types are those that combine an element of both along-slope and down-slope deposition. The principal erosional elements, as defined by Herna´ndez-Molina et al. (2008b), include erosional terraces, abrasive surfaces, contourite channels, contourite moats, marginal valleys and furrows. These also divide into two fundamental genetic types: planar and linear erosive features. Erosional terraces and abrasive surfaces are both of broad planar extent, the former occurring under maximum

148

The Nature of Contourite Deposition

current velocity with erosion and the latter under the influence of strong currents with non-deposition and erosional scouring, as well as deposition of sand sheets and sand/gravel ribbons. The linear features include larger contourite channels and smaller elongate furrows, as well as channels related to slope drifts (contourite moats), and those caused by erosion around a topographic obstacle such as a seamount (marginal valleys). These different types of depositional and erosional contourite system are controlled largely by a combination of factors: the nature and style of bottomcurrent flow (e.g. tabular versus multicore flow of Herna´ndez-Molina et al., 2008b); the slope gradient and other topographic features; and the sediment supply. The information they yield on the nature of contourite deposition is briefly reviewed below and illustrated in Herna´ndez-Molina et al. (2008b; their Figure 19.17), with particular reference to slope contourite systems. Sheeted drifts typically represent relatively slow rates of deposition of fine-grained contourites over a large area of sea floor (slope plastered drifts and abyssal sheet drifts). The flow is mostly simple and tabular, broad and regionally stable, although it may also include strands of more intense flow and giant eddy circulation. Contourite deposition appears to take place directly from suspension more or less evenly across the whole flow width. There is probably a complete gradation in depositional style between sheeted drifts (with very low mounded geometry) and thinner sheets (or beds) of contourites closely interbedded with other deepwater facies (hemipelagites, turbidites, etc). Contourite sand sheets are a depositional/erosional body linked with more localised zones of higher energy flow, as found in erosive terraces and abrasive surfaces below very active bottom currents. Erosion, winnowing and deposition as bedload alternate continuously. Mounded drifts represent relatively enhanced rates of deposition of fine-tomedium-grained contourites, commonly focused into slope-parallel, elongate sediment bodies over moderate to large areas of sea floor. The flow may be either simple (as above) or part of multiple current pathways (multicore flow), and tends to be markedly intensified at drift margins, in some cases causing narrow erosional zones: contourite channels, moats and marginal valleys. Slower flow and large eddies dominate over the drift itself and lead to enhanced deposition and hence to gradual build up of the drift mound. It remains somewhat unclear on which side of the current core drift build-up is most likely to occur, and whether or not this is primarily a result of Coriolis deflection. There are a number of apparently contrasting examples in the literature, so that we consider it more likely that drift deposition is focused in-between different strands of a multicore flow pattern, whereas erosion, non-deposition and coarse-grained contourite facies occur directly beneath the high velocity cores. Over time, and in response to sea level, climate or other forcing factors, the flow pattern varies both spatially and in intensity, so that individual drifts build up by differential aggradation, progradation and erosion (Herna´ndez-Molina et al., 2006c; Llave et al., 2001, 2006). Channel-related drifts are those associated with deepwater channels, passageways or gateways through which bottom circulation is constrained, which leads to an increase in flow intensity and velocity. Where channels are broad, they are typified by multicore flow pathways that may result from topographic interaction

D.A.V. Stow et al.

149

and channel margin effects. Smaller channels by contrast may show simple flow, although flow meandering and edge effects are also common. The channel region is characterised by erosion, non-deposition and coarse-grained contourite facies, together with deposition of finer-grained contourites in localised patch drifts (either mounded or sheet-form). The channel exit region experiences flow broadening and deceleration, together with re-combination of distinct current strands. Deposition occurs as a sheet-like contourite-fan, typically with downflow decrease in contourite grain size (Fauge`res et al., 1998, 2002b: Masse´ et al., 1998). Confined drifts are among the lesser known systems which, to some extent, appear similar to the broad channel systems (above). They are probably affected by multicore flow pathways and typified by zones of both erosion and deposition. Data from the Sicilian gateway in the central Mediterranean have suggested an interesting aspect of flow behaviour and contourite deposition (Reeder et al., 2002). There are a number of confined basins through which the Levantine Sea bottom water flows and in one of these the nature of sediments and high rate of sedimentation suggest that a process of bottom-current flow lofting has occurred. The result is one of a mixed contourite/hemipelagite sheet-like deposit. Infill drifts are even less known and mixed systems are extremely variable. For both these types, which are not discussed here, additional information may be found in Fauge`res and Stow (2008).

9.5.

B OTTOM -C URRENT B EDFORMS

At a medium scale, the impact of bottom currents in shaping the deep sea floor over both depositional (drift) and erosional elements is well known (Hollister and Heezen, 1972; McCave and Tucholke, 1986; Masson et al., 2004). The sea floor is smoothed and/or sculpted into a wide variety of bedforms at a range of different scales that can provide important insights into both flow characteristics and depositional and erosional mechanisms of contourites. Surface lineation and ripples are ubiquitous at the centimetric scale, sand waves and dunes are common metric scale features, and mud furrows and sand ribbons occur with a spacing of tens of metres and lengths up to several kilometres. In a recent paper, Stow et al. (in press) have synthesised a large amount of this data into a bedform/velocity matrix (Figure 9.3), from which one can derive information on flow direction, velocity, variability and continuity. The nature and distribution of bedforms also allows the following comments on various aspects of contourite deposition. 1. The widespread fine-grained contourites of many drifts, with smoothed sediment surfaces and/or surface lineation, represent deposition of silt and clay directly from suspension through a laminar boundary layer, which in places is subject to a series of sub-parallel, small-scale, helical flow vortices creating the linear sediment fabric. 2. The common presence of ripple bedforms at all scales on silt/sand substrates indicates that tractional movement of bedload at the base of flow is the normal

150

3.0

2.0

Flow velocity ( m s–1)

The Nature of Contourite Deposition

Irregular scour Gravel furrows

Erosional bedforms

Sand furrows

1.0 Mud furrows

Gravel waves Sand ribbons

Barchan dunes

Giant mud waves

0.1 ±surface lineation

Smooth surface

Groove and ridge

Obstacle + comet scour

Sand waves Linguoid ripples

Depositional bedforms

0.5

Gravel bars and irregular gravel patches ± scour

Gravel ribbons

Undulatory ripples

Sand/gravel lineation

Obstacle and scour

Straight ripples Smooth sand sheet ± surface lineation ± crag and tail

Contourite Gateways + channels

Sand sheets

Contourite drifts 0.05 clay + silt

0.063

0.125

0.25

0.5

sand

1.0

2.0

Grain size (mm) gravel

Figure 9.3 Bedform-velocity matrix: a schematic distribution of bedform types with respect to flow velocity and grain size (from Stow et al., in press; with permission from the Geological Society of America).

mode of transport and deposition of fine-to-medium-grained granular material in bottom currents. 3. In zones subjected to higher velocity currents, tractional movement of coarser materials is evidenced by sand waves, barchan dunes and, more rarely, gravel waves and bars. That these bedforms are often covered by smaller-scale ripples is evidence of bottom-current variability, probably over timescales of hours (tidal influence) to weeks (benthic storm effects). Periods of intense bedload transport therefore alternate with periods of lesser transport and deposition. 4. Some linear bedforms, known as furrows, are formed at relatively higher flow velocities in both mud and sand/gravel substrates. These represent mainly erosive conditions at the base of flow leading to sediment entrainment and transport. Groove and ridge structures (also known as longitudinal triangular

D.A.V. Stow et al.

151

ripples (see also Martin-Chivelet et al., 2008) appear to be a smaller scale equivalent of mud furrows, formed at moderate flow velocities over fine substrates and involving both deposition and erosion. 5. The lateral juxtaposition of bedform types occurs over a horizontal scale of metres, indicating the variability in velocity of strands of flow (or regions of flow) at this order of magnitude. There is also the larger-scale of variation over hundreds of metres from more dominantly erosive to mainly depositional. This can occur in an across-flow sense, from erosive marginal moat to central depositional drift, and in a down-flow sense, away from a gateway or channel exit. 6. The development of large fields of giant sediment waves and their persistence in time through the sedimentary record (thousands to a few million years), reflects the broad tabular flow and long-term stability of low-velocity bottom currents in their region of formation. Deposition of the fine clays and silts that make up these bedforms is almost certainly directly from suspension, under the influence of internal lee waves in a weakly stratified bottom current (Flood, 1988).

9.6.

C ONTOURITE FACIES AND FEATURES

At the scale of the sediment itself, a wide range of different contourite facies are now recognised (Stow et al., 1996b, 1998a, 2002b; Stow and Fauge`res, 2008). These range in grain size from fine muds, through silts and sands, to sand and gravel lag deposits, and are often poorly sorted mixtures of different grain-size fractions. In composition they are equally varied, including siliciclastic, bioclastic (calcareous, siliceous), volcaniclastic and chemogenic (manganiferous) varieties, commonly displaying a mixed composition. Their principal characteristics, as illustrated and tabulated in the above referenced papers (including Figures 13.9 and 13.10 in Stow and Fauge`res, 2008), give further important insights into the mechanism of contourite deposition.

9.6.1.

Sedimentary structures

Based on the many observations of smoothed sediment, surface lineation and regular bedforms (ripples, dunes, etc.) on the sea floor beneath bottom currents (see above), we might expect contourite sediments to show extensive parallel lamination as a result of fluid flow processes and depositional sorting mechanisms, as well as crosslamination at different scales as a result of bedload traction (Martin-Chivelet et al., 2008). However, most contourites recovered from drift systems beneath extant contour currents are characterised by a notable absence of clear, distinct lamination and by the presence of common to abundant, pervasive bioturbation. In some cases, they appear completely homogeneous, whereas in other cases they show indistinct and discontinuous parallel lamination, partial grain alignment, sub-horizontal to irregular erosion surfaces, thin layers and lenses of coarser material. Cross-lamination is only rarely present in silts and fine sands, and slightly more common in medium to coarse-grained sands.

152

The Nature of Contourite Deposition

The presence of pervasive bioturbation rather than lamination might be explained by relatively low rates of contourite accumulation, so that bioturbation is able to keep pace with deposition and effectively destroy most primary lamination, leaving only remnants as indistinct lamination. Furthermore, for thick, muddy contourite accumulations, the low to very low current velocities and sediment concentrations are insufficient to result in primary lamination through the depositional sorting mechanism that develops silt/mud lamination in fine-grained turbidites, for example. The minor erosion/non-depositional surfaces as well as coarser lenses and layers within muddy contourites provide evidence of repeated and alternating phases of erosion, winnowing and deposition. Laminated and cross-laminated sandy contourites are known from beneath higher-velocity bottom currents with large-scale bedforms (e.g. dunes) evident on the sea floor. The lamination is distinct, but can also be diffuse, and may be associated with bioturbation. The presence of such structures clearly indicates bedload tractional movement of granular sediments by the bottom current. Preservation of the primary lamination is probably due either to intermittently rapid sedimentation from high-velocity currents and/or dearth of organic matter inhibiting extensive bioturbation.

9.6.2.

Sedimentary texture and fabric

The dominant grain size of mud-rich contourite drift deposits is clayey silt and silty clay, generally ranging between 3 and 11 (125–0.5 mm). They commonly show poor sorting (1.4–2.5 ), bimodal or polymodal grain-size distribution, and may contain significant (up to 15%) sand-size material ( >63 mm) (Figure 9.4). Grain alignment of silt and fine magnetic particles (using anisotropy of magnetic susceptibility measurements) shows flow-parallel trends. These characteristics are all commensurate with transport of a mixed-composition load within the bottom current and deposition directly from suspension. The finest material (<10 mm) is most likely carried in the form of larger aggregates, flocs and faecal pellets, which are then disaggregated during grain-size analysis. These are more or less hydrodynamically equivalent to the silt fraction between 10 and 63 mm, which is referred to as sortable silt and used as an indicator of flow velocity (McCave et al., 1995b; McCave, 2008). Sand-sized material in muddy contourites is typically made up of biogenic tests (calcareous or siliceous), which may either be hydrodynamically light and so be transported within the ambient current suspended load, or represent the direct fall of pelagic material through the current. At higher latitudes, ice-rafted debris is a common addition, which almost certainly represents glaciomarine hemipelagic fallout. Sandy contourites are mostly fine to medium-grained, more rarely coarsegrained or with a gravel component. In many cases, they are only moderately to poorly sorted (0.8–2), partly as a result of bioturbational mixing with mud grade material, whereas the laminated sands may be moderately well sorted (0.5–0.7). There is little good fabric data available for this facies. Grain-size distribution spectra are more or less unimodal and, on cumulative frequency plots, commonly show a tripartite subdivision into a coarser-grained bedload fraction moved by

153

D.A.V. Stow et al.

Extra fine tail due to bioturbation

60 Saltation load coarse silt and fine sand part traction/part suspension

40

20 Bedload

0

0 1 2 3 4 5 6 7 1000 250 63 16

Medium and coarse sand deposition from traction

8 4

40

20

0 2 250

4 63

6 16

8 4

10 1

φ Grain µm size

Suspended load clay, silt, v.fine sand sorting by current deposition from suspension

60

40

20 Saltation load minor

0 6 16

8 4

10 1

φ Grain µm size

80

60

SortableSand silt fraction fraction dominantly part current sorting part biogenic component components sorting most in suspension minor saltation

Suspended load

Cumulative weight (%)

80

4 63

60

100%

Sortable silt fraction

2 250

Suspended load (100%) clays and silts (>90%) minor biogenic very fine sand deposition from suspension

(d) Biogenic-rich contourites

(b) Contourite silts 100%

80

9 10 φ Grain 1 µm size

Cumulative weight (%)

Cumulative weight (%)

80

(c) Contourite muds 100%

Suspended load clay, silt, v.fine sand deposition from suspension

Cumulative weight (%)

(a) Contourite sands 100%

40

20

Saltation load minor

0 2 250

4 63

6 16

8 4

10 1

φ Grain µm size

Figure 9.4 Grain-size distribution and characteristics of a range of different contourite facies, plotted as smoothed cumulative frequency curves. (a) Sandy contourites, (b) silt and silt ^ mud contourites, (c) muddy contourites and (d) biogenic-rich contourites. A multicolour version of this figure is on the enclosed CD-ROM.

traction, an intermediate fraction moved as saltation load, and a finer-grained fraction transported wholly in suspension (Figure 9.4). Each fraction may show more than one compositional sub-population separated by hydraulic sorting during transport and deposition. Still coarser grained contourites (coarse sand and gravelrich) are moved wholly and intermittently as bedload, whereas in gravel-bearing muddy contourites, the gravel clasts generally have an ice-rafted origin.

9.6.3.

Contourite composition

As noted earlier, there is a wide range of different components in contourite sediments, which themselves yield some clues as to the nature of deposition. The most common components are shown in Figure 9.5, together with their typical range of grain size. Although some contourites have a more or less single uniform composition, such as mid-ocean drifts with >90% pelagic biogenic material and high-latitude drifts with >90% glaciomarine hemipelagic material, most contourites show a characteristically mixed composition. This indicates a range of sources and supply

154

The Nature of Contourite Deposition

J

Medium

F

A S

Fine

Very fine

C

Fine

N

Sand

M

Silt Very fine

Clay

P Q M

Kaolinites

H J

C

Faecal Particulate

Micrite

Carbonate silt/sand

Cohesive properties

Figure 9.5

Granular properties

Common components of contourites and their typical grain-size range.

routes coupled with a depositional process (or processes) that tends to mix rather than segregate components (see also Figure 9.2). Certain processes and hence component inputs will dominate in different contourite settings. In most cases, the sand-sized fraction will show partial fragmentation, rounding and iron staining, which is all indicative of bottom-current transport as saltation load and bedload.

9.7.

C ONTOURITE C YCLICITY

Variation in contourite characteristics and hence in the nature of contourite deposition, appears to be the norm at all scales of observation, from the seismic cycles recognised by Llave et al. (2001, 2006) and Stow et al. (2002b), to sediment facies cycles depicted in the standard facies model for contourites (Stow and Fauge`res, 2008; Hu¨neke and Stow, 2008), and to the less regular alternation of coarser-grained lenses and layers with fine muddy contourites (see above). The origin of both the seismic cycles and the facies sequence can be related to long- and medium-term fluctuations in the mean current velocity, and/or to variation in sediment supply (Knutz et al., 2001, 2002a, b). Stacked sequences and repeated

155

D.A.V. Stow et al.

partial sequences indicate cyclic variation in the forcing variables, while cycle irregularity is likely to result from interaction of sediment supply and current velocity. The periodicity ranges from 1000 to 2000 years for some of the shorter period facies cycles to 100,000–200,000 years for some of the regular seismic cycles. At the smallest scale of observation, alternation of coarser- and finer-grained contourites in partial and rather indistinct layers and lenses is the result of multiple intermittent depositional, non-depositional and erosional phases (over a period of hours to weeks) within an overall continuous sedimentation regime.

9.8.

SUMMARY

There is now a considerable body of evidence on the nature of contourite deposition, derived from direct observations of bottom currents, deep-sea nepheloid layers, contourite drifts, sea-floor bedforms and contourite sediment facies. The principal characteristics of that deposition reflect: 1. in part, the long-term stability of bottom-current systems, allowing for significant sediment focusing and large-scale drift accumulation coupled with zones of pronounced erosion and non-deposition; 2. in part, spatial and temporal variability in mean current velocity and/or sediment supply; semi-permanent bottom currents lead to semi-continuous deposition of contourites at rates some 10–20 times the normal background rate for pelagic accumulation; 3. sediment influx into the bottom current, which occurs via vertical settling of material from the surface, slow horizontal downslope advection, direct input from low-density turbidity currents and other down-slope spillover processes, and by direct bottom-current erosion; 4. drift deposition beneath simple tabular flow operating over broad regions of a gently sloping seafloor, and under more complex current systems (including multicore current strands), over steeper slopes and through narrow gateways and channels; 5. deposition, non-deposition and erosion, which occur repeatedly over both short and long timescales as a result of variation in bottom-current velocity, flow location and sediment influx; 6. for the coarser-grained to finer-grained facies, bedload traction, winnowing, re-suspension, sediment saltation and direct fallout from suspension; 7. mostly relatively slow and semi-continuous, intense post-depositional bioturbation, which is the norm for most contourites.

ACKNOWLEDGEMENTS We thank many colleagues for their time and discussion at different stages of this research. We also acknowledge the technical and administrative support at our respective institutes. This work has been carried out as part of two research stages

156

The Nature of Contourite Deposition

funded by the Secretarı´a de Estado de Educacio´n y Universidades (Spanish Ministry of Education and Science). These awards enabled D.A.V. Stow to work at the Instituto Espan˜ol de Oceanografia, Malaga (Reference SAB2005-0182) and F. Javier Herna´ndez-Molina to work at the National Oceanography Centre, Southampton (NOCS) (Reference PR2006-0275). Funding from the UK NERC Rapid Climate Change directed research programme (grant number NER/T/S/2002/00453) is gratefully acknowledged and allowed financial support for the PhD projects of Sally Hunter and David Wilkinson. The Spanish Comisio´n Interministerial de Ciencia y Tecnologı´a (CYCIT) supported this research through the Project CTM 2008-06399-C04-01/ MAR (CONTOURIBER Project). Paul Knutz and the volume editors deserve special mention for their hard work and helpful comments.

C H A P T E R

1 1

B IOTURBATION AND B IOGENIC S EDIMENTARY S TRUCTURES IN C ONTOURITES A. Wetzel1, F. Werner2 and D.A.V. Stow3 1

Geologisch-Pala¨ontologisches Institut, Universita¨t Basel, Basel, Switzerland Institut fu¨r Geowissenschaften, Universita¨t Kiel, Kiel, Germany 3 National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, UK 2

Contents 11.1. Introduction 11.2. Effects of Contour Currents on Benthic Habitats 11.2.1. Biota 11.2.2. Burrowing activity in current-affected settings 11.2.3. Bioturbation, trace fossils and ichnofabrics 11.2.4. Hiatuses 11.3. Examples of Bioturbation in Contourites 11.3.1. Gulf of Cadiz 11.3.2. Nova Scotia Rise 11.3.3. Iceland–Faeroe Ridge 11.3.4. Faeroe–Shetland Channel 11.4. Discussion and Conclusions 11.4.1. Strong currents: non-deposition horizons and sand-dominated contourites 11.4.2. Weak currents – mud-dominated contourites 11.5. Perspective Acknowledgements

11.1.

183 185 185 186 187 189 190 190 192 193 193 196 196 198 201 202

I NTRODUCTION

Modern contourite drifts exhibit a typical overall geometry and internal stratal configuration that can be distinguished from other sedimentary deposits based on seismic-reflection data analysis (Stow et al., 2002c). Because of a lack of information about large-scale drift geometry, as well as concerning the local palaeoceanographic and palaeobathymetric conditions, contourite deposits are, however, much more difficult to recognize in cores or outcrops (Stow et al., 1996b, 1998a). In addition, small-scale primary sedimentary structures that might otherwise be diagnostic are generally scarce or lacking due to bioturbation, except for specific cases, for Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00211-2

Ó 2008 Elsevier B.V. All rights reserved.

183

184

Bioturbation and Biogenic Sedimentary Structures in Contourites

Figure 11.1 Completely bioturbated Late Pleistocene sediments interpreted as contourites. Ice-rafted debris (IRD) has been reworked by bottom currents and bioturbated subsequently. Institute of Geosciences Kiel (Germany) core 16820 -2, 137^152 cm, from the southern side of the Iceland ^ Faeroe Ridge, taken at the NW slope of a channel parallel to the ridge crest (62°35.560 N, 14°18.620 W, 1649 m water depth). (a) Fresh core surface. (b) X-ray radiograph (negative; coarse material is light, fine-grained material dark) of the same interval. Note how much more detail is observed in X-ray radiographs compared with fresh core. P = Planolites; Th =Thalassinoides;Th-Gy = Gyrolithes-likeThalassinoides; vt = vertical tube.

instance, when the sedimentation rate exceeds the bioturbation rate (Figure 11.1; see also Martin-Chivelet et al., 2008). The overall accumulation rate of contourites is low because contour currents are generally very low-concentration flows and because episodes of winnowing and erosion alternate with deposition (Gross and Williams, 1991). Slowly accumulating, well-oxygenated deep-sea sediments are normally completely bioturbated (Wetzel, 1991; Bromley, 1996). Burrowing animals displace and, eventually, destroy primary sedimentary structures. In contrast, the resulting biogenic sedimentary structures provide valuable information about environmental conditions at the sea floor, because endobenthic organisms respond sensitively to oceanographic and sedimentary changes (Scha¨fer, 1962; Ekdale et al., 1984; Wetzel, 1991). Numerous studies deal with strong bioturbation of contourites as evident in visual examination of fresh cores (several contributions in Stow et al., 2002f). Only a few studies, however, address bioturbation structures in more detail based on analysis through high-quality X-ray radiograph images. These reveal much more detail than is possible with visual core analysis alone (Fu and Werner, 1994; Baldwin and McCave, 1999; Lo¨wemark et al., 2004b). The purpose of this chapter is to present a compilation of information available on bioturbation, bioturbation structures and the resultant ichnofabrics in contourites.

A. Wetzel et al.

185

11.2.

EFFECTS OF C ONTOUR C URRENTS ON B ENTHIC H ABITATS

Contour currents are typically produced by movement of well-oxygenated cold water masses formed at high latitude in the Pacific and Atlantic Oceans (Stow et al., 1996b; Salon et al., 2008; Shanmugam, 2008; Zenk, 2008). The mean flow velocity of contour currents is in many cases between 10 and 30 cm s1 (McCave et al., 1980; Stow et al., 2002c). On a short-term scale (a few days to several weeks), however, the current speed can fluctuate between a state of quiescence and that of benthic storms ( >70 cm s1; Richardson et al., 1981). As a consequence, contour currents may re-distribute sediment particles on the sea floor (winnowing and erosion) and/or take up suspended particles settling in the water column. Bottom currents may thus carry in suspension a considerable amount of fine material (<2–12 mm) and particulate organic matter (McCave, 1985a; Thomsen et al., 2002), which form the so-called nepheloid layer (Ewing and Thorndike, 1965). Because suspended organic matter is often adsorbed onto suspended clay minerals (Mayer, 1994; but see also Thomsen et al., 2002), contour currents supply food to deep-marine benthic organisms (Thistle et al., 1985; Lavaleye et al., 2002). In contrast, the changing bottom-current velocity and resultant alternation of erosion, non-deposition, and deposition, exert physical stress on the benthic habitat. Finally, it is important to note that burrowing leads to a rough sediment surface, which is important for the initiation of sediment transport by contour currents through substrate erosion (Hollister and Nowell, 1991).

11.2.1.

Biota

In the absence of horizontal bottom water flow, benthic fauna is fuelled only by the vertical flux of organic matter, which decreases exponentially with water depth (Suess, 1980). Consequently, the benthic biomass decreases with increasing water depth (Rowe, 1983). Where, however, the vertical particle flux is supplemented by lateral current-carried supply, the benthos exhibits a higher biomass and population density than on adjacent tranquil sea floor (Thistle et al., 1985; Lavaleye et al., 2002). On the continental slope off Northwest Spain, contour currents move on average at less than 10 cm s1. The nepheloid layer carries between 2 and 4 g m3 of suspended material, of which 40–100 mg m3 are composed of particulate organic matter (Thomsen et al., 2002). The abundance of benthic megafauna across the continental slope does not decrease with depth, as would be expected in the absence of bottom currents (Lavaleye et al., 2002). The bottom currents sustain a benthic population dominated by filter feeders rather than deposit feeders. Northwest of Ireland, at the BENBO (BENthic BOundary Layer) experimental sites, at water depths of 1100 m (Rockall Plateau), 1925 m (Feni Drift) and 3750 m (Porcupine Basin), local hydrodynamic conditions, controlled by contour currents of variable intensity, affect the relative contribution of the mega-, macro- and meiofauna to the benthic biomass (Gage, 1979; Hughes and Gage, 2004). At the HEBBLE (High Energy Benthic Boundary Layer Experiment) site off Nova Scotia in about 4800 m water depth, also affected by bottom currents, the

186

Bioturbation and Biogenic Sedimentary Structures in Contourites

macro- and meiofauna are on average much more abundant than expected and compare to that 2000 m shallower (Thistle et al., 1985). Due to the repeated physical disturbance induced by bottom water flow, the faunal parameters resemble those typically obtained in re-colonization experiments (Thistle et al., 1991). With respect to behavioural groups, suspension feeders are not abundant, probably as their filter apparatus can easily be plugged when suspension concentration is periodically very high (Thistle et al., 1991). Only the taxa that passively extract drifting particles from sea water by maintaining a relief on the sea floor exhibit increased abundance (Aller and Aller, 1986). Furthermore, mobile organisms that live both on and within the sediment occur at a higher proportion than at tranquil sites (Thistle et al., 1985). A high organic-matter flux and physical disturbance by bottom currents stimulate the production of bacterial biomass on the sea floor, being 6–8 times higher than at a reference site not affected by bottom currents (Yingst and Aller, 1982). The following four different hydrodynamic situations can be distinguished (Aller, 1989, 1997): 1. High current velocity ( >25 cm s1) removes sediment, organic matter and many organisms, in particular epifauna, micro-organisms, larvae and juveniles. Bacterial biomass is low. High-velocity currents support filter feeders and, to some degree, surface feeders. After a high-energy event, the fauna recovers during a long time span dependent on the size of organisms. 2. Flow deceleration leads to rapid deposition of up to several centimetres of sediment. Bacteria and meiofauna can rapidly respond to surplus trophic levels by increasing the standing stock within days to weeks (Hughes and Gage, 2004). The benthic meio- and macrofauna are at a maximum during these periods. 3. Intermediate current velocity (5–15 cm s1) results in deposition of fresh organic matter and removal of metabolites. Microbial production is enhanced. Epifaunal abundances decrease, but numbers of burrowing and tube-building organisms increase. 4. Low current velocity (<5 cm s1) leads to a depositional regime. Addition of labile particulate organic matter favours a significant increase in the standing stock of the meio- and macrofauna; mean body sizes are larger, and there is greater diversity in feeding strategies and life habits than encountered in most deep-sea habitats.

11.2.2.

Burrowing activity in current-affected settings

Benthic organisms mix the sediment so effectively that even short-lived radioisotopes exhibit a roughly constant concentration within the so-called surface mixed layer (Erlenkeuser, 1980). Expressing benthic mixing as a diffusion-like process, calculated mixing parameters and organic-matter flux appear to be positively correlated (Trauth et al., 1997). At the HEBBLE deep-water site, sediment mixing is as rapid as in near shore environments. The mixing rates vary considerably in space and time because repeated physical reworking occurs before equilibrium has been reached (DeMaster et al., 1991). The mixing profile is often discontinuous (Thomson et al., 2000), as

A. Wetzel et al.

187

burrowing organisms transfer material from the surface of the sediment to distinct depth (Hughes et al., 2005). Probably due to repeated reworking and deposition, radiotracers do not outline any obvious link between faunal parameters and mixing modes (Hughes and Gage, 2004).

11.2.3.

Bioturbation, trace fossils and ichnofabrics

The amount of labile organic matter within the sediment decreases with depth below the sea floor under steady-state conditions (Soetaert et al., 1998), followed, in general, by the benthic biomass (Rowe, 1983). Also the amount of oxygen provided by molecular diffusion from above decreases with depth in the sediment (Glud et al., 1994), forcing deep-burrowing organisms to maintain an open connection to the sea floor. Furthermore, the sediments consolidate and become stiffer downward because of the increased overburden (Einsele, 1977). It follows that the ecospace within the sea floor exhibits a partitioning, and various burrowing animals occupy specific depth intervals, the so-called ‘‘storeys’’ (Wetzel, 1981) or ‘‘tiers’’ (Ausich and Bottjer, 1982). A fossil tier comprises trace fossils that cross-cut each other while being cross-cut by deeper ones (Wetzel, 1981). The fossilization potential of a trace is higher the deeper it is produced within the sediment, where about 50 cm is the maximum depth at which a complete tier can be produced in abyssal sediments (Werner and Wetzel, 1982). It is therefore important to analyse deep-tier burrows to understand whether they represent responses to environmental conditions at the time of deposition of this layer or at the time of burrowing. Two types of bioturbation structures are distinguished by Scha¨fer (1956): (1) biodeformational structures have indistinct outlines and destroy pre-existing structures (Figure 11.2); they are, as a rule, produced near the surface, for

Figure 11.2 Ice-rafted debris (IRD) has been reworked by bottom currents and bioturbated subsequently. The biodeformational structures (bds) destroy pre-existing structures and have no sharp outlines, but an eddy-like appearance. Institute of Geosciences Kiel (Germany) core 16820-2, 417^425 cm, from the southern side of the Iceland^ Faeroe Ridge, taken at the NWslope of a channel parallel to the ridge crest (62°35.560 N, 14°18.620 W, 1649 m water depth). (a) Fresh core. (b) X-ray radiograph (negative; light: coarse material; dark: fine grained material).

188

Bioturbation and Biogenic Sedimentary Structures in Contourites

instance, by grazing animals in soft sediments (Thayer, 1979); (2) trace fossils have distinct outlines and a defined shape allowing classification in terms of paleontological nomenclature. Some trace fossils occurring in contourites are shown in Figure 11.3.

Figure 11.3 Trace fossils commonly encountered in contourites. Arenicolites represents a vertical U-tube without spreiten; the tube diameter varies between 1 and 2 mm, and the limbs are 20^60 mm apart. The depth of the burrow is in the range of 20^40 mm. Asterosoma consists of elongate (several centimetres long, a few centimetres wide) bulbous segments with a terminal or excentrical, laminated fill. The segments are oriented (sub)horizontally and are arranged around a central vertical, somewhat twisted axis. Astersoma occurs in impure muds, in vertical sections very similar to Patagonichnus (Olivero and Lo¤pez-Cabrera, 2005). Chondrites is a three-dimensional, regularly branching tunnel system consisting of an open connection to the surface and numerous tunnels, which ramify at acute angles to form a dendritic network. The producers of Chondrites appear to tolerate low-oxygenation conditions (Bromley and Ekdale, 1984). Cylindrichnus is a subconical burrow. It consists of a slightly curved tube, 2^5 mm in diameter, which is surrounded by a thick wall composed of concentric layers that thicken towards the sediment surface. Here the burrow is 10^20 mm in diameter. Diplocraterion is a U-shaped, vertical burrow with protrusive (common) or retrusive (seldom) spreite. Openings to the sea floor are often funnel-shaped. The tube diameter is 5^15 mm, the burrow width is 10^70 mm, and the vertical extension may

A. Wetzel et al.

11.2.4.

189

Hiatuses

When bottom currents prevent deposition for a considerable time span, and/or erode sediments, submarine hiatuses develop, represented by semi-consolidated firm- or hardgrounds or stable cohesive partially dewatered muddy substrates. In addition, iron and manganese oxides may be precipitated and cause partial induration of the sediment surface (Chester, 2000; Fauge`res et al., 2002b). In these cases, distinct, mainly vertical to sub-vertical, sharp-walled, unlined burrows are produced that exhibit scratch marks along the walls. They may subsequently be passively filled (Pemberton et al., 2001; Figure 11.4). When levels representing a hiatus become buried by new sediments, a marked contrast in consistency occurs between the two layers, which can be deciphered from burrow morphology (Wetzel and Aigner, 1986). In addition, the change in lithology from under the hiatus horizon to the contouritic sediments above renders these transitions even more evident in the sedimentary record. When sandy material fills open tubes across a hiatus horizon, the ichnofabric can be ascribed to the Glossifungites ichnofacies. Such intervals are clearly recognizable in the fossil

Figure 11.3 (Continued ) exceed 300 mm. Glyphichnus/Spongeliomorpha represents an arcuate, vertical and probably U-shaped burrow with 10^15 mm long scratching marks (bioglyphs) sub-parallel to the burrow axis. Four to six grooves are arranged in fan-shaped groups (Bromley and Goldring, 1992). Nereites is a winding to regularly meandering trace, consisting of a median back-filled tunnel enveloped by an even to lobate zone of reworked sediment (for details, see Uchman, 1995; Wetzel, 2002). Planolites is a normally unbranched, smooth or ornamented, lined, essentially cylindrical, predominantly horizontal burrow of variable diameter.The fill is typically structureless and of similar lithology as the host rock. Phycosiphon consists of a compound system of U-shaped burrows in an antler-like arrangement, with horizontal or inclined spreiten. A single spreiten is 1^5 mm wide, the marginal tube is thinly lined and 0.5^1 mm wide (Wetzel and Bromley, 1994). Scolicia is a large, bilaterally symmetrical, sub-cylindrical burrow having meniscate lamellae often divided into two concave sets. In traverse cross-section, a concentric structure of bilobate lamellae surrounds an excentric axis.Today, Scolicia is produced by Echinocardium-like sea urchins that prefer to live in sandy to silty sediments (Wetzel, 1984; Fu and Werner, 2000). Skolithos is a straight, simple, unbranched, normally vertical tube having a uniform diameter (4^10 mm). Teichichnus is a vertical, blade-like elongate spreiten structure resulting from the upward or downward displacement of the causative tube.Teichichnus shows some preference to occur in sandy to silty muds. Thalassinoides consists of horizontally branched networks connected to the surface by steeply inclined or vertical shafts; swellings may occur at points of branching. An eccentric fill structure often results from active filling by the burrowing organism or collapse of the burrow walls.The tubes are 5^20 mm in diameter and smooth-walled. Shrimps or shrimp-like animals can produceThalassinoides.Trichichnus is a thread-like and rarely branching cylindrical burrow up to several tens of centimetres in length. Other small, hair-like, often horizontally oriented burrows have been informally called ‘‘mycellia’’ (Blanpied and Bellaiche, 1981). Zoophycos is a 1^20 -mm-thick spreiten structure that surrounds a central shaft in distinct levels or coilings. The regular fill structure appears as crescentic en echelons in vertical sections. A marginal tube (open or stuffed) borders the spreite; the whole trace is connected to the sediment surface through a vertical shaft. The vertical extension of a Zoophycos burrow may exceed 1 m (Wetzel and Werner, 1981). The modern Zoophycos producers feed a least episodically from the surface (Lo«wemark et al., 2004a).

190

Bioturbation and Biogenic Sedimentary Structures in Contourites

Figure 11.4 Modern hiatus horizon produced by bottom currents that traverse a seamount chain within the South China Sea (14°52.40 N, 118°35.40 E; 3940 m water depth; core SO 114 -16; for details see Wetzel, 2008). The sediment has been partly eroded. Staining by Fe and Mn oxides [Mn in (b)] indicates the hiatus. A stiff sediment consistency is indicated by numerous open burrows and ornamentation of burrow margins by claw sculptures, typical for Glyphichnus (Gly) burrows. (a) Overview. (b) Detail of the Glyphichnus margin.

record and represent possible candidates to generate high seismic reflectivity because of the density and acoustic velocity contrast they represent.

11.3. 11.3.1.

E XAMPLES OF BIOTURBATION IN C ONTOURITES

Gulf of Cadiz

The Mediterranean Outflow Water (MOW) flows into the Atlantic Ocean through the Strait of Gibraltar as bottom current and then moves to the northwest along the southern margin of the Iberian Peninsula (Tomczak and Godfrey, 1994; Herna´ndez Molina et al., 2008a). The outflow is warm ( >13°C), saline ( >36.4‰), and relatively low in oxygen (4.1–4.6 ml O2 l1) (Zenk, 1975). Flow velocities decrease from about 300 cm s1 in the Strait of Gibraltar to 180 cm s1 just west of the Strait, to 30–40 cm s1 in vicinity of Faro Drift, and finally to 10–20 cm s1 south of Cape St Vincent (Baringer and Price, 1999). The sea floor swept by this current is in places free of sediment. Where the MOW decelerates, a series of sediment drifts have developed (Stow et al., 2002b; Herna´ndez-Molina et al., 2008a; Mulder et al., 2008). The MOW density and intensity are thought to have varied through time since the Messinian (5.5 Ma) in response to climatic changes (Baas et al., 1998; Stow et al., 2002b; Lo¨wemark et al., 2004b). Turbidites constitute 20–40% of the deposits outside the Faro drift complex, but only 2% within the drift (Stow et al., 2002b). At a proximal location within the upper MOW vein, a sand-rich (40%) interval covering Marine Isotopic Stage (MIS) Termination I is ascribed to increased winnowing, and is intercalated between greyish silty muds (Lo¨wemark et al., 2004b). The sedimentation rate averages <20 cm ky1. This interval is completely bioturbated, with dominant Planolites and Thalassinoides burrows (Figure 11.5a).

A. Wetzel et al.

191

Figure 11.5 Sediments affected by bottom currents of the Mediterranean Outflow Water, Gulf of Cadiz (Institute of Geosciences Kiel, Germany, core M39008; 36°22.80 N, 7°04.30 W; 577 m water depth); X-ray radiographs (negative; coarse material ^ light, fine material ^ dark). For sedimentological and ichnological details, see Lo«wemark et al. (2004b). (a) Sand-rich facies (sand content 35^45%, sedimentation rate 17 cm ky1; 0.3% Corg) representing the Blling^Allerd time span (404^420 cm). (b) Mud-rich facies (sand content <1%, rapidly deposited at 250 cm ky1, organic-rich 0.8% Corg); Holocene in age (254^270 cm). bds = biodeformational structures; Ch = Chondrites; my = ‘‘mycellia’’; P = Planolites; py = pyrite, in many instances probably pyritizedTrichichnus; vt = vertical tube;Th =Thalassinoides.

The muds, which contain about 1% sand and 0.8% Corg, have accumulated rapidly due to sluggish bottom currents (250 cm ky1) and are completely bioturbated by a low-diversity ichnofauna mainly consisting of Chondrites, Trichichnus, ‘‘mycellia’’ and pyritized microburrows (Figure 11.5b). Further downflow, the Faro Drift Complex formed at an average rate of up to 15 cm ky1 in the late Pleistocene. Hiatuses due to winnowing occur in 40% of the cores and are most likely directly correlatable with peak sand concentrations in adjacent areas (Stow et al., 2002a). Clay and silt particles (<63 mm) dominate, and intense bioturbation has been continuous with deposition (Gonthier et al., 1984). The following three interbedded lithologies have been distinguished.

192

Bioturbation and Biogenic Sedimentary Structures in Contourites

1. Sand- and silt-dominated intervals have mostly irregular, sometimes gradational or sharp contacts. Current bedforms on the modern sea floor suggest that the sandy layers were originally laminated and have been mixed later by bioturbation. Many small-diameter bioturbational tubes occur. 2. Mottled silts and muds have sharp to gradational, often mottled contacts. Some remnants of indistinct wavy lamination may occur. Chondrites, Planolites, Teichichnus and ?Thalassinoides, in addition to unidentified small burrows, have been encountered. 3. Homogeneous muds have sharp to erosional, but commonly gradational contacts. Faint lamination occurs rarely; otherwise, the muds are thoroughly bioturbated. The low-diversity ichnofauna is dominated by Chondrites, Trichichnus and ‘‘mycellia’’. Large burrows have not been observed. Pyritization is common.

11.3.2.

Nova Scotia Rise

The continental rise off Nova Scotia is affected by southward flowing North Atlantic Deep Water (McCave and Hollister, 1985). At the HEBBLE site, radiotracers document a long-term average sedimentation rate of about 5 cm ky1 (DeMaster et al., 1985). However, intermittent benthic storms generate faster flowing currents (<70 cm s1) capable of eroding and re-suspending the sediment about once a year (Gross and Williams, 1991). When the current velocity decreases again, up to several centimetres of sediment can be deposited within a few days or weeks (McCave, 1985b). Burrowers mix the sediment so effectively that only 6% of the surface mixed layer carries a primary fabric (Flood et al., 1985). Aspects of bioturbation at the HEBBLE site have been studied by Baldwin and McCave (1999) using X-ray radiographs. They recognized four tiers from top to bottom. In Tier 1 (<6 cm thick), the degree of bioturbation depends on the time since deposition, but normally exceeds 50%. Typically, biodeformational structures occur in addition to ‘‘mycellia’’. In Tier 2 (2–4 cm thick), the degree of bioturbation reaches 60%, but some lamination is still present. Barrel-shaped burrows like Bergaueria or Conostichus, funnels like Monocraterion, tubes like Skolithos and spirals like Helicodromites can occur. In Tier 3 (4–15 cm thick), the degree of bioturbation is 70–100%. Palaeophycus/ Planolites occur throughout, as well as Scolicia in the upper part and Chondrites in the lower part. Tier 4 ( >20 cm) is completely bioturbated, in particular by Thalassinoides. The recurrent funnels and barrel-shaped burrows may record a nutritional strategy of passive collection of food particles (Aller and Aller, 1986). The common vertical shafts and the openings at the surface support the observation that many benthic organisms inhabiting the deposits are also active on the sea floor (Thistle et al., 1985). Because the number of shafts below Tier 2 is low, most of these organisms are thought to burrow in Tiers 2 and 3 only. Scolicia occurs in reworked (turbiditic) silts and sands in an intermediate tier position. The burrow producers are very active and, hence, Scolica dominates in the fossil record in silt- and sand-rich deposits. In muds, Palaeophycus/Planolites dominate; Phycosiphon and Chondrites may occur in addition. In deep tiers, Thalassinoides may occur abundantly and Zoophycos more rarely.

A. Wetzel et al.

193

Further to the south on the Northwest Atlantic Margin, off New Jersey, a late Pleistocene succession interpreted to include contourites was studied ichnologically by Savrda et al. (2001). During times of warm climate and high sea level, sediments at the sea floor have been reworked by bottom currents to form contourites. Their grain-size composition varies between muddy sands and sandy muds; brownish grey banding indicates temporarily enhanced oxidation (e.g., Mangini et al., 2001). The sediments are completely bioturbated and exhibit abundant trace fossils, in particular Thalassinoides, occurring in two variants. When associated with firm- or hardgrounds, these burrows are filled with coarse material and, hence, can be ascribed to the Glossifungites ichnofacies (see above), typical of non-deposition horizons. Other Thalassinoides have been produced in soft substrates. The occurrence of Thalassinoides matches the observations made at the HEBBLE site that crustaceans temporarily exploit the sediment surface, but are dwelling within the deposits (see above).

11.3.3.

Iceland–Faeroe Ridge

The Iceland–Faeroe Ridge separates the Norwegian Sea in the north from the North Atlantic in the south. Cold Norwegian Deep Water flows into the North Atlantic partly over the crest, but mainly through the Faeroe Bank Channel that continues into the Faeroe–Shetland Channel (Borena¨s and Lundberg, 2004). During the late Pleistocene and Holocene, bottom currents reworked the mainly muddy terrigenous sediments, with abundant coarse-grained ice-rafted debris, forming intensely bioturbated glacigenic contourites (Stoker et al., 1998a; Akhurst et al., 2002; Van Weering et al., 2008). The currents were relatively slow, leading to the accumulation of finegrained, poorly sorted, organic-rich muds, with a monospecific Chondrites assemblage and some (pyritized) micro-burrows (Fu and Werner, 1994; see Figure 11.6). On the southern slope of the Iceland–Faeroe Ridge, the Norwegian Sea overflow is deflected by the Coriolis force northwestwards into a deep-sea channel system oriented parallel to the ridge (Meincke, 1983). Within the channel system, there is an asymmetric distribution of contouritic sediments (Dorn and Werner, 1993). In response to frequent fluctuations of the current velocity, the sedimentary regime varied between erosion, non-deposition, and deposition of sand and mud. On the northeastern side of the channel system, a thin, mostly coarse sediment cover documents repeated reworking and deposition under high-velocity conditions. An alternation of trace fossil assemblages in the channel system comprises Planolites, Teichichnus/Thalassinoides, Trichichnus and Zoophycos (Fu and Werner, 1994; Figures 11.6a and b). On the southwestern side of the channel system, currents favour deposition of well sorted deposits (Fu and Werner, 1994). Within sand-rich intervals, Scolicia prevails (Figure 11.6c); Chondrites, Palaeophycus, Phycosiphon, Planolites and Thalassinoides are present in lower abundance. As the mud content increases, Scolicia becomes sparser and finally disappears, whereas Chondrites, Phycosiphon, Planolites, Thalassinoides or Teichichus, Trichichnus and Zoophycos are present (Figure 11.6d; Fu and Werner, 1994).

11.3.4.

Faeroe–Shetland Channel

Within the Faeroe–Shetland Channel, the uppermost part of the drift succession is dominated by muds and sands that accumulated at an average rate of 3–7 cm ky1,

194

Bioturbation and Biogenic Sedimentary Structures in Contourites

Figure 11.6 Bottom-current-affected sediments from the southern flank of the Iceland ^ Faeroe Ridge, Late Pleistocene (X-ray radiographs, negative; dense and/or coarse material ^ light; fine-grained, water-rich sediment ^ dark). (a) Bioturbated glacigenic sediments interpreted as contourite. Current-reworked ice-rafted debris (IRD) has subsequently been mixed by bioturbation; sand content >40% (Fu and Werner, 2000), NE flank of the crestparallel channel. Institute of Geosciences Kiel, Germany, core 16397-2 (460^473 cm), 1145 m water depth (61°52.350 N, 11°10.720 W). P = Planolites; Th =Thalassinoides; Z = Zoophycos. (b) Muddy sand facies, NE flank of the crest-parallel channel. Institute of Geosciences Kiel, Germany, core 16397-2 (132^145 cm). P = Planolites; Ph = Phycosiphon; Sk = Skolithos; Th =Thalassinoides; Sk = Skolithos. (c) Silty facies from the SWside of the crest-parallel channel system; in the silty deposits. Scolicia dominates, Sct = Scolicia in traverse section, Scl = Scolicia in longitudinal section (for details, see Fu and Werner, 2000). Institute of Geosciences Kiel, Germany, core 16396 -1 (135^144 cm), 1145 m water depth (61°52.330 N, 11°14.540 W). (d) Rapidly accumulated, organic-rich muds. Institute of Geosciences Kiel, Germany, core 16384 -2 (342^351 cm), 1255 m water depth (61°57.550 N, 11°53.150 W). Faint Chondrites-like burrows crosscut by iron sulphide (py).

A. Wetzel et al.

195

with a peak rate of 10 cm ky1. Muddy sands, sandy muds and muds, representing 8%, 60% and 26% of the cored intervals, respectively, are interpreted as contourites (Akhurst et al., 2002). 1. Muddy sands are dark grey to brownish (oxidized), and colour-banded with individual layers of 0.1–0.7 m thick. They appear structureless and thoroughly bioturbated; small-scale homogenizing bioturbation is accentuated by some large, distinct burrows (Figure 11.7a). Relicts of primary lamination are very rare. The basal contacts are gradational or relatively unbioturbated, the upper ones are gradational. 2. Sandy muds are olive grey and occur as beds up to 3 m thick. They appear mostly structureless and intensely bioturbated. Small-scale (<5 mm), homogenizing burrows are cross-cut by some large, unidentified trace fossils. Bedding contacts are gradational, or locally sharp at the base of coarser horizons (Figure 11.7b).

Figure 11.7 Glacigenic sediments interpreted as contourites from the Iceland ^Shetland Channel. Fresh-core photographs (for details, see Akhurst et al., 2002). (a) Sand-rich facies, current-reworked sediments have been mixed by bioturbation to muddy sand (ms); subsequently, sand-enriched burrows (s) were produced. British Geological Survey core 61-04/39 (61°03.50 N, 3°25.10 W; 1125 m water depth) 274^290 cm (Late Pleistocene). (b) Sandy mud facies; sands have been mixed into muds, after a early phase of homogenization producing uniform sandy mud (m), distinct burrows containing some more sand (ms) have been formed, which may be ascribed toTeichichnus orThalassinoides. (c) Muddy facies, light mud resting on grey mud, the contact has been heavily bioturbated, vertical tubes and halo burrows (Palaeophycus, Planolites, Thalassinoides) are common; British Geological Survey core 60 - 05/50 (60°49.70 N, 4°12.50 W;1100 m water depth), 247^254 cm (Late Pleistocene).

196

Bioturbation and Biogenic Sedimentary Structures in Contourites

3. Muds are dark olive to olive grey, and form beds up to 2.3 m thick. These deposits appear structureless and thoroughly homogenized by bioturbation; distinct burrows may occur, for instance Zoophycos. Rarely cross-lamination is preserved. Bed contacts are gradational (Figure 11.7c).

11.4.

D ISCUSSION AND C ONCLUSIONS

On a small scale (cores and outcrops), the recognition of contourites is problematic because diagnostic criteria such as primary sedimentary structures and grain-size variations have mostly been obliterated by intense bioturbation. In response to the fluctuating current velocity, periods of erosion, non-deposition, winnowing and deposition may alternate through time at a specific site. The effects of strong and weak currents on bioturbation are addressed separately.

11.4.1.

Strong currents: non-deposition horizons and sand-dominated contourites

Strong bottom currents may lead to • erosional surfaces covered with a pebble lag deposit (Figure 11.8a); • sharp erosional surfaces covered by sand (Figure 11.8b); • deep erosion or long-term non-deposition, yielding indurated discontinuity surfaces marked by a stiff- to hardground ichnofauna; where overlain by sand, a typical Glossifungites ichnofacies is present; where covered by mud a sharpwalled piped zone (Figure 11.8c); • coarse sediment, mostly bioturbated throughout, but in some cases with primary structures preserved (Figure 11.8d). These features are especially well recognizable where the coarser-grained unit has been covered by fine-grained contourites. Normal grading needs to be analysed to distinguish between a turbiditic and a contouritic origin (Table 11.1). Most contouritic sands are poorly sorted or impure and appear to be structureless or mottled (Figures 11.5–11.7). This results from intense bioturbation driven by the organic-matter supply from bottom currents. As the bottom-current velocity fluctuates, especially during low-velocity stages, benthic animals may utilize preferentially organic-rich fine particles and mix them with sands. Detailed ichnological analyses are still too few to define the association and tiering position of any trace fossils present. Theoretically, the upper parts of the sand layers may contain biodeformational structures resulting from ploughers and passively ventilated tube systems. These burrows become overprinted by the deeper penetrating ones like Skolithos, Scolicia and Planolites in addition to burrows having a U-shaped causative tube such as Arenicolites, Diplocraterion or Teichichnus. Furthermore, burrows produced by deepdwelling crustaceans such as Thalassinoides and Gyrolithes can be expected (Fu and Werner, 1994). Interestingly, Ophiomorpha has not been described yet from contourites, probably because previous bioturbation mixed so much mud into the sand that

A. Wetzel et al.

197

Figure 11.8 Discontinuity surfaces in bottom-current-affected settings. (a) Hiatus horizon covered with mud pebbles representing the base of the Late Pleistocene within the Iceland ^ Shetland Channel in 1100 m water depth. British Geological Survey core 60 - 05/50 (60°49.70 N, 4°12.50 W), 346^358 cm (for details, see Akhurst et al., 2002). Mud pebbles (mc), one with Zoophycos (Ze), are embedded in coarse sand (cs) that in turn rests on mud with autochthonous Zoophycos (Za). The minimum amount of erosion is 40 cm, based on the burrowing depth of the Zoophycos producer (Wetzel and Aigner, 1986). (b) Holocene sand sharply overlying mud within the Iceland ^Shetland Channel in 1125 m water depth. British Geological Survey core 61-04/39 (61°03.50 N, 3°25.10W), 9^17 cm (for details, see Akhurst et al., 2002). (c) Sand (light) resting on mud (dark); sand has been infilled into open burrows (Gl) typical for Glossifungites surfaces; the sand has been intensely bioturbated. X-ray radiograph (negative). Institute of Geosciences Kiel, Germany, core 16384 -2 (61°57.550 N, 11°53.150 W), 310^320 cm (Late Pleistocene), southern flank of the Iceland ^ Faeroe Ridge at 1255 m water depth. Th =Thalassinoides; Sk = Skolithos. (d) Bioturbated contact of an initially sharp contact between glacigenic sands (with IRD clasts) and underlying muds interpreted as contourites. Gy = Gyrolithes-like burrow; P = Planolites; Th =Thalassinoides; Sk = Skolithos. X-ray radiograph (negative; dense and/or coarse material ^ light; fine-grained, water-rich sediment ^ dark). Southern flank of the Iceland ^ Faeroe Ridge at 1255 m water depth. Institute of Geosciences Kiel, Germany, core16384-2 (61°57.550 N, 11°53.150 W), 290^302 cm (Late Pleistocene).

198

Bioturbation and Biogenic Sedimentary Structures in Contourites

Table 11.1 Bioturbation in sandy turbidites and sandy contourites Turbidite

Contourite

Sediment accumulation Organic matter supply (long-term) Organic matter imported by currents Fauna

Instantaneous, very rapid

Intermittent deposition/ reworking; consecutive Vertical flux supplemented by lateral input Mainly non-refractory

Faunal abundance

Average, dependent on water depth Survivors (deep burrowers), colonizers

Faunal structure Specialists Bioturbation Bioturbation rate Degree of bioturbation Basal layer

Mainly vertical flux Mainly refractory Partially killed by turbidite deposition

Pre- and post-depositional burrowers Post-depositional Normally low, dependent on local conditions Complete reworking rare Often not completely bioturbated, exploited by deep burrowers

Partially removed by currents, but many deep burrowing survivors (Very) high due to additional food supply Permanently, adapted burrowers, larval import by currents Deep-dwelling surface utilizers Continuous, syn- and post-depositional Very high Complete reworking normal Normally completely bioturbated, vertical tubes frequent

Thalassinoides is produced instead of Ophiomorpha (see Kern and Warme, 1974). Where the mud content increases, Teichichnus, Planolites, Asterosoma (or Patagonichnus, see Olivero and Lo´pez-Cabrera, 2005) and Chondrites may also occur. Coarse-grained contourites (either sandy and/or with gravel horizons) can sometimes be mis-interpreted in small-scale core or outcrop studies as being of turbidite origin. Although we cannot yet provide a definitive ichnofacies association and tiering structure for sandy contourites, certain marked differences between the two facies are apparent (Table 11.1). Bioturbation is generally stronger and the rate of bioturbation is very high in sandy contourites compared with that of turbidites, leading to complete or near-complete reworking of the sediment. Biodeformational structures are common, and the trace fossil assemblage quite varied (see above).

11.4.2.

Weak currents – mud-dominated contourites

Fine-grained particles and organic matter are deposited by weak bottom currents. Thin sand layers may form when currents temporarily accelerate. Within such organic-rich muds, oxygen consumption by benthic animals and bacteria may lead to anoxic pore waters at shallow sediment depths. The small size of individuals and the low biodiversity (including monospecific populations) characterize the ichnofauna (Fu and Werner, 1994; Figures 11.5b, 11.6a). Chondrites trace fossils often dominate this environment at the sea floor, and are used as an indicator of low

A. Wetzel et al.

199

oxygenation within the sediments (Ekdale et al., 1984). Such fine-grained contourites are, however, very difficult to recognize in outcrops without knowing the environmental context from independent evidence. In many cases, colour banding occurs. Brownish sediments pass downwards to greenish muds. First sediments are rapidly deposited by slow currents and they are often greenish. Subsequently, when currents accelerate and prevent sedimentation organic matter is oxidized from the surface downward (<2 cm ky1; Jung et al., 1997; Figure 11.9). Faster currents may also lead to thin silty/sandy horizons and/or to a surface of a minor hiatus. A distinct ichnocoenosis should occur in such environments, theoretically dominated by U-shaped and vertical burrows (Arenicolites, Teichichnus, Skolithos) accompanied by those related to surface utilization, including Thalassinoides and Zoophycos. In addition, Planolites, Phycosiphon and Chondrites may occur. Nereites may form along the redox boundary (Wetzel, 2002). Alternation between reworking and deposition leads to relatively low average sedimentation rates in muddy contourites, which therefore become prone to complete bioturbation (Wetzel, 1991). When bottom currents supply additional food, the fauna becomes more abundant, and burrowing rates become higher than at reference sites without significant bottom-current influence. However, at present there is no unequivocal ichnological criterion that documents the endobenthic response to fluctuating food levels (Wetzel, 1991; Wetzel and Uchman, 1998a, b), because the organic matter content within the sediment is controlled mainly by the sedimentation rate (Mu¨ller and Suess, 1979; Mangini et al., 2001). The effects of short-term fluctuations in sedimentation on the final ichnofabrics are not yet known.

Figure 11.9 Muds deposited by a bottom current (fresh core); slowly accumulated brownish mud (bm) rests on rapidly deposited greenish mud (gm). Brownish mud has been transferred by burrowing into the greenish interval. The change in sedimentation rate probably results from changing bottom-current activity. Institute of Geosciences Kiel, Germany, core 16820 -2 (336^348 cm), southern side of the Iceland ^ Faeroe Ridge, taken at the NWslope of a channel parallel to the ridge crest (62°35.560 N, 14°18.620 W, 1649 m water depth).

200

Bioturbation and Biogenic Sedimentary Structures in Contourites

Table 11.2 Bioturbation in hemipelagites, muddy turbidites and muddy contourites

Sediment accumulation Organic matter supply (long-term) Organic matter imported by currents Fauna

Faunal abundance Faunal structure

Hemipelagite

Muddy turbidite

Muddy contourite

More or less continuous, slow Mainly vertical flux, dependent on water depth None

Instantaneous, very rapid

Intermittent deposition and reworking Vertical flux supplemented by lateral input Refractory and non-refractory

Continuously living

Partially killed by deposition of thick turbidite

Average, dependent on organic matter flux Several tiers

Varying (partially killed, re-colonization)

Suspension-carried material and vertical flux Mainly refractory

Surviving deep burrowers, sequential colonizers

Specialists

Each tier with special adaptation

Pre- and postdepositional burrowers

Bioturbation

Continuous

Post-depositional

Bioturbation rate

Dependent on organic matter flux (average) Complete reworking normal

Normally low, dependent on local conditions Complete reworking rare

Degree of bioturbation Basal layer

Often not completely bioturbated, exploited by deep burrowers

Partially removed by currents, deep-burrowing survivors (Very) high, due to additional food supply Permanent living, adapted burrowers; fauna import by currents Burrowing surface utilizers; lowoxygen pore water Continuous, synand postdepositional Very high Complete reworking normal Complete bioturbation common

Fine-grained contourites (muddy and silty facies) are, in some cases, mis-interpreted as silt/mud turbidites, although they are actually much more similar to hemipelagites in their general appearance. As for their coarse-grained counterparts, our knowledge is currently insufficient to produce a definitive ichnofacies association and tiering structure for muddy contourites, but certain characteristics are nevertheless distinctive (Table 11.2). Bioturbation is strong and continuously leading to complete reworking of the sediment, with a generally varied trace-fossil assemblage (see above). The rate of bioturbation is typically high, but does vary with

201

A. Wetzel et al.

fluctuations in sedimentation rate and supply of organic matter. Periodically higher bottom-current velocities may lead to a silty or sandy horizon and/or to a nondeposition horizon, together with associated vertical and U-shaped burrows.

11.5.

P ERSPECTIVE

Fluctuations of bottom-current intensity affect the burrowing fauna considerably (Figure 11.10). Bioturbation in such settings is fuelled by organic matter that is Effects of bottom currents on burrowing fauna Hypothetical steady-state conditions Energy

High

Intermediate

Low

Current velocity Sedimentary regime Substrate Laterally advected POM

Erosional Hiatus horizon

By-pass winnowing

Depositional

Sandy

Muddy

Suspended Deposited

Microbial activity Current effects

Removal of POM, juveniles

Nutritional strategy

Filter feeding Interface feeding (deposit feeding)

Import of juveniles, larvae, POM Filter feeding Interface feeding Deposit feeding

Temporary Interface feeding Deposit feeding

Fluctuating bottom curent velocity • erosional and depositional phases alternate • on-average low sedimentation rates, but short-term high • high benthic food level • various nutritional strategies • high bioturbation rates common: bioturbation rate > sedimentation rate - impure lithologies prevail - deep-penetrating burrows dominate the ichnofabric documenting short-term favourable conditions seldom and/or local: sedimentation rate >> bioturbation rate - multiple reworking by high-speed currents that deposit coarse, fraction-transported material -primary sedimentary structures dominate

Figure 11.10 Theoretical compilation of how contour currents affect the burrowing fauna. POM = particulate organic matter.

202

Bioturbation and Biogenic Sedimentary Structures in Contourites

re-suspended and/or laterally advected, in addition to that directly provided by the hemipelagic rain. This high-trophic mode of bioturbation results in an enhanced rate of burrow production, distinct depth of bioturbation and a specialized behaviour. In particular, observations on the modern sea floor suggest that passively ventilated burrows or traces are produced at high abundances by animals that collect or trap food particles on the surface while living within the sediment. The geometry and taxonomy of such burrows and their position within the bioturbated zone is not really known as yet, although it is clear that Thalassinoides commonly belongs to the contourite ichnofacies. Although a clear set of ichnological criteria recording the contourite mode of bioturbation is not yet available, we can identify certain bioturbational characteristics for both sandy and muddy contourite facies that will help with their distinction from related turbidite and hemipelagites facies in deep-water environments (Tables 11.1 and 11.2). Without doubt, further investigation of bioturbation in contourites is needed, especially using high-resolution X-ray radiographs.

ACKNOWLEDGEMENTS M. Akhurt (Edinburgh, GB) provided access to BGS cores. L. Lo¨wemark (Stockholm) and T. van Loon (Poznan) critically read the manuscript and made helpful suggestions. S. Lauer and A. Reisdorf (Basel) helped with figure preparation. All these contributions are gratefully acknowledged.

C H A P T E R

1 2

S OME A SPECTS OF D IAGENESIS IN C ONTOURITES P. Giresse Laboratoire d’E´tudes des Ge´o-Environnements Marins, Universite´ de Perpignan, Perpignan, France

Contents 12.1. Introduction 12.2. Some Examples of Sedimentary Rhythms Associated with Changes in Colour 12.3. Diagenesis of Fine-Grained Contourites 12.3.1. Manganese oxidation processes 12.3.2. Iron oxidation processes 12.3.3. Carbonate diagenesis 12.4. Diagenesis of Coarse-Grained Contourites 12.4.1. Glauconite growth Acknowledgements

12.1.

203 204 207 208 210 212 212 213 221

INTRODUCTION

Colour changes in sediment are caused by a change in the relative abundance of different sedimentary phases. For example, changes in abundance of CaCO3 affect the lightness of the sediment colour. Rhythmic variation in the lithology of deep-water sediment is frequently recorded as alternating colours and cycles in the spectrophotometric lightness factor, bulk density, magnetic susceptibility and other physical parameters determined either in the laboratory or downhole (Yokokawa, 2001). Such alternations are noted within both hemipelagic and contouritic episodes, and even within turbiditic episodes. The cyclic variations are controlled by physical mechanisms as well as by geochemical processes. Various examples of alternating laminae of light minerals and mica within a stratum and/or concentrations of heavy minerals along lamination surfaces or within laminae have been interpreted in terms of physical processes at the sea f loor. These deposits reflect increased influence of tractive transport and typically characterise environments affected by bottom currents (Stanley, 1993). Some laminae formed by heavy minerals suggest placer concentrations and winnowing of fines. Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00212-4

Ó 2008 Elsevier B.V. All rights reserved.

203

204

Some Aspects of Diagenesis in Contourites

Rhythmic colour changes are known to occur within both hemipelagic and contouritic sediments, and even within turbidites. However, in contrast with contourites, which are characterised by low sediment-accumulation rates and often by winnowing, turbidites are not favourable sediments for the development of diagenetic processes. Winnowing is the cause of the long-lasting ion exchange and mineralogical evolution at the water/sediment interface. In this chapter, an attempt is made to distinguish between (1) occurrences of rhythmic variations that can be linked to primary sedimentary processes, such as turbidite, contourite and hemipelagite deposition, (2) diagenetic processes associated with fine-grained contourites and (3) diagenetic processes associated with coarse-grained contourites.

12.2.

SOME EXAMPLES OF SEDIMENTARY R HYTHMS ASSOCIATED WITH C HANGES IN C OLOUR

Samples from the Blake–Bahamas Outer Ridge and the Sohm Abyssal Plain collected as part of the Ocean Drilling Program (ODP) Leg 172 (Keigwin et al., 2001) contain distinct sedimentary structures derived from episodic events such as turbidity currents and benthic storms (Yokokawa, 2001). In sediments of ODP Site 1062, sharp colour changes occur, where colour is lighter as the sediment becomes coarser and as the nannofossil (CaCO3) content increases (Figure 12.1). Upward gradation from light to dark colours indicates fining upwards. It is suggested that these changes in colour may be accompanied to some degree by intense bottom-water flow within the North Atlantic Western Boundary Currents in these areas. Late Miocene and younger sediment drift deposits drilled offshore of Prydz Bay (East Antarctica) (O’Brien et al., 2001) show colour alternations within both hemipelagites and contourites. Darker coloured units contain a higher terrigenous fraction, whereas lighter coloured beds have a higher biogenic component. At ODP Site 1167, drilled mid-slope at the trough mouth, diffuse reddish-brown colour bands are widespread within several thin intervals of hemipelagic Holocene deposits. The colours at Site 1165, drilled in the Wild Drift, on the continental rise, reflect the sediment source and sediment distribution by oceanic currents. Darker units generally show a higher bulk density, a higher magnetic susceptibility, and a higher Corg content. The lightness of the green and grey sediment colour may be related to the oxidation state of iron Fe (II) and Corg contents. The greenish facies are structureless diatombearing clays with a higher biogenic content. Generally, greenish facies have lower values of bulk density, magnetic susceptibility, and Corg than dark grey facies (Shipboard Scientific Party, 2001a). The deeper deposits at Site 1165 are darker and harder with local diagenetic cementation (e.g., opaline silica and authigenic carbonate) (Figure 12.2). Diagenesis of silica is documented at Site 1165 by the transition from sediments with diatoms (opal-A) to sediments with silicified horizons (inferred opal-CT and/or cherts) (O’Brien et al., 2001).

205

P. Giresse

(a)

(b)

Figure 12.1 The 115^125 cm interval containing thin carbonate turbidites (intervals i ^ v) at ODP Site 1062 (after Yokokawa, 2001): (a) close-up and (b) X-radiograph. A multicolour version of this figure is on the enclosed CD-ROM.

Glacial/interglacial cycles of the Late Pleistocene were analysed on the continental rise of the Pacific Margin of the Antarctic (Sagnotti et al., 2001). On the gentle slope of a sediment drift, brown muds were deposited during climatic conditions of minimum continental ice extent. The lack of terrigenous supply favoured a reduced accumulation rate. In contrast, grey muds accumulated under conditions of considerable supply of terrigenous sediment from an ice sheet front that was at or near the shelf edge during glacials. Sediments re-suspended by turbidity current were deposited under the influence of weak contour-following currents. Magnetic signals reflect a disproportion in the mixture of detrital origin and pyrrhotite, an authigenic product along the pathway of pyrite formation. The low amount of detrital organic matter during glacial times limited the reaction leading to the formation of pyrite and allowed the persistence of pyrrhotite in the sediments. During interglacials, however, the rise in oceanic productivity provided large amounts of reactive organic matter and induced a higher production of pyrite at the expense of intermediate

206

Some Aspects of Diagenesis in Contourites

(a)

Figure 12.2 Cyclic variations in lithology at ODP Site 1165 (from Shipboard Scientific Party, 2001a). (a) Minor silt laminae and stringers and strong bioturbation in dark diatom-bearing clay in unit II. (b) One of the locally occurring 1-cm-thick stiff green clays beds. (c) Dark-grey thinly bedded fissile claystones with planar lamination and cross-stratification, fine green laminae with chert layers and locally calcareous intervals. A multicolour version of this figure is on the enclosed CD-ROM.

sulphides. Pyrrhotite is also rare in the dark grey glacial beds that correlate to the Heinrich events, because the reduced input of organic matter during the deposition of these beds inhibited sulphide formation. The magnetic signal was therefore interpreted in terms of changes in organic-matter input that controlled iron/sulphur diagenesis. These changes were linked to the past sea-ice extent, induced by cooling events.

207

P. Giresse

(b)

Figure 12.2

12.3.

(c)

(Continued)

D IAGENESIS OF FINE-G RAINED C ONTOURITES

The mechanisms of element redistribution along a redox gradient may only be active under certain conditions (Sighinolfi and Tateo, 1998). These conditions are frequently found together in fine-grained contourites, indicating (1) that

208

Some Aspects of Diagenesis in Contourites

sufficient organic carbon is available to act as a reducing agent driving a redox reaction (the fine-grained deep-water sediment is normally Corg-rich); (2) that sufficient time is available to allow relocation to take place (a low sedimentation rate and/or erosional hiatuses are more common in muddy contourites than in gravity-induced deposits) and (3) that the redox-sensitive elements – especially Mn(II) and Fe(II) – are present in a form which allows their transport to major redox boundaries (Raiswell et al., 1994).

12.3.1.

Manganese oxidation processes

On the slope of the northwestern Weddel Sea, the concentrations of dropstones observed in sea-floor photographs indicate a combination of localised ice-rafted sediment input and subsequent winnowing, which concentrates the clasts at the sea floor (Howe et al., 2004). The presence of Mn-covered dropstones implies that Mn is carried either by an upward flow from the sedimentary column or by a lateral flow of oxygenated bottom water. A higher sediment supply would rapidly bury the clasts and no Mn would precipitate. The trace elements Mn and U are the product of redox diagenesis and are indicators of the degree of organic-matter degradation leading to reducing conditions within the sediment. Mn(II) is soluble under anoxic conditions and precipitates as Mn(IV) oxyhydroxides under oxidising conditions. Values of the Mn/Li ratio between 15 and 20 in cores from the continental slope and abyssal plain of the South Sandwich Trench (Figure 12.3a) indicate that the sediments are suboxic, with little Mn oxyhydroxide in the sediments. The Mn redoxcline is present at only 10 cm core depth as a consequence of a lithological variation between 16 and 30 cm from the glacial hemipelagite (above) and the lower contourite (below). In sediments that are poor in Corg, the high Mn/Li (70–140) and U/Li ratios show a strong correlation of Mn oxyhydroxides. U(VI) is soluble under oxic conditions, but precipitates as a solid particulate U(IV) under reducing conditions. In the organic-poor sediment of stations 142 (Figure 12.3b) of the forearc region, high Mn/Li and U/Li ratios indicate a strong concentration of Mn. High Ba/Li ratios suggest a limited primary productivity. In the interplain channel of the Ulleung Basin (East Sea), erosion by turbidity currents (truncated reflectors in channel walls) coexists with the deposition of hemipelagites, low-sedimentation-rate muddy contourites and manganiferous contourites (Lee et al., 2004). Laminated Mn carbonate mud (facies CaM) layers most likely reflect the abrupt oxygenation of bottom water during the Last Glacial Maximum (LGM), implying a significant reduction of the sedimentation rate. The Late Quaternary deposits on the Sao Thome deep-sea channel–levee system of the South Brazilian Basin provide evidence of contour-current activity only by sediment wave fields and manganiferous-rich layers, truncated successions and foraminiferous sand layers (Gonthier et al., 2003). Mn-rich sediments are muddy and occur both at the top of turbidite sequences and interbedded within hemipelagic–pelagic deposits. They form a 310-cm-thick layer with typical darkish to brownish millimetre-thick laminations with black Mn oxides nodules alternating

P. Giresse

209

Figure 12.3 Depth profiles of Mn (a) and U (b) ratios normalised to Li (from Howe et al., 2004, with permission from Elsevier). MC and BC refer to multicore (MC) and boxcore (BC) samples for each station. A multicolour version of this figure is on the enclosed CD-ROM.

with greenish-yellow hemipelagic–pelagic muds (Figure 12.4). Erosional surfaces are frequently observed inside and at the top or the base of the succession. The Corg content is less than 0.3%. Gonthier et al. (2003) conclude that the winnowing of fine-grained sediments by bottom currents during time-spans between turbiditic events is responsible for the low sedimentation rates that favour diagenesis. Manganese nodules with millimetre to centimetre diameters constitute more than 80% of the sediment surface of the Vema Channel, which connects the Argentine Basin with the Brazilian Basin. Bottom-water flow in this channel is constrained and the flow velocity is high. The nodules merge into a broad manganese pavement covering most of the sea floor. The channel-floor sediments are silts and clays with low carbonate contents and cyclic variations in silt content. The silt and clay sedimentation rate averages 2–3 cm ka1, decreasing with depth and changing progressively into carbonate-free manganiferous brownish-coloured clays (Masse´ et al., 1991; Fauge`res et al., 1993). A study of the Columbia Channel fan drift (Southern Brazilian Basin) (Masse´ et al., 1998) indicates that, on the southern flank (core KS 8823), the very homogeneous yellowish-brown muds are associated with manganese nodules. These deposits, similar to red clays, are free of turbidites and are interpreted as contouritic muds rather than pelagic clay. Occasionally, erosional surfaces indicate evidence of

210

4000 ppm

3000

2000

Mn

1000

(b) 0

4000 ppm

3000

2000

Mn

0

(a)

1000

Some Aspects of Diagenesis in Contourites

132 342 134 344 136

Core depth (cm)

346 138 348 140 350 142

Figure 12.4 Manganiferous silty ^ clayey succession in core KS 8813 in the Sao Thome deepsea channel ^ levee system (South Brazilian Basin). Significant increases in Mn concentration are located in the dark layers (after Gonthier et al., 2003). (a) Mn sequence showing erosional surfaces and very black and Mn-rich layers. (b) Mn sequences showing more diffuse black laminations. A multicolour version of this figure is on the enclosed CD-ROM.

the activity of bottom currents during deposition. In the neighbouring core (KS 8826), the colour of the sediment allows the definition of two facies: (1) yellowishbrown bioturbated muds, with frequent manganiferous laminae and (2) grey– green homogeneous muds with frequent millimetre-thick beds of micropebbles. The occurrence of micropebbles is thought to reflect short episodes of enhanced bottom currents (Figure 12.5). Similar colour alternations have been described from the contouritic muds at the northern exit of the Vema Channel where no turbidites occur.

12.3.2.

Iron oxidation processes

Iron diagenesis in the fine-grained contourite deposits has been described from North and South Atlantic continental margin settings. Viana et al. (1998b) report yellowish-brown, rarely reddish (rusty) iron-rich fine sands with planar lamination marked by an alternation of fine sand and silts, on the middle slope off Brazil, in water of 550–1200 m deep. Planolites-type bioturbation (Caddah et al., 1998) is common. This iron-rich sandy layer is generally 10 cm thick, often encrusted, and overlies sandy/silty laminated muds. This sedimentary facies resembles the bottomcurrent reworked sands described by Shanmugam et al. (1993a) from the Plio–Pleistocene of the Gulf of Mexico. The iron-rich crusts in the contourites

211

P. Giresse

G sw 4600 4800 5000 5200 m

Columbia Channel I-B

H III-D

NE

III-B

KS 8826

0

Yellowish-brown mud Yellowish-grey mud

1 2

Manganiferous laminae Yellowish-grey mud Micropebbles Green mud

3 4 5 Erosional contact Sharp contact Gradational contact

Yellowish mud with manganiferous laminae

6 7

Green mud with beds of Micropebbles Silt (foraminifera) Green mud Yellowish mud

8 9 10

Green mud with beds of micropebbles Beds of diatom debris Green mud with beds of micropebbles

11 12 m

Figure 12.5 Lithology of core KS 8826 and echofacies distribution near the Columbia Channel (after Masse¤ et al., 1998).

of the Brazilian Margin are thought to form in response to oxidation of the sea floor – between 550 and 1200 m by the flow of oxygen-rich Antarctic Intermediate Water (AAIW). The same water mass is probably responsible for the growth of deep-water coral banks (Viana, 1994). The crusts are Holocene on the middle slope, becoming older (Pleistocene) on the lower slope. In the Newfoundland Basin (North Atlantic Ocean), Turonian–Santonian reddish-brown, intensely burrowed muddy sandstones (Unit 4) are thought to have been reworked by bottom currents on a well-oxygenated sea floor (Shipboard Scientific Party, 2004). Goethite, responsible for the reddish colour, is more abundant in these muddy sandstones than in the sediments above and below. In addition, manganese minerals are found in concretions, both as manganite in one concretion and as minor manganese hydroxide. These reddish sediments correspond to a stage of exceedingly slow sedimentation of 2 m per million years or less, followed by intervals of non-deposition. The boundary with the underlying sediments (Unit 5) is an erosional unconformity that could have been created by scouring due to vigorous bottom currents. It is suggested that bottom

212

Some Aspects of Diagenesis in Contourites

currents were established for the first time in the Turonian during the first deepwater connections between the North and South Atlantic.

12.3.3.

Carbonate diagenesis

Plio–Pleistocene muddy contourites in the West Alboran Basin (Western Mediterranean) are richer in clays, quartz and dolomite (5–55%) (Comas et al., 1996). Most of the dolomite grains are detrital and are derived from eroded emerged areas. Diagenesis of the dolomite-rich levels led to the formation of smaller authigenic dolomite crystals with euhedral morphology. Circulation of interstitial fluids becoming enriched in Mg at these (probably) higher primary-porosity levels may have contributed to additional dolomitisation of any primary calcite content, which may in turn explain the notable depletion in calcite (<5%) in some intervals.

12.4.

D IAGENESIS OF C OARSE -GRAINED C ONTOURITES

The deposition of sandy and silty contourites is largely controlled by winnowing processes. Laminations in the silty contourites of the Weddel Sea represent variability in the current, producing preferential winnowing of silty coarsening-up (increasing current speeds) and fining-up (decreasing current speeds) sequences (Stow et al., 2002c). Generally, these contourites are formed at slow accumulation rates. In the Lofoten Drift, the sedimentation rate of Late Pleistocene glaciomarine deposits was 190 cm ka1. Then, during the Holocene, the only sediment input was from winnowing, falling to 9 cm ka1 (Howe et al., 2004; Laberg and Vorren, 2004). Petrological analysis of geological sections at St Croix in the Caribbean, the Niesenflysch in Switzerland and the Annot Sandstone in the French Maritime Alps sheds light on multiple process transport in deep marine settings (Stanley, 1993). A model depicting a turbidite-to-contourite continuum of stratal types is applied to these three rock units. Discontinuous bedding, including starved ripples and laminae formed by heavy-mineral concentrations suggest placer concentration and winnowing of fines. These laminae, sometimes highlighted by dark heavy minerals, are similar in appearance to strata that are entirely current reworked and to contourites. Well-sorted tephra layers with sharp basal contact are often observed in the Iceland Basin (Lacasse et al., 1998), where bottom currents have gradually built up a contourite succession over a long time. Similarly, changes in the concentration of high-density ferromagnetic minerals brought about by current winnowing are recorded from the Barra Fan (UK North Atlantic Margin) (Knutz et al., 2002a, b). Such concentrations enhanced the formation of bacterial Fe sulphides such as pyrrhotite and greigite (Mann et al., 1990).

213

P. Giresse

12.4.1.

Glauconite growth

One of the most significant consequences of these grain-winnowing concentrations concerns glaucony.1 The mineralogical evolution of glaucony needs a long period of cation exchange at the sea water/sediment interface. Generally, such conditions are met on the outer part of shelves where terrigenous supply is scarce or absent, but winnowing action by contour currents may also induce the same effect of glauconite growth. Below we review two important case studies. 12.4.1.1. Western Falkland Trough Late Pliocene and Middle Pleistocene glaucony-rich sandy contourites have been described from water depths of 1200–3200 m and from a variety of settings within the central and Western Falkland Trough (Howe et al., 1997). These contourites clearly indicate changes in the intensity of the Antarctic Circumpolar Current (ACC). The glaucony-rich greenish-black layers range in thickness from 10 to 50 cm. Contacts with sediments above and below are bioturbated or sharp. Contour currents reworked the sediment and winnowed the fine fraction, and bioturbation subsequently homogenised the deposits. The glaucony grains are wellrounded and well-sorted and may be derived from a combination of contemporary authigenic formation on the slope and erosion of locally outcropping Cretaceous and Tertiary strata. This last assumption is supported by dinoflagellate analyses indicating abundant reworked Cretaceous and Tertiary cysts throughout the sediments. Two types of grains have been identified: (1) light green or evolved grains with a rosette internal structure and a maximum K2O content of 6.5% and (2) dark green or highly evolved grains with a lamellar internal structure and a maximum K2O content of > 8%. Howe et al. (1997) took into account previous works (Odin and Matter, 1981) and considered that a typical time for the formation of the light green evolved grain is 0.1 Ma, whereas darker, highly evolved grains are at least 1 Ma old. In our opinion, these are overestimations of the minimum age of the grains. A K2O content of 6.5% is commonly reached after 10,000–20,000 years, and higher amounts such as 8% are not really a matter of time. Rather, further enrichment in K2O is linked to late-diagenetic conditions induced by increased burying and pressure (Giresse et al., 1988; Wiewio´ra et al., 1996). Glaucony is present within the Holocene sediments across all core sites in the Western Falkland Trough, although in smaller quantities than in the Pleistocene sandy contourites. This suggests that the degree of winnowing and in situ erosion during the Holocene was less than during deposition of the Pliocene–Pleistocene condensed sandy contourites. Consequently, the ACC flows were less vigorous than during the Pliocene–Pleistocene.

1

The term glaucony is used in this chapter to designate the grains and f ilms formed by the process of glauconite formation. It is not a mineral name, but rather a morphological or facies term. The term glauconite is used to identify the potassium-rich green clay mineral found in the glaucony facies.

214

Some Aspects of Diagenesis in Contourites

12.4.1.2. Coˆte d’Ivoire–Ghana Marginal Ridge ODP Site 959 is located on a small plateau on the northern flank of the Coˆte d’Ivoire–Ghana Ridge, at a depth of 2100 m (Figure 12.6a). Because of its topographic position, this site was isolated from turbidity currents or debris flows originating from the Ghanaian Slope. The upper 20 m of the hole comprises dark green foraminiferal ooze and nannofossil ooze variably mixed in distinct sediment beds. Microstructural analysis reveals that the sediments were formed by 1-mm- to 1-cm-thick laminae with relatively high concentrations of mainly planktonic foraminifers. Foraminifer-rich laminae are lighter in colour and contain green pigments in clay filling the chambers of pelagic foraminifers. Only rarely are the green minerals found as grains. This material is devoid of any littoral evidence and cannot be interpreted as transported from the outer shelf to the bottom of the slope. The very low sedimentation rate (varying from 1 to 2 cm ka1) reflects the persistent activity of forced deep-water circulation. Giresse et al. (1998) were able to calculate that the dominant period of recurrence for successive laminae is between 1000 and 2000 years. The water depth at ODP Site 959 corresponds roughly to the interface between southward-flowing North Atlantic Deep Water (NADW) and northward-flowing South Atlantic Intermediate Water (SAIW) (Figure 12.6b) (Sarnthein et al., 1982). Based on the oxygen-isotope record, two main stratigraphic gaps are present, viz from 1.6 to 0.99 Ma and from

Figure 12.6 Topography and hydrology of ODP Site 959 (Giresse et al., 1998). (a) Locations of Site 959 drilled during leg 159 (contours in metres). (b) Scheme of interference circulation between North Atlantic Deep Water (NADW) and South Atlantic Intermediate Water (SAIW). According to Sarnthein et al. (1982).

P. Giresse

215

0.20 to 0.12 Ma. Each hiatus was preceded by a decreasing accumulation rate: 0.61 cm ka1 at 11.9 m below the sea floor (mbsf) and 0.13 cm ka1 at 1.5 mbsf, respectively. In conclusion, significant changes in sediment winnowing by bottom currents have determined the deposition of laminated contourites with increased concentration of glauconitic infillings of foraminifers. The onset of contouritic deposition coincides with a climatic deterioration at about 1.3 Ma, and intensification of the cold-water circulation pattern at about 0.9 Ma (Giresse et al., 1998). 12.4.1.3. Green grain concentrations Studies of the green grain concentration compared to the bulk sand fraction suggest that glaucony abundance is not constant through time (Figure 12.7b) (Giresse and

Figure 12.7 Age/depth and weight percentages of green grains (for sand fraction) in the upper 21.4 mbsf at ODP site 959 (after Giresse and Wiewio¤ra, 2001). (a) Stable-oxygen isotope record and position of interglacial stages (striped area) on the left. (b) Peaks of green grain concentration represent the end of a winnowing interval between two active deposition intervals. (c) Green grains/white grains ratio. (d) Relative abundance of dark grains conventionally calibrated from 1 to 5.

216

Some Aspects of Diagenesis in Contourites

Wiewio´ra, 2001). At depths over 12 mbsf green grains represent less than 10%, while they always exceed 10% above 12 mbsf, although their concentration is more irregular there. In particular, between 11 and 12 mbsf, which coincides with a 530 ka hiatus, the green grains reach 20–45% of the total sand fraction. Closer to the sea floor, concentrations over 10% are found, along with more irregular contents. Maxima of green grain concentration correlate with maxima of the relative abundance of cracked dark green grains (Figure 12.7b–d). In contrast, minima of green grain abundance characterise sediments with paler green glauconitic infillings in calcareous foraminifers. Assuming that the superficial redistribution of the sediment was favourable for longer contacts with sea water and to glauconite mineralisation, the variable concentration of green grains after the hiatus suggests that the restoration of sediment accumulation varied with time. Another observation is that the sediments richest in green pigments are usually very poor in pyrite. Pyrite moulds and framboids are present in the sediments with the lowest concentration of green pigments. In the early Pleistocene strata, framboids can even be abundant. Framboid crystallisation took place later than the pale green infilling and occurred in the chambers of calcareous foraminifers. It is suggested that pyrite began to form when easily reduced iron species were subjected to sulphidisation with increasing burial depth. It follows that the pyrite-rich layers are the result of relatively high sedimentation rates. Similarly, in the Oligo–Miocene of southeastern Australia, pyrite occurs rimming and in partial replacing glauconitic pellets, postdating the development of glaucony (Kelly and Webb, 1999). 12.4.1.4. Genesis of the green grains The mineral assemblage of the mud filling the foraminifer chambers is kaolinite (45–50%), smectites (35–40%), illite (10–15%), and traces of mica, calcite and quartz. This was a precursor material. X-ray diffraction (XRD) analyses were carried out on the bulk green grains representative of the average composition of a heterogeneous grain population with white and dark green as end members in five reference levels (11.14, 11.04, 10.94, 10.74 and 10.44 mbsf) (Figure 12.8a–c). After saturation of the samples with ethylene glycol, the 00l peak shows lattice expansion typical of smectites, but other orders do not present integral 00l series. Figure 12.8d indicates that d values for second-order reflection vary from 8.718 A˚ ˚ for level 10.94 to 9.0 A ˚ for level 11.14 mbsf. This for level 10.44, via 8.78 A ˚ indicates a growing number of the closed (10 A) layers in the structure of the smectite. Along with dissolution of kaolinite, crystallochemical modifications of smectites were observed both on the sequential concentration scale (Figure 12.8d) and on the individual grain scale (from white to dark green). The growing of the closed layers can be observed as the maturation process proceeds (mixed smectite– glauconite layer 80/20). In the green grain-rich level (9.54 mbsf), a Fe3þ-rich montmorillonite, which may have more than 50% expandable layers, corresponds to the ‘‘montmorillonite way’’ of Amouric (1990). In the other depth cores, the K-rich nontronite genesis process prevails and resembles the ‘‘nontronite way’’ previously described from various shelves of the Gulf of Guinea (Giresse et al., ˚ peak from the 22.4 mbsf is firm evidence of the appearance of a 1988). The 10 A

217

P. Giresse

10.4

(a)

10.44 m

(b)

(c)

10.6 10.74 m 10.94 m

11

11.04 m 11–14 m

Greens/whites

Green grains %

30

40

0

1

2

3

4

Dark green

50

1

2

17.14

20

3

4

–7.199

K

–3.343

K

Q

Q –3.574

–4.519

(d)

–8.718

16000.00

10

10.44 mbsf

8.78

counts

11.2 Depth 0 (mbsf)

–4.257

10.8

10.74 mbsf

0.00

9.0

10.94 mbsf

6

8

10

12

11.04 mbsf 11.14 mbsf 14

16

18

20

22

24

26

28

30 2°θ Cok α

Figure 12.8 Green grain vertical distribution between 11.2 and 10.4 mbsf (after Giresse and Wiewio¤ra, 2001). (a) Green grain concentration, (b) green/white grains ratio, (c) dark green (cracked) grain relative abundance and (d) expanded XRD patterns of the ethyleneglycolated samples from a series of sediment at various levels (mbsf ): 10.44, 10.74, 11.04, 11.14; Q = quartz, K = kaolinite.

discrete although poorly organised glauconite phase among the more mature green grains from the deeply buried sedimentary layer. During this development, the substitution of aluminium by iron is the most significant process. The negative correlation between the two elements is more and more evident from the polymineralic whole-green grain population to the probably monomineralic dark green grain population end-member of the evolution. The negative correlation between Al and K is medium and might correspond to the significant entry of potassium, developed during the last steps of the evolution. Carbonate remnants are either dissociated when acidic conditions develop by the release of hydrogen ions during the fixation of K (Lundegard and Land, 1986) or broken down by the pressure initiated by crystal growth. A positive correlation between Si–Mg and Si–Al is generally evidenced. Si–Mg is regarded as a signal of the substitution of inherited smectite (low-charge beidellite) by authigenic montmorillonite, whereas Si–Al more generally emphasises the change from the polymineralic to the monomineralic assemblage.

218

Some Aspects of Diagenesis in Contourites

In each nascent white infilling, characteristic crystal growth with nannostructures of thermal origin like authigenic smectite are confined to restricted voids. The rapid first steps of evolution are believed to involve solution of the precursor mineral, allowing its cations to become part of the newly formed mineral. In medium green to dark green grains, maturation is emphasised by the markedly greater size and density of the authigenic flake assemblages. High concentrations of green grains (10–40%) in the sand fraction usually occur at water depths of over 2000 m, where the scarce detrital input and more importantly winnowing, lead to slow burial. One of the questions in the glauconitisation process is the time needed for ion exchange at the water/sediment interface. These exchanges take place between the overlying fluids and a granular substrate acting as a semi-confined microenvironment. As stated before, both the outer-shelf environment, with low terrestrial supply and turbulent water, and any other deep-sea environment characterised by sediment winnowing by bottom currents, provide favourable conditions for long-lasting cation exchange and mineralogical evolution at the water/sediment interface. Constantly moving water, which replenishes the ions needed for cation exchange, is another factor in favour of glauconitisation in these environments. According to this hypothesis, the fluctuations in green grain concentration probably correspond to microsequences linked to the energy pulses of the bottom water. Generally, the identification of the evolutionary stage of glaucony allows the duration of non-deposition before burial to be estimated (Figure 12.9). It is

Figure 12.9 Step by step condensing process of the green grains accumulation (after Giresse and Wiewio¤ ra, 2001). Bottom-current-controlled green grains deposition and evolution, based on mud-matrix reworking, green infilllings mechanical concentration and mineralogical maturation reflecting the time of residence of the infillings at the sea bottom before burial (empty disks: white grains; stripped disks: medium green grain, black disks: dark green grains). Foraminifer tests were gradually removed or destroyed from 1 to 3.

P. Giresse

219

suggested that initially (pristine deposit), empty shells and nascent glaucony infillings are present in low concentrations. Winnowed foraminiferal sandy deposits are matrix-poor and green grain-rich with slightly evolved glaucony. With increasing time of exposure at the sea floor, green grains can undergo more pronounced modification. The content evolves and mature glaucony increases but simultaneously some new nascent glaucony can still accumulate. Finally, the true green grain layer, with the highest content of highly evolved dense grains, accumulates. The glauconitisation process in these tropical regions was especially favoured by abundant terrigenous iron and by relatively high temperatures. It is likely that bottom currents controlled the availability of iron and its diagenetic deposition in different mineral phases. High availability of iron favours the deposition of K-rich nontronite. In contrast, in the strongly winnowed 9.54 mbsf with Fe3þ-rich montmorillonite, part of the very fine-sized fraction of iron-rich material might be eliminated by repetitive winnowing. The lengthy dispersal of iron microparticles played an ultimate role in the development of the less iron-rich montmorillonite rather than the more iron-rich nontronite (Wiewio´ra et al., 2001). Finally, and of more general importance, it is still an open question whether the smectite-rich argillaceous matrix material of this site favoured the formation of Fe3þ-rich montmorillonite much more than those initiated in the common prevailing kaolinitic substrate of the Gulf of Guinea Shelf. 12.4.1.5. Blue–green grains and associated volcanic glass debris Blue–green grains occur in association with volcanic glass debris in laminated sediments at the same ODP Site 959 (Giresse and Wiewio´ra, 1999). Glass shards are rare (one piece in 1 g of sand) in some layers near 10.2 mbsf and especially at about 11.5–11.8 mbsf (Figure 12.10) coinciding with the hiatus in contourite layers with the highest concentrations of glauconite grains. Most volcanic glass belongs to the transition field of sub-alkalic rocks from basalt to trachyte. The fragments are highly vesiculated. In a total versus silica (TAS) diagram, most of the points lie in the field of alkaline to tholeiitic basalts or more siliceous rocks analysed from the Cameroonian trend (Gouhier et al., 1974). Among light to dark green grains, the most striking ones were rare blue–green grains which looked like certain celadonite grains (Seyfried et al., 1978). Blue–green grains were observed only in a few layers, between 4 and 9.5 mbsf with low concentrations (1–5 grains per gram sandy fraction), but they do not necessarily correlate with the glass shards (Figure 12.10a). In particular the blue– green grains are found distinctively later. Both glass shards and blue–green grains disappear above 4 mbsf. The mineralogical composition of the blue–green grains is as follows: (1) authigenic K, Fe–smectite associated with zeolite (like phillipsite) and different amounts of quartz and anorthite and (2) feldspathic grains dominated by albite but including quartz, volcanic glass and smectites. On the basis of their chemical composition, a genetic relationship between blue–green grains and volcanic glass is suggested. Only a few of the grains with heterogeneous composition seem to be primary ignimbrite and not the result of glass weathering. Palagonitisation, and especially phillipsite formation, may result from a relatively rapid reaction during burial diagenesis in deep-sea sediments with

220

Some Aspects of Diagenesis in Contourites

2

85,000

520,000

Figure 12.10 Plots of blue ^ green grains and accumulation rate of glauconitic grains vs Hole 959 depth (after Giresse and Wiewio¤ra, 1999). (a) Distribution of blue ^ green grains and volcanic glass (number of particles for 1 g of the sand fraction). (b) Accumulation rate of glauconitic grains. Glauconitic grain percentages were calculated after sand-sized wet sieving and 0.1 N HCl treatment.

a relatively low sedimentation rate. The most probable origin of these pyroclastic ejecta are explosive events from the Cameroon Volcanic Ridge, especially from the Sao Thome and Principe islands and the Mount Cameroon area. They were introduced into the marine environment after a 1200-km-long westward eolian

P. Giresse

221

transport and consequent fallout. Subsequent winnowing by bottom currents concentrated the glass shards on the sea floor. The sediments immediately following the non-deposition between 1.6 and 0.99 Ma (11 mbsf) contain the highest amount of coarse volcanic debris and blue– green grains. Repetitive winnowing seems the most probable explanation for the partial or total disappearance of alteration products of palagonitisation at the surface of the glass.

ACKNOWLEDGEMENTS The present contribution benefited from critical reviews by Angelo Camerlenghi, Tom van Loon and Michele Rebesco on an earlier version. I thank them especially for suggesting many improvements to the original English version. I also thank Andrzej Wiewio´ra for fruitful collaboration spread over many years of research on the ‘‘green grains’’.

C H A P T E R

1 3

C ONTOURITE F ACIES AND THE F ACIES M ODEL D.A.V. Stow1 and J.-C. Fauge`res2 1

National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, UK De´partement de Ge´ologie et Oce´anographie, Universite´ Bordeaux1, Talence cedex, France

2

Contents 13.1. Introduction 13.2. Historical Context 13.3. The Range of Contourite Facies 13.3.1. Muddy contourites 13.3.2. Silty contourites 13.3.3. Sandy contourites 13.3.4. Gravel-rich contourites and gravel-bearing contourites 13.3.5. Shale-clast or shale-chip layers 13.3.6. Volcaniclastic contourites 13.3.7. Calcareous muddy and silty contourites 13.3.8. Calcareous sandy contourites 13.3.9. Calcareous gravel-lag contourites 13.3.10. Siliceous bioclastic contourites 13.3.11. Chemogenic contourites 13.3.12. Other contourite-related facies 13.4. Contourite Facies Model and the Contourite Sequence 13.4.1. Sequence origin and interpretation 13.5. Lamination Versus Bioturbation in Contourites 13.5.1. Laminated sandy contourites 13.5.2. Laminated muddy/silty contourites 13.5.3. Mixed turbidite/contourite systems 13.6. Contourite-Related Facies Acknowledgements

224 224 225 226 226 228 230 230 231 231 232 233 233 234 234 234 237 238 239 239 239 240 250

Appendix 1 13A1.1. NW UK Continental Margin 13A1.2. Norwegian Continental Margin 13A1.3. Greenland Continental Margin 13A1.4. Gulf of Cadiz, Iberian Margin 13A1.5. Mediterranean Sea 13A1.6. East North American Continental Margin 13A1.7. East South American Continental Margin 13A1.8. East New Zealand Continental Margin, SW Pacific 13A1.9. Antarctic Continental Margin and Weddell Sea

251 251 251 252 252 253 253 253 254 255

Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00213-6

Ó 2008 Published by Elsevier B.V.

223

224

Contourite Facies and the Facies Model

13.1.

INTRODUCTION

Tremendous advances have been made in just over 40 years since contourites were first recognised and described, and their global significance in the present-day oceans began to be unravelled. Today, we have an enormous body of data documenting the nature and characteristics of contourite facies, derived from modern sedimentary systems throughout the world. These are the primary data used to construct the standard contourite facies models, as embraced by most contourite workers. Much less consensus exists with regard to the recognition and interpretation of contourites from the ancient record. In this chapter, we briefly review the historical context, and then describe in some detail the different contourite facies recognised, outline the contourite facies models and discuss the interpretation of contourite sequences. We further consider some hybrid contourite facies and summarise criteria for the distinction between contourites and associated facies in deep-water systems. The focus is very much on what we have learned from studying modern contourite systems. The problems and controversy surrounding the recognition of fossil contourites in the ancient record is covered by Hueneke and Stow (2008).

13.2.

