Late Palaeozoic and Mesozoic ecosystems in SE Asia

Non-marine Late Palaeozoic and Mesozoic formations are widespread in mainland SE Asia. Although the first reports on fossils from some of these formations were published as early as the 1890s, it is only since 1980 that floras and faunas from the Permian, Triassic, Jurassic and Cretaceous of SE Asia have received the attention they deserve. Fieldwork in various parts of Thailand and Laos has revealed a succession of fossil assemblages that now allows a reconstruction of the evolution of continental ecosystems in that part of the world during the Late Palaeozoic and the Mesozoic. The first papers in this book present the geological background of these floral and faunal successions, as well as historical aspects of their discovery. Descriptions of new taxa and review papers deal with plants, sharks, bony fishes, turtles, crocodilians, dinosaurs and mammal-like reptiles. Papers about the Mesozoic palaeobiogeography, environments and climates of Asia conclude the volume.

Late Palaeozoic and Mesozoic Ecosystems in SE Asia

The Geological Society of London Books Editorial Committee Chief Editor

BOB PANKHURST (UK) Society Books Editors

JOHN GREGORY (UK) JIM GRIFFITHS (UK) JOHN HOWE (UK) PHIL LEAT (UK) NICK ROBINS (UK) JONATHAN TURNER (UK) Society Books Advisors

MIKE BROWN (USA) ERIC BUFFETAUT (FRANCE ) JONATHAN CRAIG (ITALY ) RETO GIERE´ (GERMANY ) TOM MC CANN (GERMANY ) DOUG STEAD (CANADA ) RANDELL STEPHENSON (UK)

Geological Society books refereeing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society’s Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society Book Editors ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees’ forms and comments must be available to the Society’s Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. More information about submitting a proposal and producing a book for the Society can be found on its web site: www.geolsoc.org.uk.

It is recommended that reference to all or part of this book should be made in one of the following ways: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) 2009. Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315. PHILIPPE , M., DAVIERO -GOMEZ , V. & SUTEETHORN , V. 2009. Silhouette and palaeoecology of Mesozoic trees in Thailand. In: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 83 –94.

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 315

Late Palaeozoic and Mesozoic Ecosystems in SE Asia

EDITED BY

E. BUFFETAUT CNRS, France

G. CUNY The Natural History Museum of Denmark, Denmark

J. LE LOEUFF Muse´e des Dinosaures, France and

V. SUTEETHORN Department of Mineral Resources, Thailand

2009 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 9000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc.org.uk ) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ 44 (0)20 7434 9944; Fax þ 44 (0)20 7439 8975; E-mail: [email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to [email protected] Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ 44 (0)1225 445046, Fax þ 44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. # The Geological Society of London 2009. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/09/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-275-5 Typeset by Techset Composition Ltd., Salisbury, UK Printed by MPG Books Ltd., Bodmin, UK Distributors North America For trade and institutional orders: The Geological Society, c/o AIDC, 82 Winter Sport Lane, Williston, VT 05495, USA Orders: Tel. þ 1 800-972-9892 Fax þ 1 802-864-7626 E-mail [email protected] For individual and corporate orders: AAPG Bookstore, PO Box 979, Tulsa, OK 74101-0979, USA Orders: Tel. þ 1 918-584-2555 Fax þ 1 918-560-2652 E-mail [email protected] Website http://bookstore.aapg.org India Affiliated East–West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel. þ 91 11 2327-9113/2326-4180 Fax þ 91 11 2326-0538 E-mail [email protected]

Contents BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. Late Palaeozoic and Mesozoic continental ecosystems of SE Asia: an introduction

1

METCALFE , I. Late Palaeozoic and Mesozoic tectonic and palaeogeographical evolution of SE Asia

7

STEYER , J. S. The geological and palaeontological exploration of Laos; following in the footsteps of J. B. H. Counillon and A. Pavie

25

BATTAIL , B. Late Permian dicynodont fauna from Laos

33

RACEY , A. Mesozoic red bed sequences from SE Asia and the significance of the Khorat Group of NE Thailand

41

RACEY , A. & GOODALL , J. G. S. Palynology and stratigraphy of the Mesozoic Khorat Group red bed sequences from Thailand

69

PHILIPPE , M., DAVIERO -GOMEZ , V. & SUTEETHORN , V. Silhouette and palaeoecology of Mesozoic trees in Thailand

85

CUNY , G., SRISUK , P., KHAMHA , S., SUTEETHORN , V. & TONG , H. A new elasmobranch fauna from the Middle Jurassic of southern Thailand

97

DEESRI , U., CAVIN , L., CLAUDE , J., SUTEETHORN , V. & YUANGDETKLA , P. Morphometric and taphonomic study of a ray-finned fish assemblage (Lepidotes buddhabutrensis, Semionotidae) from the Late Jurassic –earliest Cretaceous of NE Thailand

115

CAVIN , L., DEESRI , U. & SUTEETHORN , V. The Jurassic and Cretaceous bony fish record (Actinopterygii, Dipnoi) from Thailand

125

TONG , H., CLAUDE , J., SUTEETHORN , V., NAKSRI , W. & BUFFETAUT , E. Turtle assemblages of the Khorat Group (Late Jurassic – Early Cretaceous) of NE Thailand and their palaeobiogeographical significance

141

TONG , H., CLAUDE , J., NAKSRI , W., SUTEETHORN , V., BUFFETAUT , E., KHANSUBHA , S., WONGKO , K. & YUANGDETKLA , P. Basilochelys macrobios n. gen. and n. sp., a large cryptodiran turtle from the Phu Kradung Formation (latest Jurassic– earliest Cretaceous) of the Khorat Plateau, NE Thailand

153

LAUPRASERT , K., CUNY , G., THIRAKHUPT , K. & SUTEETHORN , V. Khoratosuchus jintasakuli gen. et sp. nov., an advanced neosuchian crocodyliform from the Early Cretaceous (Aptian– Albian) of NE Thailand

175

SUTEETHORN , S., LE LOEUFF , J., BUFFETAUT , E., SUTEETHORN , V., TALUBMOOK , C. & CHONGLAKMANI , C. A new skeleton of Phuwiangosaurus sirindhornae (Dinosauria, Sauropoda) from NE Thailand

189

KLEIN , N., SANDER , M. & SUTEETHORN , V. Bone histology and its implications for the life history and growth of the Early Cretaceous titanosaur Phuwiangosaurus sirindhornae

217

BUFFETAUT , E., SUTEETHORN , V. & TONG , H. An early ‘ostrich dinosaur’ (Theropoda: Ornithomimosauria) from the Early Cretaceous Sao Khua Formation of NE Thailand

229

LE LOEUFF , J., SAENYAMOON , T., SOUILLAT , C., SUTEETHORN , V. & BUFFETAUT , E. Mesozoic vertebrate footprints of Thailand and Laos

245

vi

CONTENTS

LOCKLEY , M. G., McCREA , R. T. & MATSUKAWA , M. Ichnological evidence for small quadrupedal ornithischians from the basal Cretaceous of SE Asia and North America: implications for a global radiation

255

AMIOT , R., BUFFETAUT , E., LE´ CUYER , C., FERNANDEZ , V., FOUREL , F., MARTINEAU , F. & SUTEETHORN , V. Oxygen isotope composition of continental vertebrate apatites from Mesozoic formations of Thailand; environmental and ecological significance

271

FERNANDEZ , V., CLAUDE , J., ESCARGUEL , G., BUFFETAUT , E. & SUTEETHORN , V. Biogeographical affinities of Jurassic and Cretaceous continental vertebrate assemblages from SE Asia

285

Index

301

Late Palaeozoic and Mesozoic continental ecosystems of SE Asia: an introduction ERIC BUFFETAUT1*, GILLES CUNY2, JEAN LE LOEUFF3 & VARAVUDH SUTEETHORN4 1

Centre National de la Recherche Scientifique, UMR 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure, 24 rue Lhomond, 75231 Paris Cedex 05, France

2

Natural History Museum of Denmark, University of Copenhagen, Øster Voldgade 5 – 7, 1350 Copenhagen K, Denmark 3

Muse´e des Dinosaures, 11260 Espe´raza, France

4

Bureau of Fossil Research and Museums, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected])

The papers in this volume concentrate on the terminal Palaeozoic and Mesozoic non-marine formations of Thailand and Laos. Similar formations are also known from other countries in mainland SE Asia, such as Malaysia (where several nonmarine plant-bearing formations of Jurassic to Cretaceous age are known: Lee et al. 2004), Vietnam and Cambodia (where red beds similar to those of the Khorat Group of Thailand are widespread: Workman 1977; Fontaine & Workman 1978). However, the non-marine Late Palaeozoic and Mesozoic fossils of Thailand and Laos have received much more attention than those of neighbouring countries, from which little has been reported. What we know of the Late Palaeozoic and Mesozoic ecosystems of SE Asia is therefore very largely based on the Thai and Lao records. The study of these Late Palaeozoic and Mesozoic non-marine assemblages began in the 1890s when French geologists took part in expeditions to the Lao principalities that served both a scientific and a political purpose (the latter being clearly predominant, as the more or less avowed aim was to bring that area under French influence). Counillon’s 1896 report of a dicynodont skull from the vicinity of Luang Prabang thus is a landmark in the history of vertebrate palaeontology in SE Asia (Counillon 1896). In later decades, the palaeontological exploration of the Late Palaeozoic and Mesozoic nonmarine formations proceeded in different ways in the French Indochinese colonies and in independent Siam (later Thailand). Although the Thai Department of Mineral Resources was established as early as 1891 by order of king Rama V, the explorations it conducted long had a predominantly economic character and concentrated on gems,

ores and other useful substances. Although a large part of its work also was of an economic nature, the Hanoi-based Service ge´ologique de l’Indochine, founded in 1894 (following the Service des Mines de la Cochinchine, 1868, and the Service des Mines de l’Indochine, 1884), conducted research of a more academic nature, which led to important palaeontological discoveries. In the course of geological mapping in southern Laos in the late 1920s and early 1930s, Hoffet thus found fossil bones and wood at several localities in the Mesozoic red beds of that area (Hoffet 1933). Hoffet’s major finds took place a few years later, when he discovered fairly abundant Cretaceous dinosaur remains near Muong Phalane, in southern Laos (Hoffet 1936, 1942, 1944; see also Buffetaut 1991). After Hoffet was killed by the Japanese in 1945, war conditions prevailed for decades in Laos, and it was not until the 1990s that a French–Lao group could resume research at Muong Phalane (Allain et al. 1999). The vertebrate-bearing red beds (‘Gre`s supe´rieurs’) of southern Laos are now considered a lateral equivalent of the Aptian Khok Kruat Formation of Thailand (Buffetaut 1991). In Thailand, Ho¨gbom (1914) was among the first to mention the red sandstone series of the Khorat Plateau, which he thought was Triassic in age and considered as a ‘Gondwana Formation’. Lee (1927, p. 411), discussing these sandstones, noted that ‘no fossils were found, with the exception of a few pieces of badly petrified wood and small broken fragments of bone’. It was only in the 1950s that Thai and foreign geologists began detailed studies of the stratigraphy of the nonmarine rocks of NE Thailand (see Sattayarak 1983, for a review). The thick pile of non-marine

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 1– 5. DOI: 10.1144/SP315.1 0305-8719/09/$15.00 # The Geological Society of London 2009.

2

E. BUFFETAUT ET AL.

Mesozoic sediments of the Khorat Plateau, long thought to be Triassic in age, was subdivided into several formations, the age of which long remained uncertain (and to some extent still is). In the early 1960s, detailed field investigations resulted in both an important paper on the detailed stratigraphy of the Khorat Group (Ward & Bunnag 1964) and in the first study of vertebrate remains from these formations, which had been sent to Japan for identification. Isolated teeth of a spinosaurid dinosaur and a crocodilian were erroneously identified as belonging to an ichthyosaur and a plesiosaur of Jurassic age by Kobayashi et al. (1963), which resulted in enduring misconceptions about both the depositional environment (supposedly marine) and the age (supposedly Jurassic or older) of the Khorat Group. Starting in 1980, Thai –French joint projects led to the discovery of vertebrate assemblages from various formations of the Khorat Plateau, which indicated non-marine environments and ages ranging from Late Triassic to mid-Cretaceous (e.g. Buffetaut & Ingavat 1986). Although the continental character of all the formations included in the Khorat Group was rapidly accepted, many uncertainties remained concerning the exact age of these formations, although in some cases fossil vertebrates provided useful information. In the 1990s, petroleum exploration in NE Thailand provided seismic and palynological data that were of considerable use for dating the non-marine formations in that area (Racey et al. 1994, 1996) and led to a reconsideration of the definition and content of the Khorat Group, now thought to consist mainly of Early Cretaceous (and probably terminal Jurassic) rocks, comprising, in ascending stratigraphic order, the Phu Kradung, Phra Wihan, Sao Khua,

Phu Phan and Khok Kruat Formations. The Khorat Group is unconformably overlain by the Cenomanian evaporitic Maha Sarakham Formation, and unconformably overlies Late Triassic (to basal Jurassic?) non-marine formations (in ascending stratigraphic order, the Huai Hin Lat and Nam Phong Formations). Vertebrate remains, either as body fossils or as ichnites, are known from all the formations of the Khorat Group, as currently defined, as well as from the underlying Triassic formations. They include freshwater sharks (reviewed by Cuny et al. 2007), bony fishes (reviewed by Cavin et al. 2007), temnospondyl amphibians (Buffetaut et al. 1994a), turtles (reviewed by Tong et al. 2006 and in this volume), phytosaurs (Buffetaut & Ingavat 1982), crocodilians (Lauprasert et al. 2007), pterosaurs (Buffetaut et al. 2003), dinosaurs (reviewed by Buffetaut et al. 2006) and birds (Buffetaut et al. 2005a). The record of Mesozoic nonmarine vertebrates from Thailand is summarized in Figure 1. Fossil invertebrates are also present in many of these formations, as are plant remains, mostly in the form of fossil wood (Philippe et al. 2004). Mesozoic non-marine formations are also widespread in northern Thailand, but have yielded far fewer fossils than in NE Thailand. The general succession seems to be similar to that of the Khorat Plateau, and vertebrate remains, including dinosaur bones (Buffetaut et al. 2006) have been found at a few localities. The southern peninsula of Thailand, which tectonically belongs to the Shan-Thai (or Sibumasu) block (whereas the Khorat Plateau is part of the Indochina block), shows Mesozoic formations, deposited in environments ranging from marine to

Fig. 1. Distribution of main groups of vertebrates in the non-marine formations of Thailand. The Indochina Block comprises NE Thailand (Khorat Plateau). The Sibumasu (or Shan-Thai) Block includes western and southern (peninsular) Thailand. Figure courtesy of L. Cavin (Muse´e d’Histoire Naturelle, Geneva).

INTRODUCTION

brackish and continental (Meesook et al. 2005), which are not easily correlated with those of NE Thailand. Both plant and animal remains are known from some of these formations. The Khlong Min Formation, which is probably Middle to Late Jurassic in age, has yielded a diverse vertebrate assemblage (Buffetaut et al. 1994b) including sharks, bony fishes, temnospondyl amphibians, turtles, crocodiles and dinosaurs (Buffetaut et al. 2005b). It also contains plant remains, including amber (Philippe et al. 2005). The papers in this volume deal with various aspects of the geology and palaeontology of the Late Palaeozoic and Mesozoic non-marine deposits of SE Asia. Metcalfe first provides an introduction to the tectonic and palaeogeographical evolution of that part of the world, during a period marked by the northward drift and collision with mainland Asia of several terranes or ‘microcontinents’ with various origins and histories. This provides a framework for the evolutionary and biogeographical history of non-marine plants and animals in SE Asia during the Late Palaeozoic and Mesozoic. The Permian continental vertebrate-bearing beds of Luang Prabang, in northern Laos, are discussed by Steyer, who tells the story of early palaeontological investigations in that area by French geologists in the 1890s, within the framework of colonial politics, and by Battail, who reviews the question of the dicynodont mammal-like reptiles found in these beds, and shows that, contrary to some previous identifications, Lystrosaurus is not present there. The non-marine Mesozoic formations of Thailand and their abundant and diverse fossils are discussed in a series of papers. Racey reviews the stratigraphy, depositional environment and tectonic setting of the Mesozoic red beds of SE Asia, with special reference to the Khorat Group of NE Thailand. The stratigraphy of the Khorat Group is then discussed in more detail by Racey & Goodall on the basis of palynological evidence suggesting that the Khorat Group is largely Early Cretaceous in age. Philippe et al. provide a better understanding of Thai Mesozoic landscapes by reconstructing the silhouettes of fossil trees from both the Indochina and Shan-Thai blocks. Cuny et al. describe a new elasmobranch fauna supporting a Bathonian–Callovian age for the lower part of the Khlong Min Formation in the southern peninsula of Thailand. This fauna also includes the oldest record of a ray from Asia. Deesri et al.’s detailed study of the semionotiform fishes from Phu Nam Jun (Phu Kradung Formation, NE Thailand) demonstrates that this exceptional site, which has yielded more than 200 fish specimens, is the result of a single mass mortality event. Cavin et al. provide a comprehensive review of the diversity of bony fishes in the Jurassic and Cretaceous formations of Thailand, whereas

3

Tong et al. cover the turtle record for the same time interval, which suggests that SE Asia may have played an important part in the diversification of the Trionychoidea. This is illustrated by a second paper by Tong et al., describing a new large cryptodiran turtle from the Phu Kradung Formation as a basal member of the Trionychoidea. Lauprasert et al. describe a new slender-snouted neosuchian crocodyliform from the Khok Kruat Formation, representing the youngest and most advanced Mesozoic crocodyliform known from Thailand. Suteethorn et al. describe the most complete skeleton ever found of the sauropod Phuwiangosaurus sirindhornae, a dinosaur that is abundantly represented in the Early Cretaceous Sao Khua Formation. Klein et al. present the results of their histological investigations on bones of Phuwiangosaurus sirindhornae, which allow them to reconstruct the growth pattern and life history of that dinosaur. Buffetaut et al. provide a description of a new ornithomimosaur, also from the Sao Khua Formation, which shows advanced characters in its foot structure, although it is one of the oldest known ‘ostrich dinosaurs’. The paper by Le Loeuff et al. is a full review of the rather extensive record of tetrapod (mostly, but not exclusively, dinosaur) footprints found in the Mesozoic of Thailand and Laos. A second ichnological paper, by Lockley et al., discusses in more detail the tracks of small quadrupedal ornithischian dinosaurs, which are found not only in Thailand, but also in other parts of Asia and in North America. Amiot et al. discuss the results of their geochemical investigations on the oxygen isotopes of vertebrate remains from Thailand, which provide information about Mesozoic climate evolution in SE Asia and suggest semi-aquatic habits for spinosaurid dinosaurs. Fernandez et al. conclude the volume with a study of the biogeographical relationships of the Mesozoic vertebrate assemblages of SE Asia, which suggests strong provincialism of the faunas of the Indochina Block during the Jurassic and Cretaceous. It is hoped that the 19 papers in this volume will illustrate the significance of the fossils from the non-marine formations of SE Asia for our understanding of late Palaeozoic and Mesozoic terrestrial ecosystems. They may also point the way to future studies. Systematic research on the fossil floras and faunas of this part of the world began relatively recently, and large parts of SE Asia remain little explored from this point of view. What is already known, principally from Thailand and Laos, suggests that this part of the world may have played an important part in the evolution of various groups during the late Palaeozoic and Mesozoic, but much remains to be done to gain a fuller understanding of the evolution of floras and faunas in the context of changing geography and environments.

4

E. BUFFETAUT ET AL.

References A LLAIN , R., T AQUET , P. ET AL . 1999. Un nouveau genre de dinosaure de la Formation des Gre`s Supe´rieurs (Aptien– Albien du Laos). Comptes Rendus de l’Acade´mie des Sciences, Se´rie II, 329, 609– 616. B UFFETAUT , E. 1991. On the age of the dinosaur-bearing beds of southem Laos. Newsletters on Stratigraphy, 24, 59–73. B UFFETAUT , E. & I NGAVAT , R. 1982. Phytosaur remains (Reptilia, Thecodontia) from the Upper Triassic of north-eastern Thailand. Geobios, 15, 7 –17. B UFFETAUT , E. & I NGAVAT , R. 1986. The succession of vertebrate faunas in the continental Mesozoic of Thailand. Geological Society of Malaysia Bulletin, 19, 161–172. B UFFETAUT , E., T ONG , H. & S UTEETHORN , V. 1994a. First post-Triassic labyrinthodont amphibian in SouthEast Asia: A temnospondyl intercentrum from the Jurassic of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 7, 385– 390. B UFFETAUT , E., T ONG , H., S UTEETHORN , V. & R AKSASKULWONG , L. 1994b. Jurassic vertebrates from the southern peninsula of Thailand and their implications. A preliminary report. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 253– 256. B UFFETAUT , E., S UTEETHORN , V., T ONG , H., C UNY , G. & C AVIN , L. 2003. A pterodactyloid tooth from the Sao Khua Formation (Early Cretaceous) of Thailand. Mahasarakham University Journal, 23, 92–98. B UFFETAUT , E., D YKE , G., S UTEETHORN , V. & T ONG , H. 2005a. First record of a fossil bird from the Early Cretaceous of Thailand. Comptes Rendus Pale´vol, 4, 681– 686. B UFFETAUT , E., S UTEETHORN , V., T ONG , H. & K OS˘ IR , A. 2005b. First dinosaur from the Shan-Thai Block of SE Asia: A Jurassic sauropod from the southern peninsula of Thailand. Journal of the Geological Society, London, 162, 481– 484. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y. N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19–37. C AVIN , L., D EESRI , U. & S UTEETHORN , V. 2007. The Holostei fish record (Actinopterygii) from Thailand. In: T ANTIWANIT , W. (ed.) Proceedings of the International Conference on Geology of Thailand: Towards Sustainable Development and Sufficiency Economy. Department of Mineral Resources, Bangkok, 344 –348. C OUNILLON , J. B. H. 1896. Documents pouvant servir a` l’e´tude ge´ologique des environs de Luang Prabang (Cochinchine). Comptes Rendus de l’Acade´mie des Sciences, 123, 1330–1333. C UNY , G., S UTEETHORN , V., K AMHA , S., L AUPRASERT , K., S RISUK , P. & B UFFETAUT , E. 2007. The Mesozoic record of sharks in Thailand. In: T ANTIWANIT , W. (ed.) Proceedings of the International Conference on

Geology of Thailand: Towards Sustainable Development and Sufficiency Economy. Department of Mineral Resources, Bangkok, 349–354. F ONTAINE , H. & W ORKMAN , D. R. 1978. Review of the geology and mineral resources of Kampuchea, Laos and Vietnam. In: N UTALAYA , P. (ed.) Proceedings of the Third Regional Conference on Geology and Mineral Resources of Southeast Asia. Asian Institute of Technology, Bangkok, 539– 603. H OFFET , J. H. 1933. Etude ge´ologique sur le centre de l’Indochine entre Tourane et le Me´kong (Annam Central et Bas Laos). Bulletin du Service Ge´ologique de l’Indochine, 20, 1 –154. H OFFET , J. H. 1936. De´couverte du Cre´tace´ en Indochine. Comptes Rendus de l’Acade´mie des Sciences, 202, 1867– 1868. H OFFET , J. H. 1942. Description de quelques ossements de titanosauriens du Se´nonien du Bas-Laos. Comptes Rendus des Se´ances du Conseil des Recherches Scientifiques de l’Indochine, 1942, 51–57. H OFFET , J. H. 1944. Description des ossements les plus caracte´ristiques appartenant a` des Avipelviens du Se´nonien du Bas-Laos. Bulletin du Conseil des Recherches Scientifiques de l’Indochine, 1944, 179– 186. H O¨ GBOM , B. 1914. Contributions to the geology and morphology of Siam. Bulletin of the Geological Institution of the University of Upsala, 12, 65–128. K OBAYASHI , T., T AKAI , F. & H AYAMI , I. 1963. On some Mesozoic fossils from the Khorat Series of East Thailand and a note on the Khorat Series. Japanese Journal of Geology and Geography, 34, 181–192. L AUPRASERT , K., C UNY , G., B UFFETAUT , E., S UTEETHORN , V. & T HIRAKHUPT , K. 2007. Siamosuchus phuphokensis, a new goniopholidid from the Early Cretaceous (ante-Aptian) of northeastern Thailand. Bulletin de la Socie´te´ Ge´ologique de France, 178, 201 –216. L EE , W. 1927. Outline of the geology of Siam with reference to petroleum. AAPG Bulletin, 11, 407– 415. L EE , C. P., S HAFEEA L EMAN , M., H ASSAN , K., N ASIB , B. M. & K ARIM , R. 2004. Stratigraphic Lexicon of Malaysia. Geological Society of Malaysia, Kuala Lumpur. M EESOOK , A., S AENGSRICHAN , W. & T EERARUNGSIGUL , N. 2005. Stratigraphy and faunal aspects of the marine Jurassic rocks in southern peninsular Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 558–569. P HILIPPE , M., S UTEETHORN , V., L UTAT , P., B UFFETAUT , E., C AVIN , L., C UNY , G. & B ARALE , G. 2004. Stratigraphical and palaeobiogeographical significance of fossil woods from the Mesozoic Khorat Group of Thailand. Geological Magazine, 141, 319 –328. P HILIPPE , M., C UNY , G. ET AL . 2005. A Jurassic amber deposit in Southern Thailand. Historical Biology, 17, 1– 6. R ACEY , A., G OODALL , J. G. S., L OVE , M. A. & J ONES , P. D. 1994. New age data on the Khorat Group of Northeast Thailand. In: A NGSUWATHANA , P., T., T ANSATHIEN , W., W ONGWANICH ,

Late Palaeozoic and Mesozoic tectonic and palaeogeographical evolution of SE Asia IAN METCALFE School of Environmental and Rural Science, University of New England, Armidale, NSW 2351, Australia (e-mail: [email protected]) Abstract: SE Asia comprises a collage of continental terranes derived directly or indirectly from the India–Australian margin of eastern Gondwana. The Late Palaeozoic and Mesozoic evolution of the region involved the rifting and separation of three elongate continental slivers from eastern Gondwana and the successive opening and closure of three ocean basins, the Palaeo-Tethys, Meso-Tethys and Ceno-Tethys. The Sukhothai Island Arc System, including the Linchang, Sukhothai and Chanthaburi terranes, is identified between the Sibumasu and Indochina– East Malaya terranes in SE Asia and was formed by back-arc spreading in the Permian. The Jinghong, Nan– Uttaradit and Sra Kaeo sutures represent the closed back-arc basin. The Palaeo-Tethys is represented to the west by the Changning–Menglian, Chiang Mai/Inthanon and Bentong– Raub suture zones. The West Sumatra and West Burma blocks rifted and separated from Gondwana, along with Indochina and East Malaya in the Devonian, and together with South China formed a composite terrane ‘Cathaysialand’ in the Permian. They were translated westwards to their positions outboard of the Sibumasu Terrane by strike-slip tectonics in the Late Permian–Early Triassic at the zone of convergence between the Meso-Tethys and Palaeo-Pacific plates. SW Borneo is tentatively identified as possibly being the missing ‘Argoland’ that separated from NW Australia in the Jurassic. Palaeogeographical reconstructions for the Late Palaeozoic and Mesozoic illustrating the tectonic and palaeogeographical evolution of SE Asia are presented.

SE Asia is a unique natural laboratory for studying collisional plate tectonics, suture formation and mountain building (Hall 2002, 2009). The region is also one of the best for studying long-lived terrane dispersion and accretion processes (Metcalfe 1999) and the effects of rapid changes in continent– ocean configurations on biogeographical patterns and ecosystems (Hall & Holloway 1998; Kershaw et al. 2001; Metcalfe et al. 2001). The tectonic evolution, terrane accretion, and assembly of SE Asia during the Late Palaeozoic and Mesozoic has resulted in complex changes in the palaeogeography of SE Asia and in the juxtaposition of contrasting but temporally coincident palaeo-ecosystems in the region. Changing continent–ocean configurations, palaeogeography and climate changes in the Late Palaeozoic and Mesozoic of what is now SE Asia have resulted in complex patterns of both marine and terrestrial ecosystems of SE Asia both in space and time. This paper presents an overview of the tectonic and palaeogeographical evolution of SE Asia in the Late Palaeozoic and Mesozoic.

Tectonic framework of SE Asia SE Asia is located at the zone of convergence between the Eurasia, Pacific and Indo-Australian plates (Fig. 1). The region comprises a complex assemblage of continental terranes (bounded by

suture zones or other major tectonic features), island arcs and small ocean basins. The older Palaeozoic and Mesozoic portions of the region comprise allochthonous continental terranes (small to medium-sized allochthonous continental blocks that have distinctive tectonostratigraphic histories) that were assembled prior to the current collision of Australia with the region. The Cenozoic evolution of the region has been well studied, and the evolution and reconstruction models of Hall (2002) are considered the most reliable. This paper focuses on the Late Palaeozoic and Mesozoic evolution.

Continental terranes and their origins The principal continental terranes and sutures of SE Asia are shown in Figures 2 and 3. Terranes are grouped according to their interpreted origins and are discussed briefly under these origin groupings below. Descriptions of the terranes have been given by Metcalfe (2006). All the continental terranes of East and SE Asia are interpreted to have been directly or indirectly derived from the eastern Gondwana margin at different times and some small terranes were subsequently derived indirectly from larger Gondwana-derived terranes or composite terranes (see Metcalfe 1986, 1988, 1990, 1991, 1993, 1996a, b, 1998, 2001, 2002, 2005, 2006; Table 1). I have previously described the successive rifting and northwards drift of three

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 7– 23. DOI: 10.1144/SP315.2 0305-8719/09/$15.00 # The Geological Society of London 2009.

8

I. METCALFE

Fig. 1. Principal tectonic plates of SE Asia. Arrows show plate motions relative to Eurasia.