H ISTORICAL C ONTEXT

Much of the early pioneering work on contourites was carried out along the eastern continental margin of North America, and served to document the strong influence of the deep Western Boundary Undercurrent (WBUC) in shaping the lower slope and rise (Heezen and Hollister, 1964; Heezen et al., 1966; Schneider et al., 1967). As there are no distinctive, mounded, contourite drifts along much of this margin, the contourite facies are found closely intercalated with associated deep-water sediments mainly deposited by turbidity currents, pelagic and hemipelagic processes. This led to early problems with the distinction between these different deep-water facies, as illustrated in the contrasting contourite characteristics outlined by Hollister and Heezen (1972), compared with those of Stow (1979) from the same margin. Resolution of these differences has been achieved through the study of contourites from many more drifts worldwide, and especially those of incontrovertible bottom-current construction (see summary in Stow and Lovell, 1979). Furthermore, the Nova Scotian continental rise was selected for the HEBBLE (High Energy Benthic Boundary Layer Experiment) programme of the early 1980s, which provided a long-term in situ record of sea-floor conditions, bottom-current flow and sediment characteristics directly under the path of the WBUC (Nowell and Hollister, 1985; McCave et al., 1988). Hollister (1993), and Stow and Fauge`res (1993) agreed that many of the features originally attributed to contourites (e.g. Hollister and Heezen, 1972) were, in fact, those of bottom-current-reworked turbidites and/or fine-grained turbidites. As many more descriptions of modern contourites emerged in the literature, more reliable facies models were developed for both muddy and sandy contourites

D.A.V. Stow and J.-C. Fauge`res

225

(Stow and Lovell, 1979; Stow, 1982). Following a scientific cruise on the Faro Drift in the Gulf of Cadiz, linking good core recovery with bottom photographs and high-resolution seismic records, the separate facies models were combined into the now standard contourite sequence model (Fauge`res et al., 1984; Gonthier et al., 1984; Stow and Piper, 1984). This has been the basis for most subsequent contourite work, at least on modern systems, with documentation of partial sequences (e.g. Howe et al., 1994; Stoker et al., 1998a) and few further changes (Stow et al., 2002b; Stow, 2005). Some further refinement to the model is presented here, but we must also note that there continues to be some dissent, especially from Shanmugam (2000, 2006a) and Shanmugam et al. (1993a, 1995).

13.3.

T HE R ANGE OF C ONTOURITE F ACIES

The most widely accepted definition of contourites is both simple and flexible (Stow et al., 2002b): ‘‘contourites are the sediments deposited by or significantly affected by the action of bottom currents’’. As bottom currents are semi-permanent currents that act in deep water, they naturally interact with other processes, including turbidity currents, hyperpycnal flows, pelagic and hemipelagic settling. They incorporate fine-grained material from these processes, which they may transport for long distances before its ultimate deposition. They are also capable of winnowing and eroding the sea floor, and of preventing deposition, thereby causing hiatuses and/or hardgrounds in the sediment record. A wide range of different facies fall within the spectrum of contourites now recognised in deep-water deposits, and the database used here in summarising the characteristics of these contourite facies is a very extensive one. We have drawn on data published in the following: 1. early work as synthesised in Stow and Lovell (1979), and collated by Nowell and Hollister (1985), and McCave et al. (1988); 2. collected works published in the 1990s by Stow and Fauge`res (1993, 1998), Gao et al. (1998), Mienert (1998), Stoker et al. (1998b), and Maldonado and Nelson (1999a); 3. most recent compilations by Rebesco and Stow (2001), Stow et al. (2002f ), Wynn and Stow (2002a), and Viana and Rebesco (2007). In addition to these sources, we draw from a big synthesis of contourite drifts cored during some 50 DSDP (Deep Sea Drilling Program) and ODP (Ocean Drilling Program) legs at over 100 deep-water sites (see Table 1 in Stow et al., 1998a), and have compiled data on contourites from a number of the more recent and scattered publications, which are documented in Appendix 1 of this chapter. This represents detailed information from more than 200 piston and gravity cores on some 30 different drifts. The range of contourite facies now recognised and well documented include the facies groups listed in Table 13.1. A full description is given for each of the individual facies, with illustrations where possible, except for the volcaniclastic contourites. In most regions well supplied with volcaniclastic material, this simply

226

Contourite Facies and the Facies Model

Table 13.1 Range of different contourite facies Siliciclastic contourites Muddy, silty, sandy and gravel-rich variations Shale-clast/shale-chip contourites All compositions possible Volcaniclastic contourites Muddy, silty, sandy and gravel-rich variations Calcareous bioclastic contourites Muddy, silty, sandy and gravel-rich variations, also known as calcilutite, calcisiltite, calcarenite and calcirudite contourites in fossil contourite systems Siliceous bioclastic contourites Mainly sand grade recognised Chemogenic contourites (within mud or calcilutite) Include manganiferous layers, nodules, pavements Other contourite-related facies ‘‘Shallow-water’’ contourites, reworked turbidites

forms a component of either siliciclastic or bioclastic contourites. Where it is the dominant component, the facies are analogous with those of their siliciclastic equivalents. Contourite-related facies are discussed in the final section.

13.3.1.

Muddy contourites

Muddy contourites (Figures 13.1 and 13.2) are homogeneous and they commonly appear as thick featureless units. They are poorly bedded, in some cases showing centimetre- to decimetre-scale banding marked by subtle colour changes, and also noted during core logging as slight compositional variations. They are generally highly bioturbated, often with an indistinct mottled appearance, and may further show distinct burrows of varied (deep water) ichnofacies (see Wetzel et al., 2008). There may be rare primary lamination present (partly destroyed by bioturbation), diffuse and indistinct, in places marked by colour change and in places by irregular winnowed concentrations of coarser material. They have a silty–clay grain size (typical range 5–111) and poor sorting (typically 1.4–2 ). The composition is dominantly siliciclastic, but commonly mixed with some biogenic fraction. The components are in part local, including a pelagic contribution, and in part far-travelled.

13.3.2.

Silty contourites

Silty contourites (also referred to as mottled silty–muddy contourites) (Figures 13.1 and 13.2) are similar in many ways to the muddy contourites but have a larger silt-sized component and, therefore, greater potential for revealing some internal structure. They are often gradationally interbedded with both muddy and sandy contourite facies. 1 Sorting is a measure of the standard deviation of grain size distribution about the mean size, and is generally expressed in  units, where  = log2 (grain size).

227

D.A.V. Stow and J.-C. Fauge`res

(a)

(b)

(c)

Figure 13.1 Muddy and silty contourite facies. Rockall Trough, NW UK Continental Slope. Core width 8 cm. (a) Silts and muds interlayered. (b) Mottled silt horizon within structureless muds. (c) Bioturbated muds with very indistinct lamination apparent in parts. A multicolour version of this figure is on the enclosed CD-ROM.

Figure 13.2 Silty and muddy contourite facies, showing homogeneous, bioturbated muds, mottled silts and indistinctly laminated silty mud facies. Gulf of Cadiz contourite depositional system. Core width 8 cm. A multicolour version of this figure is on the enclosed CD-ROM.

228

Contourite Facies and the Facies Model

They commonly show extensive bioturbational mottling as well as a range of more distinct burrows. There is some evidence of indistinct discontinuous lamination (partly destroyed by bioturbation), typically with sharp to irregular tops and bases of silty layers, together with thin lenses of coarser material. Rarely, remnants of thin cross-laminated beds are preserved. They have a poorly sorted clayey–sandy silt size and a mixed siliciclastic/ biogenic composition, as for muddy contourites. The range of grain sizes (3–11 ) may be still greater than for muddy contourites, so that the sorting is in some cases very poor (>2 ).

13.3.3.

Sandy contourites

Sandy contourites (Figures 13.3–13.5) occur as both thin irregular layers and as much thicker units within the finer grained facies, and may display either distinct (abrupt) or gradational bed contacts. They are generally thoroughly bioturbated throughout, and may appear massive (structureless) at first viewing, or display a range of distinct burrows. In some cases, rare primary horizontal and cross-lamination is preserved (though partially destroyed by bioturbation), together with irregular erosional contacts and coarser concentrations or lags. The mean grain size normally does not exceed fine sand (apart from coarsergrained horizons and lags), and sorting is mostly poor to moderate (0.8–2 ), in part

Figure 13.3 Bioturbated sandy contourite facies. These are poorly sorted, muddy sands, with some indication of indistinct parallel lamination. Brazilian Continental Slope (Campos Margin). Core width 5 cm (from Viana et al., 2002a; with permission from The Geological Society, London). A multicolour version of this figure is on the enclosed CD-ROM.

229

D.A.V. Stow and J.-C. Fauge`res

C4

C4

C3

C3

C2

C1

C2

Figure 13.4 Muddy, silty and sandy contourite facies, showing part of standard C1 to C5 contourite facies sequence. From base to top: C1 mud, C2 mottled silty mud, C3 muddy sand, C4 mottled silty mud and C5 (this division not shown) mud. Note white bioclastic shell debris in parts of C3, bioturbation throughout, and partly disrupted discontinuous lamination with some sharp contacts. Faro Drift, Gulf of Cadiz contourite depositional system. Core width 8 cm (from Stow et al., 2002b; Stow 2005; with permission from The Geological Society, London). A multicolour version of this figure is on the enclosed CD-ROM.

due to bioturbational mixing. Both positive and negative grading may be present. A mixed siliciclastic–biogenic composition is typical, with evidence of abrasion, fragmented bioclasts and iron-oxide staining. Laminated sandy contourites (Figure 13.5) are less common than their bioturbated counterparts and have been rarely documented, but do occur where high-energy (high-velocity) bottom currents are especially dominant and larger-scale bedforms (e.g. dunes) are evident on the sea floor. The few examples observed to date are thick to very thick-bedded and distinctly laminated. The lamination is relatively broad and diffuse, enhanced by slight colour variation, and parallel at the scale of the cores, although this may also be part of large-scale cross-bedding. Bioturbation is rare, but large sub-vertical burrows have been noted.

230

Contourite Facies and the Facies Model

Figure 13.5 Laminated sandy contourite facies, showing parallel and inclined lamination as a result of sand wave migration. Gulf of Cadiz contourite depositional system, from sand sheet in proximal scour and ribbon sector. Core width 10 cm. A multicolour version of this figure is on the enclosed CD-ROM.

The mean grain size is medium-grained sand, with moderately good sorting (0.5–0.7 ). The sediment has a mixed siliciclastic/biogenic composition, with evidence of abrasion, fragmented bioclasts and iron-oxide staining.

13.3.4.

Gravel-rich contourites and gravel-bearing contourites

Gravel-rich contourites and gravel-bearing contourites (Figure 13.6) are common in drifts at high latitudes as a result of input from ice-rafted material. Under relatively low-velocity currents, the gravel and coarse sandy material from ice rafting remains as a passive input into muddy, silty or sandy contourite deposits and is not subsequently reworked to any great extent by bottom currents. This facies is often indistinguishable from glaciomarine hemipelagites. Concentration of the coarser fraction occurs under higher-velocity currents and more extensive winnowing, yielding irregular layers and lenses of poorly to very poorly sorted (1 to >2 ), sandy gravel-lag. Similar coarse-grained concentrations and gravel pavements are locally developed in response to high-velocity bottom-current activity in shallow straits, narrow contourite moats and passageways.

13.3.5.

Shale-clast or shale-chip layers

Shale-clast or shale-chip layers (Figures 13.7c and 13.8b) can be developed in both muddy and sandy contourite facies, but have been recognised to date from relatively few locations only. They result from substrate erosion by strong bottom

D.A.V. Stow and J.-C. Fauge`res

231

Figure 13.6 Pebbly sand and gravel-lag contourite facies ^ structureless with irregular concentration of coarse-grained clasts. Faeroe ^Shetland Channel (gateway), NW UK continental margin. Core width 10 cm (from Akhurst et al., 2002; with permission from The Geological Society, London). A multicolour version of this figure is on the enclosed CD-ROM.

currents (and perhaps during benthic-storm events), under conditions where erosion has reached a firmer substrate and, in some cases, burrowing on the nondeposition surface has helped break up the semi-firm mud. The shale clasts are generally millimetric in size, and occur with long axes sub-parallel to bedding and, presumably, also sub-parallel to the current direction.

13.3.6.

Volcaniclastic contourites

These are generally identical to the siliciclastic facies described above, except that their composition is dominated by volcaniclastic material. They are not, therefore, described separately.

13.3.7.

Calcareous muddy and silty contourites

Calcareous muddy and silty contourites (Figures 13.7a, b and 13.8c), together with calcareous sandy contourites with which they are commonly interbedded, commonly occur in regions of dominant biogenic input, including open-ocean sites, beneath areas of upwelling, and down-current from a source of biogenic/bioclastic material (such as a carbonate shelf or bioherm). In most cases, bedding is indistinct, but may be enhanced by cyclic variations in composition and/or grain size. Primary sedimentary structures are poorly developed or absent, in part due to thorough bioturbation as in siliciclastic contourites, but some parallel to sub-parallel

232

Contourite Facies and the Facies Model

(a)

(b)

(c)

Figure 13.7 Bioclastic, chemogenic and mud-chip contourite facies. Vema Channel and Contourite Fan, Brazilian Basin. Core width 7 cm. (a) Lower part of thick manganiferous crust at top of section (current sea floor) overlying calcareous marl contourites (yellowish) and further interbedded black manganiferous deposits. (b) Sharp contact (erosive bottom-current event) between brownish-yellow calcareous marl contourites, with thin manganiferous horizons, and greenish muddy contourites. (c) Muddy contourites (as in b) with thin micro-brecciated horizons of mud-chip contourites (from Fauge' res et al., 2002a; with permission from The Geological Society, London). A multicolour version of this figure is on the enclosed CD-ROM.

indistinct primary lamination may be preserved. The mean grain size is most commonly silty clay to clayey (and/or sandy) silt, poorly sorted (grain-size parameters as for muddy contourites) and in some cases with a distinct sand-sized fraction representing coarser pelagic biogenic particles that have not been too fragmented during transport. The composition is typically pelagic to hemipelagic, including nannofossils and foraminifers as dominant elements, but in some cases the deposits may be largely composed of reworked shallow-water carbonate debris from off-shelf or off-reef supply. There is a variable admixture of siliciclastic or volcaniclastic material.

13.3.8.

Calcareous sandy contourites

Calcareous sandy contourites are the calcareous equivalent of sandy contourites, typically occurring in successions with both calcareous silty and muddy contourites. Contacts between these different facies may be either gradational, yielding indistinct bedding, or much sharper due to the erosive action of relatively strong bottom currents, thereby yielding more distinct bedding. In thinner beds, primary sedimentary structures may be largely obscured due to strong bioturbation, but

233

D.A.V. Stow and J.-C. Fauge`res

(a)

(b)

(c)

Figure 13.8 Bioclastic, chemogenic and muddy contourite facies.Columbia fan ^ drift system, Brazilian Basin. Core width 7 cm. (a) Yellow and black manganiferous contourites, with dark manganiferous crust (centre) and diatomaceous contourite mud (near top). (b) Greenish muddy contourite with thin disrupted layers (near centre) of shale-chip contourites. (c) Yellow and black manganiferous/calcareous contourites with irregular lamination and marbling structure. (from Fauge'res et al., 2002b; with permission from The Geological Society, London). A multicolour version of this figure is on the enclosed CD-ROM.

thin-bedded cross-laminated foraminiferal contourites are also known. Thicker beds may preserve more structures, although lenticularity, non-depositional surfaces, hardgrounds and burrowing also appear common in this facies. The mean grain size is sand, and both poorly sorted and well-sorted examples are known. These coarse-grained biogenic particles may be derived from pelagic, benthic, off-shelf and off-reef sources, and may have a variable admixture of siliciclastic, volcaniclastic and siliceous biogenic material. The bioclasts are often fragmented and iron-stained as a result of transport and oxidation.

13.3.9.

Calcareous gravel-lag contourites

Calcareous gravel-lag contourites, including those comprising calcilutite microclasts or chips derived from erosion of the substrate, and are not well known from the modern record. They have been inferred and described from ancient contourite successions (see Hu¨neke and Stow, 2008).

13.3.10.

Siliceous bioclastic contourites

Siliceous bioclastic contourites (Figures 13.8a) of mud, silt or sand grade are also rarely described from modern systems. Both muddy (siliciclastic) and calcareous

234

Contourite Facies and the Facies Model

(bioclastic) contourites may be relatively rich in diatomaceous and radiolarian material, particularly at higher latitudes, but are rarely dominated by siliceous bioclastic material. Cross-laminated radiolarian-rich contourite sands are known only from ancient contourite successions.

13.3.11.

Chemogenic contourites

Chemogenic contourites (Figures 13.7a, b and 13.8a, c) are those in which chemical precipitation directly from sea water occurs in association with contourite deposition and/or erosion and hiatus surfaces. 13.3.11.1. Manganiferous contourites Manganiferous contourites are those in which manganiferous or ferro-manganiferous horizons are common. This metal enrichment may occur as very fine dispersed particles, as a coating on individual particles of the background sediment, as fine encrusted horizons or laminae, or as micronodules. These features have been observed in both muddy/silty and calcilutite/calcisiltite ancient contourites from several drifts, and can also occur in association with non-deposition surfaces, hardgrounds and hiatuses. Bioturbation and burrowing are particularly evident in such cases, forming a tiered ichnofacies assemblage in the sediment below the non-deposition surface. Extensive areas of sea-floor with larger ferro-manganese nodules and pavements are well known, but the role of bottom currents in their formation is still controversial. 13.3.11.2. Chemogenic gravel-lag contourites Where deep-water chemoherms (chemical–biogenic precipitates) of metal– carbonate chimneys, mounds and encrustations occur in the path of bottom currents, the sea floor, particularly in contourite channels, is seen to be strewn with the fallen and/or eroded debris of chemoherm material. In places, this has clearly been winnowed and aligned into chemogenic gravel-lag contourites. However, the role of bottom currents in the original growth and development of such chemoherms is unknown.

13.3.12.

Other contourite-related facies

Other contourite-related facies, including modified hemipelagites, reworked turbidites and ‘‘shallow-water contourites’’ are considered in the ‘‘Discussion’’ section below.

13.4.

C ONTOURITE FACIES MODEL AND THE C ONTOURITE SEQUENCE

The facies models for both muddy and sandy contourites were originally based on data from many examples of contourites that had been cored from modern contour-current deposits up to the late 1970s (Stow and Lovell, 1979). Subsequent work demonstrated that these muddy and sandy facies, together with intervening

D.A.V. Stow and J.-C. Fauge`res

235

silty contourites, commonly occur in composite sequences or partial sequences a few decimetres in thickness (typical range 0.2–3 m) (Fauge`res et al., 1984; Gonthier et al., 1984; Stow and Piper, 1984). It was also recognised that such facies sequences can be of siliciclastic, bioclastic, volcaniclastic or mixed composition. The ideal or complete sequence shows overall negative grading from muddy through silty to sandy contourites and then positive grading back through a silty to a muddy contourite facies (Figure 13.9). Well-defined sedimentary structures are generally absent, in part because they have been thoroughly destroyed by bioturbation. There may be an indistinct and discontinuous parallel lamination and lenses of coarser material. Primary structures, including rare cross-lamination, are more evident in the coarse silts and sands than in finer grained facies. Since the first publication of the model, such sequences of grain size and facies variation have been widely recognized from many more drifts throughout the world, and they have also been used in the recognition and interpretation of ancient contourites (e.g. Stow et al., 1998a; Hu¨neke and Stow, 2008). In the same way as for the ideal turbidite sequences, partial sequences of different thickness are equally or more common than the full sequence. Following the turbidite analogy of notation for the Bouma turbidite sequences (see Stow, 2005), a shorthand

Figure 13.9 Composite contourite facies model showing grain-size variation through the standard mud ^ silt ^ sand contourite sequence, linked to variation in contour-current velocity (from Stow et al., 2002b; Stow, 2005; with permission fromThe Geological Society, London).

236

Contourite Facies and the Facies Model

description of the contourite sequence using the notation C1–C5 has been introduced (Stow et al., 2002b). This lists the five principal divisions as follows: C5 = C4 = C3 = C2 = C1 =

upper muddy contourite division; upper mottled silty contourite division; middle sandy contourite division; lower mottled silty contourite division; lower muddy contourite division.

Thus a complete sequence of any composition is referred to as C1–5 (Figure 13.9). In a vertical succession of repeated sequences, there is a simple gradation from C5 of the underlying sequence to C1 of the overlying sequence. Separation of two superimposed sequences should be arbitrarily taken at the mid-point of the C5/C1 couplet. We have constructed for this contribution a further refinement of the standard model that shows more clearly the variety of partial sequences and facies most commonly encountered in contourite successions (Figure 13.10). All these partial sequences have been recorded from modern contourite successions (see Appendix 1), in which the omission of certain divisions can commonly be related to an increase in bottom-current velocity resulting in a prolonged phase of erosion and/or nondeposition. Base-to-middle sequences reflect the abrupt truncation of the full sequence as a result of increased bottom-current velocity and subsequent

Sandy contourites

rites ontou

ut c

cut-o

Base

C5

C5

C4

C4

C4

C5 C5

C5

C4

Omission surface

C3

C3 C2

C3

C1

C2

C5

Hiatus/ Omission C4 ± Erosion Hiatus/ C2 Omission

C3 Shale-chip horizon

C1

C1/C5 Omission surface

Fe–Mn horizons C3

C2

C1

Topc u

t-out

C3

C1 C1

conto

urite

s

Bioturbation ± Omission surfaces ± Scours

Gravel-lag horizon

C1 C2

Bioturbation Shale-chip horizon Dispersed Thin bed shale-chips + crosslamination

Gravel-lag horizon Dispersed gravel ± Scours and lenses

Mid-only sandy contourites

Thick bed + crosslamination

Figure 13.10 Variations on the standard contourite facies models showing the range of contourite facies, sequences and partial sequences commonly encountered in contourite successions.

D.A.V. Stow and J.-C. Fauge`res

237

non-deposition. Middle-to-top sequences reflect the gradual onset in deposition after a period of erosion. Partial sequences can also reflect velocity decrease downstream of a narrow gateway or channel, with middle-only sequences deposited proximally and top/base only sequences more distally. Base-only partial sequences are referred to as C1, C1–2 or C1–3, and top-only sequences as C3–5, C4–5 or C5, as appropriate. Rather than introducing a new division notation for the occurrence of other contourite facies within the sequence, it seems more sensible to highlight these departures from the standard sequence verbally. Base-only sequences that pass up into a gravel-lag and non-depositional surface are described as C1–3 with a gravel top. Likewise, top-only contourites are referred to as C3–5 with a sharp erosive base. Mid-only contourites can be referred to as C3 bioturbated, C3 laminated, or simply as bioturbated/laminated sandy contourites. Manganiferous horizons, non-deposition surfaces and hiatuses within contourite successions are all separately indicated.

13.4.1.

Sequence origin and interpretation

The origin of the C1–5 sequence can be related to long-term fluctuations in the mean current velocity, and/or to variation in sediment supply. Stacked sequences and repeated partial sequences indicate cyclic variation in the forcing variables. A growing body of data now exists, based on the best biostratigraphic, oxygenisotope and/or radiometric dating available in each case, to allow approximation of the mean time span of these cycles. This yields a periodicity of 5000–20,000 years for a variety of marginal drifts of terrigenous to mixed composition. In bioclastic successions, the cyclic facies pattern has a longer time (20,000–40,000 years) in the few examples from which we have good biostratigraphic dating, and is closely analogous to the Milankovitch cyclicity recognised in many pelagic and hemipelagic successions. It is therefore believed to be driven by the same mechanism of orbital forcing of climate that then affects changes in bottom-current velocity. Hu¨neke and Stow (2008) have estimated cycle periodicities for the few ancient successions of fossil contourites where dating has made such estimates possible. Their examples are all calcareous bioclastic in nature, the data suggesting cycles of between 20,000 and 200,000 years duration. These are therefore also possibly of Milankovitch origin. As yet, physical oceanographers know insufficient about the medium and longer timescale (decadal to millennial) variations in bottom currents (see e.g. Bacon, 1998), so we are less able to interpret the 5000–20,000 cycles as a direct function of climate or current-related variation. The shorter (sub-annual) variations resulting from benthic storm events are likely to cause some of the specific features within the contourite sequence, such as sharp or erosional contacts, coarse-grained lenses and non-deposition surfaces. Subtle colour banding, ichnofacies changes and variation in certain proxy measurements (e.g. magnetic-susceptibility characteristics), appear to occur with a mean centennial-to-millennial periodicity. We can only speculate about links to, for instance, bottom-current changes due to eddy generation and cycling, or to periodic spillover of bottom water through oceanic gateways. In addition, it is not easy to differentiate the relative importance of current velocity versus sediment supply in the development of contourite sequences, so that some of the cycling may be wholly due to supply variability that is not linked with

238

Contourite Facies and the Facies Model

bottom-current fluctuations. The most thorough approach to untangling this problem is to analyse variations in grain-size properties (including bulk-sediment mean size, sortable silt, carbonate-free mean size), and then to consider the co-variance or not of compositional attributes (e.g. terrigenous/biogenic ratios, benthic/planktonic ratios, percentage of coarse sand/gravel and shale chips, presence/absence of far-travelled components, clay/silt ratio), current indicators (e.g. scour surfaces, lamination, lag horizons, coarse-grained lenses, shale-chip concentrations) and bioturbation intensity coupled with ichnofacies types. These data need to be collated and compared for different sites over the same drift to distinguish regional from local effects, and to observe down-current trends (Stow et al., 2002b). Cycle irregularity is likely to result from interaction of sediment supply and current velocity.

13.5.

L AMINATION VERSUS BIOTURBATION IN C ONTOURITES

The great majority of contourites recovered from drift systems beneath extant contour currents are characterised by a notable absence of clear, distinct lamination and by the presence of common to abundant, pervasive bioturbation (see Appendix 1 of this chapter; see also Table 1 in Stow et al., 1998a). However, in some cases, they also exhibit an indistinct and discontinuous parallel lamination, partial grain alignment, sub-horizontal to irregular erosion surfaces, winnowed concentrations of thin layers and lenses of coarser material. These features are all reflected in the standard contourite model and sequence described above. Intuitively, we might expect the sediments deposited by any bottom-current system to show lamination as a result of fluid-flow processes and depositional sorting mechanisms. At moderate and higher flow velocities, we observe bedforms (ripples, dunes, etc.) on the sea floor beneath bottom currents, and would therefore expect their deposits to show cross-lamination at different scales. Such crosslamination is only rarely described from modern examples of these facies, however, and extensive bioturbation is generally dominant. It was this expectation of primary current lamination that led initially to a discussion about the diagnostic sedimentary structures in contourite facies. This disagreement still persists in the current literature, where well-laminated deposits are interpreted as contourites (see Hu¨neke and Stow, 2008; Martı´n-Chivelet, 2008; Shanmugam, 2008). There may be several reasons for the lack of clear lamination in contourites: • contourite accumulation rates are low and continuous, so that bioturbation is more than able to keep pace with deposition and so effectively destroys most primary lamination; indistinct lamination and other current-related features are all that remain; • for thick, muddy contourite accumulations, the low to very low current velocities are insufficient to result in primary lamination; • the very low sediment concentrations associated with most contour-current nepheloid layers are insufficient to permit the depositional sorting mechanism that develops silt/mud lamination in fine-grained turbidites.

D.A.V. Stow and J.-C. Fauge`res

239

There are, however, a number of notable exceptions in which lamination in present-day contourites has been described, and for which we have included new categories of contourite facies as described in the first part of this chapter (Figure 13.5). They are not as common as their bioturbated counterparts and are therefore best referred to as specific facies and/or depositional settings, as discussed below.

13.5.1.

Laminated sandy contourites

Laminated sandy contourites (Viana et al., 1998a; Nelson et al., 1999; Habgood et al., 2003) include siliciclastic facies where high-velocity bottom currents are dominant and large-scale bedforms (e.g. dunes) are evident on the sea floor. The lamination is distinct, but can also be diffuse, and may be associated with bioturbation. They also include bioclastic sandy contourites (biogenic calcareous and siliceous), in which both parallel and cross-lamination have been observed. Preservation of primary lamination in both siliciclastic and bioclastic facies is probably due to the high current velocity and dearth of organic matter inhibiting extensive bioturbation.

13.5.2.

Laminated muddy/silty contourites

The distinct parallel laminated facies in muddy contourites observed on the Baffin Sea slope (Yoon and Chough, 1993) are associated with continuous bioturbation, and can be clearly distinguished from interbedded turbidites and hemipelagites. They are interpreted to be related in part to down-slope flow of relatively high-concentration bottom currents following the generation of cold, dense waters near the surface, and in part to the preservation of lamination under harsh high-latitude conditions where bioturbation is insufficient to completely destroy primary structures. Distinctly laminated muds from the Baltic Sea (parallel lamination diffuse and abundant) have been interpreted as shallow-water contourites (Sivkov et al., 2002). If their interpretation and terminology is correct, the preservation of primary lamination can be related to relative lack of oxygen as well as to high rates of sediment input, both acting to suppress bioturbation. The laminated muddy facies presented in compilations by Stow (Stow, 1994; Stow et al., 1996b, 2002f) has always indicated that these are shallow-water contourites, but the better known examples are interpretations from the ancient record and must therefore remain more suspect.

13.5.3.

Mixed turbidite/contourite systems

With respect to mixed turbidite/contourite systems (Rebesco et al., 1996, 1997, 2002; Pudsey, 2000; Escutia et al., 2002), much debate has centred around the interpretaion of thick elongate sediment accumulations (drifts) off the western Antarctic Peninsula, which have now also been recognised from the northern margin off Wilkes Land. Most researchers would now accept that these are turbidity-current supplied systems that have been variously pirated and reworked by contour currents. The clear, but somewhat irregular and discontinuous lamination, locally associated with bioturbation, is best interpreted as the result of a combined turbidity-current/ contour-current process, although the nature of this process remains unresolved.

240

Contourite Facies and the Facies Model

Laminated, barren sediments containing ice-rafted debris (IRD) layers are observed on much of the Antarctic margin and are interpreted as contourites although they are atypically not bioturbated (Lucchi et al., 2002). In this case, the presence of IRD layers within laminated sediments is an indication against the hypothesis of a turbiditic origin, as turbidity currents are too short-lived to incorporate distinct layers of IRD. This particular type of glacigenic contourite facies appears associated to glacial times only, and is interpreted to result from unusual, climate-related, environmental conditions of suppressed primary productivity and oxygen-poor deep waters (Lucchi and Rebesco, 2007).

13.6.

CONTOURITE -RELATED F ACIES

As is often stressed, contourites are not an easily recognised deep-water facies, nor is the process of deposition a simple one. It is important to recognise that bottom currents will influence to a greater or lesser extent other deep-water sediments – particularly pelagic, hemipelagic, turbiditic and glacigenic – both during and after deposition. Where the influence is marked and deposition occurs in a drift, the sediment is termed ‘‘contourite’’. Where the influence is less severe, such that features of the original deposit type remain dominant, the sediment is said to have been influenced by bottom currents. Many pelagic and hemipelagic successions may have experienced some bottomcurrent influence, creating local winnowing of foraminiferal sands, regional variations in thickness or hiatuses. Reeder et al. (2002) interpreted the relatively high rates of hemipelagic accumulation in the Sicily Gateway as the result of bottom-current lofting and re-suspension of the bottom-current load into the overlying water column. This can be considered as an example of bottom-current-influenced hemipelagites. Marked thickness variations in mid-ocean pelagic successions, as noted by Dennielou (1997), might also be interpreted as weak bottom-current influence on an otherwise normal pelagite facies. Distinguishing features of contourites, hemipelagites, pelagites and other closely related facies have been discussed by Stow and Tabrez (1998). Part of the table from that work is reproduced here (Table 13.2). The recognition and interpretation of bottom-current-reworked turbidites has been notoriously more difficult. Some good modern examples have been described from the Columbia Gateway in the SW Atlantic (Masse´ et al., 1998; Fauge`res et al., 2002a, b) and the Sicily Gateway in the Mediterranean (Reeder et al., 2002). Rather more ambiguous, however, are the thin, clean, cross-laminated sands originally described by Hollister and Heezen (1972) from the NE American margin, as these might also be interpreted as normal fine-grained turbidites (e.g. Stow, 1979). The features that we suggest best characterise reworked turbidites are summarised in Table 13.3, together with those of normal (depositional) contourites and sandy contourites. These differ from the criteria proposed subsequently by several workers that are based purely on ancient turbidite successions (e.g. Stanley, 1988a,b; Mutti, 1992; Shanmugam et al., 1993a, 2000). However, we believe that criteria developed purely from known modern drift systems (as herein) are more reliable, because of the inherent uncertainties of interpretation when dealing with ancient series.

Diagnostic criteria for the recognition and distinction of various fine-grained deep-water facies

Turbidites (fine grained, thin bedded)

Bedding Usually well defined, continuous, thin bedded, regular

Structures Lamination lenticular and parallel, regular or indistinct; micro-crosslamination, low-amplitude climbing ripples, fading ripples and convolute lamination common

Turbidites ^ Unifitesa (fine grained, very thick bedded)

Hermiturbiditesb (thick, bioturbated muds, with mixed characteristics)

Contourites (fine grained, depositional)

Hemipelagites

Pelagites (biogenic ooze)

Pelagites (abyssal red clay)

Usually well defined, regular, extensive; beds >1 m, and may exceed 25 m thick

Poorly defined bedding

Poorly defined and irregular, may be absent; irregular, variation of very thin to very thick beds

Poorly defined or moderate, regular, but may be absent; beds when present may be even bedded and medium thick

Poorly defined to absent

Poorly defined to absent

Structureless or with very indistinct parallel lamination near base of bed

Primary structures absent, but may cap finegrained laminated turbidites

Lamination in parts only, mainly irregular, wavy, indistinct or lenticular, crosslamination only rarely present in silt or fine-sand layers; irregular mottling very common

No primary structures, but may develop fine fissile lamination where deposited under anoxic conditions

No primary structures

No primary structures

241

(Continued)

D.A.V. Stow and J.-C. Fauge`res

Table 13.2

(Continued )

242

Table 13.2

Turbidites ^ Unifitesa (fine grained, very thick bedded)

Hermiturbiditesb (thick, bioturbated muds, with mixed characteristics)

Contourites (fine grained, depositional)

Hemipelagites

Pelagites (biogenic ooze)

Pelagites (abyssal red clay)

Contacts usually sharp at bases and sharp or gradational at top of laminae; micro-scours, loading and injection structures

Contacts sharp at bases and sharp or gradational at top; little or no scouring evident

Contacts gradational and bioturbated

Contacts between beds always gradational and bioturbated

Contacts gradational and bioturbated

Contacts gradational and bioturbated

Bioturbation episodic, concentrated near tops of beds, often small scale, sometimes absent, rarely destroys all primary structures

Bioturbation in upper part of beds

Bioturbation continuous throughout bed; may be distinctive monospecific ichnofacies

Contacts can be sharp or gradational at tops and bases of layers and laminae; often gradations between the two along same contact; often irregular, sometimes erosive Bioturbation continuous and intensive, throughout sequence; several tiers of burrows, types vary according to contourite facies; can markedly alter or destroy primary structure

Bioturbation continuous and intensive throughout sequence; several tiers of burrows, uniform ichnofacies; may become homogenized

Bioturbation continuous and intensive as for hemipelagites

Bioturbation continuous; may be less evident than in oozes and hemipelagites; mottling and homogenization common

Contourite Facies and the Facies Model

Turbidites (fine grained, thin bedded)

Grain size typically from silt to clay

Grain size very fine silt and clay

Grain size from sand to clay grade

Distribution and moderate to good sorting indicate current deposition; silt and mud laminae usually well separated; silts often positively skewed (fine tail)

Sorting moderate to poor

Sorting moderate to poor

Grading positive, often in regular gradedlaminated units Fabric Grain alignment (silts) parallel to downslope currents

Grading absent or very poor slight positive grading

Grading generally absent

Sorting usually poor to moderate but distribution does indicate current deposition; silt and mud often irregularly mottled; silts often have low positive or negative skew (i.e. both coarse and fine tails) Grading irregular, both positive and negative sequences

Fabric not studied; presumed similar to thin-bedded turbidites

Fabric not studied

No grain alignment

Grain size from sand to clay grade

Grain size, clay and fine silt

Sorting poor to moderate, no current indications

Sorting, poor to moderate

No grading

No grading

No grain alignment

No grain alignment

243

Grain alignment (silts) may be parallel to alongslope currents; more often distributed by bioturbation

Grain size from sand to clay grade, with >40% of terrigenous fraction being silt sized; locally coarser Sorting usually poor to moderate, but distribution does indicate current deposition; silt and mud often have low positive or negative skew (i.e. both coarse and fine tails) No true grading, but irregular grainsize fluctuation may be present

D.A.V. Stow and J.-C. Fauge`res

Textures Grain size from fine to sand to clay grade

(Continued)

244

Table 13.2

(Continued )

Turbidites (fine grained, thin bedded)

Turbidites ^ Unifitesa (fine grained, very thick bedded)

Hermiturbiditesb (thick, bioturbated muds, with mixed characteristics)

Mud fabric may show large particle clusters (flocs) with random orientation

Allochthonous composition

Mainly allochthonous composition as for associated turbidites, plus admixture of hemipelagic input

Hemipelagites

Pelagites (biogenic ooze)

Pelagites (abyssal red clay)

Mud fabric may show small particle clusters, with horizontal orientation where not bioturbated Magnetic fabric (?) parallel to alongslope currents

Mud fabric may show small particle clusters and isolated particles; bed parallel

Fabric with small clusters and isolated particles; bed parallel

Fabric with small clusters and isolated particles, bed parallel

No magnetic fabric

No magnetic fabric

Random magnetic fabric

Uniform composition, local and fartravelled in surface currents, winds; varied inputs

Uniform composition derived from primary productivity in surface waters

Uniform composition

Uniform composition at scale of drift or margin deposit; part may be fartravelled, but most derived locally from pelagic and turbidite input and bottomcurrent re-suspension

Contourite Facies and the Facies Model

Magnetic fabric (?) parallel to down-slope currents Composition Allochthonous elements introduced into an area, so that turbidite composition often differs markedly from that of interbedded sediments

Contourites (fine grained, depositional)

Distribution Vertical sequence often a regular succession of positively graded beds, or gradedlaminated units (2–20 cm thick); these can form part of thicker coarsening- or fining-upward sequences

Terrigenous, biogenous, volcanigenic or mixed; may show compositional grading

Typically occur as isolated megabeds

Typically occur within distal turbidite basinplain succession

Nature usually a mixture of terrigenous and biogenic; can be >80% one or other, can also include volcanigenic, debris; reworked biogenic material common, often as broken and iron-stained debris; Fe–Mn rich in parts

Nature usually a mixture of terrigenous and biogenic (dominantly pelagic) elements, can include volcanigenic debris or glaciomarine input

Nature, calcareous, siliceous or mixed; with rare volcanigenic, terrigenous and cosmogenic input; Fe–Mn modules and crusts locally

Terrigenous and volcaniclastic clays and fine silts; wind-blown dust, cosmogenic input, microtektites, Fe–Mn nodules and crusts

Vertical sequence often an irregular succession of positively and/or negatively graded intervals (10–100 cm thick); largescale sequences not yet clearly defined

Vertical sequence absent or as regular cycles of more or less biogenic-rich compositions

Vertical sequence absent or as regular cycles of more or less biogenic-rich compositions

No vertical sequence

245

(Continued)

D.A.V. Stow and J.-C. Fauge`res

Nature can be terrigenous, biogenic, volcanigenic, or mixed, often containing shallow-water elements; may show compositional grading

246

Table 13.2

(Continued )

Turbidites (fine grained, thin bedded)

Hermiturbiditesb (thick, bioturbated muds, with mixed characteristics)

Contourites (fine grained, depositional)

Hemipelagites

Pelagites (biogenic ooze)

Pelagites (abyssal red clay)

Horizontal trends generally absent or very subtle

Horizontal trends absent

Horizontal trends of sedimentary features (e.g. grain size, composition) along bottomcurrent pathways (i.e. along-slope trends parallel to the margin or drift) Current evidence (ripples, fabric), where preserved, also show alongslope trends

Horizontal trends not present or weakly developed over large area

Horizontal trends not present or weakly developed over large area

No horizontal trends

No bottom-current evidence

No bottom-current evidence

No bottom-current evidence

Continuous sedimentation, no hiatus

Continuous sedimentation, but may be very reduced in places

Continuous sedimentation

Episodic, typically less frequent than thin-bedded turbidites

Episodic events but with very long settling periods (e.g. 0.5–1 a)

Semi-continuous sedimentation, with irregularly spaced, often prolonged hiatuses when bottom currents particularly strong

Contourite Facies and the Facies Model

Sedimentation rates Horizontal trends of sedimentary features (e.g. bed thickness, grain size, composition) along turbidity current pathways, i.e. down-slope trends Current evidence (ripples, flutecasts, fabric) also shows down-slope trends Episodic turbidite sedimentation, background sedimentation continuous, hiatuses uncommon except when associated with coarser-grained turbidites

Turbidites ^ Unifitesa (fine grained, very thick bedded)

Rates variable, low to moderate, <2 to 15 cm ka1

Rates relatively constant, commonly low, <10 cm ka1; may vary with carbonate cycles; locally may be moderate to very high ( > 100 cm ka1)

Rates very low, typically <1 cm ka1; very rarely up to 5 or 10 cm ka1

Rates extremely low, <<1 cm ka1; may be <0.1 cm ka1

D.A.V. Stow and J.-C. Fauge`res

Rate very variable, 0 to 1000 cm ka1

Source: Modified from Stow and Tabrez (1998). a Unifites (Stanley, 1981) are defined as thick, structureless muds of gravity flow origin that are the result of slow (hemipelagic) settling from detached turbidity currents in well-stratified basins. b Hemiturbidites (Stow and Wetzel, 1990) are defined as thick, bioturbated muds with partly turbiditic and partly hemipelagic characteristics. They are deposited from suspension clouds formed at the distal end of turbidity current pathways by a process of upward buoyancy.

247

248

Contourite Facies and the Facies Model

Table 13.3 Distinguishing features of muddy contourites, sandy contourites and bottomcurrent-reworked turbidites

Occurrence

Structure

Muddy contourites (terrigenous or biogenic)

Sandy contourites (terrigenous or biogenic)

Reworked turbidites (any composition)

Thick uniform sequences of finegrained sediment in deep-water settings

Thin to medium beds in muddy contourite sequences, rarely thick/very thick units Reworked tops of sandy turbidites in interbedded sequences Coarse lag in deepsea channels and straits Generally bioturbated and burrowed throughout with little primary structure remaining

In any normal turbidite setting where strong permanent bottom currents have been active

Interbedded with turbidites and other resedimented facies on inferred continental margins Dominantly homogeneous, bedding not sharply defined but cyclicity common

Bioturbational mottling generally common to dominant

Parallel and crosslamination more rarely preserved (often with bioturbation)

Distinct burrows (typical deepwater assemblage) present in many places

No regular structural sequence as in turbidites

Coarse lag concentrations (especially biogenic) reflect composition of course fraction in mud Primary silt/mud lamination – rare, but no regular sequence as in turbidites Sharp and erosive contacts common in parts

May show reverse grading near top, with sharp/erosive contacts common

Lower divisions of turbidite may be preserved, with the upper divisions either removed completely or modified by reworking Bioturbation/ burrowing common through reworked top reverse grading and irregular lag concentrations Bi-directional crosslamination, may be clean microcross-laminated silts with bioturbation Sharp erosive contacts may occur within turbidite sequence

249

D.A.V. Stow and J.-C. Fauge`res

Table 13.3

(Continued )

Texture

Fabric

Composition

Muddy contourites (terrigenous or biogenic)

Sandy contourites (terrigenous or biogenic)

Reworked turbidites (any composition)

Dominantly silty mud Frequently high sand content (0–15%) of biogenic tests in clastic contourites Medium to poorly sorted, ungraded, no offshore textural trends May show marked textural difference from interbedded turbidite if transport distances are different Mud-fabric – typically more parallel alignment of clays than for turbidites, but not well present in fossil contourites Primary silt laminae or coarse lag deposits show grain orientation parallel to the current (alongslope) Mixed contourites have combination of biogenic and terrigenous material (may be distinct from interbedded turbidites) Terrigenous material dominantly reflects nearby land/shelf source with some alongslope mixing and small amount of far travelled material (no down-slope trends)

Silt to sand-sized, more rarely gravel May be relatively free of mud and well sorted in some cases Tendency to low or negative skewness values

Removed/nondeposition of fines Significant textural differences from underlying turbidite (e.g. cleaner, better sorted, reverse grading þ lag, negative skewness)

No offshore trends

Indication of grain orientation parallel to bottom current (along-slope) or more randomised by bioturbation Other features (e.g. structures) also indicate alongslope flow, where preserved

Interbedded, reworked turbidite layers may show widely bimodal orientations or a more random polymodel fabric

Mixed biogenic/ terrigenous composition, typical terrigenous composition dependent on local source

Composition entirely reflects that of turbidite, with part of fine fraction removed

Biogenic material from pelagic, benthic and re-sedimented sources, typically fragmented and iron-stained Organic carbon content very low

Long exposure and winnowing may lead to chemogenic precipitation (probably rare) Organic carbon content very low

(Continued)

250

Contourite Facies and the Facies Model

Table 13.3

Sequence

(Continued ) Muddy contourites (terrigenous or biogenic)

Sandy contourites (terrigenous or biogenic)

Reworked turbidites (any composition)

Typically arranged in decimetric cycles of grain-size and/ or compositional variation with sandy contourites

Typically in decimetric cycles of grain-size and/ or compositional variation with muddy contourites See model (Figure 13.9) – partial sequences also common

Presents a typical turbidite sequence (i.e. top-absent or top reworked) Does not occur within standard cyclic contourite sequence

Source: Modified from Stow et al. (2002b).

Some of the sediments that have been described recently from mixed drift systems, such as those on the Antarctic Peninsula Margin (Rebesco et al., 2002) have been interpreted as the result of sediment supply from turbidity currents, followed by capture of the fine suspended cloud by active bottom currents. This material is then deposited at varying distances from the supply channel across the adjacent levees or mixed-drift bodies in a downcurrent direction. They show clear, but somewhat irregular lamination coupled with bioturbation where lamination is less pronounced. Bulk grain-size analysis shows a very poorly sorted silty clay grain size, although individual silty laminae are no doubt slightly coarser grained. A mixed turbidite/contourite facies type seems reasonable. In our opinion, there is not yet enough evidence to fully characterise and interpret various facies described from shallow-water, upper-slope to outer-shelf settings (Viana and Fauge`res 1998; Roveri, 2002; Sivkov et al., 2002; Viana et al., 2002a, b). Viana et al. (1998a) have made an important advance in this direction by synthesising data on all bottom-current-controlled sands, from deep-water true contourite sands to shallow-water bottom-current-influenced sands. However, the so-called shallow-water contourite facies, of all compositions and grain sizes, still requires further work.

ACKNOWLEDGEMENTS We thank many colleagues for giving generously their time and discussion over the years – in the field, at sea and in the laboratory. We also acknowledge the technical and administrative support at our respective institutes. DAVS further acknowledges receipt of a ‘‘Mobility Award’’ from the Spanish Ministry of Education and Science (ref: SAB2005-0182) during the time he was working at the Instituto Espan˜ol de Oceanografı´a (Malaga). Paul Knutz and the volume editors deserve special mention for their hard work and helpful comments.

A P P E N D I X

1

D ETAILS OF C ONTOURITE F ACIES R ECOVERED FROM M ODERN D RIFT S YSTEMS U SING C ONVENTIONAL C ORING T ECHNIQUES AS P UBLISHED O VER THE P AST 15 Y EARS 13A1.1.

NW UK C ONTINENTAL M ARGIN

Bottom currents: North Atlantic Deep Water (NADW), NE Atlantic Water slope current (NEAW), Norwegian Sea Overflow Water (NSOW). Drift system: complex drift system in NE Rockall Basin. Contourite facies: muds, sandy muds, muddy sands, + dispersed IRD; well bioturbated and mostly lacking primary structures, some indistinct lamination, rare cross-lamination, irregular winnowed concentrations of IRD; mixed composition, mainly siliciclastic, some bioclastic; distinct cyclicity, partial and complete sequences, 30–120 cm thick. References: Howe et al. (1994, 2002); Stoker et al. (1998a); Morri (2004). Drift system: Hebridean Slope, Barra mixed fan, Barra contourite sand sheet. Contourite facies: muds, mottled silts, muddy and silty sands, + dispersed IRD, and bottom-current-influenced glaciomarine sandy gravels; well bioturbated and mostly lacking primary structures, some indistinct lamination, irregular winnowed concentrations of IRD and relict IRD sandy gravels; mixed composition, mainly siliciclastic, some bioclastic; indistinct cyclicity in lower part (with masking by interbedded turbidites), partial base-only reversed-graded sequence common at surface, 20–400 cm thick. References: Armishaw et al. (1998, 2000); Stow et al. (2002a); Knutz et al. (2001, 2002b). Drift system: sheeted drift system in Faroe–Shetland Channel. Contourite facies: muds, sandy muds, muddy sands + dispersed IRD, thin microshale-clast horizons; well bioturbated and mostly lacking primary structures, some indistinct lamination, irregular winnowed concentrations of IRD; mixed composition, mainly siliciclastic, some bioclastic; distinct cyclicity, partial and complete sequences, 40–180 cm thick, good lateral correlation over 40–80 km. References: Akhurst et al. (2002); Howe et al. (2002).

13A1.2.

N ORWEGIAN C ONTINENTAL MARGIN

Bottom currents: Norwegian (slope) current, Arctic Intermediate Water (AIW). Drift system: Lofoten Drift, mounded elongated separated drift.

252

Contourite Facies and the Facies Model

Contourite facies: silty muds, sandy muds + dispersed IRD; well to less well bioturbated, some parts lacking primary structures, others with indistinct to faint lamination, irregular winnowed concentrations of IRD, including some microshale clasts; mixed composition, mainly siliciclastic, some bioclastic; indistinct cyclicity, partial and complete sequences, 60–180 cm thick. References: Laberg et al. (1999, 2002); Stoker et al. (1998a); Tore et al. (1998). Drift system: Barents Sea slope, sheet contourites intercalated with down-slope and hemipelagic deposits. Contourite facies: silty muds, sandy muds + dispersed IRD; well to less well bioturbated, pervasive through contourite facies, some parts lacking primary structures, most with indistinct to faint lamination (diffuse and discontinuous); mixed composition, mainly siliciclastic, some bioclastic. Note: diffuse laminated contourites interpreted as result of down-slope bottom currents from bottom-water formation near the surface. References: Yoon et al. (1991); Yoon and Chough (1993).

13A1.3.

G REENLAND C ONTINENTAL M ARGIN

Bottom currents: Norwegian Sea Overflow Water (NSOW). Drift system: Eirik Drift, mounded elongated drift. Contourite facies: muds, silty muds, sandy muds/silts + dispersed IRD; well to less well bioturbated, rare indistinct lamination; mixed composition, mainly siliciclastic, some bioclastic; indistinct cyclicity, 10–200 cm thick, and smaller-scale colour banding in parts. References: Hunter et al. (2007a, b).

13A1.4.

GULF OF C ADIZ , IBERIAN M ARGIN

Bottom currents: Mediterranean Sea Overflow Water (MOW) divides into several slope current strands. Drift system: erosive systems close to Gibraltar Gateway, leading to complex contourite depositional system, including Faro–Albufeira Drift. Contourite facies: muds, mottled silts and sands in depositional system; sands, laminated sands and sandy gravel-lag contourites in erosive system; mainly well bioturbated, faint lamination in parts, silt/sand lenses; some thick-bedded sands, well laminated (diffuse laminae); mixed siliciclastic/bioclastic composition; distinct cyclicity, partial and complete sequences, 20–140 cm thick, good lateral correlation of 40–50 km over some parts. References: Llave (2003); Nelson et al. (1999); Habgood et al. (2003); Stow et al. (1986, 2002b).

D.A.V. Stow and J.-C. Fauge`res

13A1.5.

253

M EDITERRANEAN SEA

Bottom currents: Mediterranean Bottom Water (MBW), Levantine Sea Intermediate Water (LIW). Drift system: Ceuta Drift, SW Alboran Sea. Contourite facies: muds, silty clays and sandy muds; mainly homogeneous (well bioturbated), faint lamination in parts, thin concentrations of bioclastic material; mixed siliciclastic/bioclastic composition; indistinct cyclicity, partial and complete sequences, 20–100 cm thick. References: Ercilla et al. (2002). Drift system: Corsica Basin Margin, Corsica Channel drifts, N Tyrrhenian Sea. Contourite facies: muds and silty muds; mainly homogeneous (well bioturbated), rare thin concentrations of bioclastic material; mixed siliciclastic–bioclastic composition; indistinct cyclicity, partial and complete sequences, 50–200 cm thick. References: Roveri (2002).

13A1.6.

EAST N ORTH AMERICAN C ONTINENTAL M ARGIN

Bottom currents: WBUC, plus influence of deep Gulf Stream and Antarctic Bottom Water. Drift system: HEBBLE area, Nova Scotian continental rise. Contourite facies: muds, sandy muds, some IRD gravelly muds; well bioturbated to moderately bioturbated; mainly siliciclastic composition, but sand fraction mostly bioclastic material; near surficial sediment study only (sequences unclear), good photographic evidence of current bedforms; note: top reworked turbidites present. References: McCave (1985b, 1988); McCave et al. (2002). Drift system: Greater Antilles Outer Ridge. Contourite facies: muds, silty clays and clayey silts; homogeneous and well bioturbated, thin irregular silty horizons; mainly siliciclastic composition, some bioclastic carbonate, silty horizons (minor sand) of bioclastic material, ash and Mn micronodules; no clear cyclicity, variation in sedimentation rates, plus hiatuses; distinct interbedded thin turbidites present. References: Tucholke (2002).

13A1.7.

E AST SOUTH AMERICAN C ONTINENTAL MARGIN

Bottom currents: NADW current, Brazil Intermediate Current, Brazil Current, Southern Ocean Water, AABW. Drift system: Campos middle-lower slope, Brazilian continental margin. Contourite facies: muds, silty muds and sandy muds, muddy sands; mainly well bioturbated, rare faint lamination, rare partly preserved cross-lamination (though

254

Contourite Facies and the Facies Model

extensive current bedforms at surface); mainly siliciclastic composition, some bioclastic material; distinct cyclicity, partial and complete sequences, 20–110 cm thick. References: Viana et al. (1998a, b, 2002a). Drift system: Campos upper slope, Brazilian continental margin. Contourite facies: fine- to coarse-grained sands (some pebbles), sandy muds and muddy silts; fine-grained facies mainly well bioturbated, rare faint lamination; sandy facies homogeneous to laminated (extensive current bedforms at surface); mainly siliciclastic composition, some bioclastic material; distinct but irregular cyclicity, overall coarsening to fining-upward sequence, top sands up to 10 m thick on upper slope. References: Viana and Fauge`res (1998); Viana et al. (1998a, b, 2002a). Drift system: Vema Channel contourite fan, Brazilian continental margin. Contourite facies: muds and silty muds, with micro-shale-clast layers and manganiferous horizons; mainly homogeneous (well bioturbated), rare faint lamination and Mn lamination; dominant siliciclastic composition, plus manganiferous horizons; distinct cyclicity, mostly small-scale colour sequences, 5–20 cm thick. References: Me´zerais et al. (1993); Fauge`res et al. (1998, 2002a). Drift system: Columbia mixed fan drift, Brazilian continental margin. Contourite facies: contourite muds and silty muds, with manganiferous horizons, and interbedded turbidites; contourite facies homogeneous (well bioturbated), with Mn lamination and micronodules; dominant siliciclastic composition, plus manganiferous horizons; no cyclicity evident; turbidites occur both as thin laminated siltto-mud units, and as top-truncated thin silt beds. References: Masse´ et al. (1994, 1996, 1998); Fauge`res et al. (2002b).

13A1.8.

EAST NEW Z EALAND C ONTINENTAL MARGIN, SW P ACIFIC

Bottom currents: SW Pacific Deep Western Boundary Current, and complex water-mass movement. Drift system: Campbell Skin Drift, thin sheet drift over elongate mound/wedge overlying pelagic–hemipelagic succession. Contourite facies: mixed biogenic sands, Mn-rich sands, interbedded with muds; well bioturbated, lacking primary structures; Mn nodule concentration and extensive reworking of older sediments; mixed terrigenous–biogenic composition, forams and Mn nodules/micro-nodules. References: Carter and McCave (1997, 2002); McCave and Carter (1997). Drift system: North Chatham Drift, plastered drift along the Chatham Margin. Contourite facies: muddy sands and sandy muds (sandy biogenic oozes), interbedded with silty muds; well bioturbated, lacking primary structures; mixed terrigenous/biogenic composition, foram-rich sand component. References: Carter and McCave (1997, 2002); McCave and Carter (1997). Drift system: Hikurangai Fan–Drift, bottom-current deflected accumulation modified from major down-slope turbidite fan.

D.A.V. Stow and J.-C. Fauge`res

255

Contourite facies: surficial muds and silty muds, well bioturbated, lacking primary structures, overlying thick unit of distinct turbidite muds and silty muds interbedded with bioturbated units; dominantly terrigenous composition as drift lies near or below Carbonate Compensation Depth, some airfall volcaniclastic silts. References: Carter and McCave (1997, 2002); McCave and Carter (1997).

13A1.9.

ANTARCTIC C ONTINENTAL MARGIN AND W EDDELL SEA

Bottom currents: Antarctic Circumpolar Current (AACC), AABW (Weddell Gyre). Drift system: North Weddell Sea drifts, plastered and mounded drifts transitional to hemipelagic basin fill. Contourite facies: silty muds, sandy muds, muddy sands + dispersed IRD; well bioturbated, mainly lacking primary structures, others with indistinct to faint lamination, Mn lamination and irregular winnowed concentrations of IRD; mixed composition, mainly siliciclastic, some bioclastic; distinct glacial/interglacial cyclicity in texture and biogenic content, sequences 30–150 cm thick. References: Pudsey (1992, 2002); Gilbert et al. (1998). Drift system: South Weddell Sea mixed drift/turbidite system. Contourite facies: complex mix of fine-grained silty muds and sandy muds + dispersed IRD, fed by turbidity currents and reworked by bottom currents; mostly laminated (from original turbidite?), some bioturbated; mixed composition, mainly siliciclastic, some bioclastic; distinct glacial/interglacial cyclicity in texture and biogenic content, but sequences confused by turbidite input. References: Weber et al. (1994); Melles et al. (1995); Diekmann and Kuhn (1999); Michels et al. (2002). Drift system: Scotia Sea, plastered, detached and sheeted drift system. Contourite facies: silty muds and muddy ooze + dispersed IRD; mostly bioturbated, some with distinct colour lamination; alternating siliciclastic/bioclastic composition; distinct glacial/interglacial cyclicity in texture and biogenic content, sequences 0.5–2.4 m thick. References: Howe and Pudsey (1999); Pudsey and Howe (1998, 2002). Drift system: Falkland Trough, plastered and confined mounded drift system. Contourite facies: silty muds, sandy muds, muddy sands and sands + dispersed IRD; mostly bioturbated, with rare silty laminae; mixed composition, including siliciclastic muds with biogenics, bioclastic sands and glauconitic sands; distinct cyclicity in texture and composition, partial and complete sequences (including hiatuses) 15–110 cm thick. References: Cunningham and Barker (1996); Cunningham et al. (2002). Drift system: Western Falkland Trough, surficial sandy contourites. Contourite facies: foraminiferal sands over sandy glauconite-rich contourites; foram sands are pale cream coloured, no visible burrows but homogenized by bioturbation; moderately well sorted, medium to fine grained; mixed composition of 60% biogenic material (forams, diatoms, radiolarians, nannofossils) and 40% terrigenous material.

256

Contourite Facies and the Facies Model

Glauconite-rich sands, greenish black, heavily bioturbated, with rare lamination and thin horizons of dropstones; fine to very fine grained and well sorted; well rounded glauconite grains. References: Howe et al. (1997). Drift system: Antarctic Peninsula, mixed drift–turbidite system. Contourite facies: fine-grained silty muds and sandy muds + dispersed IRD, fed by turbidity-current overspill, hemipelagic plumes and reworked by bottom currents; some laminated (diffuse laminae), some bioturbated; mainly siliciclastic composition, rare bioclastic material; indistinct cyclicity between bioturbated and laminated, sequences 50–150 cm thick. References: Camerlenghi et al. (1997); Pudsey and Camerlenghi (1998); Pudsey (2000); Rebesco et al. (1996, 1997, 2002). Drift system: Wilkes Land continental margin, mixed drift–turbidite system. Contourite facies: silty muds and sandy muds + dispersed IRD, fed by turbiditycurrent overspill and reworked by bottom currents; some laminated (diffuse parallel laminae), some bioturbated; mainly siliciclastic composition, rare bioclastic material; cyclicity between bioturbated and laminated. References: Escutia et al. (2002).

P A R T

5

MORPHOLOGY, GEOMETRY AND PALAEOCEANOGRAPHIC RECONSTRUCTIONS

C H A P T E R

1 4

C ONTOURITE D RIFTS : N ATURE , E VOLUTION AND C ONTROLS J.-C. Fauge`res1 and D.A.V. Stow2 1

De´partement de Ge´ologie et Oce´anographie, Universite´ Bordeaux1, Talence cedex, France National Oceanography Centre, Southampton (NOCS), Southampton, UK

2

Contents 14.1. Introduction 14.2. Drift Distribution and Characteristics 14.2.1. Tectonic setting 14.2.2. Water depth 14.2.3. Drift size 14.2.4. Drift shape 14.2.5. Sediment type 14.2.6. Bottom-current type 14.3. Growth History of the Blake Outer Ridge Drift System 14.3.1. The first sedimentary unit 14.3.2. The second unit 14.3.3. The third unit 14.4. Factors Controlling Drift Location, Morphology and Depositional Pattern 14.4.1. Bathymetric framework 14.4.2. Current conditions 14.4.3. Sediment supply 14.4.4. Process interaction 14.4.5. Sea level 14.4.6. Climate change 14.4.7. Tectonic setting and activity 14.4.8. Time constraints 14.5. Contourite-Drift Types 14.5.1. Sheeted drifts 14.5.2. Mounded drifts 14.5.3. Mixed drift systems 14.6. Discussion 14.6.1. Spatial evolution of drifts 14.6.2. Evolution of drifts in time 14.6.3. Complex contourite systems 14.6.4. Distinction from turbidite systems 14.6.5. Buried contourite drifts in modern ocean successions Acknowledgements Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00214-8

259 262 262 262 262 263 263 263 264 264 267 267 268 268 268 269 269 269 270 271 271 271 272 272 281 283 283 284 285 285 286 288

Ó 2008 Published by Elsevier B.V.

259

260

Contourite Drifts: Nature, Evolution and Controls

14.1.

INTRODUCTION

Since the mid-1960s, following the pioneering work of Bruce Heezen and colleagues (Heezen et al., 1966; Schneider et al., 1967; Heezen and Hollister, 1971; Hollister and Heezen, 1972), the significance of bottom currents for sediment transport and deposition has been well documented. Such deep-water bottom currents, flowing in response to major thermohaline and wind-driven circulation, are known to construct large accumulations of sediments in the deep sea. These sediment bodies were first described from the North Atlantic Ocean (Figure 14.1), 80°

75°

70°

65°

40°

40° 108 107

H.C.106 35°

35°

105

H.O.R.

104 102 103

30°

Bl.O.R. 30°

101 98

25°

100

25°

Bah.O.R.

99

20°

20° 80°

75°

70°

65°

Figure 14.1 Physiography of a portion of the western North Atlantic (from Hollister et al., 1972), showing the shaping of the continental rise by deep geostrophic currents. Bl.O.R. = Blake Outer Ridge; Bah.O.R. = Bahamas Outer Ridge; H.O.R. = Hatteras Outer Ridge; H.C. = Hudson canyon. Numbers 100, 101, . . . refer to the location of DSDP Leg 11 drilling sites. A multicolour version of this figure is on the enclosed CD-ROM.

261

J.-C. Fauge`res and D.A.V. Stow

and variously called ‘‘outer ridges’’ or ‘‘sediment drifts’’. They are now generally referred to as ‘‘contourite drifts’’ and are well-known throughout the world oceans, occurring everywhere from the abyssal floor, through deep continental margins, to mid-slope and even upper-slope settings (Figure 14.2). The term ‘‘shallow-water contourite drift’’ is preferred for those bodies constructed along the uppermost part of the continental slope and over the outer shelf edge. Contourite drifts are mainly composed of sediments deposited by contour currents (contourites), but may also contain associated deepwater facies, particularly hemipelagites, pelagites and glaciomarine sediments, and be variously intercalated with turbidites and debris-flow deposits. In the past 40 years of research, a large number of drifts have been studied using a combination of seismic profiling, side-scan sonar, swath bathymetry, sea-floor photography, coring and drilling techniques. Some of the more recent syntheses of this work, particularly focusing on the nature and construction of contourite drifts, include publications by Stow and Fauge`res (1993, 1998), Stoker et al. (1998a), Fauge`res et al. (1999), Rebesco and Stow (2001), Stow et al. (2002f), Wynn and Stow (2002a), Rebesco (2005) and Viana and Rebesco (2007). These have demonstrated a wide variation in drift location, morphology, size, sediment patterns, construction mechanisms and their controls. In this chapter, we briefly outline drift distribution and characteristics, consider the nature of drift development and the various controls that operate, and then

80°W

40°W

0°

7

60°N

0

400 km

60°N

5

6

1

3

8 4

2

50°N

50°N 9

10

40°N

40°N 15

11 12

30°N

14

30°N

13

80°W

40°W

0°

Figure 14.2 Contourite distribution and major giant contourite-drift locations in the North Atlantic Ocean (modified from Fauge' res et al., 1999; with permission from Elsevier). 1 = Hebrides drifts; 2 = Feni; 3 = Hatton; 4 = Gardar; 5 = Bjorn; 6 = Gloria; 7 = Snorri; 8 = Eirik; 9 = Sackville Spur; 10 = New Foundland; 11 = Hatteras; 12 = Blake; 13 = Bahama; 14 = Northern Bermuda;15 = Faro.

262

Contourite Drifts: Nature, Evolution and Controls

summarise our current classification of drift types. Subsequent chapters will focus more specifically on contourite drifts from the abyssal regime (Herna´ndez-Molina et al., 2008b), continental slopes (Herna´ndez-Molina et al., 2008a), shallow water (Verdicchio and Trincardi, 2008a) and high latitudes (van Weering et al., 2008). A further chapter (Mulder et al., 2008) is devoted to mixed turbidite/contouritedrift systems.

14.2.

D RIFT D ISTRIBUTION AND C HARACTERISTICS

Several attempts have been made over the years to classify contourite drifts (e.g. McCave and Tucholke, 1986; Fauge`res et al., 1993, 1999; Stow et al., 1996b, 2002d), although all run into the problem of just how best to sub-divide a continuous spectrum of deposit types (Rebesco, 2005). Furthermore, any individual drift may evolve through time from one type to another – for example from a small patch drift, to a larger sheeted drift to an elongated mounded drift.

14.2.1.

Tectonic setting

Contourite drifts are perhaps most common along the passive continental margins of the North Atlantic, Southwest Atlantic, Antarctic and southwest Indian Oceans. However, they also occur in mid-ocean settings (for instance, associated with the Mid-Atlantic Ridge, abyssal gaps or gateways and on abyssal plains), as well off active margins, such as the eastern New Zealand, southern Indonesian or Aleutian Margins and close to active margins, such as the northeastern Australian Margin.

14.2.2.

Water depth

Although contourite drifts have been found at almost all depths in the oceans, there may be some differences in their origin, facies, or other features, related to the depth at which they occur. A threefold division has therefore been proposed (Viana et al., 1998a; Stow et al., 2002c) into deep-water (>2000 m), mid-water (300–2000 m) and shallow-water (<300 m) drift types. This is of little relevance for non-recent and fossil contourite drifts, for which the palaeodepth is largely unknown (Rebesco, 2005).

14.2.3.

Drift size

Contourite drifts can be as large in size as many sedimentary systems built by turbidity currents and related down-slope processes. Drift sizes range from small patch drifts (about 100 km2), equivalent in size to isolated turbidite lobes or debris-flow masses on slopes, to giant elongated drifts (>100,000 km2), which match many of the world’s large muddy elongate fans. In some cases, with the Argentine Basin as an example, the contourite sheet deposits may cover an area of more than 1,000,000 km2. Although they show a large morphological variability, the most easily recognised drifts have an elongated, mounded shape and variable dimensions, ranging

J.-C. Fauge`res and D.A.V. Stow

263

up to the really giant drifts, tens to hundreds of kilometres long (up to 1000 km), 10 to more than 100 km wide, and with sediment thicknesses from some tens to 2000 m. These may show a relief up to 1500 m above the adjacent sea floor.

14.2.4.

Drift shape

The most famous contourite drifts in the present-day oceans were first identified on the basis of their typical mounded morphology, and their elongation more or less parallel to the continental margin, and hence to the contour-current flow direction. However, a range of different shapes is now recognised, including less regular patches and sheets, as well as fan-shaped bodies downstream of deep-water gateways. In addition, it is clear that contourites also occur closely interbedded with other deep-water facies types, and do not necessarily form unique or distinctive sedimentary bodies.

14.2.5.

Sediment type

Giant contourite drifts are predominantly formed of muddy and silty contourites, often as relatively thick and uniform successions, with less abundant sandy contourite horizons. Regarding their composition, they include siliciclastic, biogenic and volcaniclastic components, either mixed or alone, and may also contain ferromanganese horizons and nodules. They may be interbedded with pelagites, hemipelagites and fine-grained turbidites, and at high latitudes with glaciomarine hemipelagites and dropstones. Sandy contourites do not generally form whole drift systems, but are relatively restricted to more active bottomcurrent environments (contourite channels and gateways). The thin sandy horizons within muddy drift deposits correspond to episodes of higher current velocity. Shallow-water contourite drifts, deposited on the outer-shelf/upper-slope (50–300 m water depth), typically have more reduced dimensions and may comprise coarser sediments (silty and sandy contourites) than those of the larger drifts. They are relatively closer to continental sources of sediment, and the currents involved in their construction may have a higher velocity.

14.2.6.

Bottom-current type

In this chapter, we are principally concerned with drifts formed under the influence of bottom currents driven by the major thermohaline and wind-driven circulation systems. We also recognise, however, that a number of other bottom-current processes can deposit isolated contourite facies, often interbedded with other sediments, including upwelling and downwelling currents (Yoon and Chough, 1993; Seranne and Abeigne, 1999), up- and down-canyon currents (Stow et al., 1996b) and internal tides and waves (Rebesco, 2005). Sandy contourite sheets and patch drifts can be deposited in outer-shelf to upper-slope settings by a combination of related processes, collectively termed sea-floor polishing and spillover (Viana and Fauge`res, 1998; Stow and Mayall, 2000).

264

Contourite Drifts: Nature, Evolution and Controls

The classification of drifts followed in this chapter uses a combination of drift distribution and shape. First, however, we outline the growth and development of drifts using the well-known Blake Outer Ridge as an example.

14.3.

GROWTH HISTORY OF THE BLAKE OUTER R IDGE D RIFT SYSTEM

As contourite-drift deposition is mainly controlled by deep-water thermohaline circulation, drift initiation and growth history are closely related to global changes in the pattern of deep oceanic circulation. For many existing drifts in present-day ocean basins, drift development typically begins with – and overlies – a more or less flat, major erosional horizon, which is coeval and widespread throughout the basin. This hiatus corresponds to an important hydrological event involving the initiation of active bottom-water circulation in the basin. It is then overlain by contourite-drift deposits that begin to accumulate following some reduction in the vigorous circulation. Following their initiation, the construction of most contourite drifts is semicontinuous over a very long time span, but is also typically marked by an alternation of intervals of deposition with intervals of erosion, dissolution or some other changes in the depositional regime. This periodicity reflects significant changes in circulation, which are, in turn, controlled by global climatic changes and polar ice-sheet development, often associated with plate-tectonic events (Kennett, 1982). Once contourite accumulation has ceased, following a last major change in bottom circulation, cessation for example, the drift is progressively covered by onlapping turbidites and/or by draping pelagite/hemipelagite deposits. One of the first series of detailed studies of a major drift system was that of the Blake Outer Ridge (Figures 14.3 and 14.4) located along the eastern US Margin and associated with the Western Boundary Under Current (WBUC) (Ewing and Hollister, 1972; Shipboard Scientific Party, 1983; Mountain and Tucholke, 1985; Tucholke and Mountain, 1986; McMaster et al., 1989; Locker and Laine, 1992). This giant drift (about 600 km long, up to 100 km wide, between 1000 and 2500 m thick, and up to 1000 m relief) presents a mounded elongated morphology, as a prominent SE-trending extension of the continental rise, somewhat oblique to that of the continental margin. The drift overlies a major regional unconformity, Au, which is marked by a widespread seismic reflector that corresponds approximatively to a zone of silicified sediments at depth within the margin succession. It has a complex origin but, as it is distributed along the whole of the American North Atlantic Margin and is associated with a large depositional hiatus, there is little doubt that it reflects widespread erosion by a newly initiated bottom-current system near the Eocene/Oligocene boundary.

14.3.1.

The first sedimentary unit

The first sedimentary unit (Figure 14.4-1), Oligocene to Early–Middle Miocene in age, was deposited on the lower rise between Horizon Au and a late Miocene

265

J.-C. Fauge`res and D.A.V. Stow

UC WB

SE

NW A

B

Blake Outer Ridge

3

A

5 B

7 9 3

Md1

Md3

Base of gas hydrates

0

40 km

Md2

5 A 7 9

J1

J2 Basement

β Hummocky Sediment Backslope Hummocky Sediment reflectors waves deposits reflectors waves

(b)

(a)

Figure 14.3 The Blake and Bahama Outer Ridges (SE US Margin). (a) Morphology of the contouritic drift (detached drifts) and localisation of seismic line (a, b) and DSDP drilling sites (102, 103, . . .; modified from Shipboard Scientific Party, 1983). (b) Seismic line and interpretation showing three major discontinuities (Md1 = Au ; Md2 = ‘‘Merlin’’; Md3 = ‘‘Blue’’) linked to global hydrological events (multi-channel seismic line, modified from Mountain and Tucholke, 1985).

unconformity (‘‘Merlin’’ reflector), dated as about 12–10 Ma. The Merlin hiatus is related to a significant hydrological event that was synchronous with a sharp rise in sea level (Haq et al., 1987). The episode of drift growth between these horizons is characterized by a large sediment supply and predominant down-slope processes on the continental slope and upper rise. On the lower rise, along-slope processes were dominant. The contour currents were less intense than those that created horizon Au, but still sufficiently strong to influence sedimentation on the rise. An active ‘‘Gulf Stream’’ (warm, winddriven, surficial current) at this time led to marked erosion of the Blake Plateau on the adjacent upper continental slope (Pinet and Popenoe, 1989). Interaction of the deeper parts of this ‘‘Gulf Stream’’ with the underlying WBUC is believed to have been instrumental in causing contourite deposition and drift construction (Figure 14.4-5a and b). The location of the ‘‘Gulf Stream’’ is believed to oscillate from NW to SE in response to eustatic sea-level changes. During the early Miocene sea-level lowstand, its shift towards a more southeasterly location would have been responsible for strong erosion of the Blake Plateau and for eastward deflection of the WBUC, leading to deposition of the Blake Outer Ridge drift with a trend oblique rather than parallel to the margin. Subsequently, during the latter part of the early and middle Miocene, the ‘‘Gulf Stream’’ veered back to a more northwesterly location as the sea level rose. At that time, it played no further significant role in drift sedimentation, which was instead wholly

(3)

Upper Pliocene–Present

WBUC

35°

(c)

Blake Outer Ridge

30°

80°

75°

WBUC

(b)

Bahama Outer Ridge

(2)

Middle Miocene– Upper Pliocene 35°

WBUC

(a) 30° Blake Outer Ridge

80°

75°

GS GS

Bahama Outer Ridge

(5)

(1)

Lower Oligocene– Middle Miocene 35°

Zero thickness

................. ................. ................. ................. ................. ................. ................. ................. ................. ................. .................

Outcrop

200

Bathymetry

1000

Isopach (m)

.

Drill site

30° Sediment waves 80°

75°

Bahama Outer Ridge

Blake Outer Ridge

102

(4)

Figure 14.4 Interpretation of the Blake Outer Ridge growth. (1^3): Isopach maps showing the distribution and thickness variations (modified from Tucholke and Mountain, 1986). (1) = Lower Oligocene to Middle Miocene; (2) = Middle Miocene to Upper Pliocene; (3) = Upper Pliocene to Present). (4) = captions of the maps. (5) = hypothetical reconstruction of the ridge development above horizon Au (from Ewing and Hollister, 1972). The initial Oligocene building (5a) is characterized by a surface (the ‘‘Gulf Stream’’: GS) and deep current interaction (Western Boundary Under Current: WBUC), predominant sediment aggradation with low southeastward deposit migration; (5b) corresponds to a time of predominant contour-current depositional processes, strong southeastward deposit migration, and initiation of a secondary drift with a westward migration, during the early Miocene. (5c) represents the last development of the ridge (Quaternary) characterised by contour-currentcontrolled erosion (northern flank of the ridge) and deposition (top and southwestern flank). A multicolour version of this figure is on the enclosed CD-ROM.

J.-C. Fauge`res and D.A.V. Stow

267

controlled by the WBUC. However, the initial eastward deflection of the drift was more or less mirrored during the following stages of deposition, with only a gentle shift of the drift towards the southeast, as the deposits migrated into the deeper basin. The sediments are composed dominantly of silts and clays derived from an Appalachian source, as well as from a pelagic input rich in siliceous microfossils. The sedimentation rate was relatively high (19 cm ka 1), as a result of turbid water flowing off the northern part of the Blake Plateau and feeding the southward flowing undercurrent. This succession was deposited by active contour currents, as underlined by active construction of the Blake drift and deposition of well developed sediment waves. During the early Miocene, part of the WBUC began to cut across the drift and to build a spur of sediment towards the west (Figure 14.4-5b).

14.3.2.

The second unit

The second unit (Figure 14.4-2), late Miocene to late Pliocene in age, is bounded at the base by the Merlin unconformity, and at the top by a major late Pliocene unconformity (‘‘Blue’’ reflector). This latter is believed to be related to an episode of bottom-current intensification following active glaciation in the Arctic region, which was in turn associated with the closing of the Central American gateway at Panama. Isopach maps of deposit thickness and seismo-facies both show strong evidence of contour-current deposition. The Blake drift experienced a sedimentation regime fairly similar to that of Unit 1 and continued to grow upwards and seaward with a sedimentation rate only slightly lower than during the previous period (14 cm ka 1). However, contourite deposition was thicker on the southwestern flank of the drift (where the incipient spur developed further into a secondary E–W-oriented drift) than on the northeastern flank.

14.3.3.

The third unit

The third unit (Figure 14.4-3), late Pliocene to Holocene, was deposited after the pulse of bottom-current erosion correlating with the ‘‘Blue’’ reflector. It forms a perched lens of sediment on the top of the ridge where the Pleistocene deposits had completely filled the trough between the crests of the main ridge and the secondary drift (Figure 14.4-5c). The sediments are dominantly silty clay, probably derived from enhanced glacial supply, and accumulated with a similar fairly high rate (14 cm ka 1 during the Pleistocene, at the very top of the drift). Sediment distribution still suggests a strong control by active contour currents. Along the southwestern flank, the deposits show patterns similar to those of the underlying succession. On the northeastern flank, sea-floor erosion linked to the ‘‘Blue’’ reflector has re-excavated the Merlin reflector, and the contour currents have remained strong enough to prevent significant sediment accumulation up to the present day.

268

Contourite Drifts: Nature, Evolution and Controls

14.4.

FACTORS C ONTROLLING D RIFT LOCATION , MORPHOLOGY AND D EPOSITIONAL P ATTERN

The example given above for the Blake Outer Ridge is only one of many, but does serve to show some of the key controls on drift development. Contourite drifts form in many different locations and water depths, and with different morphologies, sediment composition and depositional patterns. These large-scale features of drifts are controlled by a number of interrelated factors (Fauge`res et al., 1993; Rebesco, 2005), including (1) the bathymetric framework (water depth and morphological context), (2) the current conditions (velocity, variability, and Coriolis force), (3) the sediment supply (amount, type, source, input, variability), (4) interaction with other depositional processes (in time and space), (5) sea level and sea-level fluctuations, (6) climate and climate change, (7) tectonic setting and activity and (8) the length of time over which these various processes and controls have operated and varied. It is not a simple matter to disentangle these various controls as many clearly overlap and are interrelated. Neither is it always certain just what effect a particular control exerts. However, the following sections briefly review some of the above key controls and show some of the ways in which they may influence contourite drifts.

14.4.1.

Bathymetric framework

To some extent, the water depth influences the size of the drift and the type of contourite deposits: at increasing depths, there is greater potential to develop larger drifts, and these are generally composed of finer-grained deposits. The removal of carbonate material below the carbonate compensation depth can also be significant. Bottom-current velocities in deep, flat basins are generally lower and less focused, so that deposits spread out as very large sheet drifts. The morphological context plays a still more fundamental role. According to the slope gradient of the sea floor, the current velocity and confinement vary due to the Coriolis effect, which in turn affects the presence of erosional versus depositional processes and hence drift growth and sediment distribution. Variations of the trend of the margin strongly influence the flow pathway, flow separation and meandering, and hence the overall shape, position and number of individual drifts. Particular morphological contexts, such as slope terraces, channels, narrow oceanic gateways and confined basins, are responsible for specific drift morphologies and sediment distribution.

14.4.2.

Current conditions

The velocity and intensity of the current is variable in time and space, as determined by a range of global factors and by the Coriolis force. The nature and distribution of currents thus determine (1) the type of contourite facies deposited, (2) the grain size of sediment transported, deposited or eroded over a drift system, (3) the amount of sediment carried, and the overall rate of accumulation and (4) the

J.-C. Fauge`res and D.A.V. Stow

269

development of various depositional bedforms, erosional features, and seismic facies. Current variability creates cyclicity in sediment and seismic facies over different scales, and particularly intense current conditions lead to lack of deposition and/or erosion. Hydrological events marked by drastic erosion may remove large volumes of sediment already deposited on the drift and form major discontinuities and sediment hiatuses. Long-term current conditions, together with other controls, ultimately determine what type of drift is deposited – sheet drifts tend to derive from slower, more spread-out currents, whereas elongated mounded drifts are formed under higher velocities. Short-term current changes, such as the benthic storms caused by high-energy disturbances (eddies), also have pronounced (but more local) control on contourite facies, deposition and erosion.

14.4.3.

Sediment supply

The amount of sediment available partly controls the drift size, relief and deposit thickness. The source and input points (e.g. pirating of upstream turbidity currents, pelagic/hemipelagic contribution from the surface, erosion of contourite channels and transfer to drift) in part control the location of drift growth, and further determine drift form, development and composition. The type of sediments (terrigenous, biogenic, volcaniclastic) influences the depositional bedforms and seismic facies, while variability in input affects contourite cyclicity (sequences) and facies. Sediment supply is, in turn, significantly affected by other variables such as tectonic activity, sea level and climate.

14.4.4.

Process interaction

Contour currents rarely act alone in the marine environment, so that both the currents and their deposits will be affected by interaction with other processes. Sediment supply is significantly influenced by pelagic, hemipelagic, glaciomarine and turbidity-current input into the contour current. In some cases it may be that a drift origin at a particular location requires a specific sediment input from one or several other related processes, and certainly the rate and variability of drift growth is affected by such interaction. Interaction of one or more bottom currents or between different strands of the same current, and interaction with major surfacecurrent systems, are all considered as significant controls in drift location, deposition and shape. Particularly close interaction with shallow-water processes, in the outershelf to upper-slope environment, leads to a variety of shallow-water drift types (Verdicchio and Trincardi, 2008a), whereas turbidity-current processes lead to the development of mixed drift systems (Mulder et al., 2008); high-latitude drifts have their own specific characteristics, in part related to interaction with glaciomarine processes (van Weering et al., 2008).

14.4.5.

Sea level

Eustatic sea-level fluctuations also influence drift growth and morphology indirectly, as they partly control the nature and volume of sediment supply, the nature

270

Contourite Drifts: Nature, Evolution and Controls

and generation of different water masses (surface and deep), and the oceanic circulation pattern (wind-driven and thermohaline). Sea-level fluctuations are closely linked also with global climatic fluctuations (see below). However, there are no unequivocal data that directly link the sea-level with rates of drift accumulation or destruction, and certainly not at a timescale approaching that of glacial/ interglacial sea-level fluctuations. Detailed study of the complex contourite depositional system in the Gulf of Cadiz region, related to Mediterranean Outflow Water, reveals a complex relationship between sea level, climate and drift growth (Herna´ndez-Molina et al., 2008a). Whatever the relative intensity of bottom circulation, the influx of large volumes of continental sediment into the deep-sea – during a major sea-level lowstand – generally results in the masking of contourite sedimentation and hence dominance of down-slope deposits during that time span. Such a situation is particularly clear for the continental margin off the eastern USA (Tucholke and Mountain, 1986). Episodes of active bottom-current circulation (as during a global hydrological event) typically generate a widespread surface of erosion or nondeposition in drift systems. Therefore, significant accumulation on contourite drifts is favoured, on the one hand, by a moderate intensity of bottom currents and, on the other hand, by relatively low rates of sediment supply via turbidity currents or other mass flows. In such a case, contourite-drift development therefore does not fit neatly into highstand, lowstand or an intermediate position in sequencestratigraphic models as proposed by some authors (Vail et al., 1977a, 1991; Posamentier et al., 1988; Haq, 1991).

14.4.6.

Climate change

The lack of a good sequence-stratigraphic model incorporating sea-level change and contourite-drift development is closely linked with uncertainty in relating particular climatic conditions to the pattern and intensity of bottom circulation. For example, for bottom circulation linked to the North Atlantic Deep Water (NADW), the episodes of greatest intensity are found by some authors (Boyle and Keigwin, 1982, 1987; Duplessy et al., 1988; Lehman and Keigwin, 1992; Dowling and McCave, 1993; Howe et al., 1994) during the interglacials, but during the glacials by others (Robinson and McCave, 1994; Revel et al., 1995), or during deglaciation (Dowling and McCave, 1993). In fact, it appears that different water masses may behave differently during the same climatic episode. In the North Atlantic, there was an intensification of intermediate water and reduction of deep water during the Last Glacial Maximum, followed by a brief changeover and then a return to this pattern during the main deglaciation (Bond et al., 1992; McCave et al., 1995a). The same pattern of intermediate- (Antarctic Intermediate Water) and deepwater circulation (NADW) was found for the Southwest Atlantic (Viana, 1998). We might therefore conclude that, at the scale of interglacial/glacial cycles or longer (i.e. 100,000–1,000,000 years) during the Neogene, the data from the drifts that have been studied best would appear to show a more or less random variation in drift growth related to climate and sea-level fluctuations.

J.-C. Fauge`res and D.A.V. Stow

14.4.7.

271

Tectonic setting and activity

Many aspects of the overall tectonic framework act, directly or indirectly, as fundamental controls on drift development. The tectonic setting affects the slope gradient and morphology, the opening and closing of oceanic gateways, sediment supply (volume and input points), slope stability and, hence, more local sediment supply and generation of morphological features. Furthermore, neotectonic activity in a more local context can help produce physiographic obstacles to bottomcurrent flow as well as favoured pathways or contourite channels. This, in turn, affects the nature of the bottom-current system (one or multiple pathways, for example) and hence of the style and size of drift development. Such an interaction between neotectonics, morphology and contour currents has been clearly demonstrated for the complex contourite depositional system in the Gulf of Cadiz, (Herna´ndez-Molina et al., 2006c, 2008a; Llave et al., 2007, among others).

14.4.8.

Time constraints

The length of time over which any of the above controls operates, and the periodicity of change or cycles in such controls obviously will affect drift nature, development and growth history distinctly. To construct drifts of significant thickness, size and with a mounded morphology, the contour-current processes involved must have remained semi-continuous in time and space (albeit possibly periodically interrupted by important erosive events) over several to tens of millions of years.

14.5.

CONTOURITE-D RIFT T YPES

The following classification of contourite drifts based on those of Fauge`res et al. (1999) and Stow et al. (2002c) uses a combination of drift distribution (i.e. location) and morphology, and also illustrates the drift development in the context of a particular hydrological background. We have refrained from dealing with additional drift types based largely on specific morphological and/or tectonic settings, as published by Rebesco and Stow (2001) and Rebesco (2005), as these make for greater complexity in classification and terminology but add little by way of fundamentally different genetic systems. Furthermore, we recognise that such classifications are never mutually exclusive, so that gradation and overlap between all types are the norm. At some still ill-defined point of morphology, a more or less flat drift with low mounded geometry (a ‘‘sheeted drift’’) will be called an ‘‘elongate mounded drift’’. There is a similar gradation between large (and quite long) ‘‘channel-related patch drifts’’ and ‘‘elongated mounded drifts’’, and between drifts dominated by contour-current deposition and those increasingly affected by other processes; therefore, we call them all ‘‘mixed drifts’’. Most drifts are either sheeted or mounded in overall morphology, but the mounded forms are further classified on the basis of their location with respect to other physiographic features.

272

14.5.1.

Contourite Drifts: Nature, Evolution and Controls

Sheeted drifts

The overall geometry of contourite sheets or sheeted drifts differs only very subtly from turbidite sheets on basin plains or their cover of lower-slope and interchannel regions. They are characterised by a wide, mounded geometry, covering a large area with a fairly uniform thickness, showing a very slight decrease in thickness from the central region towards its margins. The internal seismofacies is typically one of low-amplitude, discontinuous reflectors or, in some parts, is more or less transparent. The depositional units that form the sheet have a fairly regular thickness over the whole area swept by the currents. They show a predominantly aggradational stacking pattern and no significant migration. They may comprise or be covered by large fields of sediment waves. The following three kinds of sheeted drift are identified: (1) abyssal sheets, which cover basin plains whose margins trap the bottom currents within a complex pattern of gyre-like circulation; (2) slope sheets, which are spread out across continental margins where a gentle gradient and smooth topography favour a wide non-focused current; and (3) channel-related patch sheets (see channel-related drifts below). 14.5.1.1. Abyssal sheeted drifts Abyssal sheeted drifts (Figure 14.5) are the most impressive as they can cover large areas with deposits up to hundreds of metres thick. Examples include those of the South Brazilian Basin (Damuth, 1975; Damuth and Hayes, 1977; Me´zerais et al., 1993), the Mozambique Basins (Kolla et al., 1980; Ben-Avraham et al., 1994), the Irminger Basin (Egloff and Johnson, 1975, 1978), the Argentine Basin (Flood and Shor, 1988) and the North Rockall Trough (Richards et al., 1987; Howe et al., 1994; Stoker, 1995, 1998b). These drifts may be capped by giant elongate bifurcated drifts like in the Irminger Basin (Gloria Drift) or the Argentine Basin (Zapiola Drift). 14.5.1.2. Slope sheeted drifts Slope sheeted drifts (Figures 14.6 and 14.7) occur either near the foot of slopes where upwelling or downwelling bottom currents exist such as in the Gulf of Cadiz (Kenyon and Belderson, 1973; Fauge`res et al., 1985c; Nelson et al., 1993; Habgood et al., 2003), and the Faeroe–Shetland Channel (Howe et al., 2002), or plastered against the slope or rise, particularly where gentle relief and smooth topography favour a broad non-focused bottom current, such as on the Hebrides Margin (Howe et al., 1994; Stoker, 1998b; Stoker et al., 1998a), the Chatham rise (Wood and Davy, 1994) and the Brazilian Margin (Viana et al., 1998a).

14.5.2.

Mounded drifts

These drifts are characterised by their distinctly mounded and more or less elongated geometry. Three kinds of mounded drifts are identified (Figure 14.8): (1) giant elongated drifts that are of spectacular dimensions and that are, because of their common elongation parallel or sub-parallel to contours, easily recognised as being of contourite origin, (2) channel-related drifts specifically related to deep channels, gateways or contourite moats and (3) confined drifts deposited in relatively small confining basin.

J.-C. Fauge`res and D.A.V. Stow

Figure 14.5 The abyssal Gloria sheeted drift (from Egloff and Johnson, 1975; with permission from the Canadian Research Council Press). (1) Morphological map. (2) Seismic line. A multicolour version of this figure is on the enclosed CD-ROM.

273

274

Contourite Drifts: Nature, Evolution and Controls

North-East Rockall Trough Buried sediment waves

Wyville-Thomson ridge

Sheeted drift

ED

M MRD

twt

NE

(1)

SW

0.4 s

+ 1.4 s

0

10 km

2.4 s

twt

5 km

(2)

Figure 14.6 Slope sheeted drift in the Hebrides slope, adjacent to the Wyville-Thomson Ridge, Rockall Trough (single-channel air gun, see Howe et al., 1994; Stoker, 1995, 1998; Stoker et al., 1998b). (1) Seismic line showing a buried wavy sheeted drift (PP = base of Plio ^ Pleistocene), and a sheeted drift merging into a plastered/separated drift (ED = elongated low mounded relief ) associated with moat-related drift (MRD). (2) Detail of the buried sediment wave field overlain by a debris-flow package (d.b.).

14.5.2.1. Giant elongate drifts Their dimensions are variable: from a few tens of kilometres to over 1000 km long, with length/width ratios from 2:1 to 10:1, and thicknesses of up to several hundreds of metres. Both the elongation direction and directions of progradation can vary with respect to the contours of the continental margin or basin, and are dependent on the interaction between the morphology (i.e. slope gradient and regularity of the sea floor), the current system and intensity, and the Coriolis force. Three principal types are recognised (McCave and Tucholke, 1986): plastered, separated and detached drifts. 14.5.2.1.1. Plastered drifts Plastered drifts (Figures 14.8 and 14.9) are located along a gentle slope swept by fairly low-velocity currents. Deposition occurs on one side, both sides and/or directly below the current pathway, together with lateral migration of the drift axis. Examples include the Gardar and Bjorn Drifts in the North Atlantic, the Guadalquivir Drift in the Gulf of Cadiz and other drifts like along the New Zealand Margin and in the Norwegian Sea (e.g. McCave et al., 1980; McCave and Tucholke, 1986; Fauge`res et al., 1999; Laberg et al., 2001). 14.5.2.1.2. Separated drifts Separated drifts (Figures 14.8, 14.10a, 14.11, 14.12) are elongated parallel to the slope and can occur at any depth, particularly associated with steeper parts of the slope where the contour current is restricted due to Coriolis force. The elongated body is separated from the adjacent margin by a distinct contourite moat along

275

J.-C. Fauge`res and D.A.V. Stow

2 (G.C.)

36° 30 Sheeted sandy drift

G.B.

36°N

MOW 1(G.E.C.)

8° 750

NW

825

7°

7°W

35° 35 (a)

SE G.E. Channel Acoustically transparent sediments in levee-like flank 2 km

900 975 High amplitude reflections beneath channel axis

Depth (m)

1050

825

NW Wave-covered levee Wave-covered levee G.E. Channel

(b)

SE

900 975 1050

3 km

(c)

Figure 14.7 Contourite drift and bottom-current channel in the slope of the Gulf of Cadiz. (a) High-resolution bathymetric map (acquired with a SIMRAD EM300 multi-beam echo sounder) showing a sandy slope sheeted drift swept by the Mediterranean Outflow Water (M.O.W.) and channels that funnel the downflowing M.O.W (see Mulder et al., 2003b). 1 = ‘‘free-standing’’ bottom-current channel (G.E.C. = Gil Eanes Channel: Habgood et al., 2003); 2 = channel controlled by tectonic structures (G.C. = Guadalquivir Channel; G.B. = Guadalquivir Bank). (b) Single-channel air-gun seismic section. (c) 3.5 kHz profiles across the Gil Eanes Channel, showing contouritic lateral levees built by the M.O.W. downcurrent funnelled in the channel (from Habgood et al., 2003).

which the principal flow is focused. Erosion and non-deposition are dominant near the current axis, while deposition occurs laterally, where the velocity decreases (to the left of the current on the northern hemisphere, and to the right on the southern hemisphere). The location of drifts with respect to the current axis can, however, be variously affected by other factors, including interacting currents and topographic obstacles. Such drifts show an up-slope lateral migration marked by oblique or sigmoidal reflector patterns in seismic lines.

Mounded drifts:

migration and aggradation any type of reflections, except horizontal/parallel reflections Low current speed gradient

Plastered drift - along-slope migration (downstream of the current flow)

Giant elongated drifts

- down-and up-slope migration

Gentle slope

Example: Gardar drift

Separated drift

Moat channel

- along-slope migration

Drift levee

(downstream of the current flow)

High current speed gradient

Steep slope with a slope break

- up-slope migration E.g. Faro drift

200 m

Detached drift - predominant down-slope migration

2000 m

Confined drifts

channel-related drifts

Example: Eirik drift

- predominant down-current migration

Channel

Channel

Channel

- random lateral migration Example: Vema contouritic fan

Downstream of a deep channel issue

Channel

- predominant down-current migration

Channel

- limited lateral migration Example: Sumba drift

In between high tectonic or volcanic reliefs

Figure 14.8 Summary of the different types of mounded contourite drift (see McCave and Tucholke, 1986; Fauge' res et al., 1993, 1999; Stow et al., 2002d), showing the drift general geometry and trend of migration ^aggradation as well as inferred bottom-current pathways.

+ NW

SE

Lofoten Drift 75°

1.5

1000 Water depth (m) 1100

0

Paleo-moat

1 km 3°

Erosional scarp 2.0

(s) twt

2.5.

Intra Miocene VB-32–89

Figure 14.9 The Lofoten drift in the northern Norwegian Sea, a plastered contourite drift (from Laberg et al., 2001; with permission from Springer).

277

J.-C. Fauge`res and D.A.V. Stow

Turbiditic levee Perpendicular, oblique to parallel to the slope trend

Contouritic levee Parallel to the slope trend

(a)

(b)

(c) Turbidity current

Contour current

Progradation

Progradation

Progradation Example: Cap Ferret deep-sea fan

Example: Faro Drift

Turbidity current

Example: Var deep-sea fan

Figure 14.10 The relationship between the directions of contouritic and turbiditic levees and the direction of the margin along which the drifts are developed (from Fauge' res et al., 1999); black dashed arrow indicates deposit migration. (a) Usual trend for a contouritic drift. (b) Usual trend for a turbiditic levee. (c): Possible variations in direction during the growth of turbiditic levees.

Faro Drift S

N

Channel

Levee

twt (s)

1

2 Drift upslope progadation

Separated drift 1

md 3

2

M.d.

5

4

6

2 1

B.E.D.

Channel

migration

Plastered drift Sheeted drift

Substratum

5 km

Figure 14.11 The Faro Drift in the Gulf of Cadiz, a separated contourite drift: multi-channel seismic line and interpretation (modified from Fauge' res et al., 1985a). Note that the contourite sediments are deposited first as a plastered drift (units 1 and 2) and then as a separated drift (units 3^6). The separated drift consists of a ‘‘channel ^ levee’’couplet and is characterised by an up-slope direction of progradation. BED = basal erosive discontinuity of the drift; M.d. = major discontinuity between the plastered drift and the overlying separated drift; md = discontinuity bounding the major seismic units.

The steep slope along which the drift is deposited may have various origins. In the case of giant elongate drifts, the relief of the continental margin is inherited from its long tectonic and sedimentary history. Locally, steep slopes may result, however, from fault activity or erosional scars associated with major slides or slumps (Figure 14.13). These can lead to the deposition of smaller elongate drifts, which have been termed ‘‘fault-controlled drifts’’ and ‘‘infill drifts’’, respectively (Rebesco and Stow, 2001; Rebesco, 2005).

278

Contourite Drifts: Nature, Evolution and Controls

+

(a)

+ (b)

e.s.

+ (c)

e.s. +

+

(d)

Figure 14.12 The Faro drift (Gulf of Cadiz). Interpretation of the separated drift deposit geometry and sedimentary processes according to the modifications of the sea-floor morphology and variations of the bottom-current velocity (modified from Fauge'res et al., 1985a). (a) Lowvelocity current irrespective of the sea-floor morphology: draping deposit. (b) Medium to high current velocity on a flat or slightly dipping sea floor, with a fairly large gradient in the flow velocity (Coriolis Force): mostly deposition with gently downlapping reflectors; the moundeddrift relief is not well developed (cf. unit 3, Figure 14.11). (c) The mounded-drift relief is well developed and the currents tend to have a higher velocity, as they are funnelled in the channel; if the currents have a high velocity, oblique to sigmoid deposits prograde on the left (southern) flank of the channel, and erosion is prevailing on the channel bottom and the right (northern) flank (e.s. = erosional surface), due to the sharp current-velocity gradient (cf. unit 5, Figure 14.11). (d) Morphological pattern similar to the previous case, but unstable currents with alternating lowand high-velocity currents: the channel is partly filled by alternating downlapping and onlapping deposits, and the up-slope migration of the drift channel slows down compared to the previous case (e.s. = erosional surface; cf. unit 6, Figure14.11).

14.5.2.1.3. Detached drifts Detached drifts (Figures 14.1, 14.3, 14.4, 14.8, 14.14a, b) typically present an elongation that deviates at a larger or smaller angle from the adjacent slope against which it first began to form. Such a drift development can result from a change in the margin’s trend (Eirik Drift: Arthur et al., 1989; Hunter et al., 2007b), or from the interaction between surface and bottom currents (Cape Hatteras, Gulf Stream and Blake-Bahama Drifts: Tucholke and Laine, 1982; McCave and Tucholke, 1986). 14.5.2.2. Channel-related drifts Channel-related drifts are specifically related to narrow conduits (deep channels, gateways or contourite moats) where the bottom circulation is constrained and flow velocities consequently markedly increased. Examples include, among others, the Faeroe–Shetland Channel (Bulat and Long, 2001), the Florida Strait (Denny et al., 1994), the Kane Gap (Mienert, 1986) and the Vema Channel (Me´zerais et al., 1993) drifts in the Atlantic Ocean; the Amirante Passage drift in the Indian Ocean; the Samoan Passage and the Sand Dune Valley drifts in the Pacific Ocean (Lonsdale and Malfait, 1974; Lonsdale, 1981;

279

J.-C. Fauge`res and D.A.V. Stow

1.1

Erosional scar

Faeroe contourite drift

.