Gondwana-derived continental slivers (now disrupted into various terranes), with the successive opening and closure of the Palaeo-Tethys, Meso-Tethys and Ceno-Tethys ocean basins. This broad scenario is still advocated here (Fig. 4), but recent new data demand modification of the terrane make-up of these continental slivers, and also reinterpretations of the origins and boundaries of some of the terranes in the region. Terranes derived from Gondwana in the Devonian. A group of East and SE Asian terranes are interpreted to have rifted and separated from Gondwana as an elongate continental sliver in the Devonian (Fig. 4) and comprise North China, South China (including Hainan), Indochina–East Malaya (including the Qamdao– Simao/Simao and disrupted West Sumatra and West Burma terranes) and Tarim (including the disrupted Kunlun, Qaidam and Ala Shan terranes). The West Sumatra terrane was proposed by Hutchison (1994) and Barber & Crow (2003), and is interpreted to have been translated westwards from ‘Cathaysialand’ (combined South China –Indochina–East Malaya composite terrane in Permo-Triassic times) as suggested by Barber et al. (2005) and Metcalfe (2005). A Devonian separation of these terranes from Gondwana is still advocated here based on palaeomagnetic and biogeographical data from the terranes themselves and also from Devonian age data for oceanic radiolarian cherts in the

Palaeo-Tethys (for details, see Metcalfe 1988, 1990, 1998, 2005). Jablonski & Saitta (2004) have, however, on the basis of transgressive–regressive sequences in western Australian basins, argued for a later Early Carboniferous (Vise´an) separation of South China, Indochina and Simao terranes. This timing seems too young in view of the fact that the Palaeo-Tethys suture zone includes oceanic sediments of Devonian age and no post-Devonian Gondwana biota is reported from the terranes in question. Terranes derived from Gondwana in the Early Permian. The second continental sliver that rifted and separated from Gondwana in the Early Permian (with opening of the Meso-Tethys) was the Cimmerian continent of Sengor (1979, 1984), which included the Sibumasu terrane (Metcalfe 1984) of SE Asia and the Baoshan and Tengchong terranes of Yunnan in western China. I have presented evidence for the NW Australian origin and Early Permian rifting and separation of the Sibumasu portion of the Cimmerian continent from Gondwana in a series of papers (e.g. Metcalfe 1986, 1988, 1990, 1991, 1993, 1996a, b, 1998, 2001, 2002, 2005, 2006) and this will not be repeated here. Recent studies (Sone & Metcalfe 2008) have led to the recognition of an island arc system, the Sukhothai Island Arc system, through SE Asia (Fig. 5) situated between Sibumasu and Indochina–East Malaya (from which it was derived by back-arc spreading). This interpretation,

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION

9

Fig. 2. Distribution of principal continental terranes and sutures of East and SE Asia. WB, West Burma; SWB, SW Borneo; S, Semitau Terrane; L, Lhasa Terrane; QT, Qiangtang Terrane; QS, Qamdo–Simao Terrane; SI, Simao Terrane; SG, Songpan Ganzi accretionary complex; KL, Kunlun Terrane; QD, Qaidam Terrane; AL, Ala Shan Terrane; LT, Linchang Terrane; CT, Chanthaburi Terrane.

together with recognition of the West Sumatra terrane and extension of the Bentong–Raub suture through the tin islands of north Sumatra, dictates some modification of the boundaries of the Sibumasu Terrane (Figs 2 and 3). The newly proposed

Sukhothai Island Arc system includes the Linchang, Sukhothai and Chanthaburi terranes (Sone & Metcalfe 2008) and is broadly equivalent to the ‘Sukhothai zone’ originally defined as the ‘Sukhothai fold belt’ by Bunopas (1982, fig. 139)

10

I. METCALFE

Fig. 3. Distribution of continental blocks, fragments and terranes, and principal sutures of SE Asia. 1, East Java; 2, Bawean; 3, Paternoster; 4, Mangkalihat; 5, West Sulawesi; 6, Semitau; 7, Luconia; 8, Kelabit–Longbowan; 9, Spratly Islands– Dangerous Ground; 10, Reed Bank; 11, North Palawan; 12, Paracel Islands; 13, Macclesfield Bank; 14, East Sulawesi; 15, Bangai-Sula; 16, Buton; 17, Obi-Bacan; 18, Buru-Seram; 19, West Irian Jaya; LT, Lincang Terrane; SI, Simao Terrane; ST, Sukhothai Terrane; CT, Chanthaburi Terrane; C-M, Changning– Menglian Suture; C.-Mai, Chiang Mai Suture; Nan-Utt., Nan –Uttaradit Suture.

in Thailand. The previously recognized Inthanon Zone (¼ Inthanon Fold Belt of Bunopas) in Thailand is here interpreted as part of the Palaeo-Tethys suture Zone (Fig. 5) following Sone & Metcalfe (2008). Use and abuse of the terms ‘Shan-Thai’ and ‘Gondwana–Tethys/Cathaysia Divide’. The Sibumasu Terrane (Metcalfe 1984) is the Gondwanaderived terrane in SE Asia that included parts of western Thailand, Burma, western Peninsular

Malaysia and NW Sumatra, characterized by the presence of late Carboniferous and early Permian glacial–marine diamictites (Fig. 6) and late Palaeozoic strata with Gondwana affinity faunas and floras. This terrane, as defined, is not equivalent to the Shan-Thai Terrane of Bunopas (1982), which was defined as including ‘eastern Burma, western Thailand and northwestern Malay Peninsula’, but some workers have equated, and continue to equate Shan-Thai with Sibumasu. It is here stressed that these are not equivalents and the terms should not

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION

11

Table 1. Suggested origins for the East and SE Asian continental terranes Terrane North China South China Sibumasu Indochina East Malaya West Sumatra West Burma Lhasa Qiangtang Simao Lincang Chanthaburi Kunlun Qaidam Ala Shan Tarim Hainan Kurosegawa (Japan) South Kitakami (Japan) East Java Bawean Paternoster West Sulawesi Mangkalihat East Sulawesi Banggai-Sula Buton Obi-Bacan Buru-Seram West Irian Jaya SW Borneo Semitau Luconia Kelabit–Longbowan Spratley Islands–Dangerous Ground Reed Bank North Palawan Paracel Islands Macclesfield Bank

be used interchangeably. Matters have been made worse recently with some proposals to apply the term ‘Shan-Thai’ to include Cathaysian elements of Thailand (e.g. Hirsch et al. 2006; Ishida et al. 2006; Ferrari et al. 2009), introducing further confusion of the originally defined Gondwana Shan-Thai Terrane. The ‘Shan-Thai Block’ of Hirsch et al. (2006) and Ferrari et al. (2009) in fact includes both continental terranes and suture zones (see Hirsch et al. 2006, fig. 2; Ferrari et al. 2009, fig. 5), which is an unacceptable oversimplification and composite grouping of very different tectonic units. Confusion has also arisen relating to the major Late Palaeozoic biogeographical boundary recognized through East and SE Asia that separates Gondwana faunas and floras from Cathaysian

Origin N. Australia Himalaya – Iran region of Gondwana NW Australia Eastern Gondwana Eastern Gondwana Eastern Gondwana (Cathaysialand) Eastern Gondwana (Cathaysialand) Himalayan Gondwana Himalayan Gondwana Eastern Gondwana: South China Indochina – East Malaya (Cathaysialand) Indochina – East Malaya (Cathaysialand) NE Gondwana? Originally part of Tarim? NE Gondwana? Originally part of Tarim? NE Gondwana? Originally part of Tarim? Australian Gondwana? NE Gondwana? Australian Gondwana? North China margin NW Australia (Argoland) NW Australia (Argoland) NW Australia (Argoland) NW Australia (Argoland) NW Australia (Argoland) New Guinea region of the Australian margin New Guinea region of the Australian margin New Guinea region of the Australian margin New Guinea region of the Australian margin New Guinea region of the Australian margin New Guinea region of the Australian margin Cathaysialand or NW Australia (Argoland)? Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin) Cathaysialand (South China – Indochina margin)

faunas and floras. This major biogeographical divide has been termed the ‘Gondwana–Tethys Divide’ or ‘Gondwana–Cathaysia Divide’ by some workers and has been used to mark the boundary between Gondwana-derived continental terranes in the west, with Early Permian cold- or coolclimate sediments and biota, from warm-climate equatorial Cathaysian continental terranes to the east (Ueno 2003; Metcalfe 2005). It was the recognition of this major biogeographical divide, coupled with Late Carboniferous –Early Permian diamictites interpreted to be of glacial–marine origin, that led to models of Gondwana dispersion and Asian accretion of terranes derived from Gondwana (e.g. Metcalfe 1988, 1990). The Gondwana– Cathaysia biogeographical divide has been taken by some workers to indicate the boundary between

12

I. METCALFE

Fig. 4. Schematic diagram showing times of separation and subsequent collision of the three continental slivers or collages of terranes that rifted from Gondwana and translated northwards by the opening and closing of three successive oceans, the Palaeo-Tethys, Meso-Tethys and Ceno-Tethys.

Gondwana terranes and Cathaysian terranes (e.g. Ueno 1999; Ueno & Hisada 1999, 2001) and also used to identify the position of the Palaeo-Tethys suture as corresponding to the Mai Yuam Fault in Thailand (Ueno & Hisada 2001; Ueno 2003). This is an unfortunate interpretation, because Cathaysian faunal elements in seamounts occurring within the Palaeo-Tethys suture zone have been misinterpreted to indicate that the suture lies further to the west. This was particularly illustrated by Hirsch et al. (2006, fig. 2), who placed the Gondwana–Tethys divide in Peninsular Malaysia west of stable continental margin limestones (Kanthan Limestone, Kinta Valley) that contain Gondwana faunas. Hirsch et al. also showed a ‘Pattani Suture’ in the Gulf of Thailand and extending into Peninsular Malaysia delineating the eastern margin of their ‘Mae Sariang Zone’ yet did not provide any description of or justification for this suture zone. Terranes derived from Gondwana in the Late Triassic–Late Jurassic. Metcalfe (1990) interpreted western Burma, which was named the Mount Victoria Land block, as a terrane derived from NW Australia that represented the missing ‘Argoland’ continental fragment that must have rifted from

that region in the Jurassic. This terrane was later named West Burma (Metcalfe 1996a, b) to avoid nomenclatural confusion with Mount Victoria in Antarctica. Correlation of the Mogok Belt in Burma (which forms the boundary between Sibumasu and West Burma) with the Medial Sumatra Tectonic Zone separating Sibumasu from the Cathaysian West Sumatra Block suggests that the West Burma terrane represents a continuation of the West Sumatra Block (Barber & Crow 2009). Barber & Crow (2009) also pointed out that Middle Permian fusulinids from Karmine, Burma (Oo et al. 2002) are Cathaysian in nature and similar to Middle Permian faunas of the West Sumatra Block, supporting their interpretation. The West Burma Block is therefore here regarded as a Cathaysian terrane that was derived from the Cathaysialand superterrane together with the West Sumatra Block in the Permian. If this interpretation is accepted, then the West Burma Block must have been derived from Gondwana in the Devonian as part of the Indochina–East Malaya terrane. This leaves the identity of ‘Argoland’ yet to be established. One possibility (A. J. Barber, pers. comm.; Hall 2009; Hall et al. 2009) is that SW Borneo might be a candidate. I previously considered this

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION

13

Fig. 5. Tectonic subdivision of mainland SE Asia proposed by Sone & Metcalfe (2008), showing the Sukhothai Island Arc System, Palaeo-Tethys and Sukhothai back-arc sutures, and occurrence of deep-sea sediments in each local suture. C-M S. Z., Changning–Menglian Suture Zone; BKK, Bangkok. After Sone & Metcalfe (2008).

but ruled it out on the basis that Cathaysian faunas were known from the Carboniferous –Lower Permian Terbat Limestone on the Sarawak– Kalimantan border. If, however, these limestones form part of the Kuching zone (accretionary

complex) or the small Semitau Block, and not the core of the SW Borneo Block, then SW Borneo becomes a candidate for the ‘Argoland’ block. This would be supported by the occurrence of diamonds in headless placers (placer diamond deposits

14

I. METCALFE

Fig. 6. Map of mainland SE Asia, showing the distribution of Late Carboniferous– Early Permian glacial– marine sedimentary rocks and major alluvial diamond deposits. Inset photograph shows dropstone in glacial –marine diamictite oriented vertical to bedding, Singa Formation, Langkawi Islands, Peninsular Malaysia. Abbreviations as in Figure 2.

without any obvious local or regional diamond source) in SW Borneo (Fig. 6), which could well have been derived from NW Australia. New information is also now available on the small Gondwana-derived continental fragments located in eastern Indonesia. Recent provenance studies have identified an Australian Gondwana-derived East Java terrane (Smyth et al. 2007). The Bawean Arch and Paternoster Platform pre-Cenozoic continental blocks (Manur & Barraclough 1994) are also possibly of Australian Gondwana origin but hard data supporting this are at present lacking. Other small continental blocks postulated to have had their origin on the Mesozoic margin of Australian Gondwana include the West Sulawesi Block (which has been linked with the East Java

terrane) and the Mangkalihat Block in NE Borneo. It is possible that these microcontinental blocks (numbered 1– 5 in Fig. 3) may in fact represent a single disrupted terrane derived from NW Australia (Hall 2009).

Tectonic and palaeogeographical evolution Late Palaeozoic Palaeomagnetic, biogeographical and tectonostratigraphic data suggest that a continental sliver comprising North China, Indochina (including East Malaya, West Sumatra and West Burma), Tarim and South China rifted and separated from eastern

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION

Gondwana in the Devonian, and Palaeo-Tethys opened between this sliver and Gondwana (Metcalfe 1996a, b). By Late Devonian –Early Carboniferous times, significant spreading had occurred in the Palaeo-Tethys ocean but some connection with Gondwana probably remained in the east (Metcalfe 2001) and endemic shallow-marine faunas characterized by Chuiella developed on the western portion of this separating continental sliver (Chen & Shi 1999; Fig. 7). Tarim collided with Siberia in the Late Carboniferous to Early Permian and was firmly welded to Asia by the Middle Permian (Carroll et al. 1995). Late Early Carboniferous floras of Indochina– East Malaya and South China are very similar (Laveine et al. 1999), suggesting a continental connection between these terranes at that time. Tectonostratigraphic data also indicate that these terranes collided and fused in the Carboniferous along the Song Ma Suture (Metcalfe 2001) to form a superterrane, here termed Cathaysialand (Fig. 8). Palaeomagnetic data indicate that North China remained in equatorial to low northern latitudes during the Carboniferous– Permian between Cathaysialand and NE Pangaea (Nie 1991). During the Permian, Cathaysialand and North China, which were situated within the Tethys ocean and largely isolated from the rest of Pangaea, developed the characteristic terrestrial flora and shallow-marine faunas of the Cathaysian biogeographical province. The

15

Cathaysian equatorial flora is distinctly different from the cooler climate contemporaneous northern Angara and southern Gondwana floras, and coeval equatorial American and Euramerican floras of Pangaea (Fig. 9). Shallow-marine faunas are also distinctive and include endemic forms (e.g. the conodont Pseudosweetognathus; see Fig. 8b). During the Permian a second continental sliver, the Cimmerian continent (Sengo¨r 1979, 1984), the eastern portion of which is the Sibumasu Terrane, rifted and separated from eastern Gondwana and drifted northwards to collide with Cathaysialand (Fig. 8). As this terrane separated and moved northwards, it exhibits a progressive change in marine provinciality from peri-Gondwanan Indoralian Province faunas in the Asselian –Sakmarian to endemic Sibumasu Province faunas in the middle Permian to Cathaysian Province faunas in the Late Permian (Shi & Archbold 1998; Ueno 2003). The northwards latitudinal change, isolation from Gondwana, then amalgamation with Cathaysialand is reflected in the change from cool-climate or -water conditions to warm-wate or -climate conditions and in the changing biogeographical affinities of faunas and changing ecosystems. During the northwards drift of Sibumasu, the Palaeo-Tethys was subducted beneath northern Pangaea, North China and Cathaysialand. Subduction beneath Cathaysialand resulted in the opening of a back-arc basin and development of the Sukhothai Island Arc terranes probably

Fig. 7. Reconstruction of eastern Gondwana in Late Devonian to Early Carboniferous (Tournaisian) times showing the postulated positions of the East and SE Asian terranes. Also shown is the distribution of the endemic Tournaisian brachiopod genus Chuiella. NC, North China; SC, South China; T, Tarim; I, Indochina –East Malaya– West Sumatra– West Burma; QI, Qiangtang; L, Lhasa; S, Sibumasu; SWB, SW Borneo; WC, Western Cimmerian Continent.

16

I. METCALFE

Fig. 8. Palaeogeographical reconstructions of the Tethyan region for (a) Early Early Permian (Asselian–Sakmarian), (b) Late Early Permian (Kungurian) and (c) Late Permian (Changhsingian) showing relative positions of the East and SE Asian terranes and distribution of land and sea. Also shown is the Late Early Permian distribution of

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION

17

Fig. 9. Distribution of Early Permian floral provinces plotted on (a) present-day geographical map, and (b) Early Permian palaeogeographical map. KT, Kurosegawa Terrane. Other abbreviations as in Figure 2.

developed over thin continental crust (Lincang, Sukhothai and Chanthaburi terranes) that originally formed the margin of Indochina (Fig. 8c). The back-arc basin then collapsed and closed to form the Jinghong, Nan–Uttaradit and Sra Kaeo sutures (Sone & Metcalfe 2008). Collision of the Sibumasu

terrane with the Sukhothai Island Arc terranes and Cathaysialand closed the southeastern Palaeo – Tethys in the Permian–Triassic producing the Changning– Menglian, Inthanon and Bentong – Raub suture zones. The precise timing of collision of Sibumasu with Indochina– East Malaya is

Fig. 8. (Continued ) biogeographically important conodonts, and Late Permian tetrapod vertebrate Dicynodon localities on Indochina and Pangaea in the Late Permian. SC, South China; T, Tarim; I, Indochina; EM, East Malaya; WS, West Sumatra; NC, North China; SI, Simao; S, Sibumasu; WB, West Burma; QI, Qiangtang; L, Lhasa; SWB, SW Borneo; WC, Western Cimmerian Continent. Land and sea shading as in Figure 7.

18

I. METCALFE

debated. Metcalfe (2000) suggested a Late Permian–Early Triassic collision for the Bentong–Raub suture segment based on a range of constraining data, and based on interpretation of radiolarian cherts of Triassic age belonging to the Semanggol Formation and equivalents in western Peninsular Malaysia and Sumatra having formed in a successor basin rather than being deposited on Palaeo-Tethys ocean floor. This timing was challenged by several workers, who suggested a Late Triassic or even Jurassic collision based on interpretation of the Semanggol cherts and equivalents as Palaeo-Tethyan deposits (e.g. Sashida et al. 1995, 1999, 2000a, b; Kamata et al. 2002; Hirsch et al. 2006; Ishida et al. 2006; Ueno et al. 2006). Barber & Crow (2009) supported a latest Permian collision and suggested that the Semanggol and equivalent Triassic cherts formed in a successor basin(s) over the foreland fold-and-thrust belt as suggested by Metcalfe (2000). There are, however, valid arguments for a younger (late Triassic) collision and suturing to the north along the Changning–Menglian suture in SW China (Liu et al. 1996). During the Late Permian–Early Triassic collision of Sibumasu and Indochina–East Malaya, and as a result of Cathaysialand being located at the zone of interaction between the north-moving Meso-Tethys and west-moving Palaeo-Pacific plates, it is postulated here, following the suggestion of Barber & Crow (2003) and Barber et al. (2005), that the West Sumatra and West Burma terranes (as an initial single unit later split by the opening of the Andaman Sea) were translated westwards

to their current biogeographically unexpected positions outboard of the Sibumasu Terrane by largely strike-slip translation. A land connection (which may have been temporary) between Indochina and Pangaea in the Late Permian is indicated by the confirmed presence of the Late Permian tetrapod vertebrate Dicynodon in Laos (Fig. 8c). It is not known if this connection was via South and North China (most likely), or via the western Cimmerian continent, or both. The timing of collision between South and North China, along the Qingling– Dabie –Sulu suture zone has long been controversial, with Mid-Palaeozoic, Late Palaeozoic and Late Triassic –Jurassic timings being proposed. A Permian to Early Triassic collision is here interpreted. This is based on studies of low-grade metamorphic rocks in the Sulu belt (Zhou et al. 2008), and geochronological and structural data (e.g. Faure et al. 2003) indicating Permian subduction of South China beneath North China. In addition, identification of a Devonian –Triassic accretionary wedge that includes eclogites, and that formed a coeval volcano-plutonic arc that stretches from the Longmen Shan to Korea, supports subduction beneath the Qinling– Sino–Korean plate and a Permian– Triassic collision (Hacker et al. 2004).

Mesozoic During the Triassic, welding of Sibumasu and Cathaysialand continued and collision with North China was largely completed by Late Triassic times (Fig. 10). Comparisons of apparent polar wander

Fig. 10. Palaeogeographical reconstructions of the Tethyan region for the Late Triassic (Rhaetian) showing relative positions of the East and SE Asian terranes and distribution of land and sea. Abbreviations and sea shading as in Figures 7 and 8.

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION

19

Fig. 11. Palaeogeographical reconstructions for Eastern Tethys in (a) Late Jurassic, (b) Early Cretaceous and (c) Late Cretaceous showing distribution of continental blocks and fragments of SE Asia–Australasia and land and sea. SG, Songpan Ganzi accretionary complex; SC, South China; QS, Qando–Simao; SI, Simao; QI, Qiangtang; S, Sibumasu I, Indochina; EM, East Malaya; WSu, West Sumatra; L, Lhasa; WB, West Burma; SWB, SW Borneo; NP, North Palawan and other small continental fragments now forming part of the Philippines basement; M, Mangkalihat; WS, West Sulawesi; P, Paternoster; B, Bawean; PA, incipient East Philippine arc; PS, Proto-South China Sea; Sm, Sumba. M numbers represent Indian Ocean magnetic anomalies.

20

I. METCALFE

paths of North and South China imply that collision between these blocks did, however, continue in the Jurassic but was complete by the Late Jurassic. The time of rapid (18 Ma21) relative angular velocity between the two plates (225 –190 Ma) coincides with a peak in U – Pb and Ar–Ar dates obtained from metamorphic rocks in the Qingling–Dabie –Sulu suture (Gilder & Courtillot 1997). Thus, the initial consolidation of what is now East and SE Asia took place in Late Triassic – Jurassic times. Palaeo-Tethyan ocean crust trapped between the western Cimmerian continent, Cathaysialand, North China and Siberian Pangaea was covered by thick Triassic deposits eroded from adjacent collisional orogens and became the Songpan Ganzi giant suture knot (Fig. 10). The Meso-Tethys was probably at its widest at this time and further rifting of the Indian–Australian margin of Gondwana was initiated and continued into the Jurassic. Separation of a further collage of small terranes took place in the Jurassic, the Ceno-Tethys ocean opening behind these terranes as they drifted northwards. The continental fragments that separated from Gondwana at this time are somewhat uncertain but are here interpreted to include the Lhasa, SW Borneo, East Java, Bawean, Paternoster, Mangkalihat, West Sulawesi and Sumba blocks (Fig. 11). The timing of rifting and separation of the Lhasa block from Gondwana has seen much debate over the years. An early separation in the Permian, along with other elements of the Cimmerian continent as part of a ‘Mega-Lhasa Terrane’ has been proposed by some workers (e.g. Alle`gre et al. 1984; Dercourt et al. 1993, 2000). Others have argued for a later separation in the Late Triassic– Early Jurassic (e.g. Metcalfe 2002; Golonka 2007), and this timing is still advocated here and supported by information on oceanic cherts from the Yarlung– Zangbo suture (Matsuoka et al. 2002) and recent palaeomagnetic data (Otofuji et al. 2007). Few data are available from the Banggong suture zone between the Lhasa and Qiangtang terranes to constrain the age range of the ocean that the suture represents. An earlier separation in the Permian may be supported by Permian limestone blocks interpreted as possible seamount caps in the Indus –Yarlung suture zone (Shen et al. 2003). Little is known of the precise age of separation and northwards drift of the other small Australian-derived continental blocks now located in Java, Borneo and Sulawesi, but Jablonski & Saitta (2004) have suggested that these microplates separated and migrated successively in the Hettangian, Oxfordian, Kimmeridgian and Tithonian based on megasequence studies and transgressive–regressive cycles in the Perth Basin and Westralian superbasin. Interestingly, Jablonski & Saitta (2004) did not refer to the Lhasa block at all and did not show this on their

palaeogeographical reconstructions. They provided little evidence for the identification of the blocks rifting at different times. By Late Cretaceous times the Lhasa block was welded to East Asia and SW Borneo and the other small continental fragments now found in the Java –Borneo– Sulawesi region had approximately reached their current positions relative to Indochina and other SE Asia blocks (Fig. 11c).

Conclusions The Late Palaeozoic and Mesozoic evolution of SE Asia involved the rifting and separation of three collages of continental terranes (probably as elongate slivers) from eastern Gondwana and the successive opening and closure of three ocean basins, the Palaeo-Tethys, Meso-Tethys and Ceno-Tethys. The Palaeo-Tethys is represented in SE Asia by the Changning–Menglian, Chiang Mai/Inthanon and Bentong –Raub suture zones. The Sukhothai Island Arc System, including the Linchang, Sukhothai and Chanthaburi terranes, is identified between the Sibumasu and Indochina– East Malaya terranes in SE Asia and was formed by back-arc spreading in the Permian. The Jinghong, Nan–Uttaradit and Sra Kaeo sutures represent the closed back-arc ocean. The West Sumatra and West Burma blocks rifted and separated from Gondwana, along with Indochina and East Malaya in the Devonian, and formed a composite terrane ‘Cathaysialand’ with South China in the Permian. West Sumatra and West Burma were translated westwards to their positions outboard of the Sibumasu Terrane by strike-slip translation in the Late Permian– Early Triassic at the zone of convergence between the Meso-Tethys and Palaeo-Pacific plates. SW Borneo is tentatively identified as possibly being the missing ‘Argoland’ that separated from NW Australia in the Jurassic. I would like to thank R. Hall and M. Faure for their helpful reviews. I would also like to thank A. Barber for discussions and comments on the draft manuscript. Facilities provided by the University of New England are gratefully acknowledged.

References A LLE` GRE , C. J. ET AL . 1984. Structure and evolution of the Himalaya-Tibet orogenic belt. Nature, 307, 17– 22. B ARBER , A. J. & C ROW , M. J. 2003. An evaluation of plate tectonic models for the development of Sumatra. Gondwana Research, 6, 1 –28.

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION B ARBER , A. J. & C ROW , M. J. 2009. The structure of Sumatra and its implications for the tectonic assembly of Southeast Asia and the destruction of Paleotethys. Island Arc (in press). B ARBER , A. J., C ROW , M. J. & D E S MET , M. E. M. 2005. Tectonic evolution. In: B ARBER , A. J., C ROW , M. J. & M ILSOM , J. S. (eds) Sumatra: Geology, Resources and Tectonic Evolution. Geological Society, London, Memoirs, 31, 234– 259. B UNOPAS , S. 1982. Palaeogeographic history of Western Thailand and adjacent parts of Southeast Asia—a plate tectonics interpretation. Department of Mineral Resources, Bangkok, Geological Survey Paper, 5. C ARROLL , A. R., G RAHAM , S. A., H ENDRIX , M. S., Y ING , D. & Z HOU , D. 1995. Late Paleozoic tectonic amalgamation of northwestern China; sedimentary record of the northern Tarim, northwestern Turpan, and southern Junggar basins. Geological Society of America Bulletin, 107, 571 –594. C HEN , Z. Q. & S HI , G. R. 1999. Chuiella gen. nov. (Brachiopoda) and palaeoecology from the Lower Carboniferous of the Kunlun Mountains, NW China. Alcheringa, 23, 259–275. D ERCOURT , J., R ICOU , L. E. & V RIELYNCK , B. (eds) 1993. Atlas Tethys Paleoenvironmental Maps. Gauthier-Villars, Paris. D ERCOURT , J., G AETANI , M. ET AL . (eds) 2000. Atlas Peri-Tethys Paleoegeographical Maps. CCGM– CGMW, Paris. F AURE , M., L IN , W., S CHARER , U., S HU , L., S UN , Y. & A RNAUD , N. 2003. Continental subduction and exhumation of UHP rocks. Structural and geochronological insights from the Dabieshan (East China). Lithos, 70, 213–241. F ERRARI , O. M., H OCHARD , C. & S TAMPFLI , G. M. 2008. An alternative plate tectonic model for the Palaeozoic-Early Mesozoic Palaeotethyan evolution of Southeast Asia (Northern Thailand-Burma). Tectonophysics, 451, 346– 365. G ILDER , S. A. & C OURTILLOT , V. 1997. Timing of the North-South China collision from new middle to late Mesozoic paleomagnetic data from the North China Block. Journal of Geophysical Research, 102, 17713– 17727. G OLONKA , J. 2007. Late Triassic and Early Jurassic palaeogeography of the world. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 297–307. H ACKER , B. R., R ATSCHBACHER , L. & L IOU , J. G. 2004. Subduction, collision and exhumation in the ultrahighpressure Qinling-Dabie orogen. In: M ALPAS , J., F LETCHER , C. J. N., A LI , J. R. & A ITCHISON , J. C. (eds) Aspects of the Tectonic Evolution of China. Geological Society, London, Special Publication, 226, 157–175. H ALL , R. 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computerbased reconstructions and animations. Journal of Asian Earth Sciences, 20, 353–434. H ALL , R. 2009. The Eurasia SE Asian margin as a modern example of an accretionary orogen. Geological Society, London, Special Publications, 318. H ALL , R. & H OLLOWAY , J. D. (eds) 1998. Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden.

21

H ALL , R., VAN H ATTUM , M. C. A. & S PAKMAN , W. 2008. Impact of India-Asia collision on SE Asia: the record in Borneo. Tectonophysics, 451, 366– 389. H IRSCH , F., I SHIDA , K., K OZAI , T. & M EESOOK , A. 2006. The welding of Shan-Thai. Geosciences Journal, 10, 195– 204. H UTCHISON , C. S. 1994. Gondwana and Cathaysian blocks, Palaeotethys sutures and Cenozoic tectonics in Southeast Asia. Geologische Rundschau, 82, 388– 405. I SHIDA , K., N ANBA , A., H IRSCH , F., K OZAI , T. & M EESOOK , A. 2006. New micropalaeontological evidence for a Late Triassic Shan-Thai orogeny. Geosciences Journal, 10, 181 –194. J ABLONSKI , D. & S AITTA , A. J. 2004. Permian to Lower Cretaceous plate tectonics and its impact on the tectono-stratigraphic development of the Western Australian margin. APPEA Journal, 2004, 287 –328. K AMATA , Y., S ASHIDA , K., U ENO , K., H ISADA , K., N AKORNSRI , N. & C HARUSIRI , P. 2002. Triassic radiolarian faunas from the Mae Sariang area, northern Thailand and their palaeogeographic significance. Journal of Asian Earth Sciences, 20, 491– 506. K ERSHAW , P., D AVID , B., T APPER , N., P ENNY , D. & B ROWN , J. (eds) 2001. Bridging Wallace’s Line: The Environmental and Cultural History and Dynamics of the SE-Asian–Australian Region. Advances in GeoEcology, 34. L AVEINE , J.-P., R ATANASTHIEN , B. & A ZHAR H AJI H USSIN . 1999. The Carboniferous floras of Southeast Asia: Implications for the relationships and timing of accretion of some Southeast Asian Blocks. In: M ETCALFE , I. (ed.) Gondwana dispersion and Asian accretion, Final Results Volume for IGCP Project 321. A. A. Balkema, Rotterdam, 229–246. L IU , B., F ENG , Q., F ANG , N., J IA , J., H E , F., Y ANG , W. & L IU , D. 1996. Tectono-paleogeographic framework and evolution of the Paleotethyan archipelagoes ocean in Changning–Menglian belt, Western Yunnan, China. In: F ANG , N. & F ENG , Q. (eds) Devonian to Triassic Tethys in Western Yunnan, China. China University of Geosciences Press, Wuhan, 1– 12. M ANUR , H. & B ARRACLOUGH , R. 1994. Structural control on hydrocarbon habitat in the Bawean area, East Java Sea. Proceedings of the 23rd Annual Indonesian Petroleum Association, Convention, Jakarta. Jakarta, Indonesian Petroleum Association, 129– 144. M ATSUOKA , A., Y ANG , Q., K OBAYASHI , K., T AKEI , M., N AGAHASHI , T., Z ENG , Q. & W ANG , Y. 2002. Jurassic–Cretaceous radiolarian biostratigraphy and sedimentary environments of the Ceno-Tethys: Records from the Xialu Chert in the Yarlung– Zangbo Suture Zone, southern Tibet. Journal of Asian Earth Sciences, 20, 277– 287. M ETCALFE , I. 1984. Stratigraphy, palaeontology and palaeogeography of the Carboniferous of Southeast Asia. Me´moires de la Socie´te´ Ge´ologique de France, 147, 107–118. M ETCALFE , I. 1986. Late Palaeozoic palaeogeography of Southeast Asia: Some stratigraphical, palaeontological and palaeomagnetic constraints. Geological Society of Malaysia Bulletin, 19, 153–164. M ETCALFE , I. 1988. Origin and assembly of Southeast Asian continental terranes. In: A UDLEY -C HARLES ,

22

I. METCALFE

M. G. & H ALLAM , A. (eds) Gondwana and Tethys. Geological Society, London, Special Publications, 37, 101–118. M ETCALFE , I. 1990. Allochthonous terrane processes in Southeast Asia. Philosophical Transactions of the Royal Society of London, Series A, 331, 625–640. M ETCALFE , I. 1991. Late Palaeozoic and Mesozoic palaeogeography of Southeast Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 211– 221. M ETCALFE , I. 1993. Southeast Asian terranes: Gondwanaland origins and evolution. In: F INDLAY , R. H., U NRUG , R., B ANKS , M. R. & V EEVERS , J. J. (eds) Gondwana 8—Assembly, Evolution, and Dispersal (Proceedings Eighth Gondwana Symposium, Hobart, 1991). Balkema, Rotterdam, 181 –200. M ETCALFE , I. 1996a. Pre-Cretaceous evolution of SE Asian terranes. In: H ALL , R. & B LUNDELL , D. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 97– 122. M ETCALFE , I. 1996b. Gondwanaland dispersion, Asian accretion and evolution of Eastern Tethys. Australian Journal of Earth Sciences, 43, 605– 623. M ETCALFE , I. 1998. Palaeozoic and Mesozoic geological evolution of the SE Asian region: multidisciplinary constraints and implications for biogeography. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 25– 41. M ETCALFE , I. 1999. Gondwana dispersion and Asian accretion: An overview. In: M ETCALFE , I. (ed.) Gondwana Dispersion and Asian Accretion, Final Results Volume for IGCP Project 321. Balkema, Rotterdam, 9 –28. M ETCALFE , I. 2000. The Bentong-Raub Suture Zone. Journal of Asian Earth Sciences, 18, 691–712. M ETCALFE , I. 2001. Palaeozoic and Mesozoic tectonic evolution and biogeography of SE Asia– Australasia. In: M ETCALFE , I., S MITH , J. M. B., M ORWOOD , M. & D AVIDSON , I. (eds) Faunal and Floral Migrations and Evolution in SE Asia–Australasia. Balkema, Lisse, 15– 34. M ETCALFE , I. 2002. Permian tectonic framework and palaeogeography of SE Asia. Journal of Asian Earth Sciences, 20, 551–566. M ETCALFE , I. 2005. Asia: South-East. In: S ELLEY , R. C., C OCKS , L. R. M. & P LIMER , I. R. (eds) Encyclopedia of Geology, vol. 1. Elsevier, Oxford, 169–198. M ETCALFE , I. 2006. Palaeozoic and Mesozoic tectonic evolution and palaeogeography of East Asian crustal fragments: The Korean Peninsula in context. Gondwana Research, 9, 24–46. M ETCALFE , I., S MITH , J. M. B., M ORWOOD , M. & D AVIDSON , I. (eds) 2001. Faunal and Floral Migrations and Evolution in SE Asia– Australasia. Balkema, Lisse. O O , T., H LAING , T. & H TAY , N. 2002. Permian of Myanmar. Journal of Asian Earth Sciences, 20, 683– 689. O TOFUJI , Y., M U , C. L. ET AL . 2007. Spatial gap between Lhasa and Qiangtang blocks inferred from Middle Jurassic to Cretaceous paleomagnetic data. Earth and Planetary Science Letters, 26, 581–593. S ASHIDA , K. & I GO , H. 1999. Occurrence and tectonic significance of Paleozoic and Mesozoic Radiolaria in

Thailand and Malaysia. In: M ETCALFE , I. (ed.) Gondwana dispersion and Asian accretion, Final Results Volume for IGCP Project 321. A. A. Balkema, Rotterdam, 175–196. S ASHIDA , K., A DACHI , S., I GO , H., K OIKE , T. & A MNAN , I. B. 1995. Middle and Late Permian radiolarians from the Semanggol Formation, Northwest Penisular Malaysia. Transactions and Proceedings of the Palaeontological Society of Japan, New Series, 177, 43– 58. S ASHIDA , K., I GO , H., A DACHI , S., U ENO , K., K AJIWARA , Y., N AKORNSRI , N. & S ARDSUD , A. 2000a. Late Permian to Middle Triassic radiolarian faunas from Northern Thailand. Journal of Paleontology, 74, 789–811. S ASHIDA , K., N AKORNSRI , N., U ENO , K. & S ARDSUD , A. 2000b. Carboniferous and Triassic radiolarian faunas from the Saba Yoi area, southernmost part of peninsular Thailand and their paleogeographic significance. Science Reports of the Institute of Geoscience, University of Tsukuba, Section B: Geological Sciences, 21, 71–99. S ENGO¨ R , A. M. C. 1979. Mid-Mesozoic closure of PermoTriassic Tethys and its implications. Nature, 279, 590–593. S ENGO¨ R , A. M. C. 1984. The Cimmeride orogenic system and the tectonics of Eurasia. Geological Society of America, Special Papers, 195. S HEN , S., D ONGLI , S. & S HI , G. R. 2003. A biogeographically mixed late Guadalupian (late Middle Permian) brachiopod fauna from an exotic limestone block at Xiukang in Lhaze county, Tibet. Journal of Asian Earth Sciences, 21, 1125– 1137. S HI , G. R. & A RCHBOLD , N. W. 1998. Permian marine biogeography of SE Asia. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 57–72. S MYTH , H. R., H AMILTON , P. J., H ALL , R. & K INNY , P. D. 2007. The deep crust beneath island arcs: Inherited zircons reveal a Gondwana continental fragment beneath East Java, Indonesia. Earth and Planetary Science Letters, 258, 269– 282. S ONE , M. & M ETCALFE , I. 2008. Parallel Tethyan sutures in mainland SE Asia: New insights for PalaeoTethys closure and implications for the Indosinian Orogeny. Comptes Rendus Ge´osciences, 340, 166–179. U ENO , K. 1999. Gondwana/Tethys divide in East Asia, solution from Late Paleozoic foraminiferal paleobiogeography. In: R ATANASTHIEN , B. & R IEB , S. L. (eds) Proceedings of the International Symposium on Shallow Tethys 5. Chiang Mai University, Chiang Mai, 45– 54. U ENO , K. 2003. The Permian fusulinoidean faunas of the Sibumasu and Baoshan blocks: Their implications for the paleogeographic and paleoclimatologic reconstruction of the Cimmerian Continent. Palaeogeography, Palaeoclimatology, Palaeoecology, 193, 1– 24. U ENO , K. & H ISADA , K. 1999. Closure of the PaleoTethys caused by the collision of Indochina and Sibumasu. Chikyu Monthly, 21, 832–839 [in Japanese].

TECTONIC AND PALAEOGEOGRAPHICAL EVOLUTION U ENO , K. & H ISADA , K. 2001. The Nan– Uttaradit– Sa Kaeo Suture as a main Paleo-Tethyan suture in Thailand: Is it real? Gondwana Research, 4, 804–806. U ENO , K., N ISHIKAWA , S., VAN W AVEREN , I. M. ET AL . 2006. Early Permian fusuline faunas of the Mengkarang and Palepat formations in the West Sumatra Block, Indonesia: their faunal characteristics, ages and geotectonic implications. Abstract Volume and Proceedings, 2nd International Symposium on Geological Anatomy of East and South Asia:

23

Paleogeography and Paleoenvironment in Eastern Tethys (IGCP 516), Quezon City, University of the Philippines, 98–102. Z HOU , J. B., W ILDE , S. A., Z HAO , G. C, Z HENG , C. Q., J IN , W., Z HANG , X. Z. & C HENG , H. 2008. Detrital zircon U-Pb dating of low-grade metamorphic rocks in the Sulu UHP belt: evidence for overthrusting of the North China Craton onto the South China Craton during continental subduction. Journal of the Geological Society, 165, 423–433.

The geological and palaeontological exploration of Laos; following in the footsteps of J. B. H. Counillon and A. Pavie J. SEBASTIEN STEYER CNRS, UMR 5143 ‘Pale´obiodiversite´ et Pale´oenvironnements’ and De´partement ‘Histoire de la Terre’, Muse´um national d’Histoire naturelle, CP 38, 57 rue Cuvier, 75231 Paris cedex 05, France (e-mail: [email protected]) Abstract: Historical studies on the exploration of SE Asia reveal interesting points concerning the geology and palaeontology of this part of the world: the first geological description of northern Laos was published in 1896 by the French geologist J. B. H. Counillon, a member of the famous Pavie Mission (1879– 1895). Although scientific notes are rare in the literature about the Pavie Mission, which dealt with diplomacy and politics, Counillon’s studies are very informative: they contain accurate observations on the geomorphology, hydrography, stratigraphy and palaeontology of the Mekong River and its banks. Counillon explored the dangerous region of Luang Prabang (northern Laos) with Captain Cupet. He discovered the first fossil tetrapods from Laos, and, with the help of Vasseur and Repelin, correctly referred them to dicynodont mammallike reptiles.

The explorer Auguste Pavie [1847 (Dinan, France) – 1925 (La Raimbaudie`re, France)] was sent by the French Telegraph Company in 1879 to explore SE Asia. Pavie gave his name to various French expeditions grouped under the title ‘Pavie Mission’ (1879–1895) and became General Commissioner of Laos. At that time, the French had been in possession of Cochinchina (southern Vietnam) since the 1860s and had expanded their colonial domain in Indochina to include Annam and Tonkin farther north. Cambodia had also fallen under their rule as a ‘protectorate’. The regions forming present-day Laos, including the Luang Prabang area, were semiautonomous principalities officially under the suzerainty of the king of Siam. The so-called ‘colonial party’ in France pushed for annexation of these territories (and possibly of Siam itself) to the French colonial empire in Indochina. Pavie’s expeditions to this region must be seen in this context, and they ultimately contributed to the takeover by France of the Lao principalities in 1893. Behind the political and commercial purposes of the French government, which were therefore to expand the borders of Indochina, Pavie’s personal motivation was also to explore this part of the world, its wild and wonderful nature, and to meet the various cultures and ethnic groups living there. In his book, equivocally titled The conquest of the hearts (A la conqueˆte des cœurs), Pavie wrote: ‘In permanent contact with the natives, I am trying to live fully among them’ (Pavie 1925). Pavie is considered mostly as a peaceful and humanistic explorer, the famous so-called ‘barefoot explorer’ or the ‘peaceful conqueror’ described by many historians or biographers (e.g. Vilbert 1996), even

though he instituted the blockade of Bangkok harbour against Siam in 1893, a decision that could have led to war between France and Britain. This ‘peaceful’ image of Pavie has been used to justified a posteriori the French colonization of Indochina (Larcher-Goscha 2005). The expeditions of the Pavie Mission were carried out by representatives of the French Telegraph Compagny (e.g. Biot, Controller of the Telegraph, in 1882–1883), civil servants and other state representatives (e.g. Gautier, French Consul, in 1887–1888); administrators, interpreters and secretaries (e.g. the Cambodian Ngin, Pavie’s faithful companion, from 1885 to 1895), numerous army officers (e.g. P. Cupet, a Captain at the 3rd Zouaves, from 1887 to 1892), and some scientists. Unfortunately, very little information about the scientists is available in the literature: the scientific notes, or the biographical notes taken by the scientists during the Pavie Mission are scarce in the literature, which mainly deals with diplomacy and politics rather than natural history. One can often read, however, that Pavie was collecting shells, plants and insects: ‘The collections of insects, shells, the search for ancient writings, the study of the local customs, photography . . . are my distractions’ (Pavie, quoted by Simon 1997, p. 113). Among the scientists, some naturalists or other researchers went to SE Asia within the framework of the Pavie Mission (Pavie 1904). This was true for Emile Lefevre, who studied the ethnography and anthropology of the Asian tribes for years, and for the geologist Jean-Baptiste-Henri Counillon. Counillon was born on 19 June 1860, at Agonges (not Orgonges, as given by Brebion 1935, p. 99), a

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 25–32. DOI: 10.1144/SP315.3 0305-8719/09/$15.00 # The Geological Society of London 2009.

26

J. S. STEYER

small village near Souvigny (Allier, France) where unfortunately none of his descendants has lived for several decades (according to personal research in situ by the author). As his birthplace is geologically located in the Carboniferous– Permian basin of Bourbon-l’Archambault (Steyer et al. 2000), it would be interesting to know whether the young Counillon worked in the field locally, before going to Marseilles and SE Asia, and whether he published observations or notes on the geology of his native region (biographical research is in progress by the author). Unfortunately, the archives of the Bulletin de la Socie´te´ d’Histoire Naturelle d’Autun contain no article by Counillon (Gand, pers. comm.). Before going to SE Asia, Counillon was Preparator at the Faculty of Science of Marseilles, but not in the Geological Department, according to the Archives of the Faculty of Sciences (Villier, pers. comm.). Counillon was sent to Indochina as a teacher at Chasseloup-Laubat High School, Saigon. Attached to the Pavie Mission from 1889, he joined the exploration team as a geologist at Luang Prabang (northern Laos), on 8 June 1890, and remained until 1892. However, despite his important role compared with the other members of the Pavie Mission, Counillon is rarely mentioned in the literature: Pavie (1919) mentioned several persons: the army officers Nicolon and Cupet often, Massie (‘Major Pharmacist’, p. 152, and ‘friend’, p. 238), or even the geographer McCarthy, his British counterpart whom he met in the field (e.g. pp. 210, 308– 310). Regarding Counillon, Pavie (1919) mentioned very briefly his arrival at Luang Prabang (footnote, p. 289), and described him as a ‘companion of Massie’ (p. 315). This absence of Counillon in the literature suggests a relatively self-effacing personality. This was confirmed by McCarthy (1900, who met Counillon but did not really see him) and Durand-Delga (1990, when Counillon was forced to retire; see below). Finally, Cupet (1900) mentioned Counillon as a ‘naturalist’ (p. 179) and wrote (p. 181): ‘I am leaving KhamMuoˆn at this date [26 February 1890], taking Mr Counillon with me, for whom a new country imposes an apprenticeship to a new life’. The geological exploration of Laos was beginning.

Counillon, a geologist in Laos (1890– 1892) When J. B. H. Counillon arrived in Laos in 1890, Pavie had already crossed the country several times and was famous at Luang Prabang for having diplomatically brought an end to the control of the city by ‘pirates’ (i.e. groups that had no sympathy for the French) during the famous ‘sack’ of the city in May 1887. In 1890 the French Vice-consulate at Luang Prabang had recently

been established, and Pavie had started his second period in SE Asia 4 years before, having been freshly appointed General Consul in Bangkok and Commissioner of the Republic for Border Affairs by the French Government. With these new titles and responsibilities, and probably thanks to his close links with freemasonry (Simon 1997), Pavie was able, from this date, to organize more important expeditions, and for a longer time than before. He chose Counillon and others (Fig. 1), to explore in detail the dangerous Province of Luang Prabang under the supervision of Cupet, Head of the Mission, who arrived in the region 2 years before Counillon (i.e. in 1888). (The first Frenchman to have reached Luang Prabang, and to have died there, was Henri Mouhot in July 1861.) Besides king cobras, tigers and fevers, this strategic region, crossed by the Mekong River, was considered as wild and dangerous for other reasons: it remained much disputed between Laos (under French influence), Siam (supported by the British) and China, and was the scene of frequent conflicts. The aim of this large exploration mission was therefore both commercial and strategic: (1) to map the region in detail so as to open and control new commercial French routes between the north and the south of SE Asia; (2) to counter James McCarthy, the British counterpart of Pavie and Counillon, who also mapped this region at the same time and for the same economic reasons, but for the British Empire. McCarthy was General Director of the Siamese Government Surveys and explored Siam from 1881 to 1893 for the Royal Geographical Society. The aim of the French Government was therefore to compete with the British Empire for the colonization of this part of the world. Interestingly, McCarthy and Pavie themselves became good friends once they returned to Europe. In his book Surveying and Exploring in Siam, McCarthy described his meeting with the scientists of the Pavie Mission: ‘We called on Dr Massie and M. Cavillion [very probably Counillon], and found them very interesting. Dr Massie was an enthusiastic geologist’ (McCarthy 1900, p. 174). According to McCarthy, the main French geologist working in the Luang Prabang region was Massie (Fig. 2) rather than Counillon. This point of view is surprising, because Massie, as official Head Pharmacist of the Pavie Mission from 1882 to 1892, was involved in linguistics rather than in geology according to his publications (e.g. Massie 1894). However, McCarthy (1900, p. 177) wrote: ‘Dr Massie’s manner was brusque, but he had a kindly disposition. In his geological researches on Pu Sai, near Luang Prabang, he found coal, and said, besides, he was on the track of a salt-field. His pockets were always full of fossils of the reptile and fish period. . . . But he unnecessarily

THE EXPLORATION OF LAOS BY COUNILLON

27

Fig. 1. The French geologist J. B. H. Counillon (right, in white, in profile) during an expedition in northern Laos (1889– 1891) with other members of the Pavie Mission, including P. Cupet, Captain at the 3rd Zouave (background, in the middle, at the table). From Cupet (1900, fig. 2, p. 15).

exposed himself to danger, and in his geological excursions often got soaking wet, without taking the slightest precautions to avoid fever’. Interestingly, McCarthy mentioned here fossil reptiles found by Massie and very probably described by Counillon (see below). Massie mysteriously committed suicide after 4 years in SE Asia, in 1892.

According to McCarthy (1900, p. 178), the careless Massie was overtaken by strong fevers in descending a river near Nakawn Panom, Thailand, on his way back to Saigon. Did the strong fevers push Massie to become delirious and to commit suicide? The task of Counillon, helped by Massie, was to map the geology of the region and to find as many

28

J. S. STEYER

Fig. 2. Massie (in white, in the middle), Head Pharmacist of the Pavie Mission and active geologist in the Luang Prabang region according to McCarthy (1900). From Pavie (1919, fig. 45, p. 127).

mineral resources as possible. For this exploration mission, his companions came from various backgrounds, but most of them were army officers, such as Cupet (1888–1992), Cogniard (1889– 1891) from the Foreign Legion, Rivie`re (1889– 1891, 1894, 1895), and Captain de Malglaive (1889–1891) from the Marine Infantry. Captain Cupet was one of the leaders of the expedition, after Pavie of course (the latter met the team several times at Luang Prabang or sent them orders). Some political agents such as M. Lefe`vrePontalis (1889–1891) or M. Lugan (1889–1895) from the French Embassy of (very probably) Hanoi, as well as M. de Coulgeans, telegraph principal agent (1890–1895), also belonged to the team (Fig. 1). Finally, a few naturalists joined Pavie in the exploration of Laos: Le Dantec, an embryologist from the Sorbonne in Paris, stayed until 1890 only. Counillon mapped the region either alone or with the exploration team. Cupet (1900, pp. 183– 186) mentioned his prospections of mineral resources between the Kham-muoˆn and the Mekong rivers. Pavie took along a total of 15 tonnes of goods for this mission (which was a lot compared with his prior exploration missions), and the progress of the team was very slow: the subtropical climate, the jungle, possible encounters with ‘pirates’

(i.e. Siamese or Chinese ‘irregulars’), the mountain peaks to be climbed, the streams to be crossed (they navigated by pirogue or raft along the Mekong River and its tributaries), and various diseases made travel conditions difficult in northern Laos (the so-called ‘Upper Laos’ of that time). Counillon became sick and had to stay for several days in convalescence at the Catholic Mission of Kham-Kheum, near Lakhoˆn, and de Malglaive also had to stop temporarily because of foot wounds caused by leech bites (Cupet 1900, p. 194 and p. 200, respectively). De Malglaive wrote about Counillon: ‘At B. Na Hat, I meet M. Counillon who came by foot from Tourakom. He got a fever and wants to come back to the River. We share our drugs and I leave him to gain time for tomorrow’s trip’ (De Malglaive & Rivie`re 1902, p. 52). Cupet also suffered from a high fever, but he became delirious, so that his companions started to look for a place in the jungle to bury him (Cupet 1900, p. 211). From 1888 to 1892 (the time period during which Cupet was Head of the Pavie Mission), 9000 km of routes were mapped in Laos. From a strategic and political point of view, the French Government, which could thereafter start to instal outposts between northern and southern SE Asia, was now ready to complete the colonization of Indochina. It is interesting here to note that, when Counillon placed

THE EXPLORATION OF LAOS BY COUNILLON

Luang Prabang in Cochinchina in the title of his report on the geology of the area (Counillon 1896), he was taking liberties with geography, as northern Laos can definitely not be considered as part of the Mekong delta region, which is Cochinchina proper. His motivation may have been to emphasize that Luang Prabang (and the rest of Laos) were now under French administration and governed from Saigon. From a scientific point of view, all these explorations made on foot and by boat provided Europeans with better knowledge of this terra incognita: hundreds of Asian ethnic groups had been met and described, as well as hundreds of new plant or animal species. As to the Earth Sciences, hundreds of maps had been drawn and fieldbooks had been filled concerning the geography, geology, hydrography and other characteristics of the region.

The work of Counillon and the Luang Prabang Formation After his exploration missions and fieldwork, Counillon published several works on the mineral

29

resources (Counillon 1899a, b, 1900a– c, 1907) and palaeobotany (Counillon 1914) of SE Asia. He also published, for the first time, palaeontological observations on Laos and notes on the stratigraphy and sedimentology of the country: in 1896, Counillon published a description of the Luang Prabang Formation, in which he mentioned fossilized invertebrates (e.g. foraminifers, Spirifer, etc.) and in which he reported, for the first time, the discovery of plants and tetrapods in continental sedimentological layers. ‘This publication is the result of observations made in collaboration with Mr Massie, during our stay at Luang Prabang, as members of the Pavie Mission’ (Counillon 1896, p. 1330). Counillon (1896) described and subdivided the thick Luang Prabang Formation into five Zones (or Units, Fig. 3). He also mentioned large (indeterminate) reptiles in the (second) ‘Zone of the red claystones’, and mammal-like reptiles in the (fourth) ‘Zone of the purple claystones’. This fourth Zone is composed of red siltstones and purple claystones of several metres thickness alternating with yellow sandstones of

Fig. 3. Geological sketch of the Luang Prabang area, northern Laos, from Counillon (1896, p. 1331), showing the lithological Zones (or Units) of the Luang Prabang Formation, Upper Permian of northern Laos.

30

J. S. STEYER

several metres thickness (author’s personal observation). Counillon studied the mammal-like reptiles found in the (fourth) ‘Zone of the purple claystones’: with the help of Gaston Vasseur and Joseph Repelin (Faculty of Science, Marseilles), Counillon (1896) correctly assigned the anterior portion of a skull to a dicynodont. (Interestingly, Counillon had retained contact with his colleagues from Marseilles, G. Vasseur who occupied from 1886 to 1915 the Chair of Geology and Mineralogy at the Faculty of Science and his Preparator/Assistant J. Repelin who replaced Vasseur after World War I.) The dicynodont skull portion found by Counillon and Massie in Laos was briefly described by Repelin (1923), who assigned it to a new species, Dicynodon incisivum. Although Woodward (1932) considered this specimen as belonging probably to the genus Lystrosaurus, Piveteau (1937a, b, 1938) and Battail et al. (1995), after closer examination, confirmed its initial identification as Dicynodon. This debated specimen was unfortunately considered as lost by Battail et al. (1995), but the author is still looking for it in the historical collections of the Faculty of Science of Marseilles with the help of Loı¨c Viller, and in the historical collections of the Natural History Museum of Marseilles. At the time when Counillon was writing (1896), all known dicynodont reptiles were from rocks considered as Early Triassic in age (mainly in India, Scotland and South Africa) and therefore Counillon supposed that the dicynodont-bearing rocks at Luang Prabang also belonged to the Triassic; the age of this ‘Zone of the purple claystones’ has recently been corrected as Tatarian, Late Permian by Battail et al. (1995). However, Counillon did not map only the Luang Prabang area (e.g. Counillon 1903, 1909). For his explorations and geological fieldwork in SE Asia, particularly in Annam, he was appointed, once he returned to Hanoi after the Mission, ‘Second President of the Society of Indochinese Studies’ in 1898, and ‘Head of the Geological Survey of Indochina’ from 1902. He became Engineer of Public Works on 24 April 1907, and Principal Controller (Third Class) of Mines. Counillon never returned to France; he stayed in Hanoi, married a Vietnamese woman, and became more and more sick and isolated, while Pavie collected all the national honours in France (from the Socie´te´ Ge´ographique de Paris, the Socie´te´ de Ge´ographie commerciale, the French Government and, of course, from freemasonry). Durand-Delga (1990, p. 125) wrote about Counillon: ‘this self-effacing man, who had a permanent tropical anaemia and a dysentery . . . squats in his house’. Pushed by the powerful Honore´ Lantenois to vacate the position of Head of the Geological Survey of Indochina, just before

Jacques Deprat’s arrival in Hanoi and before the ‘Deprat Affair’, Counillon was forced to retire on 10 June 1910 (Durand-Delga 1990; Osborne 2000). He was therefore one of the first victims of the ‘Deprat Affair’. All we know of Counillon at the end of his career, and after his retirement, is through Deprat’s autobiographical roman a` clef Les Chiens Aboient (Wild 1926– 7; McBirney & Janvier 2005). Deprat became Counillon’s successor at the Geological Survey of Indochina. Soon after his arrival at Hanoi, he was told by both Lantenois and H. Mansuy that Counillon was incompetent and therefore had been asked to take early retirement. Deprat made a brief visit to Counillon, who warned him against Lantenois and Mansuy. However, in 1917, after Deprat had been accused by Mansuy of having salted Indochinese outcrops with European trilobites, Counillon told him that his private trilobite collection, which had strangely disappeared after his retirement, had probably been used for this forgery, by Mansuy himself. We will probably never know the truth. According to Deprat, Counillon lived in a very small house in the ‘native quarter’ of Hanoi. After Deprat’s return to France and after the trial that put an end to his career in geology, Counillon was murdered by a thief at his home (see Les Chiens Aboient). At the end of his novel, and to make Lantenois and Mansuy appear even more evil, Deprat alluded to the fact that Counillon’s murderer may not have been an ordinary thief. According to Henri Fontaine (pers. comm.), Counillon indeed died at his home, but was poisoned by his own cook, who probably prepared for him a delicious mixed salad containing Gelsemium elegans, a very toxic plant, well known in Vietnam as Caˆy la´ ngo´n (N. Thu, pers. comm.). Counillon was buried in the small colonial cemetery of Hanoi not far from the Service Ge´ologique d’Indochine (this cemetery no longer exists; the remains of the French colonists have been officially removed to Europe). Counillon may have been the only one who knew everything about the trilobites of the ‘Deprat Affair’. He died in 1923 at the age of 63 years, leaving behind him proofs of honesty and professionalism; his works on the geology and palaeontology of SE Asia are among the first, and they remain exemplary. Many later geologists and palaeontologists, such as Hoffet (1935), discoverer of the first dinosaurs of Laos (Taquet 1994), and Saurin (1962) used the works of Counillon to draw geological maps of SE Asia. History therefore shows that, during his career, Counillon did not receive the honours he actually deserved. Almost a century after the Pavie Mission, during the 1990s, and within the framework of a French– Lao scientific convention, palaeontologists from

THE EXPLORATION OF LAOS BY COUNILLON

31

Fig. 4. Three-dimensional reconstruction of Dicynodon from the Upper Permian of Laos. Courtesy of M. Boulay and S. Lorrain, Hox Studio, La Roque and Paris (World wide web address: www.hox.fr).

the Muse´um National d’Histoire Naturelle (Paris; MNHN), headed by P. Taquet and supported by the Fondation de France, rediscovered the Permian fossil localities described by Counillon: they discovered in the Luang Prabang Formation numerous cranial and postcranial remains of dicynodonts, and fragmentary and rare remains of amphibians (Battail et al. 1995). More recently, in 2005, an international team, including Lao participants and the author of the present paper, also followed in Counillon’s footsteps and discovered, in the Upper Permian deposits of Luang Prabang, well-preserved dicynodont remains and an exquisitely preserved skull of a new, carnivorous tetrapod (description in progress by Steyer et al.). These newly discovered specimens have been prepared by R. Vacant (CNRS Paris), B. Le Dimet and P. Richir (MNHN). A 3D reconstruction of a Dicynodon from the Permian of Laos (Fig. 4) has already been made by the 3D modellers M. Boulay and S. Lorrain (members of the 2005 field expedition), under scientific control. The rest of the material will also provide very interesting data on the Permian continental faunas of Asia, and has implications for the palaeobiogeography and palaeoclimates of Pangaea at the end of the Palaeozoic. I sincerely thank P. Taquet (MNHN, Paris) for his invitation to Laos in 1999, and the Committee of Research and Exploration of the National Geographic Society for a Grant (No. 7655-04), which allowed me to organize an expedition to Laos in 2005. Many thanks go to

M. Boulay and S. Lorrain (Hox Studio, Paris), R. Damiani (Staatliches Museum fu¨r Naturkunde Stuttgart), C. Sidor (University of Washington, Seattle), J.-P. Cuomo (Naturalia Moulage, Luang Prabang), M. Ve´ran, P. Richir and B. Battail (MNHN, Paris), J.-P. Porte, S. Vongphamany (Department of Museums and Archaeology, Vientiane), B. Khentavong (Muse´e des Dinosaures, Savannaketh), and B. Phanmongkhoune (Culture Office, Luang Prabang) for their help in the field. I thank the Minister of Information and Culture of Lao PDR for his field authorization. I also acknowledge the assistance of L. Vachier (Centre des Archives d’OutreMer, Aix-en-Provence), F. Engelmann (Luang Prabang), P. Janvier (CNRS, Paris), C. Pereira, P. Richir, M. Ve´ran (MNHN, Paris) and G. Gand (University of Burgundy) for their help with literature, and P. Debriette (Charbonnages de France) for his help in finding the birthplace of Counillon. I also thank L. Villier (Universite´ de Provence, Marseilles) for his help in the research on the historical Dicynodon specimen from Laos, P. Janvier (CNRS, Paris) and E. Buffetaut (CNRS, Paris) for their valuable contributions. Many thanks go to E. Buffetaut for his invitation to contribute a paper to this Special Publication, and for reviewing, with R. Osborne, a first draft.