1.2

1.3

1.4

1.5

Slump deposits

twt (s)

1.6 2200 m

Unconformity

1.7

Figure 14.13 The Faeroe separated drift on the southern margin of the Norwegian Sea (from Nielsen and van Weering, 1998; Nielsen et al., 1998c, with permission from The Geological Society, London).The current-scoured channel and associated contourite levee are typical of a separated contourite drift.The location of the channel coincides with an earlier slump scar, the basal unconformity has been formed by mass movements, and the contourite deposits are overlying slump deposits.

Johnson et al., 1983). Significant erosion and scouring commonly occur on the channel floor and flanks. Irregular discontinuous sediment bodies may, however, be deposited both within the channel and at the down-current exit of the channel. 14.5.2.2.1. Axial and lateral channel-patch drifts These are typically small (a few tens of square kilometres in surface area, 10-150 m thick) contourite accumulations preserved on channel floors and flanks. They may have either a gently mounded or sheeted form capped by wavy bedforms (Figure 14.15), and be either irregular in shape or elongated in the direction of flow. They can be reflector-free or with a chaotic seismic facies very similar to debris-flow lobes and masses, but more commonly show good parallel to sub-parallel seismic reflectors. 14.5.2.2.2. Contourite fans Contourite fans (Figure 14.8; see also Fauge`res et al., 2002b) are typically much larger fan-shaped deposits deposited downstream of the channel exit. A good example is the Vema contourite fan, in the South Brazil Basin (Me´zerais et al., 1993; Fauge`res et al., 2002c). This drift is up to 100 km or more in width and radius, and 300 m in thickness. It is composed of an aggradation of flat irregular lenticular depositional units of limited extent, which are sedimentary relicts, bounded by major erosional surfaces. There is little clear or consistent evidence of migration, although the topmost unit normally shows down-flow progradation. As such it is similar to some small- and medium-sized turbidite fans, and may even contain distinct channel–overbank units. Their relative thin character, in combination with laterally extensive erosional discontinuities, can help distinguish them from purely turbiditic systems (Fauge`res et al., 1998).

280

Contouritic Levees Perpendicular to the slope trend

(a)

Contour currents along a margin corner

(b)

Interactive surface and deep contour currents

0m

(c) 0m –200

Along-slope contour current (c.t.) Interfering with turbiditic currents (t.c.)

(d)

Contour current Downwelling

–3000

–200

c.t –2000

t.c. Bott

om c curr ontour ent

Example: Eirik drift

Figure 14.14

Example: Blake–Bahama drift

t.c.

t.c.

Example: Bellingshausen Antarctic margin mixed turbiditic and contouritic levees

Example: Gulf of Cadiz slope contouritic levees

Various scenarios for contouritic levees perpendicular to the slope direction (modified from Fauge' res et al., 1999).

Contourite Drifts: Nature, Evolution and Controls

Surfacet curren

281

J.-C. Fauge`res and D.A.V. Stow

Elongate contourite mounds

10 km

A

A′

NW

SE

Moats Elongate contourite mounds

(Record missing)

100 ms

0

Sediment waves

100

m

120

00

14

0

1 km

(a)

(b)

Figure 14.15 The Faroe ^Shetland Channel (Bulat and Long, 2001; with permission from Springer). (a) Image performed from a 3-D seismic data set showing the lower part of the eastern flank and the bottom of the channel (1000^1500 m water depth). Note the current bedforms interpreted as elongate contourite mounds and sediment waves, and the interference pattern between these bedforms oriented NE ^SWand NW^SE, respectively. (b) Seismic line crossing the elongate mounds: these bedforms have been interpreted as built by an along-slope northeastward flowing current (profile location in a). A multicolour version of this figure is on the enclosed CD-ROM.

14.5.2.3. Confined drifts This drift type (Figures 14.8 and 14.16) is characterised by a mounded geometry elongated parallel to the axis of a relatively small confining basin or passage where fairly slow contour currents are flowing. They have been described with distinct contourite moats along both flanks, suggesting that flow is confined on both margins, or perhaps develops into some kind of circulatory pattern within the basin. Relatively few examples are currently known of such drifts, typically within morphotectonically active areas, such as the northern Corsica Basin (Roveri, 2002) and Sicilian gateway (Reeder et al., 2002) in the Mediterranean Sea, the Louisville Drift in the deep part of the eastern New Zealand Margin (Carter and McCave, 1994), the Sumba Drift in the Sumba forearc basin of the Indonesian arc system (Reed et al., 1987), the Meiji Drift in the Aleutian Trench (Scholl et al., 1977) and an unnamed drift in the Falkland Trough (Cunningham and Barker, 1996; Cunningham et al., 2002). Apart from their topographic confinement, the gross seismic character appears similar to elongatemounded drifts. They may show a complex stacking of convex-upward lenticular depositional units, partly in relation to active basin subsidence.

14.5.3.

Mixed drift systems

Mixed drift systems are those that involve the significant interaction of along-slope contour currents with other depositional processes in the building of the drift body. As mentioned above (Sections 14.2.5 and 14.4.3), sediment supply to contour currents is

282

Contourite Drifts: Nature, Evolution and Controls

(a) N

NE

Sill Gap

e nel

r

nda

Bou

ge

or

an y Ch

sin

Ba umba

S

dary

Boun

Rid

-Tim

rust

u Th

a Ridg

Sumb

Sawu

Basin

Saw

wu Sa

nel

Chan

a1 1

N

S Deformed strata along Sawu_Timor ridge

Sumba Basin

Boundary channel

Mud diapir

Buried boundary channel

2 ene lioc ep cen -Mio Mid

Sawu thrust

2-way (s)

3

Sumba ridge

a2

10 km

Sicily Gateway Mounded drift

0.5 s

Boundary Channel

Boundary Channel

5 km

(b)

Figure 14.16 Confined drifts. (a) In a forearc basin (Sumba drift, modified from Reed et al., 1987). a1 = schematic diagram showing the morpho-structural and oceanic circulation background of the drift; a2 = water-gun seismic profile crossing the drift. (b) In a large gateway (Sicily Gateway, from Reeder et al., 2002; with permission of The Geological Society, London).

variously influenced by pelagic, hemipelagic, glaciomarine and turbidity-current input. Where pelagic/hemipelagic input is particularly significant, the drift system will tend to be less pronounced morphologically – many of the sheeted drifts are probably of this type, although they are not generally termed ‘‘mixed drifts’’ because of the inherent difficulty in distinguishing between the processes involved. Particularly

J.-C. Fauge`res and D.A.V. Stow

283

close interaction with shallow-water processes, in the outer-shelf to upper-slope environment, leads to a variety of shallow-water drifts or mixed drifts (Viana et al., 1998a, 2002a; Verdicchio and Trincardi, 2008a). Interaction with glaciomarine processes at high latitudes may be clear from the sediment facies (scattered and winnowed ice-rafted debris), but does not generally produce distinct mixed drift types (see Van Weering et al., 2008). However, where glaciomarine input has been instrumental in triggering significant down-slope sediment movement, the resulting mixed drift systems are characterised by both down-slope and along-slope processes, and hence the normal contourite-drift morphologies and development may be markedly modified. Such down-slope/along-slope mixed drifts have been described from a number of locations, including the high-latitude Antarctic Margin (a.o. Rebesco et al., 1996, 1997, 2002, 2007), the middle-latitude NW European Margin (Armishaw et al., 1998, 2000; Stow et al., 2002b), the middle US Atlantic Margin, east off Cape Hatteras (a.o. Tucholke and Mountain, 1986; McMaster et al., 1989; Locker and Laine, 1992), the eastern New Zealand Margin (Carter and McCave, 1994, 2002; Shipboard Scientific Party, 1999a; Carter et al., 2004) and the low-latitude east Brazilian Margin (Masse´ et al., 1998; Fauge`res et al., 2002b). The variable characteristics of these drifts are outlined at more length in Mulder et al. (2008).

14.6.

D ISCUSSION

Although the classification system outlined above neatly categorizes different drift types, it is important to emphasize again that the distinctive morphologies described are simply type members within a continuous spectrum. Looking at the drift morphology and overall deposit geometry, a high spatial and temporal variability can be found at the scale of the same drift. This results primarily from the depositional background and controlling factors (which change through time), and secondly from the influence of the drift relief, which is constantly growing and evolving in shape, thereby inducing changes in current flow and associated sedimentary processes during drift building.

14.6.1.

Spatial evolution of drifts

Spatial morphological variations are exemplified by the Feni drift along the western and southern margins of the Rockall Trough (Figure 14.17). Bathymetric data (Kidd and Hill, 1986) show that this sediment body is initiated as a slope-plastered sheeted drift at its northern location. Further south, the drift evolves into a separated, elongated mounded drift, then becomes a plastered sheeted type again and, finally, just before the northwestern bend in drift trend, there is a complex section including formation of a secondary crest between the moat and the main drift to the east. These changes are believed to be due to variations in the slope gradient and to the fact that several superimposed contour currents are involved in the sediment deposition (Dickson and Kidd, 1986; Stoker et al., 1998a). Similar morphological drift variations are also observed along the Hebrides slope in the northern Rockall Trough (Stoker, 1998b).

284

Contourite Drifts: Nature, Evolution and Controls

W

E

(a)

Drift crest (b)

55°N (c)

(d)

(e)

52°N

Secondary crest

(f)

(a)

(b)

Figure 14.17 The Feni drift (modified from Kidd and Hill, 1986). (a) Bathymetric map of the eastern margin of the Rockall Plateau and Feni drift and location of the bathymetric crosssections [a ^ f] seismic lines (A, B, C) and DSDP site 610. (b) Bathymetric cross-sections showing the morphological evolution of the drift with profiles typical of plastered sheeted or plastered drift (sections a, e) and separated drift (sections b ^ d).

14.6.2.

Evolution of drifts in time

Drift evolution at a geological timescale and the complexity of the evolution controls may be illustrated by the Faro Drift (Fauge`res et al., 1984; 1985a) along the northern margin of the Gulf of Cadiz (Figures 14.11, 14.12). This drift is of medium size (50 km long, 10–25 km wide) and was constructed during the Pliocene–Quaternary on a large plateau located at the foot of a steeply faulted upper slope. Its development began following the opening of the Gibraltar Strait an important oceanographic event recorded along the whole margin by a major erosional surface on which lies the Faro Drift. The first stage of contourite deposition involved deposition of a sheeted drift (units 1 and 2; Figure 14.11) onlapping the continental slope. The second stage involved upward growth into a low-relief slope-plastered drift (units 3 and 4; Figures 14.11 and 14.12b) with gently downlapping reflections, up-slope migration and the beginning of moat development (non-deposition). The third stage involved its evolution into an elongated mounded and separated drift (units 5 and 6; Figures 14.11, 14.12c) with a deep moat. The drift thus became a prominent high-relief feature with sigmoidal to oblique reflectors, strongly downlapping onto the erosive moat floor. The whole system shows migration up-slope (and partly along-slope,) and significant erosion into the foot of the slope. Such drift evolution may be explained in part by increase of the contour-current velocity as the drift relief increased.

J.-C. Fauge`res and D.A.V. Stow

14.6.3.

285

Complex contourite systems

Still more marked variations in time and space are noted in the Gulf of Cadiz region, where a complex contourite depositional system has been evolving for the past 5 million years since the opening of the Gibraltar Strait at the end of the Miocene. This system is now very well known following studies over the past 25 years; many of these have been summarised (Nelson et al., 1999; Llave et al., 2001, 2007; Stow et al., 2002b; Herna´ndez-Molina et al., 2003, 2006c). Both large-scale depositional (drift) and erosional features occur in relation to a strong mid-depth bottom current that shows a general decrease in intensity from east to west along the margin, but which also divides into a number of flow strands in response to the morphotectonic framework. Depositional features include sedimentary wave fields, sedimentary lobes, mixed drifts, elongated mounded drifts (plastered and separated drifts) and sheeted drifts, whereas the principal erosive features are contourite channels, furrows, marginal valleys and moats. The range of interacting controls that have led to this complex development over the whole margin are summarised in Herna´ndez-Molina et al. (2008a). A similar complex contourite system with various types of drifts deposited at different water depth is present along the continental slope of the South Brazilian Basin, where it results from complex interaction of hydrological and morphological factors (Duarte and Viana, 2007; Figure 14.18).

14.6.4.

Distinction from turbidite systems

Contourite drifts perpendicular to the margin may also depend on down-slope and along-slope current interaction or downwelling bottom-current activity (Figures 14.14c, d). In the last case, bottom currents could be wholly responsible for the erosion of down-slope channels and, along the channel, the deposition of lateral ‘‘contourite levees’’. These may be true contourite drifts, as proposed for the Chattam Rise off NE New Zealand (Barnes, 1992, 1994) and in the Gulf of Cadiz (Fauge`res et al., 1985c), where they have been called ‘‘free-standing’’ bottom-current channels and associated levees by Habgood et al. (2003) (Figures 14.7 and 14.14d). However, this now seems less viable than a mixed turbidite/contourite system (Mulder et al., 2008). Confusion between elongate-mounded drifts and turbidite levees (Figure 14.10) can, therefore, occur on several counts: (1) both have a similar elongate mounded geometry, (2) drifts that are commonly elongated parallel to the slope, can be elongated down-slope as in the cases outlined above, (3) turbidite channel–levee systems – being usually elongated in down-slope direction (Figure 14.10b) – can, in part, be elongated along-slope (Figure 14.10c), where migration has been influenced by the Coriolis force or any morpho-tectonic control, as in the case of the Var levee (Migeon et al., 2000) and (4) true mixed turbidite/contourite levees exist (Figure 14.14c). Turbidite and contourite levees can, therefore, not always be distinguished on the basis of mounded geometry or elongation trend, except where the mounds are clearly isolated from down-slope supply, as in the case of separated drifts. Clearly, though, an along-slope orientation is typical of many contourite drifts and is an important pointer towards their interpretation. In addition, mounded contourite drifts commonly lie on a more or less flat, major erosion surface that corresponds to

286

Contourite Drifts: Nature, Evolution and Controls

Figure 14.18 The Santos Basin (southern Brazilian continental slope; from Duarte and Viana, 2007, with permission from The Gological Society, London). Note the occurrence of two superimposed channel (moat) drift systems: (1) the Santos Channel ^ Drift system located on the upper slope and associated with a northward flowing current, and (2) the Sao Paulo Channel ^ Drift system on the lower slope (Sao Paulo Plateau, where salt diapirs are active) associated with a deeper, north-flowing current. f = furrows; sw = sediment waves; scfc = salt crestal fault channels; R = Rupelian unconformity; U2 = early Miocene unconformity; U4 = mid-late Miocene unconformity; U6 = Mio ^ Pliocene unconformity.

an important hydrological event associated with the initiation of active bottom-water circulation in the area. Such basal erosion surfaces overlain by contourite deposits are found in most of the drifts of the Atlantic and Pacific Oceans (Fauge`res et al., 1999). This is not normally the case for turbidite levees.

14.6.5.

Buried contourite drifts in modern ocean successions

The different contourite-drift types that are well individualized on the bottom of the modern oceans have also been identified, with similar patterns, in sub-surface sediments, where they are buried by shelf to deep-sea deposits. Some examples (among many others) that come from various domains of the continental oceanic margins are as follows:

287

J.-C. Fauge`res and D.A.V. Stow

(a) 0

NW

twt (s)

1. in the Neogene shelf sediment prism off the SE New Zealand Margin (Fulthorpe and Carter, 1991; Carter, 2007), shallow-water drifts (25 km long and 15 km wide) that mimic separated drifts (Figure 14.19); 2. in the northeastern Australian continental slope, the Marion separated drift (Figure 14.20a), built along a carbonate platform during the Pliocene–Early Quaternary (Shipboard Scientific Party, 1991), is covered by a recent sediment drape; 3. in the eastern US continental rise, the buried Chesapeak Drift and wave field (middle Miocene to early Pliocene) is overlain by the Pliocene-Quaternary deposits of the Norfolk-Washington fan (Mountain and Tucholke, 1985; Locker and Laine, 1992); 4. on the eastern New Zealand Margin (Figure 14.20b), the Chatham separated drift (Wood and Davy, 1994) was deposited during the Neogene and buried by recent onlapping sediments that could be associated to bottom currents of reduced activity and/or gravity currents, and the south moat filled up by probably gravity sediments (debris-flow and turbidite);

Pink

1

SE

Yellow Green

2

0

Landw ard ch

annel

0 NW

20 km

migrati

on

0

20 km

SE

twt (s)

1

(b) 2

Pink Yellow Green

Figure 14.19 Buried shallow-water drifts in the Neogene shelf sediment prism of the SE New Zealand Margin (modified from Fulthorpe and Carter, 1991; with permission from the Geological Society of America; see also Carter, 2007). (a) Isolated shallow-water drift showing a clear northwestward migration. (b) Superimposed shallow-water drift showing a similar migration direction.

288

Contourite Drifts: Nature, Evolution and Controls

NE 0

1 km

SW

Marion Drift

N

0.7 3

3

0

Chatham Drift

10 km

Drift up-slope Onlapping post-drift migration Onlap fill (g?) df sediments

S

t w t (s)

1 4 2

5

1

4

1.0

(b) (a) 5 km

Florida Drift

Drape Buried moats Last Buried Drifts buried moat

t w t (s)

1

0

N

NW 3.5

Sao Paulo Plateau Buried Drift Moat

t w t (s)

S

Furrows

SW Buried Drift Recent drape

4.0

2

2 1

4.5 Salt

(d)

1 km

(c)

Figure 14.20 Buried drifts. (a) The Marion buried separated drift on the northeastern Australian Margin (modified from Shipboard Scientific Party, 1991); the drift (black arrow 4) was built along a carbonate platform (black arrow 5) by contour currents flowing northwestward. Note: (1) narrow and deep channels suggesting strong current activity; (2) is a large and shallow channel related to slow currents; (3) draping surficial deposits could be associated with a strong deceleration of the bottom currents as soon as the platform relief has been overflown. (b) Northeastern Chatham Drift on the eastern New Zealand Margin (air-gun seismic line; modified from Wood and Davy, 1994). Up-slope migrating drift buried by onlapping sediments possibily associated with bottom currents of reduced activity and/or gravity currents, and moat infill probably due to gravity processes (CEN1 and CEN2 = late Oligocene, Miocene to Recent successions; g? = possible gravity-flow deposits; d.f. = : debris-flow). (c) Sao Paulo Plateau buried drifts (modified from Viana, 2001; with permission from Springer). Note the two generations of buried drifts built along a diapiric relief and covered by a recent sediment drape (1 = erosional basal discontinuity; 2 = buried moat). (d) Florida Drift in the Florida Strait (modified from Denny et al., 1994). Note the up-slope migration of two superimposed buried moat ^ drift systems (1 to 2), the seaward shift of the most recent system (3 = last moat axis), following a change in the slope morphology (erosion), and the modern sediment drape.

5. in the Brazilian Basin (Sao Paulo Plateau), two generations of superimposed drifts that fill a depression between diapiric reliefs have been described by Viana (2001) (Figure 14.20c); 6. in the Florida Strait (Figure 14.20d), successive Neogene moat–drift systems are buried below modern sediments along the northern flank of the strait (Denny et al., 1994).

ACKNOWLEDGEMENTS The authors thank E. Gonthier and A. Viana for valuable discussions and useful comments. They are also grateful to Thierry Mulder for his helpful critical review and to M. Rebesco, A. Camerlenghi and T. van Loon that suggest many improvements for the final version of this chapter.

C H A P T E R

1 5

S EDIMENT W AVES AND B EDFORMS R.B. Wynn and D.G. Masson National Oceanography Centre, Southampton (NOCS), Southampton, UK

Contents 15.1. 15.2. 15.3. 15.4. 15.5.

Introduction Location, Morphology and Genesis of Fine-Grained Sediment Waves Location, Morphology and Genesis of Coarse-Grained Sediment Waves Related Large-Scale Features Generated by Bottom Currents Applications to Bottom-Current Reconstruction: A Case Study From the NW UK Acknowledgements

15.1.

289 290 294 296 298 300

INTRODUCTION

Large-scale sediment waves are some of the most distinctive and frequently described depositional features generated by bottom currents, and can cover huge areas of sea floor (>1000 km2). They occur in a wide range of deep-water environments (Figure 15.1) and are highly variable in terms of their morphology, dimensions and sediment composition. Wynn et al. (2000) defined a sediment wave as ‘‘a large-scale (generally tens of metres to a few kilometres wavelength and several metres high), undulating, depositional bedform, generated beneath a current flowing at, or close to, the sea-f loor’’. An overview of deep-water sediment waves, generated by both bottom currents and turbidity currents, can be found in Wynn and Stow (2002b). The present study will build upon that review by incorporating recently published data, and will have the following specific aims: 1. to outline the morphology, genesis, identification and depositional environment of both fine- and coarse-grained sediment waves formed by bottom currents; 2. to describe and illustrate some of the related large-scale features generated by bottom currents, e.g. sand ribbons and erosional furrows; 3. to investigate the application of data obtained from sediment waves and related features to bottom-current reconstruction. Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00215-X

Ó 2008 Elsevier B.V. All rights reserved.

289

290

Sediment Waves and Bedforms

Figure 15.1 Global bathymetric map showing location of well-documented examples of bottom-current sediment waves (white circles), and location of the case-study area (black rectangle; Figure 15.2). Numbered white circles refer to the following areas: (1) Norwegian Sea ^Atlantic Ocean gateway (e.g. Roberts and Kidd, 1979; Dorn and Werner, 1993; Manley and Caress, 1994; Howe, 1996; Kuijpers et al., 2002; Masson et al., 2002; Wynn et al., 2002a; Howe et al., 2006); (2) Mediterranean Sea ^Atlantic Ocean gateway (Kenyon and Belderson, 1973; Habgood et al., 2003); (3) Central Mediterranean Sea (Marani et al., 1993); (4) Blake ^ Bahama Outer Ridge (Flood, 1994; Flood and Giosan, 2002); (5) Gulf of Mexico (Kenyon et al., 2002); (6) Carnegie Ridge (Lonsdale and Malfait,1974); (7) Argentine Basin (e.g. Flood and Shor,1988; Flood et al.,1993; Manley and Flood,1993a, b;Von Lom-Keil et al., 2002); (8) Antarctic ^Atlantic Ocean gateway (Cunningham and Barker, 1996; Howe et al., 1998); (9) Chatham Ridge (e.g. Lewis and Pantin, 2002). Base bathymetric map obtained from USGS website: http:// walrus.wr.usgs.gov/infobank/gazette/html/bathymetry/gl.html. A multicolour version of this figure is on the enclosed CD-ROM.

Although bottom-current sediment waves are geographically widespread (Figure 15.1), this study will focus only on case studies from the NW UK continental margin (Figures 15.1 and 15.2), as this is a well-studied region with abundant high-quality data.

15.2.

L OCATION, MORPHOLOGY AND G ENESIS OF F INE -G RAINED S EDIMENT W AVES

Fine-grained sediment waves generated by bottom currents are generally found draping the flanks and crests of sediment drifts (Figures 15.3 and 15.4) (Fauge`res et al., 1999), in basinal and lower continental-rise environments (Roberts and Kidd, 1979; Richards et al., 1987; Flood and Shor, 1988; Flood et al., 1993; Marani et al., 1993; Manley and Flood, 1993a, b; Flood, 1994; Manley and Caress, 1994; Cunningham and Barker, 1996; Howe, 1996; Howe et al., 1998; Flood and Giosan, 2002; Lewis and Pantin, 2002; Masson et al., 2002; Von Lom-Kiel et al., 2002; Howe et al., 2006).

R.B. Wynn and D.G. Masson

291

Figure 15.2 Location map of case-study area (for location: see Figure 15.1). Modified from Masson et al. (2002); with permission from Elsevier. Surface and deep-water circulation shown by filled and open arrows, respectively. Grey-shaded arrow illustrates inferred path of intermediate depth water deflected to the west by the Wyville-Thomson Ridge. Contour intervals: 200 m. Locations of Figures 15.3, 15.4, 15.8 and 15.9 are indicated.

Fine-grained sediment waves are dominantly composed of mud, silt and fine sand (often poorly sorted), and may be siliciclastic, volcaniclastic and/or bioclastic in composition (Stow et al., 1998a; Wynn and Stow, 2002b). They typically show features diagnostic of contourite deposition, e.g. poorly developed laminae and intense bioturbation resulting from steady quasi-continuous sedimentation (Stow and Lovell, 1979). Wave dimensions are impressive, with wave crests often >10 km in length, wave heights of up to 50 m (and occasionally 150 m: Flood et al., 1993), and wavelengths ranging from 1 to 10 km (Figures 15.3 and 15.4). Wave dimensions appear to be related to flow velocity and/or sedimentation rate, and often decrease towards the margin of wave fields where bottom-current influence is reduced (Flood and Shor, 1988; Cunningham and Barker, 1996). In plan view, wave crests usually appear straight or slightly sinuous, and bifurcation is rare (Figure 15.4a). This planform geometry is a reflection of the typically unidirectional, steady bottom-current flow responsible for generating the waves.

292

Sediment Waves and Bedforms

NW (a) 880 1 km

SE Sediment waves

Elongate drift Moat

940 m 1000 1060

970

(b)

1 km

Wave-migration direction

1000 m 1030

1060

Figure 15.3 Profiles across the f lank of an elongate drift adjacent to the Hebrides Slope (for location: see Figure 15.2). Images modified from Masson et al. (2002); with permission from Elsevier. (a) Deep-tow boomer profile across a small sediment wave field. (b) Enlarged image showing up-slope wave migration, resulting from asymmetric sediment deposition across the wave crest. Note the reduction in wave-migration rate during deposition of the middle transparent unit, which is interpreted to represent the last glacial lowstand.

Most examples of fine-grained bottom-current sediment waves on f lat basin f loors have wave crests aligned roughly perpendicular to the f low, and wave migration is in an upcurrent direction (i.e. opposite to flow). Waves developed on open slopes typically have wave crests aligned at a low angle (10–50°) to the flow. Wynn and Stow (2002b) concluded that, where the mean bottom-current flow is roughly slope-parallel, examples of oblique waves usually migrate in an up-slope and upcurrent direction (Figures 15.3 and 15.4). The lateral wave-migration rate is largely controlled by the sedimentation rate (Figure 15.3b), with areas of relatively high sedimentation rate displaying faster lateral migration (up to 1.0 m ka 1: Masson et al., 2002). However, it seems likely that flow velocity and/or wave-crest orientation also affect migration rates, with higher flow velocities and higher angles between wave crest and flow direction leading to higher wave-migration rates (Flood, 1988; Blumsack and Weatherly, 1989). Sediment-wave initiation is still a poorly understood process, but is thought to involve irregular deposition from a bottom current passing over a pre-existing sea-floor perturbation, e.g. the crest of a sediment drift (Howe et al., 1998; Von Lom-Keil et al., 2002; Wynn and Stow, 2002b). Various models have been proposed for wave growth and migration once the bedform is established, with the most widely accepted being the lee-wave model introduced by Flood (1988), and later modified by Blumsack and Weatherly (1989) and Hopfauf and Speiss (2001). The lee-wave model suggests that internal lee waves can develop within a weakly stratified bottom current as it passes over a sediment wave, leading to increased flow velocities on the downcurrent (lee) flank. This flow pattern leads to asymmetrical sediment deposition across the wave, with enhanced deposition on

293

R.B. Wynn and D.G. Masson

Y (a)

Sediment core locations

1 km

54 53

X Wavemigration direction

SW 1100 (b) 1 km

NE X

53 54

Y

1150 m 1200 1250

Figure 15.4 TOBI 30 kHz side-scan sonar image (a) and 3.5 kHz profile (b) crossing a sediment-wave field on the flank of a broad sheeted drift in the northern Rockall Trough. For location see Figure 15.2. Comparison between image and profile shows that high backscatter stripes (white) correspond to the lee (downcurrent) wave flanks. Low backscatter stripes (black) correspond to upcurrent wave flanks with thicker accumulations of well-sorted Holocene contourite sand (as confirmed by cores 53 and 54). X^Y locates section of profile crossing the side-scan image. Both images modified from Masson et al. (2002); with permission from Elsevier.

the upcurrent flank, leading to the observed upcurrent migration of the wave (Figures 15.3 and 15.4). The model predicts that flow velocities of approximately 0.09–0.3 m s 1 are required for active migration to occur, with aggradation at lower velocities and localised erosion or non-deposition at higher velocities. A number of studies carried out in the Argentine Basin during Project MUDWAVES successfully tested the lee-wave model against field data. Wave morphology and sediments were measured and used to predict active bottom-current flow direction and velocity, which was tested using long-term current moorings and found to be in general agreement once alongwave flow components were taken into account (Blumsack, 1993; Flood et al., 1993; Manley and Flood, 1993b; Weatherly, 1993). Wynn and Stow (2002b) outlined several criteria for distinguishing between fine-grained sediment waves generated beneath bottom currents and turbidity currents. These included regional setting, wave regularity, sediment type, crest alignment and sequence thickness. Key characteristics of bottom-current sediment waves include the following: (1) environmental location away from turbiditycurrent input, e.g. on contourite-drift flanks, (2) no consistent up- or down-slope

294

Sediment Waves and Bedforms

trend in wave dimensions, (3) contouritic sediments with associated high levels of bioturbation, (4) crests aligned oblique to bathymetric contours (when located on slopes) and (5) no consistent spatial trend in sediment-wave sequence thickness. However, it should be noted that, without integrated datasets comprising both geophysical and sedimentological data, specific identification of the wave-forming process may not always be possible, especially as some sediment wave fields are formed by a combination of both bottom-current and turbidity-current activity (Rebesco et al., 1996; Kenyon et al., 2002). Lee et al. (2002) and Wynn and Stow (2002b) also described how fine-grained sediment waves can be distinguished from soft-sediment deformation features, e.g. creep folds and extensional faults (Kenyon et al., 1978; Hill et al., 1982; Mulder and Cochonat, 1996; O’Leary and Laine, 1996; Lee and Chough, 2001; Lee et al., 2002). Soft-sediment deformation features can normally be identified on the basis of the following: (1) they do not show lateral migration, (2) individual reflectors may be difficult to trace from crest to crest, (3) they show no spatial trends in dimensions and (4) they are often discontinuous in plan view. However, again, there may be examples where multiple processes are operating, involving both soft-sediment deformation and near-bottom currents (Fauge`res et al., 2002c; Cattaneo et al., 2004).

15.3.

L OCATION, MORPHOLOGY AND G ENESIS OF C OARSE -G RAINED S EDIMENT W AVES

Coarse-grained sediment waves generated by bottom currents are relatively common in shelf environments, but in deep water they are scarcer than their fine-grained counterparts. Most published examples are located in areas of enhanced bottom-current flow such as topographic ridges or gateways between basins (Lonsdale and Malfait, 1974; Dorn and Werner, 1993; Kenyon et al., 2002; Kuijpers et al., 2002; Wynn et al., 2002a; Habgood et al., 2003). Coarse-grained sediment waves are dominantly composed of sand-sized sediments, although sampled examples are rare (Wynn et al., 2002a; Habgood et al., 2003). They can occur as linear waves, with crests aligned roughly perpendicular to the flow and spaced a few tens to hundreds of metres apart (Dorn and Werner, 1993; Habgood et al., 2003), or, more commonly, as distinctive barchan dunes up to 200 m wide and a few metres high (Figure 15.5) (Kenyon and Belderson, 1973; Lonsdale and Malfait, 1974; Lonsdale and Speiss, 1977; Kenyon, 1986; Dorn and Werner, 1993; Kenyon et al., 2002; Kuijpers et al., 2002; Wynn et al., 2002a; Habgood et al., 2003). In contrast to their fine-grained counterparts, coarse-grained barchanoid waves migrate in a downcurrent direction, in a similar fashion to subaerial barchans. Sediment is transported across the upcurrent flank and then avalanches down the lee face before moving along the barchan horns towards the tips. Barchanoid wave forms (Figure 15.5) only occur in areas of sparse sediment supply, where flow velocities exceed 0.4 m s 1, and they often sit on a coarse (gravel or sand) substrate (e.g. Kenyon et al., 2002; Wynn et al., 2002a). The bottom-current

295

R.B. Wynn and D.G. Masson

(a) Gently rippled sea-floor (straight-crested ripples) Sinuous/ linguoid ripples Linguoid ripples Sinuous/ linguoid ripples

Linguoid ripples Smooth sea-floor with faint lineations

Not to scale Accumulation of pale sediment

Linguoid ripples

Smooth sea-floor without ripples

Closely spaced ripples arranged in a fan shape

Gently rippled sea-floor with gravel patches

(b) Current flow direction

Not to scale

No current

Figure 15.5 Bottom-current flow field and ripple distribution across coarse-grained sediment waves in an active contourite system. (a) Interpreted distribution of ripple types across a typical barchanoid wave in the Faroe ^Shetland Channel, based on sea-floor video and photographs. For location see Figure 15.8. (b) Interpreted current flow over the dune surface based on ripple distribution. Arrow size schematically represents flow velocity. Both images modified fromWynn et al. (2002a); with permission from Elsevier.

f low f ield across deep-water barchans has been reconstructed by Lonsdale and Malfait (1974) and Wynn et al. (2002a), on the basis of ripple distribution across the barchan surface (Figure 15.5). Linguoid ripples on the upper barchan surface reflect higher flow velocities than the sinuous and straight-crested ripples on the barchan flanks and surrounding sea floor, while an area of smooth sea floor with no ripples is found just beyond the lee face (Figure 15.5). Further details of ripple formation and distribution in contourite systems can be found in Martin-Chivelet et al. (2008).

296

Sediment Waves and Bedforms

15.4.

R ELATED LARGE -SCALE FEATURES G ENERATED BY B OTTOM C URRENTS

Sediment waves form part of a continuum of bottom-current-generated features, with the largest being contourite drifts and sheets (Figure 15.3a) – often on the scale of tens of kilometres – and the smallest being current ripples (Figure 15.5) – usually a few centimetres across. In this section, we will examine those features that are between 1 m and 10 km in size, such as erosional scours, furrows, sand ribbons and comet marks, and that are resolvable with high-resolution geophysical instrumentation. Identification of such features can be important in the reconstruction of bottom-water flow fields, and an example from the NW UK continental margin is discussed later. Identification of certain erosional features, such as scours, is also important in deep-water geohazard assessment. These features are generated beneath currents with velocities higher than 1.0 m s 1, which may be capable of damaging sea-floor infrastructure, including pipelines and telecommunications cables. Large-scale erosional scours may be tens of metres deep and several kilometres across (Figure 15.6), and are only found in areas of topographic constriction where (a) 1 km

Scarp Sand ribbons?

Y Furrows X

Artefacts

970 m 1020

(b) X

Y 1 km Sediment drift ?

Erosional scarp

1070

Figure 15.6 Large-scale erosional features in an active contourite system (images modified from Masson et al., 2004; with permission from Blackwell Publishing) (a) TOBI 30 kHz side-scan sonar image showing part of a large-scale bottom-current scour. High backscatter is white. The image shows an irregular headwall scarp, backscatter banding parallel to the inferred flow direction (sand ribbons?) and limited development of furrows. For location see Figure 15.8. (b) 3.5 kHz profile showing a sediment drift and inferred deposition within the scour, suggesting recent infill. X^Y locates section of profile crossing the side-scan image.

R.B. Wynn and D.G. Masson

297

bottom currents are strongly focused and flow velocities reach 1.0–2.5 m s 1 (Kenyon and Belderson, 1973; Bulat and Long, 2001; Masson et al., 2004). Morphologically, they are very similar to large-scale scours found in environments swept by high-energy turbidity currents, such as channels and channel mouths (e.g. Wynn et al., 2002b). They typically display roughly rectangular or oval planform morphology, with a steep headwall and sidewalls, and a shallower downcurrent opening (Figure 15.6). Erosional furrows, in the form of relatively narrow lineations, often occur within, or adjacent to, scours (Figure 15.6); they may be several kilometres in length, a few tens of metres wide, and a few tens of centimetres deep, and are usually cut into coarse gravel and sand substrates (Masson et al., 2004). Similar features have been described from shelf environments where strong tidal currents sweep across gravel substrates (Belderson et al., 1988); they are believed to occur under flow velocities of 1.0–1.5 m s 1. Furrows cut into fine-grained cohesive sediments have also been described by Flood (1983), and occur under lower flow velocities (<0.3–0.7 m s 1). Sand ribbons are depositional bedforms that also frequently occur in association with the erosional features described above (Figure 15.6). They are up to 500 m wide and several kilometres long, and are believed to form under flow velocities of 0.75–1.5 m s 1 (Dorn and Werner, 1993; Kuijpers et al., 2002). Comet marks are found around obstacles, such as boulders, in areas of strong bottom-current flow (0.6 to more than 1.0 m s 1), and take the form of deflated zones around the upcurrent and lateral margins of the obstacle (Figure 15.7). Shadow zones or sand tails, where sands and gravel are deposited in an elongate ribbon, are found in the lee of the obstacle. The genesis of comet marks was described by Werner et al. (1980). There is some evidence to show that longer sand

Figure 15.7 Well-defined comet mark developed around a large boulder on a sand and fine-gravel sea floor in the Faroe ^Shetland Channel (modified from Masson et al., 2004; with permission from Blackwell Publishing). Black arrow indicates interpreted flow direction. Location is stationWTS13 in Figure 15.8, at a water depth of 1095 m.

298

Sediment Waves and Bedforms

tails behind obstacles are related to a higher velocity flow (Kuijpers et al., 2002). In the Faroe–Shetland Channel, sand tails vary in length from 1 to 700 m (Dorn and Werner, 1993; Kuijpers et al., 2002; Masson et al., 2004).

15.5.

APPLICATIONS TO B OTTOM -C URRENT RECONSTRUCTION: A CASE STUDY FROM THE NW UK

Recent studies by Kuijpers et al. (2002) and Masson et al. (2004) have used bedforms and related erosional features to map the distribution and strength of bottom-current flow through the Faroe–Shetland and Faroe Bank Channels (Figures 15.8 and 15.9). These channels form a broad conduit that is a major gateway for the flow of deep cold water between the Norwegian Sea and the North Atlantic (Hansen and Osterhus, 2000). Erosional scours, furrows, comet marks, barchan dunes, sand sheets/ribbons and sediment drifts have all been identified and mapped in the study area, using geophysical data combined with sea-floor photographs and sediment cores (Figure 15.8). Published data relating to bedform type and flow velocity (Table 15.1) were then used to reconstruct the bottom-current flow through this gateway (Figure 15.9).

Figure 15.8 Summary interpretation of bottom-current features in the Faroe ^Shetland and Faroe Bank Channel, based on side-scan sonar images and 3.5 kHz profiles (modified from Masson et al., 2004; with permission from Blackwell Publishing). Symbols showing surficial sediment classification are based on core/dredge samples and/or sea-floor photographs. FSC = Faroe ^Shetland Channel; FBC = Faroe Bank Channel; WTR =Wyville-Thomson Ridge. For location: see Figure 15.2. A multicolour version of this figure is on the enclosed CD-ROM.

R.B. Wynn and D.G. Masson

299

Figure 15.9 Reconstructed bottom-current flow field through the Faroe ^Shetland and Faroe Bank Channel based on the distribution of bedforms and related bottom-current-generated features shown in Figure 15.8 (modified from Masson et al., 2004; with permission from Blackwell Publishing). For location: see Figure 15.2. Red arrows indicate currents related to northeastward transport of near-surface water masses; blue arrows relate to southwestward transport of Norwegian Sea Deep Water (NSDW); orange arrows show recirculation of NSDW (and intermediate water masses?) at the southern and eastern margins of the Faroe Bank and Faroe ^Shetland Channels, respectively. Black dots show the locations of sea-floor photography stations. A multicolour version of this figure is on the enclosed CD-ROM.

The results obtained were in broad agreement with direct measurements obtained from short-term current meter deployments (Hansen and Osterhus, 2000), but with the advantage that they represent longer timescales and are more spatially extensive. Masson et al. (2004) discussed in detail the timescales at which the described features may be operating, and concluded that larger features, such as fine-grained sediment waves and drifts, may require thousands to a few million years to form, whereas small-scale features such as current ripples may only record the last significant event, e.g. a benthic storm. It is thought that most bedforms and related features represent the peak current flowing across an area. In the Faroe–Shetland and Faroe Bank Channels, the fresh appearance of most observed features (Figures 15.5–15.7) was provided as evidence for active bottom currents at the present-day. The high-velocity core of southwest-directed Norwegian Sea Deep Water (NSDW) travels across a sill at the northern edge of the Faroe–Shetland Channel before being directed by Coriolis forces along the western channel margin at water depths of 800–1200 m (Figure 15.9). In this area, strong bottom currents produce numerous erosional scours, scarps and furrows, indicating flow velocities in excess of 1.0–1.5 m s 1 (Kuijpers et al., 2002; Masson et al., 2004). On the opposite side of the channel, at equivalent water depths, small-scale features, including comet marks, show that the flow was actually to the northeast with velocities mostly in the range of 0.4–1.0 m s 1 (Figure 15.9); this flow is interpreted to result from a recirculation cell within the NSDW (Masson et al., 2004).

300

Sediment Waves and Bedforms

Table 15.1 Bottom-current velocities associated with various types of bedforms observed in the deep ocean Bedform

Sediment type

Peak current velocity (m s 1)

Reference

Sediment waves, contourite drifts

Fine-grained, often pelagic/ hemipelagic Silt

0.05–0.2

Manley and Flood (1993a, b)

0.05–0.15

Fine-grained, cohesive mud Sand, often rippled

<0.3 0.7 0.3–0.4

Foraminiferal sand Clastic sand

>0.3 0.4–1.0

Comet marks

Sand/gravel lag

0.6 to >1.0

Sand ribbons

Sand

0.75–1.5

Furrows

Gravel

1.0 to >1.5

Erosional scours

Gravel, rock

1.0–2.5

Hollister and McCave (1984) Flood (1983) Belderson et al. (1982) Southard and Boguchwal (1990) Baas (1999) Masson (2001) Lonsdale and Malfait (1974) Kenyon and Belderson (1973) Kenyon (1970, 1986) Kuijpers et al. (2002) Kenyon (1986) Belderson et al. (1982) Kuijpers et al. (2002) Kenyon (1970) Kenyon and Belderson (1973) Belderson et al. (1988) Flood (1983) Belderson et al. (1988) Belderson et al. (1982) Kenyon and Belderson (1973) Belderson et al. (1982)

Lineations Furrows Contourite sheet

Barchan dunes

ACKNOWLEDGEMENTS This study contains data obtained from a large number of scientific research cruises, and we thank the scientists, officers and crews for their assistance in data collection and subsequent processing and interpretation. We are grateful to Peter Talling and Michael Frenz for their review comments.

C H A P T E R

1 6

S EISMIC E XPRESSION OF C ONTOURITE D EPOSITIONAL S YSTEMS T. Nielsen, P.C. Knutz and A. Kuijpers Geological Survey of Denmark and Greenland (GEUS), Øster Voldgade, Copenhagen, Denmark

Contents 16.1. Introduction 16.2. Seismic Identification and Characteristics of Contourites 16.2.1. Large scale – first-order seismic elements 16.2.2. Medium scale – second-order seismic elements 16.2.3. Small scale – third-order seismic elements 16.3. Seismic Methods and Interpretation Concepts in Contourite Studies 16.3.1. Seismic scale 16.3.2. Reflection-seismic methods 16.3.3. Seismic mapping 16.3.4. Seismic interpretation 16.4. Summary Acknowledgements

16.1.

301 302 302 304 305 305 305 308 314 315 320 321

INTRODUCTION

Since the early days of recognition of contourites in the marine sedimentary record, reflection seismics has been used to identify and map these deposits, and this is now considered as a standard method in most contourite studies. In fact, the initial recognition of a contourite deposit in the marine setting is most often by means of seismics because the geometry of these deposits is an important diagnostic criterion. Throughout the last decades, reflection-seismic investigations of contourite deposits have benefited greatly from the increased interest in deep-water areas by the petroleum industry, which has lead to improved quality of the seismic data and to sophisticated interpretation techniques. A full reflection-seismic study of contourite deposits encompasses both seismic identification based on individual seismic profiles and more accurate mapping of the geometry of the deposits using either a 2-D grid or a 3-D volume of seismic data. Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00216-1

Ó 2008 Published by Elsevier B.V.

301

302

Seismic Expression of Contourite Depositional Systems

Further objectives of seismic studies of contourites are the prediction of the lithology and the reconstruction of the geological and (palaeo)oceanographic history. Clearly, this does not only pose demands to the interpreter, but also to the quality of the seismic data and its tie to borehole information. In this chapter, we give an overview of reflection-seismic characteristics useful for the identification of contourite deposits (see also Howe, 2008), followed by a review of some general acoustic and physical properties relevant to seismic studies of contourites. In addition, the chapter treats the significance of the seismic source, acquisition and processing parameters and the seismic-to-geology conversion. Interpretation concepts and terms used in seismic contourite studies are also discussed. From this, some guidelines and a summary of the seismic expression of contourites are presented.

16.2.

SEISMIC I DENTIFICATION AND C HARACTERISTICS OF C ONTOURITES

The seismic appearance of large, current-controlled sediment deposits in relation to bottom-current regime and slope and basin morphology has previously been described by Jones et al. (1970), McCave and Tucholke (1986), Kidd and Hill (1986), Eiken and Hinz (1993), Howe et al. (1994) and Stoker et al. (1998a). In 1999, Fauge`res et al. introduced a triple-scale approach that involves a set of key attributes to be used in the seismic definition of contourite drifts. This approach has later been expanded by other authors (e.g. Rebesco and Stow, 2001; Stow et al., 2002c). These seismic criteria, at times slightly modified, are summarised and discussed below. To place the criteria into a seismic-scale context, we further correlate the key attributes to a hierarchic range of elements that we call ‘‘seismic elements’’. We suggest dividing the seismic characterisation of contourites into ‘‘orders of seismic elements’’, where increasing order numbers refer to an increasing level of seismic interpretation details. Accordingly, the ‘‘first-order seismic element’’ refers to the overall drift geometry, while, for instance, ‘‘third-order seismic elements’’ refer to internal drift stratigraphic details. In Figure 16.1, we propose a conceptual model based on the type (1) and (2) drift systems of Fauge`res and Stow (2008) (see below) illustrating the case of increasing bottom-current activity along a continental margin. The model considers the whole sedimentdrift accumulation, but is independent of the spatial dimension and thus valid for scales of tens to thousands of metres of thickness and length.

16.2.1.

Large scale – first-order seismic elements

The seismic identification of a contourite deposit at the large scale is based on the overall architecture of the drift depositional system, i.e. the external geometry of the drift, the lower and upper boundaries confining the drift system, and the configuration of larger internal seismic units.

303

T. Nielsen et al.

Mounded elongate drift (first-order seismic element)

Large-scale depositional units Post-drift sediments

(first-order seismic element)

Depositional sub-units (second-order seismic element)

Seismic facies (third-order seismic element)

Moat

Large unconformities (first-order seismic element)

Pre-drift sediments Sheeted frift (first-order seismic element)

Figure 16.1 Conceptual model of principal seismic characteristics of contourite deposits. (Modified from Stow et al., 2002c, reproduced with permission from the Geological Society Publishing House.)

16.2.1.1. External geometry Based on their overall morphology, contourite drifts are either sheeted or mounded. They can be further classified based on different geological and oceanographic settings, particularly the mounded ones. Today, drift systems are grouped into five main types (Fauge`res et al., 1999): (1) sheeted drifts, (2) giant elongated drifts, (3) channel-related drifts, (4) confined drifts and (5) mixed drift systems. The characteristics of the different drift systems are dealt with in detail in Fauge`res and Stow (2008). It can be difficult to distinguish a contourite drift system from other deep-sea deposits, but a guiding characteristic is that contourite deposits form elongated along-slope geometries that follow the direction of geostrophic bottom currents. In contrast, turbidite–fan systems are driven by gravity-induced mass transport, favouring elongation in a down-slope direction. However, the two end-member processes may interact to form more complex sediment patterns on continental slopes (Mulder et al., 2008), which in some cases cannot be separated solely on the basis of the geometry of the deposits. 16.2.1.2. Bounding reflectors The proximal part of a contourite drift deposit normally records a major change in the depositional style from a non-current-dominated to a current-dominated regime, and vice versa when the contourite deposition ceases. These changes result in regional unconformities that confine the upper and lower boundaries of the drift as a whole. The basal unconformity is commonly revealed by a continuous highamplitude reflector of semi-regional to regional extent, that may extend beyond the limits of the drift system, and which represents non-deposition or erosion produced by strong along-slope currents with average current velocities well above the

304

Seismic Expression of Contourite Depositional Systems

threshold for deposition (unconformities that have developed more locally and display deep incision into the substratum are more likely the product of slide scars or erosion by turbidite channels). Internally, reflectors often display low-angle downlap onto the basal unconformity. If the drift is buried, the upper bounding unconformity is also characterised by a continuous high-amplitude reflector of semi-regional to regional extent. If the contourite drift is still active, the sea-bed reflector forms the upper boundary. 16.2.1.3. Gross internal character The overall internal seismic character of a drift as a whole is that of a uniform pattern of continuous, low- to medium-amplitude reflectors that tend to follow the gross drift morphology. This pattern reflects the long-lasting, stable conditions that are a prerequisite for building up a large contourite drift. However, large temporary changes in current strength and sediment supply may occur during the lifetime of the contourite, causing shifts between erosional and depositional environments. Such major shifts are revealed by continuous, high- to moderate-amplitude internal reflectors that are either unconformable or conformable, and which bound largescale, first-order seismic units within the drift (Figure 16.1). The shape of these first-order units tends, however, to follow the overall geometry of the drift, indicating the temporary character of the changes in the depositional regime.

16.2.2.

Medium scale – second-order seismic elements

In terms of seismic interpretation, the ‘‘medium scale’’ deals with the internal drift architecture, i.e. the internal character of the large-scale units recognised as firstorder seismic elements. These are composed of second-order seismic sub-units, commonly displaying: • a lens-shaped, upward-convex geometry; • a more or less uniform stacking pattern (reflects long intervals of relatively stable conditions typical for most contourites); • a progradational stacking pattern that shows migration in a down-current direction or an aggrading stacking pattern (the latter being most common for sheeted drifts); • downlapping reflector terminations (toplapping might occur in connection to internal erosional unconformities). While first-order seismic units reflect larger temporary changes in the depositional environment, the presence of the second-order seismic sub-units results from smaller fluctuations causing variations in sediment characteristics like composition, homogeneity, compaction, bedding, biotubation and similar aspects, as described by, among others, Stow and Fauge`res (2008). Second-order seismic sub-units can be observed independent of the resolution of the seismic profile (see below) (Laberg et al., 2001, 2002), showing that minor fluctuations in the depositional regime in many cases are more frequent than revealed from conventional multi-channel seismic data only.

305

T. Nielsen et al.

16.2.3.

Small scale – third-order seismic elements

Once the second-order seismic sub-units have been established at the mediumscale level, examination of their internal acoustic character can be performed on the basis of seismic facies and seismic attribute analysis. Using seismic facies analyses, information can be gained on the gross lithology and depositional environment of the individual sub-units. As for the contourite sedimentary facies (Stow and Fauge`res, 2008), no seismic facies or attribute is unique for contourites. Furthermore, the seismic facies and attributes depend strongly on the seismic acquisition and processing parameters. Hence they do not provide ‘‘stand-alone’’ diagnostic tools, but may be useful in combination with geometry of seismic units and correlation with core information. The most common seismic facies configurations in contourite drifts are: • continuous, (sub)parallel reflection configurations; • wavy reflection configurations; • structureless or reflection-free configurations. Seismic attributes can be used to map current-induced bedforms (sediment waves and ripples, erosional furrows, moats and channels, etc.), which in turn can be related to different current regimes and sediment types. Attribute analysis can also help to extract information on lithology by providing information on amplitude, phase and dip-azimuth of a seismic reflector. In general, seismic facies analysis are used to interpret variations in the contourite depositional environment, but a few attempts have been made to correlate the seismic facies to specific sediment facies observed in cored sections (e.g. Knutz et al., 2002a; Stow et al., 2002c). More recently, seismic attribute analysis has been introduced also in contourite studies to interpret the depositional environment and sedimentary facies (e.g. Knutz and Cartwright, 2004; Hohbein and Cartwright, 2006; Viana et al., 2007). However, an unequivocal correlation between contourite sediment facies and seismic facies or attributes does yet not exist and more work on this topic is needed.

16.3. 16.3.1.

SEISMIC M ETHODS AND INTERPRETATION C ONCEPTS IN C ONTOURITE S TUDIES

Seismic scale

For correct interpretation of seismic data, it is important to acknowledge the difference between a seismic profile and a geological profile. Hence the seismic reflectors do not uniquely correspond to actual bed interfaces. Moreover, the horizontal scale on a conventional seismic profile is displayed in the metric system, while the vertical scale is displayed as two-way travel times (twtt). Thus, seismic profiles tend to be highly exaggerated on the vertical scale, leading to distortion of thickness and dip of layers.

306

Seismic Expression of Contourite Depositional Systems

16.3.1.1. Seismic reflectors In sedimentological terms, a layer is defined by the lithology, whereas seismically a unit is defined by its bulk density and the velocity of sound propagation. Changes in the acoustic impedance (velocity multiplied by density; Sheriff, 1984) will cause the seismic signal to be reflected, and the strength of the acoustic impedance contrast between two units determines the amplitude of the reflected signal. Thus, the appearance and character of a seismic reflector is the result of changes in the physical properties of the strata. However, changes in these properties usually, but not always, correspond to lithological contrasts, which are often determined by changes in the depositional conditions. Possible causes of such chances in a contourite environment are numerous, but they are generally related to bottom-current strength (controlling sedimentation rate and flux, non-deposition or erosion) and sediment sources (Stow and Fauge`res, 2008). The most confident way to determine the origin of a reflector evidently is by sediment sampling, or ground-truthing. 16.3.1.2. Resolution and penetration Seismic resolution is the key to extraction of stratigraphic details from the seismic data. Seismic resolution comprises two aspects: vertical and horizontal resolution. Vertical resolution refers to the minimum thickness of a layer that can be distinguished on the seismic profile, whereas the horizontal resolution is the ability to recognise two laterally separated features as two distinct reflections. The seismic resolution is proportional to the wavelength of the seismic signal, and thus to the frequency component of the seismic source (wavelength equals velocity divided by frequency; Sheriff, 1984). The seismic resolution is therefore tied to the type of seismic system used to attain the data. As a rough estimate, both the vertical and horizontal resolutions correspond to 1/4 of the dominant wavelength of the seismic signal. Thus, the higher the frequency of the seismic source, the higher the resulting seismic resolution, and vice versa. Accordingly, the resolution will decrease with increasing penetration depth because the earth filters the highest frequencies, while the seismic velocity increases (Telford et al., 1976). The seismic resolution is therefore not fixed, but corresponds to different depth levels (Figure 16.2). This is particular relevant in seismic studies of very thick or deeply buried contourite deposits. The seismic penetration is expressed as the maximum depth from which seismic reflections can be picked with reasonable certainty (Sheriff, 1984). The penetration depends on the size of the seismic source (the bigger the source volume, the deeper the penetration) but also on the type of seismic source as well as on sedimentological factors such as grain-size distribution, degree of compaction and diagenesis. Because the earth filters the higher frequencies, there is an inverse relationship between penetration and frequency, i.e. high-frequency seismic sources give low penetration, whereas low-frequency sources provide high penetration. 16.3.1.3. Seismic processing The basic objective of seismic processing is to improve the quality of the seismic profile by enhancing what are regarded as useful seismic signals, and removing or

307

T. Nielsen et al.

(a)

twtt (s)

3

SW

3 km

ne

4

5 (b) 3 Pre-drift sediments

twtt (s)

Migrating moat 4

Large-scale depositional units

Bounding reflectors Post-drift sediments

5

Figure 16.2 Buried contourite deposits off NE Greenland.The drift is only crossed by a single seismic line and the designation as a contourite is based solely on the seismic characteristics. (a) Multi-channel, low-frequency seismic profile. Acquisition frequency range of the air-gun array was 5^70 Hz. To improve the vertical resolution of the resulting seismic profile, a postcruise deconvolution operation designed to compensate for the effects of the Earth’s natural filtering of the high frequencies was applied, keeping the dominant frequency range at 10^60 Hz. Assuming a likely velocity increase from 1600 m s 1 to 2300 m s 1, the nominal seismic resolution varies from about 6.5 m in the upper part to about 10 m at the base of the displayed section. This indicates an average maximum vertical resolution for the contourite deposits of 7^8 m. (b) Interpretation of the seismic profile. The designation as a contourite drift is primary based on recognition of first-order seismic elements (see Figure 16.1), i.e. the mounded cross-sectional shape, the uniform internal pattern of continuous low- to mediumamplitude reflectors which bound large-scale upward-convex depositional units that tend to follow the overall shape, and the palaeomoat development to the left. The base of the drift is marked by a high-amplitude, continuous reflector and the top by a medium-amplitude continuous reflector onto which the overlying post-drift unit is onlapping. (Data courtesy of GEUS, Denmark.)

attenuating what is seen as noise. Seismic processing involves complicated mathematic algorithms (Yilmaz, 1987), which is beyond the scope of this chapter. It is, however, important to bear in mind that data processing involves subjective choices, meaning that a seismic profile processed by one person can differ substantially from that processed by another person. For instance, a reflector described as ‘‘high-amplitude’’ on one seismic profile might appear as ‘‘medium-amplitude’’ on another seismic profile across the same drift deposit (Figures 16.3 and 16.4). A summary of seismic expressions of contourites, as given in the previous section, is therefore a guideline rather than a rule.

308

twtt (s)

(a) 2

Seismic Expression of Contourite Depositional Systems

a

b

d

e

3 e d

a c

4 (b) 2

twtt (s)

c

10 km

b

Large-scale depositional units (first-order seismic elements) Internal erosional unconformities

3

10 km

4

Figure 16.3 Seismic loop across the Eirik Drift off South Greenland composed of seismics of various vintage and different acquisition and processing settings, illustrating how the seismic database influences the seismic characteristics of the drift. (a) Multi-channel low-resolution profiles. Profiles a,b and d-e are post-2000 data, profile c is from the 1970s. Note the better resolution of the younger data sets and the mismatches at the extremities of profile c, due to the inaccurate navigation of the older data. Profiles a and b have similar acquisition settings but different processing settings causing minor differences in the seismic appearance, while profiles d and e are part of the same data set and have similar seismic looks. Profiles a and e are acquired the same year, but have different acquisition and processing settings resulting in different seismic appearances. (b) Interpretation of the seismic loop. The first-order seismic elements (see Figure 16.1) can be followed around in the loop with confidence. Also some of the second-order seismic elements can be followed around, but with less certainty and clear variations in appearances. The largest variation in the seismic expression is of the third-order seismic elements, where the seismic facies appearance of the sub-units clearly differs with the different seismic settings.

16.3.2.

Reflection-seismic methods

The characteristics of the seismic system are as critical to the resulting seismic profile as is the subsurface geology. Thus, the seismic expression and level of interpretation, i.e. the level of detail of the stratigraphic and sedimentological information obtainable from seismics, is dependent on the technique used to acquire the data. In general, there are two types of recording arrays in reflection-seismic systems: single-channel seismics (SCS) and multi-channel seismics (MCS). Still some ultrahigh-frequency seismic data (see below) are recorded as analogue data, but digital recording of both SCS and MCS has become standard nowadays. This allows processing of the data to enhance the quality of the resulting seismic profile.

T. Nielsen et al.

309

Because most seismic sources generate sound with a distinct frequency content, there is a trade-off in all reflection-seismic systems between high penetration, which requires low frequencies, and high resolution, which demands high frequencies (Figure 16.5). Reflection-seismic systems can be grouped into categories (see below), corresponding to the relationship of the source frequency, level of resolution and depth of penetration. 16.3.2.1. Low-resolution seismics Low-resolution (LR) seismics applies a source frequency of 5–50 Hz. Surveys using LR seismics are usually devoted to investigating the deeper parts of the substratum. The seismic systems consist of a low-frequency, high-energy source (or array of sources) and multi-channel hydrophone streamer(s). The signal is digitally recorded and require subsequent processing to produce a seismic profile for interpretation. This, in turn, facilitates computer-based interpretation allowing adaptation of the display modes and analysis of various seismic attributes, which is a valuable tool in seismic studies of contourites, even in detail. 16.3.2.2. High-resolution seismics High-resolution (HR) seismic sources have a frequency range of approximately 50–2000 Hz, and encompass seismic sources like small water guns, air guns and sleeve guns in the lower end of the frequency range and sparker, boomer, chirp and parametric echo sounders in the higher end. These seismic techniques are relatively simple to handle, can be obtained from even small vessels, and are usually based on single-channel data acquisition, and thus do not require extensive processing. HR seismic systems are widely used for shallow to medium sub-bottom investigations, in both shallow- and deep-water areas. In the latter case, the source may be built into a deep-towed vehicle to avoid the loss of energy over a thick water column. Because the acquisition of HR data is relatively simple, concurrent acquisition of ultrahigh-resolution (UHR) seismics is common, providing both good penetration and resolution in one survey (Figures 16.6 and 16.7). 16.3.2.3. Ultrahigh-resolution seismics For detailed information of the sea floor and near-sea floor conditions, ultrahighfrequency sources like sub-bottom profilers (3–10 kHz), multi and single-channel echo sounders (10–500 kHz) and side-scan sonars (10–500 kHz) are used. Due to the high frequency of these systems, the penetration ranges from none (i.e. only the sea floor is mapped) to a few tens of metres below the sea floor. Thus, identifying contourite deposits exclusively using this type of seismic systems is difficult, considering the triple-scale approach described above. Therefore, the use of UHR seismics in contourite investigations is mainly to identify and map sea-floor features related to bottom-current activity (Wynn and Masson, 2008). However, the UHR seismic systems lying at the lower end of the high-frequency range (3–5 kHz) can be useful for investigations of the shallower, most recent part of a contourite drift accumulation (Figure 16.7) and for tying of high resolution sediment cores used in palaeoceanographic studies (Knutz, 2008).

310

Seismic Expression of Contourite Depositional Systems

311

T. Nielsen et al.

Figure 16.4 The effect of processing on the seismic character of the drift illustrated by different versions of the same multi-channel LR seismic profile (air-gun source, maximum frequency in data below 100 Hz) across the Eirik Drift off South Greenland. The upper panel shows the impact of deconvolution parameters upon seismic resolution; the top-left panel shows the original data without deconvolution (no Decon), the top-central panel focuses on deeper-lying structures using a 28/200 deconvolution parameter setting, and the top-right panel enhances the resolution in the shallow part using a 22/200 deconvolution parameter setting. Notice that first-order seismic elements (see Figure 16.1) can be followed on all three versions, whereas marked dissimilarities occur at the second order and, in particular, third-order seismic-element level. The lower-left panels A1 and A2 are close-ups of the ‘‘no Decon’’ and ‘‘Decon 22/200’’ versions, respectively, illustrating how high-frequency wavy features are enhanced on the latter (A2). Detailed examinations prove that these are real features, probably caused by sediment waves. The lower-right panels B1 and B2 are close-ups of the ‘‘no Decon’’and ‘‘Decon 22/200’’ versions, respectively, illustrating how the seismic facies of the subunits varies with different deconvolution parameters. The seemingly clear and continuous reflection pattern of the ‘‘no Decon’’ version (B1) is due to the low-frequency content, whereas the more crispy and less prominent reflection pattern of the ‘‘Decon 22/200’’ version (B2) is the result of finer geological details being enhanced by the higher frequency content. A similar difference in the seismic-facies appearance is observable between A1 and A2. (Data courtesy of GEUS, Denmark. Data processed by R. Rasmussen, GEUS.)

twtt (ms)

1400

e

W

1600

1800 1 km

Figure 16.5 Single-channel high-resolution (HR) seismic profile off Southwest Portugal. The seismic system consisted of a Geo-spark 800 source and a 48 -element streamer, resulting in a high vertical resolution (>30 cm) and a relative shallow penetration (<500 m). The profile shows a series of up-slope migrating sediment deposits, interpreted as mega sediment waves that form part of a larger current-controlled accumulation deposited by the Mediterranean outflow water. Due to the high-frequency character of the seismic system, the base of the current-controlled deposits cannot be mapped, and classification as a specific contourite type is therefore not feasible on the basis of this type of seismics alone. In contrast, the HR of the data permit detailed investigations of the relationship between the individual mega waves that build up the youngest, near-surface part of the accumulation. Also this type of seismics is very suitable for ground-truthing by shallow coring. (Data courtesy of GEUS, Denmark.)

312

Seismic Expression of Contourite Depositional Systems

Profile 1 Profile 2

twtt (ms)

(a) W Internal unit? post-drift sediments? 1000

? Top contourite ? e

1500

2000

Bedrock

7 km

2500

(b)

Profile 1

se

Profile 2

twtt (ms)

1000 1000 1500

2000

55°W

50°W

67°N

Davis Strait

2000 65°N

t

00

15

1

d Ice Shee

Greenlan

1000

twtt (ms)

5 km 1500

2

63°N

2000

2500 Labrador Sea

Figure 16.6 Drift deposit off West Greenland, first recognised and mapped using a 2-D grid of multi-channel low-resolution (LR) seismic data, and subsequently studied in detail using highresolution (HR) seismics (Nielsen et al., 2001, 2003). Inset map (lower right) shows the study area and the sea-floor bathymetry in metres. The bathymetric high located in the central part shows the elongated outline of the drift. Also shown is the position of seismic profiles 1 and 2. (a) Seismic profile 1. LR seismic data showing the seismic characteristics (see Figure 16.1) identifying the deposits as a contourite drift. (b) Seismic profile 2. HR seismic data revealing the details of the contourite drift (sediment waves confirming the influence of strong bottom currents).

Using UHR seismics favours the resolution needed for detailed studies of the shallower part of the contourite, but lacks the penetration necessary to study the deeper parts (Figure 16.8). Therefore, as mentioned earlier, a full-scale seismic study of thick contourite deposits requires the use of different types of seismic systems in order to achieve both the high penetration needed to map the drift geometry, and the high resolution necessary for sediment-core correlation.

313

T. Nielsen et al.

(a)

twtt (s)

(b) Internal unconformities Large-scale units Bottom-current direction away 1.8

2.3 Basal erosional unconformity 2.8

Basaltic basement ( Lateral migration direction of the channel system) sw

(c)

Outcropping basalts

ne

Sand-wave fields

nw

se - Current direction Lens-like sand body

Figure 16.7 Contourite depositional system at the outlet of the Faroe Bank Channel, the North Atlantic. The deposits were surveyed during the TTR 7 cruise (Nielsen et al., 1998a ^ b). Identification of the drift deposits was based on high-resolution seismic data, whereas areal mapping and reconstruction of the current pattern involved long-range and deep-towed sidescan sonar data. Further details of the current pattern also involved bottom-sampling results (Akhmetzhanov et al., 2007). (a) Map of the study area with outline and interpretation of long-range side-scan mosaic. Position of profiles in figures b and c is also shown. (b) Singlechannel HR air-gun seismic profile illustrating the seismic characteristics of the drift deposits (see Figure 16.1). (c) Deep-towed side-scan profile image (upper) showing bottom-current sea-floor features and built-in UHR Pinger profile (lower) revealing details of the most recent deposition at the surface of the contourite. (Modified from Kenyon et al., 2004 with permission of Springer Science and Business Media.)

314

Seismic Expression of Contourite Depositional Systems

(a)

twtt (s)

(b) 1.1

DANA97/9-04

1.2

2 km 1.3

Figure 16.8 Comparison of two different seismic profiles across a contourite drift system northeast of the Faroe Island, illustrating how the use of different seismic systems is a trade-off between penetration and resolution. (a) Multi-channel HR sleve-gun seismic profile. The relative low frequency allows penetration below the base of the drift deposit, making it possible to map the overall drift morphology. However, the resolution is less than 10 m s 1 twtt, i.e. only layers more than 8^10 m thick can be distinguished. Thus the internal reflection pattern reveals only a rough picture of the interior of the drift. (b) Analogue UHR Pinger profile across the same part of the drift system. The penetration is max. 60 m s 1 twtt, preventing the base of the drift to be recognised, but the resolution is at least 1 m s 1 twtt, i.e. internal layers less than 1 m are revealed, allowing detailed investigation of the shallow part of the drift and tying to sediment cores.

16.3.3.

Seismic mapping

In addition to the vertical and horizontal seismic resolution, the density of the seismic profiles, or the seismic grid, is of major importance for the resolution of the mapping. As the recognition of a contourite deposit is highly depending on the overall drift morphology, a seismic study of contourites is ideally based on a grid of data (2-D) or a volume of data (3-D). 16.3.3.1. 2-D mapping The conceptual profile shown in Figure 16.1 is constructed as a dip line, i.e. it illustrates a seismic profile perpendicular to the slope and bottom-current direction. If the model profile was directed parallel to the slope (i.e. a strike direction), it would have appeared quite different. For instance, the

T. Nielsen et al.

315

characteristic moat in the up-slope direction would not be expressed on a slope-parallel profile. Moreover, the internal reflection pattern would be more or less uniform and parallel, and therefore difficult to distinguish from other types of marine deposits. Thus, when attempting to identify and map sediment packages induced by bottom currents, not only the grid density, but also the direction of the seismic profiles with respect to the overall drift geometry should be considered. Examples of 2-D seismic mapping of contourites are shown in Figures 16.6 and 16.7. 16.3.3.2. 3-D mapping 3-D seismic surveys provide the most accurate data for mapping contourite deposits in a process-oriented context. The ability to visualise geological surfaces from seismic reflectors is particular relevant for identifying small-scale morphological features such as moats, channels and migrating bedforms associated with contourites. 3-D seismic studies are further enhanced by software tools that allow the extraction of different attributes from the seismic data (see below) and the ability to perform horizontal slicing of the 3-D cube. However, acquisition of 3-D seismic data is very costly and mainly associated with hydrocarbon exploration. Thus, it is not readily available for academic use and it provides only limited coverage of deep-sea areas. Another problem is that the seismic processing is often aimed at enhancing the resolution at depths typically varying between 1 and 4 s twtt at the expense of the acoustic clarity in the top 500 ms interval. Despite the access limitations of commercial 3-D seismic data, the recent focus on enhanced imaging techniques in marine research means that such data will become more available for basic research in the near future. Also, a non-commercial, HR 3-D seismic platform under development (Marsset et al., 2002; Praeg et al., 2006; Vanneste et al., 2007) should stimulate considerable progress in 3-D sub-sea-floor imaging of deep-sea contourites. The ability to map stratigraphic features in three dimensions at a detail otherwise unattainable has already provided new insights on processes and deposits related to contour currents (Figure 16.9) (Knutz and Cartwright, 2003).

16.3.4.

Seismic interpretation

A prime motive in the application of reflection-seismic methods in contourite studies is to unravel the sediment-drift stratigraphy. Previously, this task was pursued simply by identifying the continuity, amplitude and stacking pattern of the reflections seen on the seismic profile. The recent advances in seismic technology, in particular the use of digital 3-D volumes, provide new opportunities for seismic facies and seismic attribute analysis, and thus extracting more detailed subsurface information from contourite depositional systems. Concurrent with the development of seismic methods, the terminology within seismic interpretation has become increasingly confusing. Terms like ‘‘sequence’’, ‘‘facies’’ and ‘‘attributes’’ appear in a multitude of contexts to the extent that the scientist must clarify the meaning of these terms within the context of the specific study.

316

Seismic Expression of Contourite Depositional Systems

Figure 16.9 Detailed morphological reconstruction of a buried contourite-slide depositional system based on 3-D seismic data from the continental slope west of Shetland (see multicolor version of this figure on the enclosed CD-ROM). (a) Seismic cross-section with reflectors representing the Intra-Neogene Unconformity (INU) and two internal horizons demarcating the top of unit B1 (mounded drifts of Pliocene age) and of unit B2 (multiple slide events of early Pleistocene age) (Knutz and Cartwright, 2003). (b) Amplitude attribute extraction of the Top B2 surface, illustrating lobate slides and thin down-slope channels outlined by light contrasts presumably representing sand-rich features (Knutz and Cartwright, 2004). (c) 3-D view of time-depth maps representing the INU and the Top B1 surface. The series of mounded drifts separated by moat ^ channels have developed in response to southward-flowing bottom currents that were active during the Pliocene ^ early Pleistocene.The basal reflector represents the regional erosional unconformity (INU) of late Miocene ^ early Pliocene age. Sea-floor depressions created by slumping are thought to have promoted the initial accumulation of contourites by acting as sediment traps for fine-grained clastic material supplied by alongslope currents.

T. Nielsen et al.

317

16.3.4.1. Seismic units Since the concepts of ‘‘seismic stratigraphy’’ and ‘‘seismic sequence stratigraphy’’ were introduced in the 1970–1980s (Vail et al., 1977a; van Wagoner et al., 1988), a whole set of terms has developed for seismic interpretation that are now widely used. However, the conventional concept of seismic sequence stratigraphy – and the terminology involved – is not fully applicable to contourite studies. While the conventional seismic sequence-stratigraphic concept implies that the sequences and their stratal components form in response to base-level changes and down-slope shelf-to-basin sediment transport, contourites are primarily controlled by bottomcurrent regimes and sediment pathways orientated parallel to the continental margins. Also, in conventional seismic sequence stratigraphy the unconformities that bound the sequences are inferred to be autocyclic and the results of predictable changes in accommodation space, whereas unconformities in contourite deposits tend to occur in a non-predictable pattern as a result of changes in current regime. Hence, the use of the conventional seismic sequence stratigraphic terminology in deep-water current-controlled depositional environments may lead to misinterpretations. In recent years, there has been an increasing demand for the development of a new, more comprehensive sequence-stratigraphic concept encompassing deep-water current-controlled settings (Shannon et al., 2005). Although some deep-water depositional processes have been included, this mainly concerns gravity-driven depositional settings (Posamentier and Kolla, 2003). At the present stage, a ‘‘seismic sequence’’ is still expected to have a depositional affinity closely tied to down-slope-directed sediment input and base-level changes. We thus suggest the use of the simple term ‘‘seismic unit’’ that is not associated with any specific depositional environment to denote a stratigraphic subdivision of a sediment-drift succession. The breakdown into seismic units, in contrast, may be based on analysis of reflection terminations and internal reflector patterns using the traditional technique introduced by Mitchum and Vail (1977) among others. 16.3.4.2. Seismic facies Analysis of the seismic facies is another method commonly used in the seismic interpretation of contourites, and deals with investigation of vertical and lateral variations of internal reflections. Synonymous terms are ‘‘seismic reflection pattern’’, ‘‘acoustic facies’’ or ‘‘echo character’’ analysis, the latter mostly used in connection with UHR seismic data. Seismic-facies analysis in relation to contourite studies mostly deals with investigation of the contourite sub-units, i.e. at the small-scale level. At the larger-scale level, one should refer to the ‘‘internal reflection pattern’’ rather than the ‘‘seismic facies’’. The seismic facies is relative to the type of seismic method employed. This means that a seismic facies will display differently on different types of seismic data (Figure 16.3), and will also depend on processing parameters (Figure 16.4). Seismic facies sometimes reflect sedimentary structures like waves and progradation.

318

Seismic Expression of Contourite Depositional Systems

However, because the reflections result from changes in the physical parameters through the sedimentary succession, there is no unequivocal correlation between seismic facies and sedimentary structures within the facies. A seismic facies characterised by a parallel reflection configuration, for instance, need not necessarily indicate the existence of fine parallel banding or stratification of the sediments.

16.3.4.3. Seismic attributes The advent of digital recording of seismic data and 3-D seismic technology has facilitated interpretation using seismic attribute analysis. A seismic attribute is a measure of the character of the seismic data. As seismic reflections derive from changes in the physical parameters of the sub-surface, the character of the reflections contains valuable clues of geological significance. An attribute can be a measure of a single reflector, an interval (e.g. a sub-unit) or – in 3-D seismics – a horizon. Basically there are two categories of seismic attributes: those that quantify the geometric aspects and those that quantify the reflectivity components of the seismic data. The geometric attributes reveal information on dip, azimuth and termination of a reflector or horizon, which can in turn be related to current-induced bed forms like waves, ripples, furrows, moats and channels. The reflectivity attributes reveal information on reflector amplitude, frequency and phase, which in turn might be related to lithology. As the types of attributes are numerous, so are the computational methods and a variety of seismic attribute analysis techniques exist (Barnes, 2001, 2006). In the context of contourite studies, the morphological attributes have proven to be particular useful (Figure 16.9; see section ‘‘3-D mapping’’ above).

16.3.4.4. Seismic inversion (‘converting seismics into geology’) Seismic inversion, broadly defined, is the study of acoustic information like velocity, impedance and amplitude to extract geological information of the subsurface layers like density, porosity and compaction. It is a difficult process, since the seismic measurements are limited and the earth extremely complex, and there are many different inversion methods (Yilmaz, 1987). In seismic depth conversion, the twtt is converted to depth. This process requires estimation of the seismic velocity, which in fact is the most uncertain link between seismics and geology. Therefore, contourite interpretation which makes use of core and borehole information in combination with seismics requires information on the seismic velocity used for the correlation (Figure 16.10).

319

T. Nielsen et al.

(a)

se MD95-2006

nw

30 ms –1 twtt 1 km IA DF 3

IB II III

(b) Magnetic susceptibility m

Lithol. units Lithofacies ka BP

MD95–2006

1

10.89

2

14.86

5

I-A

10 3

17.66

15

I-B II

4

~23 m

24.31

20

25.81 25

5

33.48

III ~500 m

40 90140 SI units

Figure 16.10 Interpretation of muddy contourites from the Barra Fan on the UK Margin, using core information in combination with HR seismics.The correlation is based on a seismic velocity of 1525 m s ^1. (a) High-resolution seismic profile (DT Boomer) from the lower part of the fan. (b) Close-up of the profile and correlation with core lithostratigraphy (MD95-2006). The weakly layered upper unit IA, representing Late Glacial ^ Holocene deposition, is interpreted as a glacial ^ marine hemipelagic facies with a gradual transition to a silty ^ muddy contourite in the upper 1.5 m. The parallel reflectors in unit II correspond to sandy turbidite layers in the core record. Seismic unit III approaching the limit of acoustic penetration is interpreted as a silty ^ muddy contourite in the lower core section. (Adapted from Knutz et al., 2002b, reproduced with permission from Elsevier.)

320

Seismic Expression of Contourite Depositional Systems

16.4.

S UMMARY

The seismic character of contourite features is to a large extent depending on the methods used in acquisition and processing of the seismic data. Prerequisites for an adequate seismic study of contourites are therefore: • identification of the type of seismic sources (LR, HR or UHR), acquisition parameters (SCS or MCS; 2-D or 3-D), processing sequences (e.g. filtering, deconvolution) and display parameters (e.g. twtt versus depth, axes scale, polarity); • recognition of how these factors affect the interpretation and level of detail resolved by the seismics in the contourite study; • careful attention for the seismic-to-geology conversion when using core and borehole information (seismic resolution versus core data, velocity estimation); • specification of the definitions and use of interpretational methods and terms (seismic sequence or unit, seismic facies) relevant for the current-controlled depositional environment.

Large-scale depositional units

First-order seismic elements (overall drift architecture)

Seismic Characteristics

Large-scale

Moat Large unconformities

Depositional sub-units

Second-order seismic elements (internal architecture)

Medium-scale

Seismic facies

Small-scale

e.g. Reflection strength map

Third-order seismic elements (seismic facies and attributes)

UHR Resolution

10–500 m

0.5–1.5 km

Air gun Water gun Sleever gun

0–10 m

arrays of:

Water gun Air gun Sleeve gun

Penetration

LR HR Sparker Boomer Topas Chirp

Sources

Pinger Echo sounder Side-scan sona

Seismic Methods

Frequency

>1.5 km

Figure 16.11 Correlation between the seismic methods, frequency of the seismic source and the seismic characteristics of contourite deposits.

T. Nielsen et al.

321

Seismic identification and stratigraphic breakdown of a contourite depositional system embodies three aspects, which reflect the increasing level of detail of interpretational seismic elements: • first-order seismic elements (large-scale features): the overall architecture of the drift focusing on the gross geometry and largescale depositional units; • second-order seismic elements (medium-scale features): the internal architecture (drift structure) resolving the sub-units that build up the large-scale depositional units; • third-order seismic elements (small-scale features): the seismic attributes and seismic facies configuration of the depositional subunits. An overview of the correlation between the seismic methods, frequency of the seismic source and the seismic characteristics is shown in Figure 16.11.

ACKNOWLEDGEMENTS We acknowledge the good cooperation with the editors of this book and the review by Dr J.S. Laberg. We thank The Geological Survey of Denmark and Greenland (GEUS) for allowing the reproduction of seismic data, and R. Rasmussen (GEUS) for providing good processing examples.

C H A P T E R

1 7

I DENTIFICATION OF A NCIENT C ONTOURITES : P ROBLEMS AND P ALAEOCEANOGRAPHIC S IGNIFICANCE ¨neke1 and D.A.V. Stow2 H. Hu 1

Institute of Geography and Geology, University of Greifswald, Greifswald, Germany National Oceanography Centre, Southampton (NOCS), Waterfront Campus, Southampton, UK

2

Contents 17.1. Introduction 17.2. Examples of Fossil Contourites 17.2.1. Neogene contourites, Miura–Boso region, SE Japan 17.2.2. Carbonate contourite drift, Oligocene palaeoslope, Cyprus 17.2.3. Carbonate contourite drifts, Late Cretaceous–Palaeocene Chalk Group, Danish Basin 17.2.4. Carbonate contourites, Devonian pelagic successions, Europe and North Africa 17.2.5. Carbonate contourites, Ordovician Jiuxi Drift, China 17.3. Discussion 17.3.1. Recognition of ancient contourites 17.3.2. Sea level and preservation of contourite drifts 17.3.3. Sequence interpretation and non-deposition surfaces Acknowledgements

17.1.

323 324 324 326 329 332 338 340 340 343 344 344

INTRODUCTION

Contourites are well established and well described from recent and sub-recent deposits throughout the world’s deep ocean environments. In over 40 years of research, they have been amply described in terms of contourite drift architecture and seismic facies (e.g. Fauge`res et al., 1993, 1999; Rebesco and Stow, 2001), the erosional features and depositional bedforms associated with contour current flow pathways, and their specific sediment characteristics (e.g. Stow and Fauge`res, 1993, 1998; Stow et al., 2002c, f ). The same cannot be said for their identification in ancient sedimentary series now exposed on land (Pickering et al., 1989; Shanmugam, 2000), although some further significant progress has been Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00217-3

Ó 2008 Elsevier B.V. All rights reserved.

323

324

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

made in this direction since the problem of fossil contourites was critically reviewed for the last time (Stow et al., 1998a). Many of the existing large-scale drifts present in the various oceans have now been drilled, principally during DSDP and ODP scientific expeditions spanning the past 35 years. Nearly 100 contourite drift sites were listed in the review by Stow et al. (1998a), the deposits drilled ranging in age from Eocene to recent. Together with a much larger number of gravity and piston cores recovered from these and other drift systems, they provide clear evidence of the nature and variability of contourite facies. Such data have been variously reported and synthesized in numerous papers (e.g. Stow and Fauge`res, 1993, 1998; Stow et al., 2002f ) and form the basis of documenting the contourite facies in this volume (Giresse, 2008; Herna´ndez-Molina et al., 2008a, b; Martı´n-Chivelet et al., 2008; Stow et al., 2008). There also exists a growing body of examples of contourites in ancient rock series, ranging in age from Cambro-Ordovician to Neogene. Here, we review those case studies with which we are most familiar and for which the body of evidence used in their interpretation as contourites is most persuasive. We then discuss some of the important common attributes from these studies, preservation conditions of contourite deposits, and the paramount need for caution when trying to interpret ancient deep-water successions. It is important to note that controversy still surrounds the recognition and interpretation of (contour-current) reworked turbidites in ancient series (see Stanley, 1988b; Stow et al., 1998a, 2002a; Shanmugam, 2000). This problematic discussion is further detailed in other chapters of the present volume (see Martı´n-Chivelet et al., 2008; Shanmugam, 2008; Stow et al., 2008), but will not be addressed here.

17.2. 17.2.1.

EXAMPLES OF FOSSIL C ONTOURITES

Neogene contourites, Miura–Boso region, SE Japan

The thick middle Miocene to Pliocene sedimentary succession, which is particularly well exposed along the coastline of the Miura and Boso peninsulas, south of Tokyo Bay in Japan, has been recently accreted to the main Honshu Arc from its former position as part of the Izu–Bonnin forearc (Soh et al., 1989, 1991; Taira and Ogawa, 1991). Within the mixed assemblage of predominantly hemipelagic and volcaniclastic sediments, which is known locally as the Misaki Formation and was originally deposited at bathyal depths on the forearc slope–apron system, detailed study has revealed the marked influence of bottom currents (Stow and Fauge`res, 1990; Stow et al., 1998b, 2002e). The intervals of muddy or calcilutitic contourites occur most notably in the light-coloured, fine-grained, background slope sediments, which are associated with distinct beds of dark-coloured, scoriaceous, volcaniclastic sediments, mainly deposited as turbidites, debrites and direct pyroclastic, sub-aqueous fall-out.

¨neke and D.A.V. Stow H. Hu

325

The principal characteristics of these contourite facies are as follows (Stow et al., 2002e): • a notable absence of clear primary sedimentary structures, but presence of indistinct and discontinuous parallel lamination in parts; • irregular concentrations and lenses of coarser-grained (silt/sand grade) scoriaceous and bioclastic material; some of these thin lenses show remnants of micro-cross-lamination and parallel lamination, now partially obscured by a bioturbational fabric; • sharp and erosional contacts in places associated with the coarser-grained horizons, and locally as minor discontinuities (non-deposition surfaces) within the fine-grained facies; • an ichnofacies assemblage dominated by Chondrites, Helminthoides, Planolites and Zoophycos within an intensely bioturbated sediment; below localized omission surfaces, the ichnofacies include small-scale vertical and sub-vertical traces. • a mean grain size of clayey silt, poorly sorted with dispersed sand-sized material and localized concentrations of slightly better sorted coarse silt and sand; cyclic grain-size variations evident over intervals of approximately 10–100 cm; • a mixed composition of biogenic and volcaniclastic material; the biogenic fraction is dominated by pelagic foraminifers, nannofossils, diatoms and radiolarians, mixed with relatively fewer benthic and shallow-water fossils, in some cases fragmented and iron-stained; the volcanic material is mainly pumiceous glass, devitrified glass and resultant clays, with minor amounts of basaltic scoriaceous grains. Although the Misaki Formation contourites are well preserved in terms of detailed sedimentary characteristics, the accretion process generated considerable structural complexities and sequence repetition onto the Honshu Arc. This has made it impossible to reconstruct regional distribution of facies associations or thickness variations and hence to reconstruct a meaningful original depositional geometry. The obvious slope setting and intercalated deposits of turbidites and debrites suggest a closely interbedded system of down-slope and along-slope facies. Down-slope facies were evidently derived from the Izu–Bonin Arc, which lay to the west of the slope system, based on dominant palaeocurrent directions and magnetic-fabric measurements (Kanamatsu, 1995; Lee and Ogawa, 1998). Similar magnetic-fabric measurements in the fine-grained facies show mixed current directions to the SSE, NNW and NE, which are not incompatible with an element of along-slope transport by bottom currents. 17.2.1.1. Palaeoceanography For the late Neogene, the palaeoceanographic reconstruction is generally very good so that we can be confident that the Misaki Formation sediments were originally deposited on the Pacific-facing slope of the Izu–Bonin Arc. The older part of the succession would have started out several hundred kilometres further south, followed by slow, steady northward movement. Ancestral precursors of both the southward flowing North Pacific Deepwater and (probably) northward-flowing Antarctic Bottom Water would have been in existence at this time, and both would, as can be deduced from comparison with measurements of the equivalent present-day deep-water current systems (Lee and Ogawa, 1998), have been capable

326

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

of moving at least relatively fine-grained sediment along the Izu–Bonin slope. Whereas the ancestral Kuroshio surface current would also have existed and been capable of affecting sedimentation at depths in excess of 2000 m (Taft, 1978; Stow et al., 2002e, f), its influence is unlikely to have extended far enough south for the onset of contourite deposition as noted in the Misaki Formation. 17.2.1.2. Problems We recognize three principal problem areas with respect to this case study. 1. The sedimentary features of these fine-grained contourites are very subtle, as is generally the case in ancient contourites, and although the authors have followed the full three-stage approach recommended by Lovell and Stow (1981) and Stow et al. (1998a); evidence based on regional and large-scale considerations is not unequivocal. These lines of enquiry indicate the potential of, rather than requirement for, contourite deposition. 2. Whereas most of the detailed facies characteristics strongly suggest contourite deposition, the cause of cyclicity in grain size remains uncertain. It is unclear to what extent it may be the result of long-term fluctuations in the mean bottomcurrent velocity, and to what extent a function of variable input of the coarsergrained volcaniclastic fraction. There is evidence that both were involved. 3. Most of the coarse-grained scoriaceous beds show clear evidence of primary deposition from down-slope processes, but as bottom currents were also active at this time, it is possible that the top parts of some turbidite beds have been reworked by along-slope processes. Direct evidence for this has been proposed by Lee and Ogawa (1998), but has been disputed by Stow et al. (1998b). The question remains unresolved to date.

17.2.2.

Carbonate contourite drift, Oligocene palaeoslope, Cyprus

The late Cretaceous to early Miocene Lefkara Formation is well known to have been deposited over newly formed oceanic crust (Cretaceous ophiolite sequence) of the Tethys Ocean (Malpas et al., 1990; Robertson, 1990). It forms a mostly continuous succession, up to some 600 m thick, and is relatively free from tectonic deformation, except where it abuts most closely the now exposed crustal basement rocks. Whereas pelagic biogenic sediments dominate throughout, they are intercalated with a series of both carbonate and siliceous turbidites that are more common in the lower two-thirds of the formation. The Lefkara Formation passes gradationally upwards into the Pakhna Formation, which was deposited on the flanks of the Troodos ophiolite complex that was being uplifted, and therefore shows progressive upward shallowing (Roberston, 1990; Stow et al., 2002d). First proposed by Robertson (1976) and later identified by Kahler (1994), contourites are recognized within the upper part of the Lefkara Formation. Considerable work has now been undertaken to fully characterize these deposits (Kahler and Stow, 1998; Stow et al., 2002e; Turnbull, 2004), and so designate them as a type example of mainly Oligocene contourites (Figure 17.1). The associated facies are principally calcareous biogenic pelagites, and some rare, more silty

327

¨neke and D.A.V. Stow H. Hu

(m) 2 12

1

11

10 0

12

9

11 8 10

7

9 8 7

6

6 5

5

4 4

Figure 17.1 Oligocene calcareous sediments interpreted as calcicontourites, Petra Tou Romiou, southern Cyprus. Detailed log of section shown in photograph, together with representative microfacies at sample points 4^12 (from Turnbull, 2004, with permission). Cyclic grain-size variations evident over approximately 50 cm of the section, between bioturbated calcilutite and lenticularcalcarenite macrofacies.Thin-section micrographs showaclearsequence fromwackestone with pelagic foraminifers and nannofossils (4), through fine-grained, well-sorted packstone, with lenses including benthic foraminifers, shallow-water algal and echinoid debris (5^8), and back to wackestone (9 and 10). A multicolour version of this figure is on the enclosed CD-ROM.

clay-rich hemipelagites. Some thin-bedded calcareous turbidites and probably reworked turbidites have been identified in parts of the contourite succession. The principal characteristics of these calcareous contourite facies are as follows (Stow et al., 2002d; Turnbull, 2004): • there are three main facies: a fine calcilutite, generally structureless but with some faint discontinuous laminae, and lenses of calcisiltite; a more clay-rich calcilutite (or marl), which is too strongly weathered to still exhibit detailed features; and a

328

•

•

•

•

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

calcarenite/calcisiltite with a distinctive thin-bedded lenticularity and only very rare internal lamination and cross-lamination (Figure 17.1); in addition to an ubiquitous small-scale lenticularity, the calcarenite facies shows in places what may be interpreted as shallow scours (2–10 cm deep) into the underlying calcilutites, and larger-scale lenses (up to 30 cm thick and several metres in lateral extent); for the most part, the sediments are thoroughly bioturbated and burrowed; distinct burrow traces include small-scale Chondrites, ?Helminthoides, and less regular micro-traces, as well as larger-scale forms including Planolites, Thalassinoides and sub-vertical Ophiomorpha, the latter typically developed below a probable omission surface; original grain-size attributes are partially obscured by carbonate diagenesis, but the size variation between calcilutite and calcarenite most likely reflects the original conditions, and hence the clear but irregular cyclicity that occurs on a decimetre scale between these two facies and is interpreted as the result of fluctuations in the mean bottom-current velocity, leading to winnowing, cleaning and sorting of the calcarenites; the contourite microfacies are dominantly packed biomicrites (including wackestones and foraminiferal packstones, with the finer beds and pelagites being sparse biomicrites). Much of the finer fraction is now diagenetic micrite with relict coccolith plates, whereas the coarser fraction, dominant in calcarenite facies, is better preserved. This includes dominant pelagic foraminifers (Globigerina sp.), together with a variable admixture of benthic foraminifers, radiolarians, echinoid spines, calcareous red algae and terrigenous grains (quartz, mica, feldspar and ferromagnesian minerals). Bioclast fragmentation and iron staining is common.

At a regional scale, certain aspects of the upper parts of the Lefkara Formation are significant in supporting the contourite interpretation. The Upper Marl unit shows a notable variation in thickness across southern Cyprus from little more than 10 m to over 200 m, which Stow et al. (2002d) interpret as most likely due to deposition as part of a contourite drift system. There is also a widespread hiatus (or hiatuses) in sedimentation, which is variable in duration but everywhere present for the midOligocene interval, below or within the Upper Marl unit (Kahler, 1994). The distinctive lenticular calcarenite facies is present immediately above the Upper Marl unit in all sections so far examined across southern Cyprus over a lateral extent of some 80 km. This is interpreted as the result of bottom-current intensification, simultaneously over an extensive area. 17.2.2.1. Palaeoceanography Tectonic reconstruction of the region shows that the Lefkara Formation as a whole was deposited in the closing Tethys ocean, in a basin plain to distal slope–apron setting (Robertson et al., 1991; Kahler and Stow, 1998; Robertson and Comas, 1998). Palaeodepths have been estimated as between 2000 and 3000 m, and land was at least 50–80 km away to the north. The nature of bottom circulation in the closing Tethys during the Oligocene is not well known. Earlier in the Palaeogene,

¨neke and D.A.V. Stow H. Hu

329

almost certainly warm, saline bottom waters flowed from east to west through the Tethys Seaway, as did a strong surficial flow of the Tethyan Current. It is a matter of debate just how long this circulation pattern lasted, and to what extent the growing development of cold, deep waters generated at high latitudes influenced the region. Stow et al. (2002d) believe that continued constriction of the Tethys at this time led to intensification of both surface and bottom-current systems and that this is reflected in the development of regional hiatuses and contourite drifts. 17.2.2.2. Problems The overall interpretations of this Oligocene succession on Cyprus that we suggest are supported by evidence at the small, medium and large scale, and constitute one of the better examples of ancient contourites. The main problems remaining are largely related to the nature and effects of diagenetic modifications to the original facies: 1. diagenesis of carbonates typically yields dissolution seams of softer marly layers between well-cemented horizons, and it is uncertain to what extent the lenticular bedding characteristic of the calcarenite contourites results from dissolution and to what extent it reflects an original depositional fabric; similarly, the larger-scale scours, calcarenite lenses and omission surfaces may have all been affected in part by diagenetic dissolution. 2. the exact nature of grain size characteristics is also obscured by subsequent diagenesis, but the cyclic variation in facies and grain size is believed to reflect the original distribution.

17.2.3.

Carbonate contourite drifts, Late Cretaceous–Palaeocene Chalk Group, Danish Basin

Distinct Santonian–Maastrichtian seismic units of the Late Cretaceous chalk in the Danish Basin have, based on seismic profiles in the Kattegat and Øresund areas (Lykke-Andersen and Surlyk, 2004; Esmerode et al., 2007; Surlyk and Lykke-Andersen, 2007), been interpreted as contourites deposited in an epeiric sea. The seismic stratigraphic sequence boundaries were tied to lithostratigraphic and chronostratigraphic information obtained from boreholes and the coastal cliff at Stevens Klint outside the principle contourite area. Topographic elements such as moats, channels, sediment wave fields, and elongate mounded and sheeted drifts have been identified. Large erosional channels are also known from outcrops of the Normandy coast of NW France (Quine and Bosence, 1991) and southern England (Evans et al., 2003). The Cenomanian–Danian Chalk Group in the North Sea and the Danish Basin was deposited in a relative deep epeiric sea (300–800 m), considerably deeper than the 300 m water depth generally suggested as a minimum for contourite deposition (Stow et al., 2002c). The NW–SE trending Sorgenfrei–Tornquist Zone formed an important and influential elongate topographic ridge with a SW-facing slope, separating a northeastern shallow shelf from the deeper Danish Basin towards the SW (Figure 17.2).

330

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

58°

Cretaceous:

Skagerak

Land Shallow marine Pelagic chalk Contourite deposits

Kattegat

So

rg

en

fre

i-T or

nq

uis

t

Dan

ishB

56°

Rin

gko

bing

Zo

asin

ne

-Fy

n

Hig

h

Nor th G

erman

54° 8°

10°

Basin

100 km

12°

14°

Figure 17.2 Map of Denmark and adjacent areas indicating the Late Cretaceous epeiric sea with the inferred contourite systems and the position of main structural elements (from Surlyk and Lykke-Andersen, 2007, and Esmerode et al., 2007, with permission from the Geological Society, London). Light grey background indicates modern geography; the Sorgenfrei ^ Tornquist Zone is shown in dark grey.