References B ATTAIL , B., D EJAX , J., R ICHIR , P., T AQUET , P. & V ERAN , M. 1995. New data on the continental Upper Permian in the area of Luang-Prabang, Laos. Journal of Geology, Series B, 5– 6, 11– 15. B REBION , A. 1935. Dictionnaire de bio-bibliographie ge´ne´rale ancienne et moderne de l’Indochine franc¸aise. Acade´mie des Sciences coloniales, Paris.

32

J. S. STEYER

C OUNILLON , J. B. H. 1896. Documents pouvant servir a` l’e´tude ge´ologique des environs de Luang Prabang (Cochinchine). Comptes Rendus de l’Acade´mie des Sciences, 123, 1330–1333. C OUNILLON , J. B. H. 1899a. Notes sur les gisements de gypse dans l’arrondissement de Taˆnan. Bulletin Economique de l’Indochine, 14, 453–455. C OUNILLON , J. B. H. 1899b. Les mines de Quangnam. Bulletin Economique de l’Indochine, 16, 575–579. C OUNILLON , J. B. H. 1900a. Les mines de charbon de Nong-Son (Annam). Bulletin Economique de l’Indochine, 19, 5–14. C OUNILLON , J. B. H. 1900b. Note sur une collection offerte par M. Beauverie, Ing. des mines au Tonkin. Bulletin Economique de l’Indochine, 24, 303– 306. C OUNILLON , J. B. H. 1900c. Les gisements aurife`res du Haut-Laos. Bulletin Economique de l’Indochine, 27, 459– 469. C OUNILLON , J. B. H. 1903. Note sur la ge´ologie de la re´gion de Noˆngson, Annam. Bulletin Economique de l’Indochine, 19, 478–490. C OUNILLON , J. B. H. 1907. Re´sultats de la Mission ge´ologique et minie`re du Yunnan me´ridional septembre 1903 janvier 1904. Note sur la ge´ologie de la re´gion de Po Si, Lou Nan, Mi Leu, Tou Tza, A Mi Tche´ou. Annales des Mines, 8, 1 –20. C OUNILLON , J. B. H. 1909. Sur le gisement liasique de Huu-Nien, Province de Quang-Nam (Annam). Bulletin de la Socie´te´ Ge´ologique de France, 8, 524– 532. C OUNILLON , J. B. H. 1914. Flore fossile des gıˆtes de charbon de l’Annam. Bulletin du Service Ge´ologique de l’Indochine, 1, 1 –36. C UPET , P. 1900. Voyages au Laos et chez les sauvages du Sud-Est de l’Indo-Chine. Mission Pavie—Indo-Chine 1879–1895—Ge´ographie et Voyages, tome III. E. Leroux, Paris. D E M ALGLAIVE , J. & R IVIE` RE , A.-J. 1902. Voyages au centre de l’Annam et du Laos et dans les re´gions sauvages de l’Est de l’Indo-Chine. Mission Pavie— Indo-Chine 1879–1895, tome IV. E. Leroux, Paris. D URAND -D ELGA , M. 1990. L’affaire Deprat. Travaux du Comite´ Franc¸ais d’Histoire de la Ge´ologie, 4, 117– 212. H OFFET , J.-H. 1935. Carte ge´ologique de l’Indochine au 500 000e`me: Feuille de Tourane. N.12, avec notice explicative, d’apre`s les travaux de MM. H. Counillon, R. Bourret et J.-H. Hoffet. Publication Service Ge´ologique d’Indochine. L ARCHER -G OSCHA , A. 2005. Auguste Pavie et la conqueˆte pacifique du Laos dans les e´crits coloniaux franc¸ais entre les deux guerres. In: Le Pays de Dinan (eds) Rencontres Auguste Pavie—Cambodge Laos Vieˆt Nam. Bibliothe`que de la Ville de Dinan, Dinan, 129– 137. M ASSIE , M. 1894. Dictionnaire laotien. Mission Pavie—Exploration de l’Indo-Chine—Me´moires

et Documents, vol. II, Litte´rature et Linguistique. E. Leroux, Paris. M C B IRNEY , A. & J ANVIER , P. 2005. The Trilobite Affair. Bostok Press, Corvallis, OR. M C C ARTHY , J. 1900. Surveying and Exploring in Siam. J. Murray, London. O SBORNE , R. 2000. The Deprat Affair. Ambition, Revenge and Deceit in French Indo-China. Pimlico, London. P AVIE , A. 1904. Mission Pavie—Indo-Chine 1879–1895, tome III, Recherches sur l’Histoire naturelle de l’IndoChine orientale. E. Leroux, Paris. P AVIE , A. 1919. Mission Pavie—Indo-Chine 1879– 1895—Ge´ographie et Voyages, tome VII, Journal de marche (1888– 1889)—Eve`nements du Siam (1891– 1893). E. Leroux, Paris. P AVIE , A. 1925. A la conqueˆte des cœurs. Le Pays des Millions d’Elephants et du Parasol Blanc. Bossard, Paris. P IVETEAU , J. 1937a. Note pre´liminaire sur un reptile the´rapside´ d’Indochine et sa signification pale´oge´ographique. Comptes Rendus de la Socie´te´ Ge´ologique de France, 6, 70–72. P IVETEAU , J. 1937b. Un reptile dicynodonte d’Indochine. Les reptiles the´romorphes et la notion de continent de Gondwana. Annales de la Socie´te´ Ge´ologique du Nord, 62, 122– 127. P IVETEAU , J. 1938. Un The´rapside´ d’Indochine. Remarques sur la notion de continent de Gondwana. Annales de Pale´ontologie, 27, 139– 152. R EPELIN , J. 1923. Sur un fragment de craˆne de Dicynodon recueilli par H. Counillon dans les environs de Luang Prabang (Haut-Laos). Bulletin du Service Ge´ologique d’Indochine, 12, 1 –7. S AURIN , E. 1962. Luang Prabang Est. Carte ge´ologique Viet Nam, Cambodge, Laos 1:500 000. Service Ge´ographique National, Dalat, Vietnam. S IMON , H. 1997. Auguste Pavie, explorateur en Indochine. Collection ‘Des Hommes de l’Ouest’. Ouest France, Rennes. S TEYER , J.-S., E SCUILLIE´ , F., P OUILLON , J. ET AL . 2000. New data on the flora and fauna from the Lower Permian of the Bourbon-l’Archambault basin (Buxie`res-les-Mines, Allier, France). Bulletin de la Socie´te´ Ge´ologique de France, 171, 239– 249. T AQUET , P. 1994. Chercheur d’os au Laos. In: Commune d’Oberhausbergen (eds) Josue´-H. Hoffet, d’Oberhausbergen au Laos. Commune d’Oberhausbergen, Oberhausbergen, 49– 61. V ILBERT , L.-R. 1996. Auguste Pavie, le conque´rant aux mains nues. Le Me´morial des Bretons, T.V., 1870– 1940. Breiz, Rennes. W ILD , H. (for Deprat J.) 1926–1927. Les chiens aboient: Roman de mœurs contemporaines. Albin Michel, Paris. W OODWARD , S. A. 1932. Dicynodontidae. In: VON Z ITTEL , K. A. (ed.) Text-book of Palaeontology, Vol. II. Macmillan, London, 257– 260.

Late Permian dicynodont fauna from Laos BERNARD BATTAIL Muse´um National d’Histoire Naturelle, De´partement Histoire de la Terre, USM 0203—UMR, 5143 du CNRS Pale´obiodiversite´ et Pale´oenvironnements, C.P. 38, 57 rue Cuvier, 75231 Paris cedex 05, France (e-mail: [email protected]) Abstract: In Laos, dicynodonts have long been known only from one specimen, now lost, a partial skull discovered by Counillon in the purple beds of the area of Luang Prabang, and initially described by Repelin as Dicynodon incisivum. Subsequent researchers attributed the specimen either to the Late Permian genus Dicynodon or to the Early Triassic genus Lystrosaurus. Recent Franco-Laotian expeditions have gathered, from the same purple beds, a collection of tetrapods composed mainly of dicynodonts. They can all be ascribed to Dicynodon, and the available evidence suggests that the purple beds are Late Permian in age. The genus Lystrosaurus remains unknown in Laos.

The presence of dicynodonts north of Luang Prabang, Laos, was first reported almost one century ago by Counillon (1896). However, it is only fairly recently that a relatively important collection of dicynodont remains could be gathered from that area, during annual Franco-Laotian expeditions led by P. Taquet (Muse´um National d’Histoire Naturelle, Paris, France) between 1993 and 2003. This paper is only a preliminary account of the dicynodont fauna of Laos, as the specimens are still currently under study.

Counillon’s dicynodonts In a short note on the geology of the area of Luang Prabang, Laos, Counillon (1896) briefly described two distinct sedimentary units, which he named ‘zone des argiles rouges’(red clays) and ‘zone des argiles violettes’ (purple clays), the latter being stratigraphically higher than the former. In fact, they are not composed of clays, in the strict sense of the word, but of various clastic sediments. Therefore, they will be designated hereafter as ‘red beds’ and ‘purple beds’. Counillon had noticed the presence, in the red beds, of a few scattered remains of ‘large reptiles’. Our recent observations confirm that the red beds contain rare, disarticulated and often broken elements belonging to large tetrapods. None of those elements could be determined very accurately, but most of them can be attributed to large dicynodonts. From the purple beds, which he attributed to the Early Triassic, Counillon mentioned the discovery of an incomplete dicynodont skull, which consisted primarily of the snout and orbital region. Counillon’s specimen was first studied by Repelin (1923), who described it as a new species of the

genus Dicynodon, D. incisivum. Repelin, however, considered D. incisivum as resembling D. orientalis, a species from the Panchet Formation of India, which had been described by Huxley in 1865. Later, D. orientalis was transferred to the genus Lystrosaurus (Das Gupta 1922), and it was eventually considered as a junior synonym of L. murrayi (Tripathi & Satsangi 1963; King 1988). Without discussing the anatomical characters of Counillon’s specimen, Woodward (1932, p. 259) chose to attribute it to Lystrosaurus (or to a genus closely related to the latter). He may have been influenced either by its supposed Early Triassic age, or by its alleged resemblance to ‘Lystrosaurus orientalis’. Yuan & Young (1934) referred the Laos fossil to Lystrosaurus, basing their opinion on the supposed high position of the nares and on the breadth of the snout. Piveteau (1938) redescribed the specimen, and concluded that it did belong to Dicynodon. Moreover, he suggested that it could be attributed to the species D. lacerticeps, known from South Africa. Since then, the specimen has unfortunately been lost, and attempts to determine its systematic position can be made now only on the basis of bibliographical data. In a paper on the distribution of Lystrosaurus in Pangaea, Colbert (1982) mentioned the specimen from Laos as a possible Lystrosaurus, but he stressed that ‘the status of this fossil is very much in doubt. The illustrations accompanying the original description leave much to be desired: indeed it is difficult to interpret them with any degree of confidence’ (Colbert 1982, p. 377). Many recent workers, however, have referred, without comment, the fossil from Laos to Lystrosaurus (Keyser & Cruickshank 1979; King 1988). As already noted by Colbert (1982), very little can be inferred from Repelin’s description, which

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 33–40. DOI: 10.1144/SP315.4 0305-8719/09/$15.00 # The Geological Society of London 2009.

34

B. BATTAIL

is too brief and accompanied by poor, inadequate illustrations. However, the paper by Piveteau (1938) deserves more attention. The text is a slightly more detailed account of the anatomy of the specimen, and the figures, although composed only of simple sketches, provide, I think, interesting data, which were discussed in some detail by Battail (1997). The comparative analysis of the skull structure of Dicynodon and Lystrosaurus shows major differences regarding the shape of the snout and the position of the nostril. The snout is high and slopes steeply downwards in Lystrosaurus, whereas it is shallow in Dicynodon. The nostril is closer to the inferior border of the snout than to the orbit in Dicynodon. In Lystrosaurus, the nostril is displaced dorsally and is situated at a short distance in front of the orbit; it is closer to the orbit than to the lower border of the snout (Fig. 1). If we orient the figure given by Piveteau (1938) of the lateral view of the incomplete skull from Laos with the canine in vertical position, the snout looks high and is therefore slightly evocative of that of Lystrosaurus. However, with that orientation, the basicranial axis of the specimen is abnormally steep, and the nostril is much lower than the orbit instead of being situated in front of the latter; a character that would indeed be observed if the specimen belonged to Lystrosaurus. Let us consider now Piveteau’s figure as oriented in the publication; that is, with the dorsal borders of the nostril and of the orbit on the same horizontal line (Fig. 2). If we compare that figure with that of the homologous portion of an undistorted Dicynodon skull similarly oriented, we observe very few differences between the two: the specimen from Laos is more flattened dorsoventrally and its canine slopes more forwards,

two features that are not necessarily natural, but could also be the result of post-mortem distortion. From the above discussion, several conclusions can be drawn: (1) the specimen of dicynodont discovered in 1896 by Counillon in Laos does not belong to the genus Lystrosaurus; (2) none of its known characters is incompatible with the anatomy of Dicynodon; however, because the existing descriptions of D. incisivum provide little useful anatomical information and the only known specimen has been lost, D. incisivum must be considered as a nomen dubium; (3) there is no reason to consider the specimen as Early Triassic, rather than Late Permian, in age.

Recent findings of dicynodonts in Laos Between 1993 and 2003, a good collection of dicynodont remains from the Luang Prabang area has been gathered. Only a few very fragmentary specimens have been collected from the red beds, all of them belonging to large individuals. The purple beds are much more fossiliferous and have yielded more complete and better preserved material, and skulls and cranial fragments are more numerous than postcranial elements. The material collected by the Franco-Laotian expeditions is temporarily kept, for preparation and study, in the palaeontology unit of the Muse´um National d’Histoire Naturelle, Paris. It will eventually be returned to Laos, and will be housed in the Palaeontology Museum of Savannakhet. All the skulls and skull fragments found in the purple beds correspond to mediumsized specimens: the skull length, measured from the tip of the snout to the posterior edge of the occipital condyle, hardly exceeds 15 cm in the largest individuals. The state of preservation of the fossils

Fig. 1. Skull shapes of Lystrosaurus and Dicynodon compared, both in lateral view. (a) Lystrosaurus, simplified after Cluver (1971). (b) Dicynodon, simplified after Cluver & Hotton (1981).

DICYNODONTS FROM LAOS

Fig. 2. Reconstruction of the anterior part of the dicynodont skull discovered by Counillon (1896) in the area of Luang Prabang, Laos. After Piveteau (1938, fig. 1). c, canine; fr, frontal; mx, maxillary; n, nostril; na, nasal; or, orbit; pal, palatine; pmx, premaxillary; pt, pterygoid; smx, septomaxillary.

is variable; they are usually distorted, and most sutures cannot be clearly seen. The dicynodont skulls from the purple beds conform, in their observable characters, with the definition of Dicynodon as given by Cluver & Hotton (1981, pp. 106 –107), or, in a somewhat more detailed manner, by Cluver & King (1983): ‘Medium-sized to large dicynodonts (average skull length 100 mm to over 400 mm), single pair of maxillary tusks in upper jaw, lower jaw edentulous. Postorbitals tend to cover parietals behind pineal foramen. Septomaxilla merges smoothly with outer surface of snout, does not meet lacrimal. Low boss formed over external nares by nasals.

35

Palatal rim sharp edged, uninterrupted by notch. Palatal portion of palatine large and flat, marking short contact with premaxilla. Vomers form long, narrow septum in interpterygoidal fossa. Anterior border of interpterygoidal fossa formed by a crest that joins the vomerine septum. Ectopterygoid small, displaced laterally. Labial fossa present between maxilla, palatine and jugal. Pterygoid makes short contact with maxilla. Basioccipital tubera separated by intertuberal ridge. Fused dentaries carry narrow dentary tables. Dorsal edge of dentary carries deep sulcus behind dentary tables. Rear of dentary extended dorsally to form weak posterodorsally directed process. Mandibular fenestra large, bounded dorsally by lateral dentary shelf. Occipital surface of opisthotic carries depression above paroccipital process’ (Cluver & King 1983, pp. 234, 237). Since that time, more recent studies on late Permian dicynodonts, undertaken in a cladistic perspective, have dealt with many more characters (see, e.g. Angielczyk 2001; Maisch 2002; Angielczyk & Kurkin 2003b; Fro¨bisch 2007). A detailed analysis of the characters listed by these recent workers helps to distinguish the dicynodonts from the purple beds of Laos from other forms whose dentition is reduced to a pair of upper tusks. The skulls of the dicynodonts from Laos have the overall shape of Dicynodon skulls, with relatively shallow snouts (Fig. 3), a feature that differs sharply from the one observed in Lystrosaurus, which has a high snout sloping steeply downwards (Fig. 1). In Lystrosaurus, the parietals are widely

Fig. 3. Dicynodont skull from Laos, specimen LPB (Luang Prabang) No. 1993– 2, lateral view. Scale bar 5cm.

36

B. BATTAIL

exposed on the skull roof. In the dicynodonts from Laos, the postorbitals slope ventrolaterally and overlap the parietals nearly completely behind the pineal foramen (Fig. 4), as in Dicynodon lacerticeps. In Lystrosaurus, Aulacephalodon and Idelesaurus, there is a median ridge on the anterior surface of the snout; such a ridge is absent in the described species of Dicynodon, as well as in the specimens from Laos. The dicynodonts from Laos have a much narrower skull than Aulacephalodon and Idelesaurus (Figs 4 and 5). Unlike Aulacephalodon, their pineal foramen is not surrounded by a boss. Unlike Idelesaurus, they have no postcaniniform crest. The nasal bosses of the dicynodonts from Laos, rather weakly developed, have a continuous posterior margin, as in Dicynodon, Delectosaurus and some species of Lystrosaurus. In Elph and Interpresosaurus, there are no nasal bosses. In Vivaxosaurus, the nasal bosses are paired swellings that meet in the midline to form a swollen anterodorsal surface of the snout. The labial fossa, present in the dicynodonts from Laos as well as in Dicynodon and many other forms, is absent in Elph and Interpresosaurus. Anatomical differences between Dicynodon and Delectosaurus are rather inconspicuous, and one can even question the validity of the latter genus. According to Kurkin (2001), in Dicynodon, the palatine is in contact anteriorly and medially

Fig. 4. Dicynodont skull from Laos, specimen LPB No. 1995– 9, dorsal view. Scale bar 5cm.

Fig. 5. Dicynodont skull from Laos, specimen LPB No. 1993–3, ventral view. Scale bar 5 cm.

with the premaxillary, whereas in Delectosaurus the palatine has only a very short anterior contact with the premaxillary. Unfortunately, among the specimens from Laos, the sutures remain unclear even when the palate is well preserved (Fig. 5). Another difference between Delectosaurus and Dicynodon seems to be the shape of the snout, which tapers forward more in Delectosaurus than in Dicynodon. The snouts of the dicynodonts from Laos are clearly of the Dicynodon type (Figs 4 and 5). To sum up, it appears that the dicynodonts from the purple beds of Laos can indeed be distinguished from Lystrosaurus, Aulacephalodon, Idelesaurus, Interpresosaurus, Elph and Delectosaurus, but not from Dicynodon. But is Dicynodon a monophyletic or a paraphyletic genus? Indeed, as Dicynodon is the only unspecialized dicynodont genus represented at the very end of the Permian, it is tempting to believe that it could have included the stem of all subsequent, Triassic dicynodonts. Angielczyk & Kurkin (2003a, b) published the results of a phylogenetic analysis of dicynodonts, in which three species of Dicynodon, the two Russian species D. amalitzkii and D. trautscholdi, and the South African species D. lacerticeps, have been treated as separate entities. These results suggest that those three species do not form a clade to the exclusion of all other dicynodonts, whereas the alternative hypothesis of a monophyletic Dicynodon is more weakly supported.

DICYNODONTS FROM LAOS

Comparing the dicynodont material from Laos with the already described species of the genus Dicynodon is not an easy task. The genus was erected by Owen (1845), with Dicynodon lacerticeps, from the Late Permian of South Africa, as type species. Owing to the fact that Dicynodon lacks obvious specialized characters, and taking into account the erection of the genus at an early date, it is not surprising that many species have been attributed to it. Some of those species were ultimately transferred to other genera, mainly to Diictodon Broom 1913 and Oudenodon Owen 1860 (see Keyser 1975; Cluver & Hotton 1981). In contrast, a few species were initially attributed to genera (Daptocephalus, for example) that are usually regarded nowadays as synonyms of Dicynodon. In the list established by King (1988, pp. 89– 93), 59 Dicynodon species were still found, but King noted that they are not necessary all valid. The main problems encountered while studying the Dicynodon specimens from the purple beds of Luang Prabang can be summarized as follows. (1) The specimens differ in size and proportions. Do those differences reflect the existence of distinct species, or are they attributable to intraspecific variability, to different individual ages, to sexual dimorphism, to distortion during the fossilization process, or to several of those factors combined? (2) Do the Dicynodon specimens from Laos belong to one or several new species, or can they be attributed, partly or totally, to already described species? The answers to those two sets of questions are not obvious, for the following reasons. (1) The specimens are not numerous enough to allow growth series to be established. (2) The modalities of a possible sexual dimorphism are not well known. By comparison with what is known in many tetrapods, we can, however, admit

37

that sexual dimorphism can relate to the size and robustness of the skull (thickness and relief of the bones), that it can affect one character (the development of the canines, for example), but that it does not show itself in major differences in the shapes and proportions of the skull bones. (3) Distortion is obvious each time, as the initial skull symmetry has been affected, but its extent is often difficult to determine. (4) Many of the described Dicynodon species are known only by incomplete and badly preserved specimens, making detailed comparisons difficult. In spite of the above-mentioned problems, it appears that new species of Dicynodon from Laos, now being described, can be established on the basis of series of characters whose association is not found in any other species. The anatomical study of the cranial material recovered from the purple beds revealed many differences of varying importance between the specimens. Differences in dimensions must be considered with caution, as they could merely reflect differences in individual age. Some differences in the skull proportions can also be related to different ontogenetic stages within the same species; it must be remembered, for example, that in all therapsid species in which a growth series could be established, the temporal fenestra is relatively larger in the largest individuals. However, other anatomical differences cannot be explained in terms of individual variability, and have a taxonomic significance: the relief of the skull roof, the structure of the occiput (with a more or less developed dorsal notch of the squamosal) (Fig. 6), the presence or absence of a maxillary antrum, or the shape of the canines (rounded in transverse section, compressed anteroposteriorly, or compressed laterally with sharp anterior and posterior edges) (Fig. 7) fall in that category. If, as I believe, Dicynodon is

Fig. 6. Occipital views of two Dicynodon specimens from Laos. (a) Specimen LPB No. 1993– 3: the dorsal notch of the occipital part of the squamosal is very deep. (b) Specimen LPB No. 1993– 2: the dorsal notch of the occipital part of the squamosal is very shallow.

38

B. BATTAIL

Fig. 7. Ventral views of snouts of three Dicynodon specimens from Laos. The canines, when preserved, have not been figured. (a) Specimen LPB No. 1993– 2: canine base rounded. (b) Specimen LPB No. 1993–3: canine base compressed anteroposteriorly. (c) Specimen LPB No. 1996–11 A: canine base compressed laterally.

represented in Laos by new, distinct species, we would have another example of endemism in Permian dicynodont faunas, a phenomenon already stressed by Angielczyk & Kurkin (2003b, pp. 184 –185).

Age of the dicynodont fauna from the purple beds of Luang Prabang It can now be assumed that the purple beds of the area of Luang Prabang do not contain Lystrosaurus, but specimens of Dicynodon. Until recently, this sole fact would have been regarded as sufficent to consider the purple beds as representing the very end of the Permian. In South Africa, Dicynodon has been used to define the last biozone of the Permian, the Dicynodon Assemblage Zone (Kitching 1995; Rubidge 1995). Also, considering the cosmopolitan distribution of Dicynodon, Lucas (1998, 2001) suggested that the genus could be used to define a Late Permian biochron in Africa, Europe and Asia. Nevertheless, in South Africa, at least one specimen of Dicynodon lacerticeps was collected in the underlying Cistecephalus Assemblage Zone strata (Rubidge, pers. comm. in Angielczyk & Kurkin 2003a, p. 365). Also, it appears now that Dicynodon might be paraphyletic, and would therefore not correspond to a biologically real entity. As a consequence, Dicynodon should be used with caution in biostratigraphic correlations, especially between widely separated basins (Angielczyk & Kurkin 2003a, b). In spite of these

reservations, it seems that all the specimens attributed to Dicynodon that could be dated with some degree of confidence are Late Permian in age. At least in South Africa, where the end-Permian extinction has been carefully studied, the last appearence of Dicynodon marks the Permian –Triassic boundary (Smith 1995; MacLeod et al. 2000; Ward et al. 2000; Smith & Ward 2001). Consequently, even if Dicynodon can no longer be regarded as a very accurate stratigraphic marker, a Late Permian age of the purple beds of the area of Luang Prabang can be considered as established; a few additional data, although not very informative, are not in contradiction with such an age. Coral remains, unfortunately not found in situ, were collected towards the base of the purple beds. They had most probably been reworked, incorporated as pebbles into the lower conglomerate that marks the beginning of the succession, and then freed from the conglomerate by weathering. They were determined by Fontaine as Ipciphyllum laosence, Ipciphyllum subelegans and Multimurinus kmerianus, an assemblage that is characteristic of the Late Murgabian (H. Fontaine, pers. comm.); therefore, the purple beds cannot be older than Late Murgabian (¼ Late Kazanian) (Battail et al. 1995). In addition to the dicynodonts, which constitute the bulk of the fauna, rare remains of Anthracosauromorpha, now being studied, have also been collected from the purple beds. Their closest relatives seem to be the Chroniosuchida described from the Late Tatarian (Latest Permian) of Russia.

DICYNODONTS FROM LAOS

Conclusions The genus Lystrosaurus has never been recorded from Laos, in spite of many mentions of it in the literature. In the purple beds of the area of Luang Prabang, dicynodonts are represented by the Late Permian genus Dicynodon. Locating the Dicynodon-bearing strata of Laos in a broader geological context is of particular interest: the outcrops are situated on the northwestern edge of the Indochina block, which has often been regarded as having no land connection with Pangaea in the Late Permian (Metcalfe 1996, 2002; Shi 2003). The presence in the Indochina block of Dicynodon, a Pangaean terrestrial genus, leads us to reconsider this palaeogeographical model. I thank the people who took part in the field work: R. Allain, J. Dejax, F. L. Duparcmeur, K. Manikham, O. Mateus, P. Richir, J.-S. Steyer, P. Taquet, R. Vacant and M. Ve´ran. I am also indebted to P. Sayarath, from the STEA (Science, Technology and Environment Agency), Vientiane, Laos, for his support in the project. I am grateful to C. Letenneur, who drew the figures, and to C. Lemzaouda, who prepared the photographs for publication. Comments by reviewers M. J. Benton and K. D. Angielczyk led to significant improvements of the final version of this paper.