In the Kattegat area, a deep moat adjacent to the topographic ridge formed over the inverted NW–SE trending Sorgenfrei–Tornquist Zone (Surlyk and LykkeAndersen, 2007). During the Maastrichtian, the moat was up to 120 m deeper than its southwestern flank that was formed by an internally complex elongate drift passing distally into a drift sheet (Figures 17.2 and 17.3). Moat-related patch drifts, channels, and elongate multicrested drifts and waves occur in addition to the large-scale features. The current is interpreted, on the basis of the internal architecture of the elongate drift and the NW-ward branching and decrease in moat relief, to have flowed from the SE towards the NW. The region of highest current velocity was gradually shifted NE-wards towards the inversion zone ridge, resulting in the formation of the deep moat in the Kattegat area flanked by the elongate drift along its southwestern margin. In the Øresund area, a system of major WNW–ESE oriented ridges and valleys reflects the influence of strong bottom currents (Lykke-Andersen and Surlyk, 2004). The ridge/valley system can be traced into the succession exposed at Stevens Klint, where it corresponds to the relief of the Cretaceous/Palaeogene boundary. High-resolution seismic structures are interpreted to represent real topographic

331

¨neke and D.A.V. Stow H. Hu

SW 3200

NE 3400

3600

3800

4000

4200

4400

4600

4800

5000

5200

5400

5600

5800

6000

6200

0

0 9

200

200

4

6

300

300

400 5

400 500

(a)

3

500

2

600

1

700

3200

3400

3600

3800

4000

4200

4400

4600

4800

5000

5200

5400

5600

5800

6000

100 200 300 400 500

6200 0

0 8

Two way time (ms)

100

7

Approx. depth (m)

Two way time (ms)

8

100

9

100

7

200 4

6

300 400

5 3

500

2

600

1

Approx. depth (m)

(b)

700 5 km

STZ

Figure 17.3 Section K90 -4b crossing northeastern parts of the Danish Basin and parts of the inverted Sorgenfrei ^Tornquist zone in the Kattegat area. Note the alternation between relatively sheet-like units (2, 4 and 6) and units showing marked wedging and clinoform bedding towards NE (1, 3 and 5) considered diagnostic of contourite drift formation. STZ = Sorgenfrei ^Tornquist Zone. Seismic units 1^6: mainly Maastrichtian (and uppermost Campanian?); 7: Danian; 8: Selandian; 9: Pleistocene and Holocene. (b) Seismic section. (a) Its seismic stratigraphic interpretation (after Surlyk and Lykke-Andersen, 2007, with permission from Blackwell Publishing). A multicolour version of this figure is on the enclosed CD-ROM.

elements of the Late Cretaceous sea floor, including: a complex mounded drift (late Maastrichtian), a sediment wave complex (late Santonian–early Maastrichtian), an elongate moat–drift system (Campanian) and erosional scours or channels of a major unconformity (late Campanian–early Maastrichtian) (Esmerode et al., 2007). The regional-scale unconformity separates two intervals of dominant drift sedimentation and indicates the presence of strong or more focused bottom currents during the late Campanian to early Maastrichtian, coinciding with a stage of significant Late Cretaceous transgression (Haq et al., 1987). The erosive event was more pervasive in the northeastern part of the Øresund area, i.e. closer to the southwesterly dipping slope of the Sorgenfrei–Tornquist Zone. This supports the idea of a main contour current flowing northwestward through the area with a positive lateral velocity gradient towards the slope, generated by the Coriolis force (Esmerode et al., 2007). The complex mounded drift (15 km long and up to 160 m thick) comprises wedge-shaped seismic units that show opposing migration directions. The units are interpreted to represent intervals during which the current axis shifted laterally across the Øresund area (Esmerode et al., 2007). The internal architecture and the

332

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

crests of the sediment waves (2 km in wavelength and up to 30 m in amplitude) trend roughly parallel to the basin margin. 17.2.3.1. Palaeocirculation Lykke-Andersen and Surlyk (2004), Esmerode et al. (2007) and Surlyk and Lykke-Andersen (2007) consider the Late Cretaceous epeiric current system of the NW European ‘‘chalk sea’’ as analogous to thermohaline current systems flowing parallel to the contours of continental margins and being responsible for modern contourite deposition. The slope of the Sorgenfrei–Tornquist Zone acted as an analogue to a continental margin during the Late Cretaceous high sea-level stand. The rise in sea level caused a worldwide overstepping of the Early Cretaceous basin margins and the opening of seaways linking the colder water Arctic and Barents Shelf with the warmer water North Atlantic and the warm water Tethys ocean (Ziegler, 1990). The bottom currents affecting the Danish Basin probably had their origin in warm shallow shelf waters along the northern margin of the Tethys Ocean. Strong thermohaline circulation would be initiated with the formation of northwestward flowing bottom currents that crossed the European epeiric sea due to the density difference between the warm salty waters of the northern Tethys and the colder but probably less saline waters of the North Atlantic. 17.2.3.2. Problems Identification of these Late Cretaceous contourites is based mainly on mediumand large-scale features such as regional variations in thickness, unconformities, drift geometry, elongation and propagation trends, palaeoceanographic features, continental reconstructions and basin location. Unfortunately, hardly any small-scale criteria such as sediment facies characteristics and distinctive microsequences are provided by the authors. Neither gamma-ray logs nor sedimentary logs indicate a facies change at the Campanian/Maastrichtian regional unconformity in the Øresund area. Self-potential and resistivity curves vaguely indicate the presence of slightly coarser material at and above the unconformity surface, which may represent reworked and winnowed deposits (Esmerode et al., 2007). Referring to erodibility experiments on modern pelagic carbonate ooze (Black et al., 2003), however, Surlyk and Lykke-Andersen (2007) argue for a very low threshold stress for entrainment of chalk ooze due to the non-spherical nature of coccolith platelets. Non-deposition and erosion of pure chalk ooze probably resulted from current velocities of 8–20 cm s–1 at approximately 20 m above the sea floor.

17.2.4.

Carbonate contourites, Devonian pelagic successions, Europe and North Africa

Oczlon (1990) highlighted the fact that condensed sections and stratigraphic hiatuses are widespread and extensive in Middle (to Late) Devonian successions of Europe

¨neke and D.A.V. Stow H. Hu

333

and North Africa, and interpreted the related disconformities as erosional features caused by strong bottom currents. He claimed various siliciclastic and carbonate formations in Algeria, Morocco, France, Spain, Germany and Czechia as examples of contourites that are bounded by such disconformities. Unfortunately, the sedimentary characteristics have been insufficiently specified and the criteria used by Oczlon (1991, 1994) for recognition were those of Hollister and Heezen (1972), which date from the early, erroneous interpretations of many fine-grained turbidite successions as contourites (see Stow and Lovell, 1979; Pickering et al., 1989; Stow et al., 1998a). Nevertheless, recent work by Hu¨neke (2001, 2006, 2007a, b) has demonstrated that there are contourites bordering such hiatuses within otherwise pelagic and hemipelagic successions. Based on a combination of sedimentological evidence, facies and microfacies interpretation, taphonomic observation, biostratigraphic age control and correlation, Hu¨neke (2001, 2006, 2007a, b) provides clear evidence of ancient carbonate contourites in the Givetian–Frasnian successions of the Moroccan Central Massif, the Harz Mountains in Germany and the Carnic Alps in Austria/Italy. In all cases, the contourites are associated with phosphatic hardgrounds and phosphorites. Grain-size analysis and sediment characterization are based on thin sections of closely sampled successions. Locally, the formation of neomorphic sparite is a limiting diagenetic phenomenon, but non-carbonate particles provide a point of reference for the primary grain size of calcite particles (Figure 17.4). By analogy with modern contourite facies, calcarenites, laminated calcisiltites as well as mottled calcisiltites and calcilutites have been distinguished (Hu¨neke, 2007a). Calcarenites are mainly represented by styliolinid grainstones to packstones with rarely preserved horizontal and cross-lamination (Figure 17.5). Laminated calcisiltites are particularly rich in noncarbonate components with a higher density than calcite (conodonts, phosphatic intraclasts). Grain fragmentation and reworking of older bioclasts are evident. Both calcarenites and laminated calcisiltites show generally faint structures indicative of current-induced deposition, together with bioturbation throughout. Burrow mottled calcisiltites and calcilutites formed under conditions of weak currents. Parts of coarsening-upward and fining-upward micro-sequences are best preserved in the Moroccan record. These micro-sequences are mostly between 2 and 10 cm thick, and in many cases include an omission surface, which separates the top of the coarsening-upward interval from the base of the fining-upward interval (Figure 17.6). The vertical facies variation can be interpreted in terms of fluctuations in velocity of the transporting bottom currents or shifts of the current axes, whereas the nature and distribution of components result from both longlasting phases of high current strength and the fragility of the biogenic particles (Hu¨neke, 2007b). The transition from mud- to grain-supported textures within the coarsening-upward unit obviously represents the waxing phase of the current. As the current continues to accelerate, the non-cohesive bioclastic sediment may easily be reworked. Repeated and long-lasting high current velocities favour sediment bypassing and the formation of pristine or condensed phosphorites (see Fo¨llmi, 1996). Concurrently, thin-shelled biogenic tests such as styliolinids became more and more fragmented. As the current waned, there were hardly any sand-sized particles left and silt-sized bioclasts were first deposited with subsequent calcareous mud deposition.

334

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

(b)

2 1

3

1

(a) (a)

1

2

Figure 17.4 Early Frasnian limestones interpreted as calcareous contourites, Gara de Mrirt, Moroccan Central Massif. Thin sections show microfacies of mottled calcisiltites and calcilutites (bioclastic wackestones) and laminated calcisiltites at sample point M52 (see Hu«neke, 2007a). Indistinct lamination and non-deposition surfaces are partly obscured by bioturbational mottling. (a) Bioclastic wackestones overlain by laminated calcisiltite. (b) Styliolinid packstone blanketed by calcisiltite showing normal grading. Black triangles at photograph margin indicate non-deposition surfaces. Erosional character is well preserved since pressure-dissolution seams apparently do not mask facies boundaries. The laminated calcisiltites are particularly rich in phosphate intraclasts (1), quartz (2), feldspar and conodonts (3). Grain-size of non-carbonate particles indicates that carbonate diagenesis has not substantially modified the original grain size of the coarser carbonate fraction, whereas much of the finer fraction is now diagenetic micrite.

Regional trends and other medium-scale characteristics are compatible with a contourite interpretation. Overall, very low net accumulation rates prevailed during bottom-current-induced deposition, bypassing or erosion. Strongly condensed successions with distinctive reduced intervals and biostratigraphic hiatuses, which occur mainly at the base of the bottom-current deposits, characterize the Givetian and early Frasnian record, in contrast to the more expanded and

335

¨neke and D.A.V. Stow H. Hu

(a) 3 4

2

1 mm

1

5 6

5 mm

Figure 17.5 Early Frasnian limestones interpreted as calcareous contourites, Gara de Mrirt, Moroccan Central Massif. Cross-laminated calcarenite (styliolinid grainstones ^ packstones) overlain by mottled calcisiltites and calcilutites (styliolinid wackestones, bioclastic wackestones) at sample point MG1 (see Hu«neke, 2007a). Note burrowed/mottled omission surface (1) and palimpsest biogenic traces (2). (a) The calcarenite is represented by styliolinid grainstones ^ packstones. The thin section is oriented perpendicular to the bedding and traces of the cross-lamination run at an angle from upper right to lower left. Nearly all the conical styliolinid shells have their long axis oriented perpendicular to the slip-face dip direction (3; O-shaped sections) but an orientation with the frustum of the cone in dip direction (downcurrent) is common, too (4; V-shaped sections). Pore space is mainly filled by radiaxial acicular calcite. Phosphorite intraclasts (5) and conodonts (6) commonly occur within the calcarenite facies.

complete sequences of older and younger deposits. In addition, erosional surfaces and hardgrounds are widespread. Marked variations in thickness, accumulation rates, volumes of eroded sediment, and microfacies of the principle fossil contourite units have been documented across the three regions. Stratification types of the associated phosphatic sediments give additional evidence of increased hydraulic energy. These features are best compatible with moat areas of contourite drifts that experienced bottom currents with higher velocities and more marked erosion than the drift crests (see Fauge`res et al., 1993; Stow et al., 2002c).

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

Mottled calcilutites and calcisiltites 15

Laminated calcisiltites

Omission:

Mean grain size mm

Depositional fabric Bioclastic wackestone

e ty de c r eas Veloci

Facies cm types

0. 0 0. 02 0 0. 04 0 0. 08 0 0. 16 0 0. 31 0 0. 62 1 0. 25 2 0. 50 50 0

336

Bioclastic wackestone and Intraclastic calcisiltite

Particle abrasion, erosion and phosphatization

10 Styliolinid grainstone

5

Veloc ity

increase

Calcarenites Styliolinid packstone

Styliolinid wackestone

Styliolinid packstone

Mottled calcisiltites and calcilutites

Bioclastic wackestone

0

Figure 17.6 Composite facies model for Devonian bottom-current deposits showing the complete vertical superposition of a coarsening-upward and a fining-upward micro-sequence (see also Hu«neke, 2007b). These idealized micro-sequences are between 2 and 15 cm thick and include in many cases an omission surface in between. The variation in mean grain size, composition and depositional texture is interpreted as recording an increase and decrease in bottom-current velocity, respectively, for the coarsening-upward and fining-upward micro-sequence. Oblique hatching (light blue in the coloured figure in the CD-ROM) indicates calcarenites that are cemented by coarse sparry calcite (packstones ^ grainstones). A multicolour version of this figure is on the enclosed CD-ROM.

17.2.4.1. Palaeocirculation Bottom-current-induced deposition occurred contemporaneously in different settings of the narrow oceanic passageways between the approaching continents of Gondwana and Laurussia (Figure 17.7). The areas affected were the southeastern

¨neke and D.A.V. Stow H. Hu

337

Figure 17.7 Palaeogeographic reconstruction of Laurussia and northern Gondwana during the Givetian ^ Frasnian (370 Ma), showing expected pattern of oceanic surface circulation as well as distribution of deep-sea, shelf-sea, land and mountain ranges (from Hu« neke, 2006, with permission from the Geological Society, London).The map is based on reconstructions by Kiessling (in Copper, 2002) and Golonka (2002). Pelagic successions of Europe and northern Africa that include hiatuses at the Givetian/Frasnian boundary are sedimentary records of Gondwana (MC = Moroccan Central Massif; AA = Anti-Atlas; AB = Ahnet Basin), the Armorican terrane assemblage (MN = Montagne Noire; FW = Frankenwald), the Noric terrane (CA = Carnic Alps) and Laurussia (RS = Rhenish Slate Mountains; HM = Harz Mountains). A multicolour version of this figure is on the enclosed CD-ROM.

Rhenish Sea shelf, which occupied the distal passive margin of Laurussia, the disintegrated northern continental margin of Gondwana, whose sedimentary record is now preserved in the Moroccan Meseta, and deep marginal plateaus of the Noric Terrane in the western part of the Prototethys (Hu¨neke, 2006, 2007a). While the contourite accumulation in the Harz Mountains and the nappe of Ziar–Mrirt occurred on pelagic carbonate platforms, the reworking in the Carnic Alps occurred in proximity to a carbonate slope–apron system. The contemporaneous occurrence and the likelihood of bottom currents according to the palaeogeographic reconstruction both support, independently, the interpretation as ancient contourites. The nature of bottom circulation within the oceanic passages between Gondwana and Laurussia during the Devonian is not well known. Contrasting palaeotectonic–palaeogeographic models are the main reason (Ziegler, 1989; Scotese and Mc Kerrow, 1990; Golonka, 2002; Stampfli and Borel, 2004; Torsvik and Cocks, 2004). Among them, close-fit reconstructions explain the Givetian/ Frasnian circulation event most convincingly. Under such conditions, the time span witnessed constriction of the main surface flow from the Prototethys to the Protoatlantic (Hu¨neke, 2006, 2007a). It seems most likely that continued constriction of the oceanic seaways led to an intensification of the westward-directed oceanic current, which was therefore capable of influencing sedimentation at

338

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

greater depths. Alternatively, there may have been a deeper counter flow to the east. According to the principle of continuity, a broad and thin surface current which has to pass a narrow oceanic strait may accelerate or swell to greater depth (Brown et al., 1991). Under both circumstances, the sediment accumulation in pelagic depositional environments would have been strongly affected, particularly on the top of topographic highs. This intensified exchange of warm saline water and cold less saline water between the equatorial ocean in the northeast and the sub-polar oceanic realm in the southwest ended during the middle Frasnian, when continued convergence between Gondwana and Laurussia led to firm contact of their continental margins (e.g. Ziegler, 1989; Golonka, 2002). 17.2.4.2. Problems In general, the interpretation of these Devonian fossil contourites is well supported by evidence at three different scales of observation. The principal problems mainly relate to the relative age of the deposits as follows. 1. Diagenesis in older carbonate successions becomes progressively more intense so that many of the original depositional features may become obscured. Nevertheless, the centimetre-scale cycles of grain size, structure and compositional variation are very similar to the standard mud–silt–sand contourite sequence of Stow et al. (2002c). The coarsening-upward micro-sequences of the Moroccan case may be considered as base-only partial sequences (C1–3), and the fining-upward micro-sequences may be referred to as top-only partial sequences (C3–5). The main differences compared to the standard sequence are those in the Moroccan example: (1) the coarseningupward and fining-upward micro-sequences are commonly separated by a sharp contact or erosional surface, which may be phosphatized; and (2) the overall sequence is on a rather smaller scale than that of the sequence model. However, both these aspects can be related to deposition under relatively high-energy currents. 2. Further work is required on these examples as well as on comparable Devonian successions elsewhere in related depositional areas to reconstruct the shape and 3-D geometry of the entire sedimentary bodies and to find reliable indications of palaeocurrent directions. However, the tectonically disturbed stratification – due to the Variscan deformation in the studied areas – makes a geometric reconstruction of the contourite drift bodies very difficult. 3. Precise palaeoceanographic reconstructions are notoriously difficult for the Palaeozoic, and hence the discussion of bottom-current patterns and influence is less rigorous than for the Cenozoic, for example. Nevertheless, we are confident that the inferred palaeocirculation pattern is correct.

17.2.5.

Carbonate contourites, Ordovician Jiuxi Drift, China

An ancient carbonate contourite drift has been identified within an Early Ordovician deep-water succession in the Hunan province of China, near Jiuxi on the Yangtze Platform (Duan et al., 1993; Luo et al., 2002). Its recognition is based on small-, medium-, and large-scale criteria as set out by Stow et al. (1998b, 2002c).

¨neke and D.A.V. Stow H. Hu

339

Analysis of field data, thin sections and polished slabs reveal that the general sedimentary characteristics of the Jiuxi Drift deposits are typical of those described from many modern drift systems (Luo et al., 2002). The main contourite facies are bioturbated calcilutites and burrow-mottled calcisiltites, both of which show some irregular discontinuous lamination, together with a lesser proportion of irregular laminated and highly bioturbated calcarenites. These occur in repeated coarsening-upward to fining-upward micro-sequences typically 30–80 cm thick, locally up to 2 m. Possible calcirudite contourite lag deposits and coarser-grained bioclastic contourites are also identified. The bioclastic contourites usually form lenticular beds and comprise grains of a varied assortment of biogenic debris (pelagic, benthic and resedimented origin) and some non-biogenic admixtures (terrigenous clays and rounded quartz grains). Based on detailed logging and on fourteen sections, the geometry of the contourite body has been reconstructed (Duan et al., 1993). The Jiuxi contourites form a distinct mound-like drift some 350–450 m thick elongated parallel to the southern continental margin of the Yangtze Terrane. It accumulated at a sedimentation rate of 35– 40 m Ma 1. Features of traction flow coupled with intense bioturbation and alongslope current indicators support a bottom-current origin. Large-scale cross-stratified units found in parts of the calcilutitic contourite section are believed to result from sea-floor development of mud waves and/or erosion furrows under the influence of a semi-permanent bottom-current regime (Duan et al., 1993). In addition, pelagites, hemipelagites and distal turbidites make up a small proportion of the drift. The Jiuxi Drift formed during the Tremadocian along the relatively stable passive margin in an inferred mid-slope position south of the Yangtze shallow-water carbonate platform (Luo et al., 2002). The depositional area had been a deep-water environment dominated by gravity flows since the Middle and Late Cambrian, and continued to receive some carbonate shallow-water debris during the early Ordovician global sea-level rise. The drift formation may have been prompted both by the rise in sea level and by a change in strike orientation of the slope from SW–NE to more or less E–W. 17.2.5.1. Palaeocirculation The Yangtze Terrane and related continental blocks of China were probably incorporated within or close to the Australian–Antarctic section of east Gondwana, which spanned a large part of the southern hemisphere and was continuous from the south pole to equator (Metcalfe, 1996). The Jiuxi contourite drift formed in close proximity to the western Gondwanan shelf margin at low palaeolatitudes. The low palaeolatitude of the Yangtze Terrane gave way to bryozoan, lithistid and Calathium reef growth on the bordering shallow, warm-water Yangtze platform during the Tremadocian (Webby, 2002). The palaeogeographic reconstructions suggest that the slope of the Yangtze Terrane was influenced by the Eastern Boundary Current of the Palaeo-Asian Ocean at low northern latitudes (Golonka, 2002; Rowland and Shapiro, 2002; Webby, 2002). This southward-directed surface current transported cool water towards the equator, similar to the California Current in the western Pacific today.

340

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

By analogy with recent conditions (e.g. Cornuelle et al., 2000; Bograd et al., 2001), we should take the view that much of the transport occurred as ‘‘squirts and jets’’: upwelling centres associated with capes and promontories, offshore transport in intense jets deep into the epeiric seas, and a vigorous field of meso-scale eddies that affected the slope of the Yangtze continental margin. 17.2.5.2. Problems The three principal concerns we have with respect to this interpretation of the Ordovician contourites relate in part to their age and in part to some of the facies interpretations. 1. As already noted, diagenesis in older carbonate successions obscures many primary features. Most of those described from the Jiuxi Drift appear compatible with a contourite interpretation, although the presence of largescale cross-stratification and its interpretation as the result of sediment waves or erosional furrows remain speculative. We know of little modern data of comparable type that would help support the authors’ interpretation (see Martı´n-Chivelet et al., 2008). 2. The regional context and variation in unit thickness certainly appears to conform with a mounded-drift geometry. However, there must remain some caution concerning the rigour of dating and correlation between measured sections, and whether or not tectonic factors may have influenced variations in section thickness. 3. More crucial for contourite-drift formation, the nature of bottom circulation during the Ordovician is not well known. The period is part of Fischer’s (1981, 1984) global late Cambrian to Devonian greenhouse state, with oceanic modes represented by warmer seas, gentler latitudinal and vertical gradients, sluggish circulation, an expanded oxygen minimum zone below the thermocline, and higher sea levels (Fischer and Arthur, 1977; Martin, 1996). Berry and Wilde (1978) emphasized the greatly different oceanic circulation conditions during the Early–Middle Ordovician. Almost the entire oceanic water column, except for the oxygenated wind-mixed upper surface waters (100 m), was represented by warm, poorly ventilated anoxic waters.

17.3. 17.3.1.

D ISCUSSION

Recognition of ancient contourites

It is first important to reaffirm the point made in several previous publications on fossil contourites (e.g. Stow et al., 1998a), that it is both difficult and time-consuming to recognize the influence and deposits of bottom currents in ancient series. Whereas many turbidites or debrites, for example, are readily identified as such even with cursory field examination, the same is not true for contourites. Several different lines of evidence must be carefully gathered, following the three-stage approach most recently summarized in Stow et al. (2002c) and repeated here in Table 17.1. With regard to the small-scale criteria and the sedimentary characteristics of contourite facies, the examples discussed in the present chapter all conform to the specific

¨neke and D.A.V. Stow H. Hu

Table 17.1

341

Criteria for the recognition of contourites in both modern and ancient systems

Stage 1: Small-scale (field, borehole or lab) Do the sediments have the range of features as summarized in the discussion or as described by Stow et al. (2002d)? Where there is a possibility of mixed turbidite/contourite sequences, can a distinction be made between the two facies on the basis of character and/or palaeocurrent evidence? Is there sufficient evidence to discount deposition from fine-grained turbidity currents? Particular care must be taken for inferred reworked turbidites. Where there is a possibility of mixed hemipelagite–pelagite/contourite sequences, is there sufficient evidence for the influence of bottom currents during sedimentation? Can any cyclicity present be related to variation in bottom current velocity rather than to variations in terrigenous input or biogenic productivity? Stage 2: Medium-scale (drift, formation or region) Do regional trends in facies occurrence, palaeocurrent directions, textures, mineralogical or geochemical tracers exist that would support a bottom-current origin? Is there any other evidence of bottom-current activity such as unconformities, condensed sequences, regional variation in thickness, drift geometry, etc.? Is it possible to reconstruct the shape and 3-D geometry of the whole sedimentary body? And, if so, are the elongation and propagation trends parallel or perpendicular to the inferred margin? Are the associated facies, palaeontological data and rates of accumulation compatible with a contourite interpretation? Stage 3: Large-scale (system, ocean or continent) Do the conclusions from Stages 1 and 2 above fit with what is known from other independent lines of evidence concerning major oceanographic or palaeoceanographic features and continental reconstructions? What kind of bottom-current systems exist at present or might have existed in the study area at the time of deposition, taking into account constraints imposed by known palaeoclimatic conditions and inferred basin location and geometry? Source: Modified from Stow et al. (1998a, 2002d).

characteristics as encompassed by the standard facies model for contourites and tabulated in previous works (e.g. Stow et al., 1998a, 2002c; see also Martı´n-Chivelet et al., 2008). We will not repeat this summary here, but note that all the examples highlighted are, in fact, carbonate-rich contourites, and that some of their features may be more closely associated with the calcicontourite facies. These include: Occurrence and facies: within thick successions of pelagic and hemipelagic sediments, possibly associated with regional hiatuses and/or condensed sequences with phosphorite or ferromanganese hardgrounds; calcilutite, calcisiltite and calcarenite facies most common. Structures and ichnofacies: mostly very subtle structures and ghost primary structures after intense bioturbation; calcarenites can show distinctive lenticular, very thinbedded character, and less bioturbation; range of normal deep-water burrow traces including some vertical burrows from omission surfaces. Texture and sequences: diagenesis tends to preserve the original calcilutite to calcarenite grain-size differences and hence grain-size/facies sequences; such sequences can be of limited thickness (<20 cm) and incomplete.

342

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

Microfacies and composition: packed to sparse biomicrites (including wackestones, packstones and some grainstones) are common as microfacies; dominant bioclasts, mostly of pelagic type but including admixture of benthic and reworked older biogenic material, and variable proportion of terrigenous grains/clays; diagenetic micrite, microspar and sparite all present. Bioclast fragmentation and iron staining is common. The sedimentary characteristics are best discernible where contourites are found embedded in typically fine-grained pelagic sediments, whereas in basin fills substantially supplied by sub-aqueous density flows, which may transport coarse or fine sediment, these small-scale criteria are commonly of limited value. The identification of ancient contourites becomes even more speculative where diagenesis, metamorphism and tectonic deformation have largely modified the depositional structures. Diagenetic alteration can be intensive in particular in calcareous and siliceous biogenic contourites. Therefore, investigations should always include careful examination of the diagenetic effects on particle size by neomorphic processes or tectonic stress. Admixtures of siliciclastic components and more stable phosphatic grains may serve as a point of reference for the primary particle size (Figures 17.4 and 17.5). With careful fieldwork and observation, where exposure permits, medium-scale criteria can be extremely useful in helping to elucidate a contourite interpretation. For the examples cited above, the recognition of hiatuses and condensed sequences has been important, as well as regional variation in the thickness of depositional units allowing reconstruction of mound-like geometry. The associated facies are dominantly of deep-water pelagic and hemipelagic type, although more work is required to observe any regional trends in detailed facies characteristics. Where the palaeowater depth is not well constrained and the type of current system unknown, it is crucial to analyse the geological context. Very deep tidal currents, storm waves, internal waves and other clear-water currents may operate on outer shelves, upper slopes and in straits. The resulting traction deposits are not contourites sensu stricto (Fauge`res and Stow, 1993; Stow et al., 2002c). In ancient series, above all, it is not necessarily possible to distinguish the seismic features or sediment facies that result from these currents from those of bottom currents. Where a distinction can be made, this should be made clear in the terminology applied. Large-scale criteria for the recognition of contourites, including palaeoceanographic features and continental reconstructions, are generally more problematic and difficult to rely on too heavily, especially in older systems. Most Palaeozoic–Mesozoic platetectonics models only roughly outline the real palaeoceanographic conditions. A possible affirmation of a contourite origin is even more complicated in cases of different palaeogeographic models showing conflicting interpretations. In such cases, the disposition of fossil contourites may be used to evaluate palaeotectonic– palaeogeographic reconstructions and one should be aware of circular reasoning. The Devonian situation as discussed by Hu¨neke (2006) may serve as an example. The close-fit reconstructions (Pangaea-A reconstructions) of Golonka (2002) and Kiessling (in Copper, 2002) explain the Givetian/Frasnian circulation event most convincingly. The disposition of the fossil calcareous contourites and faunal data corroborate palaeogeographic reconstructions that show an advanced convergence between Gondwana

¨neke and D.A.V. Stow H. Hu

343

and Laurussia and the smaller continental plates such as the Armorican and Noric terranes during the Givetian and Frasnian.

17.3.2.

Sea level and preservation of contourite drifts

All the cases of Palaeozoic and Mesozoic contourites exemplified above reveal the influence of bottom currents in times of global sea-level highstand and greenhouse conditions (see Copper, 2002; Johnson et al., 2002; Rowland and Shapiro, 2002). At least the second is surprising since greenhouse conditions should be associated with more sluggish bottom currents due to a less strong thermohaline oceanic circulation. The Meso- and Palaeozoic examples document, however, that bottom currents have had the competence to rework pelagic sediments at least where they traversed shelf and epeiric seas. The Ordovician is characterized as a period of maximum eustatic high within the older (Palaeozoic) first-order cycle with consequent maximum flooding, peaking in the Early and again in the Late Ordovician (Vail et al., 1977a; Hallam, 1992; Ross and Ross, 1995). During the Early Ordovician, the Jiuxi contourite drift formed in close proximity to the western Gondwanan shelf margin at low palaeolatitudes (Luo et al., 2002). The sea level was probably at one of its highest points of the entire Phanerozoic. The resultant Ordovician epeiric seas are depicted as covering at least 40%, possibly even 60%, of earlier land areas. During the Devonian, the Palaeozoic first-order sea-level rise culminated again in Middle/Late Devonian times ( Johnson and Sandberg, 1988; Hallam, 1992; Golonka and Kiessling, 2002). In this episode, bottom currents reworked pelagic carbonate oozes along southern margin of Laurussia, the disintegrated northern continental margin of Gondwana, and on deep marginal plateaus of several adjoining micro-continents (Hu¨neke, 2007a). The Late Cretaceous is the time of maximum eustatic high within the younger (Mesozoic-Cenozoic) first-order cycle (Haq et al., 1987). The Albian– Cenomanian was an interval of increasing continental submergence. The highest Phanerozoic sea level was reached during the early Turonian. Following a temporary lowstand, the Campanian sea level was again high, which slowly lowered during the Maastrichtian and than dropped dramatically during the Palaeocene. During the Early Ordovician, the Middle/Late Devonian and the Late Cretaceous, the continents were extensively flooded. At high sea-level stands, the combination of reduced source areas for terrigenous sediments and the expansion of epeiric seas with water depths of more than 300 m may, therefore, have created generally favourable conditions for deposition of contourite drifts on the continental lithosphere. The preservation potential of these drifts is relatively high since the continental lithosphere is non-subductible, and this may lead to a stratigraphic bias in the record of ancient contourites. Pre-Cenozoic cases are preferentially preserved from periods of high-stand sea level, which favour the deposition of contourites on the continental lithosphere and thus its preservation. At least, pre-Jurassic contourite drifts that were originally deposited on oceanic lithosphere have been destroyed on a large scale due to global subduction processes. However, the same does not appear to be true regarding the Cenozoic examples discussed. Sea levels during the Oligocene and then during the mid to late Miocene

344

Identification of Ancient Contourites: Problems and Palaeoceanographic Significance

were relatively low, and climatic conditions moved towards an icehouse world. The Cyprus contourites were deposited over oceanic crust and later preserved by ophiolite obduction and uplift during subduction of the late Tethys Ocean. The Japanese examples are also inferred to have been deposited in a distinctly oceanic setting and their preservation on land related to arc–arc accretion.

17.3.3.

Sequence interpretation and non-deposition surfaces

The standard contourite sequence may reflect at least sub-regular periodicity in mean bottom-current velocity, related to climatic variations or other external factors. For many recent and sub-recent examples cited in the literature, little attempt has been made to calculate the duration of the deposition of such sequences. This is partly because sequences show very variable absolute thicknesses and also because there is considerable evidence for the presence of hiatuses and intervals of reduced sedimentation. However, based on data provided for various examples in the recent contourite compilation (Stow et al., 2002f), we have calculated approximate sequence periodicities for several examples (Gulf of Cadiz, Rockall Margin, West Shetlands, and Norwegian Margin), and in each case obtain a figure of between 5000 and 20,000 years. By comparison, the ancient examples cited above, for which reliable estimates can be made, yield an average periodicity of 200,000 years for the Ordovician Jiuxi Drift (Luo et al., 2002), and 20,000–250,000 years for the Cyprus examples (Stow et al., 2002d). For the Devonian contourites, various non-deposition surfaces are documented, representing periods of sediment bypassing or even temporary minor erosion (Hu¨neke, 2007a). Omissions surfaces, which mark temporary halts in deposition but involve little or no erosion (Heim, 1924; Bromley, 1975), are most widespread. Burrowed hardgrounds, pristine and condensed phosphates, erosional surfaces and encrusted surfaces commonly occur. Although many of the omission surfaces record long time spans of non-deposition, encrusted surfaces with micro-stromatolitic crusts may record breaks of no longer than annual or even noctidiurnal length. The duration of the depositional hiatuses associated with the formation of hardgrounds has been estimated at somewhere between months to hundreds or even thousands of years (Flu¨gel, 2004). Fluorapatite precipitation is generally relatively slow, having growth times between 500 and 2000 years in modern phosphorites within the top few tens of centimetres of the sediment/water interface, where microbial activity is greatest (Glenn, 1990). Taking these omission surfaces into account, the periodicity can be roughly estimated as 50,000–100,000 years for the Devonian case.

ACKNOWLEDGEMENTS We thank reviewers Jean-Claude Fauge`res, Michele Rebesco and Tom van Loon for providing valuable suggestions to improve this manuscript.

P A R T

6

DOMAINS

C H A P T E R

1 8

A BYSSAL P LAIN C ONTOURITES F.J. Herna´ndez-Molina1, A. Maldonado2 and D.A.V. Stow3 1

Facultad de Ciencias del Mar, Universidad de Vigo, Vigo, Spain Instituto Andaluz de Ciencias de la Tierra. C.S.I.C./ Universidad de Granada, Granada, Spain 3 National Oceanography Centre, Southampton (NOCS), Southampton, UK 2

Contents 18.1. Introduction 18.2. Terminology 18.3. Case Studies of Abyssal Contourites 18.3.1. Southern hemisphere basins 18.3.2. Northern hemisphere basins 18.3.3. Equatorial abyssal plains 18.4. Oceanic Gateways 18.5. Principal Oceanographic and Sedimentary Processes 18.5.1. Long-term hydrological conditions 18.5.2. Nepheloid layer 18.5.3. Large eddies 18.5.4. Benthic storms 18.5.5. Effect of sea-floor obstacles 18.6. Main Characteristics of Abyssal Plain Contourites 18.6.1. Large-scale depositional and erosional features 18.6.2. Depositional units, discontinuities, and seismic facies 18.6.3. Sedimentary facies 18.7. Final Considerations Acknowledgements

18.1.

347 350 351 351 359 359 364 365 365 365 367 367 369 370 370 374 374 375 378

I NTRODUCTION

Ocean basins situated between the continental margins and mid-oceanic ridges represent the deepest of the marine domains. They are more or less synonymous with the (oceanic) basin plains as defined by other authors (Stow, 1985; Pilkey, 1987; Pickering et al., 1989; Pilkey and Holkinson, 1991). These basins or basin plains (Figure 18.1) contain up to 10 distinct morphological elements: (1) continental rise; (2) abyssal plains; (3) oceanic rises, (4) distal fans and their distributary channels; (5) sediments drifts; (6) abyssal hills; (7) seamounts; (8) transfer fracture zones; (9) mid-ocean channels; and (10) oceanic trenches (Kennett, 1982; Boillot, 1984; Stow et al., 1996b). Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00218-5

Ó 2008 Published by Elsevier B.V.

347

348

Abyssal Plain Contourites

180°

80°N 90°E

90°W

60°N 40°N

0°

20°N 0° 20°S 40°S 60°N

120°E

160°E

160°W

120°W

80°W

40°W

0°

40°E

80°E

Figure 18.1 Distribution of abyssal plains and general water-mass circulation model in the ocean basins (modified and adapted from Bearmon, 1989; Pickering et al., 1989; Garrison, 1996). Legend for the water masses ^ a = areas of deep-water formation; b = main upwelling areas; c = areas of intermediate water formation. Legend for the abyssal plains (in black) ^ Antarctic Ocean: 1 = Amundsen; 2 = Bellinghausen; 3 = Enderby; 4 = South Indian; 5 = Valdivia; 6 = Weddell. Arctic Ocean: 7 = Barents; 8 = Boreas; 9 = Canada; 10 = Chukchi; 11 = Dumshaf; 12 = Fletcher; 13 = Greenland; 14 = Mendeleyev; 15 = Northwind; 16 = Pole; 17 = Wrangel. Atlantic Ocean (N): 18 = Barracuda; 19 = Biscay; 20 = Blake Bahama; 21 = Ceara; 22 = Dermerara; 23 = Gambia; 24 = Guinea; 25 = Hatteras; 26 = Hispaniola; 27 = Horseshoe; 28 = Iberian; 29 = Madeira;30 = Nares;31 = Para;32 = Porcupine;33 = Seine; 34 = Sierra Leone; 35 = Silver; 36 = Sohm; 37 = Tagus; 38 = Vidal. Atlantic Ocean (S): 39 = Angola; 40 = Aghulas; 41 = Argentine; 42 = Burdwood; 43 = Cape; 44 = Namibia; 45 = Pernambuco; 46 = Town. Caribbean: 47 = Columbian; 48 = Grenade; 49 = Jamaican; 50 = Panama;51 =Venezuela;52 = Yucatan. Gulf of Mexico: 53 = Florida;54 = Sigsbee. Indian Ocean: 55 = N Australian; 56 = Mid Indian; 57 = Cocos; 58 = Cuvier; 59 = Gacoyne; 60 = Mascarene; 61 = Perth; 62 = Somali, 63 = S. Australian. Sea of Japan: 64 = Japan. Mediterranean: 65 = Adriatic; 66 = Alboran; 67 = Balearic; 68 = Sicilia; 69 = Sidra; 70 = Tyrrhenan. North Sea: 71 = Norway. North Pacific: 72 = Alaska; 73 = Aleutian; 74 = Cacadia; 75 = Tufts; South Pacific: 76 = Mornington; 77 = Raukumara. Sea of Okhotsk: 78 = Okhotsk. South of China: 79 = South China Sea.Tasman Sea: 80 = Tasman.

Abyssal plains were discovered in 1947 in the North Atlantic (Heezen and Laughton, 1963), and around 80 major abyssal plains are recognised nowadays in the oceans (Pickering et al., 1989, Figure 18.1). They cover approximately 4.4% of Earth’s surface and represent the deepest, flattest parts of ocean basins that extend to the distal boundaries of the continental rise or submarine fans, between 3000 and 6000 m water depth on average. They have a slope of less than 1:1000 (1 m km 1), a shape from elongated to highly irregular, and are very variable in surface area, from <3000 km2 to > 3,000,000 km2. Their distinctive flatness results from a thin

F.J. Herna´ndez-Molina et al.

349

to thick (0.5–5 km) blanket of sediments, which smoothes out the original topographic irregularities. Smoothing out of abyssal plains fed by mid-ocean channels, or at the distal termination of submarine fans, has been achieved largely by sheetlike turbidite fills (Heezen and Laughton, 1963; Kennett, 1982; Weaver and Thomson, 1987; Pickering et al., 1989). In other cases, smoothing has occurred, at least in part, through the action of bottom currents and deposition of large sheeted drifts (Stow et al., 1996b). Abyssal hills, usually located in groups, represent elevated parts of the oceanic basement, with relief not higher than about 1000 m and slopes of 1–15°. Seamounts, or groups of seamounts, represent isolated volcanoes (conical shape) and/or linear volcanic ridges with slopes similar to those of hills but reliefs of over 1000 m above the sea floor (Kennett, 1982). Prominent transverse ridges may extend for several hundreds of kilometres across ocean basins and their abyssal plains (Bonatti, 1978; Maldonado et al., 2000; Livermore et al., 2004). Ocean basins in general and abyssal plains in particular contain extensive sheeted depositional systems characterised by many thin-bedded distal turbidites, rare but very large-scale megaturbidites, and interbedded hemipelagic and pelagic sediments (Pickering et al., 1989; Lebreiro et al., 1997; Weaver et al., 2000). The occurrence of interbedded contourite sheets (abyssal sheet drifts or sheeted drifts), resulting from deposition by bottom currents, has been widely recognised more recently (Fauge`res et al., 1999; Rebesco, 2005). Sheeted drifts have, in fact, been recognised in many abyssal plains, including the Argentine Basin (Ewing et al., 1971; Flood and Shor, 1988), Labrador Sea (Egloff and Johnson, 1975), Weddell Sea (Maldonado et al., 2005), and Sierra Leona Basin ( Jones and Okada, 2006). The principal bottom-water masses forming part of the global thermohaline circulation (THC) (Figure 18.1) are the North Atlantic Deep Water (NADW) and Antarctic Bottom Water (AABW), which spread across the deepest marine domains, mainly as Deep Western Boundary Currents (DWBC) (Broecker, 1991; Talley, 1999; Rahmstort, 2006). The NADW and the AABW circulation and related processes are influenced by the size of the basins, connectivity between basins, and proximity to the high-latitude source of bottom-water formation. Some of the largest and most prominent abyssal plains are located around the Antarctic continent, and are therefore directly influenced by the AABW circulation (Figure 18.1). Abyssal plains are common throughout the Atlantic Ocean, but significant differences exist between the northern and southern basins. The plains located in the North are large, well-connected, and mainly affected by the NADW in the northern part, but towards the Central Atlantic they are also influenced by the AABW. In the South Atlantic, the abyssal plains are mostly smaller, bounded by topographic barriers, and only locally connected by narrow gateways (Kennett, 1982; Fauge`res et al., 1993; Van Aken, 2007). In the Pacific Ocean, abyssal plains are less common (Figure 18.1) because sediment supply from turbidity currents is mostly trapped in deep trenches and marginal seas. This ocean is influenced by the AABW in the west and by the Pacific Deep Water (PDW) in the east (Kennett, 1982; Owens and Warren, 2001; Van Aken, 2007). In the Indian Ocean, there are several relatively small abyssal plains (Figure 18.1) bounded by many topographic barriers and affected by the AABW in the west and by the Northeast Indian Deep Water (NIDW) in the east (Van Aken, 2007).

350

Abyssal Plain Contourites

Deep-water processes related to water circulation, such as internal waves, tides, benthic storms, eddies, and vortices, are poorly known but believed to be capable of generating depositional and erosional features in abyssal plains (Gao et al., 1998; He et al., 2008). Nevertheless, the effects of bottom currents and related deep-water processes on sedimentation and erosion in ocean basins and abyssal plains are not so well understood compared with their effects on continental slopes and rises. In the near future, however, new findings in this field are anticipated from the application of new deep-water technologies including high-resolution geophysical techniques, manned submersibles, remotely operated vehicles (ROVs), sediment coring, and ocean drilling (Integrated Ocean Drilling Program, IODP). The rapid increase in deep-water exploration by oil companies and the new deep-water swath bathymetry systems will lead without doubt to new findings related to the deep-water oceanic circulation, in terms of processes, drift morphology, sediment stacking patterns, and facies. This chapter aims to summarise much of our current knowledge of abyssal plain contourites, first by presenting key case studies from different abyssal-plain settings and then by discussing some general and specific characteristics of different types of abyssal sediment drift.

18.2.

TERMINOLOGY

Because so little is known about the processes that act in ocean basins, and about the nature of the sediment facies and depositional elements that result from these processes, we have opted to use a more general, less specific terminology, as far as possible. The term bottom current is preferred as the general term for those semi-permanent deep-water currents capable of eroding, transporting, and depositing sediments on the sea floor. Bottom currents are the result of both THC and the major wind-driven circulation pattern of the oceans (Rahmstort, 2006). Generally, these currents are semi-permanent in nature with a net flow along-slope, but can be extremely variable in direction and velocity, plus exhibit giant eddies, and local down-slope, up-slope, or oblique-to-slope flow, especially near the basin’s entrance or exit gateways (Gao et al., 1998, Stow et al., 2002c; Rebesco, 2005). The term contourite is now generally accepted as the term for those sediments deposited or substantially reworked by bottom currents and contour currents sensu stricto. The major accumulations of contourite deposits are referred to as ‘‘drifts’’ or ‘‘contourite drifts’’, for which several classifications have been proposed mainly based on their morphological, sedimentological, and seismic characteristics (McCave and Tucholke, 1986; Fauge`res and Stow, 1993; Fauge`res et al., 1993, 1999; Gao et al., 1998; Rebesco and Stow, 2001; Stow et al., 2002f; Rebesco, 2005). An association of various drifts and related erosional features has been termed a ‘‘contourite depositional system’’ (CDS), by analogy with – and as important as – turbidite depositional systems (Stow et al., 1986, 2002f; Herna´ndez-Molina et al., 2003; 2006a). In the same way, where different CDSs are connected laterally (and vertically) and associated with the same water mass in the same or adjacent basins, we can consider this as a ‘‘contourite depositional complex’’ (CDC).

F.J. Herna´ndez-Molina et al.

18.3. 18.3.1.

351

CASE STUDIES OF ABYSSAL C ONTOURITES

Southern hemisphere basins

Active deep-water circulation in ocean basins and abyssal plains on the southern hemisphere, dominated by the AABW and associated water masses, generates multiple depositional and erosional features that together characterise huge CDCs. 18.3.1.1. Weddell Sea and Scotia Sea CDCs Several connected and small basins bounded by extended continental blocks are recognised in the southern Scotia Sea (Ona Basin, Protector Basin, Dove Basin, and Scan Basin) and northern Weddell Sea (Powell Basin and Jane Basin), near the Scotia–Antarctic plate boundary (Figure 18. 2a). The opening of gateways, which connect the aforementioned basins, occurred around the Middle Miocene (Maldonado et al., 2003). Since then, sedimentary processes have been dominated by the effects of deep-water circulation generating a huge CDC composed of large depositional and erosional features (Figure 18.2). Although each basin has local particularities, three major sedimentary units showing very close similarities can be correlated between basins (Maldonado et al., 2003, 2005, 2006). Two main components can be distinguished among the deep-water masses of the region (Naveira Garabato et al., 2002a; Herna´ndez-Molina et al., 2006b) (Figure 18.2a): (1) the Circumpolar Deep Water (CDW), which flows mostly to the east along the Scotia Sea, and (2) the AABW, which is composed of the Weddell Sea Bottom Water (WSBW) and the Weddell Sea Deep Water (WSDW). WSBW is the deepest water mass and is bathymetrically constrained to circulate exclusively within the Weddell Sea. The WSDW located above the WSBW also flows within the Weddell Gyre preferentially along the north-western Weddell Sea (Antarctic Peninsula slope). This is associated with an active nepheloid layer, which produces a large plastered drift along that slope (Pudsey et al., 1988; Gilbert et al., 1998; Howe et al., 2004). Further north-eastward, the WSDW is divided into two main cores (or ‘‘current paths’’). One core is channelled through Jane Basin and overflows into the Scotia Sea beyond the South Orkney Microcontinent, mainly through the Orkney Passage (OP), Bruce Passage (BP) and Discovery Passage (DP) (Figure 18.2a). The other core moves around the South Sandwich Trench, where it is also associated with an active nepheloid layer, flowing north and exiting into the South Atlantic (Figure 18.2a). Within the depositional complex, large mounded and sheeted drifts are the dominant depositional features (Figure 18.2) produced by the AABW and CDW in the Weddell and Scotia Basins, respectively (Maldonado et al., 2003, 2006). They are 100–600 m thick, commonly with sediment waves at the surface, in some cases (as in Powell Basin) with wavelengths up to 3.7 km and heights up to 80 m (Rodrı´guezFerna´ndez et al., 1997; Howe et al. 1998). Other types of large drift have been described as: (1) ‘‘confined drifts’’, identified in elongated basins (as in Jane Basin or West Scotia Ridge, Figure 18.2d), (2) ‘‘basement-controlled drifts’’ associated with morphologic linear irregularities in the oceanic crust (Figure 18.2e), and (3) a huge ‘‘contouritic fan’’ located in the Scan Basin and the central sector of the southern

352

Abyssal Plain Contourites

F.J. Herna´ndez-Molina et al.

353

Scotia Sea (Maldonado et al., 2003, 2005; Herna´ndez-Molina et al., 2007, Figure 18.2a). All these drifts have similar seismic facies, represented by mainly well-stratified sub-parallel reflections with moderate to high amplitude and good lateral continuity. These are interstratified in a more or less cyclic pattern with thinner units of chaotic, transparent, or weak reflections, as well as minor discontinuities (Maldonado et al., 2006). Erosional features, such as moats, channels, furrows, and large ‘‘sub-circular depressed structures’’, are also recognised adjacent to sea-floor irregularities. Major internal regional unconformities, with high-amplitude and continuous reflections, can be correlated basin-to-basin (Maldonado et al., 2003, 2005, 2006). Pliocene to Quaternary deposits drilled by ODP 697 in the Jane Basin (Barker et al., 1988) show clays and fine silts with subordinate siliceous oozes. Short sediment cores from the abyssal plain of the Weddell Sea reveal homogeneous hemipelagic fine-grained deposits, intensely bioturbated with little or no primary structures that indicate a low-energy depositional environment for the most superficial deposits (Howe et al., 2004). 18.3.1.2. Argentine Basin CDC The Argentine Basin is bounded by the Rio Grande Ridge (north), the Mid-Atlantic Ridge (east), the Malvinas/Falkland Escarpment (south), and the South American continental margin (west) (Figure 18.3). Within this basin, the Argentine Abyssal Plain has a surface area of around 2,00,000 km2, and its deepest part along the western and southwestern margins (known as the ‘‘Abyssal Gap’’) reaches a maximum water depth of 6212 m (Ewing and Lonardi, 1971; Lonardi and Ewing, 1971; Parker et al., 1996, 1997). The Argentine Basin is characterised by very active oceanic circulation, with probably the highest kinetic energy of any ocean basin (Figure 18.3a). There is a complex interaction of several different water masses including the Brasil/Malvinas Confluence (BMC), the Antarctic Intermediate Water (AAIW), and the CDW from the south, and the NADW from the north (Ewing and Lonardi, 1971; Lonardi and Ewing, 1971; Piola and Rivas, 1997; Arhan et al., 2002a, b). In addition, circulation of the AABW is partially trapped in the basin generating a huge cyclonic gyre below 3500–4000 m depth within a water mass up to 2 km thick. This oceanographic regime is clearly significant in controlling sedimentary processes across the entire ocean basin (Le Pichon et al., 1971a; Reid, 1989). The AABW in this basin is composed of WSDW, which enters the basin from the south through the Falkland/Malvinas gaps and around the Falkland/Malvinas Plateau in the east, before turning westwards and passing the Malvinas/Falkland Escarpment

Figure 18.2 Example of the depositional and erosional contourite features in theWeddell Sea and Scotia Sea generated by the action of the Weddell Sea Bottom Water (WSBW) and Weddell Sea Deep Water (WSDW). (a) Location map showing the main identified types of drifts (adapted and modified from Maldonado et al., 2003, 2005; Herna¤ ndez-Molina et al., 2007). Legend for the water masses: 1= surface circulation of the ACC; 2 = CDW; 3 = surface circulation of the Weddell Gyre; 4 = WDW; 5 = WSDW; 6 = WSBW. (b) Multi-channel seismic profiles exhibiting an example of sheeted drift. (c) Elongated-mounded and separated drift. (d) Part of a confined drift. (e) Basement/tectonics-controlled drift. (f ) Patch drift.

354

Abyssal Plain Contourites

F.J. Herna´ndez-Molina et al.

355

until it then turns to the north at the Argentine continental slope. The AABW flows continuously north as an intensified western boundary current penetrating into the Brasil abyssal plain through the Vema and Hunter Channels (Wu¨st, 1957; Le Pichon et al., 1971a; Speer et al., 1992; Fauge`res et al., 1993; Speer and Zenk, 1993; Von Lom-Keil et al., 2002). It is this long-term cyclonic movement of AABW that is principally responsible for the generation of three huge drifts (Zapiola, Argyro, and Ewing Drifts, Figure 18.3a) mainly composed of silts and muds (Ewing et al., 1971; Flood and Shor, 1988). The most prominent is the Zapiola Drift, a giant elongated-bifurcated sediment drift centred in the southern portion of the basin, filling and spilling over the abyssal plain in the planform shape of a giant letter ‘‘H’’ (Figure 18.3). It has a diameter of around 350 km, a maximum thickness of about 3 km, and curvilinear crests between 4750 and 5950 m water depths, and a relief of some 1200 m above the adjacent sea floor (Ewing and Leonardi, 1971; Le Pichon et al., 1971a; Flood and Shor, 1988; Manley and Flood, 1993b; Von Lom-Keil et al., 2002). The Argyro Drift has been interpreted as a series of overlapping smaller drifts across the central part of the plain (Heezen and Tharp, 1977). The Ewing Drift (Figure 18.3d, e) extends east–southeast from the South American rise into the abyssal plain towards the Zapiola Drift. Its main crest-line is arc-shaped (Figure 18.3a), while a secondary crest lies eastward of, and runs parallel to, the primary crest, following the 4000–5000 m contour. This pathway is consistent with the general pathway predicted for a return flow of abyssal water (Flood and Shor, 1988). Large parts of these drift complex are covered by sediment waves, including some of the largest in the world (Von Lom-Keil et al., 2002). Average wave heights are 25–30 m (maximum 150 m) and wavelengths are 5–6 km (maximum 10 km) (Figure 18.3f, g). The wave field covers an area of around 1,000,000 km2, extending beyond the recognised drifts and over still flatter parts of the basin floor (Figure 18.3a). The drift has been developing over a very long time period, sufficient to generate a superimposed wave thickness of around 0.4–0.5 s two-way travel-time (twt), which implies a thickness of approximately 400 m. 18.3.1.3. Transkei CDC The Thanskei Basin is located in the south-westernmost Indian Ocean, and acts as a critical gateway between the Indian and Atlantic Oceans (Figure 18.4a). The deepest bottom currents flowing through this basin are the NADW and AABW

Figure 18.3 Argentine Basin. (a) Depositional and erosional contourite features on the Argentine Basin generated by the action of the WSDW (adapted from Flood and Shor, 1988; Speer et al., 1992; Reid, 1989; Klaus and Ledbetter, 1988). (b, c) Profiles across the Zapiola Drift and the Malvinas/Falkland Escarpment (data courtesy of the Argentine Navy Hydrographic Service, Buenos Aires, Argentina). (d) and (e) Sheeted deposits of the Ewing Drift (data courtesy of Argentine Navy Hydrographic Service, Buenos Aires, Argentina). (f ) Seismic profile perpendicular to the wave strike direction in the central part of the sediment-wave field on the western Zapiola Drift (adapted from Von Lom-Keil et al., 2002; with permission from Elsevier). (g) Profile across the Zapiola Drift crest. The crest appears to migrate to the southwest (adapted from Von Lom-Keil et al., 2002; with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

356 Abyssal Plain Contourites

Figure 18.4 Agulhas Drift in the Transkei Basin. (a) Transkei Basin location along the southeast African continental margin with main bottom circulation pattern of the AABW and the SW Indian Ocean sub-gyre. (b) Interpretation of a seismic line, where the Agulhas Drift position is located within the Transkei Basin. (c) Detailed interpretation and seismic profile of the Agulhas Drift. Pulses in the intensity of the bottomwater current probably caused alternating intervals of deposition by sediment draping and erosion (modified from Niemi et al., 2000; with permission from Elsevier).

F.J. Herna´ndez-Molina et al.

357

(Tucholke and Embley, 1984; Rogers, 1987). Three large elongated mounded drifts generated by the AABW have been identified: the Oribi, M, and Agulhas Drifts (Figure 18.4b). They are divided by three regional unconformities (O, M, & P) (Figure 18.4c) from the Early Oligocene to present (Dingle et al., 1987; Niemi et al., 2000; Uenzelmann-Neben, 2001, 2002). The Oribi and M Drifts are buried; the first was developed from Oligocene to Early Miocene, and the second from the Early/Middle Miocene to Early Pliocene. The Agulhas Drift is located on the present sea floor between 4000 and 4500 m water depth, and is elongated in a general east–west direction, between the Mozambique Ridge and the Agulhas Plateau (Figure 18.4a). It began to develop from the Early Pliocene on and is still active at present (Niemi et al., 2000; Uenzelmann-Neben, 2001, 2002). These three drifts are bounded at the bottom by basal discontinuities which represent regional erosional surfaces related to major changes in the palaeoceanographic regime and/or current velocity. They have a mounded external shape and asymmetric geometry, with erosion on the steep flank and continuous deposition onto the gentler side of the drift, which is typically partly covered with sediment waves (Figure 18.4b, c). The Agulhas Drift appears to be closely related with the nepheloid layer associated with the AABW. As the current flow veers southwards, the slower moving portion on the outside of the bend allows sediment to be deposited over the drift (Figure 18.4a). The total thickness of the drift above the basal unconformity is 0.5 s twt (450 m), although the drift itself rises approximately 190 m above the surrounding sea floor. It has a steeper southern face and gentlersloping northern face (Figure 18.4c). Erosion of the southern margin of the Agulhas Drift during episodes of intensified bottom-water flow is believed to have created this asymmetrical morphology. Repeated cycles of drape deposition and erosion are jointly responsible for the drift formation since the Pliocene. Internal highamplitude reflections with good continuity are inclined northwards. The depositional successions are also thinner at the southern side of the drift. The internal reflector geometry of the drift suggests that there have been episodic changes in bottom-current strength leading to alternating phases of deposition and erosion (Niemi et al., 2000; Uenzelmann-Neben, 2001, 2002). 18.3.1.4. Eastern New Zealand CDC In the southwest Pacific Ocean, the huge Eastern New Zealand Oceanic Sedimentary System (ENZOSS) depositional drift complex extends for 4500 km around the Eastern New Zealand margin (Figure 18.5a) (Carter et al., 1996, 2004). The deposits have been produced by the effect of the main inflow of deep water into the Pacific Ocean, the Pacific DWBC, flowing in part over the continental slope and rise and in part along the adjacent basin plains. This drift has been active since the Oligocene (Carter et al., 1996, 1999, 2004; Carter and McCave, 2002). The principal depositional elements include various types of large-scale drift, which are from SE to NW: the Campbell ‘‘skin’’ Drift, the North Bounty Drift, Chatman Terrace, Chatman Deep, and North Chatman Drifts, the Louisville Moat Drift, the Rekohu Drift, and also the huge Hikurangi Fan Drift in the Chatmam– Kermadec region (Figure 18.5b–d). The principal erosional features include moats, channels, and scours. As with the Weddell and Scotia Sea drifts, one single

358

Abyssal Plain Contourites

F.J. Herna´ndez-Molina et al.

359

water mass has produced different types of depositional and erosional elements depending on the local physiography. Mounded elongated drifts and sheeted drifts are the most common, whereas confined drifts are developed in locally restricted areas and mixed drifts occur where there has been significant interaction between downslope and along-slope processes (Carter and McCave, 1994, 2002; Carter et al., 2004).

18.3.2.

Northern hemisphere basins

18.3.2.1. Northern Atlantic CDSs In the northern Atlantic, several CDSs have been identified in a complex area that contains deep-water troughs, banks, and plateaus, which are associated with sedimentary basins and intervening structural highs (Fauge`res et al., 1993; Laberg et al., 2005; Stoker et al., 2005b). In the Rockall Trough and the Iceland, Eminger, and Labrador Basins, most of the large drifts are located along the continental slope (Feni, Hatton, Gardar, Bjorn, Snorri, Eirik, Gloria, Labrador, and Flemish Cap Drift), but they commonly reach onto and over the basin plains (Figure 18.6). All these drifts are generated under the influence of the NADW (Egloff and Johnson, 1975, 1978; Kolla et al., 1980; Fauge`res et al., 1993; Ben-Avraham et al., 1994; Howe et al., 1994; Wold, 1994). Over geological time, the deepening of the Icelandic gateways has controlled the formation of NADW, specifically the subsidence of the Greenland–Iceland– Scotland Ridge ( Jansen and Raymo, 1996; Wright and Miller, 1996). The waters flowing southwards from the Norwegian Sea through the gateway are divided in two branches. One branch flows over the Iceland–Faeroe Ridge and enters the Iceland Basin. The other branch flows over the WyvilleThompson Ridge, enters the Rockall Trough, and flows around the Rockall–Hatton Bank and then into the Iceland Basin (Figure 18.6a). These flows move along the sea floor where they entrain and transport sediment. When the current velocity decreases, the sediment is deposited parallel to the deep-sea topography forming the aforementioned drifts. These sedimentary drifts are hundreds of kilometres in length, relatively narrow, and they may have been covered by wave-like bedforms, which have been affected by important palaeoceanographic changes (Fauge`res et al., 1993; Wold, 1994; Shipboard Scientific Party, 1998; Laberg et al., 2005; Stoker et al., 2005b; Hassold et al., 2006).

18.3.3.

Equatorial abyssal plains

18.3.3.1. Western equatorial Atlantic CDC Some of the largest drifts described in the literature are recognised in the western north Atlantic basins (Figure 18.6a), including the Hatteras, Blake Bahama Outer

Figure 18.5 New Zealand contourite deposits. (a) Location with water-mass circulation (adapted from Carter et al., 2004). (b) Single-channel seismic profile of the Louisville confined drift with a detail of sediment waves shown by a high-resolution (3.5 kHz) profile. (c) Single-channel seismic profile of the Rekohu Drift. (d) Airgun single-channel seismic profile of the Rekohu Drift and the Hikurangi Fan Drift (modified from Carter and McCave, 2002; with permission from the Geological Society, London).

360

Abyssal Plain Contourites

Figure 18.6 Major contourite deposits in the North Atlantic. (a) Distribution of deposits showing the deep-water circulation (modified from McCave and Tucholke, 1986; Fauge' res et al., 1993; with permission from Elsevier). (b) High-resolution seismic profile of the abyssal contourite sheet of the Irminger Basin Gloria drift linked to bottom currents trapped on a basin sea floor (from Fauge' res et al., 1999 after Egloff and Johnson, 1975; with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

Ridge (BBOR), Northeast Bermuda Rise (BR), Caicos Outer Ridge, and Greater Antilles Outer Ridge (GAOR) (Hollister and Heezen, 1972; Fauge`res et al., 1993; McCave and Tucholke, 1986; Tucholke, 2002; Giosan et al., 2002; Jones and Okada, 2006). All these drifts are elongated and mounded drifts. They are generated by suspended

F.J. Herna´ndez-Molina et al.

361

sediment load carried by the Western Boundary Undercurrent (WBUC). Deposition generally occurs when WBUC (made up of both NADW and AABW) interacts with the sea floor. Most of these drifts are developed in part on the very low gradient continental rise and in part on the adjacent abyssal plains. Two well-known examples of CDSs are considered briefly: the BBOR and the GAOR (Figure 18.7). The BBOR is a detached, elongated, and mounded drift of around 103 105 km2 located between the rise and the abyssal plain ( Jones and Okada, 2006). The GAOR is a more isolated or separated, elongated, and mounded drift around 900 km long and up to 1000 m thick. It is located between the Silver and Nares Abyssal Plains, in water depths >5100 m, and partly covered by large fields of sediment waves (Figure 18.7b, c). The GAOR drift is composed of clay to fine silt-sized terrigenous sediments with less than 30% pelagic foraminiferal carbonate. Sediments exhibit some indistinct lamination, thin concentrations of ash and extensive reworking by burrowing organisms (Tucholke, 2002). These large drifts have been generated since the end of the Eocene, when the deep THC started in the North Atlantic; however, more rapid growth probably began in the early Miocene (Tucholke, 2002). For the BBOR, the rate of growth increased partly in response to interaction of the northward flowing Gulf Stream surface current with southward flowing deep WBUC off eastern North America (McCave and Tucholke, 1986). 18.3.3.2. Eastern equatorial Atlantic: Sierra Leone and Gambia Basins Deep circulation in the eastern equatorial Atlantic is dominated by the influence of NADW circulation and the underlying AABW (Warren, 1981; Friedricks and Hall, 1993). Lower NADW flows between 2900 m and 4400 m water depth and generates dunes and other current-controlled sedimentary features in the abyssal domain ( Jacobi and Hayes, 1992). This includes slope-parallel sediment drifts, erosional features, and sediment waves (wavelengths of 0.5–2 km and heights up to 100 m) along the southern transform margin of the Guinea Plateau (Rossi et al., 1992; Westall et al., 1993). In the Romanche Fracture Zone, ponded and mounded sediment bodies in the paths of AABW and NADW are also present (Westall et al., 1993). The AABW enters the eastern equatorial Atlantic through the Romanche and Vema Fracture Zones (Figure 18.8a) and circulates in two distinct counterclockwise flow patterns in the Sierra Leone and Gambia Basins (Wu¨st, 1933; Heezen et al., 1964a; Eittreim et al., 1983; McCartney et al., 1991). Deep water ( >4500 m) sedimentary drifts are formed by the influence of the AABW at the transition between continental rise and abyssal plain near the intersection of large-offset fracture zone with the African Margin (Figure 18.8a): Ivory Coast Rise and Guinea drifts (Jones and Okada, 2006). The Ivory Coast Rise drift, lying along the northern side of the St. Paul Transform at 3°N, is an abyssal sheeted drift. It is 600 km long with a uniform thickness of about 300 m over a wide area, but with a slight decrease in thickness near its margins (Figure 18.8b, c). Seismic facies are mostly of low amplitude to transparent, whereas surface and subsurface sediment-wave fields are extensive. The Guinea drift is a smaller elongated, mounded drift on the northern side of the Guinea Transform fault at 10°N. It shows marked changes in sediment thickness due to nonuniform deposition and erosion (Figure 18.8d, e).

362

Abyssal Plain Contourites

F.J. Herna´ndez-Molina et al.

363

Figure 18.7 Greater Antilles Outer Ridge (GAOR). (a) 3-D map constructed from bathymetric data from Sandwell and Smith (1997) with general location of the drift adjacent to the Puerto RicoTrench. (b) Sketch with the specific location of the Greater Antilles Outer Ridge (GAOR). (c) Seismic-reflection profile from the base of the Bahama Banks across the southern Silver Abyssal Plain and onto the GAOR. (d) Seismic-reflection profile from the base of the Bahama Banks across the Caicos Outer Ridge, Silver Abyssal Plain, and northwestern Greater Antilles Outer Ridge to Vema Gap (adapted and modified from Tucholke, 2002; with permission from the Geological Society, London): location in Figure 18.7b. (e) Model for current-controlled deposition along a section extending northeast from the Bahama Banks to the southernmost Bermuda rise (from Tucholke, 2002; with permission from the Geological Society, London). A multicolour version of this figure is on the enclosed CD-ROM.

Figure 18.8 Ivory Coast Rise and Guinea Drifts. (a) Bathymetry of the Sierra Leone Basin and surroundings. The Guinea Drift along the southern margin of the Gambia Basin is shaded grey. Locations of main canyons on the continental slope and deep-sea channels on the continental rise are shown. Mau = Mauritania Channel, Kay = Kayar Channel. (Legend: GP = Guinea Plateau. Inset map: CV = Cape Verde Islands; GaB = Gambia Basin; K = Kane Gap; SR= Sierra Leone Rise; SB= Sierra Leone Basin; ICR= Ivory Coast Rise; GuB = Guinea Basin; Mar = Mid-Atlantic Ridge; V = Vema Fracture Zone; R= Romanche Fracture Zone; SA = South America). (b, c) Seismic-reflection profiles across the Ivory Coast Rise. (d, e) Reflection profiles over the Guinea Drift along the southern margin of the Gambia Basin (modified from Jones and Okada, 2006; with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

364

Abyssal Plain Contourites

Figure 18.8

18.4.

(Continued)

OCEANIC G ATEWAYS

Many ocean basins and abyssal plains are connected by oceanic gateways (deep channels, valleys, or oceanic seaways). Opening, deepening, and evolution of these gateways are crucial elements in the sedimentary evolution of the basins because they allow the exchange of deep and intermediate water masses, sedimentary processes, and biota distribution. There are many well-known examples of such gateways in both Northern and Southern hemispheres (Le Pichon et al., 1971a, b; Hogg et al., 1982; Kennett, 1982; Fauge`res et al., 1993; Barker, 2001; Owen and Warren, 2001; Maldonado et al., 2003, 2005, 2006; Livermore et al., 2004; Pfuhl and McCave, 2005; Jones and Okada, 2006; Fauge`res and Stow, 2008; Zenk, 2008). Based on these examples, the following observations can be made. (1) In flowing through a constricted gateway, bottom-current velocity is locally intensified. (2) Erosional elements include scours, moats, contourite channels, terraces, and broad non-depositional surfaces. (3) Depositional elements include the development of irregular to elongate patch drifts, elongate mounded drifts, sheeted drifts, and contourite fan drifts at the downstream exit of the gateway. (4) A range of sea-floor bedforms has been

F.J. Herna´ndez-Molina et al.

365

recognised, including giant sediment waves, dunes, ripples, and linear features. (5) The variety of siliciclastic, biogenic, and manganiferous contourite facies are intercalated in time and space with those of down-slope processes (slides, debrites, turbidites), as well as with background pelagic and hemipelagic sediments.

18.5.

PRINCIPAL OCEANOGRAPHIC AND SEDIMENTARY PROCESSES

The continental rises and abyssal plains of ocean basins have been traditionally considered as relatively tranquil environments, the realm of quiet pelagic sedimentation episodically punctuated by down-slope turbidity-current input. In addition to these processes, low-energy bottom currents characterised by large tabular, slowmoving water masses produce sheeted depositional systems. However, such gateways and bathymetric constructions both continental rises and abyssal plains are occasionally far from tranquil oceanic domains. Where an otherwise low-energy flow interacts with abyssal hills, seamounts, submarine banks, or tectonic disruptions of the sea floor, a turbulent flow can be generated, leading to erosion and deposition even at a regional scale. The regional sedimentary regime is hence influenced by both basin size and kinetic energy associated with water-masses circulation. Based on the previous examples of ocean-basin and abyssal-plain deposits around the world, the following main factors may be considered as characteristic.

18.5.1.

Long-term hydrological conditions

Large depositional and erosional features are generated by relatively stable hydrological conditions leading to long-term bottom-water flows. In fact, the onset of many of the sedimentary drifts located in ocean basins started at the Eocene/Oligocene boundary and became re-activated with the establishment of the present global THC by the Middle Miocene (Kennett, 1982; Sykes et al., 1998; Niemi et al., 2000; UenzelmannNeben, 2001; Flood and Giosan, 2002; Pfuhl and McCave, 2005). There is a clear relationship between increased thickness of drift systems in ocean basins and slowed bottom-current velocity (see Figure 18.7d). This relationship is illustrated by the Argentine Abyssal Plain (Le Pichon et al., 1971a), GAOR (Tucholke, 2002), Sierra Leone ( Jones and Okada, 2006) and ENZOSSs (Carter et al., 2004).

18.5.2.

Nepheloid layer

The thickness of the nepheloid layer (see also McCave, 2008) is generally between 150 and 1500 m with an average concentration of suspended matter of 0.01–0.5 mg L 1. The highest concentration levels are located below the subtropical gyres at the western boundaries of ocean basins (Figure 18.9a, b). The residence time for particulate material in deep nepheloid layers is estimated at several days to weeks for the lowest 15 m above the sea floor, and weeks to months for the first 100 m above the sea floor (Kennett, 1982; Gao et al., 1998). The concentration of suspended matter in the nepheloid layer is generally low in abyssal plains and over the adjacent continental rise,

366

Abyssal Plain Contourites

Figure 18.9 Concentration in suspended matter and eddy kinetic energy. (a) Zones with high and highest eddy kinetics energy and relation with major concentration in suspended matter (nepheloid layer) in deep water (modified from Bearmon, 1989, and Pickering et al., 1989). Legend: 1 = high eddy kinetics energy; 2 = highest eddy kinetics energy; 3 = highest abyssal suspended load. (b, c) Horizontal and vertical distribution of suspended matter on the eastern margin of North America (modified from Eittreim and Ewing, 1972; with permission from Blackwell Publishing).

but in the areas where water masses are flowing with greater speeds or where distinct bottom-current intensity, the average concentration increases up to 10 times (Figure 18.9), and particles can travel distances of thousands of kilometres (Ewing et al., 1971; Kennett, 1982; Tucholke, 2002). Concentration is not only a function of the hydrodynamic energy but also of sedimentary input, erosion, transport capability, and CDC position. Particles are fed by the superficial water mass and by resuspension of particles from the sea floor.

F.J. Herna´ndez-Molina et al.

18.5.3.

367

Large eddies

In some regions, the formation and activity of large vertical eddies over the sea floor have generated large erosional features (Flood and Shor, 1988; Maldonado et al., 2003). This erosional action appears to be a significant mechanism both for the formation of nepheloid layers and for long-distance sediment transport. Richardson et al. (1993) the ‘‘abyssal eddies kinetic energy’’ (AEKE) as responsible for this process. Good examples of where these eddies are especially effective are the Argentine (Cheney et al., 1983; Arhan et al., 2002a, Figure 18.10) and Antarctica (Andrew Coward, NOCS, 2007, personal communication) abyssal plains, where the eddies are active over thousands of kilometres.

18.5.4.

Benthic storms

One deep-water process closely related to eddy effects and still not well known at present is that of benthic storms (also known as ‘‘deep-sea storms’’ or ‘‘abyssal storms’’). These represent the periodic intensification of normal bottom-currents’ flow, typically increasing the mean flow velocity by two to five times; they are associated with the formation of eddy within the water column, especially close to boundaries of strong surface currents. The HEBBLE Project first documented the occurrence of benthic storms events and demonstrated their important role in the winnowing, transport, and redistribution of bottom sediments (Hollister et al., 1980; Nowell and Hollister, 1985; Hollister, 1993). Once ripped up by the erosional effects of increased bottom shear, sediments may be transported by bottom currents and deposited in quiet regions downstream (Gardner and Sullivan, 1981; Kennett, 1982; Hollister and McCave, 1984; Flood and Shor, 1988; Bearmon, 1989; Von Lom-Keil et al., 2002). In some cases, the flow velocity is over 20 cm s 1, with a very high concentration of suspended matter (up to 5 g L 1) and a large erosional capability. Although benthic storms typically last for a few days, their effects in the suspension of bottom sediment, production of plankton blooms, and supply of considerable amounts of organic matter to the drifts can be much longer-lived (Richardson et al., 1993; Von Lom-Keil et al., 2002). In the North Atlantic, the frequency is 8–10 storms per year, with maximum velocities of 15–40 cm s 1 at 10–50 m above the sea floor, sediment concentrations of 3.5–10 g L 1 at 1–5 m above the sea floor, and a sediment flux of about 0.2–2 m3 per day (Hollister, 1993). Benthic storms are intermittent, but they have durations of days to weeks (2–20 days, usually 3–5 days). They have enough velocity to resuspend sediments, produce sedimentary structures at the sea floor, and rework sediment to a depth of around 0.5 m (Kennett, 1982; Bearmon, 1989). In addition, a tidal influence of several cm s 1 can be superimposed on benthic storms in some areas, which may even produce a reversal of the flow direction (Kennett, 1982; Stow et al., 1996b). Some authors have suggested that regions subjected to particularly intense benthic storms (for example, areas of strong climatic contrast between high and low latitude) may show significant erosion of the continental slopes and production of large submarine slides (Sheridan, 1981; Kennett, 1982; Pickering et al., 1989; Stow et al., 1996b; Gao et al., 1998; Einsele, 2000).

Figure 18.10 Evidence of eddies in the Argentine Basin. (a) Argentine Basin with the locations of four large eddies, and the position of the Sub-Antarctic Front (SAF, continuous arrow) and Subtropical Front (STF, discontinuous arrow) patterns. Bathymetric contours are multiples of 1000 m. Dotted lines represent the vertical sections position. Legend of the physiographic reference points: MEB= Maurice Ewing Bank; M/F = Malvinas/Falkland Channel. (b) Vertical water mass density distribution along the line ABC shown in Figure 18.10a. Shading visualises eddies E1 and E2 and the deep boundary currents near 48°S and 32°S.The SAF and STF locations are indicated. (c) Expanded vertical sections of potential temperature (), salinity (S), density (p), and dissolved oxygen (O2) along line CB (see Figure 18.10a) showing eddy E2. The bold dashed lines in the O2 diagram show the isopycnals  = 45.80, 45.87, and 45.98, which were chosen as the upper bound of Lower Circumpolar DeepWater (LCDW), lower bound of North Atlantic DeepWater (NADW), and lower bound of LCDW, respectively (modified from Arhan et al., 2002a; with permission from American Geophysical Union).

369

F.J. Herna´ndez-Molina et al.

18.5.5.

Effect of sea-floor obstacles

Slow-flowing water masses, as well as distinct bottom currents, will interact with topographic obstacles on the sea floor (seamounts, abyssal hills, mounds, banks, etc.). The effect of this interaction is many-fold and may increase the bottom-flow velocity by a factor of two (Kennett, 1982). Broadly speaking, obstacles are either point-shaped (isolated and sub-circular) or linear in geometry. 18.5.5.1. Point-shaped obstacles Point-shaped obstacles, such as isolated seamounts, generate seamount effects, which influence physical processes (Roden, 1987), marine biota (Rogers, 1994), and sedimentation and erosion rates (Davies and Laughton; 1972; Roberts et al., 1974), and may result in longer-term palaeoceanographic effects (Roden, 1987). The streamline distortion around a seamount has an obvious relevance to sediment distribution around obstacles (Figure 18.11), and two end-members can be considered (Herna´ndez-Molina et al., 2006b): (1) if vorticity processes (related to conservation of potential vorticity on the rotating Earth, see also Zenk, 2008) are predominant over advection processes (related to the large-scale horizontal transport of water masses), the impinging flow generates a pair of oppositely rotating eddies on the flanks of the seamounts, with accelerated flow to the left of the seamount and decelerated flow to the right (looking downstream on the northern hemisphere); (2) if the mean velocity of the flow decreases, the advection processes Anticyclonic circulation current Cyclonic rotation

Seamount effects Anticyclonic Cyclonic upwelling downwelling

Taylor columns

Northern hemisphere

Sedimentation (Patch drift)

Doming effects

Vortex compresion Southern hemisphere Gradually changing generations Fast core of sedimentary structures > Erosion and on the seamount < Sedimentation

Slow core < Erosion > Sedimentation

Scour

Marginal troughs

Scour

ow

Fl

Left side

Right side

Fast core Major scour

Slow core

Northern H. Slow core

Southern H.

Fast core Major scour

Figure 18.11 Main hydrodynamic features related to an incoming flow with a seamount, showing the formation of scours or marginal troughs (valleys) and sedimentary tails (modified from Herna¤ ndez-Molina et al., 2006b). Point-shaped obstacles (such as isolated seamounts) frequently have marginal troughs (scours or moats) around their bases. In the northern hemisphere, lateral asymmetry in the flow processes due to Coriolis effects leads to greater erosion on the right side of the obstacle and lesser erosion to the left. Large sediment tails are also commonly developed in the lee of large obstacles or mounds.

370

Abyssal Plain Contourites

can became dominant over vorticity interaction processes. In this case, the flow around the seamount is clockwise (in the northern hemisphere), and an anticyclonic eddy-column-like upwelling cone (Taylor column) appears near its top. Water-mass stratification, in general, inhibits vertical motion, and the vertical extent of eddies atop the seamounts is then limited. Frequently, seamounts (or diapiric mounds, volcanic mounds, etc.) have marginal troughs (scours or moats) around their bases (Figure 18.11). Lateral asymmetry in the flow processes due to Coriolis effects leads to more erosion on one side of the obstacle and less erosion on the other side, right and left on the northern and southern hemisphere, respectively (Roberts et al., 1974; McCave and Carter, 1997). Large sediment tails are also commonly developed in the lee of large obstacles (Figure 18.11), such as seamounts (Davies and Laughton, 1972; McCave and Carter, 1997; Herna´ndez-Molina et al., 2006b) or mounds (Masson et al., 2003). 18.5.5.2. Linear obstacles Linear obstacles, such as diapiric ridges, fracture zones, banks, and seamount chains, influence the behaviour of the impinging water mass in different ways. When the water-mass flow is either perpendicular or oblique to the obstacle, two helicoidal cores of bottom current are generated parallel to the linear feature, one immediately in front of the obstacle (which can lead to erosion of an incised contourite channel) and the other on the lee side (which can also lead to erosion of marginal valleys). Major linear obstacles can further lead to the generation of internal waves downcurrent of the obstacle, and hence to the development of associated depositional and erosional features (Kenyon and Belderson, 1973; Roden, 1987; 1991; Nelson et al., 1993; Merrifield et al., 2001; Garcı´a, 2002; Herna´ndez-Molina et al., 2003, 2006a; Serra, 2004).

18.6. 18.6.1.

MAIN CHARACTERISTICS OF ABYSSAL PLAIN CONTOURITES

Large-scale depositional and erosional features

One very common type of sedimentary drift in basinal systems is the sheeted drifts, which forms a sedimentary drape above the irregularities of the oceanic basement. It has a more or less constant thickness of up to a few hundreds of metres that extends over large areas (Figures 18.3, 18.7, 18.8 and 18.12a). It shows a slight decrease in thickness towards the margins and displays a very broad low-mounded geometry (Fauge`res et al., 1999; Stow et al., 2002c; Rebesco, 2005). Sheeted drifts are generated when the current is flowing as a tabular water mass. Nevertheless, even within the unconfined setting of an abyssal plain, the water masses are not always tabular, and the type of drift is determined by the interplay of water mass and the local physiographic features of the sea floor (Figure 18.12), usually related to local tectonic movement, oceanic fracture zones, or ridges of the igneous crust. Other types of drift, such as plastered sheets and giant elongated-mounded

Figure 18.12 Morphological expression of contourite drifts and channels in the Central Scotia Sea. (a) Central sector of the Scotia Sea, based on the interpretation of the swath bathymetry map, complemented with the analysis of seismic profiles (general location in Figure 18.2a). Main depositional and morphological features of the area are shown. Legend: North, Middle, and Southern Antarctic channels (NAC, MAC, and SAC) are the main channels associated with the Antarctic Circumpolar Current (ACC). A1 and A3 are elongated-mounded drifts related to canalised branches of the ACC; A2 is a field of large sediment waves. The Weddell Channel (WC) is the main channel of the Weddell Sea Deep Water (WSDW). The Western, Eastern, and NorthernWeddell Channels (WWC, EWC, and NWC) are the main channels of the WSDW. W1 is a slope-plastered drift attached to the discovery bank. W2 is made up of elongated-mounded drift and sediment waves related to the WSDW. W3 is a contourite fan formed by two channels. W4 is composed of mounded-sheeted drift and sediment waves related to the WSDW. Bottom-current directions are also shown based on the drift depositional patterns, channel orientation, and direction of migration of sediment waves. WAB is the boundary between the ACC and the WSDW flows (modified from Maldonado et al., 2003; with permission from Elsevier). (b, c) 3-D imagery from multi-beam bathymetry showing erosional and depositional features due to the circulation of both the WSDWand the ACC. (d) 3-D imagery from multi-beam bathymetry showing a mounded-sheeted drift bounded by sub-circular depression structure due the large eddies generated by the lateral interaction between the WSDW and the ACC (for location of the area see Figure 18.2a). A multicolour version of this figure is on the enclosed CD-ROM.

372

Abyssal Plain Contourites

drifts, are most commonly developed on the basin margins by the local/regional behaviour of the water mass when the flow is not strictly tabular (Figure 18.12b, c). In more restricted confined basin settings, confined drifts are developed with limited lateral migration, and down-current of oceanic gateways and contourite fan drifts can be generated (Figure 18.12a). In larger, but still partially confined abyssal plains, the partially trapped water-mass circulation can involve very complex gyres (Figure 18.3) that can generate huge giant elongate (bifurcated) drifts as in the cases of the Argentine (Flood and Shor, 1988), Irminger (Egloff and Johnson, 1975), Agulhas, and Mozambique (Kolla et al., 1980; Ben-Avraham et al., 1994) abyssal plains. The drift may also be buried in the depositional successions by gravity deposits, such as in the North Rockall Trough (Richards et al., 1987; Howe et al., 1994; Stoker, 2005a). Large erosional features are also generated in parts of ocean basins as the result of periods of intensified bottom-water activity, for example, related to eddies and benthic storms and where more restricted higher-energy bottom currents occur. Both linear and sub-circular features occur (Figure 18.12). Large linear erosional features, including linear depressions, scours alignments, and contourite channels, are commonly found in association with linear bathymetric highs such as seamount chains, plateaus, banks, ridges, etc. (Figures 18.12b, c; 18.13a, b). These erosional features are generated by the formation of a distinct bottom-current core due to the interaction between the impinging water mass and the obstacle. This is the case within, for example, the Shackleton Fracture Zone (Livermore et al., 2004), the Northern Weddel Abyssal Plain (Maldonado et al., 2005), and the Eastern and Western Malvinas/Falkland channel (Le Pichon et al., 1971a), GAOR (Hollister and Heezen, 1972; McCave and Tucholke, 1986; Fauge`res et al., 1993; Giosan et al., 2002; Tucholke, 2002; Jones and Okada, 2006). Large subcircular erosional features are related to eddy formation and evolution (Figure 18.12d). Although they are not yet well known, we suggest that they represent a common feature in ocean basins where strong dynamic and well stratified water masses exist (Arhan et al., 2002a). Taking into account all the possible aforementioned scenarios, the main characteristics of abyssal drifts can be summarised as follows: • the type and size of abyssal sediment drifts are directly related to the flow characteristics and the interplay of the flow with topographic relief on the basin floor and margins; • development of extensive drifts of great thickness results from long-term hydrologic conditions; • the drifts are built either over essentially flat-lying sediments or as a sediment drape above irregularities of the igneous basement; • sedimentary deposits may be condensed or locally absent, particularly above obstacles and spreading centres or in areas of strong bottom-current activity; • the most common sediment bedforms developed on the surface of abyssal drifts are giant fine-grained sediment waves (e.g. Figure 18.3f, g) (Ewing et al., 1971; Damuth, 1980; McCave and Tucholke, 1986; Flood and Shor, 1988; Klaus and Ledbetter, 1988; Von Lom-Keil et al., 2002).

0

W

E twtt (ms)

0 50

1

2

3

5 km

4

(a)

Contourite channel

100

(b)

Large sediment-wave field

Elongated-mounded drift

Contourite channel MAC

50

W

E

twtt (ms)

0

100 0

2

4 5 km

3

E

Elongated-mounded drift

0 50 100

(c)

0

Elongated-mounded drift

SW

1

2

3

4

5 km 0

NE

50

twtt (ms)

twtt (ms)

W

1

100

(d) S

1

2

3 0

0

2 km

50

4

5 km

N

twtt (ms)

0

100

(e) Figure 18.13 High-resolution seismic profiles of topographic parametric system (TOPAS) showing examples of erosional and depositional features located in the central sector of the Scotia Sea. For location of the profiles, see Figure 18.12. (a, b) Contourite channels. (c, d) Elongated-mounded drift. (e) Large sediment waves (modified from Maldonado et al., 2003; with permission from Elsevier).

374

18.6.2.

Abyssal Plain Contourites

Depositional units, discontinuities, and seismic facies

The depositional units in abyssal drifts do not exhibit significant migration compared with continental slope contourite drifts, because the deposits tend to drape and gently build over the whole area swept by the water mass. Consequently, the depositional units show a fairly regular thickness and an aggradational stacking pattern (Fauge`res et al., 1999). Migration will only occur when the current is shifted laterally, due to oceanographic changes or local topographic obstacles, and elongated-mounded drifts are developed as, for instance, in the case of Figure 18.13c, d and of the Agulhas Drifts (Figure 18.4) (Niemi et al., 2000; Uenzelmann-Neben, 2001, 2002). Seismic facies typically exhibit low-amplitude, discontinuous reflectors, or they are more or less transparent (Fauge`res et al., 1999; Stow et al., 2002c; Tucholke, 2002), but high-amplitude reflections with very good lateral continuation have also been described from locations where the water mass has higher velocities (Uenzelmann-Neben, 2001, 2002; Maldonado et al., 2003, 2006). In some cases, internal reflectors onlap and downlap distinct discontinuities and are truncated both at the sea floor and at the discontinuities. Major discontinuities in the sedimentary drifts located on abyssal plains were produced by palaeoceanographic changes. They record changes in seismofacies, which reflect changes in the depositional characteristics (Uenzelmann-Neben, 2001; Stow et al., 2002c; Maldonado et al., 2003, 2006). These discontinuities generally have basin-wide extension and, in some cases, may correlate from basin to basin, as has been recognised in basins from both the northern (Laberg et al., 2005; Stoker et al., 2005b) and the southern hemisphere basins (Van Andel, et al., 1977; Maldonado et al., 2003, 2005, 2006).

18.6.3.

Sedimentary facies

Contourite facies of abyssal drifts are mainly composed of very fine and finegrained, poorly-sorted sediments, as described by, among other, Barker et al. (1988), Stow et al. (1998a, 2002c), Tucholke (2002), Howe et al. (2004), and Stow and Fauge`res (2008). The following are the most commonly encountered facies and characteristics. • The siliciclastic contourite facies consist predominantly of clay and silt, rarely with primary structures preserved, generally masked by omnipresent and thorough bioturbation, and rare sandy contourites. Glacigenic abyssal contourites typically contain scattered coarser-grained material derived from floating ice melt-out, and isolated larger dropstones. Winnowing of this ice-rafted debris may locally produce irregular lenses and discontinuous coarse-grained layers although, as a rule, the bottom-current velocity is insufficient to produce significant winnowing. • Biogenic contourite facies are equally or more common than siliciclastic facies, including fine-grained and mud-rich types as well as foraminiferal and diatomaceous/radiolarian sands. They are also heavily bioturbated for the most part, although some of the biogenic sand facies show horizontal and crosslamination. Foraminiferal contourite dunes and ripples have been observed at

F.J. Herna´ndez-Molina et al.

375

the sea floor from some abyssal oceanic gateways, but cores recovered from there show dominant bioturbation and only rare cross-lamination. • Manganiferous contourite facies occur locally as thin irregular laminae, partial crusts, and micro-nodules within the dominant siliciclastic or biogenic contourites. They appear to represent basin conditions with bottom currents of particularly low-velocity, nevertheless important in maintaining the supply of dissolved metal cations. Ocean basins swept by bottom currents may exhibit widespread regions of larger ferromanganese nodules and hiatuses associated either with manganese encrustation or with calcareous hard-grounds. • A general absence of well-defined cyclicity is common for abyssal drifts. Where cycles occur, however, contourite sequences are quite subtle, and mid-cut-out sequences (e.g. C1-2-4-5) (Stow et al., 2002f) are the most evident. These typically show a climate-driven Milankovitch cyclicity.

18.7.

FINAL C ONSIDERATIONS

It is now beyond doubt that deep-water circulation controls depositional and erosional features at both short and long term. The nature of the bottom-water flow also influences abyssal fauna on an inter-basin scale (Howe et al., 2004). Long-time (millions of years) stability of interconnected ocean basins results in a relatively constant water-mass exchange, which controls: (1) the development of large depositional and erosional features, causing hiatuses up to several Ma duration simultaneously in several interconnected ocean basins; (2) the local winnowing of fine deposits; and (3) the geochemical precipitation of ferromanganese nodules (Van Andel et al., 1977; Kennett, 1982; Cronan, 2003). Current velocity is heavily affected by sea-bottom stress. Major morphologic features (basin margins, oceanic gateways, mid-basin obstacles) within ocean basins are therefore highly significant. When an impinging water mass interacts with the bottom relief, it is likely to develop a particular local and regional hydrodynamic signature (cores, branches, vortices, local turbulence, internal waves, helicoidal flows, vertical columns, etc.) that controls the dominant sedimentary processes (Figure 18.14). On the basis of the available information on abyssal drifts and their growth patterns, the following summary statements can be made (Figure 18.14): • The combined information provided by a range of data sets, including seismicreflection profiles, swath bathymetry, and sedimentological and oceanographic data, can be used to effectively characterise the nature and evolutionary patterns of sediment drifts in ocean basins and abyssal plains. • Water masses can flow as relatively slow-moving tabular bodies or as highervelocity, more confined bottom currents. This is of key significance in understanding current-influenced sedimentary processes and the development of abyssal drifts. The local and regional flow velocity is increased within ocean basins by interaction with local relief, along the basin margins, around mid-basin

376 Abyssal Plain Contourites

Figure 18.14 3-D conceptual sketch showing the main depositional and erosional features due to deep-water circulation on abyssal plains. A multicolour version of this figure is on the enclosed CD-ROM.

F.J. Herna´ndez-Molina et al.

•

•

•

•

377

obstacles (linear or point-shaped), and through narrow oceanic gateways, gaps, and other channels. The streamline distortion around obstacles has an obvious relevance for the sediment distribution. Nevertheless, only a few hydrodynamic studies of the abyssal plains have focused on erosion or sedimentation processes under the influence of bottom currents. Abyssal-drift systems clearly illustrate the combined influence of regional oceanographic conditions and local physiographic constrains of the basins in which they occur. It is possible to deduce, from their characteristics, the flow pathways and the nature of the water mass, which was responsible for their development. Abyssal plain contourites are therefore of great significance for palaeoceanographic reconstructions. The depositional and erosional features generated by the water-mass dynamics in ocean basins can be very extensive and long-lived, and some may be correlatable regionally at an inter-basin scale. In this sense, they may be more significant than some of those found along the continental slope, although because of lower sediment supply and generally weaker currents in abyssal systems, they may not develop such great thicknesses as found in some slope-centred elongatedmounded drifts. Sheeted drifts represent the most common sedimentary drifts in unconfined basins and abyssal plains. They are generated under the influence of tabular water masses. Giant elongated-mounded or plastered drifts are frequently developed due to the local flow disturbance by obstacles, especially along basin margins and over continental rises. Confined drifts are developed where the basins are bounded by high relief that confines the flow and inhibits lateral migration. Fan drifts tend to occur down-current of oceanic gateways. Giant elongate (bifurcated) drifts are huge deposits in abyssal plains constricted by marginal relief, which partially blocks the water-mass circulation and influences development of a complex gyratory circulation pattern. Sediment drifts in the deeper parts of ocean basins and on abyssal plains occur far away from any terrigenous sediment source so that the siliciclastic component is mostly very fine-grained, having travelled long distances in the deep-water nepheloid layer or as dispersed plumes and windblown dust across the sea surface. This sediment supply is augmented by a pelagic rain of biogenic material and, at high latitudes, by the glaciomarine supply of ice-rafted debris. In the absence of significant sediment input, manganiferous contourite facies may be developed. Fine-grained siliciclastic, biogenic, and manganiferous contourites are the principal facies, generally showing poorly-developed, mid-cut-out contourite sequences. In some cases, a direct link between abyssal-plain drift evolution and orbital forcing has been demonstrated.

Although much work has been carried out on contourite deposits on continental slopes and rises, some of the features observed in abyssal drifts are still not very well characterised and understood. Future research should therefore focus on these abyssal domains, particularly because their sedimentary record contains great potential for studies of palaeoclimatology, sediment dynamics, hydrocarbon exploration, and mineral resources.

378

Abyssal Plain Contourites

ACKNOWLEDGEMENTS We sincerely thank the following researchers who authorised and/or submitted their original figures and data for use in this chapter: John Jones (University College London, UK); L. Carter (Victoria University of Wellington, New Zealand); Michel Arhan (IFREMER/Brest, France); Tina M. Niemi (University of Missouri-Kansas City, USA); and Zvi Ben-Avraham (Tel-Aviv University, Israel). Also, we thank Fernando Bohoyo (Spanish Geological Survey, IGME, Spain), who produced the bathymetry for Figure 18.2a, d; Martı´n de Isasi (Argentine Navy Hydrographic Service, Argentina), who elaborated the bathymetry used in the Figures 18.3a and 18.10a, d; Marcelo Paterlini (Argentine Navy Hydrographic Service, Argentina), who sent us a copy of the original seismic record of the Argentine Abyssal Plain; and Juan Tomas Vazquez (Instituto Espan˜ol de Oceanografı´a, IEO, Spain), who realised the 3-D bathymetric sketch of Figure 18.7a. The Spanish Comisio´n Interministerial de Ciencia y Tecnologı´a (CYCIT) supported this research through Projects REN2001-2143/ANT & CGL200405646 (CONTOURIBER Project). This work is also part of three research grants funded by the Secretarı´a de Estado de Educacio´n y Universidades (Spanish Ministry of Education and Science), which enabled D.A.V. Stow to work at the Instituto Espan˜ol de Oceanografia, Malaga (Reference SAB2005-0182) and F. Javier Herna´ndez-Molina to work at the National Oceanography Centre, Southampton (NOCS, UK) (Reference PR2006-0275) and the Marine Geology and Geophysics Division of the Argentine Naval Hydrographical Oceanographic Service (Reference PR2007-0138). Finally, we welcome the revisions and suggestions of reviewers of an earlier version of the manuscript, including Dr Estefanı´a Llave (Spanish Geological Survey, IGME, Spain) and Dr Hans Nelson (IACT, CSIC-Univ. Granada, Spain). They have undoubtedly helped improve the final contribution. We also thanks both Editors and Project Manager for their interest and corrections that have helped us to improve the final version of our chapter.

C H A P T E R

1 9

C ONTINENTAL S LOPE C ONTOURITES F.J. Herna´ndez-Molina1, E. Llave2 and D.A.V. Stow3 1

Facultad de Ciencias del Mar, Universidad de Vigo, Vigo, Spain Instituto Geolo´gico y Minero de Espan˜a, Rı´os Rosas, Madrid, Spain 3 National Oceanography Centre, Southampton (NOCS), Southampton, UK 2

Contents 19.1. Introduction 19.2. Key Examples of Along-Slope Processes 19.2.1. Upper slope 19.2.2. Middle slope 19.2.3. Lower slope 19.3. Lower Slope to Continental Rise Transition and Other Kinds of Slopes 19.4. Principal Characteristics of Continental-Slope Contourites 19.4.1. Large-scale depositional features 19.4.2. Large-scale erosional features 19.4.3. Depositional units 19.4.4. Facies characteristics 19.5. Final Considerations Acknowledgements

19.1.

379 382 382 386 392 395 396 396 397 399 399 405 407

INTRODUCTION

The continental slope (hereafter referred to as ‘‘slope’’) is considered to be the steepest (3–6°) part of the continental margin (Figure 19.1), having a gradient of over 1:40 ( >1.5°). Lying seaward of the continental-shelf border (Heezen et al., 1959), it has a total length exceeding 1,10,000 km worldwide, and extremely variable widths (5–500 km) (Kennett, 1982). It extends down-slope from the continental-shelf edge (100–200 m water depth) to the continental rise on passive margins (down to 1500–4000 m), and continues to trench depths, typically 6000–10,000 m on some active margins (Bouma, 1979; Kennett, 1982; Pickering et al., 1989; Galloway, 1998). The physiography of the slope is commonly complicated by tectonic activity, which produces intra-slope basins, terraces and scarps (Pratson and Haxby, 1996), including in some cases slope terrace (Figure 19.1), continental borderland, or marginal/pelagic shelf, dividing the slope into two or three different domains (Shepard, 1963; Bouma, 1979). Traditionally, following changes in the slope gradients, three domains can be defined on slopes: upper, Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00219-7

Ó 2008 Elsevier B.V. All rights reserved.

379

380

Continental Slope Contourites

Shelf edge Slope basins Seamounts

Oce

ani

c cr

ust

n Co

tin

t en

al

m

ar

gi

n

Slope terrace, (marginal shelf or pelagic shelf)

Continental rise Submarine banks

Figure 19.1

Conceptual sketch of different continental slopes settings on continental margins.

middle and lower. Despite the fact that slopes represent less than 7% (60  106 km2) of the submerged surface of the Earth (Kennett, 1982), the total thickness of sediment beneath them can locally exceed 15 km, which underlines their importance in terms of sediment volume and sediment transfer to the deepest marine domain. A great variety of sedimentary processes can be found on slopes and grouped into three generic types: lateral processes (down-slope, up-slope and along-slope), vertical processes (pelagic and hemipelagic settling) and authigenic processes (including both deep-water bioherm growth and chemical precipitation) (Pickering et al., 1989; Stow, 1994; Einsele, 2000). An interplay amongst all of these sedimentary processes is commonly found on slopes, generating mixed depositional systems. However, when one of these processes (mass wasting, turbidity current, pelagic settling or contour current) is dominant, a specific depositional system can develop. In some cases, the transition between these systems is gradational, but in other cases it presents a more clear-cut lateral change – for instance, a turbidity current interacting with a strong contour current (Mulder et al., 2006) – or vertical change, due to the dominance of mass movement versus contourite processes at different geological time spans – for instance, Quaternary glacial/interglacial cycles (Øvrebø et al., 2006). This complex scenario explains why, even today, slope depositional models of slope systems remain rare (Pickering et al., 1989; Stow et al., 1996b; Shanmugam, 2000), and why – in many cases – the relationships between slope and adjacent systems, whether shallower or deeper, remain undetermined.

F.J. Herna´ndez-Molina et al.

381

The impact of contour-following currents is particularly marked on slopes, since its nature and gradient act as a major control on thermohaline circulation. A contour current acting for a long time over a slope could be strong enough to profoundly affect sedimentation, from winnowing of fine deposits to developing large-scale depositional or erosional features (Kennett, 1982; Stow, 1994; Masson, 2001; Herna´ndez-Molina et al., 2003, 2006c). Contour currents can be a relatively homogeneous water mass flowing along the slope, or comprise several water masses flowing at different depths, and sometimes in opposite directions (Viana et al., 2002a, b; Laberg et al., 2005). The deposits generated by such currents, on any part of the slope, are known as ‘‘contourites’’ or ‘‘contourite drifts’’ (Stow et al., 1998a, 2002c; Rebesco, 2005). There are several recent classifications of them; here we will follow the latest, proposed by Rebesco (2005) (Figure 19.2), based on the previous classifications of McCave and Tucholke (1986), Fauge`res et al. (1993, 1999), Rebesco and Stow (2001) and Stow et al. (2002c). Where current strengths are generally stronger, a variety of erosional

Figure 19.2 Main depositional (drifts) features and inferred bottom-current paths (adapted from Rebesco, 2005; Rebesco and Stow, 2001, with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

382

Continental Slope Contourites

features are developed (Figure 19.2). Extensive erosion or non-deposition leads to the development of widespread hiatuses in the depositional record. Although several authors have studied erosional features in contourites (e.g. Nelson et al., 1993, 1999; Evans et al., 1998; Stow and Mayall, 2000; Masson, 2001; Garcı´a, 2002; Herna´ndez-Molina et al., 2003, 2006c), they have yet to be classified, and integrated into a classification system including standard drift categories. Knowledge of such processes is important in palaeoceanography, palaeoclimatology, slope-stability studies, and mineral and energy-resource exploration. This is why, over the last decade, investigation into slope-contourite processes has been one of the most active lines of marine sedimentological and stratigraphical research (e.g. Gao et al., 1998; Stow and Mayall, 2000; Stow et al., 2002f; Rebesco, 2005; Viana and Rebesco, 2007). In the near future, new advances in the field of slope-contourite studies are anticipated from the application of new technologies including state-of-the-art geophysical techniques, submersibles, Remotely operated vehicles (ROVs), sediment coring, ocean drilling (Integrated Ocean Drilling Program, IODP) and accompanying land investigations. The rapid proliferation of seismic geomorphology (Posamentier and Kolla, 2003; Posamentier, 2004), 3-D seismic technology (Davies et al., 2004; Cartwright and Huuse, 2005), and an increase in deep-water exploration by oil companies (Viana et al., 2007) will undoubtedly lead to spectacular findings about morphology, stacking patterns, and facies of contourite drifts on continental slopes. The present chapter aims to summarise key case studies in different slope settings, and then to discuss general and specific characteristics of different types of slope contourites.