References A NGIELCZYK , K. D. 2001. Preliminary phylogenetic analysis and stratigraphic congruence of the dicynodont anomodonts (Synapsida: Therapsida). Palaeontologia Africana, 37, 53–79. A NGIELCZYK , K. D. & K URKIN , A. A. 2003a. Has the utility of Dicynodon for Late Permian biostratigraphy been overstated? Geology, 31, 363– 366. A NGIELCZYK , K. D. & K URKIN , A. A. 2003b. Phylogenetic analysis of Russian Permian dicynodonts (Therapsida: Anomodontia): Implications for Permian biostratigraphy and Pangaean biogeography. Zoological Journal of the Linnean Society, 139, 157–212. B ATTAIL , B. 1997. Les genres Dicynodon et Lystrosaurus (Therapsida, Dicynodontia) en Eurasie: une mise au point. Geobios, M.S. 20, 39–48. B ATTAIL , B., D EJAX , J., R ICHIR , P., T AQUET , P. & V ERAN , M. 1995. New data on the continental Upper Permian in the area of Luang-Prabang, Laos. Journal of Geology, Hanoi, 5– 6, 11– 15. C LUVER , M. A. 1971. The cranial morphology of the dicynodont genus Lystrosaurus. Annals of the South African Museum, 56, 55–274. C LUVER , M. A. & H OTTON , N. 1981. The genera Dicynodon and Diictodon and their bearing on the classification of the Dicynodontia (Reptilia, Therapsida). Annals of the South African Museum, 86, 99– 146. C LUVER , M. A. & K ING , G. M. 1983. A reassessment of the relationships of Permian Dicynodontia (Reptilia, Therapsida) and a new classification of

39

dicynodonts. Annals of the South African Museum, 91, 195 –273. C OLBERT , E. H. 1982. The distribution of Lystrosaurus in Pangaea and its implications. Geobios, M.S. 6, 375– 383. C OUNILLON , H. 1896. Documents pour servir a` l’e´tude ge´ologique des environs de Luang-Prabang (Cochinchine). Comptes Rendus de l’Acade´mie des Sciences, 123, 1330– 1333. D AS G UPTA , H. C. 1922. Note on the Panchet Reptile. In: Sir Asutosh Mukherjee Silver Jubilee Volumes. Vol. 2, Science. University Press, Calcutta, 237– 241. F RO¨ BISCH , J. 2007. The cranial anatomy of Kombuisia frerensis Hotton (Synapsida, Dicynodontia) and a new phylogeny of anomodont therapsids. Zoological Journal of the Linnean Society, 150, 117–144. K EYSER , A. W. 1975. A reevaluation of the cranial morphology and systematics of some tuskless Anomodontia. Memoirs of the Geological Survey of South Africa, 67, 1 –110. K EYSER , A. W. & C RUICKSHANK , A. R. I. 1979. The origins and classification of Triassic dicynodonts. Transactions of the Geological Society of South Africa, 82, 81–108. K ING , G. M. 1988. Anomodontia. In: Handbuch der Pala¨oherpetologie, Part 17c. Gustav Fischer, Stuttgart, 1–174. K ITCHING , J. W. 1995. Biostratigraphy of the Dicynodon Assemblage Zone. South African Committee for Stratigraphy Biostratigraphic Series, 1, 29–34. K URKIN , A. A. 2001. Novye pozdnepermskie dicinodonty vjaznikovskogo kompleksa nazemnykh tetrapod Vostochnoj Evropy. Paleontologicheskij Zhurnal, 2001, 53–60. L UCAS , S. G. 1998. Toward a tetrapod biochronology of the Permian. In: L UCAS , S. G., E STEP , J. W. & H OFFER , J. M. (eds) Permian stratigraphy and paleontology of the Robledo Mountains, New Mexico. Bulletin of the New Mexico Museum of Natural History and Science, 12, 71– 91. L UCAS , S. G. 2001. Chinese Fossil Vertebrates. Columbia University Press, New York. M AC L EOD , K. G., S MITH , R. M. H., K OCH , P. L. & W ARD , P. D. 2000. Timing of mammal-like reptile extinctions across the Permian– Triassic boundary in South Africa. Geology, 28, 227–230. M AISCH , M. W. 2002. A new basal lystrosaurid dicynodont from the Upper Permian of South Africa. Palaeontology, 45, 343–359. M ETCALFE , I. 1996. Gondwanaland dispersion, Asian accretion and evolution of eastern Tethys. Australian Journal of Earth Sciences, 43, 605– 623. M ETCALFE , I. 2002. Permian tectonic framework and paleogeography of SE Asia. Journal of Asian Earth Sciences, 20, 551–566. O WEN , R. 1845. Description of certain fossil crania discovered by A. G. Bain, Esq., in the sandstone rocks at the southeastern extremity of Africa, referable to different species of an extinct genus of Reptilia (Dicynodon), and indicative of a new tribe or suborder of Sauria. Transactions of the Geological Society of London, 7, 59–84.

40

B. BATTAIL

P IVETEAU , J. 1938. Un the´rapside´ d’Indochine. Remarques sur la notion de continent de Gondwana. Annales de Pale´ontologie, 27, 137–152. R EPELIN , J. 1923. Sur un fragment de craˆne de Dicynodon recueilli par H. Counillon dans les environs de LuangPrabang (Haut-Laos). Bulletin du Service Ge´ologique de l’Indochine, 12, 1– 7. R UBIDGE , B. S. (ed.) 1995. Biostratigraphy of the Beaufort Group (Karoo Supergroup). South African Committee for Stratigraphy Biostratigraphic Series, 1, 1 –46. S HI , G. R. 2003. Aspects of Permian Gondwana in central Tibet and SE Asia: Distribution, paleoclimatic and paleogeographical implications. In: W ONG , T. E. (ed.) Proceedings of the XVth International Congress on Carboniferous and Permian Stratigraphy. Utrecht, the Netherlands, 10– 16 August 2003. Royal Netherlands Academy of Arts and Sciences, Amsterdam, 555– 564. S MITH , R. M. H. 1995. Changing fluvial environments across the Permian– Triassic boundary in the Karoo

basin, South Africa and possible causes of tetrapod extinctions. Palaeogeography, Palaeoclimatology, Palaeoecology, 117, 81–104. S MITH , R. M. H. & W ARD , P. D. 2001. Pattern of vertebrate extinctions across an event bed at the Permian– Triassic boundary in the Karoo basin of South Africa. Geology, 29, 1147–1150. T RIPATHI , C. & S ATSANGI , P. P. 1963. Lystrosaurus fauna of the Panchet Series of the Raniganj Coalfield. Memoirs of the Geological Survey of India. Palaeontologia Indica, New Series, 37, 1– 53. W ARD , P. D., M ONTGOMERY , D. R. & S MITH , R. M. H. 2000. Altered river morphology in South Africa related to the Permian– Triassic extinction. Science, 289, 1740– 1743. W OODWARD , A. S. 1932. Dicynodontidae. In: VON Z ITTEL , K. A. (ed.) Textbook of Palaeontology. Vol. 2. Macmillan, London, 257–260. Y UAN , P. L. & Y OUNG , C. C. 1934. On the occurrence of Lystrosaurus in Sinkiang. Bulletin of the Geological Society of China, 12, 575–580.

Mesozoic red bed sequences from SE Asia and the significance of the Khorat Group of NE Thailand ANDREW RACEY BG Group, 100 Thames Valley Park Drive, Reading RG6 1PT, UK (e-mail: [email protected]) Abstract: New geological data are presented and previously published information is reviewed to demonstrate that much of the Khorat Group (Phu Kradung to Khok Kruat Formations) of NE Thailand is Early Cretaceous in age. It is suggested that the Mesozoic red bed sequences of neighbouring Indochina are likely to be of similar age rather than spanning the entire Late Triassic to Early Cretaceous as previously assumed. Moreover, the Lower Nam Phong Formation dated as Late Triassic and previously included as the basal formation of the Khorat Group is now removed from this group, thus creating a hiatus within the Jurassic. There is therefore no clear relationship between the Indosinian Orogeny and the Triassic collision of the Sibumasu (also referred to as Shan-Thai) and Indochina Blocks and the subsequent deposition of the Khorat Group in a Late Triassic–Early Cretaceous thermal sag basin. It is now proposed that much of the sequence was deposited during the Late Jurassic–Early Cretaceous. Jurassic sediments may be absent across much of the Khorat Plateau whereas marine Jurassic sediments to the west and east show no sediments younger than Bajocian. Because sea levels were generally rising in the Middle and Late Jurassic it is likely that the Khorat region was uplifted at this time. It is suggested that the Khorat Group was originally deposited in a foreland basin setting rather than a thermal sag basin following Late Triassic rifting. Moreover, the original site of deposition was to the north in southern China, with the present-day location being the result of movement along the Red River Fault coupled with a clockwise rotation of the Indochina Block (on which the Khorat Group sits) with respect to the South China Block.

The Khorat Group and its lateral equivalents comprise a thick sequence of Mesozoic continental red bed sediments, which were deposited over much of NE Thailand and neighbouring parts of SE and central Laos (around Vientiane, Savannakhet and Pakse) and Cambodia (Fig. 1). It underlies much of the Khorat Plateau in NE Thailand and has an areal extent of around 200 000 km2 with a maximum present-day preserved width of around 500 km and maximum preserved sediment thickness of around 4.5 km. Possible coeval red beds also occur in southern Peninsular Thailand (Trang Group) and in Peninsular Malaysia (within the Gagau and Tembeling Groups). Up to 3 km of postKhorat Group section was removed in NE Thailand by erosion during Tertiary uplift associated with the collision between India and Eurasia and the formation of the Himalayas (Racey et al. 1996, 1997a, b). Although the Khorat Group crops out on the Indochina Block it does partly extend westwards across the Loei–Petchabun and Sukothai Fold Belts onto the Sibumasu Block, the two blocks having been sutured together since at least the Late Triassic (Figs 1, 2). The Indochina Block comprises eastern Thailand, Laos, Cambodia and Vietnam, and is also referred to in the literature as the Indosinian Block, whereas the Sibumasu Block comprises western Thailand, eastern

Burma, Peninsular Malaysia and Sumatra, and is also referred to in the literature as the Shan-Thai Block (Fig. 2). The original depocentre of the Khorat Basin appears to have been oriented roughly NW–SE and superficially it has the appearance of a typical sag basin with each formation gradually overstepping the previous formation towards the edge of the basin. The Khorat Plateau has an uplifted western and southern margin (up to 1300 m above sea level) with a mesa-like appearance formed as a result of compressional deformation along its western margin (Loei–Phetchabun Fold Belt), which marks the western margin of the Indochina Block and the approximate boundary between the Indochina and Sibumasu Blocks. The central region of the Khorat Plateau is at present only a few hundred metres above sea level. The sedimentary fill comprises a non-marine red bed sequence (Khorat Group), which is unconformably overlain by continental evaporites and clastic deposits of the Albian –Cenomanian Maha Sarakham Formation and is underlain by Late Triassic fluvial and lacustrine sediments with associated volcanic rocks of the Huai Hin Lat and Lower Nam Phong Formations and Late Palaeozoic marine sediments (Fig. 3). Traditionally the Khorat Group comprises six formations, which are from bottom

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 41–67. DOI: 10.1144/SP315.5 0305-8719/09/$15.00 # The Geological Society of London 2009.

42

A. RACEY

Fig. 1. Extent of Jurassic–Cretaceous red beds and principal basins in mainland SE Asia.

to top: Nam Phong, Phu Kradung, Phra Wihan, Sao Khua, Phu Phan and Khok Kruat (Fig. 3). Based on seismic and well data the Nam Phong Formation varies in thickness from 2500 m in the central portion of the basin to ,500 m along its flanks. The formation is clearly divisible on seismic sections into an Upper Nam Phong

Formation and Lower Nam Phong Formation separated by an unconformity. The Phu Kradung Formation varies from 1200 m in the basin centre to around 500 m in thickness on the basin flanks. The Phra Wihan Formation varies in thickness from around 50 to 300 m, and the Sao Khua Formation varies widely in thickness from around 100 to

MESOZOIC RED BEDS FROM SE ASIA

43

Fig. 2. Inferred location of the key sutures and blocks within the study area (see also Fig. 4).

700 m. The Phu Phan Formation is generally 50– 100 m thick, and the Khok Kruat Formation shows marked variations in thickness because of erosion, from 200 m in the NW to 850 m in the SE. The generally accepted model for the region is that the continental collision between the Sibumasu Block and mainland Indochina Block occurred in the Late Triassic (Indosinian Orogeny). Uplift

and orogenic faulting associated with this event led to the formation of intermontane thermal sag basins, which were rapidly infilled with Jurassic– Cretaceous continental sediments (Khorat Group and its equivalents) over a broad area from eastern Peninsular Malaysia across the South China Sea into the Khorat Basin of NE Thailand, southern Laos, Cambodia and Vietnam. However, as will

44

A. RACEY

Fig. 3. Revised stratigraphic column for the Mesozoic of NE Thailand with the main depositional environments and key tectonic events.

MESOZOIC RED BEDS FROM SE ASIA

be evident from the information presented below, this is probably an oversimplification. For a more detailed summary of the lithologies and environments of deposition of the various formations making up the Khorat Group the reader is referred to Racey et al. (1996). The biostratigraphy of the Khorat Group, especially the palynology and its importance in dating these formations, has been discussed in detail by Racey & Goodall (2009).

45

Terranes and sutures As shown by Metcalfe (1996) mainland SE Asia comprises a complex mosaic of allochthonous terranes that accreted throughout the Late Palaeozoic to Mesozoic (Fig. 4). The exact timing of these collisions and the relationship between the various tectonic fragments is by no means certain and has been the subject of much debate (Cooper et al. 1989; Metcalfe 1996, and references therein).

Fig. 4. Regional blocks and sutures referred to in text (after Metcalfe 1996). EM, East Malaysia; WB, West Burma; SWB, SW Borneo; S, Semitau Terrane; HT, Hainan Island terranes; L, Lhasa Terrane; QT, Qiangtang Terrane; QS, Qamdo– Simao Terrane; SI, Simao Terrane; SG, Songpan Ganzi accretionary complex; KL, Kumlun Terrane; QD, Qaidam Terrane; AL, Alashan Terrane. Dashed line between Indochina and EM marks the approximate southern boundary of the Indochina Block.

46

A. RACEY

There are three major cratons or blocks of relevance to the current study area. The northern margin of the study area is marked by the South China Block, bordered to the immediate south by the Indochina Block, which in turn is bordered to the west and SW by the Sibumasu Block (Fig. 4). The Indochina Block is bounded to the NE by the Song Ma Suture (Red River Fault Zone), to the NW by the Nan–Uttaradit Suture, which is partly coincident with the Dien Bien Phu Fault, and to the south by the Wang Chao and Three Pagodas sinistral strike-slip faults (dashed line in Fig. 4). The Sibumasu Block is bordered on its western side by the Western Burma block along the Sagaing Fault Zone (also referred to as the Shan Boundary Fault). The South China Block is bounded to the north by the Qinling mountains (Qinling–Dabie Suture), to the west by the Tertiary Long Men Shan Thrust Belt and to the SW by the sinistral Xian-Shui-He (Ailaoshan Suture) and Song Ma Suture (Red River Fault).

Geological history (Figs 3 and 4) The Indochina and South China Blocks collided during the Early Carboniferous, the boundary being marked by the NW– SE-trending Song Ma Suture (Metcalfe 1996). Within this paper two main Indosinian Orogenies are recognized, following Mouret (1994) and Booth (1998). Interestingly, two Indosinian phases are also recognized in China (Tin 1992) although these are related to different collisions (Wan & Zhu 1991). The Sibumasu and Indochina Blocks collided during the Indosinian I Orogeny in the Late Permian–Triassic along the NNE–SSW-trending Nan– Uttaradit Suture, which continues southwards as the Phetchabun Fold Belt and in Peninsular Malaysia as the Bentong –Raub Suture. Continued movement northwards of these two plates resulted in subduction between the blocks and a major period of uplift, erosion, peneplanation and thrusting (thin and thick skinned). However some workers (e.g. MacDonald & Barr 1984), observed that the volcanic rocks of this age in the Nan area of northern Thailand more closely resemble volcanic arc basalts than true obducted oceanic crust, and if correct this would suggest that there was no significant obduction of an intervening oceanic basin, at least not along this part of the suture between Sibumasu and Indochina. MacDonald & Barr (1984) suggested that the suture could be further east beneath the thick Mesozoic cover and that the collision between Sibumasu and Indochina was oblique. The Song Da Suture is considered to represent a region of Permo-Triassic rift basins that ceased in the Late Triassic with subsequent reactivation of this suture as a major strike-slip zone in the Late Mesozoic to Early

Cenozoic (Metcalfe 1996). During the Late Triassic a series of intracontinental rift basins developed across NE Thailand extending into western Laos. These generally trend WNW –ESE, except in the region of the Phu Phan Uplift (a mid-Cretaceous inversion feature), where they trend north– south. These are filled with fluvial –lacustrine sands, silts and mudstones of the Late Triassic (Carnian – Norian) Huai Hin Lat Formation. Within the Khorat Plateau region a second major tectonic event of only slightly younger Late Triassic age is observed on seismic sections separating the Huai Hin Lat Formation from the Late Triassic Lower Nam Phong Formation. This event is herein referred to as the Indosinian II Orogeny and represents a second period of uplift, erosion and peneplanation with more localized thin-skin thrusting (Booth 1998). The end of the Triassic is marked by an unconformity, which on seismic sections in the Khorat Plateau region separates the Lower from the Upper Nam Phong Formation (Cimmerian/Indosinian III event). This event represents an unconformity or hiatus with the subsequent deposition of overlying Khorat Group fluvial sediments of the Upper Nam Phong, Phu Kradung, Phra Wihan, Sao Khua, Phu Phan and Khok Kruat Formations, which are dated as ?Late Jurassic to Early Cretaceous. A possible Kimmeridgian (Cimmerian) event is also identified in Vietnam (Tien 1991) and this may be contemporaneous with the Yanshan I compression recognized in China (Mouret 1994). The top of the Khorat Group is marked by an unconformity between the Aptian Khok Kruat Formation and the Albian–Cenomanian continental evaporitic Maha Sarakham Formation. This unconformity represents a mid-Cretaceous inversion, which is also recognized to the north in Laos (Lovatt-Smith et al. 1996). This inversion led to the development of the Phu Phan Uplift in NE Thailand and to the subsequent formation of a rimmed basin across much of the Khorat Basin into which the Maha Sarakham Formation was deposited. This was followed by the deposition of fluvial and aeolian sediments of Latest Cretaceous –Early Tertiary age (the Phu Tok Formation). The unconformity at the base of the Maha Sarakham Formation is broadly coeval with the Yanshan II Phase of orogeny recognized in China (Mouret 1994). The continuing collision of India with Asia during the Tertiary resulted in extensive dextral movement along the Song Da Suture, causing South China to move southeastwards along the suture over implied distances of several hundred kilometres (Sato et al. 1999). It also resulted in general uplift and inversion of the Khorat Plateau (as supported by its present elevation and apatite fission-track data; see below). Such dextral slip may have caused a clockwise rotation of the Khorat Plateau such that the current location of

MESOZOIC RED BEDS FROM SE ASIA

the Khorat Basin may not reflect its original relative geographical location, with the possibility that much of the Khorat Group could have been deposited originally as far away as southern China (see Carter & Bristow 2002, and section below on Palaeomagnetic data).

Stratigraphy Khorat Group, NE Thailand Many previous workers have assumed that the Nam Phong Formation, which ‘traditionally’ forms the basal part of the Khorat Group, is Late Triassic and that the Early, Middle and Late Jurassic are represented by the overlying Phu Kradung, Phra Wihan and Sao Khua Formations respectively, with the remainder of the group (Phu Phan and Khok Kruat Formations) assigned to the Early Cretaceous. Early stratigraphic studies of the Khorat Group mainly used plant fossils coupled with guesswork to assign ages to the various formations. Subsequent age dating has been based mainly on vertebrates from the Phu Kradung and Sao Khua Formations (Buffetaut et al. 1993), many of which have subsequently been shown to range into the Early Cretaceous (see Racey et al. 1994, 1996; Racey & Goodall 2009; Buffetaut et al. 2006, 2009, and references therein). Palynological data now suggest an Early Cretaceous (Berriasian to Aptian) age for the Phu Kradung to Khok Kruat Formations, although the lowermost part of the Phu Kradung and Upper Nam Phong Formations could in part be Jurassic (most probably Late Jurassic), as discussed by Racey & Goodall (2009). These revised ages mean that a significant hiatus must exist between the Lower Nam Phong and Upper Nam Phong Formations, and that the former should therefore no longer be considered to be part of the Khorat Group. The Lower Nam Phong Formation on seismic sections occurs below the base Khorat unconformity and should therefore be excluded from the Khorat Group (J. E. Booth, pers. comm.). Many oil company geologists who have studied seismic sections across the region consider this unconformity, referred to in Figure 3 as the ‘Cimmerian Event’ (Indosinian III orogeny), to mark the boundary between an ‘upper’ Nam Phong and a ‘lower’ Nam Phong Formation, the ‘lower’ comprising mainly sandstone. Racey et al. (1994, 1996) presented new palynological evidence that reassigned the Phra Wihan and Sao Khua Formations to the Early Cretaceous and the Phu Kradung to the Late Jurassic to Early Cretaceous. Racey & Goodall (2009) have recently acquired additional palynological data that further support this reassignment and also indicate that most of the Phu Kradung Formation is Early Cretaceous. In recent years several new vertebrate

47

discoveries have also led to a revision of the ages of the formations such that they are now more in line with the palynological dating (see Buffetaut et al. 2006, 2009). Although common mention has been made of Jurassic vertebrates in many papers, the recent publications of Buffetaut et al. (2006, 2009, and references therein) indicate that the Phra Wihan to Khok Kruat Formations are now considered to be Early Cretaceous in age whereas the vertebrate dating suggests that the Phu Kradung Formation is Late Jurassic (Tithonian) or Early Cretaceous in age.

Indochina ‘red beds’ Within the broader context of Indochina, Mesozoic continental red beds form two main ‘groups’ identified by the early French workers in the 1930s. These comprise the ‘Terrain Rouge’, which consists of gently folded sandstones, marls, shales and red– purple claystones that vary markedly in thickness from 100 to 2000 m. This interval is also referred to as the ‘Indosinias Moyennes’ or ‘Middle Indosinias’. In older literature it was assigned to the Late Triassic to Middle Jurassic. However, Workman (1977) considered this interval to be equivalent to the Phu Kradung Formation of NE Thailand, which at that time was considered to be Early Jurassic (now thought to be mainly Early Cretaceous), and that it rested unconformably on folded Triassic or older rocks. In central Vietnam fossil wood of Middle Jurassic–Early Cretaceous age has been recorded in this group (by Hutchison 1989). There are older references indicating the presence of upper Carnian– Norian marine intercalations in the lower part of the Terrain Rouge in northern Laos (Fromaget 1937; Saurin 1950) and ?Norian– Liassic in eastern Cambodia (Saurin 1935) but these may not actually be part of the Terrain Rouge sensu stricto; for example, Saurin (1944) and Fontaine (1964) noted that the Terrain Rouge in central Vietnam rested unconformably on marine Liassic sediments. The second younger group occurs over most of southern Indochina and is referred to as the ‘Gre`s Supe´rieurs’ (also known as the Indosinias Supe´rieures or Upper Indosinias); it comprises mainly light grey, often cross-bedded, coarse-grained (sometimes conglomeratic) quartz-rich sandstones with intercalations of shale and coal. In southern Cambodia around Kampot and the Bokor Massif the Gre`s Supe´rieurs rest on the Terrain Rouge and comprise 160–400 m of conglomerate and sandstone overlain by 200 m of finer-grained sediments containing Early Cretaceous plants (Hutchison 1989). Workman (1977) subdivided the Indosinias (Gre`s) Supe´rieures into a lower member, which he tentatively correlated with the Phu Kradung and Phra Wihan Formations (and possibly also the Sao

48

A. RACEY

Khua but in a much more proximal, i.e. sandier and coarser-grained facies), and an upper member, which he considered broadly equivalent to the Khok Kruat and Maha Sarakham Formations of NE Thailand. The lower member comprises thick sequences of sandstone in the Cardamome Mountains and Bolovens Plateau regions of central and northern Laos and SW Cambodia, is up to 1000 m thick and contains some interbeds of conglomerate, marl and occasional lignite seams. Early Cretaceous pollen has been identified from the upper part of this lower member (SNMGP (Service National des Mines de la Ge´ologie et du Pe´trole) 1972–1973). In eastern Cambodia, Laos and neighbouring parts of Vietnam these lower member sandstones overlie (?unconformably) marine upper Liassic (Sinemurian –Toarcian) sediments. The upper member is restricted to the area around Savannakhet eastwards, and this has yielded Early to mid-Cretaceous plants from near Saravanne and reptile bones and freshwater bivalves from Muong Phalane (Hoffet 1939). These bones were originally dated as Senonian by Hoffet (1939) but are now known to be Aptian and equivalent to the Khok Kruat Formation of NE Thailand (Buffetaut et al. 2005). It is interesting to note that Jurassic marine facies across Indochina are apparently restricted to the Liassic, with the youngest interval identified being Toarcian (end of the Early Jurassic) at Ban Don on the Sre Pok in Vietnam, which has yielded the ammonites Hildoceras quadratum and Grammoceras lantenoisi (Workman 1977). Neither of these taxa have been recorded from the Liassic sequences of western Thailand.

Peninsular Thailand and Peninsular Malaysia ‘red beds’ In SW Peninsular Thailand on the western side of the Bentong–Raub and Nan– Uttaradit suture nonmarine red beds of the Middle Jurassic to Upper Cretaceous Trang Group crop out. These comprise the following four formations (from bottom to top): Khlong Min, Lam Thap, Sam Chom and Phu Phin (Meesook et al. 2002). As noted by Racey & Goodall (2009), the palynomorph assemblages recovered to date from these formations, although limited, show marked similarities to those of the Khorat Group of NE Thailand, although in some cases they are stratigraphically younger and include samples dated as Albian –Cenomanian, Berriasian –Albian and early Barremian– Berriasian. In eastern Peninsular Malaysia east of the Bentong–Raub suture, an area that also formed part of the Indochina landmass, the lower part of the Tembeling Group (Kerum Volcanic series and

Lanis Conglomerate Formations) is considered equivalent to the Terrain Rouge and its upper part (the Mankin Sandstone and Termus Shale Formations) is considered equivalent to the Gre`s Supe´rieurs. Elsewhere in Peninsular Malaysia the Tebak Formation and the Gagua Group (comprising the Badong Conglomerate and the Lotong Sandstone Formations) are considered to be broadly equivalent to the Gre`s Supe´rieurs (Hutchison 1989). The Tebak Formation in SE Peninsular Malaysia has yielded Early Cretaceous fossil plants, and the flora of the underlying Tembeling formation is similar and may for the most part be assigned to the Early Cretaceous (Smiley 1970a, b). In NE Peninsular Malaysia also on the eastern side of the Bentong – Raub Suture, red conglomerates and subordinate sandstones, siltstones and shales of the Badong Conglomerate are overlain by the Lotong Sandstone, which comprises cross-bedded sandstones with some rhyolite and coal interbeds (Hutchison 1989). These two formations form the Gagau Group, which has yielded Late Jurassic –Early Cretaceous plant fossils comparable with those of the Tebak Formation. Palynomorphs including abundant Corollina spp. with rare Aequitriradites cf. verrucosis have been recovered from plant-bearing beds within the Lotong Sandstone of the Gargua Formation with A. cf. verrucosis indicating an Early Cretaceous age (Smiley 1970a, b). Within the Termus Shale and Mangkin Sandstone Formations of the Tembeling Group, Jirin & Morley (1994) recorded a palynomorph assemblage comprising Corollina, Exisipollenites, Ephedrites, Cycadopites, Cicatricosisporites, Eucommidites and Clavatipollenites to which they assigned an Early Cretaceous age. Moreover, the occurrence of Clavatipollenites in the Termus Shale Formation indicates a Barremian age whereas its absence in the underlying Mangkin Sandstone Formation may indicate that the latter is pre-Barremian. Lithologically similar red bed sequences of presumed Mesozoic age have also been encountered in hydrocarbon exploration wells in the offshore Gulf of Thailand although to date these have yielded only rare, poorly preserved, long-ranging Mesozoic palynomorphs (pers. obs.).

Jurassic – Cretaceous sequences of mainland SE Asia (Fig. 5) Vientiane Basin – Sayabouri– Nakhon Thai Basin The Vientiane Basin represents the northern extension of the Khorat Basin into southern Laos, where the Phu Kradung to Khok Kruat Formations are represented by the lithologically similar Nam

MESOZOIC RED BEDS FROM SE ASIA

49

Fig. 5. Lithostratigraphy and possible correlation of Late Triassic to Late Cretaceous sequences of NE Thailand with those of the surrounding region.

50

A. RACEY

Set, Phu Phanang, Ban Ang, Champa and Ban Thalat Formations, which rest unconformably on a Triassic rift sequence (Huai Hin Lat Formation equivalent) and are unconformably overlain by a continental evaporitic sequence (Thangon Formation, which is equivalent to the Maha Sarakham Formation in NE Thailand). Stokes et al. (1996) observed that the Pak Lay Fold Belt, which borders the western side of the Vientiane Basin, comprises a compressively deformed imbricate wedge of ?Late Carboniferous to Late Jurassic sediments and associated Triassic to Late Jurassic lavas, agglomerates and tuffs. They concluded that volcanic arc-related activity occurred during subduction prior to collision and the age of this collision can therefore be no older than Late Jurassic. They recorded a palynological assemblage of Corollina spp., Callialasporites dampieri, Cerebropollenites mesozoicus, Uvaesporites spp. and Peripollenites elatoides to which they assigned a Middle to Late Jurassic age. However, this assemblage could equally well extend into the Early Cretaceous. Radiometric dating of extrusive basalts yielded Bajocian and Kimmeridgian ages whereas intrusive granites were dated as Aptian (Stokes et al. 1996) with youngest ages in the west. Stokes et al. (1996) concluded that subsequent major erosion was followed by the deposition of the ‘Khorat Group’, which oversteps this truncated Carboniferous–Late Jurassic sequence (and would therefore suggest an age for the ‘Khorat Group’ of no older than Late Jurassic). Granites were intruded in the mid-Cretaceous and these may have been associated with the inversion of the Khorat Basin prior to deposition of the Late Cretaceous Phon Hong Group (comprising the Thangon and Saysomboun Formations) in the Vientiane Basin (Stokes et al. 1996). The Pak Lay–Phetchabun Fold Belt was inverted and eroded during the mid-Cretaceous and possibly also during Tertiary compression, removing the Cretaceous cover from this region. The Nakhon Thai –Sayabouri Basin is located on the western side of the Nan–Uttaradit Suture and extends from the area around Nan in northern Thailand northwards into westernmost Laos and contains a Khorat Group red bed fill.

Savannakhet Basin The Savannakhet Basin represents the eastwards extension of the Khorat Basin into SW Laos and has a broadly similar stratigraphy (Fig. 5). The sequence comprises non-marine Late Triassic synrift clastic rocks of the Apok Formation (Huai Hin Lat Formation equivalent) overlain by alluvial sediments with westward-directed palaeocurrents, of the Bangbouyang Formation (Nam Phong Formation equivalent). These are unconformably

overlain by lacustrine to quasi-marine or lagoonal sandstones, mudstones and limestones of the Lamo Formation (Phu Kradung Formation equivalent), which pass upwards into mud-dominated fluvial or alluvial red beds of the Salavan Formation, which becomes depositionally more proximal up-section, passing into stacked massive sheet and channelled sandstones (Phu Kradung Formation equivalent in its lower part becoming sandier and more like the Phra Wihan Formation in its upper part). This is overlain by the Bangfai Formation (Phu Phan Formation equivalent), which comprises pale-coloured medium-grained to pebbly quartz – feldspathic sandstones deposited in a proximal braided river setting. The absence of an obvious intervening red muddy– silty equivalent of the Sao Khua Formation seen in NE Thailand may be explained by the fact that this basin is closer to the sediment source area, causing a marked thinning (and ?partial erosion) of the Sao Khua equivalent interval creating an effective ‘merging’ of the distinctive sandstone-dominated ‘Phra Wihan’ and ‘Phu Phan’ Formation equivalents into a single formation referred to here as the Bang Fai Formation. This in turn is overlain by a sequence of argillaceous sediments with thin sandstones and occasional evaporites deposited in a playa lake setting with fluvial influence and referred to the Champon Formation, which is assumed to be equivalent to the Khok Kruat and possibly in part the Maha Sarakham Formations.

Western Thailand (west of suture) In western Thailand (Mae Sariang– Kanchanaburi Basin area) the marine Mae Moei Group rests unconformably on Triassic marine sediments (Fig. 5). It comprises undifferentiated Early Jurassic (dating based on foraminifers and calcareous algae), and late Toarcian –early Bajocian sediments based on ammonites (Chonglakmani 1982; von Braun & Jordan 1984; Fontaine & Suteethorn 1988; Meesook & Grant-Mackie 1996). Sediments of late Middle Jurassic age (Bathonian–Callovian) and latest Late Jurassic (Kimmeridgian– Tithonian) have not been recorded. Marine sedimentation had ceased by the Late Jurassic and the Mae Moei Group along this western margin of the Sibumasu Block is unconformably overlain by inferred Cretaceous continental red beds of the Kalaw Formation (von Braun & Jordan 1976). This would support the idea of uplift and erosion during the Late Jurassic–Early Cretaceous (Cimmerian) and may be correlatable with the Kimmeridgian orogeny in eastern Burma, where Jurassic Loian Beds are unconformably overlain by Kalaw red beds of inferred Late Jurassic– Early Cretaceous age (Thein 1973). This Cimmerian event was

MESOZOIC RED BEDS FROM SE ASIA

accompanied by widespread granite emplacement (Klompe 1962; Kobayashi 1984).

Southern Laos, Cambodia and south Vietnam Key studies in this area are those by Fontaine (1962), Workman (1977) and Tien (1991). The Jurassic is represented by the Tholam and Bandon Suites, which rest unconformably on a variety of Precambrian to Triassic rocks. A Lower Jurassic (Hettangian –Toarcian) to Middle Jurassic (Aalenian –Bajocian) marine sequence is known from southern Laos, Kampuchea and South Vietnam whereas the Bathonian –Callovian is ‘missing’ (as in Western Thailand) and the inferred Upper Jurassic is represented by red beds (Tien 1991). These red bed sequences could be Khorat Group equivalent (?Phu Kradung). Jurassic limestones in southern Laos and Vietnam are interbedded with non-marine ?Phu Kradung equivalent clastic rocks and overlain by pale-coloured sandstones (possibly equivalent to the Phra Wihan Formation).

Peninsular Malaysia Peng (1983) recorded ?Late Jurassic to Early– Middle Cretaceous non-marine red beds from the eastern side of the Bentong Raub Suture forming the Tebak Formation and Gagua Group, and similar age red beds are known (Jirin & Morley 1994) from the upper part of the Tembeling Group (Mangkin Sandstone and Termus Shale Formations) (Fig. 5). Interestingly, the Lower and Middle Jurassic appear to be absent and the Triassic is mainly marine. This suture probably represents the southwards continuation of the Nan–Uttaraditt Suture, which separates the Sibumasu and Indochina Blocks.

Tectonic models for the evolution of the Khorat Basin The formation and evolution of the Khorat Basin is generally poorly understood and has been the subject of much debate. Saurin (1956) first suggested the presence of two orogenies within this region, an Indosinian (late Triassic) and a ‘Cimmerian’ (late Jurassic to Early Cretaceous), and Klompe (1962) suggested that the boundary between these two orogenies was located along the western margin of the Khorat Plateau. Baum et al. (1970) concluded that the boundary was further to the west along the north–south-trending ChiangMai ‘Geanticline’. The later ‘Cimmerian’ Orogeny was accompanied by the widespread emplacement of granites. Early tectonic models assumed that the

51

Khorat Group was deposited as molasse related to the Indosinian Orogeny; that is, the Late Triassic collision between Sibumasu and Indochina (Bunopas & Vella 1978: Hutchison 1989). The model most frequently applied to the area is that of Cooper et al. (1989), who suggested that the Khorat Group was deposited in a thermal sag basin following Late Triassic extension related to the crustal collapse of an overthickened crust produced by the Indosinian Orogeny. More recently, Lovatt-Smith et al. (1996) interpreted the Khorat Basin as being a foreland basin deposited at the front of a Jurassic Orogenic Belt created by the Song Ma and Song Da sutures to the north or NE, which broadly parallel the axis of the Khorat Basin and mark the suture between the southern China Block and Indochina. Racey et al. (1994, 1996, 1997a) concluded that although the Nam Phong Formation could have in part been deposited during thermal subsidence following Triassic extension the same model could not be applied to the overlying Phu Kradung to Khok Kruat Formations owing to the presence of a hiatus spanning a large part of the Jurassic. However, the unconformity is now considered to fall within the Nam Phong Formation (as discussed above) and separates a Lower Nam Phong Formation of Late Triassic age from an Upper Nam Phong Formation of probable Jurassic (?Late Jurassic) age. Lovatt-Smith et al. (1996) noted that there is little evidence on seismic data for major syndepositional faulting in the Khorat Basin, suggesting that the stretching factor was low. On regional seismic lines Khorat Group formations have a mainly layercake appearance with no significant topographic features for the sediments to infill; that is, they appear typical of deposition in a thermal sag driven setting. From the regional perspective Metcalfe (1996) concluded that from the Late Permian–Triassic onwards the various blocks that now form much of SE Asia began to collide and accrete in a series of four events. First, the South China and Indochina Blocks collided, followed by the collision and accretion of the Qiantang and Sibumasu Blocks to the Indochina Block in the Triassic. The Indochina Block then became sutured to South China in the NE along the Song Ma and Song Da Sutures and to the Sibumasu Block in the west along the Nan– Uttaradit Suture. Finally, in the Early Cretaceous the Lhasa Block, and then in the Late Cretaceous the West Burma Block collided and accreted with the Qiantang–Sibumasu terrane (Metcalfe 1996). Of particular relevance to the Khorat Plateau region is the collision between the Sibumasu– Qiantang Blocks and Indochina in the mid-Triassic with associated late kinematic granites in NE Thailand dated at 200 Ma (end Triassic)

52

A. RACEY

marking the final stages of this collision (Singharajwarapan & Berry 2000). This Triassic event is typically referred to as the Indosinian Orogeny and was originally thought to have formed the thermal sag basin into which the Khorat Group was deposited. However, there are several problems with this interpretation, many of which are discussed below, including the significance of the revised age dating of the Khorat Group, regional palaeogeographical considerations, sediment provenance and palaeocurrent data, to list a few.

Significance of the Khorat Group The following sections outline some of the key features of the Khorat Group and age-equivalent sequences elsewhere in mainland SE Asia and discuss the bearing of these observations on the new ages for the group outlined by Racey & Goodall (2009) and their significance in understanding the geological evolution of the region. A key factor here is that if the revised ages are correct and there is a major hiatus in the Jurassic, then what is the evidence from other geological data for this ‘event’?

Age dating Revised age dating based on palynology by Racey & Goodall (2009, and references therein) indicates that the Phu Kradung to Khok Kruat Formations are mainly Early Cretaceous (Berriasian to Aptian) in age. A Late Jurassic age for the lowermost part of the Phu Kradung Formation cannot be completely ruled out, as the lowermost part of the Phu Kradung Formation lacks the key Early Cretaceous palynomorph D. etruscus but also has not yielded any Jurassic-restricted palynomorphs. Moreover, recent vertebrate discoveries (see Buffetaut et al. 2006, 2009) within the Sao Khua Formation (previously reported to be Late Jurassic) are now considered to be Early Cretaceous, and those from the Phu Kradung Formation previously thought to be Early Jurassic are now considered to be Late Jurassic to Early Cretaceous, based on their degree of evolution and similarity to other similar assemblages from SE Asia. The age of the Nam Phong Formation is more problematic, because in the subsurface well and seismic data suggest that it can be subdivided into two ‘formations’, herein referred to as the Upper Nam Phong and Lower Nam Phong Formations, which are separated by an unconformity. These are up to 1500 and 1000 m thick, respectively. Although the age of the Lower Nam Phong Formation can be constrained to the Late Triassic (Rhaetian) based on palynology and vertebrate data, the age of the Upper Nam Phong Formation

is more problematic. Limited palynological data from the Phu Horm-1 well presented by Racey & Goodall (2009) suggest that the Upper Nam Phong Formation cannot be older than Pliensbachian (middle Early Jurassic) based on the presence of C. turbatus, but is unlikely to extend into the Early Cretaceous (based on the absence of the Early Cretaceous marker D. etruscus seen throughout most of the overlying Phu Kradung Formation). Consequently, the age of the Upper Nam Phong Formation could fall anywhere between the Pliensbachian and the end of the Jurassic. Because the Lower Nam Phong Formation is bounded by unconformities it cannot according to the code of lithostratigraphic nomenclature be assigned to the same formation as the Upper Nam Phong Formation. Consequently, either the Lower Nam Phong or Upper Nam Phong should be given a new formation name.

Palaeogeographical reconstructions; regional correlation If the Khorat Group was mainly Jurassic (a view still held by some Thai geologists) then it is difficult to explain the absence of marine influence and deltaic sediments within the group, as it is bordered by marine Jurassic sediments to the east and west. Although bivalves have been reported from the Phu Kradung Formation, no convincing specimens of marine taxa have yet been illustrated. It could be suggested that the Khorat Group ‘delta’ portion of this fluvial system is located further south in the Gulf of Thailand and is now obscured by a thick cover of Tertiary sediments. However, two observations are against this model. First, there are well penetrations within the Gulf of Thailand that have penetrated Mesozoic non-marine red beds and, second, similar Late Jurassic– Early Cretaceous non-marine red bed sequences occur even further south in eastern Peninsular Malaysia. In western Thailand, within the marine Mae Moei Group, late Toarcian, Aalenian –Bajocian and Oxfordian ammonites were noted by Chonglakmani (1982), and Beauvais & Fontaine (1993) noted the presence of the coral Montlivaltia numismalis, a taxon considered a marker for the Bathonian. Meesook & Grant Mackie (1993), in a study of the marine Jurassic faunas of western Thailand, noted that previous reports of Late Jurassic (Oxfordian) ages could not be supported and that the faunas were mainly Toarcian –Bajocian; that is, latest Early to middle Middle Jurassic (see also Meesook & Grant-Mackie 1996). Further west in eastern Burma, Bender (1983) also recorded marine Jurassic sediments. To the east in southern Laos, Kampuchea and central Vietnam, Early and Early– Middle Jurassic sediments have also been noted, but here

MESOZOIC RED BEDS FROM SE ASIA

the inferred Late Jurassic sediments are red beds and are associated with Kimmeridgian basalts (Hoffet 1933; Hoffet & Le Maitre 1939; Fontaine 1962; Tien 1991). Interestingly, throughout most of these areas the Callovian and Oxfordian, and possibly most of the Late Jurassic, appears to be missing. All microfossils (pollen and spores) and macrofossils (vertebrates and bivalves) found to date in the Khorat Group are interpreted as non-marine (Racey et al. 1994, 1996) and would not therefore fit with a marine delta model of any age (Jurassic or Cretaceous). Moreover, no evidence of marine diagenesis or marine sedimentary structures or trace fossils have been noted in Khorat Group sediments to date.

Recent vertebrate data Buffetaut et al. (2006) have recently noted that the vertebrate fauna from the Phu Kradung Formation of NE Thailand is very similar to that from the Upper Shaximiao Formation of Sichuan in southern China which is assigned a Late Jurassic age. Buffetaut et al. also noted that the fauna of the Sao Khua Formation is in part similar to that of the Napai Formation from Guangxi in southern China, and the Khok Kruat Formation vertebrates are very similar to the Aptian –Albian forms from China. E. Buffetaut (pers. comm.) suggested that in terms of time there is a wider gap between the Phu Kradung and Sao Khua vertebrate assemblages than between those of the Sao Khua and Khok Kruat assemblages. This may indicate that if the Khok Kruat Formation is Aptian then it is more likely that the Sao Khua Formation is Barremian (the inference being that the Phu Kradung Formation would be earliest Cretaceous (Berriasian) or Latest Jurassic (Tithonian).

Sedimentation rates If the Khorat Group was mainly Jurassic in age then it is difficult to explain how red bed sedimentation could have continued uninterrupted from Late Triassic to Early Cretaceous time over an interval of c. 90 Ma with no major breaks or unconformities yet only accumulated around 4500 m of sediment (a rate of 49 m Ma21), which appears low for this type of depositional system. On the other hand, if the Khorat Group, including the Upper Nam Phong Formation, is mainly confined to the Late Jurassic –Early Cretaceous and was deposited over a 49 Ma interval (excluding the Lower Nam Phong Formation dated as Late Triassic) then this represents a sedimentation rate of c. 92 m Ma21, which appears to be more realistic.

53

Palaeocurrents Palaeocurrent data for the Khorat Plateau from outcrops presented by Heggemann et al. (1994) and Racey et al. (1996) indicate that the Nam Phong Formation had both eastward- and westwarddirected palaeocurrents (Fig. 6), whereas in the Phu Kradung to Phu Phan Formations they are mainly directed towards the SE and west (Figs 7–9), implying sediment sourcing from the NE and east. The Khok Kruat palaeocurrents tend to be directed towards the west (Fig. 10). Thus, based on palaeocurrents, the Nam Phong Formation differs from all the overlying formations that traditionally make up the Khorat Group, possibly because of the continued influence of Triassic extensional faults causing uplift and erosion. Of the palaeocurrents shown only the three data points from the western margin of the Khorat Plateau are known to be from the Lower Nam Phong Formation and therefore of Late Triassic age, though interestingly these are all directed eastwards. It is not clear in the field whether the other data shown are from the Upper or Lower Nam Phong Formation. Mouret (1994) noted that the pale-coloured sandstones (Phra Wihan and Phu Phan Formations) thicken ‘upstream’ towards the NE with a concomitant thinning of the interbedded floodplaindominated red beds of the Phu Kradung and Sao Khua Formations. This he noted is coupled with an increase in the size and abundance of quartz pebbles in the same direction. Mouret (1994) suggested that the sandstones are probably diachronous and may be younger in Thailand than in Vietnam to the east. He further suggested that early Kimmeridgian tectonics in north and south Vietnam predates the onset of deposition of the pale-coloured sandstones of the Phra Wihan Formation, which would agree with the revised Early Cretaceous age for this formation. Heggemann (1994) also suggested that the Khorat Group sequence changes from being more proximal in the east to distal in the west of the Khorat area. The Vientiane Basin and Pak Lay area to the north have been studied by Lovatt-Smith & Stokes (1997) and Stokes et al. (1996), respectively. The Mesozoic red bed fill in this basin effectively represents the northern part of the Khorat Basin although the formation names change to use Lao terminology (Fig. 5). Palaeocurrents in the Salavan and Lamo Formations, which are broadly equivalent to the Phu Kradung and Phra Wihan Formations (Fig. 5), in the Savannakhet Basin of Laos to the immediate east of the Khorat Basin are dominantly towards the SW and west (Figs 7 and 8). An unconformity

54

A. RACEY

Fig. 6. Palaeocurrent data for the Late Triassic Nam Phong Formation and its equivalents. (Note a dominance of both eastward- and westward-directed palaeocurrents.) Data sources: Heggemann et al. (1994); A. Racey et al. (unpubl. data).

MESOZOIC RED BEDS FROM SE ASIA

Fig. 7. Palaeocurrent data for the latest Jurassic? to Early Cretaceous Phu Kradung Formation and its equivalents. Palaeocurrents appear to be dominantly coming from the north to NE. Data sources as for Figure 6.

55

56

A. RACEY

Fig. 8. Palaeocurrent data for the Early Cretaceous Phra Wihan and Sao Khua Formations and their equivalents. Palaeocurrents appear to be flowing mainly from the NE, with some also from the east. Data sources as for Figure 6.

MESOZOIC RED BEDS FROM SE ASIA

57

Fig. 9. Palaeocurrent data for the Early Cretaceous Phu Phan Formation and its equivalents. Sediment transport directions are dominantly from the NE, with a minor component from the east. Data sources Hegemann et al. (1994), Racey et al. (1994, 1996) and A. Racey et al. (unpubl. data).

58

A. RACEY

Fig. 10. Palaeocurrent data for the Early Cretaceous Khok Kruat Formation. Although outcrops of this formation are limited, thus restricting the amount of palaeocurrent data for interpretation, the dominant sediment transport direction appears to have changed to being from the west.

MESOZOIC RED BEDS FROM SE ASIA

or disconformity at the top of the Salavan Formation is associated with a change in provenance, with palaeocurrents flowing mainly to the SW within the overlying fluvial sandstones of the Bang Fai Formation (Phu Phan equivalent) as shown in

59

Figure 9. Mouret et al. (1993) and Mouret (1994) have also suggested the presence of a minor unconformity at the base of the Phu Phan Formation based on changes in lithology and palaeocurrents coupled with seismic stratigraphy.

Fig. 11. Distribution of Mesozoic red bed sequences of Late Triassic to Early Cretaceous age and the effect of major lateral displacement along the Red River Fault system coupled with a clockwise rotation of Indochina relative to China. This, coupled with other data discussed in the text, suggests that the original source area for the Khorat Group may have been far to the north in southern China as initially suggested by Carter & Bristow (2002).

60

A. RACEY

Because the main palaeocurrents in the Khorat Group flowed towards the west and SW, it is difficult to envisage how these could have been associated with the Indochina–Sibumasu collision in the Late Triassic with subsequent uplift and erosion throughout the ‘Jurassic’, as we would expect a significant number of palaeocurrents to also flow towards the east and NE; that is, sediment would be shed from both sides of the suture or collision.

Palaeomagnetic data The palaeomagnetic data of Yang & Besse (1993) suggest that only a minor rotation of Thailand has occurred with respect to South China in the Tertiary. Yang & Besse noted a large discrepancy between the poles for Indochina and other published data for the South China Block for both the Triassic – Jurassic and Jurassic –Cretaceous boundaries and calculated that 1500 + 800 km of post middle Cretaceous left lateral slip has occurred along the Red River and Xian –Shui –He fault zones with an approximately 14 + 78 clockwise rotation of Indochina relative to South China (Fig. 11).

The North China Block, South China Block, Yunnan and Indochina were probably in contact with each other since the Late Triassic and then they moved slowly northwards relative to Eurasia until the Cretaceous (Enkin et al. 1992; Yang et al. 1992). The calculated relative pole positions for both the Khorat and Yunnan Basins show no significant change in latitude during the Late Triassic to Early Cretaceous whereas postCretaceous displacement between the Khorat Basin of NE Thailand and both the central Yunnan Basin and South China Block is implied. Yang & Besse (1993) noted that the various blocks to the west of the stable South China Block (e.g. around Markam (Si Mao Basin), the Song Pan Ganze area and the Yunnan Basin) all display similar strong clockwise rotations, suggesting that they, like Indochina, were subsequently displaced, probably along left lateral strike-slip faults (the Xian –Shui –He and Red River Faults or their predecessors) as narrow tectonic slivers. This tends to broadly support the general extrusion model for the Tertiary collision of the Indian Plate with Eurasia proposed by Peltzer & Tapponier (1988).

Fig. 12. Variations in rock fragment type between the various formations of the Khorat Group and the Triassic Huai Hin Lat and Nam Phong Formations. The Huai Hin Lat and Nam Phong Formations (stippled) show a marked volcanic component compared with the remainder of the Khorat Group. Within the Phu Kradung to Phu Phan field the area overlapping with the Nam Phong and Huai Hin Lat fields dominantly belongs to the Phu Kradung Formation. The Khok Kruat Formation appears to have a different composition from the other formations.

MESOZOIC RED BEDS FROM SE ASIA

Carter & Bristow (2002), using palaeomagnetic data from Yang & Besse (1993) and subsequent estimates of left lateral Tertiary displacement of between 500 and 1300 km (from Leloup et al. 1995; Sato et al. 1999) along the Indochina– Sibumasu suture, have shown that in palaeogeographical restorations the pre-extrusion location of the Khorat Basin was more probably within South China close to the Sichuan Foreland Basin. Insamut et al. (1995) undertook a palaeomagnetic study of the Late Cretaceous –Early Tertiary Phu Tok Formation red bed sequence, which overlies the Albian –Cenomanian Maha Sarakham Formation in northern Thailand. Their data suggested that the Phu Tok Formation was deposited at a palaeolatitude of 20– 308N (i.e. close to southern China). Moreover, they interpreted a clockwise rotation of the palaeopoles for ‘Indochina’, which suggests at least a 1000 km sinistral displacement along the Red River Fault, possibly associated with the collision between India and Eurasia, further supporting the extrusion model for the Himalayas collision.

61

Provenance Racey et al. (1996) noted that, based on thinsection modal analysis and whole-rock chemistry of outcrop samples, the Huai Hin Lat and Nam Phong Formations contained a significant volcanic component unlike the overlying formations (Fig. 12). The Nam Phong Formation samples are most probably from the Lower Nam Phong Formation, which has in part yielded Rhaetian palynomorphs. In terms of trace element chemistry the Nam Phong and Huai Hin Lat Formations show a depletion in Fe2O3 þ MgO and an enrichment in Na2O possibly indicative of a calc-alkaline volcanic component (Fig. 13). A minor albeit very diluted volcanic component was also noted by Racey et al. (1996) for the lower part of the overlying Phu Kradung Formation, which may either represent a minor contemporaneous volcanism at this time or minor reworking of Nam Phong lithologies. The Phu Kradung to Phu Phan Formations are dominated by metamorphic quartz (in contrast to the underlying Nam Phong Formation). Racey et al.

Fig. 13. Variations in whole-rock chemistry between the various formations of the Khorat Group and the Triassic Huai Hin Lat and Nam Phong Formations. The depletion in Fe2O3 þ MgO and enrichment in Na2O for the Nam Phong and Huai Hin Lat Formations (stippled) supports a partial cak-alkaline volcanic origin for these sediments, as suggested in Figure 12. The Khok Kruat Formation contains more feldspar (supported by Fig. 14), and hence has a higher Na2O content, suggesting a change in provenance for this uppermost formation of the Khorat Group.

62

A. RACEY

Fig. 14. Variations in quartz, lithic fragments and feldspar content between the various formations of the Khorat Group and the Triassic Huai Hin Lat and Nam Phong Formations. Most of the formations significantly overlap in this figure although the Khok Kruat Formation stands out as being significantly enriched in feldspar.

(1996) originally suggested that the bulk of these sediments (Phu Kradung to Phu Phan Formations) may have been derived from the Kontum Massif to the east with some input from the Louangphabang and Truongson Belts and Rao Cao Massif to the north and NE (i.e. the Annamitic Mountain Belt). However, Carter & Bristow (2002) have suggested that the palaeogeographical position of the Khorat Basin in Early Cretaceous times was more probably much closer to the Sichuan Foreland Basin to the north; this is situated to the immediate south of the Qinling Orogenic Belt, which may have provided a more likely sediment source area for the Khorat Group sediments. The Qinling Orogenic Belt would form a suitable mature hinterland, being dominated by 250 Ma granitic–plutonic, rocks and this would support the petrographic data of Racey et al. (1996), which indicated that much of the Khorat Group was dominated by metamorphic quartz and metamorphic rock fragments, suggesting sourcing from an area of regionally metamorphosed crystalline basement. The Khok Kruat Formation at the top of the Khorat Group has a broadly similar provenance but additionally contains a large proportion of detrital feldspar, indicating a possible

igneous component as shown in Figure 14. The Khok Kruat Formation also differs from the underlying formations of the Khorat Group in having westerly directed palaeocurrents (Fig. 10) and in showing minor evidence of possible marine influence based on the presence of bi-directional ripple trains, which may indicate partial deposition in a marine-influenced sand flat, although the formation is dominantly fluvial in aspect. If there were a Late Jurassic–Early Cretaceous ‘Cimmerian’ event or collision then one might expect to see some evidence of contemporaneous volcanism at around Phu Kradung times. However, detrital mineralogy (and palaeocurrents) suggests that sediment sourcing was mainly from the east and NE (i.e. away from any potential continuing collision to the west, and is dominated by metamorphic quartz. Racey et al. (1997b), in a fission-track study of the Khorat Group, noted that apatites within the Phu Kradung, Phra Wihan and Phu Phan Formations were relatively rich in chlorine (up to 3 wt%), suggesting a partial volcanic source (chlorine-poor apatites would indicate a metamorphic or granitic source, which we know existed to the east and NE and supplied much of

MESOZOIC RED BEDS FROM SE ASIA

the sediment to the Khorat Basin). The chlorine-rich apatites were observed to be fairly euhedral, possibly indicating derivation from a nearby volcanic source. Moreover, many of the apatites collected from the Phu Phan Formation near the centre of the Khorat Plateau (sample Ap8 in fig. 1 of Racey et al. 1997b) showed many dislocations, a feature commonly observed in ‘chilled’ apatites from a tuffaceous source (I. Duddy, pers. comm.). Combined these observations on the apatites may indicate minor near-contemporaneous volcanism at this time (i.e. latest Jurassic to Early Cretaceous). Radiometric dating of detrital micas from the Phu Kradung and Phra Wihan Formations gives a Devonian –Carboniferous age of 330 –380 Ma for the last intrusive or metamorphic event affecting the source area (Heggemann et al. 1994). This suggests that the source area for the micas was granitic or high-grade metamorphic rocks, with the likely source area proposed by Heggemann et al. (1994) being the Devonian –Carboniferous ‘Variscides’ of North –Central Vietnam and Laos, although the Qinling Orogenic Belt in southern China to the north could also be a potential source area. This would not agree with the idea of a Mesozoic metamorphic event (‘Indosinian’) in the sediment source area as proposed by Bunopas & Vella (1978), Ridd (1978) and Maranate & Vella (1986). However, this is based on a limited database of only two samples, one from the Phu Kradung Formation and one from the Phra Wihan Formation.

63

‘limestones’ within the Phu Kradung and Sao Khua Formations are now referred to calcretes, and early reports of marine bivalves and ichthyosaurs have been discounted (see Racey et al. 1996, for details).

Fission-track analysis

K –Ar ages of igneous rocks reported by Stokes et al. (1996) from the Pak Lay Fold Belt area have shown the presence of folded and thrust Jurassic rocks including basalts dated at 152.5 + 6.3 Ma (Late Jurassic – Kimmeridgian) and 167.2 + 5.2 Ma (Middle Jurassic –Bajocian). Moreover, Stokes et al. have also redated the Ban Samxong granite (previously considered to be Triassic) which partially intrudes these Jurassic sequences as 117 + 3.0 Ma (i.e. Aptian). Consequently, there was some extrusive volcanism in the region in the Late Jurassic, whereas Aptian granites may have been associated with the mid-Cretaceous inversion event shown in Figure 3.

Carter et al. (1995) used fission-track dating to date the Phra Wihan Formation at 125 + 20 Ma (basal Berriasian to Albian). However, this was based on a single sample (T90/17). Assuming the analysis for separating single grain ages is correct, their data indicate a possible age range of 165–85 Ma (Bathonian– Santonian). Six other zircon samples from the Phra Wihan and Phu Kradung Formations were also shown by Carter et al. to have a significant spread in single grain ages, although no information on the youngest age mode of these samples was presented. No other results were discussed in detail, although a younger age mode of 133 + 13 Ma (159 –107 Ma, Callovian –Albian) was claimed for the Khok Kruat Formation and 160 + 6 Ma (172 – 148 Ma, Aalenian –Tithonian) for a Sao Khua sample, but no further data were provided to support these age assignments. Racey et al. (1997b) processed 16 outcrop samples for apatite fission-track analysis from the Phra Wihan, Phu Kradung and Phu Phan Formations, of which four yielded no apatite and a further four samples had low apatite abundance. An additional seven outcrop samples from Laos were also analysed. The samples showed that temperatures of 120 8C had been exceeded during burial and that cooling had commenced from Early –Mid Tertiary (65 –45 Ma) with cooling possibly starting c. 10 Ma earlier in the Savanakhet Basin of Laos to the east. Although it was not possible to calculate a palaeogeothermal gradient at the time of maximum palaeotemperature, the application of any realistic heat flow value would indicate the erosion of at least 3.5 km of section since early to mid-Tertiary, with much of this occurring in the mid –late Tertiary (35 –25 Ma). This amount of erosion is in line with the 3–3.5 km of erosion of post Phra Wihan sediments in the vicinity of the Phu Phan Anticline since the Palaeocene suggested by Mouret et al. (1993) based on the truncation of at least 3200 m of section observed on various seismic lines in the area.

Diagenesis

Sea-level changes

Petrographic studies of the Khorat Group have revealed a complete absence of marine cements, with the calcite and ferroan calcites observed being of clear pedogenic origin. All diagenetic processes observed in this sequence were indicative of deposition in an arid to semi-arid continental clastic setting (Canham et al. 1996). Early reports of

Mouret et al. (1993) have attempted to correlate parts of the Khorat Group with sea-level shifts, correlating the distinctly coarser and sandier Phu Phan Formation with a relative sea-level fall at 128.5 Ma (Barremian) and the Phra Wihan Formation with a relative sea-level fall at 177 Ma (Aalenian). These dates would not fit with the new

Volcanism – radiometric dating

64

A. RACEY

vertebrate and palynology dating. However, there is a wide range of tectonic and climatic factors that can cause a change in fluvial style leading to the deposition of these coarser, higher energy, more quartz-rich formations, and there is additionally no direct field evidence for incision caused by base-level changes.