19.2.

KEY EXAMPLES OF A LONG-SLOPE P ROCESSES

Selected examples of depositional and erosional features on upper, middle and lower slope-contourite depositional systems (CDS) are presented below.

19.2.1.

Upper slope

19.2.1.1. Brazilian Margin A shallow-to-deep-water CDS has been generated on the Brazilian Margin (Figure 19.3a) since at least the early Neogene (Viana et al., 1998a, b, 2002a, b; Viana, 2001). Confinement of the western boundary core of the Brazil Current by the shelf edge, has allowed bottom currents to reach speeds of up to 120 cm s–1, effectively carving the shelf-edge escarpment, and leading to the development of erosional terraces, furrows, sea-floor erosion and gravel lags below the core of the current (Figure 19.3b). Laterally, in current-deceleration zones (with velocities 60 cm s 1), southward-migrating 2-D and 3-D sand waves are produced. Along this part of the slope, a sand-rich sheeted drift has been developed, due to the interaction between shelf spillover processes and the Brazil Current. This drift comprises a coarse- to medium-grained sand tongue stretching parallel to the upper

F.J. Herna´ndez-Molina et al.

383

Figure 19.3 Depositional and erosional contourite features on the Brazilian margin. (a) Location sketch with main water-mass circulation along the margin. (b) Down-slope 3.5 kHz records showing contourite erosional features in the current funnelling zone.The area between the shelf edge and the erosional ridges is dominated by conglomerates to coarsegrained sand and by erosional/transport processes. The erosional terrace results from the activity of the Brazil Current (BC).

slope, several kilometres long, a few kilometres wide, and 10–50 m thick, becoming finer and thinner down-slope. Towards the north (i.e. up-current) (Figure 19.3a), there are elongated plastered drifts with a maximum thickness of about 350 m, and 7–20 km long and 3–15 km wide, typically containing sediment waves down-slope (Figure 19.3c and d). A general up-slope migration of this CDS can also be observed, accompanying the continuous retrogradation of the shelf edge during the late Quaternary, marked by the landward shift of the shelf-edge escarpment. 19.2.1.2. Northern European Margin In Northern Europe, there are many good examples of the effects on the upper slope of strong contour currents flowing towards the northeast (Figure 19.4a). Examples include the eastern slope of the northeast Rockall Trough (Masson et al., 2002), and the eastern slope of the Faeroe–Shetland Channel (Masson, 2001). Elongated drifts are common

384

Continental Slope Contourites

Figure 19.3 (Continued) (c) Multi-channel seismic line recording an elongated plastered drift more than 300 ms thick. Alternate phases of erosion and deposition are reflected; sediment waves are present inside the drift. (d) Multi-channel seismic line recording up-slope prograding sediment waves in an open middle-slope setting. Arrows indicate the migration trend of the deposit (modified fromViana, 2001;Viana et al., 2002b, with permission from The Geological Society, London).

depositional features with varying moat development on their up-slope flanks (Figure 19.4b–d). A thin contourite sand sheet has been described, formed beneath strong (>30 cm s 1) bottom currents (Figure 19.4d). Its composition ranges from a relatively coarse-grained layer (<25 cm) of well-sorted or very well-sorted muddy sand to sandy gravel with a high backscatter signature. The surface of this sheet deposit is ornamented with ripples. Gravel lag deposits and mobile sand bedforms are produced in those parts of the upper slope affected by strong currents (>75 cm s 1) (Figure 19.4e). The general morphology of these areas took shape mainly during the last glacial stage, when the high sediment input resulted in the deposition of glacigenic debris. Nevertheless, later Holocene modification of the late glacial sea floor by strong bottom currents has produced the remarkable depositional and erosional features observed. 19.2.1.3. Pennell coast, Antarctica A wide sheeted drift has been recently reported on the outer shelf and upper slope, at a depth of 200–1200 m offshore of the Pennell Coast, Antarctica (Rodrı´guez and Anderson, 2004). This sand tongue (mainly composed of volcanic grains) was

Figure 19.4 Depositional and erosional contourite features on the Northern European upper slope. (a) Location sketch with main water-mass circulation along the margin. Surface and deep-water circulation patterns are shown by filled and open arrows, respectively. Grey-shaded arrow illustrates inferred path of intermediate-depth water deflected to the west by the Wyville-Thomson Ridge. Contour interval 200 m. (b) and (c) Seismic profiles crossing an elongate drift on the Hebrides Slope. Note the sediment waves on the down-slope flank of the elongated drift, which are composed of three distinct seismic sequences (adapted from Howe et al., 1994; Masson et al., 2002; with permission from Elsevier). (d) 3.5 kHz profile crossing an elongate drift down-slope from the Geikie escarpment (adapted from Masson et al., 2002; with permission from Elsevier). (e) TOBI images of bedforms on the eastern slope of the Faeroe ^Shetland Channel (from Masson, 2001; with permission from Elsevier). High backscatter = light tones. On the top, barchan-like dunes in 350 m water depth. Orientation of dunes indicates formation under a NEdirected current. On the bottom, furrows are present with a NE trend, parallel to the contours (waterdepth 600 m) (for location of the profiles, see Figure19.4a).

386

Continental Slope Contourites

produced by the westward-flowing circumpolar boundary current during at least the last 9000 years. It represents an important example of a contour current’s capacity to generate thick, extensive slope sand bodies in a very short time span.

19.2.2.

Middle slope

19.2.2.1. Southwestern Alboran Sea The Ceuta CDS is located in the southwestern Alboran Sea (SW Mediterranean), close to the Strait of Gibraltar (Figure 19.5a), running parallel to the Moroccan middle slope at a water depth of 200–700 m (Ercilla et al., 2002). This CDS is mainly characterised by an elongated plastered drift deposited over a slope terrace, with a broad lenticular geometry. It is around 100 km long, 28 km wide, with a relief of up to 400 m above the surrounding sea floor, and a seismic sediment thickness up to 700 ms two-way travel time. It is formed under a westward simple tabular deep-water mass, producing deposition on the middle slope, and erosion at the toe of the slope (Figure 19.5b). The Ceuta Drift is composed of Pliocene and Quaternary deposits over the basal surface. The Quaternary sedimentary record has been divided into five prograding seaward units, with stratified facies that produce reflections with high lateral continuity and amplitude, converging both seaward and landward (Figure 19.5b). The drift is composed mainly of muds, intercalated with 40–50-cm-thick layers of sandy muds bounded by sharp surfaces and thinner (10 cm thick) silty clay layers with gradual contacts (Figure 19.5c). The thickness and vertical distribution of these sediment types suggests an evolutionary model where sandy muds were deposited during cold (sealevel lowstand) intervals and silty clays during warm (sea-level highstand) intervals. 19.2.2.2. Gulf of Cadiz A very large CDS was generated during the Pliocene and Quaternary by the action of Mediterranean Outflow Water (MOW) on the middle slope of the Gulf of Cadiz (Figure 19.6a, b), extending around the west Iberian Margin (e.g. Kenyon and Belderson, 1973; Gonthier et al., 1984; Nelson et al., 1993, 1999; Llave et al., 2001, 2006; Stow et al., 2002b; Alves et al., 2003; Habgood et al., 2003; Herna´ndez-Molina et al., 2003, 2006c; Mulder et al., 2003b, 2006). This CDS comprises both large depositional and erosional features (Figure 19.6b–d), conditioned by a strong current with speeds reaching nearly 300 cm s 1 close to the Strait of Gibraltar, slowing to 80 cm s 1 at Cape St Vincent (Kenyon and Belderson, 1973; Ambar and Howe, 1979; Cherubin et al., 2000). The main depositional features are sedimentary wave fields, sedimentary lobes, mixed drifts, plastered drifts, elongated mounded and separated drifts and sheeted drifts. The main erosional features are contourite channels, furrows, marginal valleys and moats. All of them have a specific location along the margin, and their distribution defines five morphosedimentary sectors within the CDS; details are found in Herna´ndez-Molina et al. (2003, 2006c) and Llave et al. (2007). The development of each of these five sectors at any time is related to a systematic deceleration of the MOW as it flows westwards from the Strait, due to its interaction with margin bathymetry, and to the effects of Coriolis force.

F.J. Herna´ndez-Molina et al.

387

Figure 19.5 Depositional and erosional contourite features on the middle slope of the southwestern Alboran Sea (SW Mediterranean). (a) Location sketch with main water-mass circulation along the margin. (b) Air-gun seismic profile oblique to the Ceuta plastered drift across its central sector.The drift has a positive relief feature. Internally, it is composed of up to five seismic units (Ercilla et al., 2002; with permission from The Geological Society, London). (c) Some core section examples from the proximal (TG-5) and distal (TG-6) part of the slope terrace, and the lower slope (TG-12). The three textural types (muds, sandy muds and silty clays) are shown (adapted from Ercilla et al., 2002; with permission from The Geological Society, London). For profiles and core location, see (a).

388

Continental Slope Contourites

Figure 19.6 Depositional and erosional contourite features on the middle slope of the Gulf of Cadiz. (a) Location sketch with main water-mass circulation along the margin (modified from Herna¤ ndez-Molina et al., 2003, 2006c; with permission from the Geological Society of America and Elsevier). (b) Uninterpreted multi-channel seismic (MCS) reflection profile across the Faro ^Albufeira elongated mounded and separated drift and the Alvarez Cabral moat on the middle slope (data courtesy of TGS-NOPEC Geophysical Company ASA). (c) Sparker seismic profiles indicating the abraded surface. (d) Sparker seismic profiles showing examples of contourite channels and marginal valleys. (e) Sketch of the MPC-1 borehole from the contourite depositional system on the middle slope of the Gulf of Cadiz (proximal scour and ribbons sector of Herna¤ ndez-Molina et al., 2003, 2006c) with an uninterpreted MCS reflection profile across the middle slope (line S- 81A, provided by REPSOL-YPF Oil Company for the present work). For location of the profiles, see (a). A multicolour version of this figure is on the enclosed CD-ROM.

F.J. Herna´ndez-Molina et al.

Figure 19.6

389

(Continued)

The stratigraphic architecture of these different drifts and their relationship with the major structural features in the area has been described in detail by Llave et al. (2001, 2006, 2007). In general, the drifts show laterally extensive, progradational to aggradational seismic depositional units, with low- to high-amplitude sub-parallel reflectors and widespread discontinuities (Figure 19.6b). The elongated and separated Faro–Albufeira drift (located within Sector 4, in the central and northwest areas where sedimentary processes are dominant) represents a classic example of middle-slope contourite deposits (Figure 19.6b) with a well-layered internal acoustic structure and internal reflection configuration that onlap and downlap up-slope. It is mainly composed of muddy, silty and sandy sediments, with a mixed terrigenous (the dominant component) and biogenic composition (Gonthier et al., 1984; Stow et al., 1986, 2002e). By contrast, within the large contourite channels (located in central Sector 3), sand and gravel are found (Nelson et al., 1993, 1999) as well as many erosional features (Garcı´a, 2002; Herna´ndez-Molina et al., 2006c). In the proximal sector (Sector 1) close to the Strait of Gibraltar, an exceptionally thick sandy sheeted drift (815 m thick) is located (Figure 19.6e), with sand layers that average 12–15 m in thickness (minimum 1.5 m, maximum 40 m) (Buitrago et al., 2001; Llave et al., 2007).

390

Continental Slope Contourites

This large, complex CDS started to develop after the opening of the Strait of Gibraltar, which made MOW circulation possible, since the end of the Miocene (Nelson et al., 1993, 1999). Later evolution of the system has been controlled by Pliocene and Quaternary environmental changes (climatic and eustatic), MOW palaeoceanographic changes, and changes in the morphology of the margin (Llave et al., 2001, 2006, 2007). On a smaller timescale, a grain-size cyclicity related to cyclic changes in the MOW strength has been identified, suggesting MOW intensification during cold intervals in the Gulf of Cadiz (Llave et al., 2006; Voelker et al., 2006) and along the western Iberian Margin (Scho¨nfeld and Zahn, 2000; Shackleton et al., 2000; Scho¨nfeld et al., 2003). Furthermore, post-Miocene tectonic activity has also played an important part in the morphological development of the sea floor, controlling multiple current pathways for the MOW at each evolutionary stage (Llave et al., 2008). 19.2.2.3. Northern Europe From the southwest of Ireland (Porcupine Bank) to offshore central Norway (Lofoten Islands), many processes have occurred due to persistent, vigorous bottom currents (Laberg et al., 2005). These currents have generated a wide variety of drift deposits, including (Figure 19.7a) the Rockall–Porcupine region (McDonnell and Shannon, 2001; Van Rooij et al., 2003; Øvrebø et al., 2006), the northern Rockall Trough (Stoker et al., 1993, 1998a, 2001; Howe et al., 1994, 2001, 2002; Masson et al., 2002), the Hebrides Slope (Howe et al., 2002; Knutz et al., 2002a; Masson et al., 2002), Faeroe–Shetland region (Van Weering et al., 1998; Kuijpers et al., 2001; Bulat and Long, 2001; Stoker et al., 2003; Smallwood, 2004) and the North Sea Fan–Vøring region (Laberg et al., 1999, 2001, 2002; Laberg and Vorren, 2004). Most of these drifts present classic elongated, mounded geometries (Figure 19.7b–d), with well-layered internal acoustic signatures and marked onlap and downlap up-slope. They are composed essentially of sandy and muddy contourites, but the facies indicates that deposition was dominated by a combination of bottom currents, ice-rafting and hemipelagic settling. A number of authors have cited a widely accepted model, linked to Quaternary climatic changes, for these northern European slopes (e.g. Howe, 1995; Stoker, 1998a; Weaver et al., 2000; Gro¨ger et al., 2003; Øvrebø et al., 2006). During the warm interglacials (sea-level highstands), the mixed carbonate/siliciclastic deposits were affected by strong currents, leading to active bedform growth and CDS development. Variability of sediment input was the main factor affecting sedimentation during cold glacials (sea-level lowstands). Where input was high, such as when ice sheets reached the shelf edge, down-slope sediment transport dominated (dominant siliciclastic deposits), overwhelming any bottom currents that were active. Nevertheless, active glacial bottom currents have been dated to the last glacial, although these were less energetic than those during highstands. This activity could have been due to short intervals with lower sediment input (Masson et al., 2002), or to minor interstadial climatic warming (Knutz et al., 2001, 2002a, b; Laberg and Vorren, 2004).

Figure 19.7 Depositional and erosional contourite features on the Northern European middle slope. (a) Location sketch with main water-mass circulation along the margin. (b) and (c) Sparker seismic profiles on the Porcupine Bank indicating the seismic features and unit geometry on the eastern slope of the Porcupine Seabight (Van Rooij et al., 2003; with permission from Elsevier). (d) Sparker profile (UiT 97-227) across the Lofoten Drift illustrating the layered, continuous, parallel or slightly divergent internal seismic signature of the youngest seismic units. Below, a segment of a multi-channel seismic profile (VB-32- 89) across the same drift. The base drift reflection corresponds to the intra-Miocene reflection (Laberg et al., 2002, 2005; Laberg and Vorren, 2004; with permission from Elsevier). For profile locations, see (a).

392

Continental Slope Contourites

19.2.2.4. Eastern Weddell Sea An exceptional example of a sandy plastered drift on the middle slope is located at approx. 52°W in the Weddell Sea (Gilbert et al., 1998; Pudsey, 2002) (Figure 19.8). It is mainly composed of bioturbated, very poorly sorted, fine-grained terrigenous sediments. This contourite drift represents a relatively young feature (seismic thickness 1 s two-way travel time), which could have been initiated at the onset of bottom-water flow in the Early Miocene or, alternatively, at the start of high sediment supply to the western Weddell Sea in the latest Miocene. 19.2.2.5. New Zealand On the eastern margin of South Island, New Zealand, in the Canterbury Basin, there is a narrow slope between the shelf and the Bounty Trough. It is influenced at present by northward flows of the Southland Front, and a local gyre associated with the Sub-Antarctic Front (SAF) (Figure 19.9a). Bottom currents have strongly influenced the deposition there, locally modifying the sequence architecture and leading to the development of a large CDS on the middle to upper slope during the Late Miocene–Pliocene (Fulthorpe and Carter, 1989; Lu et al., 2003). The contourite drifts are elongated mounded and separated, up to 1000 m thick, >50 km long and 20 km wide (Figure 19.9b). The internal architecture of the drifts defines two end members. One end member is formed by simple drifts that have a welldefined moat, subparallel to the palaeoshelf, and a large mound drift basinward of the moat. Their internal seismic facies (base, core and crest) reflect increasing confinement and intensification of the current, initially due to Coriolis deflection and later by physical confinement within the moat as the adjacent drift aggraded. The other end member is constituted by complex drifts, may be multi-crested or multi-stage, and formed in response to rapid lateral shifts in position of the moat, within a regime that involved multiple flow pathways (Lu et al., 2003) (Figure 19.9b). Sediment waves are also locally preserved at the base of elongated drifts on their basinward flank (Figure 19.9c).

19.2.3.

Lower slope

19.2.3.1. Eirik Drift The Eirik CDS lies on the lower slope and rise off the southern tip of the Greenland Margin to the south of Cape Farewell at depths of 1500–3500 m (Figure 19.10a). It is an elongated, mounded drift (Figure 19.10b), approx. 800 km long, resulting from two principal factors (Chough and Hesse, 1985; Arthur et al., 1989; Hunter et al., 2007b): (1) the influence of the North Atlantic Deep Western Boundary Current and (2) high sediment input. Local variations in the slope angle have caused funnelling of the current and the generation of three secondary crests along the drift. Four seismic units have been identified (Arthur et al., 1989), but the drift’s growth stage was mainly during the two younger sequences (Figure 19.10b). High- to moderateamplitude reflectors, with good lateral continuity, characterise these two sequences, with a clear cyclicity in reflector amplitude within the youngest one. This drift is mainly composed of strongly bioturbated silty clays and clayey silts, containing

F.J. Herna´ndez-Molina et al.

393

Figure 19.8 Large depositional feature (plastered drift) on the continental slope of the western Weddell Sea. (a) Location sketch with main water-mass circulation along the margin. (b) BAS multi-channel seismic Line AMG845 -14 showing an example of plastered drift on the continental slope. A large sedimentary wave field down-slope is presented within the lower slope.The chaotic seismic facies below the drift are interpreted as slumped masses. Beneath this body, SE-dipping parallel-bedded sediments are probably passive margin deposits that accumulated prior to expansion of grounded ice onto the continental shelf (modified from Pudsey, 2002; with permission from The Geological Society, London).

394

Continental Slope Contourites

Figure 19.9 Depositional and erosional contourite features on the eastern margin of the South Island in the Canterbury Basin (South Island of New Zealand). (a) Location sketch with main water-mass circulation along the margin. Locations of the modern Southland Front, Sub-Antarctic Front (SAF), and a local gyre associated with SAF are shown (modified from Lu et al., 2003). (b) Dip profile (EW00 -01-12) showing the main features of a large mounded and elongated drift developed on the lower to middle slope during the Late Miocene-Pliocene (Craig S. Fulthorpe University of Texas, USA, Personal Communication, 2007). (c) Dip profile (EW00 - 01-16) showing sediment waves (shaded area) preserved basinward at the base of elongate drifts. Wavelength is about 2 km and amplitude averages 50 m (modified from Lu et al., 2003; with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

variable proportions of biogenic material. The main phase of drift construction started after 4.5 Ma, due to the marked increase in thermohaline circulation (Hunter et al., 2007b). Pleistocene cycles reflect glacial/interglacial alternations. 19.2.3.2. Weddell Sea, Antarctica Many contourite sedimentary processes are active in the Weddell Sea, due to the interaction of the Weddell Sea Bottom Water with the lower slope (Gilbert et al., 1998; Pudsey, 2002; Howe et al., 2004). On the western slope,

F.J. Herna´ndez-Molina et al.

395

Figure 19.10 Depositional and erosional contourite features on the Eirik Drift (Greenland Margin). (a) Location sketch with main water-mass circulation along the margin. (b) Singlechannel seismic profile crossing a elongated mounded drift on the lower slope (Hunter et al., 2007b; with permission fromThe Geological Society, London).

muddy/silty contourite deposits are found mixed with glacigenic hemipelagites, rich in ice-rafted debris. A cyclicity between winnowed, better-sorted deposits, interpreted as representing warmer climatic conditions and stronger current activity, and muddominated, poorly sorted sediment interpreted as representing colder intervals with lower bottom-current strength, has been attributed to climatic fluctuations of shorter duration than the main 100 ka glacial/interglacial cycle, although the definitive chronology has yet to be established.

19.3.

LOWER SLOPE TO C ONTINENTAL R ISE T RANSITION AND O THER K INDS OF S LOPES

The transition between the toe of the lower slope and the continental rise is a common setting for CDS. The separated and detached drifts, sheeted drifts and mixed drifts are usually found in these settings (Fauge`res et al., 1999; Stow et al., 2002f; Rebesco, 2005). Good examples of the separated drifts are on the Hebrides Slope (Howe et al., 1994; Masson et al., 2002), of detached drifts on the Blake–Bahama

396

Continental Slope Contourites

(McCave and Tucholke, 1986; Stoker, 1998a) and of sheeted drifts in the Argentine Basin (Flood and Shor, 1988; Klaus and Ledbetter, 1988; Flood et al., 1993). In these settings, contourite processes are often associated with turbidite processes, generating mixed drifts (Fauge`res et al., 1999) as in the case of the large sedimentary mounds on the Antarctic Peninsula Pacific Margin (Rebesco et al., 1996, 1997, 2002, 2007; Pudsey and Camerlenghi, 1998; Pudsey, 2000; Lucchi et al., 2002), or the eastern margin of the USA off Cape Hatteras (e.g. Tucholke and Mountain, 1986; Locker and Laine, 1992; Benetti, 2006). Other kinds of slopes in marine environments can be found around seamounts, large submarine ridges and submarine banks (Figure 19.1). In some cases, large ridges and banks are located within the continental margins, but in other cases, they constitute isolated submarine structures. They commonly interact with bottom currents, as in the case of (1) most of the North Atlantic Banks, including Rockall, Hatton, Snorri and Rosemary (Stoker et al., 1998a; Fauge`res et al., 1999; Laberg et al., 2005); (2) the Galicia Bank (Ercilla et al., 2008), off the northwest Iberian coast; (3) the Guadalquivir Bank (Herna´ndez Molina et al., 2003; Llave et al., 2007), off the southern Iberian coast; (4) the Elba Ridge (Roveri, 2002), off western Italy and (5) around the slope of the Campbell Plateau, eastern New Zealand (Carter et al., 1998; Carter and McCave, 2002). In these slope settings, usually plastered drifts and erosional contourite features are frequent on the middle to lower slope, as well as its transition to the abyssal plain, where drift–moat complexes can be formed (Stoker, 1998b; Howe et al., 2006).

19.4.

P RINCIPAL C HARACTERISTICS OF C ONTINENTAL -SLOPE C ONTOURITES

Based on the findings from marine geological fieldwork (e.g. the previous examples and many others described in the literature), many proxies can be used to recognise depositional and erosional features on slopes, although essential research has yet to be carried out at diverse scales, integrating all of the results collected with different methods (e.g. acoustic, seismic, cores and submarine photography: see Howe, 2008). Moreover, in the fossil sedimentary record (Hu¨neke and Stow, 2008), observations are in many cases limited to the scale of the outcrop, with facies analysis being the main tool for slope-contourite identification; however, at this scale, contourite facies can be easily misinterpreted as turbidite or pelagic/hemipelagic facies.

19.4.1.

Large-scale depositional features

Slope drifts can develop on any part of the slope, and their size is a function of such factors as sediment supply, the velocity of the contour current, and slope morphology (Fauge`res and Stow, 2008). Drifts can be anything up to hundreds of kilometres long, tens of kilometres wide and several hundreds of metres high, relative to the surrounding sea floor. They can be generated both in settings with a high sediment supply, and in relatively sediment-starved settings as the Porcupine (Øvrebø et al., 2006) (Figure 19.7) in New Zealand (Lu et al., 2003) (Figure 19.9) and Gulf of

397

F.J. Herna´ndez-Molina et al.

Cadiz slopes (Llave et al., 2005, 2007) (Figure 19.6). Plastered, mounded and separated, and sheeted types are the most common drifts on slopes, but occasionally confined, patch, infill, and mixed drifts can develop. Basically, the type of drift formation depends directly on whether the water masses impinging against the slope have a simple or a multiple current pathway. Flow behaviour (Salon et al., 2008) is controlled by sea-floor morphology, current velocity, and Coriolis effects, which cause each particular drift’s geometric formation and specific style of progradational/aggradational stacking pattern. The main characteristics of slope drifts are summarised in Figure 19.11.

19.4.2.

Large-scale erosional features

Although depositional features of contourite systems have been the principal focus of many previous studies, large-scale erosional features are equally common in slope contourites. In some regions, erosional features characterise a wide area of a slope (Viana, 2001; Viana et al., 2002a; Herna´ndez-Molina et al., 2003, 2006c), but more 1. Drifts are generally parallel to the slope 3. No stratigraphic conformity between the top surface and their internal layers

4. Internal widspread erosive discontinuities. They represent phases of stronger currents or major paleoceanographic changes

2. Basal surface. It represents a truncation surface which marks the onset of drift formation

5. Hierrachhy of high-to low-energy large-sacle deposits is generated downstream

6. Drifts ae associated with large-scale bedforms usually sediment waves fields (but also 3-D and 2-D dune fields)

Figure 19.11 3-D sketch summarising the main characteristics of large-scale depositional features (drifts) on slopes. Generally, drifts are parallel to the slope if the impinging water mass has a simple current pathway. If not, the drifts can be parallel, oblique or even perpendicular to the slope margin. In all these cases, however, drifts exhibit down-current elongation with internal widespread erosional discontinuities reflecting intervals of stronger contour currents and major palaeoceanographic changes. Frequent hiatuses of variable durations can be associated with these surfaces.

398

Continental Slope Contourites

F.J. Herna´ndez-Molina et al.

399

commonly they are limited in extent, occurring adjacent to, and associated with, contourite drifts (Fauge`res et al., 1999; Stow and Mayall, 2000). However, as no systematic classification of these features has been published previously, we propose the following classification into six different types of large erosional contourite features on slopes (Figure 19.12), mainly based on findings from the Brazilian Margin (Viana, 2001; Viana et al., 2002a, b) (Figure 19.3), northern European slopes (Masson, 2001; Masson et al., 2002) (Figures 19.4 and 19.7) and, especially, the Gulf of Cadiz (Kenyon and Belderson, 1973; Garcı´a, 2002; Habgood et al., 2003; Herna´ndez-Molina et al., 2003, 2006c; Llave et al., 2005, 2007) (Figure 19.6). Additionally, the conceptual considerations of Evans et al. (1998) and of Stow and Mayall (2000) have been taken into account. These five types are erosional terraces, abraded surfaces, contourite channels, contourite moat and furrows (Figure 19.12).

19.4.3.

Depositional units

On a medium scale, drifts alternate between phases of sedimentation and phases of erosion or non-deposition. On the slope, lateral migration of both depositional and erosional contourite features generates lenticular aggradational sedimentary units in plastered drifts, but sigmoidal progradational units in elongated mounded and separated drifts (Figure 19.13), generally with depositional units of high lateral continuity. The stacking pattern trend of these units is a function of major palaeoceanographic and tectonic changes in the evolution of the CDS (Llave et al., 2007, 2008).

19.4.4.

Facies characteristics

In the following, we consider specific characteristics of slope-contourite systems in terms of acoustic-backscatter facies, seismic facies, sediment facies and bedform types. 19.4.4.1. Acoustic-backscatter facies Sea-floor acoustic response (intensity of the backscattered signal) depends on the angle of incidence of the acoustic pulse, the nature of the sea floor (e.g., grain size in soft sediments) and the roughness/topography of the sea floor (Blondel and Murton, 1997; Blondel, 2003; Medialdea et al., 2008). Depositional and erosional contourite facies exhibit particular acoustic facies, which can be identified in the acoustic-backscatter data obtained with a multi-beam echo sounder or side-scan sonar (e.g. TOBI or Seamap). In general, depositional systems (drifts) have medium

Figure 19.12 Main characteristics of large-scale contourite erosional features on slopes. A classification into six different types of erosional features on slopes is proposed: erosional terraces, abraded surfaces, contourite channels, contourite moat and furrows (adapted from Davies and Laughton, 1972; Garcı´a, 2002; Herna¤ ndez-Molina et al., 2003, 2006c; with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

400

Continental Slope Contourites

Figure 19.13 Lateral down-current migration and up-slope migration (false transgressive onlap) both depositional and erosional contourite features on the Faro ^Albufeira drift system (elongated mounded and separated drifts) on the middle slope of the Gulf of Cadiz (modified from Buitrago et al., 2001; with permission from REPSOL-YPF). The drift ^ moat system mapping the seismic signature of the erosional surface at the top on those sequences, where the abrupt lateral change to highest amplitude signature represents the previous location of the moat.

to low acoustic-backscatter values, whereas large-scale erosional features exhibit high backscatter values (Habgood et al., 2003; Mulder et al., 2003b, 2006; Masson, 2004; Hanquiez et al., 2007). However, local lateral variations in backscatter are common (Figure 19.14), and very useful for studying details of slope-contourite sedimentary processes. 19.4.4.2. Seismic facies Seismic facies of contourite deposits have been described by Myers (1986) and Myers and Piper (1988) using multi-channel seismic reflection profiles, and by Fauge`res et al. (1999) using intermediate-penetration/intermediate-resolution seismic profiles. A wide variety of seismic facies in slope-drift deposits has been observed in the key examples presented before, depending in part on the methods of seismic acquisition and processing used (Nielsen et al., 2008). Additional variation in seismic facies may occur due to the fact that drifts are frequently associated with other kinds of bedform (such as sedimentary waves). A preliminary attempt to relate seismic facies in contourite systems to the intensity of the contour-current regime (Stow et al., 2002f) proposed the following five seismic facies reflecting increasing current strength: (1) semi-transparent reflector-free intervals, (2) continuous sub-parallel moderate- to low-amplitude reflectors, (3) regular migratingwave moderate- to low-amplitude reflectors, (4) irregular wavy to discontinuous moderate-amplitude reflectors and (5) an irregular continuous single high-amplitude reflector. Generally speaking, where the deposits are relatively uniform (fine or coarse sediments) there is a transparent or weak seismic facies, and where extensive sheets of

–7°30′W

–7°W

–6°30′W

(a)

36°N

8°W

7°W

7°30′W

Cadiz contourite channel

(b)

36°N 7°W 36°20′N

6°50′W

6°40′W

(c)

36°10′N

36°N

Figure 19.14 Backscatter facies. (a) Regional original Seamap side-scan data from Sectors1and 2 of the contourite depositional system of the Gulf of Ca¤diz (adapted from Herna¤ndez-Molina et al., 2006c; with permission from Elsevier). (b): High-backscatter areas related to the Ca¤diz Contourite Channel obtained by Simrad EM-12S-120 multi-beam echo sounder (R. Leo¤n and M.C. Ferna¤ndez-Puga, Spanish Geological Survey, IGME, Personal Communication, 2006). (c): Sea-floor acoustic response of the abraded surface revealed by the reprocessing of the Seamap side-scan data ( Joan Gardner, Naval Research Laboratory, USA, Personal Communication, 2006). A multicolour version of this figure is on the enclosed CD-ROM.

402

Continental Slope Contourites

coarse- and fine-grained sediments are interlayered, the seismic reflectors are higher amplitude with good lateral continuity. A more detailed look shows that within slope depositional units, the same cyclicity in seismic facies trends is frequently found (Figure 19.15): (1) a transparent zone at the base; (2) smooth, parallel reflectors of moderate-to-high amplitude in the upper part and (3) a high-amplitude erosional continuous surface at the top. This facies trend is well established within the Faro–Albufeira Drift (Llave et al., 2001, 2006; Stow et al., 2002b), the Porcupine Drift (Van Rooij et al., 2003) and the Eirik Drift (Hunter et al., 2007b). Correlation of high-resolution seismic profiles with cores (Llave et al., 2001, 2006) has shown that (at least in the Gulf of Cadiz) this cyclic pattern of seismic facies response most likely represents cyclic lithological changes showing long-term coarsening-upward sequences bounded on the top by erosional surfaces related to bottom-current

Figure 19.15 Cyclicity in seismic facies in the slope-contourite deposits.The same facies trend is observed in each depositional sequence, and includes (1) a transparent zone at the base; (2) smooth, parallel reflectors of moderate to high amplitude in the upper part and (3) a highamplitude erosional continuous surface at the top. Correlation of high-resolution seismic profiles with calypso piston and gravity cores shows that this cyclic pattern of acoustic response most likely represents cyclic lithological changes showing long-period coarsening-upward sequences to an erosional top. (a) The Gulf of Ca¤diz (Llave et al., 2001; Stow et al., 2002b; Herna¤ ndez-Molina et al., 2003; with permission from Elsevier, the Geological Society of America and The Geological Society, London). (b) Greenland (Eirik drift, Hunter et al., 2007b; with permission fromThe Geological Society, London). (c) Porcupine (Van Rooij et al., 2003; with permission from Elsevier).

F.J. Herna´ndez-Molina et al.

403

changes. Seismic-amplitude maps obtained from 3-D seismic surveys are useful for sea-floor mapping and inferring sediment grain size on the present-day slope surface, since strong amplitudes are generally associated with coarse-grained sediments and low amplitudes with fine-grained sediments (Buitrago et al., 2001; Viana, 2001). 19.4.4.3. Sediment facies Specific contourite facies have been described by many authors (e.g. Stow and Lovell, 1979; Stow and Holbrook, 1984; Stow and Piper, 1984; Pickering et al., 1989; Fauge`res and Stow, 1993; Gao et al., 1998; Fauge`res et al., 1999; Stow et al., 2002f; Rebesco, 2005; Stow and Fauge`res, 2008). Slope contourites may have a wide variety of sediment facies, with different grain sizes (from clays to gravel), composition (terrigenous, biogenic, volcanic and mixed). Although, no specific facies are particularly associated with slope-contourite systems, the most common characteristics are a predominantly poorly sorted, mud-rich facies that is intensively bioturbated, intercalating with thinner horizons of fine-grained sand and silt, typically showing more or less rhythmic bedding (contourite sequences: Stow et al., 2002f ). More rarely, sand-prone slope contourites occur; deep-water massive sands have been reported on the Brazilian Margin (Viana et al., 2002a) and in Sector 1 of the Gulf of Cadiz (Buitrago et al., 2001; Llave et al., 2005). At high latitudes, in particular, slope contourites can contain much gravel-sized material brought in as ice-rafted debris. Gravel-lag contourites are also known from shallow straits, narrow moats and passageways at lower latitudes. Slope-contourite facies typically exhibit a clear cyclicity (Rasmussen et al., 1998; Armishaw et al., 2000; Roveri, 2002; Stow et al., 2002b; Llave et al., 2006; Øvrebø et al., 2006). This cyclic trend was described for the first time from the middle slope of the Gulf of Cadiz (Gonthier et al., 1984), which is composed of three main facies: (1) homogeneous mud; (2) mottled silt and mud and (3) sand and silt. These facies are typically arranged in a coarsening-up to fining-up cycle (1–2–3–2–1) that defines the standard sequence for contourites (Figure 19.16). This has been used as the general model for muddy and sandy contourites, ideally representing an increase from weaker to stronger flow, and then back again to weak (Stow et al 2002c; Hu¨neke and Stow, 2008). Partial or incomplete sequences are also common (Shanmugam, 2000; Howe et al., 2002; Stow et al., 2002f). Other authors have identified similar coarsening-up to finingup sequences, as on the upper slope of the Brazilian Margin (Figure 19.16) (Viana and Fauge`res, 1998) and on the upper, middle and lower slopes of the Porcupine Margin (Øvrebø et al., 2006). 19.4.4.4. Bedforms Slope contourites present the same hierarchy of small-scale bedforms as those drifts located in other marine settings, depending on the relative strength of the current and the nature of sediment available (Kennett, 1982; Stow, 1994; Stow et al., 1996b; Wynn and Masson, 2008).

404

3

sing crea

ity in

Mottled silt and mud. Lenses and irregular layers. Bioturbated. Contacts variable.

10

Mud Bioturbated

Normal grading

Lithology and structure

20

Reversed grading

Sandy silt. Massive irregular sandy

Maximum pockets. Bioturbated. Contacts sharp velocity to gradational.

Current veloc

Subphylocordae

30

g

2

Thalassinoides Faint parallel lamination

Last Glacial Maximum up to the onset of sea-level rise (Latest Pleistocene/Early Holocene)

Mottled silt and mud. Lense, irregular layers, horizontal alignment. Bioturbated.

asin

1

Facies 1

Sand/mud intercalations

cre

0 m

Silt with cross-lamination. Bioturbated

y de

Y

40

cm

Late Pleistocene

locit

Siliciclastic and bioclastic sand

t ve

Z2 Planolites

Mud Bioturbated Laminated

ren

Z1

Holocene

Mean grain size 4 8 16 32 64 μm Cur

Facies 2

50 Siliciclastic muddy sand

Planolites

Coarsening- to fining-upward cycle

0m Facies 3

(b)

Fine slit Coast silt Very fine sand Fine sand Medium sand

(a)

Middle Slope Lithology and structure

Upper Slope

0 Silt and sand Mud

Discontinuous silt lense Grandational contact Sharp (irregular) contact

Figure 19.16 Standard sedimentary sequences of slope contourites. (a) Upper slope of the Brazilian Margin (Viana et al, 2002a; with permission fromThe Geological Society, London). Schematic log representing the vertical facies succession.The general characteristics of these deposits are coarsening- to fining-upward cycles, separated from one another by sharp erosional contacts. Coarse-grained sediments were mainly deposited during the Late Pleistocene/Holocene. X, Yand Z are foraminiferal biozones. Sediment is extremely bioturbated, but ichnofossils are also still present. (b) Schematic sequence of facies from the Faro Drift (middle slope of the Gulf of Ca¤diz) showing the typical superposition of coarsening-upward and fining-upward sequences (Gonthier et al., 1984, Stow et al., 2002c; with permission from The Geological Society, London).

Continental Slope Contourites

Shelf fragments

Bioturbation

F.J. Herna´ndez-Molina et al.

19.5.

405

F INAL C ONSIDERATIONS

Contourite systems are common on many slopes; they show both depositional and erosional features. These contourite systems are generated on any part of the slope (that is, on the upper, middle or lower slope) by different water masses flowing at different depths and at different velocities, in either the same or different directions. Contourite depositional systems have a basal discontinuity which represents the onset of the activity of an impinging water mass (or several) along the slope. That discontinuity is commonly related to a gateway opening or deepening, due to long-term plate–tectonic evolution, and/or large-scale palaeoceanographic changes associated with climatic changes. Although no comprehensive sedimentary model has been established as yet for slope-contourite depositional systems, two end members can be considered in relation to flow behaviour of the impinging water mass. 1. If there are different water masses along the slope but with simple current pathways, different contourite depositional systems can develop. Ideally, these occur on a passive margin without a very complex sea-floor physiography (Figure 19.17a), as in the case of the Brazilian slope and Northern European Margin. In these settings, sand-sheet drifts (or ‘‘sand tongues’’) develop on the upper slope. Meanwhile, plastered drifts rich in fine sand to mud occur on the middle slope, some of which may have associated fields of sedimentary waves down-slope (migrating obliquely up-slope in many cases). On the lower slope (mainly mud-dominated), elongated mounded and separated drifts are most common, in some cases located at the transition to the continental rise. The main erosional features in these settings are erosional terraces, moats and furrows. The formation of these features depends very much on the position of the main current core, the velocity of the current and sediment supply. 2. On margins where recent tectonic activity has produced a complex slope morphology, more possibilities exist for generating multiple current pathways, including several different branches of the current, secondary flows, small fluxes (filaments), internal waves, local turbulence associated with eddies, overflows and helicoidal flows. Topographic obstacles produce local acceleration and deceleration of currents. In this context, very complex contourite depositional systems can be generated, including the many kinds of depositional (most of the drifts described by Rebesco, 2005) and erosional features (erosional terraces, abraded surfaces, contourite channels, marginal moats, erosional scours and furrows) (Figure 19.17b). This is the case on the slopes of the Gulf of Cadiz and western Iberian margin (Llave et al., 2001, 2005, 2007, 2008; Alves et al., 2003; Herna´ndez-Molina et al., 2003, 2006c; Mulder et al., 2003b), as well as on those of some presently active margins (Reed et al., 1987; Carter and McCave, 1994). Ultimately, the presence of simple or multiple current pathways depends on the particular physiography of a continental margin, and therefore on both tectonic activity and sediment supply. Tectonic activity represents a key factor in producing morphological changes on the sea floor (producing slope basins, diapirs, uplifted fault blocks, banks, ridges, etc.), thereby controlling the development of new

406

Continental Slope Contourites

F.J. Herna´ndez-Molina et al.

407

Figure 19.17 3-D conceptual sketch showing the potential contourite depositional and erosional features on a slope over: (a) margins with different water masses along the slope but with simple current pathways (e.g. the Brazilian slope and the Northern European Margin), (b) margins where recent tectonic activity has produced very complex slope morphology, increasing the possibility of generating multiple-current pathways (e.g. the slopes of the Gulf of Cadiz, western Iberian Margin and some active margins) and (c) interaction between down-slope (submarine canyons and slides) and along-slope processes (slope currents). A multicolour version of this figure is on the enclosed CD-ROM.

pathways for the core and branches of the impinging current at each evolutionary stage of the slope. Subsequently, this will condition drift stratigraphy, architectural changes and the location of large-scale erosional features on the long term. Moreover, active slope tectonics may trigger more active down-slope processes, which can partially or completely mask contourite processes (Figure 19.17c). On the contary, environmental (palaeoclimatic or sea level) and palaeoceanographic changes are other essential factors controlling slope-contourite evolution. On a short term, they control the vertical contourite stacking pattern, sequences and facies. Climate plays an important role in the position of the core of a contour current, thereby producing two different effects: (1) vertical and lateral changes in the core position and (2) changes in the intensity of the current cores. These changes are not coeval and global. On some continental margins, some water masses are stronger during cold, glacial intervals, but on other margins, other water masses are more active during warmer, interglacial intervals (Knutz, 2008). Moreover, some currents are active all the time, but their effects are masked (or not) by down-slope processes, mainly during regressive and lowstand sea-level stages (Mulder et al., 2008). Therefore, it is not possible to fit along-slope contourite processes and contourite depositional systems in a global sequence-stratigraphic model, because contour currents do not fit universally into a particular segment of the eustatic cycle. However, for a particular margin and current system, it should be possible to derive a regional sequence-stratigraphic model. Future research on slope contourites, using new and better techniques, should focus on two main objectives: (1) a more detailed understanding of water-mass circulation (especially regarding submarine obstacles), their behaviour and variability (especially tidal and benthic storms), which could help to explain the common association between drifts and other bedforms (e.g. sedimentary waves and dunes); and (2) establishing facies models, including their association with other deep-water sedimentary environments, in both the present and ancient submarine domains.

ACKNOWLEDGEMENTS This compilation has been made possible by the support of many people who sent us their original data (some of them unpublished) and figures to be included. So, we would like thank: Adriano R. Viana (PETROBRAS, Brasil); Carol Pudsey

408

Continental Slope Contourites

(BAS, UK); Craig S. Fulthorpe (University of Texas, USA); David van Rooij (Ghent University, Belgium); Douglas G. Masson (National Oceanography Centre, Southampton, UK); Fernando Bohoyo (Spanish Geological Survey, IGME, Spain); Gemma Ercilla (Instituto de Ciencias del Mar, CMIMA-CSIC, Spain); Jan Sverre Laberg (University of Tromsø, Norway); Joan M. Gardner (Naval Research Laboratory, USA); Lena; K. Øvrebø (University College Dublin, Ireland); Mari Carmen Ferna´ndez-Puga (Instituto Espan˜ol de Oceanografı´a, IEO, Spain); Ricardo Leo´n (Spanish Geological Survey, IGME, Spain); Sally Hunter (National Oceanography Centre, Southampton, UK); Tjeerd C.E. van Weering (Royal Netherlands Institute for Sea Research, The Netherlands); and Tove Nielsen (The University Centre in Svalbard (UNIS)). We are also very grateful to both TGS-NOPEC Geophysical Company (and especially to Frode Sandnes) and REPSOL-YPF (and in particular to Wenceslao Martinez del Olmo) for allowing us to use unpublished seismic records of the Gulf of Cadiz. We thank the revision and suggestions of Dr R.D. Larter (BAS, UK) of the manuscript draft who helped us to improve the present contribution. We also thanks both Editors and Project Manager for their interest and corrections that have helped us to improve the final version of our chapter. This work has been carried out as part of two research stages funded by the ‘‘Mobility Award’’ from the Spanish Ministry of Education and Science. These awards enable Dorrik A.V. Stow to work at the Instituto Espan˜ol de Oceanografı´a, Malaga (Reference SAB2005-0182) and F.J. Herna´ndez-Molina to work at the National Oceanography Centre, Southampton (NOCS) (Reference PR20060275) and Marine & Geophysical Division of the Argentine Hydrographic Institute (PR2007-0138). The Spanish Comisio´n Interministerial de Ciencia y Tecnologı´a (CYCIT) supported this research through the Project CTM 2008-06399-C04-01/MAR (CONTOURIBER Project).

C H A P T E R

2 0

S HALLOW -W ATER C ONTOURITES G. Verdicchio1,2 and F. Trincardi1 1

ISMAR-CNR, Via Gobetti, Bologna, Italy EDISON SpA, Foro Buonaparte, Milano, Italy

2

Contents 20.1. Introduction 20.2. Bottom Currents Shaping Shallow-Water Contourite Deposits 20.2.1. Thermohaline contour currents 20.2.2. Cascading currents 20.2.3. Wind-driven currents 20.2.4. Internal waves and tides 20.3. Examples of Shallow-Water Contourites 20.3.1. Outer shelf of the SE-African continental margin 20.3.2. Antarctica shelves 20.3.3. Campos Basin outer-shelf/upper-slope contourite system (offshore SE Brazil) 20.3.4. Gulf of Cadiz outer-shelf/upper-slope contourite system 20.3.5. Mediterranean shallow-water contourite deposits 20.3.6. Shallow contourite deposits within sills, gateways and slope basins 20.4. Discussion 20.4.1. Shallow-water contourite deposits and sea-level change 20.4.2. ‘‘Extremely-shallow’’ contourite deposits 20.4.3. Slope stability of shallow-water contourite deposits 20.4.4. Modern versus ancient shallow-water contourite deposits 20.5. Summary Acknowledgements

20.1.

409 412 413 413 414 415 415 415 416 418 420 423 427 430 431 431 431 432 432 433

INTRODUCTION

The term contourites was first introduced to define deep-sea deposits that are genetically linked to the flow of deep and long-lasting (i.e. remaining in the same state for an indefinitely long time) contour-parallel bottom currents of thermohaline origin (Heezen et al., 1966; Hollister and Heezen, 1972). Since the 1970s, a great number of continental margins worldwide have been investigated, and contourite systems have been increasingly recognized from abyssal Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00220-3

Ó 2008 Elsevier B.V. All rights reserved.

409

410

Shallow-Water Contourites

plains, in water depth ranging from over 5000 m (e.g. Stow and Holbrook, 1984; McCave and Tucholke, 1986) to relatively shallow, upper slopes (e.g. Viana and Fauge`res, 1998; Reeder et al., 2002; Rodrı´guez and Anderson, 2004; Verdicchio and Trincardi, 2008b). While general agreement exists on the definition of deep-sea contourite drifts (sediment accumulation formed principally by bottom currents) along continental rises and lower-slope settings, some confusion remains in the definition of contourite drifts in relatively shallow-water settings, from the outer shelves to the upper slope, including shallow sills that connect different sub-basins. The difficulty in defining shallow-water contourite deposits is evident also when considering fossil sedimentary successions exposed on land where water depth and morphological and oceanographic setting, as well as the dominant sedimentary processes active at the time of deposition, remain uncertain (Stow and Lovell, 1979; Lovell and Stow, 1981; Stow et al., 1998a). In their work about ancient sandy contourites, Lovell and Stow (1981) broadened the definition of the term contourite with respect to that originally proposed by Hollister and Heezen (1972), referring to a persistent bottom-current regime that flows under the wave base and whose origin is not critical to define the contourite system. Fauge`res and Stow (1993), returning to the original concept, proposed to use the term contourite for deposits below 500 m depth and located where a relatively steady deep geostrophic current exists. In addition, these authors suggested using the general term bottom-current deposits for those deposits that do not match strictly the definition of contourites. Stow et al. (1998a) shifted the limit of contourite deposition upward (water depth of 300 m) excluding from the contourites sensu stricto all deposits affected by surface waves (including storm waves), wind-generated currents, tidal currents, clear-water currents in canyon or upwelling currents. Stow et al. (1998a), moreover, suggested not to use the term fossil contourite in case of ancient sedimentary series deposited where the palaeocurrent regime and palaeodepth are unknown. Based on the morphological setting, Viana et al. (1998a) recognized three depths of deposition for contourites: deep-, mid- and shallow-water, corresponding to >2000 m, 300–2000 m and 50–300 m, respectively. While deposits in the first two zones fit the definition of contourites, those in the shallowest range can be referred to as shallow-water bottom-current deposits, as they may reflect also other hydrodynamic factors (shelf currents, tides and waves, including internal waves) capable of impinging the sea floor. Considering the depth variations of the bottom-current flows, like those documented for example on a seasonal basis, a mid/upper-slope bottom-current system can impact water depths shallower than 300 m; Stow et al. (2002c) suggested to use the term ‘‘shallow water contourite’’ for defining deposits in this setting, and the same should be used for contourite drift deposits in relatively shallow gateways. Finally, in their recent discussion about the economic importance of contourite deposits, Viana et al. (2007) employed the term ‘‘shallow-water contourite’’ to describe bottom-current deposits in water depths where the influence of currents derived from storms, tides and trade winds is still negligible compared with a dominant geostrophic current. Along upper slopes and outer continental shelves (i.e. in the first few hundreds of metres of water), the bottom currents that impact the sea floor derive from a variety of processes, most of which can be related to thermohaline and atmospheric forcing, show short-term changes in intensity and direction, and exhibit eddies and, in some cases,

411

G. Verdicchio and F. Trincardi

distinctly bidirectional flows (Johnson and Baldwin, 1996). These relatively shallow bottom currents impact widespread areas for significant intervals of time (up to 104 years) but can appear ephemeral if compared with the quasi-steady-state and long-lasting (up to 106 years) geostrophic currents acting in much deeper water (thousands of metres). In several examples of upper continental slopes (e.g. in the south Adriatic Sea: Verdicchio et al., 2007), bottom currents derive from a stable geostrophic circulation and can form contourite drifts resembling, in morphology, internal geometry and sediment facies, the typical contourite sediment drifts profusely described from deep-water settings (Fauge`res et al., 1999). At places, bottom currents can resuspend and deposit a thin (up to few metres) veneer of sandy sediment on the outer continental shelf (e.g. in the Campos Basin: Viana et al., 1998a, b). In this chapter, we focus on some Quaternary examples of shallow-water contourites from a wide variety of Quaternary geological settings (Figure 20.1).

6

5 1

2000 km

11

2

12

8

7

Antarctica

10

9 1000 km

1500 km The Mediterranean Sea

4

3

Figure 20.1 Location of the examples of shallow-water contourites selected for this study: 1 = Kwazulu-Natal shelf deposits, west Africa; 2 = The Andvord Drift, west Antarctic Peninsula; 3 = The Mertz Drift, George V Land, west Antarctica; 4 = sand sheet off Pennel coast, NorthVictoria Land, Antarctica;5 = Campos Basin shelf/upper slope contourite system, SE Brazil; 6 = Gulf of Cadiz shelf/upper-slope contourite system, SW Spain; 7 = Corsica Basin contourite deposits, NW Mediterranean; 8 = Southwestern Adriatic Margin shelf-edge contourite deposits, south Adriatic; 9 = Eastern Gela Basin shelf/upper-slope contourite system, Sicily Channel; 10 = The Marmara Sea contourite deposits, east Mediterranean; 11 = Denmark Strait deposits, North Sea;12 = Pelagosa sill contourite drift, central Adriatic.

412

Shallow-Water Contourites

In these settings, recent deposits can be interpreted in the context of a known physiographic setting and water depth, and interpreted taking into account the oceanographic regime under which they formed; in addition, ultrahigh-resolution seismic stratigraphy allows reconstruction of the external morphology and internal geometry of sediment drifts and related bottom-current deposits. Older deposits, defined by a lack of seismic and outcrop data information, cannot be unequivocally defined as shallow-water contourites. We included also examples from high-latitude glacigenic shelves of Antarctica that cannot be considered strictly as a shallow-water environment, because in some cases they reach on the inner shelf water depths of over 800 m. The variety of settings, characterized by marked differences in the depth of the shelf edge, morphology of the shelf and steepness of the upper slope, hinder the adoption of a strictly depth-dependent definition of the term ‘‘shallow-water contourite’’. For such reasons, we consider the term ‘‘shallow-water contourites’’ only as a general term for deposits, from fine- to coarse-grained, that are controlled by the activity of long-lasting bottom currents on shelves, upper slopes and shallow sills. Anyhow, an adjective specifying the context (if known) in which shallow-water contourites occur (e.g. upper-slope contourite; mid-shelf contourite; shelf-break contourite) would be sufficient to avoid further confusion in describing such a peculiar class of deposits.

20.2.

B OTTOM C URRENTS S HAPING SHALLOW-W ATER C ONTOURITE D EPOSITS

Bottom currents capable to impact the sea floor in shallow-water settings may derive form a variety of processes including thermohaline circulation, wind-driven circulation, tide- or wave-induced circulation and internal motion of water masses (internal tides and internal waves). Bottom currents generated by wind and tidal forcing are almost exclusive of shallow-water environments; thermohaline circulation and internal waves and tides operate at any water depth (Cacchione et al., 2002). Among the bottom currents that impact the sea floor in relatively shallow water, only long-lasting bottom currents (>104 years) are typically considered in the study of contourite systems. The term ‘‘long-lasting currents’’ entails a long interval of activity but does not necessarily imply lack of fluctuations in velocity or other physical properties of the water mass. Indeed, both in shallow and deep water, ocean currents are subjected to changes in flow depth, strength, direction and physical properties of the water mass, from seasonal to millennial scales. Seasonal variations reflect the meteorological forcing responsible of the formation of bottom currents and are particularly significant on numerous continental shelves worldwide where winter cooling generates dense water masses that flow seasonally towards the basin (Ivanov et al., 2004). Tide-related currents reach their maximum intensity during the spring, when tides reach their maximum amplitudes (Pickard and Emery, 1990). Millennial-scale variations in bottom-current strength and direction respond to basin-scale changes in oceanographic regime, caused by global climatic and

G. Verdicchio and F. Trincardi

413

sea-level fluctuations and by tectonic deformation, over longer intervals (>105 years). Tectonic deformation is also responsible of modifications of the basin shape and the degree of connection among basins. These long-term variations affect the areas where bottom currents form by influencing the preconditioning climatic factors, including changes in air temperature and shifts in wind circulation. In deep water, millennial-scale variations lead to changes in depth range and intensity of the bottom current (e.g. Christie-Blick et al., 1990; Myers et al., 1998). In general, the shallowest contexts are more drastically affected by climatic and eustatic variations. Beyond the marginal areas that become subaerially exposed during lowstands, sea-level variations lead to changes in the accommodation space available for sediment storage ( Jervey, 1988) and in the pathway of the tidal currents (Schlager, 1993). In turn, all these factors lead to substantial variations of the bottom-current regime that significantly impact sediment supply and dispersal along the continental margin. The existing examples of shallow-water contourite deposits allow recognition of four main oceanographic regimes that may result in sustained unidirectional contour-parallel flows over appreciable distances and prolonged intervals.

20.2.1.

Thermohaline contour currents

Theromohaline currents are driven by uneven density distributions of water masses that are caused by latitudinal changes in temperature and salt-content, worldwide (e.g. Stommel and Arons, 1960; Toggweiler and Key, 2001; Wunsch, 2002). Thermohaline currents are relevant on shallow settings in two main ways: (1) they can enter or exit enclosed basins, often acting as concentration basins where evaporation exceeds the combined effect of precipitation and runoff (e.g. the Mediterranean Sea: Wu¨st, 1961); or (2) they intrude outer shelf areas particularly on open continental margins (e.g. SE Africa: Flemming, 1981; and the East China Sea: Guo et al., 2006). Margin configuration, including seaward projections or re-entrances of the shelf, affects the along-slope flow of thermohaline currents generating zones of current acceleration (flow constriction) or deceleration (flow expansion). In the latter case, the flow expansion can induce the formation of gyres or eddies on the outer shelf (Figure 20.2; Viana et al., 2007).

20.2.2.

Cascading currents

Cascading currents are a specific type of thermohaline-driven currents generated by dense waters formed by evaporation, cooling or freezing in the surface layer over the continental shelf and descending across the continental slope (Shapiro et al., 2003). This process is a key component of the ocean-floor ventilation worldwide and therefore affects global oceanic circulation and climate. The impact of such currents on the sea floor has recently received attention by the sedimentological community (Canals et al., 2006; Trincardi et al., 2007). These currents appear to propagate both along and across the contours under the influence of gravity, Earth rotation and mixing with other water masses, and may result in the formation, also

414

Shallow-Water Contourites

d

a

Shelf

g Slope

Figure 20.2 Schematic representation of along-slope current behaviour with respect to the margin configuration (from Viana et al., 2007; with permission from The Geological Society, London). Zones of bottom-current acceleration (a) are related to margin outbulges and topographic obstacles that confine the flow; zones of deceleration (d) correspond to margin re-entrants where flow expansion can induce the formation of gyres or eddies (g). Size of arrows indicate current intensity.

in relatively shallow waters, of sedimentary deposits that are very similar to classical deep-sea contourites (Verdicchio et al., 2007).

20.2.3.

Wind-driven currents

The wind-driven circulation occurs principally in the upper few hundreds of metres and therefore is primarily a horizontal circulation, in contrast to the thermohaline one. Wind shear stress on the sea surfaces causes horizontal movement of the superficial layers and the subsequent propagation of the motion through the water column through the friction of different water layers. Consequently, the effect of wind stress extends through the upper few hundred metres of the water column (Pickard and Emery, 1990). Considering the vertical distribution of the wind stress, the direction of the wind-driven current differs from the main direction of the wind. In superficial layers, this difference is related to the Coriolis force associated with the rotation of the Earth. In the deeper layers (up to few hundred of metres), a wind-driven current veers in a clockwise direction with depth as expressed by the Ekman spiral effect (Ekman, 1905; Pickard and Emery, 1990). Extreme events, such as storms, can strongly amplify the wind shear strength on the sea-surface, increasing the energy and, consequently, the shear stress on the sea floor.

G. Verdicchio and F. Trincardi

20.2.4.

415

Internal waves and tides

Internal waves and internal tides (i.e. internal waves associated with tides) are gravity waves that oscillate at the density interface between distinct water masses (Karl et al., 1986; Cacchione et al., 1988). These processes are relevant in the longterm shaping of continental slopes (Cacchione et al., 2002; Shanmugam, 2008) and probably represent one of the main processes governing sediment resuspension and the formation of nepheloid layers in outer-shelf and upper-slope settings (Puig et al., 2004). Bottom currents associated with the breaking of internal waves, in particular, have the capability to shape sedimentary bodies in contour-parallel depocentres and up-slope-migrating bedforms (Karl et al., 1982).

20.3.

EXAMPLES OF SHALLOW -W ATER C ONTOURITES

This section includes a set of selected examples of Quaternary shallow-water contourites (Figure 20.1) from high to low latitudes and from margins of oceanic basins to margins of semi-enclosed basins such as the Mediterranean and the Baltic. For all these examples of contourite deposits, the physiographic setting is well constrained and the bottom-current regime known. We include also examples of contourite drift from the inner shelves of the Antarctica, where water depths are over 800 m, in response to their glacigenic evolution.

20.3.1.

Outer shelf of the SE-African continental margin

20.3.1.1. Physiographic and oceanographic setting The continental shelf off the Kwazulu-Natal (SE Africa) represents one of the bestknown examples of a shelf dominated by oceanic currents generating distinctive deposits and erosional features. The morphology of this modern shelf largely reflects the Pleistocene sea-level fluctuations, as well as sedimentary processes during the Holocene (Flemming, 1981). The width of the shelf ranges from 40 to <4 km and a shallow shelf-edge is sharply defined between 65 and 100 m water depth. Numerous canyons and gullies elongated southeastward across the steep slope dissect the shelf break (Flemming, 1978; Ramsay, 1994). The area is characterized by a microtidal/mesotidal regime with a typical tidal range of 2 m, and the inner shelf is impacted by recurrent high-energy waves and high-amplitude swells from the southeast (Ramsay, 1994). The most important oceanographic feature is the western boundary south-flowing Agulhas Current (AC), which forms off the northern Mozambique coast from the confluence of waters from the Mozambique Channel and the areas south of Madagascar (Lutjeharms et al., 1981). Upwelling of the AC is observed at the shelf-edge and may occasionally intrude onto the shelf (Lutjeharms et al., 2000), but these intrusions do not display a clear seasonal pattern (Schumann and Van Heerden, 1988; Goschen and Schumann, 1990). Marked changes in current velocities along the shelf are therefore related to large eddies diverting the general southward trend of the AC (Flemming, 1980; Ramsay et al., 1996).

416

Shallow-Water Contourites

20.3.1.2. Bottom-current deposits Large fields of dunes, composed by fine-grained sand, reach a maximum thickness of 20 m and are well documented from the outer shelf (Flemming, 1978, 1980; Ramsay, 1994; Ramsay et al., 1996). The dunes are accompanied by other distinctive bedforms such as comet marks, sand streamers, sand ribbons and smooth-top sand sheets (Flemming, 1980). The sedimentary processes responsible for the growth and migration of these bedforms (sand dunes migrate at rates of 1.25 m per year) are imputable to the south-flowing AC intruding on the outer shelf. The largest subaqueous dunes documented by Flemming (1980) have a maximum height of 17 m and a wavelength of almost 700 m. Near-bottom current velocities of up to 1.3 m s 1 were estimated by Flemming (1978) to produce such bedforms. Subordinate hummocky and swaley structures and oscillation ripples occur in water depths over 60 m, indicating that storm waves are sometimes superimposed on the unidirectional current flow (Ramsay, 1994).

20.3.2.

Antarctica shelves

20.3.2.1. Physiographic and oceanographic setting Antarctica margins are ideal sites for assessing the role of geostrophic and thermohaline currents in transporting and sorting sediment across shelf areas. Along these continental margins, there is no fluvial input, and glaciers deliver sediment directly to the continental shelves that typically dip landward because of the combined effect of the ice-sheet loading and scouring (Anderson, 1999). The continental shelves of Antarctica are deeper than mid-latitude shelves with an outer shelf that averages between 200 and 500 m, and an inner shelf that may reach water depths of over 800 m. These characteristics are produced by the combined effect of loading and scouring by the grounded ice sheet (e.g. Vega Channel: Camerlenghi et al., 2001; Palmer Deep: Rebesco et al., 1998a; Dennistoun Trough: Anderson, 1999). On Antarctica shelves, tidal currents are relatively weak, whereas wind-driven currents impact sea-floor areas shallower than 300 m (Anderson et al., 1984). South of the Antarctic Divergence, the Antarctic Surface Water (ASW) flows mainly westward (counter clockwise around the continent), along the shelf break as East Wind Drift (Davis, 1972; Pickard and Emery, 1990), reaching velocities exceeding 27 cm s 1 (Rodriguez and Anderson, 2004). Further North, within the Antarctic Polar Front, water circulation is dominated by the strong, eastward-flowing Antarctic Circumpolar Current (ACC), also named ‘‘West Wind Drift’’ because it is attributed to the stress of the strong westerly winds (Pickard and Emery, 1990). Below the surface, and down to the bottom at depths exceeding 4000 m, the Antarctic Circumpolar Deep Water (CDW) flows eastward around the continental margin as a deep component of the ACC. This water mass often rises up to the surface, either in the proximity of the Antarctic Divergence or along the Antarctic continental slopes, intruding the continental shelf ( Jacobs et al., 1978, 1985). The Antarctic Bottom Water (AABW), the densest and deepest water in the world oceans, forms in the regions where the westward-flowing surface currents approach the coast, with localized areas of enhanced dense water formation in the Weddell and Ross Seas (Weaver et al., 1999).

417

G. Verdicchio and F. Trincardi

On the continental shelf, the surface water freezes and rejects the salt to increase the salinity and hence the density of the underlying water (Foldvik and Gammelsrøld, 1988). Once formed, this water accumulates at depth within shelf depressions, and rapidly cascades across the continental slope, eventually spilling over sills or draining through canyons, and reaching the deep sea (Ivanov et al., 2004). We describe here three examples of Antarctica shelf contourites: the Andvord Drift (Harris et al., 1999), the Mertz Drift (Harris et al., 2001) and a widespread sand sheet off Pennel Coast (Rodriguez et al., 2004).

Rongé Island

Gerlache Strait

South Shetland Islands Study site

o

lag

ipe

er

lm Pa

0

ic P enin

64°45’

arct

n Sea Belling

Ant

40 50

shause

20

Palmer Deep

sula

Anvers Island

ra Erre nel n a Ch

h Arc

Larsen Ice Shelf

Weddell Sea

62°40′

62°50′

DancoIs.

20.3.2.2. The Andvord Drift The Andvord Drift is located on the shelf of the west Antarctic Peninsula between 300 and 500 m water depth (Harris et al., 1999). It is a multi-lobate deposit that covers an area of 44 km2 among the outer reaches of the Andvord Bay Fjord, Gerlache Strait and the Aguirre and Errera Channels (Figure 20.3). This deposit shows all the main features of deep-sea sediment drifts. It contains up to 40 m of post-glacial mud rich in ice-rafted debris (IRD) (Domack and Ishman, 1993). The maximum thickness of the Andvord Drift is reached seaward, within the Gerlache Strait and the adjacent channels, although the sediment input comes from

Arctowski el

or dv An

nn Ch a re

40

uir

Coast

Ag

64°50′

Peninsula 0

y Ba

Danco

d

Lemaire Island

0

5 km

Isopach interval, 10 ms

Figure 20.3 Isopach map showing thickness, above the Last Glacial Maximum diamicton, of the unconsolidated sediments that form the Andvord Drift (from Harris et al., 1999; with permission from the Geological Society of America). See Figure 20.1 for location.

418

Shallow-Water Contourites

a tidewater glacier in Andvord Bay, and might be expected to lead to thicker deposits close to the land, as suggested by studies along glaciated coasts (e.g. Syvitski et al., 1987). The observed pattern of deposition and thickness distribution is consistent with bottom currents flowing along the western side of the Antarctic Peninsula, intruding channels adjacent to the Andvord Bay. 20.3.2.3. The Mertz Drift The Mertz Drift is located on the East Antarctica continental shelf in an 850-m deep landward-dipping trough of glacial origin off George V Land (Harris et al., 2001). It covers an area of about 400 km2 and comprises Late Quaternary unconsolidated deposits up to 35 m thick (Domack, 1988; Figure 20.4). A likely oceanographic mechanism causing the observed depositional pattern in the drift area is the upwelling of the modified CDW onto the shallow outer shelf, its transformation in High Salinity Shelf Water (HSSW) by salt rejection from freezing surface waters, and its subsequent outward flow, which entrains and transports fine-grained sediment (Rintoul, 1998). Several sediment units can be distinguished within the Mertz Drift (Harris et al., 2001). During the mid-Holocene (5–3 ka B.P.), fine-grained sediment, with minimal IRD content (unit 3), was rapidly deposited (up to 290 cm ka 1), suggesting relatively sluggish bottom currents likely related to an Antarctic climate warmer than at present (Cunningham et al., 1999). Between 13 and 5 ka B.P. and after 3 ka B.P., sandy mud layers, rich in IRD (units 2 and 4), were slowly deposited (up to 10 cm ka 1), suggesting, instead, more energetic bottom currents under a climatic regime similar to the present day. 20.3.2.4. Sand sheet off the Pennel Coast A widespread (3200 km2) sand sheet, 0.1–1 m thick, with sharp top and base, is located between 200 and 1200 m water depth on the shelf and upper slope offshore the Pennel Coast, North Victoria Land (Rodriguez et al., 2004). This deposit has been interpreted as a shelf/upper-slope contourite deposit generated by the westward flowing CDW impinging the shelf. The sand sheet is composed of fine/medium sand, transported by the current up to 70 km from its source (Cape Adare) within the past 9 ka (average accumulation rate of 8 cm/1000 years). This deposit displays no appreciable vertical trend in grain size, which suggests fairly constant current velocities over the time interval of deposition.

20.3.3.

Campos Basin outer-shelf/upper-slope contourite system (offshore SE Brazil)

20.3.3.1. Physiographic and oceanographic setting The Campos Basin is situated along the Brazilian southeastern passive continental margin, presently dominated by contourite deposition. The width of the Campos continental shelf ranges between 50 km in the North and 100 km in the South. The margin presents an abrupt shelf break at a depth between 80 m and 130 m, and a steep (10°) erosional upper slope, down to 220 m deep, that displays several gullies up to the shelf edge (Viana et al., 1998a). The shelf broadens, displaying a seaward

419

G. Verdicchio and F. Trincardi

(a)

(b)

Magnetic susceptibility Magnetic susceptibility

Magnetic susceptibility

Water depth (m)

Accumulation rate

Accumulation rate

Late quaternary drift sediments

Figure 20.4 The Mertz Drift built by the flow of High Salinity Shelf Water of variable intensity (from Harris et al., 2001; with permission from Elsevier). See Figure 20.1 for location. (a) Location of the drift with the schematic circulation pattern of the Modified Circumpolar Deep Water (MCDW) and of the High Salinity Shelf Water (HSSW). (b) 3.5-kHz seismic profile across the Mertz Drift with associated core logs showing the sediment units described in the text.

bulge dissected by the Sao Tome´ Canyon separating the northern and southern upper slopes where an extensive, flat erosional terrace extends from the shelf break to a depth of 450 m (Viana et al., 2002a, b). Water circulation over the Campos Margin includes two distinct shelf and upper-slope components. The former results from meteorological forcing, including NE trade winds and SW storm fronts, and wave and tidal forcing. The latter

420

Shallow-Water Contourites

reflects the flow of two distinct water masses: the southward-flowing Superficial Tropical Water (STW) and the northward-flowing South Atlantic Central Water (SACW) (Viana et al., 1998b). The more superficial STW is carried southward by the Brazil Current (BC), driven by the wind-controlled South Atlantic Gyre (Signorini, 1978), and is best defined between the shelf break and a depth of 300 m. Locally, superficial meanders of the BC may generate bottom currents on the outer shelf (Evans et al., 1983), as a consequence of shearing and exfoliation of the BC offshore the basinward bulge of the shelf break and the thermal contrast between the upwelling cold SACW and the warm and saline BC (Viana et al., 1998a). 20.3.3.2. Bottom-current deposits In the Campos Basin, the geostrophic quasi-steady-state BC sweeps the upper slope, penetrating the outer shelf where the interaction with other shelf currents controls the shaping of the sandy deposits of bottom currents (Viana and Fauge`res, 1998). On the outer-shelf, low-amplitude, outward-migrating sand dunes spread extensively for more than 100 km2 (Viana et al., 1998a). The inner portion of the broad erosional terrace, north of the Sao Tome´ Canyon, is covered by a system of erosional ridges up to 50 m high, 1 km wide and 10 km long. They are associated with distinct high-energy erosional features (furrows, scours, comet marks) suggestive of bottom-current velocities up to 1 m s 1 (Viana and Fauge`res, 1998). Toward the south, the BC decelerates in response to shelf-slope morphology, which shows a landward inflection, and BC generates eddies that interfere with shelf processes. In this sector, a field of mud or sand waves and dunes (1 m in high, tens of m in length) develop. On the upper slope, a sand-rich slope plastered sheet drift (using the contourite classification proposed by Fauge`res et al., 1999) results from the interaction between off-shelf sand spill-over and the BC (Viana, 2001); silty- and muddy-sand dunes occur on the upper slope down to 450 m on the flat erosional terrace (Viana et al., 2002a, b; Figure 20.5). Viana et al. (2002a) ascribed the growth of these upper-slope sand deposits to the intense bottom-current regimes established during Late Pleistocene sea-level rises (Figure 20.6).

20.3.4.

Gulf of Cadiz outer-shelf/upper-slope contourite system

20.3.4.1. Physiographic and oceanographic setting The Gulf of Cadiz is located Northwest of the Gibraltar Strait in the eastern Atlantic Ocean. The first geological investigations were carried out by Heezen and Johnson (1969) and Kenyon and Belderson (1973). The margin of the Gulf of Cadiz shows the interplay between a complex tectonic history associated with the African and Iberian plate-boundary dynamics, a multi-source sediment supply and the inflow and outflow currents through the Mediterranean gateway (Maldonado and Nelson, 1999b; Herna´ndez-Molina et al., 2008a). This area is rimmed by an arcuate coast and includes a continental shelf up to 45 km wide with a 120–140 m deep shelf break. The western sector is characterized by an abrupt margin incised by the Portimao Canyon (Llave et al., 2001). The eastern sector is characterized by a higher variability in sea-floor gradient with a steep (2–3°) upper slope, between

421

G. Verdicchio and F. Trincardi

(a)

(b)

t1

(c)

t2

N

Brazil current Southern current

S

Figure 20.5 Upper-slope morphology related to the Brazil Current (fromViana et al., 2002b; with permission from The Geological Society, London). (a) Side-scan sonar mosaic showing the bedforms at the surface of a plastered sheet drift. (b) 3.5-kHz seismic profiles and core logs located in the region with sand dunes near the top of the slope where a narrow terrace is developed. (c) Block diagram showing the bottom-current regime along the Campos Margin with the associated bedforms.

130 and 450 m, and a broad slope terrace dipping less than 1° (Habgood et al., 2003). A field of NE-SW trending ridges of diapiric origin characterizes this sector (Herna´ndez-Molina et al., 2003; Marche`s et al., 2007). The present-day circulation of the Gulf of Cadiz is controlled by the exchange of water masses through the Gibraltar Strait, consisting of the superficial Atlantic inflow of the southeastward flowing North Atlantic Superficial Water (NASW), shallower than 300 m, and the Mediterranean Outflow Water (MOW), defined by the westward flow of the Mediterranean Intermediate Water and of the Mediterranean Deep Water (O’Neil Baringer and Price, 1999). Because of their greater density than the surrounding Atlantic waters, the MOW progressively sinks as it flows toward the open ocean, across the isobaths, at a depth between 200 and 1800 m (Ambar and Howe, 1979). 20.3.4.2. Bottom-current deposits The variety and distribution of sea-floor bedforms suggest that the eastward flowing NASW accelerates significantly toward the Gibraltar Strait in water shallower than 300 m, across the continental shelf and the upper slope of the Gulf of Cadiz. The

422

Shallow-Water Contourites

Figure 20.6 Seismic-stratigraphic scheme of the upper-slope sand deposition in the Campos Basin (fromViana et al., 2002b; with permission fromThe Geological Society, London). Clean sand was deposited during the early phases of the post-glacial sea-level rise, when the Brazil Current (BC) achieved its maximum intensity. The relative deceleration of the BC during the present-day highstand resulted in the deposition of muddy sands. Sand/mud intercalations correspond to deposits formed during sea-level lowering.

inner shelf is dominated by the presence of sand dunes formed by the combination of the Atlantic inflow current and storm waves (Nelson et al., 1999). The middle slope is characterized by a Holocene highstand mud-wedge reaching a thickness of 25 m off the major river mouths (Gutierrez Mas et al., 1996; Rodero et al., 1999). The deposition of the highstand mud wedge is confined inshore, above a water depth of 100 m, by energy-rich processes related to the complex pattern of shelfedge currents and internal waves (Karl et al., 1983; Stanley et al., 1983; Cacchione and Drake, 1986). Advection, associated with the SE-flowing NASW, leads the along-shelf progradation of the mud wedge (Nelson et al., 1999), similar to cases documented from other continental margins (e.g. Amazon: Nittrouer et al., 1984; Adriatic: Cattaneo et al., 2003). Transgressive coarse-grained sands are the dominant deposit on the outer shelf, where the strong Atlantic inflow prevents deposition of a modern highstand mud wedge. The steep upper slope, between 300 and 500 m, is dominated by the MOW and shows flat-lying units, except for the occurrence of small-scale sand waves (0.5 m high and up to 10 m in wavelength) that prevail in the eastern sector where the current is stronger (Nelson et al., 1993). A gradual decrease in current intensity away from the Gibraltar Strait has resulted in the deposition of sand layers that are

G. Verdicchio and F. Trincardi

423

about 1 m thick and present a lateral gradation from fine, clean sand in the southeast to muddy silt in the northwest of the margin. The bent in the margin near the Spanish/Portuguese borderland induces a change in current regime (Ambar and Howe, 1979), resulting in the deposition of the large Faro Drift (e.g. Stow and Holbrook, 1984; Fauge`res et al., 1984).

20.3.5.

Mediterranean shallow-water contourite deposits

20.3.5.1. Physiographic and oceanographic setting of the Mediterranean The Mediterranean is a semi-enclosed marginal sea that communicates with the Atlantic Ocean through the narrow Strait of Gibraltar. It is a concentration basin where evaporation exceeds the combined effect of precipitation and river runoff. Its circulation is strongly influenced by water inflow from the Atlantic and by the presence of shallow sills, which define several sub-basins (Miller, 1972). The most important sill (400 m deep) extends between Sicily and the North African coast, dividing the Mediterranean into two main basins: the Western and the Eastern Mediterranean Basins. The modern Mediterranean circulation can be schematically defined as antiestuarine, with two secondary cells located in each of the Western and Eastern Basins (Lascaratos et al., 1999). The basin-scale circulation controls the exchange between the relatively fresh and light Mediterranean Surface Water (MSW), entering from the Atlantic, and the salty and relatively warm Levantine Intermediate Water (LIW; Wu¨st, 1961), which forms in the Eastern Mediterranean Basin (Levantine Basin). The LIW flows westward at a depth of 200–600 m and represents the main contribution to the Mediterranean Outflow (Pinardi and Masetti, 2000). Two secondary cells control the exchange between surface and intermediate waters with deep waters, namely in the Gulf of Lion (Western Mediterranean Deep Water, WMDW) and in the Adriatic Basin (Eastern Mediterranean Deep Water, EMDW), respectively. Most examples of Mediterranean contourite deposits described in the literature are located at shallow to intermediate water depths and appear associated with the along-margin flow of the LIW (depth range 200–600 m; Marani et al., 1993; Reeder et al., 2002). Commonly, these deposits appear similar in shape and internal geometry to their oceanic equivalents; however, in general, they are smaller in size, encompass a shorter interval of deposition and are commonly deposited in shallower water, as a direct consequence of the margin physiography and bottom-water circulation. We describe here examples of Mediterranean shallow-water contourites from the Corsica Basin (Roveri, 2002), the southwestern Adriatic Margin (Verdicchio et al., 2007) and the Gela Basin (Verdicchio and Trincardi, 2008b), and suggest the possible occurrence of similar deposits in the shallow Marmara Sea (Kuscu et al., 2002). 20.3.5.2. The Eastern Corsica Basin shallow contourite deposits The Corsica Basin is a shallow N-S trending basin separating the Corsica shelf (to the West) from the Elba Ridge (to the East), and connecting the deep Ligurian and Tyrrhenian Basins located to the north and south, respectively (Figure 20.1). The eastern side of the basin is affected by the LIW that flows northward at a depth

424

Shallow-Water Contourites

chan

Moat axis

Elba Ridge

Sigmoid drift

p62

Elba

Shelf edge

°40

Drift axis

42

′

°50

°00

42

nel

43°

00

10

N

′

between 200 and 600 m (Astraldi et al., 1990) and with an average velocity of 8 cm s 1, reaching winter peaks up to 35 cm s 1 (Artale and Gasparini, 1990). Contourite deposits in the eastern Corsica Basin show well-defined and elongated contour-parallel moats, on the upper slope in water between 300 and 500 m deep (Figure 20.7; Roveri, 2002). These deposits fit the definition of elongated mounded contourite drifts following the classification proposed by Fauge`res et al.

10

°00

–400

9°4

Erosion

0

Canyon

LIW 9°5

0.4

Canyon wall

3

0.5

P 69

0.7

Turbidite systems

P 68

1 2 3

NE Elba ridge

Buried drift

1.0

Y

twtt (s)

0.6

Multi-cresed drift

Corsica basin

twtt (s)

Elba ridge NE Shelf edge 0.2 LIW path Deeply eroded upper slope 0.3

1 2

0.9

18

0′ Southern sector

Northern sector

P 69

SW

p67

p69 42°4 0′ p68

0′

p71

43

p73

9°4

Normal fault

42°

°00

50′

Capraia sill

Shelf edge

X

4 5

Smooth upper slope

0.2 0.3

LIW path 0.5

Separated drift Subsidiary drifts

0.6

Corsica basin

twtt (s)

0.4

0.7 Turbidite systems 3

4

5

SW 0.9 1.0

0.8

X

2

twtt (s)

1

1 2 3 4

0.9

5

Figure 20.7 Corsica Channel contourite drifts (from Roveri et al., 2002; with permission from The Geological Society, London). (a) Location of the drifts, indicated in grey, relative to the Levantine Intermediate Water (LIW) flow (black arrows), and track lines of the seismic profiles. (b) High-resolution seismic profiles (1 kJ Sparker) showing the external shape and geometry of the sediment drifts.The location map is in the inset.

G. Verdicchio and F. Trincardi

425

(1999) and Stow et al. (2002c), and are essentially composed of highly bioturbated mud and silty mud. The size and morphology of these drifts reflect the interaction between the bottom-current pathway and the local margin morphology, characterized by a pronounced northward bend of the steep upper slope. The Corsica contourite deposits have actively grown since the middle Late Pleistocene, after the opening and progressive deepening of the Tyrrhenian Sea. Their continuous growth and relatively homogeneous character suggest that the large-scale LIW circulation pattern has not changed substantially since then. 20.3.5.3. The southwestern Adriatic Margin The SW Adriatic Margin (Figure 20.8) extends roughly N-S for about 150 km from a rather deep shelf break (typically 200–300 m deep) down to more than 1100 m. Since the Last Glacial lowstand, the SW Adriatic Margin has been affected by two distinct bottom-water masses: the steady LIW, flowing southward between 200 and 600 m deep, and the seasonal North Adriatic Dense Water (NAdDW), cascading obliquely to the slope at the end of the winter (Trincardi et al., 2007). Distinctive types of Late Quaternary fine-grained contourite drifts (elongated, plastered, isolated), up-slope migrating mud waves (occasionally showing bifurcated crests), contour-parallel moats, furrows, scours and comet marks reflect the impact of strong bottom currents along this margin, particularly during the last interglacial (Verdicchio and Trincardi, 2006). All these features occur at depth between 200 and 1100 m, predominantly above 600 m, where the margin is simultaneously affected by the contour-parallel steady LIW and the contouroblique seasonal-cascading NAdDW. Moreover, the morphology and the peculiar location of the contourite deposits suggest a strong interaction between margin morphology and bottom currents (Verdicchio et al., 2007). In the northern part of the margin, several basinward-convex scarps, aligned roughly N-S, define a sharp and irregular shelf break at a depth of around 300 m, in apparent continuity with the steep (>20°) erosional slope to the south. Basinward of each erosional scarp, a contour-parallel mounded drift pinches out against the upper slope, defining a shelf-edge moat. These shelf-edge contourite drifts display thick accumulations (up to 40 m) of low-consistency mud and silty mud, and commonly present, at their top, up-slope-migrating mud waves with wavelengths of 400 m and amplitudes of 15 m (Figure 20.8). 20.3.5.4. The Eastern Gela Basin, Sicily Channel The Eastern Gela Basin (Sicily Channel) stretches NW-SE, at a depth of 180–200 m, for about 90 km, along the steep western flank of Malta Plateau, which marks a Mesozoic to Cenozoic palaeogeographic boundary between basin sequences in the West and carbonate platforms in the East. Mass-transport deposits accumulated repeatedly along this unstable slope during the Late Quaternary and especially after the Last Glacial Maximum (Minisini et al., 2007). The circulation of water masses in the Sicily Channel is driven by water exchange between the Eastern and Western Mediterranean. The Modified Atlantic Water flows eastward in the upper layers, while the LIW flows to the west in the intermediate and

(a) 43

16°55′

41

39

37

35

11

13

15

17

19

42°N –271 m

W

Seafloor erosion

(b)

Sediment waves SA80

–253 m

E –420 m

SW

Seafloor erosion Sediment drift

SA125

NE –367 m

40 m 2 km

Figure 20.8 Contourite drifts of the SW Adriatic Basins (from Verdicchio and Trincardi, 2008b; with permission from Springer). (a) Location of SW Adriatic and Gela Basins with a schematic representation of the bottom-current circulation in each area. (b) High-resolution bathymetric map of a small part of the SE Adriatic upper slope showing crescent shape moats. (c) CHIRP sonar profiles showing contourite drifts down-slope of the moats.The depositional units, separated by extensive erosional surfaces, pinch out landwards, defining a moat area of prevailing erosion. Small-scale up-slope migrating sediment waves and isolated deposits with erosional basinward flank occasionally occur in close association with the moat area.

G. Verdicchio and F. Trincardi

427

deep layers. Between the Adventure Bank and the Malta Plateau (Figure 20.8), the LIW forms a pair of subsurface eddies, the southernmost of which affects the eastern Gela Basin, with a peak current velocity of 13 cm s 1 (Lermusiaux and Robinson, 2001). Shallow-water contourite drifts (Verdicchio and Trincardi, 2008b) occur at depths between 170 and 250 m in two main adjacent contexts: a complex upperslope region with a buried slide scarp and a simpler shelf-edge region with a prominent erosional step, not related to subsurface faulting or mass wasting. Elongated contour-parallel moats, up to 50 m deep and 700 m wide, extend for several kilometres seaward of the shelf break. Late Quaternary elongated drift deposits on the basinward flank of the moats reach thicknesses of over 50 m and are composed essentially of low-consistency mud deposits (Figure 20.9). The basinward flanks of these shallow-water contourite drifts frequently show evidence of thin-skinned mass failure, suggesting a strongly unstable slope. 20.3.5.5. The Marmara Sea contourite deposits The Marmara Sea is a small (11,000 km2), land-locked but relatively deep (>1000 m) E-W trending tectonic basin between the Black Sea and the Aegean Sea (Eastern Mediterranean; Figure 20.1). It is connected to the Black Sea and the Mediterranean by two shallow channels: the Bosphorus and the Dardanelles Channels, respectively. The morpho-physiographic setting of the Marmara Sea is strictly related to the activity of the northern segment of the North Anatolian fault, an active right-lateral strike-slip transform fault marking the boundary between the Anatolian and the Eurasian plates (Alpar and Yaltrak, 2002). Water circulation in the Marmara Sea is dominated by a permanent two-layer flow system: (1) a surface outflow from the Black Sea toward the Aegean, and (2) a reverse subsurface inflow from the Aegean toward the Black Sea (Besiktepe et al., 1994; Abrajano et al., 2002). On the basis of numerous sediment cores and grain-size analyses, Ergin and Bodur (1999) indicate that the sea-floor deposits of the Marmara Sea are mainly composed of fine-grained sediment with variable silt/clay ratios. The present authors suggest that such differences in sediment composition are related to the combined effect of basin morphology, sediment supply and bottom-current activity. Numerous highresolution seismic profiles were collected to study the recent activity of the North Anatolian Fault in the 50-km long Izmit Basin, the eastern sub-basin of the Marmara Sea that reaches depths of over 200 m (Kuscu et al., 2002). These data suggest the occurrence of possible shallow-water, fine-grained contourite drift deposits that are represented in this area by faintly stratified, acoustically transparent mounds with heights up to several tens of metres, rimmed by erosional moats (Kuscu et al., 2002).

20.3.6.

Shallow contourite deposits within sills, gateways and slope basins

Shallow sills and gateways control the interconnectivity between marine basins, allowing the exchange of water masses both at shallow levels and on the sea floor. When morphologically confined, the water masses that flow over the bottom tend to accelerate, potentially affecting the sea floor through selective erosion and deposition. The resulting sedimentary bodies may represent a distinctive end-member among the shallow-water contourite deposits. Examples of shallow gateways affected

428

Shallow-Water Contourites

(a)

(b)

Figure 20.9 Contourite drifts of the Gela Basin (modified from Verdicchio and Trincardi, 2008b; with permission from Springer). See location of the Gela Basin in Figure 20.8. (a) Highresolution bathymetric map of a small part of the eastern Gela Basin, showing a marked contourparallel moat along the shelf edge; the track lines of CHIRP profiles are also shown. (b) CHIRP sonar profiles showing upper-slope contourite drifts basinward of the erosional moat adjacent to the shelf edge. Contourite drifts display occasionally multiple crests and erosional down-slope flanks. Stacked thin-skinned mass-wasting deposits occur on the basinward flanks of the sediment drifts, suggesting that these deposits are prone to failure during, or soon after, their deposition. A multicolour version of this figure is on the enclosed CD-ROM.

429

G. Verdicchio and F. Trincardi

by contour-parallel currents include the Skagerrak Basin connecting the Baltic and the North Sea, active at present, and the Pelagosa Sill, connecting the Central and South Adriatic during the Last Glacial Maximum (Figure 20.1). The Skagerrak is a current-controlled sedimentary basin that reaches depth of over 700 m, and forms the westernmost part of a narrow strait allowing water-mass exchanges between the brackish Baltic Sea and the saline North Sea. The circulation defines a counter-clockwise path of deep North Sea water within the Jutland Current and superficial, and a more intermittent, hypo-saline water flow from the Baltic Sea (Nordberg, 1991). Current-induced bedforms, such as comet marks and large sand waves are widespread along the Skagerrak margins down to a water depth of 70 m. Sand waves reach amplitudes of 7 m and comet marks are up to 50 m long, demonstrating that maximum bottom-current velocities, associated with the Jutland Current, can locally reach at least 70–100 cm s 1 (Kuijpers et al., 1993). The Pelagosa Sill is a narrow passage (100 km long and 35 km wide) currently approximately 180 m deep. It represents the gateway between the central Adriatic and the South Adriatic that existed during the last glacial (Trincardi et al. 1996; Asioli et al. 2001). At that time, the sea level was some 120 lower than nowadays (Fairbanks, 1989), and the Pelagosa Sill was less than 60 m deep. Within the Pelagosa Sill, in water of approximately 170 m deep, small elongated contourite drifts occur; they are separated from the base of the slope by a contour-parallel erosional moat (Figure 20.10). twtt (ms)

SW

NE

Pelagosa Sill

225

Moat

ES 1

250

50 km

4300′

275

2 km

Adriatic Sea 4200′

Italy 1500′

1600′

1700′

Figure 20.10 CHIRP sonar profile across the Pelagosa Sill, showing an elongate moat parallel to the base of the sill flank in water of approximately 170 m deep. The location map, including the track line of the profile, is shown in the inset.This deposit was actively growing during the last glacial cycle (since approximately 140 ka; A. Asioli, 2005, IGG-CNR, personal communication) and particularly during the last glacial lowstand (above the line marked as ES1). A multicolour version of this figure is on the enclosed CD-ROM.

430

Shallow-Water Contourites

These deposits were actively growing during the last glacial cycle, since approximately 140 ka (Alessandra Asioli, 2005, IGG-CNR, personal communication) and particularly during the Last Glacial Maximum.

20.4.

D ISCUSSION

The deposits that record the activity of contour-parallel currents on the upper slope and, occasionally, on the mid-outer shelf and in shallow gateways, range from small-scale sandy bedforms to relatively extensive mud-dominated contourite drifts. Figure 20.11 provides schematically the main processes operating in shallow-water contourite settings and their associated deposits. The variety of settings, characterized by marked differences in the depth of the shelf edge, morphology of the shelf (landward dipping on glaciated margins) and steepness of the upper slope, hamper the adoption of a strictly depth-dependent definition of a shallow-water contourite. In most cases, especially when a continental margin displays a basinward dipping shelf with a marked shelf break at a depth of around 150–250 m, the term ‘‘shallow-water contourite’’ can be used for describing fine- or coarse-grained deposits formed by long-lasting bottom currents affecting outer shelves, upper slopes and shallow sills connecting adjacent sub-basins. Nevertheless, in some cases, such as in the case of

Gyre Advective mud wedge Moat Wind-driven bottom current

Comet marks

Sand dunes

Thermohaline-driven contour current

Furrows

Sand sheet

Paleo contourite deposit

Shelf-edge contourite drift

Intruding-shelf current

Internal waves

Moat Thin-skinned mass transport deposit

Figure 20.11 Schematic representation of the main processes operating in shallow-water contourite settings and their associated deposits.

G. Verdicchio and F. Trincardi

431

contourite deposits within the landward-dipping inner Antarctic shelves, reaching depths of several hundreds of metres, definitions such as inner-shelf contourites or shelf contourites are to be preferred.

20.4.1.

Shallow-water contourite deposits and sea-level change

The global ocean thermohaline circulation affects shallow-water contourite deposits just as their deep-water counterparts. However, the pattern of such circulation can change in response to Quaternary climatic and oceanographic turnovers and impact more effectively (and repeatedly) the distribution, growth patterns and preservation potential of shallow-water contourites than those in deep water. In particular, on the shallowest margin settings a sea-level fall can lead to a drastic reduction of accommodation space with consequent subaerial or shallow-water erosion of contourite deposits accumulated during the preceding highstand. In addition, a changing sea level may also lead to substantial variations in the intensity and pathway of bottom currents, resulting in the possible demise or even cannibalization of older contourite drifts. This case is possibly documented in the Corsica Channel (Marani et al., 1993; Roveri, 2002). All shallow-water contourite examples discussed above show a Late Quaternary age, confirming that the oceanographic conditions leading to their growth have lasted over shorter time intervals than in most deep-ocean examples (e.g. Flood and Shor, 1988; Kennard et al., 1990) or have been repeatedly interrupted by intervening sea-level changes.

20.4.2.

‘‘Extremely-shallow’’ contourite deposits

The concept of shallow-water contourite deposits can be extended to include also the broad category of those shelf clinoform deposits where sediment transport is controlled by long-lasting thermohaline currents that flow sub-parallel to the coast and to the strike of the progradational foreset, as seen in the subaqueous deltas of the Adriatic (Cattaneo et al., 2003), Amazon (Nittrouer et al., 1986) and Yellow Sea (Nittrouer et al., 1984; Liu et al., 2006). These deposits are characterized by the dominance of along-shelf advection (Cattaneo et al., 2007); the resulting sediment facies does not show straightforward evidence of a fluvial supply but appear as bioturbated mud similar to the facies observed in other shallow-water contourite deposits. Consequently, if encountered in outcrops, these elongated deposits that extend over hundreds of kilometres along a margin may be interpreted as a peculiar class of shallow muddy contourites.

20.4.3.

Slope stability of shallow-water contourite deposits

The slope stability of contourite deposits in general is discussed by Laberg and Camerlenghi (2008). Fine-grained shallow-water contourite deposits can accumulate on steep upper-slope areas at remarkably high rates (Verdicchio and Trincardi, 2008b), thereby resulting in the construction of potentially unstable deposits. While the stability of such shallow-water contourite deposits has rarely been studied on a quantitative geotechnical basis (e.g. Lee and Baraza, 1999), failure of these deposits

432

Shallow-Water Contourites

is documented on high-resolution seismic profiles both in the form of ‘‘thin skinned’’ failure affecting the steepest (seaward) flank of sediment drifts (Verdicchio and Trincardi, 2008b) and as a component of more complex mass-transport deposits that mobilize the entire contourite deposit and in some cases involve also the underlying units (Minisini et al., 2007). The occurrence of these unstable shallowwater contourite deposits at the shelf break, and therefore at least in some cases close to the coast, may have significant impact on the risk evaluation for coastal areas.

20.4.4.

Modern versus ancient shallow-water contourite deposits

On outcrop scale, contourite deposits are rarely unequivocally identified on the basis of facies analysis only; the recognition of laminated intervals as diagnostic of bottom-current activity is still under debate (Hu¨neke and Stow, 2008; Martı´nChivelet et al., 2008; Shanmugam, 2008; Stow and Fauge`res, 2008). However, even when recognized, the occurrence of such deposits should not be taken as a palaeobathymetric indicator, because shallow-water contourite deposits with substantially similar geometry and facies characters are being increasingly identified in almost any water depth on modern continental margins. Palaeogeographical reconstructions from outcrop studies of small uplifted tectonic basins, like those included in the circum-Mediterranean fold-and-thrust belts, suggest that such basins had a reduced extent, reflecting the tectonic fragmentation of the area, and such a configuration is expected to hamper the development of contourite deposits (Mutti, 1992). Small basins are likely of insufficient extent to develop a strong and persistent thermohaline circulation. However, the modern Mediterranean shows that a larger-scale thermohaline circulation can encompass structurally distinct margins, including rifted margins and foredeep domains, as long as sills and shallow passageways, and can allow an oceanographic connection among them. When incorporated in growing mountain chains, such differentiated regional domains will eventually become separated from each other, but this fate does not rule out the role of large-scale thermohaline circulation at the time of sediment deposition. In fact, a growing body of examples of contourites in ancient rock series, ranging in age from Cambro-Ordovician to Neogene, is increasingly being recognized, albeit with the caution needed in interpreting ancient deep-water successions (Hu¨neke and Stow, 2008).

20.5.

S UMMARY

Quaternary examples of contourite deposits from shelves, upper slopes and shallow gateways suggest the impossibility to adopt a strictly depth-dependent definition of shallow-water contourites. In several cases (as for example the contourite drifts in the deep Antarctic inner shelves), a more descriptive term such as ‘‘inner-shelf contourite’’ is to be preferred. On Quaternary continental margins, the key element in recognizing shallowwater contourite deposits is the large-scale morphology detected on high-resolution

G. Verdicchio and F. Trincardi

433

seismic profiles, swath bathymetry or side-scan sonar imagery. In some cases, additional diagnostic elements come from the recognition of accompanying smallscale erosional or depositional features, including erosional scours, furrows fields, comet marks and sand dunes. Compared with that of their deep-water counterparts, the growth of shallowwater contourite deposits typically encompasses shorter time spans, results in a more patchy distribution and is significantly faster (several millimetres per year against few millimetres per thousand years). The processes involved in the formation of shallow-water contourite deposits are more varied and less steady than in the case of deep-water contourite deposits. Depending on the latitude and dominating oceanographic regime, shallow-water contourites may reflect the activity of geostrophic currents characterized by rather uniform flow velocities (such as the LIWs in the case of the Mediterranean Basin) or by seasonally-modulated currents related to the morphologic trapping of dense waters generated by the segregation of salt during freezing processes around polar margins or by winter cooling on mid-latitude margins. In the latter cases, bottom currents do not last year-round but, nevertheless, are the most energetic process on the margin and can flow along the contour over prolonged intervals (several weeks at least). In ancient rock outcrops, shallow-water contourite deposits are not easy to identify. However, the review of their occurrences on modern continental margins hampers using these deposits to support palaeoenvironmental inferences regarding the water depth or the size of ancient basins. Indeed, facies and external geometry of such deposits are very similar, regardless of the water depth. Moreover, tectonic basins that may have been connected through sills or passageways shearing a powerful shallow thermohaline circulation at the time of sediment deposition appear small and disconnected when incorporated in fold-and-thrust belts.

ACKNOWLEDGEMENTS We thank A. Camerlenghi, A. Cattaneo and M. Rebesco for their reviews and suggestions. We acknowledge the financial support from the EU through HERMES (G0CE-CT200551112341), EUROSTRATAFORM (EVK3-CT2002-00079) and ‘‘EASSS III-TOBI side-scan sonar’’ (HPRICT199900047). This is ISMAR (CNR) – Bologna Contribution Number 1598.