Seismic data Within the general literature there has been much confusion over the term Indosinian, with some workers using it to refer to Permian and others to Triassic events. Much of this confusion has arisen from a failure to recognize that there are two separate closely spaced Triassic orogenic events as indicated by seismic data from the Khorat Plateau area (J. E. Booth, pers. comm.). The first, referred to as ‘Indosinian I’ occurs at the top of the Permian sequence, prior to Triassic rifting and deposition of the Carnian –Norian Huai Hin Lat Formation (assigned to the Kuchinarai Group by subsurface workers). This represents major inversion and uplift of the entire basin with associated deep erosion and peneplanation together with the development of thick- and thin-skinned thrust systems within the Permian. A second event or unconformity referred to as ‘Indosinian II’ separates the Huai Hin Lat Formation from a Rhaetian ‘Lower Nam Phong unit’. This represents a second period of inversion and uplift of the basin with local folding and thinskinned thrusting. These two Indosinian orogenies are separated by a period of extension during which deep half-grabens were formed across much of the region and filled with fluvial and lacustrine sediments and volcanic rocks of the Huai Hin Lat Formation. A third unconformity (Cimmerian or Indosinian III) may also be recognized at the end of the Triassic between the ‘Lower Nam Phong’ and ‘Upper Nam Phong’ Formations, and is recognized as a significant angular unconformity along the southern edge of the basin and as a nonconformity elsewhere in the basin. To the east of the area, in western Laos, a major Cimmerian uplift event is also seen on seismic sections (E. Fenk, pers. comm.), although here there is some uncertainty as to the exact placement of the Nam Phong Formation with respect to this unconformity in that it could be picked above or below base Khorat Group. Yin & Harrison (1996) have identified a Late Jurassic orogeny in southern China (mistakenly referred to as Indosinian), and a similar Kimmeridgian age event is also recognized in Vietnam.

Conclusions There is mounting evidence, as outlined herein, that much of the Khorat Group is Early Cretaceous in

age and is separated by a hiatus spanning much of the Jurassic from the Late Triassic. In addition, there is strong evidence from biostratigraphy, provenance and subsurface data (seismic) that the Lower Nam Phong Formation should be excluded from the Khorat Group. A foreland basin setting for the Khorat Group, most probably associated with flexural subsidence at the front of a ?late Jurassic orogenic belt, would be more in keeping with the observed broad lateral extent, relative uniform thickness and mineralogical maturity of the Phu Kradung to Khok Kruat Formations. This model would also better fit the calculated sedimentation rates when the revised Early Cretaceous age for much of the Khorat Group is considered. The mechanism for basin formation is still unclear but some (e.g. Carter & Bristow 2002) have suggested that a mechanism for this marked Early Cretaceous erosion may have been the collision between the Lhasa Block and Eurasia, with a possible sediment source area being the Qinling Orogenic Belt to the north in China. E. Buffetaut, S. Polachan and K. N. Sattayarak are thanked for discussions and guidance in the field over the last 15 years. J. Booth is thanked for discussions on the stratigraphy of the Triassic. BG Group is thanked for permission to publish, although the views expressed herein are solely the responsibility of the author.

References B AUM , F., VON B RAUN , E., H AHN , L., H ESS , A., K OCH , K. E. & K RAUS , G. 1970. On the geology of Northern Thailand. Beilage Geologisches Jahrbuch, 102, 3 –24. B EAUVAIS , L. & F ONTAINE , H. 1993. Montlivaltia numismalis (D’Orbigny) a Middle Jurassic coral newly found in West Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology (BIOSEA), I. Chiang Mai University, Chiang Mai, 463. B OOTH , J. E. 1998. The Khorat Plateau of NE Thailand— exploration history and hydrocarbon potential. SEAPEX Exploration Conference, Singapore, 169–203. B UFFETAUT , E., S UTEETHORN , V., M ARTIN , V., C HAIMANEE , Y. & T ONG -B UFFETAUT , H. 1993. Biostratigraphy of the Mesozoic Khorat Group of northeastern Thailand: The contribution of vertebrate palaeontology. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology (BIOSEA), I. Chiang Mai, University, Chiang Mai, 51– 62. B UFFETAUT , E., S UTEETHORN , V., L E L OEUFF , J., K HANSUBHA , S., T ONG , H. & W ONGKO , K. 2005. The dinosaur fauna of the Khok Kruat Formation (early Cretaceous) of Thailand. In: W ANNAKAO , L., Y OUNGME , M., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources

MESOZOIC RED BEDS FROM SE ASIA of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 575–581. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y.-N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19–37. B UFFETAUT , E., C UNY , G., LE L OEUFF , J. & S UTEETHORN , V. 2009. Late Palaeozoic and Mesozoic continental ecosystems of SE Asia: an introduction. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 1– 5. B UNOPAS , S. & V ELLA , P. 1978. Late Palaeozoic and Mesozoic structural evolution of Northern Thailand, a plate tectonic model. In: N UTALAYA , P. (ed.) Proceedings of the Third Regional Conference on the Geology and Mineral Resources of Southeast Asia. Department of Mineral Resources, Bangkok, 133–140. C ANHAM , A. C., R ACEY , A., L OVE , M. A. & P OLACHAN , S. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group Northeastern Thailand: Part 2, Diagenesis and reservoir quality. Journal of Petroleum Geology, 19, 321– 338. C ARTER , A. & B RISTOW , C. S. 2002. Linking hinterland evolution and continental basin sedimentation using detrital zircon thermochronology: A study of the Khorat Plateau basin, eastern Thailand. Basin Research, 15, 271–285. C ARTER , A., B RISTOW , C. S. & H URFORD , A. J. 1995. The application of fission track analysis to the dating of barren sequences: Examples from red beds in Scotland and Thailand. In: D UNAY , R. E. & H AILWOOD , E. A. (eds) Non-biostratigraphical Methods of Dating and Correlation. Geological Society, London, Special Publications, 89, 57– 68. C HONGLAKMANI , C. 1982. The Mesozoic marine faunas of Thailand and their palaeogeographic implications. Bolletino della Societa Paleontologica Italiana, 21, 255–266. C OOPER , M. A., H ERBERT , R. & H ILL , G. S. 1989. The structural evolution of Triassic intermontane basins in Northeast Thailand. In: T HANASUTHIPITAK , T. & O UNCHANUM , P. (eds) Proceedings of the International Symposium on Intermontane Basins: Geology and Resources. Chiang Mai University, Chiang Mai, 231–242. E NKIN , R., Y ANG , Z., C HEN , Y. & C OURTILLOT , V. 1992. Paleomagnetic constraints on the geodynamic history of China from Permian to Present. Journal of Geophysical Research, 97, 13953– 13989. F ONTAINE , H. 1962. Gisements fossilife`res du bassin houillier du Quang-Nam (centre du Viet Nam). Travaux de Ge´ologie, 1, 33– 52. F ONTAINE , H. 1964. Calcaires de la re´gion de Treng (Province de Battambang, Cambodge Occidental). Archives Ge´ologiques du Viet Nam, 6, 29– 36. F ONTAINE , H. & S UTEETHORN , V. 1988. Late Paleozoic and Mesozoic fossils of West Thailand and their environments. CCOP Technical Bulletin, 20, 1 –217. F ROMAGET , J. 1937. Etudes ge´ologiques sur le nordouest du Tonkin et le nord du Haut Laos, 1e`re partie.

65

Bulletin du Service Ge´ologique de l’Indochine, 23, 1– 170. H EGGEMANN , H. 1994. Sedimenta¨re Entwicklung der Khorat-Gruppe (ober-Trias bis Paleogen) in NE- und N-Thailand. Go¨ttinger Arbeiten zur Geologie und Palaeontologie, 63, 1 –146. H EGGEMANN , H., H ELMCKE , D. & T IETZE , K.-W. 1994. Sedimentary evolution of the Mesozoic Khorat Basin in Thailand. Zentralblatt fu¨r Geologie und Pala¨ontologie, 1, 1267– 1285. H OFFET , J. H. 1933. Etude ge´ologique sur le centre de l’Indochine entre Tourane et le Me´kong (Annam Central et Bas Laos). Bulletin du Service Ge´ologique de l’Indochine, 20, 1 –154. H OFFET , J. H. & L E M AITRE , D. 1939. Sur la stratigraphic et la pale´ontologie du Lias des environs de Tchepone (Bas Laos). Comptes Rendus de l’Acade´mie des Sciences, 209, 114–116. H UTCHISON , C. S. 1989. Geological Evolution of South East Asia. Oxford Science Publications, Monographs on Geology and Geophysics, 13. I NSAMUT , S., C HARUSIRI , P., H UANG , Z. & D AOREK , V. 1995. Palaeomagnetic result of Phu Tok red beds of NE Thailand: Implication for Mesozoic tectonic history of SE Asia. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 73–78. J IRIN , S. & M ORLEY , R. J. 1994. Lower Cretaceous palynomorphs from the Termus and Mangking formations (Tembeling Group) Peninsular Malaysia: Their stratigraphic and climatic significance. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 206. K LOMPE , T. H. F. 1962. Igneous and structural features of Thailand. Geologische Mijnbouwkundige Dienst van Suriname, Mededeling, 41, 290–302. K OBAYASHI , T. 1984. Mesozoic bivalvia of the Khorat Group with a note on the Trigonoidacea. Geology and Paleontology of Southeast Asia, 25, 285–288. L ELOUP , P. H., L ACASSIN , R. ET AL . 1995. The Ailo Shan– Red River shear zone (Yunnan, China) Tertiary transform boundary of Indochina. Tectonophysics, 252, 3– 84. L OVATT -S MITH , P. F. & S TOKES , R. B. 1997. Geology and petroleum potential of the Khorat Plateau Basin in the Vientiane area of Lao P. D. R. Journal of Petroleum Geology, 20, 27–50. L OVATT -S MITH , P. F., S TOKES , R. B., B RISTOW , C. & C ARTER , A. 1996. Mid-Cretaceous inversion in the Northern Khorat Plateau of Lao PDR and Thailand. In: H ALL , R. & B LUNDELL , D. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 97– 122. M ACDONALD , A. S. & B ARR , S. M. 1984. The Nan River mafic–ultramafic belt, northern Thailand: Geochemistry and tectonic significance. Geological Society of Malaysia Bulletin, 17, 209–224. M ARANATE , S. & V ELLA , P. 1986. Palaeomagnetism of the Khorat Group, Mesozoic, Northeast

66

A. RACEY

Thailand. Journal of Southeast Asian Earth Sciences, 1, 23– 31. M EESOOK , A. & G RANT -M ACKIE , J. A. 1993. Facies, faunas and biostratigraphy of the marine Jurassic of Thailand. In: T HANASATHUPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology, I, Chiang Mai University, Chiang Mai, Thailand, 463. M EESOOK , A. & G RANT -M ACKIE , J. A. 1996. Marine lithostratigraphy of Thailand. Journal of Southeast Asian Earth Sciences, 14, 377–391. M EESOOK , A., S UTEETHORN , V., C HADUMRONG , P., T EERARUNGSIGUL , N., S ARDSUD , A. & W ONGPRAYOON , T. 2002. Mesozoic rocks of Thailand: A summary. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Reesources, Bangkok, 82–94. M ETCALFE , I. 1996. Pre-Cretaceous evolution of SE Asian terranes. In: H ALL , R. & B LUNDELL , D. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 97– 122. M OURET , C. 1994. Geological history of northeastern Thailand since the Carboniferous. Relations with Indochina and Carboniferous to Early Cenozoic evolution model. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245– 252. M OURET , C., H EGGEMANN , H., G OUADAIN , J. & K RISADASIMA , S. 1993. Geological history of the siliciclastic Mesozoic strata of the Khorat Group in the Phu Phan Range area, northeastern Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Palaeontology, I. Chiang Mai University, Chaing Mai, 23–49. P ELTZER , G. & T APPONIER , P. 1988. Formation and evaluation of strike-slip faults and basins during the India–Asia collision, an experimental approach. Journal of Geophysical Research, 93, 15085–15117. P ENG , K. H. 1983. Mesozoic stratigraphy in Peninsular Malaysia. In: N UTALAYA , P. (ed.) Proceedings of the Workshop on Stratigraphic Correlation of Thailand and Malaysia. Geological Society of Thailand and Geological Society of Malaysia, Bangkok, 370– 383. R ACEY , A. & G OODALL , J. G. S. 2009. Palynology and stratigraphy of the Mesozoic Khorat Group of NE Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 67– 81. R ACEY , A., G OODALL , J. G. S., L OVE , M. A. & J ONES , P. D. 1994. New age data on the Khorat Group of Northeast Thailand In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245– 252. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S. & P OLACHAN , S. 1996. Stratigraphy and

reservoir potential of the Mesozoic Khorat Group, North Eastern Thailand: Part 1, Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 18, 5 –39. R ACEY , A., S TOKES , R. B., L OVATT -S MITH , P. & L OVE , M. A. 1997a. Late Jurassic collision in Northern Thailand and the significance of the Khorat Group. In: D HEERADILOK , P., H INTHONG , C. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific, 2. Department of Mineral Resources, Bangkok, 412– 423. R ACEY , A., D UDDY , I. R. & L OVE , M. A. 1997b. Apatite fission track analysis of Mesozoic red beds from Northeastern Thailand and Western Laos. In: D HEERADILOK , P., H INTHONG , C. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific, 2. Department of Mineral Resources, Bangkok, 2, 200– 209. R IDD , M. F. 1978. Thailand. In: M OULLADE , M. & N AIRN , E. M. (eds) The Phanerozoic Geology of the World, the Mesozoic. Elsevier, Amsterdam, 145– 163. S ATO , K., L IU , Y., Z HU , Z., Y ANG , Z. & O TOFUJI , Y. 1999. Palaeomagnetic study of Middle Cretaceous rocks from Yunlong, Western Yunnan, China: Evidence of southward displacement of Indochina. Earth and Planetary Science Letters, 165, 1– 15. S ATTAYARAK , N., S RIGULAWONG , S. & P ATARAMETHA , M. 1991a. Subsurface stratigraphy of the non-marine Khorat Group, Northern Thailand. In: P OLACHAN , P., T HANVARACHORN , P. & P ISUTHA A RNOND , V. ET AL . (eds) Seventh Regional Conference on Geology, Mineral Resources and Energy Resources of Southeast Asia. GEOSEA VII, 5–8 November 1991, Bangkok, Thailand. Geological Society of Thailand, Bangkok, Thailand. S ATTAYARAK , N., P OLACHAN , S. & C HARUSIRISAWAD , R. 1991b. Cretaceous rock salt in the northeastern part of Thailand. In: P OLACHAN , P., T HANVARACHORN , P. & P ISUTHA -A RNOND , V. ET AL . (eds) Seventh Regional Conference on Geology, Mineral Resources and Energy Resources of Southeast Asia. GEOSEA VII, 5– 8 November 1991, Bangkok, Thailand. Geological Society of Thailand, Bangkok, Thailand. S AURIN , E. 1935. Etudes ge´ologiques sur l’Indochine du Sud-Est (sud Annam, Cochinchine, Cambodge oriental). Bulletin du Service Ge´ologique de l’Indochine, 22, 1 –419. S AURIN , E. 1944. Etudes ge´ologiques sur le CentreAnnam me´ridonal. Bulletin du Service Ge´ologique de l’Indochine, 27, 1 –210. S AURIN , E. 1950. Etudes ge´ologiques sur la province de Luang Prabang et la de´le´gation de Bun-Tai. Bulletin du Service Ge´ologique de l’Indochine, 29, 1 –99. S AURIN , E. 1956. Lexique Stratigraphique International. III. Indochina. CNRS, Paris. S INGHARAJWARAPAN , S. & B ERRY , R. 2000. Tectonic implications of the Nan suture zone and its relationship to the Sukhotai fold belt, northern Thailand. Journal of Southeast Asian Earth Sciences, 18, 663– 673. S MILEY , C. J. 1970a. Later Mesozoic flora from Maran, Pahang, West Malaysia. Part 1: Geologic

MESOZOIC RED BEDS FROM SE ASIA considerations. Geological Society of Malaysia Bulletin, 3, 77– 88. S MILEY , C. J. 1970b. Later Mesozoic flora from Maran, Pahang, West Malaysia. Part 2: Taxonomic considerations. Geological Society of Malaysia Bulletin, 3, 89– 113. SNMGP 1972– 1973. Carte ge´ologique de reconnaissance de la Re´publique Khme`re 1:200 000. (Service National des Mines, de la Ge´ologie et du Pe´trole, Re´publique Khme`re). Notices explicatives by O. DOTTIN , B. A LABOUVETTE , ET AL . Bureau de Recherches Ge´ologiques et Minie`res, Orle´ans. S TOKES , R. B., L OVATT -S MITH , P. F. & S SOUMPHNPHAKDY , K. 1996. Timing of the Sibumasu– Indochina collision: New evidence from the Pak Lay Foldbelt of the Lao PDR. In: H ALL , R. & B LUNDELL , D. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 225–232. T HEIN , M. 1973. A preliminary synthesis of the geological evolution of Burma with reference to the tectonic development of Southeast Asia. Geological Society of Malaysia Bulletin, 6, 87–116. T IEN , P. C. 1991. Geology of Cambodia, Lao and Vietnam. Explanatory note to the geological map of Cambodia, Lao and Vietnam at 1:1 000 000 scale, 2nd edn. Geological Survey of Vietnam, Hanoi.

67

T IN , H. F. 1992. The main Triassic events of East Asia. In: Proceedings, 29th International Geology Congress, Kyoto, 24 August 3 September 1992. Geological Society of Japan, 276. VON B RAUN , E. & J ORDAN , R. 1984. The stratigraphy and paleontology of the Mesozoic sequence in the Mae Sot area in Western Thailand. Geologisches Jahrbuch, 21, 5– 51. W AN T. & Z HU , H. 1991. Tectonic events of Late Proterozoic–Triassic in South China. Journal of Southeast Asian Earth Sciences, 6, 147–157. W ORKMAN , D. R. 1977. Geology of Laos, Cambodia, south Vietnam and the eastern part of Thailand. Institute of Geological Sciences, Overseas Geology Mineral Resources Bulletin, 50, 1 –95. Y ANG , Z. & B ESSE , J. 1993. Palaeomagnetic study of Permian and Mesozoic sedimentary rocks from Northern Thailand supports the extrusion model for Indochina. Earth and Planetary Science Letters, 117, 525– 552. Y ANG , Z., C OURTILLOT , V., B ESSE , J., M A , Y., X ING , L., X U , S. & Z HANG , J. 1992. Jurassic palaeomagnetic constraint on the collision of the north and south China blocks. Geophysical Research Letters, 19, 577 –580. Y IN , A. & H ARRISON , T. M. (eds) 1996. Tectonic Evolution of Asia. Cambridge University Press.

Palynology and stratigraphy of the Mesozoic Khorat Group red bed sequences from Thailand ANDREW RACEY1* & JEFFERY G. S. GOODALL2 1

BG Group, 100 Thames Valley Park Drive, Reading RG6 1PT, UK

2

Santos Ltd, Santos Centre, 60 Flinders Street, Adelaide, SA 5000, Australia *Corresponding author (e-mail: [email protected])

Abstract: New palynological results are presented and integrated with previous data for the Mesozoic Khorat Group continental red bed succession of NE Thailand. The Khorat Group is traditionally considered to comprise, from bottom to top, the Nam Phong, Phu Kradung, Phra Wihan, Sao Khua, Phu Phan and Khok Kruat Formations. The group is unconformably overlain by the continental evaporitic Maha Sarakham Formation, which has been palynologically dated as mid-Albian–Cenomanian, therefore giving a minimum age for the underlying Khorat Group. Traditionally the Nam Phong Formation is considered to be Late Triassic, the Phu Kradung, Phra Wihan and Sao Khua Formations are assigned to the Jurassic, and the Phu Phan and Khok Kruat Formations to the Early Cretaceous. This age dating has mainly been based on plant macrofossil and vertebrate studies. Palynology supports the Late Triassic age for the Lower Nam Phong and underlying Huai Hin Lat Formations but suggests an Early Cretaceous (Berriasian–Barremian) age for the Phu Kradung, Sao Khua and Phra Wihan Formations, and an Aptian age for the Khok Kruat Formation. The Phu Phan has yet to yield useful age-diagnostic palynomorphs but must also be Early Cretaceous based on the age of the under- and overlying formations. The key palynomorph present for ascribing an Early Cretaceous age is Dicheiropollis etruscus, a taxon that is restricted to this age interval and has been found in the Phu Kradung (both at outcrop and in subsurface exploration wells), Phra Wihan and Sao Khua Formations. The lithostratigraphy and age of the Nam Phong Formation is, at present, still problematic. From subsurface seismic data it is clear that the Nam Phong Formation comprises two distinct units separated by an unconformity, referred to herein as the Lower and Upper Nam Phong Formations. The age of the Lower Nam Phong Formation based on palynology is Late Triassic (Rhaetian) whereas the age of the Upper Nam Phong Formation is poorly constrained but can be no older than Pliensbachian based on data from the Phu Horm-1 well. The absence of the Early Cretaceous marker D. etruscus may be taken as indirect evidence for a Jurassic age for the Upper Nam Phong Formation in the Phu Horm-1 well. This revised age dating suggests the presence of a significant depositional hiatus within the Nam Phong Formation, and consequently the Lower Nam Phong Formation should be removed from the Khorat Group. Overall the palynomorph assemblages of the Khorat Group are dominated by the gymnosperm pollen Corollina (synonymous with Classopollis) and Dicheiropollis, both of which belong to the Cheirolepidaceae and would indicate deposition in a warm, dominantly seasonally dry subtropical climate.

The study area is located in NE Thailand (Fig. 1). For a detailed overview of the geological setting the reader is referred to Canham et al. (1996), Racey et al. (1996) and Racey (2009). The age of the formations recognized in the lower part of the Khorat Group, namely the Phu Kradung, Phra Wihan and Sao Khua Formations, have been the subject of a series of papers (Racey et al. 1994, 1996, 1997a). A stratigraphical column for the Khorat Group (Fig. 2) presents the revised ages for the various formations that make up the group based on the palynological data described herein. Previous age determinations were mainly based on plant and vertebrate studies (Buffetaut et al. 2009, and references therein) and palynology (Hahn 1982; Sattayarak 1983; Racey et al. 1994,

1996, 1997a). Traditionally the Phu Kradung, Phra Wihan and Sao Khua Formations have been considered to be of Early, Middle and Late Jurassic age, respectively, with the overlying Phu Phan and Khok Kruat Formations assigned to the Early Cretaceous and the underlying Nam Phong Formation to the Late Triassic, based mainly on vertebrate dating. This view has held sway with many Thai geoscientists despite the publications by Racey et al. (1994, 1996), which provided a palynological age of Early Cretaceous for the basal Phra Wihan (arguably topmost Phu Kradung). With subsequent new vertebrate discoveries (see Buffetaut et al. 2006, 2009, and references therein), it has gradually become accepted that the Sao Khua and Phra Wihan Formations are Early Cretaceous in age,

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 69–83. DOI: 10.1144/SP315.6 0305-8719/09/$15.00 # The Geological Society of London 2009.

70

A. RACEY & J. G. S. GOODALL

Fig. 1. Study area and sample localities.

whereas the Phu Kradung Formation is most probably of Late Jurassic to Early Cretaceous age based on the most recent vertebrate discoveries (E. Buffetaut, pers. comm.). New palynological analyses presented herein allow much (if not all) of the Phu Kradung Formation to be dated as Early Cretaceous (Barremian –Berriasian), although some of the lower part of the Phu Kradung Formation could still be Latest Jurassic (Tithonian). These analyses further support the Early Cretaceous age ascribed earlier to the Sao Khua and Phra Wihan Formations (Racey et al. 1994, 1996).

Palynology The Khorat Group comprises a continental red bed succession deposited partly in a semi-arid environment. Consequently, the general preservation of organic matter is often poor and palynomorphs are therefore rare. More than 170 samples have been collected from a wide variety of outcrops (Fig. 1). In general, less than 10% of the many samples processed by the authors have yielded identifiable palynomorphs and of these only a few per cent have contained useful age diagnostic taxa. Moreover,

it is often uncertain which part of a formation particular samples are actually from, owing to the limitations of many of the outcrops, and consequently no attempt has been made to generate range charts for the various taxa within each formation. Throughout this paper we consider Corollina to be synonymous with Classopollis. A spreadsheet showing the distribution, by formation, of the various taxa found to date is provided in Figure 3, and representative illustrations of the more common and age-diagnostic taxa are given as Figures 4–7. The formations can be dated on the basis of palynology as follows, in descending stratigraphical order.

Maha Sarakham Formation This comprises a mixed sequence of interbedded salt, anhydrite and red beds that locally unconformably overlies the Khok Kruat Formation and is, therefore, not considered to be part of the Khorat Group (see Sattayarak et al. 1991; Racey et al. 1994). The formation was deposited in a hypersaline, landlocked lake within an arid continental desert. It has been dated based on palynology from

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP

71

Fig. 2. Stratigraphy and revised age dating of the Khorat Group. Left hand column after Department of Mineral Resources, Thailand (1992) and Sattayarak & Srigulawong (2008).

interbedded mudstones and siltstones (Sattayarak et al. 1991; based on Harris 1977) as middle Albian to Cenomanian. A common background assemblage of Exisipollenites tumulus, Corollina

torosus, Corollina spp., Eucommidites troedssoni, Schizea spp., Ephedra spp. and Cycadopites spp. was recorded by Harris (1977) and would indicate a broad Early Jurassic– Cenomanian age. Rare

A. RACEY & J. G. S. GOODALL

Fig. 3. Palynomorph distribution within the Mesozoic sediments of NE Thailand. Filled boxes indicate presence and vertical lines represent ‘inferred’ presence.

72

occurrences of Verrucosisporites spp., Taurocusisporites cf. reduncus and Caliallasporites spp. were also noted. Of more importance for dating was the presence of the angiosperm pollen Tricolpites (middle to late Albian), Tricolporites (late Albian to early Cenomanian) and Triporites (Cenomanian). An Albian –Cenomanian age is, therefore, considered most likely for this formation. Hansen et al. (2002) have dated some of the salt layers within the Maha Sarakham Formation as Cenomanian using 87Sr/86Sr, K/Ar and K/Ca. Thu (1986) reported the presence of a long-ranging Mesozoic assemblage comprising Gnetaceacopollenites sp., Corollina sp. and Exisipollenites sp. from the basal part of the Thangon Formation (¼ Maha Sarakham Formation) below the first salt bed in the Vientiane area of southern Laos.

Khok Kruat Formation This comprises sandstone, conglomerate, siltstone, shale and intermittent palaeosols deposited in a dominantly fluvial environment with some local paralic or possible marine influence, based on the associated fauna, which includes bivalves, crustaceans and vertebrates. Sattayarak et al. (1991) indicated a latest Aptian age for the upper part of this formation based on palynology, although no floral lists were provided, from subsurface samples analysed by Esso. The overlying Maha Sarakham Formation has yielded Albian –Cenomanian palynomorphs. Buffetaut & Suteethorn (1992) have recorded Aptian –Albian dinosaurs from the Khok Kruat Formation. Only poorly preserved indeterminate palynomorphs have been recovered from outcrops of this formation. The formation is herein considered to be Aptian in age but may extend to the Albian based on the palynology and contained vertebrate fauna.

Phu Phan Formation This formation comprises dominantly medium- to coarse-grained sandstone beds (locally conglomeratic) and subordinate floodplain or lacustrine siltstone and mudstone. The overall depositional setting was in a high-energy, low-sinuosity braided river system. Racey et al. (1996) recorded a rare, long-ranging assemblage of Corollina spp., Cyathidites minor, ?Todisporites sp. plus indeterminate bisaccate pollen. This would indicate an age no younger than Cenomanian, and no older than Early Jurassic. The contact with the underlying Sao Khua Formation is locally erosive and therefore may be unconformable, whereas the contact with the overlying Khok Kruat Formation is conformable. Consequently, the Phu Phan Formation must fall within

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP

73

Fig. 4. Dicheiropollis etruscus large forms (a–d) and small forms (e and f). Magnifications are all 60 except where indicated by scale bar. The photographs show the typical diad form although on some broken specimens the clasps are clearly visible.

74

A. RACEY & J. G. S. GOODALL

Fig. 5. All magnifications are 60 except where indicated by a scale bar. (a) Granodiscus spp., a freshwater to brackish algae. (b) Cyathidites minor. (c) Araucariacites australis. (d) Leptolepidites verrucosus. (e) Bisaccate pollen (undifferentiated). (f) Deltoidospora minor.

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP

Fig. 6. All magnifications are 60 except where indicated by a scale bar. (a) Retriletes spp. (b) Cicatricosisporites spp. (c) Concavissimisporites punctatus. (d) Corollina spp. (e) Corollina spp. (f) Callialasporites dampieri.

75

76

A. RACEY & J. G. S. GOODALL

Fig. 7. All magnifications are 60 except where indicatted by a scale bar. (a) Exisipollenites tumulus. (b) Perinopollenites elatoides. (c) Appendicisporites distocarinatus. (d) Deltoidospora minor. (e) Cicatricosisporites spp. (f) Appendicisporites distocarinatus.

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP

the interval Aptian to Berriasian –Barremian based on the ages of the over- and underlying formations. The absence of Dicheiropollis etruscus (an important age diagnostic taxon; see below) may indicate an age younger than early Barremian. An unconformity between the Phu Phan Formation and underlying Sao Khua Formation has been suggested previously by Maranate & Vella (1986) on the basis of palaeomagnetic data. However, on oil company seismic data there is no obvious evidence of an angular discordance between the two formations (A. Racey, pers. obs.; J. Booth, pers. comm.), although such data cover only a relatively small percentage of the total study area. Mouret et al. (1993) and Mouret (1994) suggested the presence of a minor tectonic unconformity at the base of the Phu Phan Formation, based on their observed changes in lithology and palaeocurrents and on seismic stratigraphy. Based on the ages of the over- and underlying formations, the age of the formation must, therefore, fall within the interval ‘mid’-Barremian to Aptian.