C H A P T E R

2 1

M IXED T URBIDITE –C ONTOURITE S YSTEMS T. Mulder, J.-C. Fauge`res and E. Gonthier De´partement de Ge´ologie et Oce´anographie, Universite´ Bordeaux1, Talence cedex, France

Contents 21.1. Introduction 21.2. Contourite and Turbidite Alternation 21.2.1. Low-frequency alternation: morphological heritage 21.2.2. High-frequency alternations of contourites and turbidites 21.3. Redistribution of Gravity Deposits by Contour Currents 21.4. Interaction of Synchronous Contour and Turbidity Currents 21.5. Conclusions Acknowledgements

21.1.

435 436 436 437 442 446 454 456

INTRODUCTION

Down-slope gravity currents and along-slope contour currents are common phenomena along continental margins. Bottom currents are related to thermohaline and geostrophic circulation and occur along most margins, with variable intensity. Gravity currents occur regularly in areas where sediment load is important (seaward of river mouths, where large deep-sea turbidite systems can develop) and more sporadically in places where no major effluent exists (along a depositional ramp). These two phenomena have very different physical characteristics. The difference concerns mainly parameters such as flow energy, competency (i.e. the largest particle the flow is able to transport) and duration. Turbidity currents (i.e. currents in which the particle support mechanism is dominantly the turbulence: Middleton and Hampton, 1973) are energetic unsteady processes with velocities usually ranging from a few decimetres to a few metres or a few tens of metres per second. The most common process, the short-duration turbulent surge, is shortly waxing (velocity increasing with time) and then waning (velocity decreasing with time). Their duration from triggering to final deposition does not exceed a few days. Other turbulent flows such as long-duration turbidity currents sustained by retrogressive failures are less unsteady. Long-duration hyperpycnal turbidity currents generated at river mouths during floods can be considered as quasi-steady flows. They can last a few days to a few weeks. Consequently, flow competency is Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00221-5

Ó 2008 Elsevier B.V. All rights reserved.

435

436

Mixed Turbidite–Contourite Systems

high and sand can be transported over long distances. By contrast, contour currents are permanent steady flows with velocities usually in the range of few tens of centimetres per second. The temporal range of velocity changes is from seasonal to thousands of years. Competency is low. Although contour currents can transport sand, most contourites are made of silty and muddy deposits. True interactions between contour and gravity currents depend on the presence of both currents at the same place and at the same time with a balanced energy level. This is relatively rare. Most of the times, gravity and contour processes act in adjacent areas or during successive periods. When they act at the same time but in neighbouring areas, true interaction can occur at the boundary between these areas. There, interaction can imprint the sedimentary series either at the sedimentary facies level (rare) or at the sedimentary body level (frequent). When the two processes act during successive times, interaction can be recorded at low frequency mainly by variations of the sediment-body morphology (deposit distribution controlled by inherited morphologies) or at high frequency by the alternation of contourites and gravity deposits. In addition, contour currents can redistribute gravity supply or simply rework gravity deposits and construct downward any sedimentary drift bodies. This suggests that true interactions between gravity and contour processes are difficult to evidence, particularly at the sedimentary facies level. This is probably due both to the paucity of true records of interaction at the sedimentary facies level and to the difficulty to recognize the subtle differences between contourite and turbidite sedimentary facies when they are interbedded in a single sedimentary succession. The rarity is due mainly to the difference between the velocities of contour currents and gravity processes. If the difference is too large, the most energetic process, usually the down-slope turbidity current, will dominate and be recorded. In this chapter we will illustrate the following points: • • • •

low-frequency contourite and turbidite alternations (morphologic heritage); high-frequency contourite and turbidite alternations; redistribution of gravity deposits by contour currents; interaction of synchronous contour and turbidity currents.

21.2. 21.2.1.

C ONTOURITE AND T URBIDITE ALTERNATION

Low-frequency alternation: morphological heritage

A clear example of the impact of inherited morphology, the occurrence of a turbiditic canyon on the circulation path of a contour current system, is provided by the northern part of the Gulf of Cadiz (Figure 21.1; see also Herna´ndez-Molina, 2006). At this location, the Mediterranean Outflow Water (MOW) has a relatively high energy. Its velocity does not exceed 30–40 cm s 1. The upper core of the MOW is channelled along the coast. It forms a classical separated drift (Faro-Albufeira Drift) separated from the margin by the Alvarez Cabral Moat. Silty deposits dominate the drift. Conductivity– temperature–depth (CTD) measurements at the bottom of the Portimao Canyon, downstream the drift, show that the MOW is partly trapped and flows down-slope

437

T. Mulder et al.

S

N

Sp038

S

N

Sp036

Plio-Pleistocene unconformity

PC ACM

Sp

N

FAD MOW circulation

8 03 Sp

Plio-Pleistocene unconformity

LD

6

03

Figure 21.1 Multi-beam (EM 300) data showing the northern part of the Gulf of Cadiz and the Mediterranean Outflow Water (MOW) capture by the Portimao Canyon (PC) (from Marches, 2005; with permission). East of the canyon, a thick detached drift forms, the FaroAlbufeira Drift (FAD), separated from the margin by the Alvarez-Cabral Moat Channel (ACM). The sharp decrease of the MOW intensity and competency west of the canyon generates the formation of the thin plastered Lagos Drift (LD). Sparker profiles Sp036 and Sp038 illustrate the change in drift morphology due to capture of the MOW by the Portimao Canyon. A multicolour version of this figure is on the enclosed CD-ROM.

along the canyon floor (Marche`s et al., 2007). Consequently, the intensity of the MOW branch crossing the canyon is drastically reduced on the western side of the Canyon. Maximum transported grain size and sedimentation rate on this side are smaller than on the eastern side, suggesting that both competency and capacity of the MOW are reduced. Consequently, only a flat plastered drift (Lagos Drift) forms on the western side. The MOW velocity (about 20 cm s–1) is too low to form a moat channel. The control of the distribution of contourite deposits by the morphology induced by turbidity currents is also illustrated on the Late-Glacial/Holocene deposits on the Barra Fan (Knutz et al., 2002b). A small drift developed because of the diversion of the North Atlantic Deep Water (NADW) flow by the topographic high made of debrites on the fan. Local increase of NADW velocity along the scarps at the top of the debris flows might be at the origin of sandy contourite deposition on the mid-upper slope of the Barra Fan (Armishaw et al., 1998, 2000). Examples exist also of gravity systems setting on a previous morphology generated by contour current activity. A first example comes from the middle US Atlantic margin, east off Cape Hatteras (e.g. Tucholke and Mountain, 1986; McMaster et al., 1989; Locker and Laine, 1992), where down-slope and along-slope processes have been interacting since the Oligocene. From the Late Miocene to the Late Pliocene, the continental rise was subjected to simultaneous erosion and deposition by

438

Mixed Turbidite–Contourite Systems

both gravity flows and the Western Boundary Undercurrent (WBUC). The spatial alternation of sea floor under erosion and under deposition was very tight. That is particularly true for the upper and middle continental rise, where contourite drift deposits (Chesapeake Drift) seem to have partly controlled the location of the channels and preserved the turbidite record. Deposition on the lower part of the continental rise has been predominantly under the control of contour-current processes. Sedimentation by bottom currents in an area of high down-slope sediment supply generated the rapid development and high relief of the Hatteras Outer Ridge. This separated drift was bounded by an elongated depression on the western backside of the drift and covered by a field of sediment waves on its eastern flank. During the Late Pliocene and the Quaternary, down-slope processes became predominant as a result of global cooling and the development of large glacial ice caps in the sub-arctic areas. Both the deep-sea turbidite channel courses and sediment distribution are believed to be controlled by the pre-existing drift topography. Turbidites fill the depression on the backside of the Hatteras Drift, which disappears beneath the onlapping deposits of ponded turbidites. The seaward flank of the drift presents still active sediment waves, but small turbidite ponds form in the troughs of the sediment waves because turbidity currents may overflow or breach the crest of the drift. Some turbiditic transverse channels cross the whole drift and transport sediment to the abyssal plain (Figure 21.2). As a conclusion, deposit distribution controlled by previous sedimentary topography (heritage) occurs at a margin scale. Most often, it results from drastic change in the sedimentary processes of depositional environments.

21.2.2.

High-frequency alternations of contourites and turbidites

Intercalation of deposits from gravity and contour currents occurs in the sedimentary record when the two processes alternate in time, and if both have sufficient energy to transport sediment from a source to the deep basin. The intercalation of turbidites and contourites has been demonstrated at the scale of the process (high-frequency alternation), i.e. by the intercalation of individual beds in outcrops (Stow et al., 1998a) or by alternation of bed packets in seismic sections (low-frequency alternations; e.g. Fauge`res et al., 1999). Intercalated deposits in a proximal deep-sea turbidite system (canyon) are described from the South Brazilian Margin by Souza Cruz (1995), Viana (1998) and Viana et al. (1999) (Figure 21.3). In seismic-reflection profiles, deposits within the canyon show the alternation of clinoforms prograding towards the thalweg and high-reflectivity aggrading deposits that drape the canyon floor. Clinoforms are interpreted as resulting from deposition by contour currents during phases when along-slope current activity (flowing perpendicular to the canyon direction) intensified. By contrast, aggrading deposits are interpreted as the classical turbidites that drape the canyon bottom and that were deposited when down-slope gravity processes (flowing down the canyon) intensified. Other intercalated contourites and turbidites occur in the channel–levee complex of the mid deep-sea fan system in the southern Weddell Sea (Michels et al., 2002). The Weddell Sea is the main area for the formation of the Antarctic Bottom Water (AABW) (Carmack and Foster, 1977), which flows towards the Southern Atlantic at a depth of over 1500 m. The southeastern Weddell Sea margin is incised

439

T. Mulder et al.

35°36 N

Current direction

USA

Seismic profile 35°25 70°10W

NW

9°54W

Hatteras Drift Crest

Turbidites Contourites

Migrating or less well defined sediment waves

SE

twtt (s)

6

8 0

5 km

Turbidites and contourites

Figure 21.2 Pliocene^ Quaternary deposits on the middle US Atlantic margin, east off Cape Hatteras (after Locker and Laine, 1992; with permission). See top map for location. Multi-channel seismic profile showing filling of the depression between contouritic dunes by ponded turbidites.

by numerous gullies, channels and small canyons. The most important terrigeneous deep-sea system is the Crary Trough-mouth fan that formed seaward of the Crary glacial trough. The proximal fan extends in S-N direction, while the distal fan turns to a NW-SE direction, parallel to the North Antarctic Margin. Terrigeneous material is supplied to the system by both channelled turbidity currents and plumes generated by plunging Ice Shelf Water (ISW). Because of the action of the strong Weddell Gyre, and because the Coriolis force is stronger in high-latitude regions, particulate flows are deflected towards the northwestern (left) side of the Crary Channel. Overspilling generates an asymmetric channel–levee complex with a hypertrophied left levee formed by combined deposition from contour and turbidity currents. The right levee is very small due to sediment starvation on this side. Contour-current action tends to increase the lateral extent of the levee and to emphasize its wedge shape (Figure 21.4). In the Weddell Sea, intercalation of

Seismic profile

1.0

N

S

1.0

Canyon

1.5 twtt (s)

1.5

0

1 km

bc

S 1.0

N c

tc 4 c

twtt (s)

1.5

3 2 1

4

c

c

3

t t

2

c t t

c t

1 0

1 km

Figure 21.3 Multi-channel airgun seismic profile (and interpreted profile below) on the southeastern Brazilian margin (see top map for location) showing the alternation of contourites (c) forming clinoforms prograding towards the canyon axis and turbidites (t) forming high-amplitude reflectors draping the canyon bottom (after Viana et al., 1999; with permission). See top map for location. tc: turbidity current; bc: bottom current; 1^4 major reflectors on the margin (3: Middle ^ Late Miocene reflector).

441

T. Mulder et al.

Seismic profile

5

NW

SE

20 km

twtt (s)

6

NW 5

SE Levee Channel

Old eastern levee

Old channel

w5

twtt (s)

6

Eastern levee

Channel deposits w4

Preglacial sediments

Figure 21.4 Multi-channel seismic profile (see top map for location) across a mixed contourite/turbidite channel ^ levee complex on the southeastern side of the Weddell Sea at approximately 4000 m water depth (after Michels et al., 2002);W4,W5: unconformities.

442

Mixed Turbidite–Contourite Systems

contourites and turbidites is mainly due to the alternation of glacials and interglacials during the Quaternary. During the sea-level lowstand of glacials, the sedimentation rate on the Antarctic margins increased because of the input of basal glacial debris carried by the ice sheet grounded near the shelf edge (Weber et al., 1994). When the grounding line was close to the shelf break, interstitial pore pressure generated by ice loading and rapid accumulation of diamicton on the slope favoured slope failures on the upper slope and mass-wasting (debrites and turbidites) deposits down-slope. Gravitational processes were more important than ISW formation, and turbidity currents were the dominant processes in the building of the Crary levees. In contrast, the decrease of both frequency and intensity of turbidity currents during interglacials and related sea-level highstands allowed the predominance of contourite deposition in the corresponding levee deposits. A similar intercalation of turbidites and contourites is observed in the more distal depositional environment of the lobe complex off Cape Hatteras, on the eastern US margin (Tucholke and Mountain, 1986; McMaster et al., 1989; Locker and Laine, 1992; see also the previous section). In seismic profiles, layered, moderate-amplitude, gently dipping reflectors interpreted as contourites alternate with high-amplitude aggrading, sometimes draping or basin-fill reflectors interpreted as turbidites. These deposits are the prototype of the companion drift fan concept (Locker and Laine, 1992). Finally, the alternations of turbidites and contourites of the Late Eocene– Neogene Series along the Rockall Margin off NW Britain (Howe et al., 1994; Stoker, 1995, 1998b; Stoker et al., 1998a) are worth to be mentioned here. In conclusion, alternation of contourites and turbidites is frequent in the sedimentary record of continental margins and can occur at different scales, from a single decimetreto-metre-scale to large and thick sedimentary bodies. In the latter case, the surface morphology of older deposits can affect the deposition of more recent deposits.

21.3.

R EDISTRIBUTION OF GRAVITY D EPOSITS BY C ONTOUR C URRENTS

Sediment reworking and redistribution by contour-current processes are common in the deep-sea environment. As deep currents act almost everywhere in the world ocean, fine-grained pelagic and hemipelagic sediments may be reworked to various degrees by bottom currents. In that sense, a significant part of the deep-sea deposits can be interpreted as contourites (Fauge`res and Stow, 1993). Reworking and redistribution of gravity deposits by contour currents occurs both when the two processes alternate with time and when they are synchronous. Gravity currents carry most of the terrigeneous material to deep depositional areas, where the particles are swept by contour currents of sufficient energy to induce the reworking of deposits. Conversely, reworking of contourites by gravity processes, although common, cannot be easily observed, as it is mainly recorded as erosional features. As stated in the previous section, deposits from both processes may be intercalated in the sedimentary record. Reworking of deposits (including contourites) by gravity

443

T. Mulder et al.

currents tends to concentrate sediments by filling sea-floor depressions, thus increasing locally the thickness of the sedimentary pile. Deposition can be associated with erosional surfaces of limited extent (channel erosion). In contrast, the action of contour currents is marked by intensified erosion at a margin-wide scale and widespread erosional surfaces in the sedimentary record (Fauge`res et al., 1999). The particles reworked by a contour current may be transported over a very long distance before they settle to build sediment drifts downstream from the area of reworking. The Faroe–Shetland Channel acts as a conduit for southward near-bottom flow of cold Norwegian Sea Overflow Water (NSOW), along the West Shetland shelf and the Wyville-Thomson Ridge (Akhurst et al., 2002). Glacigenic debris flows on the upper continental slope are intercalated by minor turbidites and important sheeted contourite drifts (Figure 21.5). Core analysis suggests that contourites originate from the reworking of the debris-flows. Another example is provided again in the Gulf of Cadiz (Hanquiez, 2006) (Figure 21.6). At this location, small channels form within large MOW channels N 62° 0 km 30

Seismic profile N 60° 7°W

2°W

1 km

SE

100 ms

NW

Figure 21.5 Sparker seismic section across the Faroe ^Shetland Channel (see top map for location) showing the contourite sheet-drift geometry (after Akhurst et al., 2002; with permission). Erosion and bypass resulting from active contour-current activity thins the central section of the channel. Presence of lenticular debris flows thickens the basinal section.

9°W

8°W

7°W

6°W

5°W

(a) Huelva

PORTUGAL

N

Faro

SPAIN

37° 50 0

Cadix MU

36° 100 0

2000

3000

ATLANTIC OCEAN

Gulf of Cadix

NADW North-Atlantic Deep Water NASW North-Atlantic Surficial Water 7°50

(b)

Atlantic inflow water

ML Tanger

Morocco

Mediterranean outflow water (MOW) MU:Mediterranean Upper Water ML:Mediterranean Lower Water 7°00

36°00

35°50

100

10 km N

(c)

7°35′W

1000 2000 m

7°25′W

Channel

35° 45′

Lobes

35° 45′

Figure 21.6 Small channels and lobes in the Gulf of Cadiz (after Hanquiez, 2006; with permission). See top map (a) for location. Swath bathymetry map and EM 300 reflectivity map (b). These small turbidite systems initiated by failures at the front of contourite deposits are nested within large MOWchannels. The permanent MOW flow winnows the fine particles in lobe deposits. A multicolour version of this figure is on the enclosed CD-ROM.

T. Mulder et al.

445

and end with small sandy lobes with a total thickness that does not exceed 10 m. Although these channel-lobe systems are initiated by small-sized gravity flows induced by submarine failures on the front of a huge sedimentary wedge (a giant contouritic levee; Mulder et al., 2003b), the transported sediments are probably winnowed by the MOW flow both during and after deposition, and fine particles are transported down-current. The resulting lobe is composed of massive, clean, coarse silts to fine sands. Lag-deposits and hardgrounds indicating sediment bypassing can be observed in sea-floor imagery and photography on the Barra Fan (Armishaw et al., 1998; Stow et al., 2002a), where the concept of a composite slope-front fan was defined. Along the outer continental shelf and the upper slope (in water depths ranging from 140 to 300 m), the sea floor is formed by a mixture of pebbles, cobbles, boulder, gravel and shelf detritus (Figure 21.7). Because the present hydrodynamics of bottom currents cannot explain these large clasts on the sea floor, it is thought that these coarse sediments, supplied close to the shelf edge by glacial, fluvial and coastal processes, are the relict of Pleistocene sea-level lowstand depositional environments that were more energetic than today. Sediment waves with their crest oriented perpendicular to the isobaths suggest that the gravels have been reworked by shallow-water contour currents. Similar sands and gravels are also found on the sea floor at the shelf/slope transition on the US Atlantic margin (Stanley et al., 1981; Blake and Doyle, 1983) and the Scandinavian margin (Kuijpers et al., 1993; Yoon and Chough, 1993). The Sicilian gateway, located between Tunisia and Sicily (Central Mediterranean Sea), is a depression with a lateral supply by turbidity currents originating from both the Sicilian and Tunisian margins. Slope instability is mainly triggered by frequent volcanic and seismic activity. The action of the E-W-directed Levantine Intermediate Water within the gateway is responsible for reworking down-slope gravity-flow deposits (Reeder et al., 2002).

Figure 21.7 Sea-floor image showing contour-current-generated sand waves and gravel deposits resulting from redeposition by contour currents in the Barra Fan area (from Stow et al., 2002a, with permission).

446

Mixed Turbidite–Contourite Systems

To conclude, redistribution of turbiditic sediments by bottom currents can be recognized mainly at the sedimentary series scale (basin scale), while intercalation between contourites and turbidites occurs at scales from the individual bed to the sedimentary series.

21.4.

INTERACTION OF S YNCHRONOUS C ONTOUR AND T URBIDITY C URRENTS

True contour/turbidity current interaction can occur (1) over large areas if the energy of both currents is balanced; (2) at the boundary between two depositional environments if contour and gravity currents are unbalanced and act separately in adjacent areas. At a regional scale, interaction between contour and gravity currents leads to the deposition of asymmetric sedimentary bodies with preferential extension in the contour-current direction. The fan drift concept (Carter and McCave, 1994) illustrates this interaction. The fan drift has been defined in the Hikurangi Plateau area along the eastern continental margin of New Zealand (Carter and Mitchell, 1987; Carter and Carter, 1988; Carter et al., 1990; Carter and McCave, 1994; Lewis, 1994; Wood and Davy, 1994; McCave and Carter, 1997). Terrigenous sediments are transported via the Hikurangi Channel, which is located at the northern side of the Chatham Rise. At the channel mouth, a thick sedimentary accumulation forms. It includes particles brought by the turbidity currents, particles eroded from the Bounty Fan (south of the Chatham Rise) by the Pacific Deep Western Boundary Current (PDWBC) and, to the lesser extent, volcaniclastic and pelagic particles. At the Hikurangi Channel mouth, this material is rapidly deposited but the lobe is deflected eastward under the action of the PDWBC at 4850–5500 m water depth. It forms the 300-km long Hikurangi Fan Drift (Figure 21.8). Some 400 m of sediment accumulated on the drift since the channel inception in the Pleistocene (Carter and McCave, 2002). Pirating of material supplied by turbidity currents by contour currents is also demonstrated on the Barra Fan (Knutz et al., 2002b), where sediment waves generated by low-density glacigenic turbidity currents gradually fade away from the fan. When the energy of a contour current is sufficient to pirate only a part of the finest suspension of a turbidity current but is not high enough to deflect the whole depositional lobe to form a fan drift (as it happens at Hikurangi), the lobe of the deep-sea fan persists and a sediment drift forms on its side. This is what happened during the mid Miocene to Pleistocene glacial maxima at least for 1 of the 12 contourite mounds identified along the Pacific Margin of the Antarctic Peninsula by Rebesco et al. (2002) (Figure 21.9a). The steep wall of the channel separating the mounds is on the right side and is affected by important mass wasting. The mounds cannot be turbidite levees because in the southern hemisphere preferential deposition and erosion should occur on the left-hand side of the channel. Consequently, the largest mounds are interpreted as contourite drifts (Rebesco

447

T. Mulder et al.

Seismic profile

W

E

Turbidite plain Hikurangi Channel

Rehkohu Drift

twtt (s)

5

6

0

50 km

Figure 21.8 Airgun seismic profile in the Rekohu Drift (from Wood and Davy, 1994, with permission). See top map for location. It is a large depositional area extending on the left side of the southern distributary of the Hikurangi Channel.

et al., 1996). Sediment supply at the base of the continental slope is generated by failure generated by the grounded ice sheet. The frequent down-slope-channelled turbidity currents coming from the Antarctic margin are pirated by the NE-SW flowing modified Weddell Sea Deep Water (WSDW). Fine-grained particles settle down after contour-current deflection to the left side of the flow and form detached drifts on the left side of the channels (Figure 21.9b). A similar interpretation is made for contourite mounds observed on the northeastern margin of Antarctica (Wikes land, Adelie Coast: Escutia et al., 2002)

448

(b)

Mixed Turbidite–Contourite Systems

80°W

75°W

NW

SW-flowing bottom current

Seismic profile

67°S

SE

Small-scale slump Grounded ice sheet

Suspended load Turbidity current

Continental shelf

r we Lo

Continental slope

Continental rise Continental rise

(a)

Sediment drift

SW

N Mound Channel

twtt (s)

Channel 5 M-M.h.

6

Oceanic Basement

7

0 km 10

SW

N

Channel Turbidity current

ACC Migration

1

twtt (s)

5 6 7

Channel Turbidity current

2 M-M.h.

3

Oceanic Basement

Figure 21.9 Sediment drift 7 in the Antarctica Peninsula (see top map for location). (a) Multichannel seismic profile and interpretation (after Rebesco et al., 1996; with permission). M-M.h.: Middle Miocene hiatus. 1. Drift maintenance stage; 2. Drift growth stage;3. Pre-drift stage; ACC: Antarctic Contour Current. (b) Schematic depositional processes inferred for the construction of a drift by pirating the glacigenic supply by ACC (from Rebesco et al., 2002; with permission).

(Figure 21.10). Up to 490 m thick mounds are built on the left side of channels. These mounds are built both by deposition from turbidity currents (levees) and westward flowing contour currents. The period of maximum mound construction corresponds to the time of expansion of the East Antarctic ice-sheet. The important glacial erosion on the continent supplied a large amount of unsorted glacial sediment that was transported down-canyon by turbidity currents. The suspended material was pirated by the westward flowing AABW and supplied the mixed turbidite/

449

T. Mulder et al.

150°E

135°E

64°S

Seismic profile

Dibble Ice Tongue

(a)

(b)

W

Mound

Mound

Channel

Channel

Jussieu

Levee

W Channel

Adelie Coast

Wilkes Land

68°S

Mertz Glacier Ninnis Glacier

E

Jussieu Channel

Levee

Mound

(c)

Mound

Buffon Channel

Buffon E Channel

(d)

W

Crest

E

W

E

Figure 21.10 Continental rise of Wilkes Land margin (after Escutia et al., 2002; with permission). See top map for location. (a) Seismic profile showing channels, levee and sediment mounds. (b) Detailed section of (a), showing two channels ( Jussieu and Buffon), the east levee of the Jussieu Channel and a sediment mound. (c, d) Details of the sediment mound deposit and its interpretation. Ch = channels; L = levees; SW = sediment waves. WL1b, WL1c and WL2 are unconformities. A multicolour version of this figure is on the enclosed CD-ROM.

contourite mounds. Recent construction of the mounds decreased because of icesheet retreat and trapping of the terrigenous input on the continental shelf. At the sedimentary-facies scale, the identification of sedimentary structures indicating interaction of turbidity and contour currents is still under debate. In particular, the existence of a continuum between turbidites and contourites is still

450

Mixed Turbidite–Contourite Systems

an open problem (see the concept of ‘‘modified turbidites’’ of Stanley, 1993). Lee and Ogawa (1998) described an equivocal top-winnowed and reworked turbidite in pumiceous beds in Miocene contourites of the Isu forearc in Japan. More convincingly, Ito (2002) described transitional facies from turbidites to contourites in a Plio-Pleistocene formation in the area of the Kazusa forearc (Figure 21.11). The active current in this area is the powerful Kuroshio with current velocities up to – and sometimes more than – 180 cm s 1 (Taft, 1978). The transitional sediments between contourites and turbidites show the following diagnostic criteria of interaction:

Lateral variation

Te

Gradational, locally

Td sharp upper contacts,

B-1

Bouma model

A Type

B-2

B-3

C-2

C-3

Tc A

Tb Ta

C-1

Better sorted than lower ripple crosslaminated sand

B Type

B-1

Inversely graded ripple crosslaminated sand at the top

B-2

Inversely graded ripple crosslaminated sand

B-3 Ripple crosslaminated lenticular sand

Erosional or sharp upper contacts

C Type Mainly starved rippled sand

C-1

C-2

C-3

Figure 21.11 Types of turbidite-to-contourite continuums derived from a Plio-Pleistocene outcrop in the Kazusa forearc basin in Japan (from Ito, 2002; with permission). Vertical bars indicate reworking by bottom currents. The succession thickness varies from 1 to 10 cm. Type A: 1^100 cm thick turbidites (Bouma, 1962). Type B: sharp upper contacts beneath nongraded siltstone (B-1). Current directions are different for ripple-cross-lamination and for basal erosional structures. Upper parallel laminae and graded siltstone are lacking. Locally, better-sorted ripple-cross-laminated sandstone with inverse grading can appear (B-2 and B-3). Type C: beds with lack of upper parallel laminae and graded siltstone. Sharp upper contacts are associated with non-graded siltstone (C-1). Laterally, this sharp contact separates the sandstone bed from a better-sorted ripple-cross-laminated very fine to very coarse sandstone (C-2). Locally, lenticular beds of ripple-cross-laminated very fine sandstone or very coarse siltstone with sharp basal and upper contacts are interbedded between ungraded siltstone (C-3).

T. Mulder et al.

451

• better sorting than classical contourites (Stanley, 1993); • changes in the vertical facies evolution that differ from the classical Bouma sequences; • both traction structures (typical of turbidites) such as parallel and crosslamination, and • crude inverse grading (generally encountered in contourites); • internal erosion surfaces overlain by mud drapes or flaser lenticular bedding (this is interpreted as an oscillating flow resulting from the interaction of two almost perpendicular flows, i.e. down- and along-slope, with a similar energy; erosion surfaces and mud drapes result from flow fluctuations with time); • variable current directions (measurements on facies with dipping laminae indicate predominantly along-slope transport: Stanley, 1993); • lateral variation of both facies and current direction along correlated beds (Figure 21.11); • heavy-mineral concentrations suggested as a diagnostic feature for sandy contourites are difficult to interpret as they can be due both to combined events and to concentration by multi-stage processes. Interaction between contour and turbidity currents is also expected along the continental rise of the Northeastern Brazilian Margin by Gomes and Viana (2002). It is suggested by the presence of a large sediment drift (Pernambuco Drift) generated under the action of the strong N-S AABW flow and an important drainage system on the continental slope, including canyons and tributary channels related to the Sa˜o Francisco Deep-Sea Turbidite System. The strength of the AABW there is due to flow constriction between the Bahia Seamounts and in the N-S Pernambuco Channel. The system corresponds to the concept of modified drift–turbidite system (Gomes and Viana, 2002). The Columbian fan drift in the South Brazilian Basin shows clearly the evolution of the interaction between down-slope and along-slope processes following the decreasing energy of the former (Fauge`res et al., 2002a). The change is recorded aside of the Columbia Channel, which is a deep valley extending in water depths from 4200 to 5200 m. This area is also affected by the AABW that flows northwards at a water depth of over 4500 m and that is deflected eastward by the Vitoria– Trindade Seamount. In the upper part of the Columbia Channel, turbidity currents flowing down-channel have sufficient energy to prevent any impact of the contour current in the sedimentary record. A turbiditic levee forms on the left-hand, proximal side of the channel (Figure 21.12a). In the lower part of the same channel, the energy of the turbidity current decreases sufficiently to allow AABW activity to imprint sediments. The turbidite levee is replaced by a contourite levee, still forming on the left (north) hand side of the channel (Figure 21.12b). Cores collected at the junction between the turbidite and the contourite levee show clear evidence of concomitant contour- and turbidity-current activity in the deposition of some sedimentary successions. The base of these successions shows an erosional surface with thin layers or lenses of silt or silty mud interpreted as fine-grained turbidites. This turbiditic unit is capped by a sharp contact overlain by bioturbated muds (Figure 21.13). This suggests that the turbidite unit

452

Mixed Turbidite–Contourite Systems

(a) N

S Columbia levee

Columbia channel

Turbidite levee

Columbia levee

Columbia channel

Turbidite levee

tw tt ( s)

0

2

S

N 0

SU2 NU2

SU1b R1

Ch2

SU1a

Ch1

R1

R2?

tw tt ( s)

R2

NU1

2 Substratum 0

(b)

NE

SW Contourite levee

Columbia channel

Turbidite levee

1

SW Columbia

Turbidite flank Contourite levee channel

2

Contourite levee

NE

channel

SU2′ R2 SU1′

R2′ R1

0

NU2′ NU1′ 1

0

2

twtt (s)

Contourite levee

twtt (s)

0

Figure 21.12 Water-gun seismic profiles and interpretations across the Columbia Channel (from Fauge'res et al., 2002a; with permission). R1: erosional discontinuity; R2: erosional, flat and horizontal surface; SU1: Lower South Unit 1; (SU1a: base; SU1b: Top); SU2: Upper South Unit; NU1: North Unit 1; NU2: North Unit 2. (a) Upper part showing the proximal turbiditic levee formed on the left-hand side of the channel. (b) Lower part showing the distal contouritic levee formed on the left-hand side of the channel under the influence of the S-N Antarctic BottomWater (AABW).

453

T. Mulder et al.

Bioturbated mud (contourites) Gradual transition

550

Homogeneous mud Gradual transition Interbedded homogeneous muds and thin silty layers

555

Bioturbated mud (contourites)

450

Gradual transition Sharp contact

Homogeneous mud

560

Bioturbated fining-upwards laminated silt

455

Alternating millimetric laminae of silt and mud Sharp contact

Erosional contact

Bioturbated mud (contourites)

Gradual transition Fining-upward laminated silt

565

460 cm

Erosional contact Bioturbated mud (contourites)

cm

Type I

Type II

Figure 21.13 Interpretation of X-radiographs showing details of contourite/turbidite interaction at layer scale (from Fauge' res et al., 2002a; with permission). Water depth of core 4515 m. Type 1: silty/muddy turbidites with an erosional basal contact gradually passing into homogeneous non-bioturbated mud. At the top, these muds are overlain by bioturbated mud. The non-bioturbated mud suggests rapid deposition and is interpreted as the upper part of a silty/muddy turbidite. The bioturbated top muds are interpreted as muddy contourites with low deposition rate. Type 2 is a top-truncated silty/muddy turbidite with a thick lenticular basal part above an erosional contact. The turbidite is topped by alternating millimetre-scale silty and muddy laminae. The top is overlain by bioturbated muddy contourites via a sharp contact.

is top-truncated. The contour current had sufficient energy to prevent the deposition of the waning low-energy back part of the current. The material was finally deposited as a classical muddy contourite. Based on Oligocene outcrops in Cyprus, Stow et al. (2002c) suggest that mixed contourite/turbidite deposits could have been relatively frequent on the margin of the closing Tethys Ocean. Down-slope gravity processes were probably as frequent on Tethyian margins as they are today along present-day ocean margins. However, the constriction of the Tethys Ocean during the Late Paleogene due to the northward motion of Africa led to an intensification of the westward Tethys Current.

454

Mixed Turbidite–Contourite Systems

This current might have been locally increased also by a distinct tectonic relief generated by subduction, just like that present now along the Japanese Pacific margins. Thus, the Tethys Current could have been strong enough to imprint the top of turbidite successions. In summary, true interaction between contour and turbidity currents is rare at the facies level. Interaction is mainly visible in the seismic geometry of sedimentary bodies.

21.5.

C ONCLUSIONS

Several conclusions can be drawn with respect to the interaction between contour and gravity currents on the basis of the data presented in this chapter. • Interaction is not commonly identified at the sedimentary-facies level. It is thought to occur only when turbidity and contour currents have a comparable energy. In this case, the evidence comes primarily from fine-grained turbidites, i.e. turbidites without the basal intervals of the Bouma (1962) sequence (usually with only the T-c-d-e units). • As contour currents are common and long-lasting processes on continental margins, and as they are characterized by a fairly low energy, they will usually affect the tails of the turbidity currents (contour-current pirating). That will favour the deposition of turbidites without the topmost Bouma intervals. Sedimentary structures recording this kind of interaction will be difficult to observe. • Morphological resemblance exists between contourites and turbidites from the sedimentary-facies (core) scale to the system (seismic) scale. This similarity usually makes the recognition of contourite and/or turbidite systems difficult (Figure 21.14). • When morphological similarity occurs at the system (seismic) scale, some of the sediment bodies resulting from contour/turbidity currents present an elongated shape either parallel to the margin (frequently observed in mounded contourite drifts) or perpendicular to the margin (as frequently observed in deep-sea fan systems; Figure 21.14). • The partial or complete obliteration of primary sedimentary structures in sediments under the action of contour currents erases the turbiditic records. Winnowing by contour currents makes the recognition of a concomitant traction/suspension process difficult. • Sedimentary structures resulting from the interaction between contour and turbidity currents are sometimes difficult to discriminate from those resulting from the interaction of two turbidity-current systems. • Contour-current imprints are relatively well preserved during time-spans like the early Neogene, and in areas where contour-current intensity is high: in highlatitudes oceans, where the Coriolis force is stronger, and in areas where sea-floor morphology generates lateral flow constriction. • Signatures of interaction of contour and turbidity currents are better preserved when they are not the result of concomitant processes. Both processes will leave

455

T. Mulder et al.

Coast line

IC

1

Continental

TC downslope

2

AC CC direction Drift construction

A′

B

B′

B

C C′

D′

A

C

D D′

D

Lobes

C′

CC

Deep-sea turbidite system

B′

A

Channel levees

Shelf break

TC A′

Canyon

shelf

TC: High energy

AC: Active canyon IC: Inactive canyon Turbidity current

TC: Low energy CC: Current direction Contour current

Figure 21.14 Schematic examples of the interaction between down-slope turbidity currents and along-slope contour currents.

their impact on the morphology in different time intervals. Sedimentary series will then show either morphological heritage or intercalation of contourites and turbidites. In the former case, contour currents will follow paths determined by gravity process. In the latter case, contour and turbidity currents form alternations of sedimentary deposits. Consequently, interactions are more visible at the seismic-facies scale. • Because contour currents pirate particles present in the water mass, contourite deposits tend to reflect the regional sedimentation rather than local sources. For example, in mid- to high-latitudes areas, ice-rafted particles can represent a strong proportion of contourite deposits, forming glacigenic contourites.

456

Mixed Turbidite–Contourite Systems

• Finally, as the intercalation of contourites and turbidites is related to the relative strength of these currents, and because this is controlled by cyclic climatic variations, their study is a particularly efficient sedimentological tool to decipher the palaeoclimatic record contained in the sediments, particularly for the Tertiary and Quaternary (see Knutz, 2002b).

ACKNOWLEDGEMENTS This chapter represents UMR CNRS 5805 EPOC contribution number 1610. Authors thank Fre´de´rique Eynaud and Laurent Masse´ for their remarks and improvements on the manuscript.

C H A P T E R

2 2

H IGH -L ATITUDE C ONTOURITES T. van Weering1,2, M. Stoker3 and M. Rebesco4 1

Royal Netherlands Institute for Sea Research (NIOZ), Texel, Den Burg, The Netherlands Department of Paleoclimatology and Geomorphology, Free University, de Boelelaan, Amsterdam, The Netherlands 3 British Geological Survey, Murchison House, Edinburgh, Scotland, UK 4 Istituto Nazionale di Oceanografia e Geofisica Sperimentale (OGS), Borgo Grotta Gigante, Sgonico (TS), Italy 2

Contents 22.1. Introduction 22.2. Neogene and Quaternary Changes in Deep-Water Circulation of the Northern North Atlantic and in the Southern Ocean 22.2.1. Northern North Atlantic 22.2.2. Southern Ocean 22.3. Northeast Atlantic Margin Contourites 22.3.1. Rockall Trough and glacial record off Northwest Britain 22.3.2. Faroe–Shetland region 22.3.3. Norwegian continental margin 22.3.4. Greenland Margin 22.4. Antarctic Margin Contourites 22.4.1. Cooperation Sea 22.4.2. Cosmonaut Sea and Riisen-Larsen Sea 22.4.3. Antarctic Peninsula Pacific Margin 22.5. Conclusions Acknowledgements

22.1.

457 458 458 460 461 461 471 473 475 477 477 480 482 488 489

I NTRODUCTION

At least two factors favour the deposition of contourites along the margins of continents that are or have been glaciated throughout the Cenozoic: 1. Cold, glaciated poles produce the periodic formation of cold, dense waters that sink and spread towards the abyssal plains of the world oceans determining the principal engine for the global thermohaline circulation, also called the ‘‘oceanic conveyor belt’’. The formation of cold polar oceanic water began in the Oligocene only in the southern hemisphere, when the Antarctic continent became gradually covered by an ice sheet. The bipolar formation of cold water started later on, most likely in the Middle Miocene. Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00222-7

Ó 2008 Elsevier B.V. All rights reserved.

457

458

High-Latitude Contourites

2. The action of ice streams within ice sheets determines an extraordinary terrigenous sedimentary input to the high-latitude oceans. The sediment dispersal pattern is via turbidity currents, bottom-water transport, and ice rafting of glacial debris. Sediment deposition as deep-sea fans (called ‘‘glacial trough-mouth fans’’ on the polar continental margins) or as contourite drifts, both including ice-rafted debris, is therefore typical of these regions. Mixed turbidite/contourite deposition is frequent. It follows that – because in a glacially-influenced depositional system sediment flux and strength of bottom currents are both controlled primarily by changing climatic and oceanographic conditions – high-latitude contourites are candidate highresolution palaeoceanographic archives. What often limits the preservation of palaeoceanographic proxies, however, is the paucity – especially in the Antarctic waters south of the Circum-Antarctic Current – of calcareous foraminifers from which carbon and oxygen stable-isotope curves, Mg/Ca ratios, and faunal associations are extracted in the rest of the world’s oceans. The scarce palaeoceanographic proxies, added to the harsh conditions and hostile environments in which field investigations must be carried out, make the knowledge of high-latitude contourite depositional systems still rather poor with respect to middle- and low-latitude counterparts. Below we offer an overview of the best-known Neogene and Quaternary glacigenic contourite deposits in both hemispheres. We take into consideration the Antarctic continental margin and the northern North Atlantic Margin (north of the British Islands). For a description of the present-day global pattern of abyssal and bottom currents responsible for the recent deposition of high-latitude contourites, we refer to the overview by Zenk (2008). In the next section, we present a brief history of the water-masses circulation throughout the Neogene in the northern Atlantic and around the Antarctic continent.

22.2.

22.2.1.

NEOGENE AND QUATERNARY C HANGES IN D EEP-W ATER C IRCULATION OF THE NORTHERN NORTH ATLANTIC AND IN THE S OUTHERN O CEAN

Northern North Atlantic

The deepest oceanic gateway for intermediate- and deep-water masses between the North Atlantic and the Arctic is currently the Faroe Bank Channel (800–1200 m), between the eastern end of the Greenland–Scotland Ridge (GSR) and the Faroe Bank. Together with the Faroe–Shetland Channel, it forms the Faroe Conduit (Figure 22.1). The opening of the Faroe Conduit took place in the Early to Middle Miocene as a consequence of the subsidence of the GSR that until then had acted as a barrier for the exchange of deep-water masses between the Atlantic and Arctic Oceans (Blanc et al., 1980; Bohrmann et al. 1990; Eldholm, 1990; Stoker et al., 2005a–c). According to Miller and Tucholke (1983) and Davies et al. (2001), the gateway opened earlier, between the Late Eocene and the Early Oligocene.

Figure 22.1

Bathymetric setting of the NW European Atlantic Margin showing the main tectonic elements, oceanographic circulation pattern, and summary of large-scale sedimentary deposits and processes that have contributed to the shaping of the ocean margin during the Neogene (modified from Stoker et al., 2005b). The Iceland ^ Faroe Ridge forms the southeastern section of the Greenland ^Scotland Ridge (see multicolour version of this figure on the enclosed CDROM). Abbreviations: FSC = Faroe ^Shetland-Channel; IFR = Iceland ^ Faroe Ridge; WTR =Wyville-Thomson Ridge; PB = Porcupine Basin; NC = Norwegian Channel.

460

High-Latitude Contourites

The intensification of the Arctic–North Atlantic thermohaline circulation system since the Middle Miocene has triggered the formation of contourite drifts throughout the North Atlantic, the geological record of which was provided by the drilling of ODP Leg 151 across a major unconformity between Middle Miocene and Oligocene or older sediments (Thiede and Myhre, 1996). Another oceanographic change occurred at least along the continental margins of northwestern Europe in the Early Pliocene, when erosion affected the contourite deposits of the West Shetland Margin and Rockall Trough. Contourite sedimentation continued unconformably above the erosional surface, with a northward shift of the depocentres (Stoker et al., 2001). The Early Pliocene reorganisation of bottom currents persists in the present-day circulation pattern, albeit at reduced current strength as no comparable deep-water erosional surfaces exist in PlioQuaternary strata. The Early Plio oceanographic circulation change in the North Atlantic may have been triggered by the formation of the Isthmus of Panama in the Late Miocene to Early Pliocene (Haug and Tiedemann, 1998; Lear et al., 2003), which redirected warm and saline water masses to high (northern) latitudes. The replacement of the earlier circum-equatorial (longitudinal) circulation pattern by an inter-polar (latitudinal) flow strengthened the formation of North Atlantic Deep Water, the outflow of which was controlled by the changing topography of the GSR. In the Northeast Atlantic, contourites with a glacial component are found in the Feni and Hatton Drifts on the Nortwest UK Hebridean Margin and along the Rockall Bank Margins, in the West Shetland Drift on the margin of the Faeroe– Shetland Channel, and in the Lofoten Drift on the Northern Norwegian Margin (see also Stow and Fauge`res, 2008), and in the Eirik Drift off Greenland.

22.2.2.

Southern Ocean

The main control on the onset of the longitudinal circulation of the Circum Antarctic Current was the opening of the Scotia Sea that separated the southern tip of South America (Tierra del Fuego) from the Antarctic Peninsula. The initial evidence for a certain clockwise pathway of Circumpolar Deep Water through the Scotia Sea dates to the Early and Middle Miocene as a consequence of the thermal subsidence of the newly formed oceanic crust in the Scotia Sea (see review by Barker and Thomas, 2004). This generated the growth of sediment drifts in a distal position with respect to the continental margin (e.g. the Fossil Mounded Sedimentary Body described by Herna´ndez-Molina et al. (2004 and 2006b). The tectonic deepening of the Southern Scotia Sea gateways in the Middle Miocene allowed Weddell Sea Deep Water to spread counter-clockwise along the Pacific Margin of the Antarctic Peninsula and triggered the growth of the series of sediment drifts along the entire West Antarctica continental margin (Rebesco et al., 1997, 2002). Due to a progressive decrease of terrigenous sediment supply, and to a change in oceanographic conditions (Billups, 2002), the growth stage of these drifts ended in the Early Pliocene. Late Pliocene erosional unconformities and highenergy contourite deposition in the Weddell Sea suggest an increased production of

T. van Weering et al.

461

Antarctic Deep Water, perhaps as consequence of the last global cooling event that caused the intensification of the glaciation on the northern hemisphere and an expansion of the Antarctic ice sheet (Rebesco and Camerlenghi, 2008). Large mixed (contourite/turbidite) sedimentary bodies with a significant glacigenic component have been identified in many areas close to the Antarctic Margins. In this chapter, we describe drifts and other bottom-current-controlled deposits in the Cooperation to Riisen Larsen Seas and West of the Antarctic Peninsula. Many additional examples that cannot be included here are, among others, the mud waves and the sedimentary mounds in the Weddell and Scotia Seas (e.g. Hollister and Elder, 1969; Barker et al., 1975; Kuvaas and Kristoffersen, 1991; Howe et al., 1998, 2004; Gilbert et al., 1998; Michels et al., 2001, 2002; Pudsey, 2002; Maldonado et al., 2003, 2005) and on the Wilkes Land Margin (e.g. Hayes and Frask, 1975; Escutia et al., 1997, 2000, 2005; Donda et al., 2003, 2007; Close et al., 2007). A few small drifts were detected closer to the continent on the overdeepened Antarctic Shelf, where several local currents may have acted: the Andvord Drift (Manley et al., 1998; Harris et al., 1999), the Mertz Drift (Harris et al., 2001; Harris and Beaman, 2003; Presti et al., 2003), and the Wega and the Tabarin Drifts (Camerlenghi et al., 2001). Many examples of contourites produced under the influence of the Antarctic Circumpolar Current were detected in the major gateways (Drake, Tasman, SW Pacific) or close to isolated plateaus such as the Meteor and Kerguelen Plateaus.

22.3. 22.3.1.

NORTHEAST ATLANTIC MARGIN CONTOURITES

Rockall Trough and glacial record off Northwest Britain

The Rockall Trough is an underfilled deep-water sedimentary basin located on the Atlantic Margin of Britain and Ireland (Figure 22.2), which separates the Hebrides and Malin Shelves from the Rockall Plateau and adjacent George Bligh, Lousy, and Bill Bailey’s Banks. To the southwest, the trough deepens and opens out into the Porcupine abyssal plain. The Rockall Trough deepens from 1000 to 3500 m water depth towards the SW and includes the Rosemary Bank, and the Anton Dohrn and Hebrides Terrace Seamounts (Stoker et al., 1993). The Rockall Trough formed in the Late Eocene by differential subsidence on the continental margin (Praeg et al., 2005; Stoker et al., 2005c), and has acted throughout the Neogene as a gateway for north- and southward flowing intermediate and deep-water masses, in a pattern comparable to the present-day current regime (Figure 22.1). The sedimentary expression of this long-lasting oceanic circulation is an association of sediment drifts and sediment wave fields that include the giant elongate Feni Drift (Howe et al., 1994, 2001, 2006; Howe, 1996; Stoker et al., 1998a, 2001; de Haas et al., 2003). The record of ice rafting from the Hebrides Slope and Rockall Plateau begins at about 2.48 Ma (Shackleton et al., 1984; Stoker et al., 1994). However, it was only in the early Middle Pleistocene that ice sheets began to cross the Hebrides and

462

Fi gu re

22 .5

High-Latitude Contourites

Fi

gu

re

22

.4

a

Figure 22.4b

0

Glacially-fed fan Giant elongated sediment drift

300 km

Maximum limit of Pleistocene glaciation Palaeo-ice stream

Basinal drift complex

BGS borehole

Prolonged sea-bed erosion

DSDP/ODP site

Sediment waves Predominant flow paths of modern intermediate- and deep-water masses

Piston core BGS Airgun seismic reflection profile

Figure 22.2 Generalised Pleistocene glacially-influenced setting of the Rockall Trough. The map also shows locations of boreholes, cores, and profiles referred to in the text and other figures. Inset shows location of DSDP Site 610, which occurs on the Feni Ridge, SW of the main study area. Abbreviations: FB = Faroe Bank; BBB = Bill Bailey’s Bank; LB = Lousy Bank; GBB = George Bligh Bank; RB = Rosemary Bank Seamount; AD =Anton Dohrn Seamount; HT = Hebrides Terrace Seamount; HRB = Hatton ^ Rockall Basin; HB = Hatton Bank. A multicolour version of this figure is on the enclosed CD-ROM.

Malin Shelves, producing a prominent erosional unconformity (Figure 22.3) overlain by tills (Figure 22.4a; Stoker, 1995; Stoker and Bradwell, 2005), and causing the progradation of the Sula Sgeir and Barra–Donegal fans in the Rockall Trough. The ice-sheet reached the edge of the shelf on several occasions between

463

T. van Weering et al.

Hebridean margin Hebrides slope

Shelf edge

Key cores 88/7, MD95 7A 2006

Mid shelf

MIS 1 2 3

AGE (ka) 0

NOTE: Section broadly correlates with this interval, but absence of interglacial strata (MIS9) precludes assignment to specific glacial stage (MIS10, 8, or both)

Mid

50

Devensian Early 100

5e 6

Holocene

Late

4 5

British chronostratigraphy

Ipswichian 150

200

7

“Wolstonian” 250

8

NOTE: Ambiguity remains on the use of this term; the MIS6–10 record remains poorly defined in Britain

300

?

9 10 11

Glacial unconformity

Glacial debris-flow deposits

350

400

Hoxnian

12

450

Anglian

13

500

Cromerian

Only Barra-Donegal Fan

Figure 22.3 Glaciation curve for Northwest Britain, for the last 0.5 Ma, based on information derived from Stoker et al. (1994) and Knutz et al. (2002a, b). British chronostratigraphic stages are based on Bowen (1999), and the ages of marine isotope stage boundaries are based on Martinsen et al. (1987) and Williams et al. (1988). A multicolour version of this figure is on the enclosed CD-ROM.

marine isotope stages 12 and 2 (Stoker et al., 1994; Kroon et al., 2000; Knutz et al., 2001, 2002a). In the seismic record, the fans are made up of 200–400 m of continuous, high-amplitude clinoforms alternating with structureless lens-shaped units (glacigenic debris flows) that downlap into a much thinner (a few tens of meters) acoustically well-layered basinal succession. In essence, the glacially derived proximal sediment input along the eastern margin of the Rockall Trough contrasts with the sediment-starved western side of the trough.

464

High-Latitude Contourites

Submarine end-moraines Glacial unconformity

SE BP

0.4

SBM SBM

Sula Sgeir Fan

0.8 Clinoform

Glacial unconformity?

BGS83/04-6

1.6

C10

NW

SE Hebrides shelf

Rockall Trough 1

Sula Sgeir Fan

Figure 22.4

Miocene–Lower Pliocene

2

Glacial unconformity

Early Pliocene unconformity

Lower Pliocene– Holocene

Oligocene

Eocene & older 10 km

3

BASED ON DGH9508 Hebrides shelf

(b)

Two-way travel time (s)

Miocene– lower Pliocene

3

Early Pliocene unconformity

1

Peach slide

Rockall Trough W

E

2

Barra fan Lower Pliocene– Holocene

3 Late Eocene unconformity 4

4 Upper Eocene– lower Miocene

Two-way travel time (s)

Two-way travel time (s)

0

1.2

Buried elongated drift/ sediment waves 5 km

Mass-flow package

Early/mid-Miocene unconformity

Top Palaeogene volcanics

Distal edge of lowest debris-flow package

3.0 Sea bed

BGS 92/01-56

Onlap

3.1

Sea bed

3.5 3.6 10 km

Two-way travel time (s)

Two-way travel time (s)

Two-way travel time (s)

NW

Two-way travel time (s)

(a)

10 km

Figure 22.4 BGS airgun profiles across the (a) Sula Sgeir and (b) Barra fans, showing the overlapping/interdigitating relationship between the glacially-fed shelf-margin fans and the basinal contourites. Profiles located in Figure 22.2. Abbreviations: SBM = sea-bed multiple; BP = bubble pulse. A multicolour version of this figure is on the enclosed CD-ROM.

22.3.1.1. Seismic characteristics At a basin scale, the glacigenic contourites form the uppermost part of a well-defined Early Pliocene to Holocene succession of sediment drifts and waves, with the Feni Ridge being the most distinctive bedform (Laberg et al., 2005; Stoker et al., 2005b). In general, the Rockall Trough basin-floor sediments display a ponded basinfloor-fill geometry reflecting three main styles of contourite sediment-drift accumulation (Howe et al., 1994; Stoker et al., 1998a, 2001) (Figures 22.4 and 22.5):

Two-way travel time (s)

Wyville-Thomson ridge

T. van Weering et al.

Rockall basin SW

NE

Sula sgeir fan

Broad sheeted drift Figure 22.3a

10 km

BGS 85/05-4

Figure 22.5 BGS airgun profile showing the sediment drift complex that characterises the Plio-Pleistocene succession in the NE Rockall Trough, adjacent to the Wyville-Thomson Ridge, including elongated-mounded and broad-sheeted drifts and associated sediment waves, which are locally overlain by debris-flow deposits at the distal edge of the Sula Sgeir Fan. Profile located in Figure 22.2. Abbreviations: EPU = Early Pliocene unconformity; LEU = Late Eocene unconformity; SBM = sea-bed multiple.

465

466

High-Latitude Contourites

(1) broad, flat-lying to gently domed, sheeted drifts which occupy a large part of the axial region of the basin floor; (2) elongated mounded drifts onlapping the margins of the basin; and (3) the giant, elongate Feni Ridge in the southern Rockall Trough. It is important to note that these distinctive morphologies are simply type members within a continuous spectrum that forms the Rockall Trough Drift complex. The sheeted drift accumulation is up to 490 m thick and several tens of kilometres across and displays a relief of up to 60 m above the general basin floor. The sediments display a layered character in seismic profiles; the reflectors are parallel and laterally continuous with low-angle downlap to the northeast observed at the base of the drift. This is related to a migration of the drift towards the slope; the broad crestal region of the drift has migrated approximately 10 km to the northeast during the Plio-Pleistocene. A large field of sediment waves occurs on the southwestern flank of the sheeted drift (Figure 22.5). These extend over an area of about 550 km2, with a thickness up to 200 m, the basal 105 m of which is made of climbing waves passing upwards into a series of sinusoidal waves, some of which have heights of 18 m and wavelengths of over 1 km (Richards et al., 1987). Howe et al. (1994) demonstrated an upcurrent/ oblique-current migration direction towards the Hebrides Slope, consistent with the migration direction of the drift complex. An elongate mounded drift of 300 m thick and up to 20 km wide, with a maximum relief above the sea floor of 150 m, is preserved adjacent to the WyvilleThomson Ridge, from which it is separated by a wide moat (Figure 22.5). The drift displays a mounded, asymmetric profile, and its axis can be traced along the base of the Wyville-Thomson Ridge and northernmost Hebrides Slope for about 60 km, until it is found buried beneath the Sula Sgeir Fan (Howe et al., 1994) (Figure 22.4a). The crest of the drift commonly bifurcates along the strike (Stoker et al., 1998a). Active sediment waves about 2 km long, with heights above the sea floor of about 20 m, are present on the basinward side of the drift (Howe, 1996). A Neogene drift/moat complex surrounds the Rosemay Bank Seamount onlapping its flanks (Howe et al., 2006). Sediment waves reach dimensions of 150 m high with wavelengths of 1.5–2 km. In the central and southern Rockall Trough, the basinal sediments, including the Feni Ridge, display continuous, parallel to wavy-bedded reflections that largely represent aggradation of the basin infill following the Early Pliocene erosional even. At the northern end of the Feni Ridge, the elongated sediment drift onlaps the slope of Rockall Bank; farther south, however, the ridge is separated from the basin margin by a wide moat. In this area, sediment waves are common on both flanks of the Feni Ridge. Wavelengths range from several hundreds of metres up to 2 km or more; wave heights can exceed 30 m (Van Weering and De Rijk, 1991; de Haas et al., 2003) and are probably inactive at the present day. On the Barra Fan, the toe of a glacigenic debris flow (DF3) interdigitates with acoustically layered strata divided by Knutz et al. (2002a, b) into four seismic units (Figure 22.6a): The uppermost unit, A1, a stratified, reflective-to-transparent drape covering the debris flow displays a wavy character near the toe of the debris flow. The crest of each waveform appears to vary with time, with an overall up-slope migration. An even stronger up-slope-migrating waveform configuration is

T. van Weering et al.

467

preserved in the underlying unit A2, also confined to, and onlapping, the toe of the debris flow. In contrast, the lowermost units B and C display more uniform, parallel, continuous reflections and appear to underlie the debris flow. Elsewhere on the Barra Fan, sea-floor erosion by bottom currents has locally removed units A1 and A2, exhuming the underlying debris flows (Knutz et al., 2002a, b). In the area between the major fans, the slope apron has accumulated a predominantly acoustically layered succession of sediments. On the upper slope, an aggrading succession is preserved, locally disturbed by the erosive effects of iceberg ploughing (Stoker et al., 1994); on the lower slope, up-slope-migrating elongate drifts and sediment waves are developed. 22.3.1.2. Sedimentary characteristics A number of boreholes and short cores recovered from the Rockall Trough (including the Feni Ridge), Rosemary Bank, and Hebrides Slope show that the sediment drifts and waves consist of a variable composition of fine-grained mud and/or ooze, and coarser-grained sand (Shipboard Scientific Party, 1987, 1996; Stoker et al., 1989, 1993, 1998b; Van Weering and De Rijk, 1991; Howe et al., 1994; Howe, 1995, 1996; Knutz et al., 2002a, b). The predominance of individual components largely reflects the distance from the ice margin on the Hebrides Shelf, with coarser-grained facies located commonly proximal to the glacially-fed fans. Lithic clasts are commonly present across the region, either widely dispersed or in concentrated winnowed horizons (Stoker et al., 1998a; Howe et al., 2001). A brief summary of the sediments is provided below from the northern Rockall Trough, the upper Hebrides Slope, the Barra Fan, and the Feni Ridge. In the Northern Rockall Trough, three main sedimentary facies are recognised from BGS boreholes and short sediment cores: 1. A muddy sand facies (with 30–80% medium- to fine-grained sand and <45% mud) occurring in layers of up to 28 cm thick, moderately to well sorted, greyish yellow in colour, intensely bioturbated, containing shell material, found at the sea floor over a large part of the basin. The basal contact is erosive or occasionally graded. Cross-bedding is only rarely preserved. This facies is interpreted as a sandy contourite. 2. A sandy mud facies (up to 70% silt and 10–40% sands and minor gravels), poorly sorted, olive brown, intensely bioturbated occurs in layers up to 55 cm thick below the muddy sand layer on the sea floor. Centimetre-scale fining- and coarseningupward units are common and lamination is occasionally observed in cores and on X-radiographs. This facies is interpreted as a silty/muddy contourite. 3. A mud facies (95% silt and clay, <5% sand and variable but low amounts of matrix-supported gravel) made of poorly sorted, extensively bioturbated, homogeneous mud occurs in the lower sections of the cores. In most cases, the mud lacks any evidence of bottom-current influence and is best interpreted as being of hemipelagic origin (Howe, 1995). On the upper Hebrides Slope, BGS borehole 88/7, 7A penetrated the entire preserved Middle to Late Pleistocene succession, down to 67.82 m (the glacial

468

High-Latitude Contourites

unconformity). The succession consists almost entirely of grey, bioturbated, mud and sandy mud facies with subordinate thin, muddy sand facies. Abundant matrixsupported gravel clasts occur throughout the succession. The mud and sandy mud facies vary from homogenous to laminated, probably reflecting changes in the style of deposition, with laminated mud indicating intervals of relatively rapid sediment input to the slope (from the adjacent ice margin), similar in form to distal turbidites, whereas the homogenous mud represents the typical ‘‘background’’ sedimentation. The latter was subsequently reworked into normal and reversely graded units comparable to muddy contourites. The muddy sand facies is mostly preserved in isolated beds (six in total) less than 10 cm thick. They are generally massive with no indication of primary structures. The uppermost sand most likely represents the Holocene. On the Barra Fan, the upper Pleistocene to Holocene record has been sampled in short sediment cores (Howe, 1996; Knutz et al., 2002a, b) and by the 30 m long core MD96-2006 (collected immediately down-slope from the glacigenic debrisflow deposits), which extends back to Marine Isotope Stage 3 ( >33 ka BP) (Figure 22.6b). Knutz et al. (2002a, b) identified a number of sedimentary facies based on variations in carbonate content and fine clastic grain size, which, in general terms, were interpreted as: (1) silty/muddy contourites, (2) hemipelagite/muddy contourite, (3) glacimarine mud and clay, and (4) thin-bedded turbidites (Figure 22.5b). The silty/muddy contourites range from <1 m to 4.75 m in thickness and occur in both stadial and interstadial interglacial stages in seismic units A1, A2 and C (Figure 22.6b). The sediments consist of pale olive brown to olive grey, silty mud to silty to fine sandy mud, with a carbonate content of 10–25%, and generally in excess of 20% in the uppermost layer. Despite intense bioturbation, X-radiographs from adjacent short cores revealed faint planar lamination in the uppermost layer (Howe, 1996); silt lamination and thin sand layers are also noted near the base of the core (Knutz et al., 2002a), in seismic unit C. The hemipelagite/muddy contourite deposits range in thickness from about 1 to 3.5 m, and occur mainly in the interstadial/interglacial intervals (Marine Isotopic Stages 1 and 3), included within seismic units A1, B, and C. The sediments are dark grey to olive-greyish brown, silty mud, with a carbonate content of 10–20%. The mud is mostly homogenous with weak to moderate to intense bioturbation. The glacimarine mud and clay deposits range in thickness from 6 to 7.5 m, and were deposited largely within the stadial interval (Marine Isotopic Stage 2), within seismic units A1, A2, and B. The sediments range from homogenous, olive brown, clayey mud to dark greyish brown, silty/sandy mud, with common silty laminae and thin sand layers; they contain scattered gravel clasts and are weakly to nonbioturbated. Interbedded sand layers up to 20 cm thick have a sharp and erosive base and are mostly normally graded. They are interpreted as thin-bedded turbidites. Beds of coarser, pebble-sized material, showing water-escape structures, are observed between the sand and the overlying mud in some of the thicker layers (Knutz et al., 2002a).

469

T. van Weering et al.

(a)

SE Sediment waves showing upslope migration of crests

NW

MD95-2006

30 ms 1 km

A1 DF 3 A2 B

0

10.89 ka

its un ic m is Se

D

ep

(b)

th (m Li ) t in hof te ac rp ie re s ta C tio -1 n 4 ka B M P IS

C

1

5

A1

Key to lithology Silty-muddy contourite

14.86 ka

Hemipelagite/muddy contourite

10

2 17.66 ka

A2

Glacimarine mud and clay Thin-bedded turbidites

15

B 20

24.31 ka 25.81 ka 29.00 ka 29.33 ka

25

33.48 ka

3 C

30

Figure 22.6 Process interaction and sediment flux on the Barra Fan. (a) BGS deep-tow boomer profile and interpreted line drawing from the lower part of the fan. (b) Simplified log of piston core MD95-2006 (a and b modified from Kroon et al., 2000; and Knutz et al., 2002a, b). Core located in Figure 22.2. Abbreviations: ms = milliseconds; MIS = marine isotope stage; DF = debris flow. A multicolour version of this figure is on the enclosed CD-ROM.

470

High-Latitude Contourites

The Feni Ridge has been drilled at DSDP Site 610 and ODP Sites 980 and 981 (Figure 22.7). At Site 610, the upper 140 m thick succession of Late Pliocene to Pleistocene sediments contains sporadic gravel-sized lithic clasts incorporated within a cyclic sequence of olive-grey to pale brown calcareous mud (with up to 15% of sand) and white to very pale grey nanno-foram ooze ranging from very pure carbonate with less than 5% terrigenous components to marly ooze with 30–60% terrigenous material (Shipboard Scientific Party, 1987). The calcareous mud beds vary in thickness from 20 to 30 cm at the base of the succession to up to 3 m at the top; the ooze layers range from 0.5 to 1.0 m throughout the section. Farther North, a comparable succession was recovered at sites 980 and 981 (Shipboard Scientific Party, 1996). The sediments are interpreted as calcareous and siliceous biogenic,

Figure 22.7 Generalised lithological logs illustrating the spatial and temporal variation in the nature of the deep-water sediments in the Rockall Trough, based on information derived from Shipboard Scientific Party (1987): Site 610; Shipboard Scientific Party (1996): Sites 980, 981; and Stoker et al. (1994): borehole 88/7,7A. Boreholes are located in Figure 22.2. Abbreviations: LAD = last appearance datum; IRD = ice-rafted detritus.

T. van Weering et al.

471

muddy and sandy muddy, contourites, with the cyclic association related to glacial/ interglacial oscillations as a consequence of the dilution of biogenic carbonate by ice-rafted debris (Kidd and Hill, 1987; Stow et al., 1998a). This is consistent with the observation that all ice-rafted gravel clasts occur in the muddy intervals. This cyclicity has also been reported from other short cores recovered from the Feni Ridge (Van Weering and De Rijk, 1991), where interglacial units also display a positive Eu anomaly, a negative Ce anomaly, and a low Rb/La ratio with respect to glacial units. The fluctuating strength of bottom currents concomitant with climatic variation is revealed by the variation in distribution of the sortable-silt fraction (Dowling and McCave, 1993).

22.3.2.

Faroe–Shetland region

The Faroe–Shetland region consists of a sedimentary basin situated at the SE limit of the GSR. The Faroe–Shetland Channel presently is a deep-water channel that forms the (deepest) eastern critical gateway for the exchange of Norwegian Sea Deep Water between the North Atlantic and the Norwegian–Greenland Sea (Hansen and Østerhus, 2000). Following post-uplift subsidence associated with the withdrawal of the Iceland mantle plume during the Early Eocene, deep-water conditions were established throughout the basin. The Middle Eocene deposition of large submarine fans and hemipelagic sedimentation changed to contourite drift deposition along the Faroe Margin and the West Shetland slope in the Early Oligocene (Laberg et al., 2005). 22.3.2.1. The Faroe Drift The southeast Faroe Drift, with an Early Oligocene age of its base, has been described as the oldest drift deposit in the Northern Atlantic. The onset of drift deposition could be related to a major change in circulation regime, evidenced by a change from fine-grained clastic to a biosiliceous ooze sedimentation fed by a deep-water flow that originated from downwelling in the Norwegian Sea (Davies et al., 2001). The bottom-current flow since its onset has been directed to the southwest. This current also caused the deposition of parallel bedded, sheeted Miocene to Early Pliocene contourite deposits in the basinal part of the northern Faroe–Shetland Channel. Throughout the Late Pliocene to Holocene, contourites were deposited in basinal areas and lower slope sections at both sides of the Faroe– Shetland Channel, with intermittent erosional events. 22.3.2.2. The West Shetland Drift The West Shetland Drift has an areal extent of over 20,000 km2 and is located on the Shetland Margin of the Faeroe–Shetland Channel (see Hohbein and Cartwright, 2006, and references therein). Knutz and Cartwright (2003, 2004) and Hohbein and Cartwright (2006) described the seismic stratigraphy from 2-D and 3-D seismic data sets distinguishing three seismostratigraphic units: Unit 1 is floored by the intra-Neogene erosional unconformity (INU) and is made up of a lower subunit (1a) with continuous reflections cut by several U- to

472

High-Latitude Contourites

T. van Weering et al.

473

Figure 22.8 Upper panel shows 3-D visualisation of the PL 20 reflector, showing a surface marked by sediment waves with crests oriented oblique to the slope. Note the decrease of wave dimensions up- and down-slope. Picture looks to the NE, along the West Shetland Slope. Lower panel shows a seismic profile illustrating cross-sectional geometries of the sediment waves, revealing a NE migration pattern forced by the SW flow direction of the depositional current (from Hohbein and Cartwright, 2006, with permission from Elsevier). A multicolour version of this figure is on the enclosed CD-ROM.

V-shaped channels and levee complexes overlain by a series of reflections marking deposition by along-slope currents. The upper subunit (1b) is a lens-shaped body with parallel aggradational reflectons including a field of well-developed sediment waves with a spacing of 300–700 m and with trough to peak heights of 15–25 m (Figure 22.8). The sediment waves form linear, slightly sinuous, or bifurcating crests 6–8 km long that strike at an angle of 25° with the contours. Based on geometry and internal reflection configuration, they are interpreted as fine-grained sediment waves formed by contour currents (Wynn and Stow, 2002a). Following the classification of Fauge`res et al. (1999) and Rebesco and Stow (2001), this unit forms a plastered contourite drift. Unit 2 is bound at its base by a small-angle erosional unconformity (Pl-40, Figure 22.9), and at its top by a regional Glacial Unconformity (GU). Internally, the unit consists of a succession of parallel, moderate- to low-amplitude discontinuous reflections with a maximum thickness of up to 300 m. It is also interpreted as a fine-grained plastered contourite drift deposited by SW flowing contour currents (Knutz and Cartwright 2003; Hohbein and Cartwright, 2006). The uppermost unit, Unit 3, consists of a series of prograding wedges of glacially derived sediments of 50–100 m thick. Iceberg ploughmarks, considered related to the advancing major Pleistocene ice sheets on the adjacent continental shelf, cut into the entire unit and reach the basal unconformity GU. In summary, the West Shetland Slope was influenced throughout the Neogene by southwestwards-directed contour currents with an estimated velocity between 9 and 30 cm s 1. Settling of sediment is considered to result from small-scale fluctuations of the contour current, in combination with variability of the entrained sediment load. Two episodes of increased current velocity determined the INU and Pl-40 erosional unconformities.

22.3.3.

Norwegian continental margin

The elongated Lofoten sediment drift has accumulated since the Miocene on the lower slope of the Norwegian continental margin (Laberg et al., 1999, 2002, 2005). Its axis is parallel to the contour lines and to the local contour currents that flow to the North-East. The base of the drift is marked by a regional seismic reflector, and the drift above it is divided in successions bounded by unconformities (Laberg et al., 2005). The seismic units are well layered, made of medium- to high-amplitude continuous, parallel or slightly divergent reflectors. A moat in one of the

474

High-Latitude Contourites

Figure 22.9 Detail of a 3-D seismic profile and section perpendicular to the Shetland slope, showing reflection terminations (marked by arrows). Inset shows details of an up-slope prograding moat system. See text for names and meaning of reflectors (from Hohbein and Cartwright, 2006, with permission from Elsevier).

lowermost units in the upper slope suggests the location of the axis of the initial contour current, where sedimentation was reduced by a vigorous current regime (Figure 22.10). There is no evidence of sediment transport and deposition down-slope. The onset of the Lofoten Drift is considered to result from increased circulation within the Norwegian Greenland Sea, as a result of the increased water exchange between the Norwegian Greenland Sea and the North Atlantic Ocean. The sedimentary record of the uppermost part of the drift shows a strong correlation with climatic cycles, with sedimentation rates during the Last Glacial

475

T. van Weering et al.

twtt (s) NW

SE

Lofotoen Drift

VB-32-89

7.5° Paleo-moat

1.5

Intra Miocene

3°

1 km 2.0

2.5

Figure 22.10 Part of seismic profile VB 32-89 showing the seismic expression of the Lofoten Drift. The base of the drift is considered as intra-Miocene. The drift shows layered, continuous, parallel, or slightly divergent internal reflections with local expressions of a palaeomoat (modified from Laberg et al., 2005).

Maximum (LGM) that as a result of glacimarine sediment input during the late Weichselian are an order of magnitude higher compared to the Holocene sedimentation due to bottom-current winnowing of shelf and upper slope sediments (Laberg et al., 2005; Laberg and Vorren, 2004). On the Norwegian upper continental slope, mounded elongated drifts also occur intercalated between glacigenic debris-flow deposits of Late Pliocene to Pleistocene age (Dahlgren et al., 2005) forming the NAUST formation. A drift that filled the scar of the Sklinnadjupet slide (Laberg et al., 2001) started to grow between 339 and 245 ka BP, just after the slide took place, and is still forming today. Most likely, the drift depositional system of the Norwegian Margin was interrupted in the Middle to Late Pleistocene during phases of maximum glacial extension, due to the rapid settling of glacigenic debris-flow deposits.

22.3.4.

Greenland Margin

Presently, the southeastern Greenland Margin is swept by the Denmark Strait Overflow Water and other cold water masses that flow parallel to the margin between 500 and 3000 m water depth. These water masses contribute to form the deep western boundary current system after merging with the Labrador Sea Waters off the southern tip of Greenland (Hunter et al., 2007a). On the southeastern Greenland Margin, turbidity and contour currents interacted to produce a mixed depositional system since the Miocene (Rasmussen et al., 2003). A dominant turbidite sediment input during the Early Pliocene allowed the growth of channel overbank deposits oriented perpendicular to the continental slope that were diverted to a SW direction, parallel to the slope, by the contour current affecting the deeper and distal part of the margin. The result is a mounded sedimentary deposit, the depocentre of which migrates southwards through time with occasional up-slope migrating sediment waves. Elongated contourite drift deposits trending parallel to the slope formed in the Middle Pliocene, as a consequence of the enhancement of North Atlantic deep-water circulation and strengthening of the contour currents.

476

High-Latitude Contourites

Figure 22.11 Distribution of echo characters and sedimentary regimes in relation to depth and forcing mechanisms (upper panel), and distribution and effects of DWBC on sedimentation/ erosion on the Eirik Drift (lower panel). Pie diagrams show the distribution of clay minerals over the Eirik Drift (modified from Hunter et al., 2007a). Inset shows location of the Eirik Drift (see the colour version of this figure on the enclosed CD-Rom).

T. van Weering et al.

477

The Pliocene–Quaternary Eirik Drift (or Eirik Ridge), the largest contourite deposit of the Greenland Margin, is found where the Denmark Strait Overflow Water and the Labrador waters, both flowing parallel to and southwards along the eastern and western Greenland Margins, merge and release the entrained sediment load from the East Greenland Shelf and Irminger Sea as their velocity decreases away from the southern tip of Greenland (Figure 22.11). The Denmark Strait Overflow Waters flow between 1900 and 3000 m deep (Clarke, 1984) with a significant decadal variability caused by changes in the output from the Nordic Seas (Bacon, 1998). The drift crest deepens from 1500 m in the North to about 3400 m in the SW (Hillaire-Marcel et al., 1994; Hunter et al., 2007a). The drift is 800 km long, about 25 km wide and several kilometres thick. According to McCave and Tucholke (1986), the Eirik Ridge is a ‘‘detached drift’’. According to Chough and Hesse (1985), the construction of the drift was induced also by anomalies in the basement topography. Sedimentation rates on the Eirik Drift are variable, ranging between 10 and 50 cm ka 1 in the last 160 ka (Hillaire-Marcel et al., 1994; Evans et al., 2007), depending on the position relative to the core of the bottom-current flow. Where the current velocity is highest, the sediments from the Holocene to Marine Isotope Stage 5 have been nearly entirely winnowed (Hillaire-Marcel et al., 1994). Due to the limited work performed on the Eirik Drift, the seismic character is not well defined. In echo-sounder records, near-surface high-amplitude reflectors are interpreted to represent sandy, winnowed sediments and a high-energy current linked to the flow of the Denmark Outflow Water, with present-day measured velocities of up to 25 cm s 1 (Hunt et al., 2007a). Variable amplitude of reflectors coincide with areas of lower maximum current velocity observed at the bottom (20 cm s 1) and a limited amount of ice-rafted debris, while low-amplitude reflectors with scattered patches of high reflectivity correlate with sediments with significant amounts of icerafted debris (Hunter et al., 2007a). High-amplitude hazy reflectors locally overlain with a very low-amplitude sea-floor echo are thought to reflect a sea floor made of moderate amounts of sand/silt, draped by finer-grained sediments (Damuth, 1980). Among the rare sediment samples from the Eirik Drift are dark grey, intensely bioturbated silty clays with clayey sands and sandy muds, often finely laminated, and occasionally with foraminifer-richer clayey silt layers (Chough and Hesse, 1985; Turon et al., 1999; Evans et al., 2007). Based on seismic correlation with ODP Site 964, the onset of the Eirik Drift growth is defined as started in the Early to Middle Pliocene (Arthur et al., 1989).

22.4. 22.4.1.

A NTARCTIC MARGIN C ONTOURITES

Cooperation Sea

Current-controlled deposits within this sector of the Antarctic Margin have been first identified in seismic-reflection profiles offshore Prydz Bay (Kuvaas and Leitchenkov, 1992). Modern ocean circulation in the Prydz Bay–Cooperation Sea region has many components (Cooper and O’Brien, 2004). Surface circulation in the bay is a cyclonic

478

High-Latitude Contourites

gyre with cold-water inflow from the east and outflow along the west side of Prydz Bay (Smith et al., 1984; Wong, 1994). There is relatively little high-saline deep water due to Prydz Bay’s geography and bathymetry (Smith et al., 1984). These two factors lead to little down-slope bottom-water current activity beyond the shelf edge. The deep-water movements on the continental slope and rise are attributed to three largescale ocean systems: the Polar Current, moving west near the shelf edge; the Antarctic Divergence, producing cyclonic gyres over the slope and inner rise; and the Antarctic Circumpolar Current, moving east over the outer rise and beyond. The interaction of the near-sea-floor currents (e.g. down-slope density currents and along-slope ocean currents) is believed to control slope and rise sediment deposition in the Prydz Bay region (Kuvaas and Leitchenkov, 1992), as in the Antarctic Peninsula region (Rebesco et al., 1997) and elsewhere. A thick succession of Cenozoic glacimarine sediments on the Prydz Bay continental rise was deposited in a complex manner, suggesting interaction of turbidity currents and strong bottom currents (Kuvaas and Leitchenkov, 1992). Reflectionseismic profiles from the lower continental slope and rise suggest an abundance of current-influenced deposits, such as sediment waves and large sediment drifts. In addition, large channel–levee complexes indicate deposition by turbidity currents. Both the large channels and the sediment ridges trend oblique to the continental margin. The geometry and character of the seismic-reflection pattern suggest that the ridges have been deposited under the combined influence of overflow from down-slope channelised turbidity currents and strong bottom-water flows. The sediment waves and the differences along the eastern and western channel margins suggest that the bottom currents are flowing towards the west. The initiation of turbidite sedimentation is inferred to have occurred in the Late Eocene to Early Oligocene, when the Amery Ice Shelf reached the shelf edge for the first time. The onset of current-controlled deposition was possibly related to the opening of the Drake Passage at the Oligocene/Miocene boundary. The most conspicuous sediment drifts are developed on the continental rise offshore from Prydz Bay (western part of the Cooperation Sea) between the Wilkins and Wild Canyons, and are referred to as the Wilkins and Wild Drifts. By analogy with other drift deposits on the Antarctic Margin (Rebesco et al., 1997; Shipboard Scientific Party, 1999b), the drifts are composed of alternating clastic- and biogenicrich intervals that reflect alternations of glacial and interglacial conditions. The mixed pelagic and hemipelagic sediments of the central Wild Drift have been sampled at ODP Site 1165 situated in 3537 m of water depth (Figure 22.12; Shipboard Scientific Party, 2001b). Drilling at Site 1165 yielded a relatively continuous 999 m thick sedimentary section of Early Miocene to Pleistocene deposits with only few relatively minor (<2 Ma) disconformities. The sediments cored at Site 1165, fine-grained and without obvious graded bedding, may be interpreted as a succession of muddy contourites and hemipelagic deposits. The Early Pliocene to Pleistocene part of the succession has a high biogenic component, is bioturbated, and shows only minor current activity during deposition (which caused some lamination and interbedding of differently coloured beds). The Middle to Late Miocene succession consists of alternating bioturbated dark and greenish-grey clay facies, which is interpreted to represent cyclical changes

479

T. van Weering et al.

West

10 km 870

East

Site 1165 860

850

840

830

820

4.5

PP12

5.0 5.5 P3

6.0 P2

Two-way traveltime (s)

4.0

6.5 7.0

P1

Figure 22.12 Regional seismic-reflection profile 33006 across ODP Site 1165 (Prydz Bay, Antarctica) on the Wild Drift (modified from Shipboard Scientific Party, 2001b, and from Kuvaas and Leitchenkov, 1992).

from more hemipelagic facies to muddy contourites (Figure 22.13). Average sedimentation rates are up to 5 cm ka 1. These structureless, strongly bioturbated, fine-grained facies are similar to those described from Pliocene–Pleistocene Drift sediments in the North Atlantic (Stow and Holbrook, 1984). The Early Miocene succession is characterised by interbedded, dark grey, laminated claystones and thin (<50 cm thick) greenish-grey claystones which represent cyclical changes in depositional environments. Average sedimentation rates range up to 15 cm ka 1. Abundant silt laminae and ripple cross-bedding in the dark grey facies are interpreted to indicate relatively strong bottom-current activity. Bioturbation is rare and lamination is well preserved, but graded beds have not been observed. The silt ripples may represent traction surfaces where the finer fraction has been winnowed during short phases of increased current activity. The two main cyclically alternating facies at Site 1165 (the dark grey, laminated, terrigenous one interpreted as muddy contourite and the greenish, homogeneous, biogenic and coarse fraction one interpreted as hemipelagic deposits with ice rafted debris) reflect orbitally driven changes (Milankovitch periodicities) recorded in spectral reflectance, bulk density, magnetic susceptibility data, and opal-content changes (Shipboard Scientific Party, 1999b; Gru¨tzner et al., 2003; Rebesco, 2003). Superimposed on these short-term variations, significant up-hole changes in average sedimentation rates, total clay content, IRD amount, and mineral composition are interpreted to represent the long-term Early to Late Miocene transition from a temperate climate to a cold-climate glaciation. The analysis of the short-term variations, interpreted to reflect ice-sheet expansions controlled by 41-ka insulation changes, were performed on the closely spaced sampled records provided by the multi-sensor track. Among those, cycles are best described by spectral-reflectance

480

High-Latitude Contourites

Figure 22.13 Cyclic sedimentation in Core ODP-1165-1165B-10H from the Wild Drift (modified from Gru«tzner et al., 2003). From left to right: photograph and green/grey colour ratio (dotted line = 5 cm record, solid line = 25 cm average); magnetic susceptibility, green/ grey colour ratio, and Fe intensity measured for cores (triangles: samples with >10% sandsized particles, i.e. >250 mm; arrows: susceptibility spikes indicating ice-rafted debris layers); magnetostratigraphic age control; astronomically calibrated depth-to-age conversion based on correlation of orbital obliquity and filtered colour (green/grey ratio) record (unfiltered data set is shown in background). A multicolour version of this figure is on the enclosed CD-ROM.

data and, in particular, by a parameter calculated as the ratio of the reflectivity in the green-colour band and the average reflectivity (Rebesco, 2003).

22.4.2.

Cosmonaut Sea and Riisen-Larsen Sea

An interplay of turbidite and contourite deposition along the Cosmonaut Sea and Riisen-Larsen Sea has been identified from seismic data by Kuvaas et al. (2004, 2005) and Solli et al (2007a, b). Down-slope and along-slope processes thus interacted to form glacimarine deposits, mostly attributed to the advances of an ice sheet, delivering huge amounts of sediment to the shelf edge and upper slope during glacial maxima.

481

T. van Weering et al.

The seismic data reveal the presence of multiple types of contourite accumulations as well as of large channel–levee complexes. A major sediment lens is present below the upper continental rise along the entire Cosmonaut Sea Margin (Figure 22.14). The lens probably consists of sediments supplied from the shelf and slope, being constantly reworked by westward flowing bottom currents that redeposited the sediments into a large-scale plastered drift deposit prior to the main glacigenic input along the margin. On the lower continental rise, large-scale sediment bodies extending perpendicular to the continental margin were deposited as a result of down-slope turbidity transport and westward flowing bottom currents. Several drift deposits on the lower rise/abyssal plain and along the western margin of the Gunnerus Ridge indicate the activity of contour currents. High-relief, elongated, and sometimes semicircular depositional structures found on the upper continental rise mainly resulted from the action of closely spaced turbidity currents. Large palaeo-channels show evidence of multiple incisions and have prominent, multi-storey levees. Oversteepening and instability generated down-slope turbidity currents forming channel–levee complexes, whereas the contourite accumulations were probably mostly formed during interglacials. The drift deposits overlie a distinct regional unconformity, which is N

Two-way travel time (s)

1.0 S 50 km

2.0 3.0 4.0 5.0

CS-4

Regional se

diment lens

6.0 7.0

Top basement

a

RAE 4607 S

N

Two-way travel time (s)

4.5 20 km

5.0 5.5 6.0 6.5

Regional sediment lens

7.0 CS-4 b

RAE 4607

Figure 22.14 Different types of sediment bodies observed along the Cosmonaut Sea Margin, Antarctica (from Kuvaas et al., 2005, with permission from Springer). (a) Line drawing of seismic profile RAE 4607 illustrating a high-relief sediment mound on the upper continental rise (light grey), a large sediment mound on the lower continental rise (dotted dark grey), and a regional sediment lens extending along the entire margin. (b) Inset from (a), showing the seismic signature of the various sediment mounds. The thick black line indicates the thickness of the regional sediment lens.

482

High-Latitude Contourites

considered to reflect a major palaeoceanographic event, probably related to a Middle Miocene intensification of the Antarctic Circumpolar Current.

22.4.3.

Antarctic Peninsula Pacific Margin

Twelve sedimentary mounds were identified on the upper continental rise of the Pacific Margin of the Antarctic Peninsula in water depths of 2700–3700 m. These mounds are interpreted as produced by the interaction of along-slope bottom-water flow with down-slope turbidity currents, providing a complete range of intermediates between two end-members: the sediment drift and the channel–levee (Rebesco et al., 2002). One of the southernmost sediment drifts, Drift 7, is the most studied and is considered a genuine sediment drift in which southwest flowing bottom currents pirate the sediment of the glacigenic turbidity currents. The present-day bottom-current flow in the area of Drift 7 was reconstructed (Camerlenghi et al. 1997; Giorgetti et al., 2003) from data recorded in three deep moorings (Figure 22.15). The general SW flowing circulation is forced by drift topography to follow the isobaths and appears to be geostrophically adjusted at least for a large part of the year. The mean current velocity 8 m above the sea

Figure 22.15 Location map of current-moorings on Drift 7, Antarctic Peninsula Pacific Margin (modified from Rebesco et al., 2002).The mean direction of the current measured 8 m above the sea floor is indicated with arrows at the location of the three moorings. The inferred bottom-water flow path along the isobath between the moorings is shown by a dashed line. A multicolour version of this figure is on the enclosed CD-ROM.

T. van Weering et al.

483

floor is between 4 and 6.2 cm s 1, and its velocity never exceeded 20 cm s 1. This flow is capable of transporting fine sediment particles, but not of eroding the sediment. The source of the bottom water is not clear, but it is probably the southern Weddell Sea. The water mass may in fact be modified Weddell Sea Deep Water (WSDW) known to flow anticlockwise around the northern tip of the Antarctic Peninsula (Sievers and Nowlin, 1984; Nowlin and Zenk, 1988; Orsi et al., 1999; Naveira Garabato et al., 2002b) or Lower Circumpolar Deep Water (LCDW) originated in the Weddell Sea and overlying the WSDW in the Drake Passage (Herna´ndez-Molina et al., 2006b). Alternatively, the water mass may be Southeast Pacific Deep Water (SPDW), which branches off from the ACC and is clockwise retroflected into the Southeast Pacific Basin; however, information about its potential density and dissolved silica concentration is not in favour of this option (Hillenbrand et al., 2008). Some evidence for an influence of the ACC on deposition on the lower continental rise west of the Antarctic Peninsula comes from a Miocene patch drift NE of some seamounts that is interpreted to indicate a NEward flowing bottom current related to the ACC (Herna´ndez-Molina et al., 2004, 2006a). Drift 7 and the majority of the other drifts have an asymmetric external shape with a steeper, rougher SW side and a gentler, smoother NE side. Seismic profiles show that the drifts generally have gently dipping NW sides merging with the lower continental rise, and steeply dipping SE sides facing the continental slope. Swath mapping revealed a complex morphology that is considered to result from the co-existence of various sedimentary processes: erosional gullies produced by debris flows on the upper part of the continental slope; deeply incised channels at the slope base; main trunk-type inter-drift turbidity channels separating the drifts; slide scars; undulating depositional bedforms interpreted as bottom current sediment waves; and fluid-escape structures perhaps associated to deep-water coral bioherms (Rebesco et al., 2007). These data suggest (Rebesco et al., 2007) that the elongation of the drift is derived from an offset in the margin and that the asymmetric morphology originated from persisting accumulation on the up-stream NE side and reduced accumulation/occasional erosion on the down-stream SW side, with subsequent instability and steepening (Figure 22.16). The internal structure of these drifts (Rebesco et al., 1994, 1996, 1997, 1998b, 2002; McGinnis and Hayes, 1995; McGinnis et al., 1997; Herna´ndez-Molina et al., 2006b) is different beneath the gentle and the steeper sides. The steeper sides are characterised by high reflectivity, abundant lateral terminations and undulations, and sea-floor terminations of reflectors, by either erosion or non-deposition. Prominent changes in reflectivity parallel to the sea floor, probably produced by a diagenetic change deep in the sediments, are particularly evident on the steep sides of the drifts. In contrast, the gentler sides are relatively smooth and are underlain by continuous, highly reflective units, with linear, parallel, or sub-parallel internal reflectors conformable with the sea floor. The drifts are separated by deep-sea channels traversing the deep areas between drifts. Channels are only evident between the drifts and not on the drifts

484

High-Latitude Contourites

Figure 22.16 Multibeam bathymetry data from the distal part of Drift 7 shown in perspective view and cut by multi-channel seismic-reflection profile IT95-130 (modified from Rebesco et al., 2007). The data are part of a large data set acquired by R/V OGSExplora between 1995 and 2004 by projects MAGICO (Multibeam Antarctic Glacial system Integral Coverage) and SEDANO (Sedimentary Drifts of the Antarctic Offshore) funded by the Progetto Nazionale di Ricerchein Antartide (PNRA). Growth of the drift reflects variations in both the Antarctic ice sheet and Southern Ocean along-slope bottom currents. A multicolour version of this figure is on the enclosed CD-ROM.

themselves. The acoustic facies of these channel–levee systems is very different from that of the drifts. Channel floors are characterised by high-amplitude, discontinuous reflectors, surrounded by transparent facies. Levees are shown by relatively well stratified deposits abruptly wedging out at the side of the channel. Buried channels with dimensions comparable with the modern ones are detectable in seismic profiles within the three upper successions. They show a limited lateral shift and are confined to the inter-drift areas. A total of 17 gravity cores have been collected from Drift 7, and 15 from the other drifts (Figure 22.17; Camerlenghi et al., 1997; Pudsey and Camerlenghi, 1998; Pudsey, 2000; Lucchi et al., 2002; Villa et al., 2003). The sediments cored on Drift 7 are predominantly terrigenous and very fine-grained. Cores show a cyclicity between (1) brown, bioturbated (mainly by Planolites), diatom-bearing mud with foraminifers and common dispersed IRD; (2) grey, barren, laminated mud. Parallel to lenticular and wavy lamination are rather indistinct and irregular; there are neither sharp-based graded units nor any signs of erosion. Bioturbation and IRD are rare. Contacts between brown and grey units are gradational and bioturbated. Cores on the steep sides of the drift recovered a condensed succession with thinner cycles and probable hiatuses. Cores from the distal part of drift and in the adjacent channels contain coarse-grained turbidite beds. According to Lucchi and Rebesco (2007), the laminated glacigenic muddy contourites (Figure 22.18) identified in the glacial stages within Drift 7 are (atypically) not bioturbated because of unusual, climatically related, environmental conditions of suppressed primary productivity and oxygen-reduced deep waters. Similar glacial contourites were observed on most of the Antarctic Margins (e.g.

T. van Weering et al.

485

Figure 22.17 Dip and strike multi-channel seismic profiles crossing Drift 7 with location of gravity cores and selected radiograph facies of last glacial sediments (from Lucchi and Rebesco, 2007; with permission from The Geological Society, London). (a) Multi-channel seismic-reflection profile IT92-109 crossing Drift 7 in a direction parallel to the margin, showing the location of gravity cores and radiography facies of most significant sites. The radiograph intervals from each core represent the same chronostratigraphic interval (last glacial sediments). Pervasive lamination is produced by faintly laminated mud alternating with closely spaced silty laminae, lenses (millimetre-thick), and layers (up to 4 cm thick).These deposits decrease in thickness and frequency towards the crest of the drift where silty laminae are detectable only in radiographs. (b) Multi-channel seismic-reflection profile I95-135 crossing Drift 7 perpendicular to the margin and showing the location of gravity cores and radiography facies of most significant sites. The radiograph intervals reported from each core represent the same chronostratigraphic interval (last glacial sediments). Along the dip transect, millimetre-thick silty laminae are common. At the top of Drift 7, the last glacial sediments consist of homogeneous layers (2^10 cm thick), grey intervals separated by sharp and irregular boundaries.The X-radiographs also indicate the presence of 1^2 cm thick intervals of scattered fine-grained sand and silt (IRD). A multicolour version of this figure is on the enclosed CD-ROM.

Anderson et al., 1979; Pudsey et al., 1988; Mackensen et al., 1989; Grobe and Mackensen, 1992; Pudsey, 1992; Gilbert et al., 1998; Anderson, 1999; Yoon et al., 2000) with the exception of areas in which polynyas were maintained during the glacial stages. Laminated, non-bioturbated contourites have been described also from shallow-water organic-rich muddy conditions in the Baltic Sea, where high biological productivity associated with periodic stagnation of near-bottom waters generated an anoxic condition at the sea floor (Sivkov et al., 2002). Glacial contourites can be used, especially in Antarctica, as a proxy to define temporal and spatial extension of the sea ice.

486

Figure 22.17

High-Latitude Contourites

(Continued )

Figure 22.18 Cartoon-style model showing the successive climatic stages in the sediment drifts of the Antarctic Peninsula Pacific Margin (from Lucchi et al., 2002, with permission from Elsevier). The model explains the repetitive succession of the four distinctive sedimentary facies, each characterised by a typical clay mineral assemblage and bioclastic content. A multicolour version of this figure is on the enclosed CD-ROM.

487

T. van Weering et al.

Scientific drilling (ODP Leg 178) showed that the alternation of lithofacies is less regular and evident in the Pliocene and Miocene than in the Pleistocene, and that silty and muddy turbidites are present within the glacial facies in the distal part of sediment Drift 7 (Barker and Camerlenghi, 2002a). The low energy of this depositional environment is confirmed by a fine grain size throughout. The general depositional model (Figure 22.19; see also Rebesco et al., 1997; Amblas et al., 2006) implies weak SW-flowing bottom currents pirating the suspended fine material from turbidity currents flowing during glacial stages in the adjacent channels with down-stream-enhanced deposition of the entrained sediment on the NE, gentler, up-stream side of the drift. Slope instability on the SW, steeper, down-stream side generates a moderate up-stream migration of the asymmetric drift (similarly to a huge sediment wave).

b

Antarctic Peninsula

SE Greenland n

gi

ar

ft

ft

Dri

en

in nt

Dri

ta

lm

Co

Dri

ft

m al

t

en

in nt

c

Co

a 1 km

b

Bottom currents (Coriolis force pushes toward left)

SW t

gra

Mi

ion

Turbidity current

d

ft

g

ar

Dri

a

in

NE

d

c SW

Local bottom currents (Coriolis force pushes toward left)

Turbidity current

NE Turbidity current

Channel

Prog

rada

tion

Migrating channel

Figure 22.19 Schematic diagrams comparing the sediment drifts of the Pacific Margin of the Antarctic Peninsula and the sedimentary mounds of the SE Greenland Margin (from Rebesco et al., 2007, with permission from The Geological Society, London). The large grey arrow shows the general path of the bottom current, and the small black arrows show the trend of the Coriolis force. In both cases, large asymmetric sedimentary mounds separated by deep-sea channels lie within an overall along-slope-flowing contour current. However, while the setting of the SE Greenland mounds is consistent with a channel ^ levee model in which the asymmetry may be explained by the Coriolis force alone (towards the right in the northern hemisphere), that of the Antarctic Peninsula Drifts requires bottom-current influence. In both cases, the mounds are dipslope-elongated, slightly oblique with respect to the trend of the margin, but while the Antarctic Peninsula ones lean and prograde upstream (like huge sediment waves with a downstream steep side), the SE Greenland ones lean and prograde downstream (like current-deflected levees with an upstream steep side).

488

High-Latitude Contourites

22.5.

C ONCLUSIONS

• In mid- to high-latitude areas, an ice-rafted component is commonly present within sedimentary drifts deposited or significantly affected by bottom currents. These deposits are referred to as ‘‘glacigenic contourites’’. • For the formation of contourites along both the NE and NW Atlantic Ocean Margins, the GSR is, and has been, of critical importance by acting as a (temporal) barrier for the exchange of deep-water masses between the Atlantic and Arctic Oceans, mainly through the Faeroe Conduit and Denmark Strait and by overspilling between Iceland and Scotland. • Changes in the configuration of basins and tectonic highs and elevation of bathymetric thresholds most likely enhanced the intensity of deep-water currents. • Glacigenic contourites are identified on continental slopes (especially in the northern hemisphere) where prograding, distinct, acoustically structureless to chaotic, mass-flow packages (up to several tens of metres thick) are separated by thinner (generally <10 m), high-amplitude, slope-wide clinoforms. A fluctuating IRD content further reveals the oscillatory nature of glacially sourced input. The fluctuating strength of bottom currents concomitant with climatic variation is revealed by the variation in distribution of the sortable-silt fraction (e.g. Dowling and McCave, 1993) in these sediments. • Scouring of a glacial unconformity by iceberg ploughmarks is common on glaciated margins and is considered related to the advancing ice sheets. Drifts show a strong correlation with climatic cycles, with an order of magnitude higher sedimentation rates during the glacial maxima compared with the Holocene, as a result of glacimarine sediment input during glacials. • Episodic turbidity-current activity, most intense during phases of maximum glaciation, interacted with contour-current-steered sedimentation since the Miocene. The turbidite-related overbank deposits perpendicular to the continental slope were often diverted in a more slope-parallel direction by the contour current affecting the distal part of the continental slope and rise. The interaction of these two transport processes formed, in many cases, migrating, mounded deposits. • The drifts, in particular in the Antarctic, are composed of alternating clastics- and biogenic-rich intervals that reflect alternations of glacial and interglacial conditions. • Seismic profiles from high latitudes show an abundance of current-influenced deposits, such as sediment waves and large sediment drifts, while, in addition, large channel–levee complexes are abundant, suggesting deposition by turbidity currents and other mass-flow processes. • An interplay of turbidite and contourite deposition is shown in most of the glacigenic sediments of the high-latitude contourite drifts. Down-slope and along-slope processes thus interacted to form the glacimarine deposits, mostly attributed to the advances of ice sheet, and delivering huge amounts of sediment to the shelf edge and upper slope during glacial maxima.

T. van Weering et al.

489

• A general depositional model (specially for Antarctic Drifts) implies weak SWflowing bottom currents pirating the suspended fine material from turbidity currents flowing during glacial stages in the adjacent channels and down-streamenhanced deposition of the entrained sediment on the gentler, up-stream side of the drift. Slope instability on the steeper, down-stream side then generates a moderate up-stream migration of the asymmetric drift (similarly to a huge sediment wave).

ACKNOWLEDGEMENTS We thank Paul Knutz and Sally Hunter for making original figures available. Tj.C.E. van Weering expresses thanks to Henk de Haas for fruitful discussions. We thank Angelo Camerlenghi for his critical review, patience, and help in editing this chapter.

P A R T

7

IMPORTANCE

C H A P T E R

2 3

E CONOMIC R ELEVANCE OF C ONTOURITES A.R. Viana Petrobras, Research Center, R&D Exploration, Rio de Janeiro, Brazil

Contents 23.1. Introduction 23.2. Major Implications of Contourite Studies 23.3. Contourites and Petroleum Exploration 23.3.1. Coarse-grained contourites: reservoirs 23.3.2. The impact of sea-floor topography 23.3.3. Seismic and well-log characteristics of contourites 23.3.4. Fine-grained contourites: sealing and source rocks 23.4. Discussion 23.4.1. Processes and main depositional elements 23.4.2. Deep-sea unconformities 23.4.3. Trends of future contourite studies motivated by economic significance Acknowledgements

23.1.

493 494 496 498 499 503 504 506 506 509 509 510

INTRODUCTION

Recognition of the economic significance of contourites has only recently been pushed forward due to the continuous advance of hydrocarbon exploration toward deeper waters and the need to better determine the origin of some reservoir deposits lacking what is commonly referred to as ‘‘typical’’ turbidite signature. A probable explanation for the limited number of scientific publications dealing with the economic significance of contourites could possibly reside on the fact that most of the hydrocarbon reserves found in deep waters are related to turbidite reservoirs, and include several giant oil fields. Industrial 3-D seismic data, well logging and coring, and outcrop analyses have supported throughout the years an ample discussion on turbidite facies models (Normark, 1970, 1978; Bouma et al., 1985; Mutti and Normark, 1987, 1991; Normark and Piper, 1991; Mutti, 1992; Mutti et al., 1999, 2003; Piper and Normark, 2001). Unlike turbidites and other gravity-flow-derived deposits, contourites have rarely been associated with hydrocarbon-bearing reservoirs, and have remained confined to a much more academic Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00223-9

 2008 Published by Elsevier B.V.

493

494

Economic Relevance of Contourites

approach based on conventional shallow penetration, high-resolution 2-D seismic data and shallow piston coring. The increasing use by academic institutions of high-resolution swath bathymetry and sea-floor imaging techniques, coupled with the exploitation of industrial 3-D seismic data gathered in deep and ultra-deep waters, now provides a broader understanding of the contourite deposits (Hohbein and Cartwright, 2006; Nielsen et al., 2008) and confirms their role as important constituents of petroleum systems (Viana et al., 2007). In the late 1970s to early 1980s, influenced by the suggestion made by Mutti et al. (1980) that some rippled, top-truncated, fine- to medium-grained sand-rich oilbearing beds of Eocene age in the Campos Basin could be constituted by contourite sands, the interpretation of other similar deposits, mostly in Oligocene–Miocene deep-water sediments, became more frequent in the Campos Basin. Contourite reservoirs were also described from the North Sea (Enjorlas et al., 1986). During the 1990s, papers by Shanmugam et al. (1993a) and Stow et al. (1998a) tried to redeem the importance of sandy contourites as potential hydrocarbon reservoir rocks (see also Shanmugam, 2008). Shanmugam et al. (1993a) presented a detailed study on cores retrieved from a Pliocene–Pleistocene succession of the Ewing Bank area in the Gulf of Mexico. In their study, they discuss the reservoir quality of the traction-dominated deep-water sand deposits and relate their origin to a bottom-current system similar to the present Loop Current. They list a series of sedimentary structures observed in those deposits and consider them as typical of the action of bottom currents. Their analysis indicates very good reservoir characteristics, with porosities ranging from 25 to 40%, and highly variable permeability values, which locally exceed 1.8D.1 Stow et al. (1998a) presented a discussion on fossil contourites, in which they stress the point that several deep-water deposits described in the literature, such as some of those cited by Shanmugam et al. (1993a), failed to present unequivocal characteristics of contourites and would be better considered as fine-grained turbidites. These authors presented a series of criteria, which may help to clarify contourites preserved in the rock record (see also Stow and Fauge`res, 2008). Such criteria have been applied in a number of cases, some of which were presented by Stow et al. (2002f ). However, despite the increasing number of fine-grained contourite systems identified in several deep-water settings (see Fauge`res et al., 1999; Rebesco and Stow, 2001; Stow et al. 2002f, for references therein), the importance given to contourite deposits as potential reservoirs remains low, probably due to the lack of unequivocal criteria for their identification.

23.2.

MAJOR IMPLICATIONS OF C ONTOURITE STUDIES

The petroleum industry is by far the field that concentrates the greatest part of the debates dealing with the economic interest of contourite studies. Nevertheless, in parallel with the petroleum-industry applications, two other main research 1

Permeability can be expressed in darcy (D), where 1 D = 9.86923 10–13 m2 or 0.986923 mm2.

A.R. Viana

495

branches are depicted as having considerable economic impact: climate change and slope stability. Global climate disequilibria induced by the perturbation of ocean circulation can also be evaluated through the study of contourites. Palaeoceanographic oscillations related to climate changes are recorded in contourite deposits (Fauge`res and Stow, 1993; Robinson and McCave, 1994; Knutz, 2008) and may help to design environmental policies aiming to prevent climatic catastrophes affecting ocean and onshore resources. Identifying phases of oceanic current activity and the factors that induce both local- and global-scale variations in thermohaline circulation is fundamental in understanding climate changes. The increasing quality and use of 3-D seismic data introduced an important new aspect in the study of contourites: the analysis of slope stability in areas affected by contour currents (Laberg and Camerlenghi, 2008). Vigorous contour currents can locally erode the sea floor, provoking sediment instability and triggering important mass-flow events (Niemi et al., 2000; Laberg et al., 2005; Viana et al., 2007; Verdicchio et al., 2007). In some cases, sediment erosion by contour currents is responsible for sapping water from eroded aquifers (Duarte and Viana, 2007; Viana et al., 2007; Figure 23.1) or release of methane gas from gas-hydrate accumulations (Holbrook et al., 2002). Thermohaline circulation may also have played an important role in the massive methane releases from marine sediments during the

Figure 23.1 Image of the Mid-Cenozoic sea floor of the Santos Basin, SE Brazil, from 3-D seismic data, showing a mass-flow deposit (mf ) originated by upper-slope destabilization related to sea-floor erosion by the action of a vigorous surface slope-boundary current (sc). Sea-floor erosion induced the outcropping of an aquifer provoking a water sapping front (wsf ) and destabilization of the slope face, transferring sediment down-slope (slope apron), locally reworked by a deeper mid-slope contour current (dc) (modified from Viana et al., 2007, with permission from The Geological Society of London). A multicolour version of this figure is on the enclosed CD-ROM.

496

Economic Relevance of Contourites

Paleocene–Eocene thermal maximum and in the Early Toarcian (Dickens et al., 1995, 1997; Hesselbo et al., 2000). The analysis of climate changes and slope instability due to bottom-current processes – themes of outstanding scientific attractiveness and environmental importance – are the major target of Knutz (2008) and Laberg and Camerlenghi (2008). In contrast, the present chapter focuses on the economic significance of contourites for hydrocarbon exploration.

23.3.

C ONTOURITES AND P ETROLEUM EXPLORATION

The action of contour currents has an impact on the petroleum systems in many aspects, including reservoir geometry and quality, and the distribution of sealing rocks (Viana et al., 2007). The potential of contourites as source rocks is generally low as a consequence of the ventilation (abundance of dissolved oxygen in the bottom boundary layer) of the sea floor, induced by contour currents. The geometry of the continental margins at a basin-wide scale, derived from the tectonic and/or sedimentary evolution of the basin, locally affects the intensity of bottom currents (Figure 23.2) with a direct response in the depositional style (Viana and Fauge`res, 1998; Stow et al., 2002f; Viana et al., 2007). Changes of the sea-floor topography by erosion or deposition induced by bottom currents can result in a re-adjustment of the sediment accommodation space and the creation of sub-basins, which act as sediment traps or gateways for sediment transfer (Figure 23.3). Changes in contour-current flow regime, responding to climatic oscillations or to topographic changes, affect the width and position of the contour-current core. Coarse to fine-grained sediments are common in contourites, independent of water depth. The sedimentary characteristics of contourite deposits are directly related to the type and size of sediments submitted to the action of the currents and to the current regime itself. Spatially, two fining tendencies are commonly observed: (1) in the direction parallel to the slope, accompanying downstream current-intensity changes, and (2) in the direction perpendicular to the slope, related to the width of the current core and to the distance from that core. In both directions, a general decrease in current velocity results in a reduction of grain size of the transported/accumulated sediments (Viana et al., 2007). Alongcurrent re-acceleration frequently occurs, mostly related to physiographic changes, and can induce alternation of zones of sediment transport, deposition or sea-floor erosion. The ‘‘normal’’ behaviour of a deep oceanic current is marked by currentintensity changes along its path as well as over time. Deep tidal imprints and extreme high- to low-frequency oscillations in the current intensity and direction (McCave, 1976, 1986; Nowell and Hollister, 1985; De Madron and Weatherly, 1994; Huthnance, 1995; McCave et al., 2002; Habgood et al., 2003) induce the development of sedimentary structures similar to those found in tidal deposits

497

A.R. Viana

Figure 23.2 Block diagram representing the modern Campos Basin sea-floor physiography. Increase, decrease and trends of bottom-current modifications related to changes in margin physiography and sea-floor topography are indicated by arrows. Size of arrows indicates relative strength of bottom currents. 1= margin protuberance induces upstream increase of current velocity, downstream perturbation and speed decrease; 2 = sea-floor obstacles imply current deviation, locally acceleration; 3 = sea-floor passageways and corridors induce flow intensification. A multicolour version of this figure is on the enclosed CD-ROM.

tc bc

mf bc

Figure 23.3 Schematic representation of local sediment trap-ponded basins created by modifications on the sea floor induced by the action of bottom currents (bc). Gravity-flowderived deposits from turbidity currents (tc) or mass flows (mf ) are trapped into depressions related to current-carved channels or troughs between sediment waves.

498

Economic Relevance of Contourites

(see Shanmugam, 2008, and references therein) or flood-related hyperpycnal flow-derived deposits (in the sense of Mulder et al., 2003a). The distinction of contourites from sediments deposited under such conditions requires extensive sampling, palaeogeographic reconstructions and detailed vertical and lateral faciesassociation analysis.

23.3.1.

Coarse-grained contourites: reservoirs

Hydrocarbon reservoirs must present appropriate petrophysical characteristics, such as high porosity, high permeability and effective lateral and vertical transmissibility of fluids (see also Shanmugam, 2008). These conditions are more typically found in sand-rich deposits. The dominant depositional features associated with the presence of sand in contourites are channels, sheets, furrows, ribbons and patches (Viana and Rebesco, 2007, and references therein). The main factors controlling geometry and facies distribution of contourite systems are (1) intensity and duration of bottom-current regime; (2) grain-size population available to the current action; (3) sea-floor physiography; and (4) margin configuration. These factors regulate where, when and how contourites are deposited in a sedimentary basin (Viana et al., 2007) and, more specifically, they determine the locus of sand deposition in a contourite system. The persistence and efficiency of a hydrodynamic regime capable of eroding, transporting and re-depositing a large volume of sand is fundamental, as reported for the Carnegie Ridge (Lonsdale and Malfait, 1974), the Campos Basin upper slope (Viana, 1998), the lower Mississipi Fan (Kenyon et al., 2002), the Gulf of Cadiz (Habgood et al., 2003) and the Faroe–Shetland Channel (Wynn et al., 2002a; Masson et al., 2004; Akhmetzhanov et al., 2007). Two main situations are considered as favourable to the supply of coarsegrained sediments and the action of a contour current. First, the vicinity of a preexisting sand-rich accumulation whose content is partially transferred to the areas subjected to current action by mechanisms such as shelf currents (currents induced by storms, tides, trade winds, eddies) inducing sand overspill (Viana and Fauge`res, 1998); sediment density flows related to catastrophic floods in a low-gradient, ramp-like margin (hyperpycnal flows in the sense of Mulder et al., 2002); and the downwelling of dense saline waters such as those occurring in the Gulf of Cadiz (Habgood et al., 2003; Mulder et al., 2002) and in the Gulf of Lyon and adjacent areas (Canals et al., 2006). Second, zones subjected to deformation and movements (elastic rebound, thermal deformation, halokinesis, intraplate stresses, etc.) where pre-deposited sand layers are exposed at the sea floor to sweeping bottom currents (Fauge`res et al., 1999; Davies et al., 2004). In both cases, there is no restriction to water depth. In the Middle Cenozoic upper slope setting of the Santos Basin, SE Brazilian margin, strong NE flowing bottom currents encountered an E-W slope escarpment, reworked coarse-grained shelf-derived sediments at the base of the escarpment and developed ENE migrating sand dunes (barchans, barchanoid ridges and an impressive slope-parallel sand-rich channel–levee complex) (Figure 23.4; see discussion in Viana et al., 2007). The sand dunes identified in the uppermost

499

A.R. Viana

Figure 23.4 Seismic amplitude horizon slice of the Mid-Cenozoic sand-rich shelf edge/upper slope of the Santos Basin. This setting is influenced by the action of a surface slope-boundary current (sc), which erodes the uppermost slope and helps to transfer sand down-slope. Barchan dunes and barchanoid ridges are developed above a mid-slope terrace (inset a), swept by midslope contour currents (dc). These bedforms evolve downstream into a slope-parallel contourcurrent-carved channel (inset b). Arrows in the insets indicate the sedimentary features. Schematic block diagram at the top left illustrates these processes. (Modified fromViana et al., 2007, with permission from The Geological Society of London). A multicolour version of this figure is on the enclosed CD-ROM.

Paleocene/lower Eocene section of the Santos Basin indicate that intense bottomcurrent activity occurred concurrently with delivery into the basin as gravity flow of coarse-grained sediments. Extensive fields of barchan dunes, sand waves and mega-furrows generated by bottom currents flowing about 100 cm 1 are also found on the modern sea floor of the Campos Basin (Figure 23.5) and Gulf of Mexico, in water depths ranging from 2000 to 3000 m, (Bryant et al., 2000; Scott et al., 2001; Kenyon et al., 2002; Niedoroda et al., 2003; Bryant and Slowey, 2004).

23.3.2.

The impact of sea-floor topography

The physiography of the continental margin is extremely important in accelerating or decelerating contour currents. Prominent seaward projections, or regions of

500

Economic Relevance of Contourites

Figure 23.5 Side-scan mosaic (SYS3000) of the lower slope setting in the Campos Basin sector, SE Brazil, showing the development of barchan dunes (d), barchanoid and linear ridges and furrows (f ) related to the vigorous flow of bottom currents. Mosaic processed by R.C. Rechden. A multicolour version of this figure is on the enclosed CD-ROM.

disturbed relief by tectonic or adiastrophic structures, act as current restrictions, enhancing the near-bottom current velocity, whereas slope embayments and canyons reduce the current velocity and induce sediment settling (Figure 23.2). The sea-floor topography of continental slopes and rises of Atlantic rifted margins is commonly marked by salt walls, fault planes, knolls, mud/salt diapirs, submarine volcanoes, escarpments, etc. The size of such structures and their orientation concerning the current interfere with the bottom-currents pattern, either accelerating or reducing the current velocity. Under a regime of strong bottom currents, coarsegrained material introduced into the system or eroded by the currents from previously deposited sand bodies may be incorporated into flows as bedload. A wide range of bedforms and depositional geometries can be developed as a function of the intensity and duration of the flow regime and the shape of the accommodation space. The intensification of current velocity produced by the sea-floor relief and the resultant coarse-grained deposits are well recorded in many continental margins, and particularly in the Northwestern Atlantic and in the SE Brazilian margin. Between the Faroe Bank and the Faroe Bank Channel, a 400-m relief wall induces the acceleration of the Norwegian Sea Outflow Water, resulting in sea-floor erosion, deposition of elongated sand bodies and carving of large-scale bottom lineations (Akhmetzhanov et al., 2007). In the SW Atlantic, salt walls and diapirs create a complex sea-floor topography. These features can locally induce the acceleration of the bottom currents which erode and transport sediments along their path. Contourite sand sheets and ribbons passing to muddy furrows and sediment waves away of the sand source are observed in a region where uplifted Late Oligocene to Early Miocene turbidite sands crop out at the modern sea floor (Figure 23.6; Viana, 2001; Viana et al., 2007). Also, the axes of some salt minibasins acted both as fairways to gravity flows transferring sediment to deep waters and as bottom-current corridors (Figure 23.7).

501

A.R. Viana

Figure 23.6 Perspective view of the modern Campos Basin sea bottom based on 3-D seismics. Topographic highs are associated to salt halokinesis. Colours are related to seismic amplitude, corresponding to grain-size differences ground-truthed by piston cores. Blue colours (dark on the topographic highs) correspond to fine-grained sediments, and yellow and red warm colours in the basin (yellow and red) to sand. In particular, the red (dark) colour in the basin corresponds to sand sheets and ribbons (see multicolour version of this figure on the enclosed CD-ROM).

Sediment overflow (so)

Sediment avulsion (a)

Main sediment conduit

Figure 23.7 Schematic diagram representing upper margin sediment delivery through fairways developed along salt crestal grabens (scg). Sediment overflow (so) and avulsion (a) feeding the axis of salt withdrawal mini-basins are also represented. Sediments are reworked by bottom currents (SOC = Southern Ocean Current of Viana et al., 2002b) and redistributed along current-generated furrows (F). Diagram based on 3-D seismic data presented by Viana et al., 2007. A multicolour version of this figure is on the enclosed CD-ROM.

The effect of the regional-scale configuration of the slope on bottom-current intensity changes is demonstrated in the Miocene to Recent slope sedimentation in the South China Sea (Zhu et al., in press), where successive canyon excavation, migration and infilling are strongly influenced by the interplay between gravity and regional bottom currents. Similarly, in the Aptian–Albian section of the Montagne de

502

Economic Relevance of Contourites

la Lance, Vocotian Basin, SE France (Viana, 1998), a large sand-waves field was developed above an erosional terrace excavated by the bottom currents in an upperslope setting. The sediments transported by the migrating dunes reached the heads of a submarine canyon and were transferred down-slope by episodic gravity-flow events. Although the great majority of sandy contourites have been described from midwater depth zones (300–2000 m; Viana et al., 1998b), sandy contourites are not directly related to water depth. Studies from different geographic and stratigraphic settings (Mutti et al., 1980; Barros et al., 1982; Shanmugam et al., 1993a; Viana, 1998; Viana and Fauge`res, 1998; Parize et al., 1998a, b; Viana et al., 1998b, 2007; GarciaMojonero and Olmo, 2001; Kenyon et al., 2002; Cakebread-Brown et al., 2003; Rodrı´guez and Anderson, 2004; Duarte and Viana, 2007; Akhmetzhanov et al., 2007) indicate that the occurrence of potentially economic sandy contourites is mostly dependent on the vicinity of a sand-rich area prone to be swept by contour currents. The seismic-amplitude map of Figure 23.8 illustrates these ideas. A sub-cropping Late Pleistocene sandy fan complex at the foot of a slope was reworked by slope contour currents whose intensity was enhanced due to the topographic restriction created by the base of slope escarpment. The fringe of the fan was eroded and sand patches were developed following the trend of local bottom currents.

Complex of avulsion lobes

Base of slope

N

400.00

7 km

600.00 800.00

1000.0

400.00 600.00 800.00 Contour current reworked deposits detached from complex of avulsion lobes

1000.0

Figure 23.8 Seismic-amplitude image of a subcropping turbidite/contourite complex. Warm colours (yellow and red) correspond to coarse-grained sediments, and green to fine-grained sediments (see multicolour version of this figure on the enclosed CD-ROM). Two systems are observed: one constituted by low-sinuosity channels flowing along a down-slope oriented fairway crossing diagonally the image from NW to SE (the blue arrow indicates the North). Laterally to this system, to the South, occurs a fan-shaped body corresponding to a complex of avulsion lobes, distally reworked by contour currents developing a fringe of irregular patch-like small sedimentary bodies.

503

A.R. Viana

23.3.3.

Seismic and well-log characteristics of contourites

Recognition and characterization of the depositional elements of contourite systems based on seismic data have been presented by Fauge`res et al. (1999). Seismic and well-log characteristics of coarse-grained contourites with high reservoir potential have been discussed by Viana and Rebesco (2007). Strong seismic amplitudes and amplitude-versus-offset (AVO) anomalies have been reported from high-porosity (30%), gas-bearing Early Pliocene contourite sands, in the Algarve Basin, Gulf of Cadiz (Cakebread-Brown et al., 2003). In the Gulf of Cadiz, Garcia-Mojonero and Olmo (2001) tied seismic lines to one exploratory well and identified a more than 800 m thick Late Pliocene–Quaternary contourite deposit, with an average net-to-gross total sand content of 75%, and sand beds up to 40 m thick with good lateral continuity and porosities beyond 30%. Industrial 3-D seismic data, shot along the Brazilian SE margin, coupled with well data and piston cores, were used to characterize the distribution of upper-slope Middle/Late Miocene contourites in the Campos Basin (Viana et al., 2007). Strong negative amplitudes (Figure 23.9), corresponding to sand deposits, occur at the top of the slope as an over 10-km long and 3–5-km wide field of linear, parallel-to-the-slope ridges. The decrease in amplitude intensity corresponds to a decrease in grain size. Erosional features are frequent in this system. Hummocky clinoform seismic patterns were associated with the migration of the bedforms. Stacked similar features are reported, suggesting that such a sandrich contourite system was maintained during a relatively long interval (>1 Ma), alternating with continuous low-amplitude, high-frequency reflectors, interpreted as representing phases of fine-grained deposition, related to the weakening of the current regime. The stacked pattern of intercalated sand-rich and mud-rich intervals in upper-slope contourite deposits is also recorded in well-log profiles (Viana et al., 2007). The gamma-ray signature corresponds to two transitional styles: an upward grading well-sorted sand (‘‘cleaning trend’’: g-ray values decrease towards the top) and a blocky pattern of clean sands. The most common association is of a cleaning trend at the base and sediments becoming blockier at the top (Figure 23.9). These styles occur both at the bed scale (10 1–100 m thick) and in a 4th–5th order depositional sequence scale (100–102 m thick). The resistivity and sonic logs have a similar signal behaviour, suggesting a frequent increase in the reservoir quality towards the top, mostly marked in medium to distal areas. Such an increase in quality occurs both at the bed scale and at sequence scales. In the example from the Campos Basin, isopach maps corroborated by aquifer and oil-reservoir pressures indicate that the reservoirs are apparently continuous for several hundreds of meters both downstream and down-slope. Shallow and deep-water contourites show similar seismic and physiographic characteristics, which may imply a probably similar pattern of upward increase in reservoir quality. The transition from a poor- to a good-quality reservoir depends on sediment availability and, mostly, on the bottom-current regime. It responds to long-term current action transferring and accumulating progressively coarse-grained, well-sorted sediments. High-frequency oscillations, mostly observed in sonic logs, are assumed to correspond to local current oscillations, resulting in small-scale variations in grain size. Stratigraphic traps are the most recurrent in contourite systems, which requires a good knowledge in sedimentology of the geoscientists

504

Economic Relevance of Contourites

Figure 23.9 Physical properties of a contourite deposit on the upper Miocene middle slope of the Campos Basin (from Viana et al., 2007; with permission from The Geological Society of London). Data indicate a down-slope decrease in sand content and the intercalation of sand and mud deposition phases in the system. (a) Seismic-amplitude map of the basal erosional surface (bes). (b) Dip seismic line crossing three wells. (c) Close-up of the seismic line at the distal site with superposed well logs. (d) Well logs of the three wells along the dip line across a contourite deposit on the upper Miocene middle slope of Campos Basin. A multicolour version of this figure is on the enclosed CD-ROM.

involved in the exploration. Structural closure and a frequent association with localized deep-seated faults are also important elements in trapping and oil migration.

23.3.4.

Fine-grained contourites: sealing and source rocks

Strong bottom currents are local and short-lived phenomena. The dominant circulation pattern on slope and basin settings comprises low- to medium-intensity currents, capable of transporting a fine-grained sediment population. Such sediments are introduced into the basins by hypopycnal flows associated with river discharge (plumes), sea-floor erosion, sediment gravity flows and primary productivity close to upwelling zones, and then incorporated into the oceanic circulation regime. Thick and widespread fine-grained drifts are recognized along all the oceanic basins (see Stow and Fauge`res, 1993, 1998, and Stow et al., 2002f, for references therein).

505

A.R. Viana

Fine-grained contourites play an important role in the characterization of deepwater petroleum systems. They are related to both sealing facies/permeability barriers (Moraes et al., 2007) and source-rock accumulation. In the Santos Basin, the over 600-m thick Neogene Santos Drift acts as an excellent sealing rock for Paleogene oil-bearing sandstones (Duarte and Viana, 2007). In the Campos Basin, a bottom-current origin is advocated for the thick shaly-marly Mid/Late Miocene wedge which overlies the sand-rich Late Paleogene/Early Neogene section where several giant oil fields have been discovered (e.g. the Marlim, Albacora and Barracuda oil fields; Souza Cruz, 1995, 1998). Such a wedge works as an excellent seal. Less frequently documented, the internal heterogeneities in oil-bearing reservoirs that can be related to the action of bottom currents constitute an important issue in appraisal projects. One of the reservoirs studies performed in a giant oil field in the Campos Basin indicates that several meters thick packages of oil-rich sandstones deposited in structural troughs are separated by extremely bioturbated silts, interpreted by Moraes et al. (2007) as being the product of bottom-current reworking along the trough (Figure 23.10). These fine-grained sediments act as important permeability barriers, and the understanding of their distribution and thickness is fundamental in water-injection and recovery projects.

NE

SW 0 25 50 75 100 m

0 Turbidite channels

500

1000 m

Bottom current deposits

Third order sequencle boundaries

Figure 23.10 Depositional model for Palaeocene ^ Eocene deep-water, trough-filling sandstones of the Campos Basin, elaborated by Moraes et al. (2007; with permission from The Geological Society of London). The model indicates the distribution of bottom-current deposits in a turbidite channel complex, showing typical facies proportions and dimensions inferred for the sandstone bodies. Vertical connection of sand bodies is proposed to be diminished by the presence of contourite bodies. Dimensions of turbiditic channels and contourites are based on well observations and analogue data. A multicolour version of this figure is on the enclosed CD-ROM.

506

Economic Relevance of Contourites

Organic-matter deposition and preservation require hydrodynamic conditions at the sea floor that do not match strong bottom currents. However, contourite deposits have been frequently associated with gas-hydrate accumulations. Large gashydrate accumulations underlined by a diagnostic bottom-simulating reflector (BSR) are found in contourite deposits all along the Atlantic margin (Blake Outer Ridge – BOR: Dillon and Paull, 1983; Kraemer et al., 2000; Rio Grande Cone: Silveira and Machado, 2004). Gas-hydrate accumulations are favoured by a high sediment permeability. The BOR is a large, clay-rich contourite drift deposited in the western Atlantic Ocean that contains 30–40 millions of tons of carbon stored as methane hydrate and underlying free methane gas (Paull et al., 2000). Heterogeneities observed in the gas-hydrate distribution on the BOR are related to climate-induced modifications in the thermohaline circulation pattern which induced different styles of drift development (alternating accumulation of terrigenous and biogenic components; Kraemer et al., 2000).

23.4. 23.4.1.

D ISCUSSION

Processes and main depositional elements

Sands deposited in an outer-shelf setting are frequently subjected to very dynamic hydrographic regimes involving storm fronts, tidal currents, cascading of cold waters (downwelling) and on-shelf penetration of superficial slope boundary currents, meanders and eddies. These processes can induce vigorous off-shelf currents that can erode the sea floor and/or transport sand. The overspilled sands are then entrained into the along-slope circulation system. Depending on the strength of the hydrographic regime, the sediments can be incorporated into the off-shelf water flow and redistributed parallel to the isobaths, developing different types of sandrich contouritic deposits. In the upper-slope setting, sandy contourites responding to these forcing mechanisms were recognized in the Aptian–Albian series of the Vocontian Basin, SE France (Parize et al., 1998a, b; Viana, 1998), in the modern western Antarctica outer-shelf/upper-slope (Rodrı´guez and Anderson, 2004) and in the Miocene successions of the Campos Basin (Viana et al., 2007). Several other systems presented in the literature with somewhat different interpretation could be re-interpreted as shallow-water contourites (e.g. the upper-slope setting of the Brushy Canyon as presented by Mutti, 1992; the western Grand Banks of Newfoundland studied by Dalrymple et al., 1992; Verdicchio and Trincardi, 2008a). Migrating sands on the upper slope can feed canyon heads by the accumulation of relatively unstable sand wedges (e.g. the migrating sand dunes in the south-east African continental margin under the action of the Agulhas Current penetrating onto the shelf, described by Flemming, 1978, 1980, and Ramsay, 1994; and on the upper slope of the Campos Basin, SE Brazil, described by Viana and Fauge`res, 1998; Viana et al., 2002a, b). Once reaching the canyon heads, such unstable sand deposits can be easily transferred down-slope through the canyon axis. Transfer mechanisms include the progressive increase of the lee-side gradient of the

507

A.R. Viana

drift-related prograding wedges (Viana, 1998), the continuous action of internal waves and tides (Cacchione and Drake, 1986; Bogucki et al., 1997; Cacchione et al., 2002; Lima et al., 2007), sediment re-suspension by the action of storm currents (Fukushima et al., 1985; Baltzer et al., 1994), cascading of dense clear waters in oceanic water masses fronts (Canals et al., 2006), earthquakes and the connection of the canyon head to a fluvial-deltaic system. In deeper waters, contour currents interact with pre-existing coarse-grained deposits in several ways, such as sweeping the fine-grained population, partially or totally eroding the previous deposits (Stanley, 1993), modifying the original depositional characteristics or transferring sediments to a new depocentre, where coarsegrained contourites can be accumulated and preserved. Geometry, permeability (K) and porosity () of the original deposits are consequently affected (Figure 23.11a, b). The NE Atlantic is rich in examples such as those from the Faeroe–Shetland Channel, the Faeroe Bank and the northern Rockall Trough area, where barchan dunes, furrows, contourite sand sheets and channels were identified and analyzed by Cochonat et al. (1989), Masson (2001), Masson et al. (2002, 2004), Wynn et al. (2002a) and Akhmetzhanov et al. (2007). Those deposits

Reworking gravity flow deposits

st vfs ms

Time Persistency of the current regime and sand availability

Construction of contouritic deposit

Figure 23.11 Sedimentary characteristics of coarse-grained deposits affected by bottom currents. (a) Model of the construction of coarse-grained contourite bodies as a function of longevity of a bottom-current regime and persistence of intensity required to erode and transport the sand available in the system, originated from previous sand-rich gravity flows. (b) Model of general petrophysics characteristics of deep-water, coarse-grained deposits varying from those derived from pure gravity-flow deposits (facies D1) to those totally reworked by bottom currents (D4). Facies D2 and D3 correspond to continuous increase in the bottom-current strength and persistence. A multicolour version of this figure is on the enclosed CD-ROM.

508

c.sand

f.sand

silt

c.sand

f.sand

silt

Economic Relevance of Contourites

D3 - Medium intensity bottom currents D1: Gravity flow deposit,

Medium to long duration

normal grading

D3: Φ good < top (better than D1 e D2), K good

Φ < top, K regular

Good – very good reservoir

Regular – good reservoir

D4 - Strong bottom currents D2 - Weak bottom currents, short duration

Figure 23.11

Long duration

D2 ~ D1: Φ < top, K regular

D4: Φ = top and base, K good to excellent

Regular – good reservoir

Excellent reservoir

(Continued)

were reported as being the result of the action of the southward flow of cold deep water from the Norwegian Sea funnelled through narrow topographic passages that extend from Greenland to Scotland, and the incorporation into the flow of coarse-grained sediments derived from shallower settings. In the Gulf of Mexico, the reworking by bottom currents of the Mississipi deep-sea fan gives way to sand-dune fields (Kenyon et al., 2002). In the western Gulf of Mexico, a giant field of mega-furrows has developed under the vigorous action of the westward branch of the Loop Current (Niedoroda et al., 2003; Bryant and Slowey, 2004). The erosion of salt knolls and diapirs favours the incorporation of dissolved salt, locally resulting in an increase of the flow density, which is consequently accompanied by an increase in erosion and transport capacity. Density-enhanced bottom currents are expected to occur in other salt-deformed basins where evaporitic structures crop out at the sea floor, such as those occurring at both sides of the South Atlantic (Viana et al., 2007), or in areas with great exchange of heat or salt, as exemplified by the Mediterranean Outflow Water penetrating into the Gulf of Cadiz (Nelson et al., 1993; Habgood et al., 2003; Mulder et al., 2002; Llave et al., 2007; Herna´ndez -Molina et al., 2008b), or the downwelling of cold shelf waters as observed in the Gulf of Lyon (Canals et al., 2006). In all studied cases, the presence of coarse-grained contourites in deep water is commonly associated with a set of medium- to large-scale bedforms and depositional geometries. Masson et al. (2004) proposed the following sequence of depositional geometries to be formed in response to increasing current velocity: fine-grained sediment waves and mounded contourite deposits, contourites and

509

A.R. Viana

sheets, coarse-grained sediment waves, barchan-like dunes, sand ribbons, channels, comet or obstacle marks, furrows and erosional scours.

23.4.2.

Deep-sea unconformities

Diachronous, regional unconformities found at the base of slope-plastered or detached contourite drifts (the basal erosional surface of Fauge`res et al., 1999) are frequently overlain by onlapping upslope migrating layers. Interpretation of these erosional features as sequence boundary unconformities, in the sense of Vail et al. (1977b) and Posamentier and Vail (1988), is not rare, driving to sometimes erroneous fitting to relative sea-level positions. However, the large regional extent and abrasive aspect of these unconformities, presenting along-slope trending rather than down-slope scouring characteristics, suggest a contour-current-related erosion [examples from the South Atlantic Brazilian margin may be found in Viana (2001), Gomes and Viana (2002), Lima (2003) and Duarte and Viana (2007)]. Physical and temporal coincidence between some basal contourite-drift unconformities and depositional sequence boundaries is also probable (Santos Drift: Duarte and Viana, 2007). The upslope migrating package, a common feature on separated or plastered drifts (Fauge`res et al., 1999), is occasionally misinterpreted as lowstand prograding wedges in the sense of their internal reflection pattern and sequence-stratigraphic position. These two different interpretations lead to different interpretations of sedimentary facies and can make any exploratory prospects rewarding or not (Viana et al., 2007).

23.4.3.

Trends of future contourite studies motivated by economic significance

In a recent article, Viana et al. (2007) claimed that the identification of some deepwater coarse-grained deposits as contourites still remains controversial. Based on a large quantity of published research articles, they proposed that a ‘‘gravity-flow culture’’ prevails over the idea of contour-current-dominated processes in the interpretation of many depositional systems in oceanic settings. They listed three main aspects as the major causes of this disequilibrium: (1) the early establishment of conceptual models proposing facies characterization and geometry of the gravity-flow accumulations; (2) gravity flow deposits of sand are volumetrically predominant at least in the modern ocean; and (3) most sandy contourites involve upper-slope sediment transport, which was a poorly investigated aspect of oceanography from the 1960s to the 1980s. The increasing access of 3-D seismics by academic institutions and the high quality of the new tools developed to investigate the deep-sea bottom provide a new data base that is enlarging the knowledge of how bottom currents interact with the sea floor and what is the result of such processes. Coarse-grained contourites are being recognized in different types of margins and bathymetric settings, and their economic potential is only in the primeval reconnaissance stages. The continuous development of numerical models dedicated to reservoir appraisal requires a better understanding on sealing capacity

510

Economic Relevance of Contourites

and fluid baffles/obstacles induced by fine-grained deposits. New studies aimed at climate forecasting based on the palaeoceanographic record, and at CO2 sequestration and re-injection in deep reservoirs, also require feedback from contourite studies, which can be fundamental in defining the critical limits of stocking such fluids. Again, the understanding of bottom-current behaviour in moulding deepwater deposits is critical and much effort remains to be done in this field, demanding many years of research by the next generations.

ACKNOWLEDGEMENTS This approach on the economic importance of contourites would not have been possible without the encouragement over many years and the wide geological sensibility of many of my Petrobras colleagues. Discussions with J. Heinerici, M. Carminatti, W. Almeida Jr, L.C. Machado, M. Moraes and C. Duarte were ever fruitful. Renato Kowsmann and Jean-Claude Fauge`res are specially thanked for their endless examples of motivation, scientific curiosity and friendship. Michele Rebesco is particularly thanked for his encouragement, patience with the delays and suggestions to improve the original manuscript. Angelo Carmelenghi’s review added much to the final format of the text. Angelo and Michele are also thanked for inviting the author to publish this contribution. C. Silva is thanked for his comments and suggestions. C.S. Lamb is deeply thanked for the careful review in English. Petrobras granted permission to publish this paper. Veritas granted permission to publish seismic data of Figures 23.1 and 23.4. The Geological Society of London kindly allowed the reproduction, with minor modifications, of some figures published in Viana and Rebesco (2007).

C H A P T E R

2 4

P ALAEOCEANOGRAPHIC S IGNIFICANCE OF C ONTOURITE D RIFTS P.C. Knutz Geological Survey of Denmark and Greenland (GEUS), Øster Voldgade, Copenhagen, Denmark

Contents 24.1. Introduction 24.2. Oceanographic Settings of Contourites 24.3. Methods and Approaches 24.3.1. Palaeoceanographic tracers 24.3.2. Non-destructive techniques 24.3.3. Seismic imaging 24.4. Palaeoceanographic Themes Addressed by Contourite Research 24.4.1. Gateways, tectonics, and ocean circulation 24.4.2. Ocean circulation during warm climate extremes 24.4.3. Rapid ocean-climate variability in the North Atlantic 24.5. Summary Acknowledgements

24.1.

511 512 515 515 520 521 524 524 528 529 534 535

INTRODUCTION

The discovery of contourite deposits formed in the path of deep-sea currents (Heezen et al., 1966) has been fundamental for the present understanding of the oceans’ climate history and its role in Earth’s evolution. Contourite drifts and their bounding unconformities are the result of persistent geostrophic currents that, by interaction with sea-floor topography and available sediment sources, are capable of eroding, transporting, and depositing fine-grained particles into mounded and elongated sediment bodies (Hollister and Heezen, 1972; Kidd and Hill, 1986; McCave and Tucholke, 1986). The enhanced accumulation rates of these drifts, in contrast to juxtaposed condensed pelagic sequences, make them attractive for palaeoceanographic studies. Palaeoceanographic research spans across a wide range of time scales, tectonic, orbital, or centennial, depending on the processes that are being studied. Arguably, the prime interest in contourites is for their potential for high-resolution palaeoceanographic reconstructions along pathways of major oceanic currents. The global transport of water masses within and between ocean basins involves large-scale Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00224-0

 2008 Published by Elsevier B.V.

511

512

Palaeoceanographic Significance of Contourite Drifts

exchange of heat and salt, driven by lateral thermohaline density gradients and the wind-driven surface currents of the oceanic gyre systems (Stommel and Arons, 1960; Wunsch, 2002). Continuing efforts in extracting detailed palaeoclimate information from contourite drift deposits are motivated by the need for a better understanding of the oceans’ role in the global climate system (e.g. Broecker, 2003). Comprehensive studies of contourite systems are determined (and limited) by technological means and operational costs, as well as the boundaries between established scientific disciplines. Examples of parallel efforts in palaeoceanographic research are studies aimed at depositional processes and basin evolution, and studies where a quantitative understanding of past oceanographic conditions is the main goal. Accordingly, marine geophysicists and sedimentologists tend to focus on three-dimensional depositional patterns, based on acoustic-seismic mapping, while palaeoclimatologists are motivated by time-series reconstructions of water-mass properties based on faunal/geochemical proxy data. In addition, numerical-modelling studies that over the recent years have brought significant impetus to palaeoceanographic research add another disciplinary tool. This chapter gives an update of recent methods and advances across the spectrum of disciplines that are relevant for contourite-based palaeoceanography. The broad range of time scales and earth-system processes that may be addressed by contourite research is highlighted by three general themes: oceanic gateways, past warm climate extremes, and North Atlantic climate variability.

24.2.

O CEANOGRAPHIC SETTINGS OF C ONTOURITES

The bottom currents that form contourites can be generated by a variety of oceanographic processes involving tidal forces, wind-driven transport, upwelling/ downwelling along shelf margins, and/or interhemispheric water-mass exchange associated with Atlantic Meridional Overturn Circulation (AMOC). As such, contourite environments are not restricted to deep ocean basins but can also form on continental shelves, in fjords and lakes, at dimensions several orders of magnitude smaller than the largest deep-sea drifts (Rebesco and Stow, 2001). A common feature of bottom currents is that they flow along topographic contours in response to Coriolis diversion (to the right in the northern hemisphere, left in the southern hemisphere), which is distinctly different from gravity-driven down-slope currents of resuspended sediments (Stow, 1979). Given the existence of a bottom-current system that can winnow, transport, and deposit fine-grained sediments, the morphological expression of large drifts reflects pre-existing topography and proximity of sediment sources. In addition to these physical controls, the presence of benthic ecosystems (e.g. corals, sponges, algae) that thrive in fast flowing, nutrient-rich water masses may help to increase sediment stability and thus enhance build-up of drift crests. For younger Cenozoic drifts, it is often possible to infer the depositional environment in terms of approximate water depth, hydrography, and climate regime, based on the knowledge of present oceanographic and geological setting (modern analogues) and seismic correlation to nearby core sites. Elucidating the depositional environment and oceanographic context of deeply buried drifts, or

P.C. Knutz

513

ancient exposed contourite formations subject to tectonic displacement, is more complicated (Stow et al., 1998a; Hu¨neke and Stow, 2008). Several legs of the Integrated Ocean Drilling Program (IOPD; previously DSDP and ODP) have targeted the giant contourite drifts that decorate the continental margins of the Atlantic and Southern Oceans (Figure 24.1), with the upper Cenozoic interval, typically reaching 20–40 Ma back in time, as a common objective. While early borehole studies were based on fragmented, low-resolution records, the development of the advanced piston-coring facility has enabled the subsequent recovery of detailed sections from unconsolidated muddy sediments. The majority of drilling campaigns aimed at contourite archives have been carried out in the Atlantic basins, in part due to the abundance of large drifts deposited by the prevailing western boundary currents (McCave and Tucholke, 1986). The most recent deep drilling in a contouritic setting was IODP Expedition 303 on the Eirik Drift south of Greenland (Channell et al., 2006). Other major drift systems have been identified in the Southern Ocean and along the western fringe of the Indian Ocean (e.g. Uenzelmann-Neben, 2002). In particular, the bottom component of the Antarctic Circumpolar Current (ACC) is associated with giant mounded drifts developed along the base of the Tasman Rise and the southern New Zealand Margin (McCave and Carter, 1997; Carter et al., 1999). Sedimentary environments promoting build-up of contourite drifts are much less common in the Pacific Ocean, due the weakly developed geostrophic currents and lack of major fluvial sources of fine-clastic sediments along the eastern Pacific margins. Sediment drifts have been reported on the northwest Pacific fringe and in the Sea of Okhotsk, but little is known about the sedimentary environments and palaeoceanography in this region (Wong et al., 2003). High-accumulation-rate sediment drifts off the Chilean Margin were drilled during ODP Leg 202, providing records of glacial variability, ocean-circulation changes (Lamy et al., 2004), and geomagnetic variability (Lund et al., 2007). Sedimentation patterns in the Arctic Ocean are only known from widely spaced seismic profiles, but results collected over the last few years suggest that deep currents have influenced deposition along and across the Lomonosov and Alpha Ridge systems ( Jakobsson et al., 2007). In addition to an unfavourable oceanographic and depositional setting, the apparent absence of contourite drifts in many oceanic regions may simply reflect scarcity in sea-floor sampling and geophysical data. Continental margins influenced by gravity-driven re-sedimentation were for many years avoided by palaeoceanographers, due to the risk of record disturbance caused by turbidites and other down-slope processes. More recently, shallow coring and scientific drilling on continental margins has been seen as essential for gaining access to ultra-high-resolution palaeoceanographic records. Examples of such highaccumulation-rate settings are contourite drifts extending from deep-sea fans, where large down-slope fluxes of fine-clastic sediments enter the regime of along-slope currents, or contourite deposits infilling slide scars and depressions generated by mass wasting (e.g. Knutz and Cartwright, 2003). Sediment focusing conditioned by large fine-clastic fluxes and sea-floor topography (e.g. accommodation space) allow fast burial, to depths beyond the zone of bioturbation, of in situ microfossils that contain the primary source of palaeoceanographic information.

514

80° N

Arctic Ocean

643–644

North Atlantic Gateways

60°

646 113 112 192 882–884

1276 36

40°

1208

35

32

Shatsky Rise 67

1202

Kuroshio Current

20°

407 352 408 336 915-919 983–984 114 115 552–554 403–404 980–981 406 647 610 611 111

43

398 603 991–997 385 388 1063 1054–1062 387 1064 102–106 386 Bermuda 533–534 Rise Cayman 99–101 28 672 Rise 998 543

138 368

659

367

64 287 815

–20°

660

Northwest Atlantic Sediment Drifts

66 65

241 288

Atlantic Ocean

Pacific Ocean

242

286

1192–1199

515

Marion Plateau

248

Indian Ocean

243 244 245

249 1124

–40° Tasmanian 1168 Gateway 1169–1171

Chile Margin

1123

1172 1120 1121

1122

862–863

Southwest Pacific Gateways

361 1089

331 514 513 327 512 328 702 511 329 698–699 701

1090

246–247

Southern Ocean

703 1093

751 326

745–746

–60°

269 1165

Cooperation Sea

322

Southern Ocean

268

325

323

1101

697

1095 1096

324

90°E

120°

150°

180°

210°

240°

Antarctic Peninsula

270°

300°

330°

0°

30°

60°

Palaeoceanographic Significance of Contourite Drifts

289

0°

515

P.C. Knutz

Figure 24.1 Map (adapted from Rebesco, 2005) of ODP, IODP, and DSDP sites drilled in contourite deposits of the world’s oceans. Data sources: 1998 to present from ODP web site http://www-odp.tamu.edu/sitemap/sitemap.html; before 1998 from Stow et al. (1998a). Note

Calcareous microfossil shells of foraminifers are often used as the primary archive for palaeoceanographic studies. In mid- to low-latitude regions, contourite drifts tend to be carbonate-rich as they primarily accumulate on the continental slope and rise above the Calcite Compensation Depth (hence, dissolution of calcareous shells is rarely a problem). However, certain depositional environments may be unfavourable for foraminifers, such as low-salinity continental shelves and fjords, and deep-sea fans with exceptionally high fine-clastic input rates. In non-carbonate marine depositional environments of the polar and sub-polar regions, sea-surface conditions may be estimated using siliceous microfauna (e.g. diatoms) or dinocyst assemblages (de Vernal et al., 2000). Due to the susceptibility of fine biogenic particles (<100 mm) to reworking and longdistance transport by bottom currents, nanno- and microfossil stratigraphy should be interpreted with some caution in contourite sediments.

24.3.

M ETHODS AND APPROACHES

Information on ocean circulation and climate history can be extracted from contourite deposits using a wide range of research approaches. These can be generalized as (1) discrete sampling analyses by application of geochemical, faunal, sedimentological, or geophysical techniques, (2) non-destructive techniques based on visual description, or continuous geophysical-chemical logging, and (3) seismic imaging to determine drift morphology and sedimentation patterns. Dating methods and the limitations of geochronology on different time scales are common to other types of marine deposits and therefore not explicitly treated in this chapter. For a comprehensive treatment of radiometric and relative dating techniques, see Bradley (1999).

24.3.1.

Palaeoceanographic tracers

A large suite of proxy variables (Table 24.1) is available to study palaeoceanographic changes over time scales that may range from 102 to 106 years. Estimates of sea temperature and salinity are important as they determine palaeoceanic density fields that allow direct comparisons with ocean-climate models. Besides temperature and salinity, proxies may target parameters like biological productivity, nutrient content, carbon-dioxide concentration, current velocity, and sediment fluxes/sources. For a more in-depth discussion of the most common palaeoceanographic proxies, see Wefer et al. (1999). Analyses of foraminifer assemblages provide a conventional and widely applied indicator for past ocean-climate conditions as well as a biostratigraphic tool for age determination. In arctic regions, or in the vicinity of formerly glaciated margins, faunal studies are commonly combined with counts of sand-size ice-rafted debris

516

Table 24.1 Most important palaeoceanographic proxies applicable to the study of contourite sediment records Principal significance

Material

Techniques

Advantages

Disadvantages

Foraminifera fauna assemblage

T, S, bioproductivity, water mass tracer

Fraction 100–500 mm

Low-cost procedure; T quantified using transfer functions

Multiple environmental controls on fauna assemblages

18

O/16O (18O)

T, S, global ice volume

Single-species foraminifer

Microscopic counting, statistical analyses Stable isotope mass spectrometry

Routine method; automated analytical procedure

13

C/12C (13C)

Ocean ventilation

Single-species foraminifer

Stable isotope mass spectrometry

Routine method; automated analytical procedure

Mg/Ca

T

Single-species foraminifer

ICP-MS, ICP-OES

Quantitative T determination

Alkenone (Uk’37) unsaturation index 231 Th/230Pa

T

Phytoplankton algae extraction

Gas chromatography

Quantitative T determination

Ocean ventilation

Amorphous component

Thermal ionization MS, multi-collector ICP-MS

Integrates isotope signature across entire water mass column

Isotope signature is ambiguous without independent determination of T Species-specific fractionation effects; longterm changes in carbon cycle Elaborate cleaning procedure; low sensitivity at low T Calibration biased by ecological and physiological factors Elaborate and costly technique; results influenced by sediment composition

Palaeoceanographic Significance of Contourite Drifts

Proxy

P.C. Knutz

143

Nd/144Nd

Water mass tracer

Fe–Mn oxides, biogenic minerals

Sortable silt (SS)

Current flow speed

Fraction 10–63 mm

Clay mineralogy

Sediment sources, ocean currents, and climate regime Ice-sheet variability, glacial sources, and climate regime

Fraction <2–4 mm

Ice-rafted debris (IRD)

Fraction >150–250 mm

Thermal ionization MS, multi-collector ICP-MS X-ray or optical particle size detectors X-ray diffraction

Microscopic counting, petrological analyses

Detects long-term changes in bottomwater sources Direct indicator of physical sorting Widely available technique Low-cost procedures; direct indicator of glacial weathering

Elaborate and costly technique; composition influenced by aeolian input Apparatus-dependent results; influence from sediment source variations Qualitative results; multiple controls on clay mineral composition Multiple glacial sources and environmental controls on IRD flux

T, temperature; S, salinity; ICP-MS, inductively coupled plasma–mass spectrometry; ICP-OES, inductively coupled plasma–optical emission spectrometry.

517

518

Palaeoceanographic Significance of Contourite Drifts

(IRD), which provides an indicator of iceberg production and melting. Foraminifer fauna analyses are usually employed in the size fraction >150 mm, although in subpolar regions with low faunal diversity screening down to 100 mm may be necessary to embrace the size spectrum of cold-water pelagic species. By application of statistical analyses using transfer functions, it is possible to determine sea temperatures in the range of 5–30C. Quantitative faunal analyses provided by the Modern Analogue Technique (MAT) (Pflaumann and Jian, 1999) are based on the assumption that spatial relationships between faunal assemblages and their marine host environments were similar in the past. The best modern analogue for a given sample is obtained by comparison with a calibration data set from modern surface sediments. Stable-isotope studies of oxygen and carbon (18O, 13C) performed on benthic and planktonic foraminifers are standard palaeoceanographic proxies attainable at a high analytical precision (e.g. Lea, 2003). In the Atlantic Ocean, 13C measured on benthic shells is commonly applied as a tracer to distinguish North Atlantic water-mass sources from bottom waters originating from the Southern Ocean around Antarctica. In present-day surface waters, 13C of dissolved CO2 increases with decreasing nutrient concentrations due to the preferential uptake of light 12C during plankton blooms (Broecker and Peng, 1982; Kroopnick, 1985). Because the 13C signature from benthic foraminifers relates to end-member variation at the surface source as well as the chemical ‘‘aging’’ of water masses, it is considered as an indicator of ocean ventilation. The two end-member water-mass components commonly identified in Late Quaternary records from the Atlantic Ocean are North Atlantic Deep Water (NADW; 13C 0.8–1.1‰), presently occupying depths between 1.5 and 3.5 km, and Antarctic Bottom Water (AABW; 13C 0.3–0.6‰), which inundates the abyssal basins below 3.5 km depth (Sarnthein et al., 1994). During the Late Quaternary, centennial-scale excursions in 13C have been observed in high-resolution records, which are indicative of major production changes in deep and intermediate water masses between the Southern Ocean and the North Atlantic (Pahnke and Zahn, 2005). Interpretation of 13C in terms of ocean ventilation rates over long time-spans, i.e. millions of years, may be complicated by changes in source area of northern hemisphere deep-water formation (Raymo et al., 2004) and the influence of global variations in the exchange between terrestrial and oceanic carbon reservoirs. To interpret, 13C as a water-mass tracer, it is important to use benthic foraminifer species living on the sea floor (as opposed to within). Scarcity of epibenthic species, such as Cibicidoides wuellerstorfi, may result in low-resolution or fragmented 13C records. The oxygen-isotope composition (18O) of marine shells has been widely applied as a marine climate indicator related to ocean temperature, salinity, and changes in global ice volume (Shackleton, 1987; Duplessy et al., 1991). In expanded sediment archives of sub-millennial scale resolution, it is useful for tracing glacial freshwater plumes that are often associated with events of iceberg discharge (Bond et al., 1992). However, in order to be used as a quantitative indicator of ice volume or salinity index, the sea-surface temperature needs to be established by other means (e.g. faunal, Mg/Ca, or alkenone saturation index: Table 24.1). For older sediments subjected to diagenesis, recrystallization of foraminifer calcite may present a problem by altering the original 18O signature (Pearson et al., 2001). In addition to temperature, the fractionation of stable isotopes in foraminiferal carbonate is

P.C. Knutz

519

influenced by species-specific metabolism (known as ‘‘vital’’ effects), thus requiring a correction factor when comparing values measured on different foraminifer species (Kohfeld et al., 1996). A new technique in palaeothermometry is based on the abundance of 13C–18O–16O measured in CO2 from organic and inorganic carbonate (‘‘clumped isotope’’) (Ghosh et al., 2006). The advantage is that temperature can be constrained independent of the isotopic composition of sea water in which the shell grew; however, the method has yet to be calibrated for use on calcareous microfossils. The Mg/Ca ratio in foraminifer shells provides a quantitative indicator of past seawater temperature, which relies on the correlation between the Mg concentration in marine carbonates and precipitation temperature (Nurnberg et al., 1996). Although this relationship has been known for some time (Chave, 1954), its exploitation in microfossil palaeothermometry requires accurate determination of small concentrations of Mg in low-temperature biogenic calcite, which has been made possible by recent developments in mass-spectrometry techniques. Studies using modern foraminifers suggest that the Mg/Ca ratio increases with temperature according to an exponential equation that varies slightly according to species. Measurements of foraminifers from sediment traps suggest that a general relationship may be given by the expression Mg/Ca = 0.380.09T (Anand et al., 2003). The Mg/Ca method requires an elaborate cleaning procedure, involving three to four wet chemistry steps, to avoid contamination by non-lattice-bound Mg derived from clay particles, Fe–Mg oxides, and organic compounds (Martin and Lea, 2002; Barker et al., 2003). Potentially, the Mg/Ca method may resolve sea surface temperature in the ‘‘warm’’ temperature range (10–30C) with errors within +1–2C. However, when applying the Mg/Ca technique over long-term time scales, additional uncertainties are introduced due to metabolic (‘‘species’’) effects and the geochemical cycling of Mg and Ca causing excursions in the Mg/Ca ratio of sea-water (Billups and Schrag, 2003). Mg/Ca in combination with 18O analyses performed on the same sample batch allows detailed reconstructions of temperature, salinity, and sea-level changes from expanded contourite sections. One of the main challenges is to obtain palaeo-temperature estimates in the cold temperature interval <10C, encompassing deep ocean and polar marine environments, where the exponential Mg/Ca–T relationship flattens (i.e. sensitivity is reduced). Additional information and references on Mg/Ca palaeothermometry can be found in Lea (2003). Recent developments in radio-isotopic techniques have produced a suite of palaeoceanographic tracers for quantifying water mass sources and deep-water exchange between the major ocean basins. 231Th/230Pa isotopes measured on bulk mud samples have been applied as a tracer for ocean ventilation changes in the North Atlantic, where it provides an alternative to benthic 13C (McManus et al., 2004; Hall et al., 2006). 231Th and 230Pa are particle-reactive radionuclides produced by alpha decay of 235U and 234U, respectively, with an initial production ratio of 0.093. Its palaeoceanographic application arises from the difference in adsorption of dissolved Th and Pa to settling particles. Th is extremely particlereactive (residence time of 20–40 years) and is rapidly removed from the water column, whereas Pa has a residence time of 100–200 years, similar to the flushing time of NADW. These properties render the distribution of excess 231Th/230Pa

520

Palaeoceanographic Significance of Contourite Drifts

(decay corrected to the time of deposition) in sediments sensitive to deep-water age. A potential pitfall is that 231Pa is preferentially scavenged by biogenic silica, which complicates the interpretation of 231Th/230Pa in regions with high opal production, like many parts of the Southern Ocean (Yu et al., 2001). Due to its strong particle reactivity, excess 230Th in marine sediments provides a proxy for accumulation rates and sediment concentration in Late Quaternary contourite sediments (Hall and McCave, 2000; Thomson et al., 2006). This information can provide a valuable alternative to radiocarbon dating in intervals that are barren of calcareous fossils. On a longer Cenozoic time scale, neodymium isotopes (143Nd/144Nd ratio, expressed as eNd) extracted from Fe–Mn oxide, and biogenic mineral components such as fish teeth have been applied to trace major water masses from their source regions (O’Nions et al., 1998; Martin and Scher, 2004). The application is based on the short oceanic residence time (500–1000 years) of Nd relative to ocean mixing time, and the distinct signature of eNd reflecting the weathering processes and geology of continental drainage regions. Its usage in contourite archives and potential for resolving ocean-climate changes at higher resolution (<105 years) remains to be investigated. More information on palaeoceanographic applications of Nd and other radiogenic isotope systems can be found in the work of Frank (2002). Sortable silt (SS), defined as the mean grain size of carbonate-free (e.g. terrigenous) silt in the 10–63 mm range, provides an indicator of current velocity (Ledbetter, 1984; McCave et al., 1995b). The SS proxy is of special interest in contourite environments where size sorting is common, due to winnowing by bottom currents of the non-cohesive silt to fine-sand fraction. SS has been applied to study the linkages between ocean-climate change and the circulation of major deep water-mass components linked to the global ocean circulation (Hall et al., 2001). Some main critical points should be considered when using the SS technique. Measured grain-size spectra differ according to what method is used, and grain-size variations in a single core do not necessarily correspond to major hydrographic changes and geostrophic flow of a single water mass, but rather to flow conditions forced by local topography. As an example, the hydrodynamic conditions across a sediment-wave field migrating like antidunes (cf. lee-wave model) will vary laterally, so that sediments sampled on the lee slope where currents accelerate will be coarser than on the stoss side of the crest (McCave and Hall, 2006). A sortable-silt record through a sediment-wave field may thus record wave migration rather than changes in geostrophic flow. These and other issues of grain-size analysis as a palaeo-current indicator are discussed in more detail by McCave (2008).

24.3.2.

Non-destructive techniques

A lithological description is a common first-order step to extract palaeoceanographic information from marine sediments. Contourite environments, related to different dynamic regimes of bottom currents, may be recognized by sedimentary facies based on visual parameters like colour, grain size, and sedimentary structures (Stow and Fauge`res, 2008). Continuous logging of geophysical properties performed on split core sections provide more detailed and semi-quantitative information on contourite stratigraphy and sediment composition. In particular magnetic susceptibility has been

521

P.C. Knutz

used to identify sedimentary cycles related to current winnowing in contourite drifts (Moros et al., 1997). Magnetic susceptibility is often complemented by measurements of g-ray density and P-wave velocity, providing information on sedimentary properties and litho-stratigraphic boundaries. Quantitative descriptive techniques include X-ray photography linked with image-analysis facilities (SCOPIX, Migeon et al., 1999) useful for revealing sedimentary structures and continuous colour scanning to monitor compositional changes, like CaCO3 content. Recent developments in continuous core logging include X-ray fluorescence (XRF) geochemistry, which provide semi-quantitative profiles of major and minor elements (e.g. Ca, Fe, Mg Ti, Sr, Ba) (Funk et al., 2004). This information is useful for correlation purposes and as an aid in sampling strategy, but can also be used to infer changes related to diagenesis and sediment source, if combined with discrete sampling analyses.

24.3.3.

Seismic imaging

Long-term palaeoceanographic information (time scales of 104–106 years) can be gained by from reflection-seismic imaging of contourite deposits. Geophysical visualization of drift geometry, internal reflector configuration (e.g. stacking patterns), and seismic facies (Nielsen et al., 2008) provide constraints on palaeo-current pathways and depositional rates as well as changes in current energy and direction during the lifetime of the contourite drift (Fauge`res et al., 1999). Information on modern or recent (e.g. Late Glacial to Holocene) oceanographic patterns may be gained from shallow seismics or sea-floor acoustic data (Kuijpers et al., 1998; Bianchi and McCave, 2000). The clarity, resolution and burial depth by which the drift depositional patterns can be determined depends on the acoustic source, spatial coverage (seismic grid), and quality of signal processing (Nielsen et al., 2008). Extracting palaeohydrological information from seismic imaging of contourite drifts requires an accurate definition of geometries and strata patterns, which is hampered by poor spatial coverage of conventional 2-D data. The advent of commercially acquired high-quality 3-D seismic data, which in recent years has been made available to research communities, has enhanced the application of acoustic imaging as a palaeoceanographic tool. 3-D seismic volumes and specialized interpretation tools with auto-tracing and horizon interpolation facilities (e.g. Schlumberger Petrel, SMT Kingdom, Landmark Seiswork) allow a detailed morphological reconstruction of contourite drifts and associated marine deposits (Figure 24.2). In addition to horizon time-structure and time-thickness (isochrone) maps, the potential in 3-D seismic data lies in the extraction of amplitude, dip-azimuth, and coherency from the wavelet trace, which can reveal detailed drift topography (such as small-scale channels and bedforms) and spatial changes in sediment properties. The direction of bottom currents that formed the sediment drift can be established from acoustic information over a range of spatial scales (Figure 24.3). Progradation of the entire drift body is the most robust evidence of a long-term current direction. This may be inferred from the internal reflector configuration in 2-D profiles or from the planform geometry of the entire drift system in relation to a point-source of sediments (Figure 24.3a). For example, an elongate contourite drift developed in the vicinity of a deep-sea fan will extend laterally away from the

(a)

NE

F lo wd ire ctio n

Faeroe–Shetland Basin

Shetland Slope

km 0

SW

5

10

NE

SW

g

b

(b)

twtt (ms)

2000

2500

Figure 24.2 Reconstruction of palaeocurrent patterns across a Plio-Pleistocene contourite drift using 3-D seismic data (see also multicolour version of this figure on the enclosed CD-ROM). Example from the West Shetland Drift (Knutz and Cartwright, 2003). (a) Isochrone (two-way travel-time difference) map between internal Pliocene reflector and basal unconformity, marked by green (g) and blue (b) arrow, respectively, in the seismic profile (b). Dark blue colours correspond to sediment thicknesses over 150 m, whereas light blue colours are indicative of non-deposition. (b) Seismic profile (thin line in A) showing alongslope progradation of internal reflectors in a southwest direction and downlap/onlap onto an erosional unconformity of Late Miocene to Early Pliocene age.The build-up of this contourite drift was promoted by southward flow of Norwegian Sea deep water in proximity of a major sediment source extending from the North Sea Basin, and the accentuated sea-floor topography generated by mass wasting. Y-axis represents two-way travel-time (twt) shown in milliseconds (ms).

523

P.C. Knutz

(a)

(b)

Slope

Slope

100 km

10 km

Shelf Shelf

Ug = 10–20 cm s–1

(c)

5 km

Ug = 10–30 cm s–1

0.2 km

Ug = 40–80 cm s–1

Figure 24.3 Contourite geometries and bedforms indicative of current direction. (a) Elongated drift developed on the lee side of a submarine fan. (b) Topographically forced bifurcation of contourite ridges and development of oblique moats between mounded drift bodies. (c) Two examples of migrating bedforms generated in response to moderate and strong bottom-current regimes. (1) Large-scale sediment waves (wavelength 1^10 km) formed in abyssal settings tend to migrate up-current at oblique angles to the flow direction, although this is not regarded as a general case for contourites in slope settings. (2) Down-currentmigrating barchan-type dunes formed under persistent high-energy current conditions.

fan in the direction of the current (e.g. the Hikurangi fan drift) (Carter and McCave, 1994). On a smaller scale, morphological patterns of drift/moat bifurcations can provide indices of long-term current directions (Figure 24.3b). The general current direction and velocity may also be estimated by resolving the spatial configuration of migrating bedforms or erosional features (e.g. scour marks) that often drape large contourite drifts (Kuijpers et al. 1998) (Figure 24.3c). The prograding direction of sediment waves (wave-lengths of 1–10 km) has been used to infer palaeocurrent directions, but this approach is complicated by the lack of a uniform relationship between along-slope currents and sediment-wave migration patterns (Kidd and Hill, 1986; Wynn and Masson, 2008). Nevertheless, many

524

Palaeoceanographic Significance of Contourite Drifts

studies of abyssal sediment waves suggest that, in the northern hemisphere, wave crests tend to migrate in an up-current direction at an oblique angle to the left of the current pathway (Figure 24.3c) (Wynn and Stow, 2002b). Large contourite drifts often express a time-transgressive pattern of erosion and deposition with drifts forming above, or in juxtaposition to, major unconformities (e.g. Tucholke and Laine, 1982; Stoker et al., 2005b). The lack of knowledge on the tectonic and oceanographic mechanisms associated with these basin-wide time gaps represents an outstanding question in the application of contourite drifts as palaeoceanographic archives.

24.4. 24.4.1.

PALAEOCEANOGRAPHIC T HEMES ADDRESSED BY C ONTOURITE R ESEARCH

Gateways, tectonics, and ocean circulation

Palaeoceanographic information from contourite drifts in the vicinity of oceanic gateways (Figure 24.4) is of major importance for resolving the role of ocean circulation in modulating Earth’s climate. The transition from the warm greenhouse climate of the Late Cretaceous and Early Cenozoic to the predominance of northern hemisphere glaciations during the last 2.5 Ma is marked by distinct phases of plate-tectonic reconfigurations. These changes led to opening and closing of oceanic gateways, thus exerting a major control on global ocean circulation and poleward temperature gradients.

Figure 24.4 World map showing important oceanic gateways that opened/closed during the Cenozoic era. 1. Fram Strait, Miocene opening. 2. Bering Strait, Late Cenozoic shallowing (?). 3. Davis Strait. 4. Greenland ^Scotland Ridge, formed in multiple phases since Early Oligocene. 5. Gibraltar Strait, narrowed during the Late Miocene (Messinian salinity crisis). 6. Panama Isthmus, final closing 3 Ma ago. 7. Indonesian Seaway, narrowed during Late Cenozoic. 8. Agulhas Passage. 9. Tasman Gap, Early Oligocene opening. 10. Southern Ocean ^ South Pacific gateway. 11. Drake Passage, Late Mid-Eocene ^ Early Oligocene opening. A multicolour version of this figure is on the enclosed CD-ROM.

P.C. Knutz

525

Although major oceanic gateways of the Cenozoic are not equally well understood in terms of their palaeoclimatic significance and timing, three tectonic ‘‘events’’ may be emphasized as having been particularly important for the development of the modern ocean circulation (Figure 24.5). First, the opening of the Drake Passage and the Tasman Gateway (Barker and Burrrell, 1977; Gersonde et al., 2002; Livermore et al., 2005) during Late Eocene to Early Oligocene led to the establishment of the circumpolar current around the Antarctic continent. The gradual thermal insulation of the South Pole generated a major expansion of the eastern Antarctic ice sheet that is commensurate with a sharp decline in deep-ocean temperatures around 34 Ma ago (Miller et al., 1991; Zachos et al., 1996). Second, the opening of the Fram Strait and the Greenland– Scotland Ridge overflow passages, effective from the Mid-Miocene (Kristoffersen, 1990; Miller and Tucholke, 1983), allowed exchange of shallow-to-intermediate water masses between the Arctic Ocean Basin and the subtropical Atlantic Ocean. Third, the closing of the Central American Seaway 5–3 Ma ago (Keigwin, 1982) resulted in increased transport of heat and moisture to the Arctic regions, a prerequisite for the development of the AMOC. Around the same time, convergence between the Austral–Asian plate and the Eurasian plate caused a narrowing of the Indonesian Seaway, thus reducing the equatorial exchange of water masses between the Indian and the Pacific Oceans (Cane and Molnar, 2001). This development promoted a cooling of the Indian Ocean between 2 and 5 Ma ago along with African aridity, and as a consequence it may have contributed to a reduction in poleward heat transport. Seismic-stratigraphic and proxy data from contourite drifts in the southern hemisphere oceans have been used to elucidate the pathway of water masses associated with the Antarctic Circumpolar Current (ACC) and the onset of Cenozoic deep-water formation. Development of major erosional unconformities succeeded by build-up of Paleogene contourite drifts in the Southern and Indian Oceans (Ramsay et al., 1994; Diekmann et al., 2004; Schut and UenzelmannNeben, 2005) and the Equatorial Atlantic (Jones and Okada, 2006) suggests that a deep-water circulation system capable of transporting fine-grained sediments became established during Late Mid-Eocene times. An Antarctic source for these water masses may be linked to the opening of a shallow gateway (<1000 m) across the Drake Passage (Livermore et al., 2005; Scher and Martin, 2006) promoting a circum-Antarctic through flow and Cenozoic cooling since about 40 Ma ago. North Atlantic contourite drifts provide an archive of the evolution of proto-NADW (also termed Northern Component Water – NCW), generated by overflow across the Greenland–Scotland Ridge. An important issue concerns the timing of initial NCW formation and upwelling of northern water masses in the Southern Ocean, along the Antarctic continental margins, as atmospheric moisture produced by upwelling may have been essential for allowing Antarctic ice sheets to expand (Schnitker, 1980). The accumulation history of contourite drifts along the southern flanks of the Greenland–Scotland ridge (Figure 24.5) suggests that an overflow of intermediate Arctic water masses from the Nordic Seas became established during the Middle Miocene (Wold, 1994). A Middle Miocene inception of the Greenland–Scotland Ridge overflow is supported by stable-isotope data, suggesting that prior to 14 Ma ago the deep-ocean circulation was generated by southern water masses and Tethys Ocean outflow (Miller and

3

2

1

0

EPU

Compression

CAS closes

FS uplift

Late Alpine Orogeny

Miocene

20

TPU Oligocene

?

Gloria Eirik Snorri Hatton

45

50

Deep-water temperature (°C) 0 4 8 12

Full-scale and permanent

Gardar

Partially ephemeral

Feni

Bjorn

Drake Passage opens

Eocene

Antarctica

Drifts

Tasman Gateway opens

FS uplift

35

40

MMU

Tibetan Plateau uplift

25

30

Unconformity

LEU

Palaeoceanographic Significance of Contourite Drifts

Age (Ma)

NE Alantic

GU

10

15

Tectonic events Global

Northern hemisphere

5

4

NCW production

Plio-Pleistoceul

5

0

Accumulation rate (g cm2 ka) 0 10 30 50

526

Glaciation index

δ18O (‰)

P.C. Knutz

527

Figure 24.5 Cenozoic tectono-climatic evolution compared with North Atlantic drift growth and development of marine unconformities. From left to right: (1) Deep-water temperature from global 18O curve based on benthic foraminifers and (2) glaciation index (modified from Zachos et al., 2001). Contourite-drift accumulation rates (modified from Wright and Miller, 1996 and Wold, 1994). (4) Relative changes in Northern Component Water (NCW) from Greenland ^Scotland Ridge overflow (Wright and Miller, 1996; Raymo et al., 2004; Poore et al., 2006). (5) Tectonic events: CAS = Central American Seaway (Haug and Tiedemann, 1998); stars denote episodes of uplift in Fennoscandia (FS) ( Japsen et al., 2007). (6) Regional Cenozoic unconformities (U) on the NE Atlantic continental margin; GU = glacial unconformity; EP = early Pliocene; MM = Middle Miocene; TP = top Paleogene; LE = Late Eocene (Andersen et al., 2000; Davies et al., 2001; Stoker et al., 2005b).

Fairbanks, 1985; Woodruff and Savin, 1989). Identification of a southward prograding contourite drift in the Faeroe–Shetland Channel (Davies et al., 2001), tentatively correlated with onset of sediment accumulation on the Feni Ridge ( Jones et al., 1970), suggests a much earlier onset of Nordic Sea overflow during the Early Oligocene. A possible explanation for this ambiguity is that an early phase of North Atlantic overturn circulation was in place, but that the overflowing water masses were of insufficient density to influence the abyssal circulation. A key question of the Late Cenozoic evolution concerns the relative influence of climatic and tectonic forcing mechanisms for the build-up of northern hemisphere ice sheets. The change from the warmth of the Early Pliocene to the Quaternary climate marked by profound cooling coincides with the final closing of the Central American Seaway (Figure 24.5). This transition is captured by stacked benthic 18O records, indicating a gradual long-term cooling superimposed by glacial/interglacial cycles. A transition to lower 18O values 2.75 Ma ago, corresponding to increased iceberg calving in the North Atlantic (Shackleton et al., 1984), is recognized as a marker for northern hemisphere ice sheets expanding from inland Arctic regions to the shelf margins. It has been suggested that closing of the Panamanian Seaway, and the consequent increase in meridional transport of atmospheric moisture, was the main cause for glacial expansion in the northern hemisphere (Haug and Tiedemann, 1998). Other explanations have in particular focused on atmospheric cooling driven by a decrease in greenhouse gas concentrations during the late Pliocene–Pleistocene transition. However, recent studies indicating that pCO2 had already reached low levels (280–300 ppm) during the Early Neogene (Pagani et al., 2005) lends support to moisture as the limiting factor for glacial build-up in the circum-North Atlantic regions. Basin-wide unconformities at the base of Plio-Pleistocene contourite units (Figure 24.2b and 24.5) suggest that strong currents traversed the eastern passage of the Greenland–Scotland Ridge during the latest Miocene to Early Pliocene (6–4 Ma). This erosional/non-depositional phase was followed by enhanced accumulation of contourites in the Faeroe–Shetland Channel (Knutz and Cartwright, 2003) and south of the Greenland-Scotland Ridge (Eirik and Gardar drifts) (Wold, 1994), indicating that a more moderate bottom current regime became established prior to the onset of the Pleistocene glaciations (2.8 Ma). With age uncertainties taken into consideration, the seismic-stratigraphic evidence is in broad agreement with stable isotope (18O, 13C) data, indicating enhanced

528

Palaeoceanographic Significance of Contourite Drifts

ventilation of the deep Atlantic during Early Pliocene (4.4–3.1 Ma) (Ravelo and Andreasen, 2000), followed by a general decrease in the relative flux of NCW between 3.0 and 2.7 Ma (Raymo et al., 1996). A possible explanation for a Late Pliocene slow-down in NCW formed by Nordic Sea overflow is that increased precipitation over northern Eurasia and subsequent freshwater run-off to the Arctic Ocean led to reduced surface-water density in regions of deep-water formation (Driscoll and Haug, 1998). Whether freshwater forcing was responsible for reducing convective circulation in the Nordic Seas during the Plio-Pleistocene transition has yet to be confirmed by high-resolution palaeoceanographic studies. Superimposed on the longer-term Cenozoic trends, combined records of benthic 13C indicate pronounced fluctuations in the production of NCW, since the Middle Miocene (last 12 Ma) (Raymo et al., 1996; Wright and Miller, 1996; Poore et al., 2006). Phases of reduced NCW fluxes observed at 9–10, 6–7, and more generally during the Early Pleistocene (from 2.7 Ma) have been associated with dynamic motions along the Greenland–Scotland Ridge, pointing to an Icelandic mantle-plume control on Nordic Sea overflow. Other scenarios for tectonic forcing of the Cenozoic ocean-climate development build on evidence for intracontinental tectonic movements in NW Europe and East Greenland ( Japsen and Chalmers, 2000). Notably, phases of uplift-exhumation in Fennoscandia and compressional tectonic regimes along the NE Atlantic margin seem to correlate with the development of major deep-sea unconformities (Andersen et al., 2000; Stoker et al., 2005b; Japsen et al., 2007). These different lines of evidence suggest that time-stratigraphic relationships between palaeoceanographic changes, terrestrial climate development (e.g. glaciations), and tectonic controls are insufficiently understood, thus providing a scope for more integrated land/ocean research.

24.4.2.

Ocean circulation during warm climate extremes

Climate forcing inflicted by anthropogenic emission of greenhouse gases is thought by IPCC (2007) to have raised global average temperatures by about 0.7C since 1900 A.D. Concerns of accelerated warming in the near future have stimulated research on ‘‘greenhouse-type’’ climate systems of the geological past. Because the oceans constitute the largest CO2 reservoir (53 times as much as the atmospheric capacity), even minor changes in dissolved CO2 in the ocean’s interior will have an impact on atmospheric greenhouse forcing (Broecker and Peng, 1982). Contourite successions deposited during warm climates can provide insight into the function of ocean circulation and chemistry in regulating planetary temperatures during intervals of higher-than-present pCO2 levels. The most relevant stratigraphic intervals for this theme are the Middle Cretaceous (100–70 Ma), Early Eocene (55–50 Ma), Late Oligocene to Mid Miocene (26–14 Ma), and Early Pliocene (5–3 Ma). Of these time-spans, the Mid-to-Late Cretaceous climate represents the most extreme past scenario of greenhouse forcing, with atmospheric CO2 estimated to two to nine times the modern levels and global temperatures 7–14C higher than today (Berner, 1990). The meridional and vertical ocean temperature gradients were probably small compared with the Late Neogene, which begs the question what ocean-circulation mechanisms could have maintained the apparent

529

P.C. Knutz

warm polar climates of the Cretaceous (Bice et al., 2003). Large-scale circulation in the Cretaceous oceans has conventionally been viewed as ‘‘halothermal’’, i.e. driven by low-latitude evaporation rather than high-latitude cooling. However, recent studies based on climate modelling (Brady et al., 1998; Poulsen et al., 1999) and nannofossil evidence (Watkins and Self-Trail, 2005) point to an oceanic circulation system driven by high-latitude cooling, more similar to the modern AMOC. The uncertainties regarding the ocean-circulation mechanism and boundary controls (such as sea level, hydrological cycle and planetary albedo) of the Late Mesozoic climate highlight the need for regional palaeoceanographic time-slice reconstructions. Studies of Cretaceous and Early Cenozoic ocean current deposits are sparse and hampered by interpretation problems that are general for ancient contourites (Hu¨neke and Stow, 2008). Recent work in the North Sea region suggests that deposition during the Late Cretaceous (Santonian to Maastrichtian 86–66 Ma) was influenced by deep geostrophic currents in an epeiric sea with water depths of 500–800 m (Surlyk and Lykke-Andersen, 2007; Esmerode et al., 2008). The margins of the basin feature broad moat-like channels that are juxtaposed to accretionary forms of nannofossil carbonate ooze and marls. Due to the limitations of quantitative geochemical-based reconstructions, determined by the diagenetic modification of fossil material, geophysical and sedimentological methods provide the principle means to unravel the palaeoceanographic history in Mesozoic chalk deposits. Accordingly, identification and regional seismic mapping of bottom-current-related features provide information on poleward water-mass exchange ocean circulation and sea-level changes of the Mid-to-Late Cretaceous (Esmerode et al., 2008). The Early-to-Mid Pliocene represents the last geological time-span during which Earth’s climate was significantly warmer than today, with high-latitude temperatures 4–8C above present levels (Dowsett et al., 1996). A distinct advantage from a modelling perspective is that the boundary conditions of the Pliocene resemble the present, with respect to the plate-tectonic configuration and presence of ice sheets in Antarctica and in parts of Greenland. The two main factors that have been considered for the Pliocene warming are high CO2 levels (Crowley, 1996; Haywood and Valdes, 2004) and/or increased levels of oceanic heat transport (Raymo et al., 1996; Ravelo and Andreasen, 2000), possibly associated with semipermanent El Nin˜o conditions (Ravelo, 2006). Peak Pliocene CO2 concentrations have been estimated at 380 ppm (Van der Burgh et al., 1993), which is only some 100 ppm higher than pre-industrial concentrations. This would suggest that either the greenhouse effect was insignificant for the Pliocene warming or, alternatively, the Pliocene climate system responded sensitively to even small changes in pCO2. A better knowledge of these two different scenarios is highly relevant for evaluating the future course of present-day global warming. Palaeoceanographic data from high-accumulation-rate contourites provide an important key for resolving the enigma of the warm Pliocene climate.

24.4.3.

Rapid ocean-climate variability in the North Atlantic

Numerous high-resolution records of Late Quaternary ocean-climate change have been obtained from contourite deposits in the North Atlantic and Nordic Seas.

530

Palaeoceanographic Significance of Contourite Drifts

Depending on core location and water depth, proxy data from North Atlantic contourite drifts have provided detailed information on ocean-climate dynamics and relative contribution of intermediate and deep water masses during the last glacial–interglacial cycle (e.g. Sarnthein et al., 1994; McCave et al., 1995b). Palaeoceanographic information from contourites deposited during the last glacial period, 10–110 ka BP, reveals detailed information on rapid climate changes involving couplings between ice sheets, oceans, and atmosphere. A typical contourite environment on a glaciated continental margin, and the controlling factors that determine climate/ocean/sediment interactions in this setting, is illustrated in Figure 24.6. An example of the close coupling between millennial-scale (i.e. stadial/interstadial) climate variability and ocean current dynamics is demonstrated by the recognition of sedimentary cycles in glacial contourites from the northern North Atlantic (Figure 24.7). Increases in geostrophic current intensity, corresponding to the warm interstadials in the Greenland ice core records, are reflected by peaks in magnetic susceptibility measured in cores collected along the Greenland–Scotland and Reykjanes ridges at 1–2 km water depth (Rasmussen et al., 1996; Moros et al.,

Precipation

Reflected insolation

Ablation Evaporation

Meltwater wn

Bio

slo

pe

ivity

uct

d -pro

Do

tra

ns

Vertical fluxes

po

rt

ry da un s Bo rrent cu

Figure 24.6 Conceptual model illustrating the principal (palaeo)environmental factors that contribute to accumulation of high-resolution contourite archives on a glaciated continental margin. The notable differences from contourites developed in a non-glacial setting are the supplies of glacimarine sediments from down-slope resuspension, meltwater plumes and in the form of ice-rafted debris supplied from tidewater glaciers. A multicolour version of this figure is on the enclosed CD-ROM.

Depth (cm)

Marine isotope stages

ENAM93-21/MD95-2009

45

δ18O (‰)

Magnetic susceptibility

(N. pachyderma sin.)

(10–6 Sl units)

40

35

30

25

20 0

200

400

600

800

Holocene 1

1 H1

2

H2 4

H3

2 3 5

500 H4

6 7 8

9 10 11

3

12 13

H5 1000

14 15 16–17 H6

4 19–20 5a 5b

21 H8

1500

22

H9

5c

H10 5d

23 24

5e

Interstadial numbers

2000

H11

6 Heinrich events

Figure 24.7 Millennial-scale variability recorded by 18O and magnetic susceptibility measurements in cores ENAM93-21 and MD95-2009, respectively, from the northern Faeroese Slope (modified from Nielsen et al., 2007). Position of Heinrich events, marked by 18O depletions, and warm interstadials (Dansgaard ^ Oeschger events) recognized here by magnetic-susceptibility peaks, are shown (see main text).

532

Palaeoceanographic Significance of Contourite Drifts

1997). While strong along-slope currents favour deposition of magnetic Fe minerals along with other basaltic weathering products, periods of low current intensities and elevated iceberg productivity, corresponding to cold stadials and Heinrich events, produce terrigenous clay-rich intervals with coarse ice-rafted sediments. The correlation between marine interstadial intervals and the Dansgaard–Oeschger events during MIS-3 suggests that abrupt increases in atmospheric temperature were associated with intense production of overflowing water masses from the Nordic Seas (Figure 24.7). Benthic oxygen-isotope records from the eastern margin of the Norwegian Sea reveal a more complex picture of deep-water circulation, as it shows that 18O depletions during cold periods were not limited to the sea-surface but also influenced the sea floor at intermediate water depths (Dokken and Jansen, 1999). This pattern has been related to a rapid transfer by downwelling of 16O-enriched brines associated with freezing of sea-ice, thus representing a stadial mode of ocean ventilation that is different from open-ocean convection formed by cooling and subduction of water-masses (cf. Carmack and Aagaard, 1973). Palaeoceanographic studies performed on fine-grained contourite sediments on the NW European continental margin show that intervals of strong NADW production have alternated with intervals of meltwater/iceberg discharge and down-slope gravity flows (Knutz et al., 2001; Øvrebø et al., 2006; Eynaud et al., 2007). The multi-decadal-scale resolution produced by muddy contourite records, enables the recognition of brief, high-amplitude pulses of meltwater and icebergs across the late-glacial interval (Figure 24.8) (Knutz et al., 2007). The results indicate that NW European ice sheets began to collapse several thousands of years prior to Heinrich events 1 and 2, apparently triggered by abrupt sea-surface warmings. The relationship between deglacial meltwater pulses and deep-water formation has been examined in detail using 231Th/230Pa (Hall et al., 2006). The study suggests that even minor discharges of meltwater into the NE Atlantic can generate a rapid negative response in AMOC. Expanded Late Quaternary climate records have also been extracted from contourite deposits on the Portuguese Margin (Zahn et al., 1997; Llave et al., 2006; Toucanne et al., 2007). Cores from this region provide insights into variability of NADW and Mediterranean Overflow Water and the response of these water masses to climate change in the wider North Atlantic region. In general, the palaeoclimate results derived from Late Quaternary contourites support ocean-climate models suggesting that the Dansgaard–Oeschger oscillations were intricately linked with freshwater pertubations of the North Atlantic overturn circulation (e.g. Ganopolski and Rahmstorf, 2001). However, several outstanding questions pertain to the forcing mechanisms of rapid climate change, in particular the role of solar variability (Stuiver et al., 1995), interhemispheric teleconnections involving ocean/atmosphere feedbacks (Cane and Clement, 1999), triggers of icesheet instability (Hulbe et al., 2004), and the reorganization of ocean circulation that leads to rapid surface warming during interstadials (Lohmann and Schulz, 2000). These unresolved issues emphasize the need for more detailed palaeoceanographic reconstructions on glaciated continental margins and in the pathway of major ocean currents (Figure 24.6). Over the recent years, an increasing amount of palaeoceanographic studies have focused on expanded Holocene sediment archives. The main objective is to

533

P.C. Knutz

DAPC2

GISP2

δ18O (VPDB ‰) 4.5

4.0

3.5

3.0

δ18O (SMOW ‰) –40

2.5

–38

–36

–34

10

a

c

b

d

Younger Dryas

15

Age (ka BP)

H1

20 Last Glacial Maximum

H2 25

0

6 8 20 2 4 IRD flux (No.cm–2 yr–1)

100

80

60

40

20

0

N. pachyderma sin.(%)

Figure 24.8 Ultrahigh-resolution marine record from the Rockall Trough, northeast Atlantic, correlated with the oxygen-isotope record from the GISP2 ice core (Knutz et al., 2007). Multi-decadal-scale resolution of this glacimarine contourite unit enables recognition of a detailed paleoclimatic sequence through the late glacial and early deglacial interval (15^25 ka). Heinrich events 1 and 2 are defined by high concentrations of detrital carbonate delivered by icebergs calved from the Laurentide Ice Sheet. Sharp peaks in non-carbonate ice-rafted debris and planktonic 18O depletions, each lasting some 200 years, are observed prior to the Laurentide Ice Sheet collapses (H1 and H2). The sharp decrease in 18 O around 16 ka BP suggests a major early deglacial meltwater pulse from NW Europe and Scandinavia. The abundance of the polar foraminifer Neogloboquadrina pachyderma (sinistral), in combination with the detailed signature of planktonic 18O, indicates a series of centennial-scale fluctuations in ocean-surface temperatures during the late glacial interval. The palaeoclimatic sequence suggests that rapid sea-surface warmings around 17 and 18 ka BP triggered an early deglacial response in the form of massive glacimarine fluxes from the NW European Margin.

534

Palaeoceanographic Significance of Contourite Drifts

establish linkages between natural climate variability over the last 11,000 years, historical records (e.g. the last 2000 years), and climate measurements gathered over the last century. This topic has been stimulated by marine core studies, indicating that millennial-scale climate variability persisted in the North Atlantic across the last deglaciation and throughout the Holocene (Bond et al., 1997; Andresen et al., 2005). A key result based on SS data from a contourite core (Gardar drift, South Iceland Basin) suggests that increased AMOC corresponds to mild periods of the Holocene, such as the Medieval Climatic Optimum (Bianchi and McCave, 1999). The mechanism for these apparently synchronous changes in deep-water flow and ocean-surface climate, and its implication for future climate trends, are not fully understood, thus calling for further palaeoceanographic studies of interglacial contourites.

24.5.

S UMMARY

Contourites provide essential information on ocean circulation and past climate changes, which can be extracted using a wide range of geochemical, sedimentological, and geophysical research methods. This chapter provides a synthesis of the palaeoceanographic approaches and techniques that apply to the study of contourite archives. Three themes are introduced in which contourite research has brought, or has the potential to deliver, significant new insights into the ocean’s role in Earth’s climate system: (1) gateways, tectonics, and ocean circulation, (2) ocean circulation during warm climate extremes, and (3) rapid ocean-climate variability in the North Atlantic. These themes are not intended to represent all aspects of palaeoceanography in the realm of contourites but rather to demonstrate the range of time scales, from tectonic (millions of years) to human (tens of years), that may be extracted from sediments formed by persistent oceanic currents. The majority of palaeoceanographic studies on contourites have been performed in the Atlantic and Southern Ocean, while geophysical mapping and coring of contourite drifts in the Indian, Pacific, and Arctic Oceans are sparse. Thus, the potential for using bottom-current generated deposits as high-resolution archives on a global scale is far from being fully exploited. The role of ocean circulation, climate, and tectonic evolution during the Cenozoic is a prominent theme that is often addressed by deep-ocean drilling campaigns. The problems of extracting the control of ocean current dynamics from other factors like sea-level, climate regime and sediment properties (e.g. Fauge`res and Stow, 2008) demonstrate the need for multi-coring transects and more integrated seismic-sediment-proxy studies of contourite depositional systems and their bounding unconformities. High-resolution palaeoceanographic information provides ‘‘ground truthing’’ of the multitude of output scenarios generated by numerical modelling. Proxy records from rapidly accumulating muddy contourite deposits allow reconstruction of leads and lags between different parameters of ocean-climate change at multi-decadal time scales, thus approaching the resolution gained from ice-core

P.C. Knutz

535

archives. This information is crucial for a better understanding of global teleconnections, feedback thresholds and forcing mechanisms that determine the past and present climate system.

ACKNOWLEDGEMENTS The author thanks John W. Jones and Christian J. Bjerrum for constructive reviews that improved the clarity of the manuscript. This work was supported by the National Geological Survey of Denmark and Greenland.

C H A P T E R

2 5

T HE S IGNIFICANCE OF C ONTOURITES FOR S UBMARINE S LOPE S TABILITY J.S. Laberg1 and A. Camerlenghi2 1

Department of Geology, University of Tromsø, Tromsø, Norway ICREA, c/o Universitat de Barcelona, Departament d’Estratigrafia, Paleontologia i Geocie`ncies Marines, GRC Geocie`ncies Marines, C/ Martı´ i Franque`s, Barcelona, Spain

2

Contents 25.1. Introduction and Objective 25.2. Background 25.3. Case Studies of Contourites and Submarine Slope Stability 25.3.1. Northern mid- and high-latitude contourites and submarine landslides 25.3.2. Southern mid- and high-latitude contourites and submarine slope instability 25.3.3. Ocean gateway – unstable contouritic levees 25.3.4. Recapitulation of evidence 25.4. Discussion 25.4.1. Gravitational instability of submarine slopes 25.4.2. Characteristics of contourites that favour slope instability 25.5. Summary Acknowledgements

25.1.

537 538 539 539 545 550 550 552 552 554 555 556

I NTRODUCTION AND OBJECTIVE

Following the introduction of new and detailed sea-floor mapping techniques (e.g. Mienert and Weaver, 2003), sophisticated equipment for in situ measurements (Strout and Tjelta, 2005) and laboratory testing (Sultan et al., 2004; Kvalstad et al., 2005) of the physical properties of sediments, our understanding of submarine slope instabilities has increased substantially over the last decade (e.g. Locat and Lee, 2002; Locat and Mienert, 2003; Canals et al., 2004; Bryn et al., 2005a). One of the important new findings was that the distribution, composition and properties of contouritic sediments are vital for some of these events to occur. This relationship has been particularly well documented on the NW European Atlantic Margin, where some of the world’s largest submarine landslides have occurred (Bryn et al., 2005b); however, a similar relationship has been found also on other continental Developments in Sedimentology, Volume 60 ISSN 0070-4571, DOI: 10.1016/S0070-4571(08)00225-2

Ó 2008 Elsevier B.V. All rights reserved.

537

538

The Significance of Contourites for Submarine Slope Stability

margins, e.g. Antarctica and Canada (Volpi et al., 2003; Piper and Campbell, 2005). The objective of this chapter is two-fold: 1. to provide an overview of environments where the stability of the slope has been affected by the distribution and properties of contouritic sediments, and 2. to discuss how and why contouritic sediments have affected the stability of submarine slopes. Here we address submarine slope instability in muddy contourite drifts on northern, high-latitude slopes (Norwegian, Canadian, UK and Faroese Margins), of elongated drift mounds on southern high-latitude continental slopes and rises (Antarctic Margin) and in sediment drifts in areas of ocean gateways (Gibraltar).

25.2.

BACKGROUND

The effect of both thermohaline and geostrophic contour currents on erosion, transport and deposition of sediments was originally identified in the deep sea (abyssal plains and continental rise), and much work has been focused on the establishment of facies models, on the link between grain size and current strength, and on deciphering their palaeoceanographic record (see Hsu¨, 2008). Thermohaline and geostrophic circulation also affect surface-water masses, so that the continental shelf and the upper continental slope may also be influenced by the action of bottom currents, such as offshore Norway (Holtedahl and Bjerkli, 1975; Michels, 2000; Hansen and Østerhus, 2000) and eastern Canada (Piper, 2005). Changes in geostrophic and thermohaline currents may also induce the necessary conditions for gas-hydrate dissociation and slope instability (Mienert et al., 2005). The higher degree of spatial and temporal variations in sediment erosion, transport and deposition in these shallower-marine environments generates sedimentary successions that are prone to become gravitationally unstable (Laberg et al., 2005). The Antarctic continent is bound by one of the most vigorous current systems of the world’s oceans, the eastward flowing Antarctic Circumpolar Current (ACC) (see also Van Weering et al., 2008). The axis of the ACC generally lies far from the margin, except at Drake Passage, where the current axis is forced close to the continental slope by a complex sea-floor morphology and the occurrence of gateways. More proximal westward flowing polar currents affect sediment transport and deposition along the Antarctic continental margin (Camerlenghi et al., 1997; Giorgetti et al., 2003), determining the deposition of large fine-grained (low-energy) mixed turbidite/coutourite drifts on the upper and lower continental rise (e.g. Pudsey and Camerlenghi, 1998; Pudsey, 2000; Rebesco et al., 2002; Herna´ndez-Molina et al., 2004). Alternative sedimentation models have been proposed by McGinnis and Hayes (1995) and McGinnis et al. (1997). Complex patterns of thermohaline circulation are found in association with ocean gateways, such as the Strait of Gibraltar. When flowing out of the Strait of Gibraltar, the Mediterranean Outflow Water (MOW) follows major and secondary

J.S. Laberg and A. Camerlenghi

539

channels or spills over a sedimentary levee. The MOW determines the growth of a high-sedimentation-rate contourite system where the continuous east–west shearing of the flow has induced intense sediment deformations, failures and mass flows (Mulder et al., 2003b).

25.3. 25.3.1.

C ASE STUDIES OF C ONTOURITES AND SUBMARINE SLOPE STABILITY

Northern mid- and high-latitude contourites and submarine landslides

The influence of contourites on continental slope stability has been documented in a number of cases from the Norwegian Margin. Key examples are the Storegga Slide as well as older events in the same area (Evans et al., 1996, 2005; King et al., 1996; Bryn et al., 2003, 2005a; Solheim et al., 2005), the Trænadjupet Slide (Laberg and Vorren, 2000; Laberg et al., 2003), the Nyk Slide (Lindberg et al., 2004) and the Sklinnadjupet Slide (Laberg et al., 2001; Dahlgren et al., 2002). Contouritic sediments were also inferred from the continental slope immediately outside the Hinlopen Slide, a giant submarine landslide in the Arctic Ocean (Vanneste et al., 2006). In addition, cases have been presented from offshore UK and the Faeroese islands – the Afen Slide (Long et al., 2003; Wilson et al., 2004) and the NE (Van Weering et al., 1998) and SE Faeroese slides and slumps (Laberg et al., 2005) – and from the Canadian Atlantic Margin (Piper 2005; Piper and Campbell, 2005). On the Norwegian continental slope, packages of glacigenic debris flows interbedded with hemipelagic and contouritic sediments compose a thick PlioPleistocene prograding wedge (Dahlgren et al., 2002, 2005; Hjelstuen et al. 2004). The depocentres lie marginal to the former ice sheets and were affected and partly modified by the action of along-slope flowing currents that caused sediment erosion and/or deposition (Dahlgren et al., 2005; Laberg et al., 2005). The sedimentology and physical properties of these sediments are summarized in Table 25.1. Contourites have been deposited on the continental slope during both times of low-stand (glacials) and of high-stand (interglacials) of the sea level. Even though the currents were weaker during glacials, the sedimentation rates on drifts were one order of magnitude higher, due to the high glacimarine sediment supply from the ice sheets grounded on the continental shelf (Laberg and Vorren, 2004). Instead, the deposition of glacigenic debris flows took place only during glacials, as a product of subglacial sediment discharge at the continental-shelf break. The Nyk Slide, which probably happened during the maximum ice extent of the Weichselian glaciation, offers an image of how contourites and glacigenic debris-flow deposits have all been affected by mass wasting. One or more glide planes of the slide are located within the contourite drift and are parallel to the

Southern high-latitude slope and rise (Antarctic Peninsula)

Ocean gateways (Strait of Gibraltar)

Grain size

Clay (’ > 8): 30–70% Sand/Gravel (’ = –4 to 4): 3–20%

Composition

Siliceous and calcareous oozes (Early Cenozoic) – repeated sedimentation of stacked glacial and contouritic/hemipelagic units (Late Cenozoic) Kaolinite-rich ooze (Eocene to earliest Miocene), smectite-rich fossilferous clays/clayey oozes (earliest Miocene to Late Pliocene), illite-dominated sediments (Late Pliocene to present) In total up to 1.5 s (twt) of sediments deposited during the Late Pliocene to Pleistocene (including glacigenic and contouritic sediments) Water content: 20–31%, unit weight: 19–21.5 g cm 3, the sediments behave contractant during triaxial testing. In situ pore pressure of contouritic sediments: excess porewater pressure (Storegga Slide headwall) or hydrostatic (within the slide scar and at the northern flank)

Bimodal (’ = 9.65 and 10.3) or unimodal (’ about 8). Sorting is very poor (’ = 2–3) Laminated and massive bioturbated, diatom-bearing silty clays with a well-defined alternation of biogenicrich and biogenic-poor horizons

Top few metres of the upper continental slope ’ varies between 6 and 8.5 (upper 50 cm coarser) Terrigenous and biogenic clasts, mainly calcareous. About 40% carbonates, some Ice-rafted debris (IRD)

Clay mineralogy

Sedimentation rates Physical properties

Diagenesis

Two end-member assemblages: <20% smectite and >40% chlorite or >20% smectite and <40% chlorite. Illite between 30 and 50% Middle Miocene to present: from 12 to less than 5 cm ka 1 (distal); Pliocene: from 18 cm ka 1 to about 8 cm ka 1 (proximal) Porosity in ODP Site 1096: 70–75% (surface) to 50–60% (100 mbsf ), below: constant or rises slightly (to 500–600 mbsf ). Porosity in ODP Site 1095: 60 to 40–45% at 480 mbsf. Surface sediment bulk density is 1.6 g cm 3, reaching 2.0 g cm 3 below the sharp decrease in porosity An Opal A-CT diagenetic front occurs at about 500 mbsf throughout the continental rise

Highly variable (between 15 and 40 cm ka 1 in the Late Neogene) Data available from cores collected in the uppermost 2.5 m of sediments indicate scattered water content (dry weight), ranging from 35 to 80%

The Significance of Contourites for Submarine Slope Stability

Northern high-latitude slope (Storegga Slide complex)

540

Table 25.1 Summary of the composition, sedimentation rates and physical properties of contourites and associated sediments on northern highlatitude continental slopes (Berg et al., 2005; Bryn et al., 2005a; Forsberg and Locat, 2005; Kvalstad et al., 2005; Solheim et al., 2005; Strout and Tjelta, 2005), southern high-latitude slope and rises (Rebesco et al., 1996, 1997; Shipboard Scientific Party, 1999b; Lodolo and Camerlenghi, 2000; ¨tterer, 2001; Pudsey, 2001; Iwai et al., 2002; Lucchi et al., 2002; Volpi et al., 2003) and ocean Hillenbrand and Ehrmann, 2001; Hillenbrand and Fu gateways (Lee and Baraza, 1999; Nelson et al., 1999; Llave et al., 2001, 2006; Herna´ndez-Molina et al., 2003) – see text for discussion

541

J.S. Laberg and A. Camerlenghi

1.6

1.7

NNW

SSE

1200 Water depth (m)

Late Weichselian debris-flow deposits

1300

1.8

Nyk Slide

The Nyk Drift

twt (s)

Headwall

1.9 Escarpment

Slide deposits

2.0

Glide plane

0

2 km

2.1

2.2

Figure 25.1 A deep-towed Boomer profile across the Nyk Slide scar offshore Norway. The Nyk Slide affected late Weichselian glacigenic debris-flow deposits and part of the underlying Nyk contourite drift (Laberg et al., 2001). The profile has been modified from Evans et al. (2005), with permission from Elsevier.

original acoustic lamination (Figure 25.1). Therefore, the weak layers that induced the slope failure were depositional surfaces, or thin layers, of contouritic sediment (Lindberg et al., 2004). Subsequent to failure, contouritic sediments have filled the slide scars, thus re-creating the prerequisite conditions for repeated sliding in the same area, eventually resulting in events such as the Holocene Trænadjupet Slide (Laberg and Vorren, 2000). The Storegga Slide complex offshore Norway is by far the best studied case of cause/effect relationships between contouritic deposition and sediment mass wasting (Bryn et al., 2005b). Despite the huge dimensions of the slide complex, geotechnical drilling and coring for physical properties have shown that in a continental slope succession composed of glacigenic debris-flow packages interbedded with hemipelagic and contouritic sediments, the contouritic sediments have a higher clay content, water content, plasticity index and liquidity index, resulting in a lower strength and higher sensitivity than the glacial sediments at comparable

542

The Significance of Contourites for Submarine Slope Stability

Figure 25.2 Part of a seismic profile from the Storegga Slide area offshore Norway showing the palaeo-slide S, an infilling contourite drift and palaeo-slide R. The glide plane of palaeoslide S is indicated by the green horizon. Note that the glide plane of palaeo-slide R (pink horizon labelled P Slide 2 Base) follows the top of the drift deposits and is parallel with the internal reflections of the drift. CD marks contouritic drift deposits and SD denotes slide debris, TNU = top Naust unit U reflection, INS2 = intra Naust unit S reflection 2 and TNR AS = top Naust unit R reflection. See Solheim et al. (2005) for further discussion. A multicolour version of this figure is on the enclosed CD-ROM.

overburden stress (Kvalstad et al., 2005). Similarly to the Nyk Slide, the persistent contouritic deposition throughout glacial/interglacial intervals of the Late Neogene and the filling of slide scars with contourite drifts (Figure 25.2), documented from the Storegga Slide area (e.g. Evans et al., 1996, 2005), are key factors for the recurrence of slope failure in the same area over one glacial/interglacial cycle (Solheim et al., 2005). Due to the widespread distribution of contourites on the Norwegian Margin, large sectors of the continental slope have repeatedly become unstable, and large submarine slides such as the Storegga Slide may not be unique events in the geological record (Bryn et al., 2005a). The trigger mechanism for the last Storegga Slide, approximately 8100 cal. years BP (Haflidason et al., 2005), was probably an extremely strong earthquake with a very low probability of occurrence (Kvalstad et al., 2005), which altered a weak equilibrium between acting and resisting forces within fine-grained contouritic deposits. Compared with the giant submarine landslides offshore Norway, the Holocene Afen Slide was a minor event affecting the continental slope in the Faroe–Shetland channel, NE Shetland continental slope. In this case, well-sorted contouritic sands

543

J.S. Laberg and A. Camerlenghi

0

5 km

800 mbsl 900 mbsl

50 m

1000 mbsl

1100 mbsl

0m

10 m

0 km NW

1 km SE

Figure 25.3 High-resolution seismic-reflection profile across the Afen Slide, offshore UK, showing acoustically transparent lenses possibly comprising contouritic sand near the crown of a slide scar. FromWilson et al. (2004), with permission from Elsevier.

within muddy contourites (Figure 25.3) are inferred to have acted as detachment surfaces for the slides. The equilibrium was altered likely by liquefaction of the contouritic sands following seismic shaking (Wilson et al., 2004). The mud/sand transition was climatically controlled, as the sandy layers were the product of interglacial deposition by relatively stronger bottom currents. According to Wilson et al. (2004), an older, buried slide is present in the same area; the same trigger mechanism can be inferred for it. On the SE Faeroese Islands slope, evidence for mass wasting comes from the identification of late Neogene slumping (Figure 25.4) (Laberg et al., 2005; Nielsen et al., 2007). Because of the lack of lithological information, the relationship between contouritc sediments and submarine sliding on the NE Faeroese Slope is not well established yet. It is known that contouritic sediments have been deposited within a slide scar, the base of which is located within sediments deposited by dominantly down-slope density currents (Nielsen and van Weering, 1998; van Weering et al., 1998). Whether contouritic sediments in any way influenced sliding in this area is presently not known. In a very different setting, infilling of a slide scar and channel by sand of possible contouritic origin has also been inferred from the Pliocene rock record of Indonesia (Roep and Fortuin, 1996).

544

The Significance of Contourites for Submarine Slope Stability

NW

1.5

SE

Contourite

twt (s)

Slide scar

Slide deposits

2.0

2.5

10 km

Figure 25.4 OGS seismic line ISTRA-280 crossing the lower slope of the northeastern Faroe Margin and showing a Late Pliocene to recent plastered contourite affected by mass wasting. Modified from Laberg et al. (2005), with permission from Elsevier.

Offshore eastern Canada, Piper and Campbell (2005) found buried up to 100-m high escarpments and associated debris-flow deposits on the slope of the Sackville Spur (Figure 25.5). Sediment sampling of the spur revealed an upper section consisting of 8 m of sandy gravelly mud, overlying 4.5 m of grey mud and then a further 12 m of gravelly sandy mud. This is a relatively shallow-water contourite drift constructed by the Labrador Current on the upper slope in response to the NW 1000

twt (ms)

1200

SE

Sackville Spur Buried 40 m high escarpment

1400

Buried 100 m high escarpment 1600

Debris-flow deposit

1800 2000

Figure 25.5 Seismic-reflection profile through Sackville Spur, a shallow-water sediment drift located on the eastern Canadian continental margin (Flemish Pass). The profile shows buried failure scarps and debris-flow deposits. Modified from Piper and Campbell (2005).

545

J.S. Laberg and A. Camerlenghi

supply of large amounts of suspended sediments by the Laurentide ice sheet reaching out to the shelf break (Piper, 2005). These sediments are thought to have failed to produce the observed slide headwalls (escarpments) and debris flows.

25.3.2.

Southern mid- and high-latitude contourites and submarine slope instability

On the continental rise of the Pacific Margin of the Antarctic Peninsula, large elongated sedimentary mounds (100–300 km long, 50–100 km wide, elevated several hundreds of metres above the adjacent sea floor) are separated by trunk-type deep-sea channels (Figure 25.6) (Rebesco et al., 1998b, 2002; Dowdeswell et al., 2004; Amblas et al., 2006; see also van Weering et al., 2008). The mounds can be regarded as 62°S

Drift 1 Drift 2

Antarctica Drift 3

ris

e

64°S

nt

ine

nt

al

Drift 4A

Co

Drift 4B

Pacific Ocean

Drift 5 Figure 25.8

ula

Drift 6

Figure 25. 7

Antarctic

lf al nt ine

Drift 8

Trough

Co

nt

68°S

Marguerite

.9 25

Drift 7

sh e

re gu Fi

Pe n

Fi

ins

gu

re

25

.1

0

66°S

100 km 70°S 80°W

75°W

70°W

65°W

60°W

Figure 25.6 Bathymetric map of the Pacific Margin of the Antarctic Peninsula (modified after Rebesco et al., 1998b). Grey arrows indicate the sediment drainage pattern on the continental rise through trunk-type deep-sea channels between sediment drifts. The grey areas with dotted line perimeter indicate the multi-beam bathymetric coverage displayed in Figures 25.7 and 25.8.

546

The Significance of Contourites for Submarine Slope Stability

sediment drifts composed of ‘‘atypical, hybrid contourites’’ (Lucchi and Rebesco, 2007). It is known from ODP Leg 178 drilling that the Middle Miocene to Holocene part of this succession is made of mainly fine-grained alternations of interglacial bioturbated biosiliceous silts and clays and glacial laminated, largely barren terrigenous silts and clays (Barker and Camerlenghi, 2002b). Table 25.1 includes a summary of the sedimentology and physical properties of the sediments within the drifts. Sub-parallel erosional channels and gullies cut into the sediment drift succession up to the sharp crest separating the steep flank from the gentler one (Figure 25.7).

km ex Al

100

de

an

0 rC

Depth scale (m) 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 4200

nn

ha

Drift 6

el

67°S

Fluid expulsion structures

Drift 7 ODP Site 1096 Coalescent slide scars Single slide scars

Ch

arc

ot C han 68°S ne

l

e

dg

pe

l

Drift 8

ta

slo

e

n ne

l ta

i

nt

78°W

77°W

76°W

e lf

sh

en

in

Co

C

t on

75°W

74°W

Figure 25.7 Shaded relief map of the bathymetry of Drift 7 (modified after Rebesco et al., 2007). The dotted line outlines the boundary of the Late Pliocene Alexander mega debris flow buried below the Alexander Channel. For location, see Figure 25.6. A multicolour version of this figure is on the enclosed CD-ROM.

547

J.S. Laberg and A. Camerlenghi

Such crests are formed by coalescent slide scars that give the crest an overall regularly undulating shape with concavities consistently oriented towards the steep slope leading to the adjacent deep-sea channel. Such sediment failure starts at 2800–2900 m water depth on 2–3° slopes and is likely induced by the undercutting of the walls of the turbidity channels flowing between the drifts (Volpi et al., 2005; Rebesco et al., 2007). Sediment-slab detachment along gentle (<1°) depositional surfaces of the sediment drifts occurs in water depths exceeding 3500 m (Figure 25.8). Conjugate steep scarps a few tens of metres high and tens of kilometres long are present on sediment drift 4A (Amblas et al., 2006). There is no evidence of the related deposit, due to the lack of data coverage in such deep and distal areas of the margin. Such slope instability is likely induced by the presence of weak layers possibly made of thin high-porosity siliceous ooze layers (Volpi et al., 2005). Another type of slope instability is represented by the deeply rooted creep of hundreds of metres of the sedimentary section below the gentle side of the drifts favoured by the presence of a conjugate sets of high-angle listric faults at the sea floor, which sole out at depth along a silica-diagenesis front outlined by a bottom-simulating reflector (BSR) (Volpi et al., 2003) (Figure 25.9). Such down-slope sediment movement favours shortening and large-amplitude folding at the very base of the gentle slope, in water depths exceeding 4000 m.

64°S

100 km

Continental rise Slab detachments

65°S Biscoe Trough

Bis co

e Is .

Continental shelf

66°S 70°W

68°W

66°W

Depth scale (m) –5000 –4000 –3000 –2000 –1000

0

Figure 25.8 Shaded relief map of the bathymetry of Drift 4A and adjacent Biscoe Trough on the continental shelf (modified after Amblas et al., 2006). For location, see Figure 25.6. A multicolour version of this figure is on the enclosed CD-ROM.

548

The Significance of Contourites for Submarine Slope Stability

Site SE 1096 S.P. 4890

6000

7000

Site 1095 NW 1000 1390

7400 200 km 0

20

twt (s)

4.5

5.0

5.5

Site SE 1096 S.P. 4890

6000

7000

Site 1095 NW 1000 1390

7400 200 km

Fluid escape

0

20

4.5

twt (s)

Low-amplitude folds 5.0

5.5

BSR

Figure 25.9 High-angle normal faults originated by creep rooted at the bottom-simulating reflector (BSR) in a multi-channel seismic profile on Drift 7 connecting ODP Sites 1095 and 1096 (modified after Volpi et al., 2003; with permission fromWiley-Blackwell Publishing).The change of reflectivity at the BSR outlines an Opal-A to Opal-CT diagenetic change. For location, see Figure 25.6.

According to Volpi et al. (2003), intergranular contacts among whole or broken siliceous microfossils prevent normal sediment consolidation. The sediment drift succession is therefore under-consolidated, with anomalously high amounts of

549

J.S. Laberg and A. Camerlenghi

interstitial water preserved at depths of nearly 550 m below the sea floor. Diagenetic alteration of biogenic opal-A to opal-CT causes a dramatic reduction of intra- and interskeletal porosity, allowing sediments to consolidate below this depth. The resulting overpressure and reduction in effective stress favours fluid expulsion along near-vertical, small-throw, normal faults extending from the diagenetic front to the sea floor. It is possible that these faults constitute a polygonal pattern (e.g. Davies, 2005). Evidence of deep-sea fluid expulsion from the sea floor of the distal sediment drifts at small rounded sediment mounds, possibly mud volcanoes, has been reported by Rebesco et al. (2007). One last evidence of sediment failure is the presence of a mega debris-flow deposit of Late Pliocene age (nearly 3.0 Ma) between Drifts 6 and 7, in the area presently incised by the Alexander Channel (Diviacco et al., 2006; Rebesco et al., 2006; Rebesco and Camerlenghi, 2008) (Figure 25.10). In the proximal part, the deposit has been eroded by subsequent turbidity flows. The deposit has a maximum observed thickness of 175 m, a width of up to 50 km and a length of up to 200 km. The estimated total volume is of about 1800 km3, whereas the inferred provenance from the continental slope implies a run-out distance exceeding 250 km. Isolated sedimentary mounds rise above the sea floor, overlying the debris-flow deposit. These coincide with underlying, narrow zones with high seismic velocity (Diviacco et al., 2006). The mounds are interpreted as the surface expression of fluid and mud expulsion from sub-vertical conduits that include methane hydrates or authigenic carbonates. If the fluid expulsion is due only to excess pore pressure induced by the rapid emplacement of the debris-flow deposit, the fluid-expulsion conduits are expected to be inactive. However, the regional deep-seated excess pore-fluid pressure induced by the diagenesis of biogenic silica may have contributed to the continuing activity of the vents until today. Slumping also has affected the upper part of a Neogene sediment prism containing sediment drifts offshore New Zealand (Fulthorpe and Carter, 1991 – their figure 5b), but to our knowledge no details exist on the extent and origin of this slumping. Sedimentary mounds SW shot

5.0

6750

6500

Sediment drift 7

Irregular upper boundary

Debris-flow deposit

ODP Site 1095 6250

6000

5750

5500

5250

Erosional base 5000

Alexander Channel

NE 4750

4500

4250

Sediment drift 6

5.5

6.0

Silica diagenesis BSR

6.5

Oceanic basement 7.0

twt (s)

Figure 25.10 Seismic cross-section between Drifts 6 and 7 illustrating the lateral extent of the Alexander Channel mega debris-flow deposit (after Diviacco et al., 2006, with permission from Springer).

550

25.3.3.

The Significance of Contourites for Submarine Slope Stability

Ocean gateway – unstable contouritic levees

The contourite depositional system of the Gulf of Cadiz and of the entire West Iberian Margin has been in place for the past 4.5–5 million years in response to the Late Miocene onset of the MOW (e.g. Ambar and Howe, 1979; Nelson et al., 1999; Stow et al., 2002b; Mulder et al., 2002; Herna´ndez-Molina et al., 2003; Hanquiez et al., 2007; Marche`s et al., 2007; Toucanne et al., 2007; see also Herna´ndez-Molina et al., 2008a). The forcing of the Mediterranean outflow through the narrow and shallow Gibraltar Strait imposes a current velocity in excess of 2.5 m s–1. The current deflects to the north to follow the continental slope of the Gulf of Cadiz in water between 300 and 1500 m deep before detaching from the sea floor and becoming a water mass comprised between deep and intermediate Atlantic waters. The suspended sediment carried by the MOW is derived primarily from the Guadalquivir River. The area of the Gulf of Cadiz contourite system that faces the Gibraltar Strait to the west is called the ‘‘Overflow Sedimentary Lobe sector’’ (Herna´ndez-Molina et al., 2003), and constitutes the part of the sea floor which is affected by the western (left) spill over of the Mediterranean Lower Waters of the MOW from the main channel. Adjacent to the northwest of this sector is the channels and ridges sector, where the Mediterranean Lower Waters split into the Main Branch and the Southern Branch. It is in these two sectors of the gulf where abrupt slope breaks, coalescent slide scars, erosional channels and sedimentary lobes at their termination suggest gravitational instability of the contouritic sediments and generation of turbidity currents (Figures 25.11 and 25.12). Their sedimentology and physical properties are summarized in Table 25.1. Sediment instability is also associated with the widespread occurrence of pockmarks (not shown in Figure 25.12) that, in turn, indicate fluid overpressure in the shallow subsurface (Mulder et al., 2003b, their Figure 4.6). According to Lee and Baraza (1999) and Herna´ndez-Molina et al. (2003), gravitational instability of contouritic sediments also occur at the shelf break of the Gulf of Cadiz, and is evidenced in the form of slide scars at depths of about 160 m and multiple slumps that reach the upper slope at a depth of about 450 m (Figure 25.11).

25.3.4.

Recapitulation of evidence

1. Thick contourite drift accumulations show escarpments, sediment detachment and/or associated debris-flow deposits on the upper and lower continental slope, as well as on the continental rise. 2. Slide scars associated with sub-parallel, down-slope-oriented channels are probably caused by turbidity-current erosion and undercutting. 3. Sediment creep may affect hundreds of metres of contouritic sediments due to under-consolidation caused by intergranular contacts among siliceous microfossils. 4. Diagenetic alteration of biogenic opal-A to opal-CT causes a dramatic reduction of intra- and interskeletal porosity. Small-throw normal faults

551

J.S. Laberg and A. Camerlenghi

8°W

7°30′W

7°W

Northern channel Central channel

Guadalquivir Drift

rn

tation

imen

he

en

Low

an

ne

OW M

ch

sed ergy

ain

ut

M

So

l

ch Sediment Waves

l

t defo

ne

imen

36°W

Figure 25.12A

Channels

Sed

an

Inactive

Lobe

Contouritic Figure 25.12B

rmat

ion

Figure 25.12C

Ancient sediment

Levee

failure

nel

han

Lobe

ec ctiv

Giant Furrows Mud Sediment and scours Volcano deformation

A

Portugal Shelf

Spain

Continental MOW

NA

SW

AI

NAD W

Gulf of Cadiz Morocco

Figure 25.11 Water-mass flow and main depositional features in the central part of the Gulf of Cadiz (adapted from Herna¤ndez-Molina et al., 2003 and Mulder et al., 2003b). MOW = Mediterranean Outflow Water; AI = Atlantic Inflow; NASW = North Atlantic Surficial Water; NADW = North Atlantic Deep Water. Sediment mass wasting occurs in the area named ‘‘sediment deformation’’ on the contouritic levee. Evidence of ancient slope instability is also found in the deeper part of the slope.

extend from the diagenetic front to the sea floor. Evidence of deep-sea fluid expulsion along these faults to the sea floor may occur on the distal sediment drifts, from where small, rounded sediment mounds, possibly mud volcanoes, have been reported. 5. Contourite drifts interbedded with glacigenic debris-flow packages have been involved in some of the largest submarine landslides known. On the Norwegian continental margin, they most likely occurred due to the combination of rapid loading of glacigenic sediments (causing under-consolidation of the contouritic sediments) and seismic activity. 6. Up to now, evidences of slope failure in contouritic sediments have been gathered essentially from mid- to high-latitude margins. It is not known whether this reflects the actual distribution of these phenomena or is biased by the absence of specific studies on low-latitude margins.

552

The Significance of Contourites for Submarine Slope Stability

7°08 W

SAR track

7°07 W

(a)

Contouritic levee

36° 08 N

36° 08 N

Slump deposits 7°08 W

290°N

Main mow channel floor

(a)

Failure scar Small sand wave crests

7°07 W 7°08 W

SAR track Slump deposits

7°07 W

Failure scars

(b) 36° 06 N

(b) 36° 06 N

7°01 W

7°00 W

7°07 W

35° 59 N 7°01 W

7°00 W

Failure scars

6°59 W (c)

35° 59 N

Giant contouritic levee

6°59 W

Contouritic levee

7°08 W

SAR track Small sand wave crests 255°N (c)

255°N

Channel floor

Figure 25.12 Morphological evidence of surface-sediment instability (Hanquiez et al., 2007). (a ^ c) Side-scan sonar (SAR) evidence for sea-floor instability on the Giant Contouritic Levee on the western bank of the main Mediterranean Outflow Water (MOW) Channel (after Hanquiez et al., 2007, with permission from Elsevier).White arrows are current directions. For location, see Figure 25.11.

25.4. 25.4.1.

D ISCUSSION

Gravitational instability of submarine slopes

There is a growing research activity focusing on various aspects of submarine slope failure, i.e. understanding of the mechanisms of sediment mass transport, trigger mechanisms and the consequences of submarine mass movements (e.g. Nittrouer, 1999; Yalc¸iner et al., 2002; Locat and Mienert, 2003; Mienert and Weaver, 2003; Mienert, 2004). The general viewpoint on submarine slides is that they occur because of a complex cause/effect relationship between applied and

J.S. Laberg and A. Camerlenghi

553

resisting stresses. Many times an external trigger, such as the cyclic ground motion imposed by earthquakes, needs to be invoked to explain why sediment resistance has decreased or the stress applied has increased, or both, so that equilibrium was lost and mass movement initiated. The size of known submarine slides varies from 101 to 104 km2. The volume of sediment displaced amounts up to 103 km3. The run-out distance of sediments before they are deposited or transformed into a density flow and ultimately turbidity currents can be several hundreds of kilometres. Interestingly, the initial discrepancy observed by Woodcock (1979) between modern (larger) and fossil submarine slides has been overcome following detailed investigations both on land and on the ocean floor (see Lucente and Pini, 2003). The geological approach to the understanding of submarine landslides takes into account various factors leading to a decrease in shear resistance of the sediments, among which the most important are excess pore pressure (gas charging, including destabilization of gas hydrates; high sedimentation rates), diagenesis (illite to montmorillonite transition, opal-A to opal-CT transition), under-consolidation (resulting from the rigid behaviour of biogenic particles and/or rapid sedimentation) and cyclic strength degradation (liquefaction). Important factors leading to an increased applied stress are sedimentary, glacial and tectonic overloading, oversteepening of the slope (e.g. tectonic uplift and tilting at the front of accretionary prisms, basal erosion on a submarine slope), seismic loading, storm-wave loading (applicable in shallow water), biological activity and human activity. On the contrary, the strictly geotechnical approach takes into consideration the index properties and geotechnical parameters of sediments, so that – with a given stress state – different sediment types may behave differently with respect to failure. The resistance to shearing of well-sorted sediments is generally lower than that of poorly sorted sediments. The water content also affects the shear resistance negatively. It follows that saturated high-porosity sediments are geotechnically weaker than low-porosity counterparts. In cohesive sediments, not only the amount of clay within a sediment correlates negatively with shear resistance, but also the type of clay is important. Montmorillonites, with their large amount of adsorbed water, affect the shear strength negatively in comparison with illite and kaolinite. All these factors influence the plasticity and liquidity indexes (derived from the Atterberg Limits), which are good indicators of the shear strength of the sediments, being less resistant if the indexes are high. Organic matter, if above about 2% of the total sediment volume, also affects the sediment strength negatively. In addition, the rate at which the stress is applied is one of the most important factors that influence the shear strength. Rapid loading will reduce the effective stress within the sediment and reduce the resistance to shear. Permeability, grain size and structural arrangement of the grains affect the ability of sediment to dissipate excess pore pressure. Of considerable importance is also the stress history of the sediment, because sediments that have been previously under higher load (over-consolidated, e.g. sediments below an erosional unconformity) tend to resist to failure better than identical sediments under a state of normal consolidation.

554

25.4.2.

The Significance of Contourites for Submarine Slope Stability

Characteristics of contourites that favour slope instability

The above summary of geological and geotechnical factors affecting the resistance to failure of marine sediments demonstrates how difficult it is to relate slope failure to one single cause, whether it results from sediment composition or from external triggers. We will discuss the geological environment of contouritic deposits and the geotechnical characteristics of contourites underneath, in order to understand why and how they can often lead to phenomena expressing submarine slope instability. The first general consideration is that contourites often are mound-forming sediments, as opposed to sheet turbidites, which are basin-fill sediments. During deposition, contourites build a slope that is generally a gentle one, in the range of continental-rise gradients (less that 1°). Only occasionally the depositional slope of contourites is higher. Like contourites, certain types of turbidites can be moundforming sediments with similar gradients. These are the channel–levee deposits, generally made of low-energy, fine-grained turbidites. Secondly, being the product of steady oceanic-circulation systems, including Western Boundary Current systems along ocean margins, contourite deposition is focused on continental slopes. Therefore, the mounds generated by deposition of contourites tend to grow on an inclined surface, and at least in some parts the slope is increased by the drift growth. Plastered drifts are typical examples of this situation. Thirdly, the action of a steady bottom current gives contourites a general characteristic of good sorting. Good sorting implies higher water content, which is, in turn, a recurrent characteristic of contourites. The three factors above concur to a state of weakness of contourites with respect to other continental slope sediments. Compared with channel/canyon-fill turbidites, channel-lobe turbidite complexes, hemipelagites and mass-flow deposits such as those deposited by debris flows and slumps originated from previous slides (e.g. Galloway, 1998), contourites are often weaker due to their good sorting and high water content. Lastly, the location along continental margins implies a depositional environment where levels of organic carbon are high and where migration of fluids from the deep, or generated in situ, is recurrent. Several factors that favour either a decrease in shear strength or an increase in shear stress are present in contourite deposits: earthquake activity can trigger liquefaction (Sultan et al., 2004) in clay-rich, high-porosity, under-consolidated silty contourites more than in other sediment types. Increased sedimentation rates in contourites themselves (e.g. Norwegian Margin: Laberg and Vorren, 2004) or giant contourite levees of the Gulf of Cadiz (Mulder et al., 2003b) lead to a decrease in effective stress and reduced resistance to shear. Consequently, rapid loading of glacigenic sedimentation on high-latitude margins has been numerically simulated (Kvalstad et al., 2005) and appears to be capable of producing an increase in pore pressure in fine-grained contourites. Field evidence comes from piezometer readings; these confirmed the numerical simulations at two locations outside the Storegga Slide scar, offshore Norway (Strout and Tjelta, 2005).

J.S. Laberg and A. Camerlenghi

555

Excess pore pressure within contouritic sediments can also develop from gas migration and gas-hydrate dissociation. From the area of the Storegga Slide, from where some of the most detailed results have been presented, modelling results have shown that reduced gas-hydrate stability conditions following ocean warming could have facilitated or contributed to submarine slope failure (Mienert et al., 2005), even though it is not considered as the primary factor. The area affected by mass wasting can be large because the contourites may have a very large areal distribution due to the intrabasinal extent of thermohaline current systems. On glaciated continental slopes, as offshore Norway, the slope morphology has an influence on the distribution of contourites, being thickest in slope embayments between the depocentres at the mouths of the large, transverse shelf glacial troughs where some of the largest submarine landslides are located (Bryn et al., 2005b). According to Locat and Lee (2002), the mechanisms for sediment transformation from slabs into flows are not well understood, but at least one likely factor is the initial density state of the sediment. If less dense than that in an appropriate steady-state condition, sediment is more likely to flow compared with sediments denser than that in steady-state condition (Locat and Lee, 2002). As discussed above, contouritic sediments may be under-consolidated due to a high sedimentation rate, generation of biogenic gas, gas migration from below and/or gas-hydrate dissociation, and therefore less dense than might be expected from the overburden alone. This may, at least partly, explain why some of the large landslides involving contouritic sediments also show a very long run-out on gentle slopes (Bryn et al., 2005a).

25.5.

SUMMARY

Slope instabilities in areas of contouritic deposits have been identified on continental slopes, on rises and in ocean gateways, including some of the largest submarine landslides known. Contouritic sediments are prone to failure because (1) their composition, geometry and location, well-sorted muddy or sandy sediments (i.e. weaker compared with poorly sorted sediments) forming sediment mounds on the continental slope and rise may experience liquefaction in response to cyclic loading (earthquake); (2) they are often characterized by high sedimentation rates implying high water content and under-consolidation and thus low shear strength; a substantial component of microfossils may also prevent normal sediment consolidation; (3) they could, due to their location on continental slopes, be subjected to rapid loading; on high-latitude margins, excess pore pressure may develop within contourites sandwiched between glacigenic sediments; (4) excess pore pressure could also develop from gas migration and gas-hydrate dissociation due to a relatively high organic-carbon content from often productive water masses along continental margins. When failing, the area affected by mass wasting can be large because the contourites may have a very large areal distribution due to the intrabasinal extent of the thermohaline and geostrophic current systems. Their weakness and resulting

556

The Significance of Contourites for Submarine Slope Stability

low shear strength may also, at least partly, explain why some of the large landslides involving contouritic sediments also show a very long run-out (up to hundreds of kilometres) on gentle slopes.

ACKNOWLEDGEMENTS We acknowledge the support to this work through the contribution of data from D.J.W. Piper, D.C. Campbell, A. Solheim, S. Ceramicola (seismic line acquired by OGS in the framework of the EU-STRATAGEM project), D. Long and C.K. Wilson. Helpful comments for improvement of the initial draft were provided by R. Urgeles and M. Vanneste. We also thank the colleagues in the EU-funded STRATAGEM, COSTA and EUROSTRATAFORM projects and the UNESCO project TTR for many valuable discussions on the issues presented in this contribution. This is a contribution to IGCP 511 (Submarine Mass Movements and Their Consequences). GRC Geocie`ncies Marines is supported by the Generalitat de Catalunya program for excellence research groups (ref. 2005SGR00152).

(a)

Minor drifts Sediment drift

Wave-field

20 km

Wave field Parasitic cones

Terraces

Concave slide scar

Moat Water depth (m)

(b)

800.00

1500.00

5 km

sediment drift

2000.

Slide scar

Sediment waves

Parasitic cones

Moat

Terrace Slide scar (c)

Slide Scar

63°N

20 km Sediment wave field

60°N

N

57°N

54°N

51°N

Sediment drift Multiple moat Sediment wave field

25°W

20°W

15°W

10°W

5°W

0°W

Moat –2300 m

Figure 3.2 Multibeam bathymetry of the Rosemary Bank Seamount, North Atlantic. (a) Shaded relief (from the northeast) and colour contoured multibeam bathymetry (Kongsberg EM120 12 kHz, 1°  1° beams, collected during cruise 99 of the RRS James Clark Ross) of the Rosemary Bank Seamount, North Atlantic Ocean. (b) Detailed view of the western spur region of Rosemary Bank showing the sediment wave-field, multiple moat and drift complex. Also indicated is a concave slide scar on the southern flank. (c) Perspective view of the seamount viewed from the northwest. Inset map shows the location of the seamount in the northern Rockall Trough (from Howe et al., 2006, with permission from Elsevier).

(a)

(b)

Figure 3.3 TOPAS sub-bottom profiles from the Fram Strait. (a):TOPAS sub-bottom profile from the western Svalbard Margin hemipelagic and contouritic sediments interrupted by debris flows and diapirs from the Storfjorden Trough Mouth Fan. (b) Small, currentcontrolled sediment drift developed on the Vestnesa Ridge, western Svalbard Margin, a region associated with abundant gas-escape-related pockmarks (adapted from Howe et al., 2008).

(a)

(b)

(c)

(d)

(e)

(f)

Figure 3.6 Sea-floor photographs showing contouritic sediments from the North Atlantic and Arctic Oceans (all images courtesy of David Hughes, the Scottish Association for Marine Science). (a) Sea-floor photograph of inclined glass sponges bending under the influence of Norwegian Sea Deep Water flowing west along the Faeroe ^Shetland Channel, north of the Wyville-Thomson Ridge, 1100 m. (b) Rippled, bioturbated contourite sandy muds in the northern Rockall Trough, south of the Wyville-Thomson Ridge, 1050 m. (c) Glaciomarine hemipelagites under seasonal sea ice on the Yermak Plateau, northern Fram Strait, 803 m. (d) Rippled contourite silty sands from 700 m on the Hebrides Slope, northern Rockall Trough. (e) Gravelly and sandy contourites from 1000 m on the Hebrides Slope, northern Rockall Trough. (f ) Rippled silty/sandy contourites with emerging clasts from 1000 m on the Hebrides Slope, northern Rockall Trough.

De

40

o

n

ar –F

d

lan

Ice

oe

nk

°

ait ssa Str e Pa nel

60

°

60

ge

rk

a nm

a Ch

Ba

r Fa

°

40

°

eg

20° y

cil

Si

An

of

W

Strait of Gibraltar

Discovery Gap

it ra

20°

St

ind wa rd

Pa

ad ss ag Pa a–J e ss un ag gf e er n

Charlie Gibbs FZ

Vema Gap

ara

Ce

0°

nche

Roma

0°

FZ

lain

al P

yss Ab

20°

°

40

°

0°

30

°

°

60

°

90

40

60

°

South Sandwich Island Arc Gap

Sh a Pa g Ro ss ck ag s e

V Ch ema an ne l

20°

Figure 4.A_(2a) A representative selection of deep-ocean passages in the Atlantic Ocean and the Mediterranean Sea (courtesy Dr J. Whitehead, WHOI, Woods Hole, MA, USA). Water exchange across sills is called ‘‘overflow.’’ Straits show out- and inflows separated by a variable shear zone between the surface and the bottom. Green indicates depths <4000 m, blue regions mark basins with depth >5000 m.

°

°

60

60

40

°

40

°

20°

20° Red Sea Indonesian Philippine Basin

0°

0°

Ninety ea Ridge st Amirante Passage

20°

20°

40

°

40

°

60

80°

13

0°

° 60

°

°

30

Figure 4.A_(2b) A selection of deep-ocean passages in the Indian Ocean (courtesy Dr J. Whitehead, WHOI, Woods Hole, MA, USA). Green indicates depths <4000 m, blue regions mark basins with depth >5000 m.

°

°

60

60

40

°

°

Funka Bay

40

20°

20°

0°

0° Ecuador Trench

20°

e

Spencer Gulf

Samoan Passag

20°

40

°

40

60

°

60

°

°

Figure 4.A_(2c) A selection of deep-ocean passages in the Pacific Ocean (courtesy Dr J. Whitehead, WHOI, Woods Hole, MA, USA). Green indicates depths <4000 m, blue regions mark basins with depth >5000 m.

Figure 4.B Generalized representation of the three-dimensional water-mass transport in the North Atlantic (courtesy Dr V. Byfield, NOC, Southampton, UK). At the surface, the Gulf Stream and its extension advect warm tropical waters northward toward the Labrador and Nordic Seas. The considerable heat loss to the atmosphere generates dense waters at high latitudes that initiate convection of freshly ventilated water into the abyss. Two resulting cold water masses (Upper and Lower North Atlantic Deep Waters: UNADW, LNADW) leave the North Atlantic southward in compensation for the warm inflow at the surface. The two ‘‘U-turns’’ in the circulation pattern are symbolized by the two vertical arrows.

ISOW C.G.F.Z.

Latitude

54°N

RAFOS 490 RAFOS 491 RAFOS 492 RAFOS 493 APEX 289

51°N

48°N

DEBC

45°N DWBC

36°W

30°W

24°W

18°W Longitude

12°W

6°W

Figure 4.C Float trajectories indicating the Deep Western Boundary Current (DWBC) at roughly 1700 m depth along the eastern flank of the Mid-Atlantic Ridge in the eastern North Atlantic (after Machı´n et al., 2006; with permission from CSIC, Barcelona, Spain). Note that the boundary current band is superimposed by subsurface eddies. Depth contours at 2000 m. All floats were launched simultaneously east of Charlie Gibbs Fracture Zone (CGFZ). RAFOS floats collect daily positions by underwater acoustics. APEX floats cycle between the surface and their mission depth every 10 days. Other abbreviations: DEBC = Deep Eastern Boundary Current; ISOW = Iceland Scotland OverflowWater.

rea

al a

st oa

C

n nyo

Ca

Deep

-sea

fan 500~3000 m

Lobe Con

tine

ntal

mar

mp

gin

Slu

in

la

p al

s

ys Ab

Figure 4.D Morphological diversity of continental margins. Figure adapted by Blondel (2003) from an anonymous author (personal communication), reproduced with permission from Springer and Business Media. Contour currents parallel the margins. Slumps, fans, and canyons disturb their preferred flow directions.

65°

W

60°

55°

50°

45°

60°

60°

62°

62°

S 64° 65°

S

W

60°

55°

50°

64° 45°

Figure 4.E An interoceanic deep boundary current at the northeastern tip of the Antarctic Peninsula links the Atlantic with the Pacific Ocean (after Nowlin and Zenk, 1988; with permission from Elsevier). This peculiar ^ unnamed ^ westward counter contour current beneath the Antarctic Circumpolar Current (ACC) was first documented in the shown short hydrographic sections and by moored current meters.

50°

40°

30°

South Georgia Island Scotia Sea

55°

South

n S a dw c Ar

55° Current

55°

20° South Atlantic

ich

60° 60° m otto

B

a

t rc

ic

arctic Peninsu Ant la

65°

Sea

t

70°

n

65°

Weddell

A

Latitude (S)

60°

70°

Current evidence Strong Moderate Weak Murky bottom water Rock outcrops and talus Current direction from oriented photographs Sediment transport direction R Scattered rocks & boulders Large benthic animals Rare Occasional Abundant

75° 70°

60°

50°

40° 30° 20° Longitude (W)

Less than 2000 m (contours in meters)

Figure 4.F Abyssal and contour currents in the northwestern Weddell Sea and around the Antarctic Peninsula (after Hollister and Elder, 1969). Current evidence and direction are inferred from photographic images of the sea floor.

38°W

t

en

.

p Ca

St

nc Vi

36°W

Gibra ltar

f fo ul z G di Ca

800 m

34°W

1200 m

10°W

8°W Longitude

6°W

Figure 4.G Modeled outflow path of the core layer of the Mediterranean outflow in the Gulf of Cadiz (after Price and Baringer, 1994; with permission from Elsevier). Soon after leaving the Strait of Gibraltar on the right side, the outflow follows the 800 m contour for most of the course through the gulf.When approaching Cape St Vincent at the southeastern tip of Europe, it sinks to about 1100 m depth. Contours of isobaths are 200 m apart. Dots along the path are at intervals of 12 h.

Figure 4.3 Highly simplified cartoon of the global thermohaline circulation (THC), modified from the original presentation of the oceanic ‘‘conveyor belt’’ by Broecker (1991) (courtesy Dr S. Rahmstorf, PIK, Potsdam, Germany). Surface and near-surface waters (red pathways) flow toward convection regions (orange ovals) in the Labrador and Nordic Seas, and in the Weddell and Ross Seas. They recirculate as deep (light blue) and abyssal (dark blue) currents, and participate in basin-scale slow upwelling in the interior of all three oceans. Sea-surface salinity controls the convection process at high latitudes decisively.Values between 34 und 36 in the open ocean and in marginal seas are indicated by light blue regions. Anomalously low (<34) surface salinities are marked by blue areas; those >36 are shown in green color.

Brazil Basin 0 500 1000

Water depth (m)

1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 38

0.0

36

0.1

34

0.2

32

0.3

0.4

30

0.5

28 26 Longitude (°W)

0.6

0.7

24

0.8

22

0.9

2.0

20

5.0

18

16

8.0

22.0

Diffusivity (10–4 m2 s–1)

Figure 4.5 Directly observed distribution of diffusivity in the Brazil Basin of the South Atlantic (from Polzin et al., 1997; with permission from The American Association for the Advancement of Science). The zonal section runs from the continental rise off Brazil (left side) toward the western flank of the Mid-Atlantic Ridge. Note the nonlinear scale for diffusivity. High correlation between diffusivity and bottom roughness is found on the slope of the ridge on the right side. The white line marks the observed depth of the 0.8°C potential isotherm.

Pressure (dbar)

3600

1.5

2

1.5

1

1

3800

1.5

0.5

4000

0

0.5

1

4200

0.5

4400

0

4600 Potential temperature (°C)

4800

–39.5

–39.4

Pressure (dbar)

3600

–39.1

–39

5.9

5.9

5.8 5.8

4200 4400 5.9

4600 4800

–38.9

Clear water minimum 5.8

3800 4000

–39.3 –39.2 Longitude (°W)

6

–1 Attenuation Att. (0.1 m–1)(0.1 m )

–39.5

–39.4

–39.3 –39.2 Longitude (°W)

–39.1

–39

–38.9

Figure 4.7 Vertical distribution of abyssal potential temperature (top) and of red-light attenuation (bottom) in a zonal section between the Santos Plateau and the Vema Channel of the South Atlantic at approximately 29°S. The Vema Channel represents a throughflow channel between the Argentine and Brazil Basins. The graph intersects the channel extension transporting cold (yellow ^ blue, <2°C) Antarctic Bottom Water (AABW) northward at a position about 210 km downstream of the Vema Sill (for location of the Vema Channel, see Figure 4.2 in the book or Figure 4.A_(2a) on the CD-ROM). Stations are indicated by open triangles (r).The x-symbols (6) denote the area of the strongest near-bottom current, which coincides with the highest attenuation (red, >0.6 m1).This co-occurrence indicates the active role of the abyssal current as a sediment-transporting contour current (Dr A. Macrander, AWI, Bremerhaven, Germany, 2006, personal communication).

60 Latitude (°N) 50 Velocity (cm s )

200

400

2 4 6 8

40

10

50

40

Longitude (°W)

20

10 (103 m2 s–1)

Figure 4.13 Streamlines inferred from subsurface float observations in the Iceland Basin at the level where the Iceland Scotland Overflow Water (ISOW) from Faroe Bank Channel in the north encounters Labrador Sea Water (LSW) entering through Charlie Gibbs Fracture Zone in the south (53°N) (from Bower et al., 2002a; with permission from Macmillan Publishers). For position see also Figure 4.2 or Figure 4.A_(2a) on the CD-ROM. The wellresolved contour currents at a depth between 1500 and 1750 m appear along the Reykjanes Ridge where the streamlines lie closely together. The inset gives a measure for current speeds. Arrowheads indicate the circulation direction.

C6A (21.3 Ma)

(a)

50°W

40°W

30°W

20°W

56°S

Present

60°S

64°S (b)

Figure 4.15 Sketch of bottom flows in the eastern Scotia Sea and northern Weddell Sea showing pathways of the Antarctic Circumpolar Current (ACC), Weddell Sea Deep Water (WSDW), and Weddell Sea Bottom Water (WSBW) (from Maldonado et al., 2003; with permission from Elsevier). A reconstruction from 21.3 Ma ago (top) can be compared by the present situation (bottom).The interaction of the various current systems generates a wide variety of contourites.

La and teral re- trans sus p pen or t sio n

Win d upw -drive n elli ng

W MO

INL NL B

0

Ib

Sl

op

e

cu

e

ri

a

rre

20

00

wn Do ling l e w L BN

40

00 41

°

nt m 200 0m

100

9°

n nyo

Ca

4 La 0° titu de (

N)

39

10°

de

itu ong

)

(W

L

°

Figure 4.18 Cartoon of the Iberian continental margin reflecting major processes and currents affecting particle transfer over the shelf edge and off the slope (from Pingree et al., 1999; with permission from Elsevier). The shown European Slope Current and the advection of Mediterranean Outflow Water (MOW) represent contour currents. Canyons act as pathways for direct transport of particles (occasionally by turbidity currents) from the shelf to the deep sea (Courtesy: Dr H. v. Haas, NIOZ,Texel,The Netherlands).

cm 5 4 3 2 1 0

Figure 5.13 Core photograph showing double mud layers (arrow) in Pliocene sand. Edop Field. After Shanmugam (2003); with permission from Elsevier.

Figure 5.16 Satellite image of internal waves in the Sulu Sea between the Philippines (to the northeast) and Malaysia (to the southwest). Sunlight highlights delicate curving lines of internal waves moving to the north toward Palawan Island. The Sulu Sea is stratified with water layers of differing densities. Unlike surface waves, however, internal waves can stretch tens of kilometers in length and move throughout the ocean for several hours. This true-color Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) image was acquired on April 8, 2003. Image courtesy Jacques Descloitres, MODIS Land Rapid Response Team at NASA/GSFC. Uniform Resource Locator (URL): http://earthobservatory.nasa.gov/ Newsroom/NewImages/images.php3?img_id=15334 (accessed May 12, 2007).

0

144 143 142

1 141 140 139 2 138 137 3 136

4

135

137 136 Interval number in measured section

Siliceous shale

134

5m

133

Turbidite

132 131

Contourite

Figure 7.3 Measured section through part of the Pingliang contourite drift, showing the relationship between turbidite and contourite in vertical sequence. Note that turbidites are overlain by contourites in the 136^137^138 and 140^141 intervals. Siliceous shale is a shale with a substantial cherty component, whereas all other lithotypes are carbonates.

(a)

(c)

(d)

(b)

0.4 mm

(e)

0.4 mm

Figure 7.7 Lenses of contourites consisting of echinoderm clastic limestones in the lower part of the Wulalike Formation of the middle Ordovician in the Zhouzishan Mountains, northern China. (a) Occurrence of echinoderm clastic limestones. (b) Series of beds with cross-bedding in echinoderm clastic limestones. (c) Large-scale cross-bedding in echinoderm clastic limestones (lamina dipping to left of the photo, though slightly). (d) and (e) Secondary enlargement of encrinite in echinoderm clastic limestones and lime mud infill (in (e)).

Figure 8.2 Analytical flow chart for size analysis and composition slicing. This is a very complete scheme for sedimentological measurements and preparation of samples for almost all geochemical and isotopic data. Critical actions are in shaded boxes. There are four areas: (I) analysis of the bulk sample for carbonate and opaline silica content and, by difference, terrigenous matter content; (II) determination of mineral magnetic properties; (III) processing of coarse fraction (‘‘cf ’’) (sand) to yield samples for microscopic, isotopic and geochemical analysis; and (IV) processing of the fine fraction (‘‘ff ’’) to yield size distributions of the terrigenous and biogenic components, as well as the composition of the fine fraction in terms of terrigenous and biogenic (carbonate and opal) components.The principal operations are numbered and are shown in shaded boxes.The term ‘‘calcimetry’’ refers to determination of calcium carbonate content, which in this scheme is implicitly by a method which involves dissolution and removal of the carbonate, for example, by coulometry, or measurement of CO2 gas released in the Chittick apparatus. If carbonate is determined by the CHN analyser, a modification to this scheme is necessary. After Robinson and McCave (1994) and McCave et al. (1995b).

C + FS 50

SS

Framework

Grain-size component percent

Interstitial

100

Sand

0 Sorting by selective deposition

Accumulation rate (cm ka–1)

20

Selective deposition and winnowing

C + FS 10

SS 2 0 0

5

10

15

Flow speed (cm s–1)

20

25

Sand

Figure 8.8 Hypothetical variation of sedimentation rate with increasing current speed shown as deposition flux (below) and component percentage (top). C þ FS = clay and fine silt (<10 mm), SS = sortable silt (10^63 mm). At zero speed, a pelagic rate of 2 cm ka1 is assumed, and erosional winnowing is assumed to occur above 20 cm s1. Above 20 cm s1, mud will be mainly interstitial in sand, and above 30 cm s1 (not shown) the sand will be sufficiently mobile to contain little mud at all and be forming ripples and sand waves.The curve is shown as peaking between 10 and 15 cm s1 but this is not at all well known and may well be dependent on the magnitude ^ frequency structure of deposition and erosion events. A peak between 5 and 10 cm s1 would be entirely feasible as some records suggest the onset of surface erosion above 10^12 cm s1. From McCave and Hall (2006).

0.4

U ′ – U ′3

0.3

0.2

0.1

0

0

0.2

0.4

0.6

0.8

1

U′

Figure 8.9 The deposition rate function (U0 ^ U0 3) combining advection and sub-layer effects as a function of increasing flow speed up to the depositional limit U0 = 1. (U0 = Ug /Ud, the ratio of geostrophic flow speed to critical deposition flow speed).

East

West

3.5 kHz

10 kHz

Figure 8.10 Acoustic profiles across a sediment drift. Upper figure: 3.5 kHz WNW-ESE profile across northern Gardar Drift at 60°N, 23°W, showing a reduced sedimentation rate on the eastern slope of the drift by closer spaced reflectors. Lower figure: 10 kHz echo sounder record along the same track showing higher amplitude reflection (redder) of the ‘‘harder’’ sea floor that is more strongly affected by currents, corresponding to coarser size and slower net accumulation rate (from McCave, 1994). The coarser sediments on the eastern side of the drift are shown by Bianchi and McCave (2000, their Figure19) to be finer than on the western side.

s

w m

and

– w m

w

Figure 9.1

Bottom-current characteristics: a schematic summary of principal features.

Outflux in bottom current 200–250

Influx for drift construction Turbidity currents 30–50

Pelagic settling 30–50

Slope spillover 30–50

Pelagic settling 100–150

Deposition on drift 500–600

Local erosion by bottom current 30–50

Upstream influx to bottom current

Bottom-current influx 600

Slope spillover Turbidity 50–150 currents 100–150

Bottom-current erosion 200–300

Figure 9.2 Estimated sediment flux from different inputs into the Deep Water Boundary Current that flows over and feeds sediment to the Eirik Drift. Arrows indicate sediment flux, not current position.

Extra fine tail due to bioturbation

60 Saltation load coarse silt and fine sand part traction/part suspension

40

20 Bedload

0

0 1 2 3 4 5 6 7 1000 250 63 16

Medium and coarse sand deposition from traction

8 4

40

20

2 250

4 63

6 16

8 4

10 1

φ Grain μm size

80 Suspended load clay, silt, v.fine sand sorting by current deposition from suspension

60

40 20

Saltation load minor

0 4 63

6 16

8 4

10 1

φ Grain μm size

80

60

Sand fraction dominantly biogenic components most in suspension minor saltation

Sortablesilt fraction part current sorting part component sorting

Suspended load

Cumulative weight (%)

60

100%

Sortable silt fraction

2 250

Suspended load (100%) clays and silts (>90%) minor biogenic very fine sand deposition from suspension

(d) Biogenic-rich contourites

(b) Contourite silts 100%

80

0

9 10 φ Grain 1 μm size

Cumulative weight (%)

Cumulative weight (%)

80

(c) Contourite muds 100%

Suspended load clay, silt, v.fine sand deposition from suspension

Cumulative weight (%)

(a) Contourite sands 100%

40

20

Saltation load minor

0 2 250

4 63

6 16

8 4

10 1

φ Grain μm size

Figure 9.4 Grain-size distribution and characteristics of a range of different contourite facies, plotted as smoothed cumulative frequency curves. (a) Sandy contourites, (b) silt and silt ^ mud contourites, (c) muddy contourites and (d) biogenic-rich contourites.

(a)

(b)

5 cm

Figure 10.5 Outcrop examples of ripple lamination and related small-scale structures in ancient contourite deposits. (a) Cross-laminated calcareous contourite from the Ordovician of the Lachlan Fold Belt, eastern Australia ( Jones et al., 1993). Climbing ripples, flaser bedding, and erosional reactivation surfaces are present (courtesy B.G. Jones). (b) Subhorizontal to sinusoidal lamination and ripple lamination in bottom-current-reworked fine sands. Note the internal erosional surfaces. Jurassic, Neuquen Basin, Central Andes, Argentina (Palma, Martı´n-Chivelet and Lo¤pez-Go¤mez, unpublished data).

(a)

(b)

5 cm

Figure 10.8 Medium- to large-scale traction structures in calcareous contourite deposits (Maastrichtian, Subbetic Zone near Caravaca, Spain). (a) Planar horizontal lamination in fine calcarenites from a contourite bed. Note the intercalated beds with current ripples. (b) Crossbedding in calcareous contourites. Note the sharp bottom contact of the laminated contourite bed, which overlies hemipelagic carbonate.

Figure 10.9 Large-scale cross-stratified calcareous contourites from the Lower Ordovician of Jiuxi (Hunan Province, China). These fine-grained deposits are believed to result from the development of mud waves and erosional furrows on the sea floor under the influence of a semipermanent bottom-current regime (Duan et al., 1993; Photo courtesy T. Duan).

(b)

(a)

(c)

Figure 10.10 Cross-bedded, glauconitic, calcarenite sediment drifts of Late Oligocene age; Weka Pass Limestone, Waihao Forks, South Island, New Zealand (Carter et al., 2004). The drifts represent resumed sedimentation above a regional unconformity, present in strata deposited between shallow shelf and abyssal water depths, and are inferred to have been cut by vigorous bottom currents associated with the onset of the predecessor Antarctic Circumpolar Current. (Photos courtesy R.M. Carter).

(b)

(a)

10 cm (d)

(c)

2 cm

2 cm

Figure 10.11 Small-scale erosional structures and winnowing deposits. (a) Scour marks (including flute casts), reflecting erosional processes on a cohesive substratum consisting of hemipelagic carbonate mud. Maastrichtian, Caravaca, SE Spain. (b) and (c) Small-scale erosional surfaces. In both cases, erosion precludes de sedimentation of sandy contourite facies. In (b), the erosive surface cuts down into a laminated contourite, whereas it affects pelagic carbonate in (c) Maastrichtian, Caravaca, SE Spain. (d) Small remobilization level with several shell fragments of large hemipelagic bivalves (inoceramids). Upper Cretaceous, Xixona (Alicante, SE Spain).

(a)

(b)

Figure 12.1 The 115^125 cm interval containing thin carbonate turbidites (intervals i ^ v) at ODP Site 1062 (afterYokokawa, 2001): (a) close-up and (b) X-radiograph.

(a)

Figure 12.2 Cyclic variations in lithology at ODP Site 1165 (from Shipboard Scientific Party, 2001a). (a) Minor silt laminae and stringers and strong bioturbation in dark diatom-bearing clay in unit II. (b) One of the locally occurring 1-cm-thick stiff green clays beds. (c) Dark-grey thinly bedded fissile claystones with planar lamination and cross-stratification, fine green laminae with chert layers and locally calcareous intervals.

(b)

Figure 12.2

(c)

(Continued)

Figure 12.3 Depth profiles of Mn (a) and U (b) ratios normalised to Li (from Howe et al., 2004, with permission from Elsevier). MC and BC refer to multicore (MC) and boxcore (BC) samples for each station.

4000 ppm

3000

2000

Mn

1000

(b) 0

4000 ppm

3000

2000

1000

Mn

0

(a)

132 342 134 344 136

Core depth (cm)

346 138 348 140 350 142

Figure 12.4 Manganiferous silty ^ clayey succession in core KS 8813 in the Sao Thome deepsea channel ^ levee system (South Brazilian Basin). Significant increases in Mn concentration are located in the dark layers (after Gonthier et al., 2003). (a) Mn sequence showing erosional surfaces and very black and Mn-rich layers. (b) Mn sequences showing more diffuse black laminations.

(a)

(b)

(c)

Figure 13.1 Muddy and silty contourite facies. Rockall Trough, NW UK Continental Slope. Core width 8 cm. (a) Silts and muds interlayered. (b) Mottled silt horizon within structureless muds. (c) Bioturbated muds with very indistinct lamination apparent in parts.

Figure 13.2 Silty and muddy contourite facies, showing homogeneous, bioturbated muds, mottled silts and indistinctly laminated silty mud facies. Gulf of Cadiz contourite depositional system. Core width 8 cm.

Figure 13.3 Bioturbated sandy contourite facies. These are poorly sorted, muddy sands, with some indication of indistinct parallel lamination. Brazilian Continental Slope (Campos Margin). Core width 5 cm (from Viana et al., 2002a; with permission from The Geological Society, London).

C4

C4

C3

C3

C2

C1

C2

Figure 13.4 Muddy, silty and sandy contourite facies, showing part of standard C1 to C5 contourite facies sequence. From base to top: C1 mud, C2 mottled silty mud, C3 muddy sand, C4 mottled silty mud and C5 (this division not shown) mud. Note white bioclastic shell debris in parts of C3, bioturbation throughout, and partly disrupted discontinuous lamination with some sharp contacts. Faro Drift, Gulf of Cadiz contourite depositional system. Core width 8 cm (from Stow et al., 2002b; Stow 2005; with permission from The Geological Society, London).

Figure 13.5 Laminated sandy contourite facies, showing parallel and inclined lamination as a result of sand wave migration. Gulf of Cadiz contourite depositional system, from sand sheet in proximal scour and ribbon sector. Core width 10 cm.

Figure 13.6 Pebbly sand and gravel-lag contourite facies ^ structureless with irregular concentration of coarse-grained clasts. Faeroe ^Shetland Channel (gateway), NW UK continental margin. Core width 10 cm (from Akhurst et al., 2002; with permission from The Geological Society, London).

(a)

(b)

(c)

Figure 13.7 Bioclastic, chemogenic and mud-chip contourite facies. Vema Channel and Contourite Fan, Brazilian Basin. Core width 7 cm. (a) Lower part of thick manganiferous crust at top of section (current sea floor) overlying calcareous marl contourites (yellowish) and further interbedded black manganiferous deposits. (b) Sharp contact (erosive bottom-current event) between brownish-yellow calcareous marl contourites, with thin manganiferous horizons, and greenish muddy contourites. (c) Muddy contourites (as in b) with thin micro-brecciated horizons of mud-chip contourites (from Fauge' res et al., 2002a; with permission fromThe Geological Society, London).

(a)

(b)

(c)

Figure 13.8 Bioclastic, chemogenic and muddy contourite facies.Columbia fan ^ drift system, Brazilian Basin. Core width 7 cm. (a) Yellow and black manganiferous contourites, with dark manganiferous crust (centre) and diatomaceous contourite mud (near top). (b) Greenish muddy contourite with thin disrupted layers (near centre) of shale-chip contourites. (c) Yellow and black manganiferous/calcareous contourites with irregular lamination and marbling structure. (from Fauge'res et al., 2002b; with permission fromThe Geological Society, London).

80°

75°

70°

65°

40°

40° 108 107

H.C.106 35°

35°

105

H.O.R.

104 102 103

30°

Bl.O.R. 30°

101 98

25°

100

25°

Bah.O.R.

99

20°

20° 80°

75°

70°

65°

Figure 14.1 Physiography of a portion of the western North Atlantic (from Hollister et al., 1972), showing the shaping of the continental rise by deep geostrophic currents. Bl.O.R. = Blake Outer Ridge; Bah.O.R. = Bahamas Outer Ridge; H.O.R. = Hatteras Outer Ridge; H.C. = Hudson canyon. Numbers 100, 101, . . . refer to the location of DSDP Leg 11 drilling sites.

(3)

Upper Pliocene–Present

WBUC

35°

(c)

Blake Outer Ridge

30°

80°

75°

WBUC

(b)

Bahama Outer Ridge

(2)

Middle Miocene– Upper Pliocene 35°

WBUC

(a) 30° Blake Outer Ridge

80°

75°

GS GS

Bahama Outer Ridge

(5)

(1)

Lower Oligocene– Middle Miocene 35°

Zero thickness

................. ................. ................. ................. ................. ................. ................. ................. ................. ................. .................

Outcrop

200

Bathymetry

1000

Isopach (m)

.

Drill site

30° Sediment waves 80°

75°

Bahama Outer Ridge

Blake Outer Ridge

102

(4)

Figure 14.4 Interpretation of the Blake Outer Ridge growth. (1^3): Isopach maps showing the distribution and thickness variations (modified from Tucholke and Mountain, 1986). (1) = Lower Oligocene to Middle Miocene; (2) = Middle Miocene to Upper Pliocene; (3) = Upper Pliocene to Present). (4) = captions of the maps. (5) = hypothetical reconstruction of the ridge development above horizon Au (from Ewing and Hollister, 1972). The initial Oligocene building (5a) is characterized by a surface (the ‘‘Gulf Stream’’: GS) and deep current interaction (Western Boundary Under Current:WBUC), predominant sediment aggradation with low southeastward deposit migration; (5b) corresponds to a time of predominant contour-current depositional processes, strong southeastward deposit migration, and initiation of a secondary drift with a westward migration, during the early Miocene. (5c) represents the last development of the ridge (Quaternary) characterised by contour-currentcontrolled erosion (northern flank of the ridge) and deposition (top and southwestern flank).

NNW (A)

SSE

Gloria Drift

Ran ridge

Eirik drift

(A)

Imarssuak channel

4

Crest

5 6

Depth in (km)

Gloria drift

Gloria Drift NE

Crest

(B)

(1)

50 km

Mid Labrador sea ridge

NW Midocean channel

SW (B)

(2)

Figure 14.5 The abyssal Gloria sheeted drift (from Egloff and Johnson, 1975; with permission from the Canadian Research Council Press). (1) Morphological map. (2) Seismic line.

Elongate contourite mounds

10 km

A

A′

NW

SE

Moats Elongate contourite mounds

(Record missing)

100 ms

0

Sediment waves

100

m

120

00

14

0

1 km

(a)

(b)

Figure 14.15 The Faroe ^Shetland Channel (Bulat and Long, 2001; with permission from Springer). (a) Image performed from a 3-D seismic data set showing the lower part of the eastern flank and the bottom of the channel (1000^1500 m water depth). Note the current bedforms interpreted as elongate contourite mounds and sediment waves, and the interference pattern between these bedforms oriented NE ^SWand NW^SE, respectively. (b) Seismic line crossing the elongate mounds: these bedforms have been interpreted as built by an along-slope northeastward flowing current (profile location in a).

Figure 15.1 Global bathymetric map showing location of well-documented examples of bottom-current sediment waves (white circles), and location of the case-study area (black rectangle; Figure 15.2). Numbered white circles refer to the following areas: (1) Norwegian Sea ^Atlantic Ocean gateway (e.g. Roberts and Kidd, 1979; Dorn and Werner, 1993; Manley and Caress, 1994; Howe, 1996; Kuijpers et al., 2002; Masson et al., 2002; Wynn et al., 2002a; Howe et al., 2006); (2) Mediterranean Sea ^Atlantic Ocean gateway (Kenyon and Belderson, 1973; Habgood et al., 2003); (3) Central Mediterranean Sea (Marani et al., 1993); (4) Blake ^ Bahama Outer Ridge (Flood, 1994; Flood and Giosan, 2002); (5) Gulf of Mexico (Kenyon et al., 2002); (6) Carnegie Ridge (Lonsdale and Malfait, 1974); (7) Argentine Basin (e.g. Flood and Shor, 1988; Flood et al., 1993; Manley and Flood, 1993a, b; Von Lom-Keil et al., 2002); (8) Antarctic ^Atlantic Ocean gateway (Cunningham and Barker, 1996; Howe et al., 1998); (9) Chatham Ridge (e.g. Lewis and Pantin, 2002). Base bathymetric map obtained from USGS website: http://walrus.wr.usgs.gov/infobank/gazette/html/bathymetry/gl.html.

Figure 15.8 Summary interpretation of bottom-current features in the Faroe ^Shetland and Faroe Bank Channel, based on side-scan sonar images and 3.5 kHz profiles (modified from Masson et al., 2004; with permission from Blackwell Publishing). Symbols showing surficial sediment classification are based on core/dredge samples and/or sea-floor photographs. FSC = Faroe ^Shetland Channel; FBC = Faroe Bank Channel; WTR = Wyville-Thomson Ridge. For location: see Figure15.2.

Figure 15.9 Reconstructed bottom-current flow field through the Faroe ^Shetland and Faroe Bank Channel based on the distribution of bedforms and related bottom-current-generated features shown in Figure 15.8 (modified from Masson et al., 2004; with permission from Blackwell Publishing). For location: see Figure 15.2. Red arrows indicate currents related to northeastward transport of near-surface water masses; blue arrows relate to southwestward transport of Norwegian Sea Deep Water (NSDW); orange arrows show recirculation of NSDW (and intermediate water masses?) at the southern and eastern margins of the Faroe Bank and Faroe ^Shetland Channels, respectively. Black dots show the locations of sea-floor photography stations.

Figure 16.9 Detailed morphological reconstruction of a buried contourite-slide depositional system based on 3-D seismic data from the continental slope west of Shetland. (a) Seismic cross-section with reflectors representing the Intra-Neogene Unconformity (INU) and two internal horizons demarcating the top of unit B1 (mounded drifts of Pliocene age) and of unit B2 (multiple slide events of early Pleistocene age) (Knutz and Cartwright, 2003). (b) Amplitude attribute extraction of the Top B2 surface, illustrating lobate slides and thin down-slope channels outlined by light contrasts presumably representing sand-rich features (Knutz and Cartwright, 2004). (c) 3-D view of time-depth maps representing the INU and the Top B1 surface. The series of mounded drifts separated by moat ^ channels have developed in response to southward-flowing bottom currents that were active during the Pliocene ^ early Pleistocene. The basal reflector represents the regional erosional unconformity (INU) of late Miocene ^ early Pliocene age. Sea-floor depressions created by slumping are thought to have promoted the initial accumulation of contourites by acting as sediment traps for fine-grained clastic material supplied by along-slope currents.

(m) 2 12

1

11

10 0

12

9

11 8 10

7

9 8 7

6

6 5

5

4 4

Figure 17.1 Oligocene calcareous sediments interpreted as calcicontourites, Petra Tou Romiou, southern Cyprus. Detailed log of section shown in photograph, together with representative microfacies at sample points 4^12 (from Turnbull, 2004, with permission). Cyclic grain-size variations evident over approximately 50 cm of the section, between bioturbated calcilutite and lenticular calcarenite macrofacies. Thin-section micrographs show a clear sequence from wackestone with pelagic foraminifers and nannofossils (4), through finegrained, well-sorted packstone, with lenses including benthic foraminifers, shallow-water algal and echinoid debris (5^8), and back to wackestone (9 and 10).

SW 3200

NE 3400

3600

3800

4000

4200

4400

4600

4800

5000

5200

5400

5600

5800

6000

6200

Two way time (ms)

0

0 100

100

200 200 300 300

400 500

400

Approx. depth (m)

(b)

600 500

3200

3600

3800

4000

4200

4400

4600

4800

5000

5200

5400

5600

5800

6000

9

100

100 7

200 300 400 500

6200 0

8

Two way time (ms)

3400

0

200 4

6

300 400

5 3

500

2

600

1

Approx. depth (m)

(a)

700

700 5 km

STZ

Figure 17.3 Section K90 -4b crossing northeastern parts of the Danish Basin and parts of the inverted Sorgenfrei ^Tornquist zone in the Kattegat area. Note the alternation between relatively sheet-like units (2, 4 and 6) and units showing marked wedging and clinoform bedding towards NE (1, 3 and 5) considered diagnostic of contourite drift formation. STZ = Sorgenfrei ^Tornquist Zone. Seismic units 1^6: mainly Maastrichtian (and uppermost Campanian?); 7: Danian; 8: Selandian; 9: Pleistocene and Holocene. (b) Seismic section. (a) Its seismic stratigraphic interpretation (after Surlyk and Lykke-Andersen, 2007, with permission from Blackwell Publishing).

Mean grain size 0. 0 0. 02 0 0. 04 0 0. 08 0 0. 16 0 0. 31 0 0. 62 1 0. 25 2 0. 50 50 0

mm

cm

Bioclastic wackestone

ty de c r e Veloci

Mottled calcilutites and calcisiltites 15

Laminated calcisiltites

Omission:

Depositional fabric

ase

Facies types

Bioclastic wackestone and Intraclastic calcisiltite

Particle abrasion, erosion and phosphatization

10 Styliolinid grainstone

5

Veloc ity

increase

Calcarenites

Styliolinid packstone

Styliolinid wackestone

Styliolinid packstone

Mottled calcisiltites and calcilutites

Bioclastic wackestone

0

Figure 17.6 Composite facies model for Devonian bottom-current deposits showing the complete vertical superposition of a coarsening-upward and a fining-upward micro-sequence (see also Hu«neke, 2007b). These idealized micro-sequences are between 2 and 15 cm thick and include in many cases an omission surface in between. The variation in mean grain size, composition and depositional texture is interpreted as recording an increase and decrease in bottom-current velocity, respectively, for the coarsening-upward and fining-upward micro-sequence. Oblique hatching (light blue) indicates calcarenites that are cemented by coarse sparry calcite (packstones ^ grainstones).

Figure 17.7 Palaeogeographic reconstruction of Laurussia and northern Gondwana during the Givetian ^ Frasnian (370 Ma), showing expected pattern of oceanic surface circulation as well as distribution of deep-sea, shelf-sea, land and mountain ranges (from Hu« neke, 2006, with permission from the Geological Society, London).The map is based on reconstructions by Kiessling (in Copper, 2002) and Golonka (2002). Pelagic successions of Europe and northern Africa that include hiatuses at the Givetian/Frasnian boundary are sedimentary records of Gondwana (MC = Moroccan Central Massif; AA = Anti-Atlas; AB = Ahnet Basin), the Armorican terrane assemblage (MN = Montagne Noire; FW = Frankenwald), the Noric terrane (CA = Carnic Alps) and Laurussia (RS = Rhenish Slate Mountains; HM = Harz Mountains).

Figure 18.3 Argentine Basin. (a) Depositional and erosional contourite features on the Argentine Basin generated by the action of the WSDW (adapted from Flood and Shor, 1988; Speer et al., 1992; Reid, 1989; Klaus and Ledbetter, 1988). (b, c) Profiles across the Zapiola Drift and the Malvinas/Falkland Escarpment (data courtesy of the Argentine Navy Hydrographic Service, Buenos Aires, Argentina). (d) and (e) Sheeted deposits of the Ewing Drift (data courtesy of Argentine Navy Hydrographic Service, Buenos Aires, Argentina). (f ) Seismic profile perpendicular to the wave strike direction in the central part of the sediment-wave field on the western Zapiola Drift (adapted fromVon Lom-Keil et al., 2002; with permission from Elsevier). (g) Profile across the Zapiola Drift crest. The crest appears to migrate to the southwest (adapted fromVon Lom-Keil et al., 2002; with permission from Elsevier).

Figure 18.6 Major contourite deposits in the North Atlantic. (a) Distribution of deposits showing the deep-water circulation (modified from McCave and Tucholke, 1986; Fauge' res et al., 1993; with permission from Elsevier). (b) High-resolution seismic profile of the abyssal contourite sheet of the Irminger Basin Gloria drift linked to bottom currents trapped on a basin sea floor (from Fauge' res et al., 1999 after Egloff and Johnson, 1975; with permission from Elsevier).

Figure 18.7 Greater Antilles Outer Ridge (GAOR). (a) 3-D map constructed from bathymetric data from Sandwell and Smith (1997) with general location of the drift adjacent to the Puerto Rico Trench. (b) Sketch with the specific location of the Greater Antilles Outer Ridge (GAOR). (c) Seismic-reflection profile from the base of the Bahama Banks across the southern Silver Abyssal Plain and onto the GAOR. (d) Seismic-reflection profile from the base of the Bahama Banks across the Caicos Outer Ridge, Silver Abyssal Plain, and northwestern Greater Antilles Outer Ridge to Vema Gap (adapted and modified from Tucholke, 2002; with permission from the Geological Society, London): location in Figure 18.7b. (e) Model for current-controlled deposition along a section extending northeast from the Bahama Banks to the southernmost Bermuda rise (from Tucholke, 2002; with permission from the Geological Society, London).

Figure 18.8 Ivory Coast Rise and Guinea Drifts. (a) Bathymetry of the Sierra Leone Basin and surroundings. The Guinea Drift along the southern margin of the Gambia Basin is shaded grey. Locations of main canyons on the continental slope and deep-sea channels on the continental rise are shown. Mau = Mauritania Channel, Kay = Kayar Channel. (Legend: GP = Guinea Plateau. Inset map: CV = Cape Verde Islands; GaB = Gambia Basin; K = Kane Gap; SR= Sierra Leone Rise; SB= Sierra Leone Basin; ICR= Ivory Coast Rise; GuB = Guinea Basin; Mar = MidAtlantic Ridge; V = Vema Fracture Zone; R= Romanche Fracture Zone; SA = South America). (b, c) Seismic-reflection profiles across the Ivory Coast Rise. (d, e) Reflection profiles over the Guinea Drift along the southern margin of the Gambia Basin (modified from Jones and Okada, 2006; with permission from Elsevier).

Figure 18.12 Morphological expression of contourite drifts and channels in the Central Scotia Sea. (a) Central sector of the Scotia Sea, based on the interpretation of the swath bathymetry map, complemented with the analysis of seismic profiles (general location in Figure 18.2a). Main depositional and morphological features of the area are shown. Legend: North, Middle, and Southern Antarctic channels (NAC, MAC, and SAC) are the main channels associated with the Antarctic Circumpolar Current (ACC). A1 and A3 are elongated-mounded drifts related to canalised branches of the ACC; A2 is a field of large sediment waves. The Weddell Channel (WC) is the main channel of the Weddell Sea Deep Water (WSDW). The Western, Eastern, and NorthernWeddell Channels (WWC, EWC, and NWC) are the main channels of the WSDW. W1 is a slope-plastered drift attached to the discovery bank. W2 is made up of elongated-mounded drift and sediment waves related to the WSDW. W3 is a contourite fan formed by two channels. W4 is composed of mounded-sheeted drift and sediment waves related to the WSDW. Bottom-current directions are also shown based on the drift depositional patterns, channel orientation, and direction of migration of sediment waves. WAB is the boundary between the ACC and the WSDW flows (modified from Maldonado et al., 2003; with permission from Elsevier). (b, c) 3-D imagery from multi-beam bathymetry showing erosional and depositional features due to the circulation of both the WSDWand the ACC. (d) 3-D imagery from multi-beam bathymetry showing a mounded-sheeted drift bounded by sub-circular depression structure due the large eddies generated by the lateral interaction between theWSDWand the ACC (for location of the area see Figure18.2a).

Figure 18.14 3-D conceptual sketch showing the main depositional and erosional features due to deep-water circulation on abyssal plains.

Figure 19.2 Main depositional (drifts) features and inferred bottom-current paths (adapted from Rebesco, 2005; Rebesco and Stow, 2001, with permission from Elsevier).

Figure 19.6 Depositional and erosional contourite features on the middle slope of the Gulf of Cadiz. (a) Location sketch with main water-mass circulation along the margin (modified from Herna¤ ndez-Molina et al., 2003, 2006c; with permission from the Geological Society of America and Elsevier). (b) Uninterpreted multi-channel seismic (MCS) reflection profile across the Faro ^Albufeira elongated mounded and separated drift and the Alvarez Cabral moat on the middle slope (data courtesy of TGS-NOPEC Geophysical Company ASA). (c) Sparker seismic profiles indicating the abraded surface. (d) Sparker seismic profiles showing examples of contourite channels and marginal valleys. (e) Sketch of the MPC-1 borehole from the contourite depositional system on the middle slope of the Gulf of Cadiz (proximal scour and ribbons sector of Herna¤ ndez-Molina et al., 2003, 2006c) with an uninterpreted MCS reflection profile across the middle slope (line S- 81A, provided by REPSOL-YPF Oil Company for the present work). For location of the profiles, see (a).

Figure 19.6

(Continued)

Figure 19.9 Depositional and erosional contourite features on the eastern margin of the South Island in the Canterbury Basin (South Island of New Zealand). (a) Location sketch with main water-mass circulation along the margin. Locations of the modern Southland Front, Sub-Antarctic Front (SAF), and a local gyre associated with SAF are shown (modified from Lu et al., 2003). (b) Dip profile (EW00 -01-12) showing the main features of a large mounded and elongated drift developed on the lower to middle slope during the Late Miocene-Pliocene (Craig S. Fulthorpe University of Texas, USA, Personal Communication, 2007). (c) Dip profile (EW00 - 01-16) showing sediment waves (shaded area) preserved basinward at the base of elongate drifts. Wavelength is about 2 km and amplitude averages 50 m (modified from Lu et al., 2003; with permission from Elsevier).

Figure 19.12 Main characteristics of large-scale contourite erosional features on slopes. A classification into six different types of erosional features on slopes is proposed: erosional terraces, abraded surfaces, contourite channels, contourite moat and furrows (adapted from Davies and Laughton, 1972; Garcı´a, 2002; Herna¤ ndez-Molina et al., 2003, 2006c; with permission from Elsevier).

–7°30′W

–7°W

–6°30′W

(a)

36°N

8°W

7°W

7°30′W

Cadiz contourite channel 36°N

36°20′N

7°W

6°50′W

(b)

6°40′W

(c)

36°10′N

36°N

Figure 19.14 Backscatter facies. (a) Regional original Seamap side-scan data from Sectors 1 and 2 of the contourite depositional system of the Gulf of Ca¤diz (adapted from Herna¤ ndezMolina et al., 2006c; with permission from Elsevier). (b): High-backscatter areas related to the Ca¤diz Contourite Channel obtained by Simrad EM-12S-120 multi-beam echo sounder (R. Leo¤n and M.C. Ferna¤ ndez-Puga, Spanish Geological Survey, IGME, Personal Communication, 2006). (c): Sea-floor acoustic response of the abraded surface revealed by the reprocessing of the Seamap side-scan data ( Joan Gardner, Naval Research Laboratory, USA, Personal Communication, 2006).

Figure 19.17 3-D conceptual sketch showing the potential contourite depositional and erosional features on a slope over: (a) margins with different water masses along the slope but with simple current pathways (e.g. the Brazilian slope and the Northern European Margin), (b) margins where recent tectonic activity has produced very complex slope morphology, increasing the possibility of generating multiplecurrent pathways (e.g. the slopes of the Gulf of Cadiz, western Iberian Margin and some active margins) and (c) interaction between downslope (submarine canyons and slides) and along-slope processes (slope currents).

(a)

(b)

Figure 20.9 Contourite drifts of the Gela Basin (modified from Verdicchio and Trincardi, 2008b; with permission from Springer). See location of the Gela Basin in Figure 20.8. (a) High-resolution bathymetric map of a small part of the eastern Gela Basin, showing a marked contour-parallel moat along the shelf edge; the track lines of CHIRP profiles are also shown. (b) CHIRP sonar profiles showing upper-slope contourite drifts basinward of the erosional moat adjacent to the shelf edge. Contourite drifts display occasionally multiple crests and erosional down-slope flanks. Stacked thin-skinned mass-wasting deposits occur on the basinward flanks of the sediment drifts, suggesting that these deposits are prone to failure during, or soon after, their deposition.

twtt (ms)

SW

NE

Pelagosa Sill

225

Moat

ES 1

250

50 km

4300′

275

2 km

Adriatic Sea 4200′

Italy 1500′

1600′

1700′

Figure 20.10 CHIRP sonar profile across the Pelagosa Sill, showing an elongate moat parallel to the base of the sill flank in water of approximately 170 m deep. The location map, including the track line of the profile, is shown in the inset.This deposit was actively growing during the last glacial cycle (since approximately 140 ka; A. Asioli, 2005, IGG-CNR, personal communication) and particularly during the last glacial lowstand (above the line marked as ES1).

S

N

Sp038 150 m 5000 m

S

N

Sp036

Plio-Pleistocene unconformity

PC ACM

MOW circulation

8

03

6

03

Sp

N

FAD

Sp

Plio-Pleistocene unconformity

LD

Figure 21.1 Multi-beam (EM 300) data showing the northern part of the Gulf of Cadiz and the Mediterranean Outflow Water (MOW) capture by the Portimao Canyon (PC) (from Marches, 2005; with permission). East of the canyon, a thick detached drift forms, the Faro-Albufeira Drift (FAD), separated from the margin by the Alvarez-Cabral Moat Channel (ACM). The sharp decrease of the MOW intensity and competency west of the canyon generates the formation of the thin plastered Lagos Drift (LD). Sparker profiles Sp036 and Sp038 illustrate the change in drift morphology due to capture of the MOW by the Portimao Canyon.

(a)

9°W

8°W

7°W

5°W

Huelva

PORTUGAL

N

6°W

Faro

SPAIN

37° 0 50

Cadix MU

36°

ML 100 0

2000

3000

ATLANTIC OCEAN

Gulf of Cadix

NADW North-Atlantic Deep Water NASW North-Atlantic Surficial Water 7°50

(b)

Atlantic inflow water Tanger

Morocco

Mediterranean outflow water (MOW) MU: Mediterranean Upper Water ML: Mediterranean Lower Water 7°00

36°00

35°50

100

10 km N

(c)

7°35′W

1000 2000 m

7°25′W

Channel

35° 45′

Lobes

35° 45′

Figure 21.6 Small channels and lobes in the Gulf of Cadiz (after Hanquiez, 2006; with permission). See top map (a) for location. Swath bathymetry map and EM 300 reflectivity map (b). These small turbidite systems initiated by failures at the front of contourite deposits are nested within large MOWchannels. The permanent MOW flow winnows the fine particles in lobe deposits.

150°E

135°E

64°S

Seismic profile

Dibble Ice Tongue 68°S

(a)

(b)

W

Wilkes Land

Mound

Mound

Channel

Channel

Jussieu

Levee

W Channel

Adelie Coast

Mertz Glacier Ninnis Glacier

E

Jussieu Channel

Levee

Mound

(c)

Mound

Buffon Channel

Buffon E Channel

(d)

W

Crest

E

W

E

Figure 21.10 Continental rise of Wilkes Land margin (after Escutia et al., 2002; with permission). See top map for location. (a) Seismic profile showing channels, levee and sediment mounds. (b) Detailed section of (a), showing two channels ( Jussieu and Buffon), the east levee of the Jussieu Channel and a sediment mound. (c, d) Details of the sediment mound deposit and its interpretation. Ch = channels; L = levees; SW = sediment waves. WL1b, WL1c and WL2 are unconformities.

Figure 22.1 Bathymetric setting of the NW European Atlantic Margin showing the main tectonic elements, oceanographic circulation pattern, and summary of large-scale sedimentary deposits and processes that have contributed to the shaping of the ocean margin during the Neogene (modified from Stoker et al., 2005b).The Iceland ^ Faroe Ridge forms the southeastern section of the Greenland ^Scotland Ridge. Abbreviations: FSC = Faroe ^Shetland-Channel; IFR = Iceland ^ Faroe Ridge; WTR =Wyville-Thomson Ridge; PB = Porcupine Basin; NC = Norwegian Channel.

22 .5 Fi gu re

Fi

gu

re

22

.4

a

Figure 22.4b

0

Glacially-fed fan Giant elongated sediment drift

300 km

Maximum limit of Pleistocene glaciation Palaeo-ice stream

Basinal drift complex

BGS borehole

Prolonged sea-bed erosion

DSDP/ODP site

Sediment waves Predominant flow paths of modern intermediate- and deep-water masses

Piston core BGS Airgun seismic reflection profile

Figure 22.2 Generalised Pleistocene glacially-influenced setting of the Rockall Trough. The map also shows locations of boreholes, cores, and profiles referred to in the text and other figures. Inset shows location of DSDP Site 610, which occurs on the Feni Ridge, SW of the main study area. Abbreviations: FB = Faroe Bank; BBB = Bill Bailey’s Bank; LB = Lousy Bank; GBB = George Bligh Bank; RB = Rosemary Bank Seamount; AD =Anton Dohrn Seamount; HT = Hebrides Terrace Seamount; HRB = Hatton ^ Rockall Basin; HB = Hatton Bank.

Hebridean margin Hebrides slope

Shelf edge

Key cores 88/7, MD95 7A 2006

Mid shelf

MIS 1 2 3

AGE (ka) 0

NOTE: Section broadly correlates with this interval, but absence of interglacial strata (MIS9) precludes assignment to specific glacial stage (MIS10, 8, or both)

Mid

50

Devensian Early 100

5e 6

Holocene

Late

4 5

British chronostratigraphy

Ipswichian 150

200

7

“Wolstonian” 250

8

NOTE: Ambiguity remains on the use of this term; the MIS6–10 record remains poorly defined in Britain

300

?

9 10 11

Glacial unconformity

Glacial debris-flow deposits

350

400

Hoxnian

12

450

Anglian

13

500

Cromerian

Only Barra-Donegal Fan

Figure 22.3 Glaciation curve for Northwest Britain, for the last 0.5 Ma, based on information derived from Stoker et al. (1994) and Knutz et al. (2002a, b). British chronostratigraphic stages are based on Bowen (1999), and the ages of marine isotope stage boundaries are based on Martinsen et al. (1987) and Williams et al. (1988).

Submarine end-moraines Glacial unconformity

SE BP

0.4

SBM SBM

Sula Sgeir Fan

0.8 Clinoform Buried elongated drift/ sediment waves 5 km

Mass-flow package

Glacial unconformity?

BGS83/04-6

1.6

C10

NW

SE Hebrides shelf

Rockall Trough 1

Sula Sgeir Fan

Figure 22.4

Miocene–Lower Pliocene

2

Glacial unconformity

Early Pliocene unconformity

Lower Pliocene– Holocene

Oligocene

Eocene & older 10 km

3

BASED ON DGH9508 Hebrides shelf

(b)

Two-way travel time (s)

Miocene– lower Pliocene

3

Early Pliocene unconformity

1

Peach slide

Rockall Trough W

E

2

Barra fan Lower Pliocene– Holocene

3 Late Eocene unconformity

4

4

10 km

Two-way travel time (s)

Upper Eocene– lower Miocene

Early/mid-Miocene unconformity

Top Palaeogene volcanics

Sea bed

BGS 92/01-56

Distal edge of lowest debris-flow package

3.0

Onlap

3.1

Sea bed

3.5 3.6 10 km

Two-way travel time (s)

0

1.2

Two-way travel time (s)

Two-way travel time (s)

Two-way travel time (s)

NW

Two-way travel time (s)

(a)

10 km

Figure 22.4 BGS airgun profiles across the (a) Sula Sgeir and (b) Barra fans, showing the overlapping/interdigitating relationship between the glacially-fed shelf-margin fans and the basinal contourites. Profiles located in Figure 22.2. Abbreviations: SBM = sea-bed multiple; BP = bubble pulse.

(a)

SE Sediment waves showing upslope migration of crests

NW

MD95-2006

30 ms 1 km

A1 DF 3 A2 B

0

10.89 ka

its un ic m is Se

D

ep

(b)

th ( Li m) t in hof te ac rp ie re s ta C tio -1 n 4 ka BP M IS

C

1

5

A1

Key to lithology Silty-muddy contourite

14.86 ka

Hemipelagite/muddy contourite

10

2 17.66 ka

A2

Glacimarine mud and clay Thin-bedded turbidites

15

B 20

24.31 ka 25.81 ka 29.00 ka 29.33 ka

25

33.48 ka

3 C

30

Figure 22.6 Process interaction and sediment flux on the Barra Fan. (a) BGS deep-tow boomer profile and interpreted line drawing from the lower part of the fan. (b) Simplified log of piston core MD95-2006 (a and b modified from Kroon et al., 2000; and Knutz et al., 2002a, b). Core located in Figure 22.2. Abbreviations: ms = milliseconds; MIS = marine isotope stage; DF = debris flow.

Figure 22.8 Upper panel shows 3-D visualisation of the PL 20 reflector, showing a surface marked by sediment waves with crests oriented oblique to the slope. Note the decrease of wave dimensions up- and down-slope. Picture looks to the NE, along the West Shetland Slope. Lower panel shows a seismic profile illustrating cross-sectional geometries of the sediment waves, revealing a NE migration pattern forced by the SW flow direction of the depositional current (from Hohbein and Cartwright, 2006, with permission from Elsevier).

Figure 22.11 Distribution of echo characters and sedimentary regimes in relation to depth and forcing mechanisms (upper panel), and distribution and effects of DWBC on sedimentation/ erosion on the Eirik Drift (lower panel). Pie diagrams show the distribution of clay minerals over the Eirik Drift (modified from Hunter et al., 2007a). Inset shows location of the Eirik Drift.

Figure 22.13 Cyclic sedimentation in Core ODP-1165-1165B-10H from the Wild Drift (modified from Gru«tzner et al., 2003). From left to right: photograph and green/grey colour ratio (dotted line = 5 cm record, solid line = 25 cm average); magnetic susceptibility, green/ grey colour ratio, and Fe intensity measured for cores (triangles: samples with >10% sandsized particles, i.e. >250 mm; arrows: susceptibility spikes indicating ice-rafted debris layers); magnetostratigraphic age control; astronomically calibrated depth-to-age conversion based on correlation of orbital obliquity and filtered colour (green/grey ratio) record (unfiltered data set is shown in background).

Figure 22.15 Location map of current-moorings on Drift 7, Antarctic Peninsula Pacific Margin (modified from Rebesco et al., 2002).The mean direction of the current measured 8 m above the sea floor is indicated with arrows at the location of the three moorings. The inferred bottom-water flow path along the isobath between the moorings is shown by a dashed line.

Figure 22.16 Multibeam bathymetry data from the distal part of Drift 7 shown in perspective view and cut by multi-channel seismic-reflection profile IT95-130 (modified from Rebesco et al., 2007). The data are part of a large data set acquired by R/V OGS-Explora between 1995 and 2004 by projects MAGICO (Multibeam Antarctic Glacial system Integral Coverage) and SEDANO (Sedimentary Drifts of the Antarctic Offshore) funded by the Progetto Nazionale di Ricerchein Antartide (PNRA). Growth of the drift reflects variations in both the Antarctic ice sheet and Southern Ocean along-slope bottom currents.

Figure 22.17 Dip and strike multi-channel seismic profiles crossing Drift 7 with location of gravity cores and selected radiograph facies of last glacial sediments (from Lucchi and Rebesco, 2007; with permission from The Geological Society, London). (a) Multi-channel seismic-reflection profile IT92-109 crossing Drift 7 in a direction parallel to the margin, showing the location of gravity cores and radiography facies of most significant sites. The radiograph intervals from each core represent the same chronostratigraphic interval (last glacial sediments). Pervasive lamination is produced by faintly laminated mud alternating with closely spaced silty laminae, lenses (millimetre-thick), and layers (up to 4 cm thick).These deposits decrease in thickness and frequency towards the crest of the drift where silty laminae are detectable only in radiographs. (b) Multi-channel seismic-reflection profile I95-135 crossing Drift 7 perpendicular to the margin and showing the location of gravity cores and radiography facies of most significant sites. The radiograph intervals reported from each core represent the same chronostratigraphic interval (last glacial sediments). Along the dip transect, millimetre-thick silty laminae are common. At the top of Drift 7, the last glacial sediments consist of homogeneous layers (2^10 cm thick), grey intervals separated by sharp and irregular boundaries.The X-radiographs also indicate the presence of 1^2 cm thick intervals of scattered fine-grained sand and silt (IRD).

Figure 22.18 Cartoon-style model showing the successive climatic stages in the sediment drifts of the Antarctic Peninsula Pacific Margin (from Lucchi et al., 2002, with permission from Elsevier). The model explains the repetitive succession of the four distinctive sedimentary facies, each characterised by a typical clay mineral assemblage and bioclastic content.

Figure 23.1 Image of the Mid-Cenozoic sea floor of the Santos Basin, SE Brazil, from 3-D seismic data, showing a mass-flow deposit (mf ) originated by upper-slope destabilization related to sea-floor erosion by the action of a vigorous surface slope-boundary current (sc). Sea-floor erosion induced the outcropping of an aquifer provoking a water sapping front (wsf ) and destabilization of the slope face, transferring sediment down-slope (slope apron), locally reworked by a deeper mid-slope contour current (dc) (modified from Viana et al., 2007, with permission fromThe Geological Society of London).

Figure 23.2 Block diagram representing the modern Campos Basin sea-floor physiography. Increase, decrease and trends of bottom-current modifications related to changes in margin physiography and sea-floor topography are indicated by arrows. Size of arrows indicates relative strength of bottom currents. 1= margin protuberance induces upstream increase of current velocity, downstream perturbation and speed decrease; 2 = sea-floor obstacles imply current deviation, locally acceleration; 3 = sea-floor passageways and corridors induce flow intensification.

Figure 23.4 Seismic amplitude horizon slice of the Mid-Cenozoic sand-rich shelf edge/upper slope of the Santos Basin. This setting is influenced by the action of a surface slope-boundary current (sc), which erodes the uppermost slope and helps to transfer sand down-slope. Barchan dunes and barchanoid ridges are developed above a mid-slope terrace (inset a), swept by midslope contour currents (dc). These bedforms evolve downstream into a slope-parallel contourcurrent-carved channel (inset b). Arrows in the insets indicate the sedimentary features. Schematic block diagram at the top left illustrates these processes. (Modified fromViana et al., 2007, with permission fromThe Geological Society of London).

Figure 23.5 Side-scan mosaic (SYS3000) of the lower slope setting in the Campos Basin sector, SE Brazil, showing the development of barchan dunes (d), barchanoid and linear ridges and furrows (f ) related to the vigorous flow of bottom currents. Mosaic processed by R.C. Rechden.

Figure 23.6 Perspective view of the modern Campos Basin sea bottom based on 3-D seismics. Topographic highs are associated to salt halokinesis. Colours are related to seismic amplitude, corresponding to grain-size differences ground-truthed by piston cores. Blue colours (dark on the topographic highs) correspond to fine-grained sediments, and yellow and red warm colours in the basin (yellow and red) to sand. In particular, the red (dark) colour in the basin corresponds to sand sheets and ribbons.

Sediment overflow (so)

Sediment avulsion (a)

Main sediment conduit

Figure 23.7 Schematic diagram representing upper margin sediment delivery through fairways developed along salt crestal grabens (scg). Sediment overflow (so) and avulsion (a) feeding the axis of salt withdrawal mini-basins are also represented. Sediments are reworked by bottom currents (SOC = Southern Ocean Current of Viana et al., 2002b) and redistributed along current-generated furrows (F). Diagram based on 3-D seismic data presented by Viana et al., 2007.

Complex of avulsion lobes

Base of slope

N

400.00

7 km

600.00 800.00

1000.0

400.00 600.00 800.00 Contour current reworked deposits detached from Complex of avulsion lobes

1000.0

Figure 23.8 Seismic-amplitude image of a subcropping turbidite/contourite complex.Warm colours (yellow and red) correspond to coarse-grained sediments, and green to fine-grained sediments.Two systems are observed: one constituted by low-sinuosity channels flowing along a down-slope oriented fairway crossing diagonally the image from NW to SE (the blue arrow indicates the North). Laterally to this system, to the South, occurs a fan-shaped body corresponding to a complex of avulsion lobes, distally reworked by contour currents developing a fringe of irregular patch-like small sedimentary bodies.

Figure 23.9 Physical properties of a contourite deposit on the upper Miocene middle slope of the Campos Basin (from Viana et al., 2007; with permission from The Geological Society of London). Data indicate a down-slope decrease in sand content and the intercalation of sand and mud deposition phases in the system. (a) Seismic-amplitude map of the basal erosional surface (bes). (b) Dip seismic line crossing three wells. (c) Close-up of the seismic line at the distal site with superposed well logs. (d) Well logs of the three wells along the dip line across a contourite deposit on the upper Miocene middle slope of Campos Basin.

NE

SW 0 25 50 75 100 m

0 Turbidite channels

500

1000 m

Bottom current deposits

Third order sequencle boundaries

Figure 23.10 Depositional model for Palaeocene ^ Eocene deep-water, trough-filling sandstones of the Campos Basin, elaborated by Moraes et al. (2007; with permission from The Geological Society of London). The model indicates the distribution of bottom-current deposits in a turbidite channel complex, showing typical facies proportions and dimensions inferred for the sandstone bodies. Vertical connection of sand bodies is proposed to be diminished by the presence of contourite bodies. Dimensions of turbiditic channels and contourites are based on well observations and analogue data.

(a) Reworking gravity flow deposits

st vfs ms

Time Persistency of the current regime and sand availability

c.sand

f.sand

silt

c.sand

silt

(b)

f.sand

Construction of contouritic deposit

D3 - Medium intensity bottom currents D1: Gravity flow deposit,

Medium to long duration

normal grading

D3: Φ good < top (better than D1 e D2), K good

Φ < top, K regular

Good – very good reservoir

Regular – good reservoir

D4 - Strong bottom currents D2 - Weak bottom currents, short duration

Long duration

D2 ~ D1: Φ < top, K regular

D4: Φ = top and base, K good to excellent

Regular – good reservoir

Excellent reservoir

Figure 23.11 Sedimentary characteristics of coarse-grained deposits affected by bottom currents. (a) Model of the construction of coarse-grained contourite bodies as a function of longevity of a bottom-current regime and persistence of intensity required to erode and transport the sand available in the system, originated from previous sand-rich gravity flows. (b) Model of general petrophysics characteristics of deep-water, coarse-grained deposits varying from those derived from pure gravity-flow deposits (facies D1) to those totally reworked by bottom currents (D4). Facies D2 and D3 correspond to continuous increase in the bottom-current strength and persistence.

(a)

NE

Flo

wd ire ctio n

Faeroe–Shetland Basin

Shetland Slope

km 0

SW

5

10

NE

SW

g

b

(b)

twtt (ms)

2000

2500

Figure 24.2 Reconstruction of palaeocurrent patterns across a Plio-Pleistocene contourite drift using 3-D seismic data. Example from the West Shetland Drift (Knutz and Cartwright, 2003). (a) Isochrone (two-way travel-time difference) map between internal Pliocene reflector and basal unconformity, marked by green (g) and blue (b) arrow, respectively, in the seismic profile (b). Dark blue colours correspond to sediment thicknesses over 150 m, whereas light blue colours are indicative of non-deposition. (b) Seismic profile (thin line in A) showing along-slope progradation of internal reflectors in a southwest direction and downlap/onlap onto an erosional unconformity of Late Miocene to Early Pliocene age. The build-up of this contourite drift was promoted by southward flow of Nordic Sea deep water in proximity of a major sediment source extending from the North Sea Basin, and the accentuated sea-floor topography generated by mass wasting. Y-axis represents two-way travel-time (twt) shown in milliseconds (ms).

Figure 24.4 World map showing important oceanic gateways that opened/closed during the Cenozoic era. 1. Fram Strait, Miocene opening. 2. Bering Strait, Late Cenozoic shallowing (?). 3. Davis Strait. 4. Greenland ^Scotland Ridge, formed in multiple phases since Early Oligocene. 5. Gibraltar Strait, narrowed during the Late Miocene (Messinian salinity crisis). 6. Panama Isthmus, final closing 3 Ma ago. 7. Indonesian Seaway, narrowed during Late Cenozoic. 8. Agulhas Passage. 9. Tasman Gap, Early Oligocene opening. 10. Southern Ocean ^ South Pacific gateway. 11. Drake Passage, Late Mid-Eocene ^ Early Oligocene opening.

Precipation Reflected insolation Ablation Evaporation

Meltwater ro io-p

wn

y

tivit

duc

Do

B

slo

pe

tra

ns

po

rt

Vertical fluxes

ry da un Bo rents cur

Figure 24.6 Conceptual model illustrating the principal (palaeo)environmental factors that contribute to accumulation of high-resolution contourite archives on a glaciated continental margin. The notable differences from contourites developed in a non-glacial setting are the supplies of glacimarine sediments from down-slope resuspension, meltwater plumes and in the form of ice-rafted debris supplied from tidewater glaciers.

Figure 25.2 Part of a seismic profile from the Storegga Slide area offshore Norway showing the palaeo-slide S, an infilling contourite drift and palaeo-slide R. The glide plane of palaeoslide S is indicated by the green horizon. Note that the glide plane of palaeo-slide R (pink horizon labelled P Slide 2 Base) follows the top of the drift deposits and is parallel with the internal reflections of the drift. CD marks contouritic drift deposits and SD denotes slide debris, TNU = top Naust unit U reflection, INS2 = intra Naust unit S reflection 2 and TNR AS = top Naust unit R reflection. See Solheim et al. (2005) for further discussion.

km Al ex

100

de

an

0 rC

Depth scale (m) 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 4200

nn

ha

Drift 6

el

67°S

Fluid expulsion structures

Drift 7 ODP Site 1096 Coalescent slide scars Single slide scars

Ch

arc

ot C han 68°S ne

l

e

dg

pe

l

Drift 8

ta

slo

e

n ne

l ta

i

nt

78°W

77°W

76°W

e lf

sh

en

in

Co

C

t on

75°W

74°W

Figure 25.7 Shaded relief map of the bathymetry of Drift 7 (modified after Rebesco et al., 2007). The dotted line outlines the boundary of the Late Pliocene Alexander mega debris flow buried below the Alexander Channel. For location, see Figure 25.6.

64°S

100 km

Continental rise Slab detachments

65°S Biscoe Trough

Bis

coe

Is.

Continental shelf

66°S 70°W

68°W

66°W

Depth scale (m) –5000 –4000 –3000 –2000 –1000

0

Figure 25.8 Shaded relief map of the bathymetry of Drift 4A and adjacent Biscoe Trough on the continental shelf (modified after Amblas et al., 2006). For location, see Figure 25.6.