Sao Khua Formation This comprises dominantly floodplain deposits including sandstone, siltstone and mudstone, together with common calcretes, and was deposited in a low-energy fluvial setting comprising meandering channels and extensive flood plains. The contact with the underlying Phra Wihan and overlying Phu Phan Formations appears to be gradational and conformable at outcrop. However, as mentioned above, it is possible that the Phu Phan–Sao Khua contact is locally unconformable. This formation was originally considered to be Late Jurassic on the basis of a rich vertebrate fauna (Buffetaut et al. 2009, and references therein) and Late Jurassic –Early Cretaceous on the basis of palynology (Hahn 1982; Racey et al. 1996). Although palynomorphs originally recorded by Racey et al. (1996) were poor and long-ranging, comprising C. minor and Araucariacitites australis, palynomorph taxa recovered from the underlying Phra Wihan Formation were assigned to the Early Cretaceous, indicating that the Sao Khua Formation must be of a similar or younger age. Samples recently collected and analysed from this formation from Phu Phan Thong (Fig. 1) have yielded an assemblage dominated by D. etruscus, Cicatricosisporites spp., Corollina spp., Cicatricosisporites?/Appendicisporites spp., and Concavissimisporites punctatus, with rarer elements including Steriesporites spp., Cyathidites minor, Leptolepidites verrucatus and ?Pediastrum. A second sample set from this formation from Hui Sai (Fig. 1) yielded common D. etruscus and Corollina sp. (both much more common than at Phu Phan Thong) and rare

77

E. tumulus, A. australis, C. minor, Convolutispora sp., Osmundacidites spp. and Leptolepidites verrucosus. Angiosperm pollen are notably absent in all these samples, suggesting an age no younger than Barremian –Aptian. No spores or pollen indicative of sediments younger than Early Barremian were recorded. Based on the presence of D. etruscus, a Berriasian –early Barremian age is preferred for this formation.

Phra Wihan Formation This comprises fine- to coarse-grained sheet and channelled sandstone beds with rarer variegated siltstone and mudstone, with intermittent conglomerate beds, and was deposited in a fluvial environment dominated by high-energy, shallow braided rivers with subordinate lower energy meandering river systems and associated floodplains. The formation has a gradational and conformable contact with the underlying Phu Kradung and overlying Sao Khua Formations. It was previously assumed to be Middle Jurassic in age based solely on the fact that the vertebrates in the over- and underlying formations were thought to be Early and Late Jurassic in age, respectively. A Middle Jurassic– Early Cretaceous age was indicated based on plants, crustaceans and insects (Heggemann et al. 1990). A Berriasian –Barremian age was proposed on the basis of palynology (Racey et al. 1996) with an assemblage from near the base of the formation comprising D. etruscus, Corollina spp., A. australis, Ichyosporites cf. variegatus, Gleichenidites senonicus, Laevigatisporites spp., Perinopollenites elatoides, Callialasporites dampieri, Anaplanisporites dawsonensis, Apiculatisporites spp., Osmundacidites wellmanii, Todisporites minor, Kraeuselisporites sp., Concavissmisporites sp. and Cicatricosisporites augustus. The key sample was collected from a thin (few centimetres) carbonaceous horizon very near the boundary between the first pale-coloured sandstone beds traditionally assigned to the Phra Wihan Formation and the red siltstone –mudstone beds of the Phu Kradung Formation. Because the formations are conformable, it was assumed that the underlying Phu Kradung Formation was at least in part of similar age (i.e. Early Cretaceous). The Phra Wihan Formation is considered to be Berriasian –early Barremian in age based on the presence of D. etruscus, Corollina spp., Cicatricosisporites augustus and C. dampieri. C. augustus ranges no older than Early Cretaceous (Berriasian) whereas D. etruscus has its stratigraphic top in the early Barremian and its base in the Berriasian. The assemblage is very similar to that of the overlying Sao Khua Formation.

78

A. RACEY & J. G. S. GOODALL

Phu Kradung Formation This comprises fluvial channel sandstone, siltstone and mudstone with intermittent calcrete. Overall, mudstone and siltstone dominate the formation, with calcrete locally common. The formation was deposited in a mainly lake-dominated floodplain cut by meandering and occasionally braided river channels. The formation is sandier in its upper part and shows a gradational conformable contact with the overlying Phra Wihan Formation. The nature of the contact with the underlying Nam Phong Formation is considered by some (N. Sattayarak, pers. comm.) to be unconformable in the subsurface along the SW margin of the Khorat Plateau, suggesting a possible minor hiatus based on a change in interval velocity at this boundary. However, on seismic sections this boundary shows no evidence of angularity and may simply represent a change in depositional setting (J. Booth, pers. comm.) The Phu Kradung Formation was originally dated as Early–Middle Jurassic based on bivalves, fish, crocodiles and turtles (Buffetaut & Ingavat 1984, 1985). The formation is now considered on the basis of recent vertebrate discoveries to be Late Jurassic (i.e. Tithonian; Buffetaut et al. 2006, 2009, and references therein). Racey et al. (1996) originally recorded a rare, long-ranging palynomorph assemblage of C. minor, Baculatisporites commaunensis, C. simplex and indeterminate verrucate spores. Samples from Hui Sai (Fig. 1) from the upper part of the Phu Kradung Formation have yielded common D. etruscus and abundant Corollina spp. plus rarer E. tumulus, Cyathidites spp., Convolutispora spp., L. verrucatus and Osmundacidites spp. A similar assemblage was also recovered from samples collected from the recently widened road cut just east of Nong Bua Lamphu, just west of Udon Thani (Fig. 1). This assemblage is dated as Berriasian to early Barremian based on the occurrence of D. etruscus and absence of taxa younger than early Barremian. The assemblage is strongly dominated by Corollina, a taxon that is rarer in the Sao Khua Formation. E. Buffetaut (pers. comm.) has suggested that the Phu Kradung Formation is probably older than Barremian because the degree of dinosaur evolution he observed between the Phu Kradung and Khok Kruat Formations is unlikely to have occurred solely within the Barremian– Aptian. This led E. Buffetaut (pers. comm.) to conclude that the Phu Kradung Formation is at least partially Late Jurassic (Tithonian) in age. Interestingly, G. Cuny (pers. comm.) has found shark teeth of similar species of Heteroptychodus in both the Sao Khua and Phu Kradung Formations, which he believes are Early Cretaceous. The subsurface palynological data from the Phu Horm-1 well discussed below also support an Early Cretaceous age

for most, if not all, of the Phu Kradung Formation, although a Late Jurassic age cannot as yet be ruled out for its lowermost part (see below).

Nam Phong Formation At outcrop this formation comprises micaceous and conglomeratic to fine-grained sandstone, siltstone and mudstone deposited in a fluvial environment. The succession is dominated by braided and meandering channels interbedded with lacustrine or floodplain sequences. It unconformably overlies the Huai Hin Lat Formation, the contact being marked by the Indosinian II unconformity (see Racey 2009). Racey et al. (1994, 1996) recorded Ovalipollis ovalis and striate bisaccate pollen indicating a Ladinian–Rhaetian age for the Nam Phong Formation from outcrop samples collected from the lower part of the formation. In the Non Sung-1 well, the formation yielded an assemblage comprising Corollina simplex, Inaperturopollenites turbatus, Leiotriletes spp. and Verrucosisporites spp., which was originally assigned a Norian–Rhaetian age. The underlying Huai Hin Lat Formation at outcrop has yielded O. ovallis, C. minor, Corollina sp., Polycingulatisporites spp. and Staurosaccites spp. (Racey et al. 1996), and Haile (1973) recorded, but did not figure, Ovalipollis spp., Cycadopites carpenteri, Alisporites sp., Zebrasporites fimbriatus, Camerosporites sp., Concavissimisporites lunzensis and Verrucosporites sp. from the underlying Huai Hin Lat Formation, which would indicate a Carnian–Norian age. Mouret (1994) subdivided the Triassic in the subsurface in the Phu Phra-1 well into three units: the Phu Noi and Phu Phra Formations, which he grouped informally to form the Kuchinarai Group (broadly equivalent to the Huai Hin Lat Formation); the Phu Lop Group, which formed an additional ‘new’ Late Triassic unit unconformably overlying the Kuchinarai Group; and the Lower Nam Phong Formation of latest Triassic age, which unconformably overlies the Phu Lop Group. He considered the age of the Upper Nam Phong Formation to extend into the Early Jurassic but no supporting dating evidence was presented to support this conclusion. The Nam Phong Formation at outcrop has, to date, yielded only Late Triassic (late Norian to Rhaetian) palynomorphs. However, subsurface seismic and well data permit the identification of a Lower Nam Phong Formation of Late Triassic age and an Upper Nam Phong Formation of possible Jurassic age separated by an unconformity (see below for details).

Palynological data from ‘coeval’ sections in Laos and Cambodia In the Khorat Group equivalent in Laos, Stokes et al. (1996), recorded the following assemblage from red

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP

beds in the Pak Lay Fold Belt (the formation was not specified): Corollina spp., Caliallisporites dampieri, Cerebropollenites mesozoicus, Uvaesporites sp. and Perinopollenites elatoides, to which they assigned a Middle to Late Jurassic age. However, all of these taxa can extend into the Early Cretaceous. Corsin & Desreumeaux (1972), in a study of the Gre`s Supe´rieurs (part equivalent to the Khorat Group; see Racey 2009) around the area of Kampot and Bokor in western Cambodia, recorded several plants and palynomorphs. The succession they studied was c. 750 m thick and rested on inferred Late Jurassic rocks assigned to the ‘Terrain Rouge’ (¼ Indosinias Moyenne). They recorded and illustrated the following plant taxa: Gleichenoides (several species), Sphenopteris mantellii and Leckenbya valdensis. This fossil fern assemblage was considered by them to be of Neocomian age. The latter two taxa are known from the Wealden red beds (Early Cretaceous) of southern England. In addition, they recorded two intervals that were productive for palynology. The first was located 250 m above the base of the Gre`s Supe´rieurs and yielded Corollina sp., Cicatricosisporites sp. and Osmundacidites sp., which would agree with the Early Cretaceous age dating based on the associated plants. A second, stratigraphically higher sample located c. 50 m below the top of the Gre`s Supe´rieurs yielded the same flora plus Concavisisporites and Diadites spinosa, the latter suggesting an age no older than Barremian and no younger than Aptian.

Palynology of the Mesozoic Trang Group, Peninsular Thailand The non-marine clastic Mesozoic sediments of Peninsular Thailand are generally assigned to the Trang Group (Thung Yai Group), which comprises from bottom to top the Khlong Min, Lam Thap, Sam Chom and Phu Phin Formations (Meesook et al. 2002). This group is considered to range in age from Middle Jurassic to Late Cretaceous (Meesook et al. 2002). Lei (1993) has described palynomorphs from outcrops of undifferentiated Trang Group sediments at Khao Noen Poon Sri and Ban Kuan Kun between the towns of Krabi and Trang in SW Peninsular Thailand to which he assigned a Late Jurassic age based principally on the percentage abundance and diversity of the Classopollis (Corollina herein) assemblage. However, the percentage of this taxa has been shown to vary inversely with latitude and therefore its abundance cannot be used as a reliable indicator of age as shown by Vakhrameev (1987). The assemblage Lei described and illustrated comprises some 21 species of Classopollis (Corollina) and D. etruscus, which respectively form 86% and 4% of the total

79

palynoflora. Corollina and D. etruscus have also been found in the Khorat Group of NE Thailand as shown in this paper. Other rare taxa recorded by Lei comprise Cycadopites, Chasmatosporites, Sphaeripollenites, Inaperturopollenites and Concentrisporites, and very rare pteridophyte spores. However, as discussed herein, D. etruscus would indicate an Early Cretaceous age, and peaks in Classopollis (Corollina) occur not only in the Late Jurassic but also in the Early Cretaceous, specifically the Barremian. Lei noted (1993, p. 367) that, although he preferred a Late Jurassic age for these assemblages, the palynoflora has a ‘closer relation’ to that of the Early Cretaceous. Alderson et al. (1994) analysed Trang Group sediments from Ban Nah and Ban Khan in the same general area. From Ban Nah they recorded Corollina torosus, and the brackish water indicators Pterospermella sp. and Leiosphere sp. Together this assemblage would suggest a restricted marine environment. In terms of age, the presence of C. torosus indicates an age no younger than Late Cenomanian. Samples from Ban Khan yielded a spore- or pollen-dominated assemblage of C. torosus, Araucariacites australis, Perinopollenites sp., Callialasporites dampieri, Klukisporites sp., Lycopodiacidites sp. and Aequitriradites spinulosus, and were deposited in a non-marine fluvial setting. The co-occurrence of A. spinulosus and C. dampieri indicates a Berriasian –Albian age. As this sample occurs stratigraphically above the Ban Nah locality it is likely the Ban Nah samples are of similar or slightly older age. Further studies in the region by Racey et al. (1997b) described palynomorph assemblages from four other Mesozoic localities in the same area as follows. From Laern Pleow an assemblage of C. torosus, C. brasiliensis, Tricolporopollenites spp., Verrucosisisporites sp., Cyathidites australis, Klukisporites spp., Spheripollenites spp. and Deltoidospora juncta was recovered. The presence of C. brasiliensis with Tricoloporopollenites suggests an Albian –Cenomanian age for these samples. From Ban Tone only C. torosus was recovered, indicating an age no younger than Cenomanian. At Khao Khad a rich but low-diversity assemblage comprising C. torosus, C. brasiliensis, C. australis and ?Araucariacites sp. was recovered. The presence of C. brasiliensis and C. torosus suggests an Albian –Cenomanian age for these samples. Finally, from Phru Toei a similar rich but low-diversity assemblage comprising C. brasiliensis, C. torosus, C. australis and Spheripollenites sp., again indicative of an Albian – Cenomanian age, was recovered. More recently, Goodall & Racey (pers. obs.) have been analysing a series of samples collected by M. F. Ridd from the same area. Although this

80

A. RACEY & J. G. S. GOODALL

work is continuing and therefore incomplete we can report that samples collected from the Mab Ching locality from the Khlong Min Formation have yielded common Corollina spp. and the freshwater algae Botryococcus, and common small inaperturate pollen and subordinate trilete spores that have yet to be identified to species level. The presence of Corollina spp. indicates an age no younger than Cenomanian. Previous workers (summarized by Philippe et al. 2005) have dated the formation as Middle–Late Jurassic based on vertebrates (turtles, fish and hybodont sharks) and charophytes (including Porochara sublaevis).

abundant D. etruscus in the Early Cretaceous Shupuhe Formation in the Tarim Basin, Xinjiang Province, China. For the present, it is preferred to adopt the conventional ideas on the range of D. etruscus and assume that its presence indicates a Berriasian to early Barremian age. In Yemen, where D. etruscus occurs in a marine section (allowing foraminiferal and dinoflagellate calibration) the range of this taxon is late Valanginian to early Barremian (unpublished oil company reports). This, at present, is the most accurately dated occurrence of D. etruscus.

Subsurface data The stratigraphical range of the zonal fossil Dicheiropollis etruscus The presence of D. etruscus in the Phu Kradung, Phra Wihan and Sao Khua Formations is important because this dyad pollen type has worldwide stratigraphical significance. D. etruscus becomes extinct in the lower Barremian and occurs over a broad geographical area, including West Africa, North Africa, Southern Europe (Trevisan 1971), China (Li 1990) Yemen, and South America. Confidence in dating the extinction point of this taxon is, therefore, relatively high. In Egypt and Libya, Ibrahim et al. (1995) distinguished the Early Cretaceous (Neocomian) from the Late Jurassic based on the presence of D. etruscus in the former, to which they assigned a late Neocomian–Barremian age, and the first appearance of Ephedrites in the Barremian. Ibrahim & Schrank (1996), in a study of dinoflagellates and spores or pollen from the Kahraman-1 well in NW Egypt, identified a distinct zone (their Zone III) of late Hauterivian –early Barremian age that is characterized by the occurrence of D. etruscus. Zone III is overlain by a late Barremian–early Aptian assemblage (their Zone IV) and underlain by a Berriasian –early Hauterivian assemblage (their Zone II), both of which lack D. etruscus and are well dated based on their associated dinoflagellate and pollen assemblages. Within Asia, D. etruscus has been recorded by Li (1990) from the Early Cretaceous of the Tarim Basin of China. In West Africa, the taxon is commonly considered to range from ‘mid-Neocomian’ to basal Barremian (Doyle et al. 1977), and in Brazil a total range of Neocomian to Barremian is recognized (Regali & Viana 1989). Nearly all publications refer to a Neocomian, late Neocomian, or late Valanginian age for the evolution point of this dyad pollen. However, Lei (1993) indicated that this taxon has been recorded from Late Jurassic sediments from the Anning Formation of Yunnan Province, China, based on as yet unknown criteria. He also noted the presence of

Palynological analysis of cuttings samples from the Phu Horm-1 well (Fig. 1) from the Khorat Plateau are summarized below. These are based on open-file reports held at the Department of Mineral Fuels in Bangkok, mainly from Harris (1977). No plates or range charts were available and therefore identifications are taken at face value. Of particular relevance to this paper is the occurrence of D. etruscus in the Phu Kradung Formation. This is a distinctive taxon, which is not easily confused with other palynomorphs, and has now also been found at outcrop in this formation. The youngest formation encountered in this well was the Phu Kradung Formation. Sampling was based solely on cuttings and therefore there is some potential for caving. However, the first and last downhole occurrences of D. etruscus are both within the Phu Kradung Formation.

Phu Kradung Formation D. etruscus was present in several samples over a 890 m thick interval together with sporadic rare occurrences of Perinipollenites elatoides, Undulatisporites distaverrucosus, Leptolepidites crassibalteus, Callialasporites turbatus, Ischyosporites cf. variegatus, Cyathidites australis and Deltoidaspora minor. Corollina classoides was abundant to very abundant throughout the interval, and Corollina itunensis was often common. D. etruscus was also found from 1238 to 1347 m depth with a slightly different assemblage comprising C. classoides, C. itunensis and very rare Eucommidites trodessoni, Araucariacitites australis and Aratrisporites sp. A Neocomian –Tithonian age (i.e. Early Cretaceous to latest Jurassic) was originally assigned to this interval, although based on the presence of D. etruscus we would now argue for an Early Cretaceous age for reasons outlined above. A similar, although often richer assemblage was noted in several other Khorat wells within this formation. Consequently, the early Cretaceous marker D. etruscus has been

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP

found over at least a 1347 m thick interval of Phu Kradung Formation. Because the well at the surface starts within the Phu Kradung Formation it is possible that D. etruscus may occur over an even thicker interval.

Nam Phong Formation Regional seismic and well data allow the identification of a Lower and an Upper Nam Phong Formation separated to by a low-angle unconformity. This unconformity has yet to be identified at outcrop, and as the red beds straddling the unconformity are lithologically similar it would be difficult to detect at outcrop. Only the Upper Nam Phong Formation is present in the Phu Horm-1 well, with the Lower Nam Phong Formation pinching out onto the Phu Horm structure (J. Booth, pers. comm.). The absence of any Late Triassic restricted palynomorphs supports this interpretation. C. classoides and Callialasporites turbatus were recovered from the Upper Nam Phong Formation and were assigned a Kimmeridgian –Pliensbachian age by Harris (1977). However, the presence of C. turbatus would indicate an age no older than Pliensbachian, whereas the absence of D. etruscus tends to support an age older than Early Cretaceous. C. classoides was recorded from as deep as 1704 m in the Upper Nam Phong Formation. The Upper Nam Phong Formation in the Phu Horm-1 well is therefore considered to be no older than Pliensbachian. Similar assemblages were noted in the Non Sung-1, Si That-1 and Phu Phra-1 wells located in the SW, central, and eastern Khorat Plateau, respectively. Of particular interest is the distribution of the Early Cretaceous marker D. etruscus within the Phu Horm-1 well, in that it occurs over at least a 1347 m thick interval of the Phu Kradung Formation. However, based on outcrops its actual distribution throughout the Khorat sequence is poorly constrained. Meesook (2000) considered the Phu Kradung Formation to vary in thickness from 800 to 1000 m at outcrop, whereas seismic and well data suggest that the formation generally averages around 1200 m thickness in the main part of the basin, thinning to around 400 –500 m at the basin margins. Consequently, based on the distribution of D. etruscus in the Phu Horm-1 well, most of the Phu Kradung Formation is Early Cretaceous. However, the Early Cretaceous marker D. etruscus was not found in the lowermost part of the Phu Kradung Formation in the Phu Horm-1 well, and this may indicate that the formation extends into the latest Jurassic (Tithonian) although no Jurassic-restricted taxa were recorded either from the wells or from outcrops of this formation. Other wells in the area appear to

81

have a Phu Kradung Formation thickness of 1000– 1500 m. In the Yang Talat-1 exploration well in the central Khorat Plateau, a Late Jurassic to Early Cretaceous age was reported based on palynology for the upper Phu Kradung to Sao Khua formations (Department of Mineral Resources, open-file report) although no supporting faunal lists were provided to support the age assignments.

Conclusions and discussion (1) Based on the above information it is proposed to include the Phu Kradung, Phra Wihan and Sao Khua Formations in the Early Cretaceous, with a maximum age-range of Berriasian – early Barremian, although the unpublished Yemen data suggest that this may tentatively be further refined to late Valanginian– early Barremian. It is still possible, however, that the lowermost part of the Phu Kradung Formation could be Late Jurassic based on the absence of the key Early Cretaceous marker taxon D. etruscus in the lowermost part of this formation in the Phu Horm-1 well, rather than on the presence of palynomorphs restricted to the Jurassic. (2) A marked difference in the percentage abundance of particular species is observed between the formations (especially in the Phu Kradung and Sao Khua Formations), which may be climatically controlled. Peaks of Corollina (Classopollis), for example, are attractive tie points on a global scale. Corollina is inferred to have occupied a warm, seasonally dry, subtropical climate. Further detailed sampling and analysis may allow a more refined palaeoenvironmental reconstruction of these sediments. (3) Despite the earlier palynological evidence for an Early Cretaceous age for much of the Khorat Group, many researchers, especially in Thailand, continue to assume a Late Jurassic to Early Cretaceous age for the Phra Wihan and Sao Khua Formations (e.g. Meesook et al. 2002), although earlier Meesook (2000) dated the Phra Wihan and Sao Khua Formations as Early Cretaceous and the Phu Kradung Formation as Jurassic (undifferentiated) using in part the palynological data of Racey et al. (1994, 1996). Some researchers (e.g. the Department of Mineral Fuels) still use the old scheme of Early, Middle and Late Jurassic for the Phu Kradung, Phra Wihan, and Sao Khua Formations. However, many researchers working on the vertebrates (e.g. Buffetaut and his colleagues), have for the most part accepted the revised dating based on palynology and have noted that it now broadly fits the vertebrate data (especially the more recent vertebrate discoveries) although they would prefer a latest Jurassic (Tithonian) age for at least part of the Phu Kradung Formation.

82

A. RACEY & J. G. S. GOODALL

(4) The Khorat Group was originally thought to comprise several formations that spanned the Late Triassic to Early Cretaceous with no major breaks in sedimentation. However, this study (together with Racey 2009) integrates additional data that support the original view of Racey et al. (1994, 1996) that a hiatus separates the Late Triassic Lower Nam Phong Formation from the remainder of the Khorat Group. Consequently, the Lower Nam Phong Formation should be excluded from the Khorat Group. The age of the Upper Nam Phong Formation is poorly constrained with palynological data, suggesting an age no older than Pliensbachian and no younger than Late Jurassic. Based on the the age of the overlying Phu Kradung Formation a Late Jurassic age is tentatively proposed for the Upper Nam Phong Formation. Buffetaut et al. (1993) had already suggested that the Nam Phong Formation may extend into the Jurassic. E. Buffetaut, M. Philippe and G. Cuny are thanked for providing samples for analysis over the last decade and for various discussions. S. Polachan and K. N. Sattayarak of the Department of Mineral Fuels are thanked for their expert guidance in the field. J. Booth is acknowledged for his comments and adivice on the stratigraphy of the Nam Phong Formation. A. R. and J. G. thank BG Group and Santos Ltd respectively for permission to publish. All of the opinions expressed herein are solely those of the authors and in no way reflect the views of BG Group or Santos Ltd.

References A LDERSON , A., B AILEY , N. J. L. & R ACEY , A. 1994. Mesozoic source rocks from peninsular Thailand: Implications for petroleum exploration. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 276– 281. B UFFETAUT , E. & I NGAVAT , R. 1984. The lower jaw of Sunosuchus thailandicus, mesosuchian crocodilian from the Jurassic of Thailand. Palaeontology, 27, 199– 206. B UFFETAUT , E. & I NGAVAT , R. 1985. The Mesozoic vertebrates of Thailand. Scientific American, 253, 80–87. B UFFETAUT , E. & S UTEETHORN , V. 1992. A new species of ornithischian dinosaur Psittacosaurus from the Early Cretaceous of Thailand, Palaeontology, 35, 801– 812. B UFFETAUT , E., S UTEETHORN , V., M ARTIN , V., C HAIMANEE , Y. & T ONG -B UFFETAUT , H. 1993. Biostratigraphy of the Mesozoic Khorat Group of northeastern Thailand: The contribution of vertebrate palaeontology. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia, II. Chiang Mai University, Chiang Mai, 51–62.

B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y.-N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, China, 19–37. B UFFETAUT , E., C UNY , G., LE L OEUFF , J. & S UTEETHORN , V. 2009. Late Palaeozoic and Mesozoic continental ecosystems of SE Asia: an introduction. In: B UFFETAUT , E., C UNY , G., LE L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 1 –5. C ANHAM , A. C., R ACEY , A., L OVE , M. A. & P OLACHAN , S. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, Northeastern Thailand: Part 2, diagenesis and reservoir quality. Journal of Petroleum Geology, 19, 321–338. C ORSIN , P. & D ESREUMEAUX , C. 1972. De´couverte d’une flore Ne´ocomienne dans les ‘gre´s supe´rieurs’ de Bokor (Cambodge Me´ridional). Annales de la Socie´te´ Ge´ologique du Nord, XCII, 199– 212. DEPARTMENT OF M INERAL R ESOURCES . 1983. Yang Talat– 1 exploration well report. Open file report, Department of Mineral Resources, Bangkok, Thailand, 1– 20. D EPARTMENT OF M INERAL R ESOURCES . 1992. Lexicon of stratigraphic names of Thailand. Department of Mineral Resources, Bangkok, Thailand, 1– 129. D OYLE , J. A., B IENS , P., D OERENKAMP , A. & J ARDINE , S. 1977. Angiosperm pollen from the pre-Albian Lower Cretaceous of Equatorial Africa. Bulletin du Centre de Recherche Exploration– Production, Elf Aquitaine, 1, 451–473. H AHN , L. 1982. Stratigraphy and marine ingression of the Mesozoic Khorat Group in northeastern Thailand. Geologisches Jahrbuch, 43, 7 –35. H AILE , N. S. 1973. Note on Triassic fossil pollen from the Nam Pha Formation, Chulabhorn (Nam Phron) Dam, Thailand. Newsletter of the Geological Society of Thailand, 6, 15– 16. H ANSEN , B. T., W EMMER , K. ET AL . 2002. Isotopic evidence for a Late Cretaceous age of the potash and rock salt deposit at Bamnet Narong, NE Thailand. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Resources, Bangkok, 120. H ARRIS , R. W. 1977. Palynology of the Phu Horm-1 well. Open-file report by R. W. Harris of ESSO. Department of Mineral Resources, Bangkok. H EGGEMANN , H., K OHRING , R. & S CHLU¨ TER , T. 1990. Fossil plants and arthropods from the Phra Wihan formation, presumably middle Jurassic of northern Thailand. Alcheringa, 14, 311– 316. I BRAHIM , M. I. A. & S CHRANK , E. 1996. Palynological studies on the Late Jurassic– Early Cretaceous of the Kahraman-1 well, northwestern desert, Egypt. In: J ARDINE , S., D E K LASZ , I. & D EBENAY , J.-P. (eds) Ge´ologie de l’Afrique et de l’Atlantique Sud, Compte-rendu des Colloques de ge´ologies d’Angers 1994. Bulletin de Centres de Recherche Exploration– Production, Me´moires, 16, 611– 629. I BRAHIM , M. I. A., S CHRANK , E. & A BDUL -K IREEM , M. R. 1995. Cretaceous biostratigraphy and

PALYNOLOGY AND STRATIGRAPHY OF KHORAT GROUP paleogeography of North Egypt and Northeast Libya. Petroleum Research Journal (Libya), 7, 75– 93. L EI , Z. 1993. The discovery and significance of the Late Jurassic sporopollen assemblage in Peninsular Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Asia: Facies and Palaeontology, II. Chiang Mai University, Chiang Mai, 361– 380. L I , W. 1990. Cretaceous System: Biostratigraphy and geological evolution of Tarim Basin. Oil and gas geology in Tarim Basin. Science Press, Beijing, 163 –172. M ARANATE , S. & V ELLA , P. 1986. Palaeomagnetism of the Khorat Group, Mesozoic, Northeast Thailand. Journal of Southeast Asian Earth Sciences, 1, 23–31. M EESOOK , A. 2000. Cretaceous environments of Northeastern Thailand. In: O KADA , H. & M ATEER , N. J. (eds) Cretaceous Environments of Asia. Elsevier, Amsterdam, 207– 223. M EESOOK , A., S UTEETHORN , V., C HAODUMRONG , P., T EERARUNGSIGUL , N., S ARDSUD , A. & W ONGPRAYON , T. 2002. Mesozoic rocks of Thailand: A summary. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Resources, Bangkok, 82–94. M OURET , C. 1994. Geological history of northeastern Thailand since the Carboniferous. Relations with Indochina and Carboniferous to Early Cenozoic evolution model. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245–252. M OURET , C., H EGGEMANN , H., G OUADAIN , J. & K RISADISIMA , S. 1993. Geological history of the siliciclastic Mesozoic strata of the Khorat Group in the Phu Phan Range area, northeast Thailand. In: T HANASUTHIPITAK , T (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology. University of Chiang Mai, Chiang Mai, 23– 49. P HILIPPE , M., C UNY , G. ET AL . 2005. A Jurassic amber deposit in Southern Thailand. Historical Biology, 17, 1–6. R ACEY , A. 2009. Mesotoic red bed sequences from SE Asia and the significance of the Khorat Group of NE Thailand. In: B UFFETAUT , E., C UNY , G., LE L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 39–65. R ACEY , A., G OODALL , J. G. S., L OVE , M. A. & J ONES , P. D. 1994. New age data on the Khorat Group of Northeast Thailand. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245–252. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S. & P OLACHAN , S. 1996. Stratigraphy and

83

reservoir potential of the Mesozoic Khorat Group, North Eastern Thailand: Part 1, Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 18, 5– 39. R ACEY , A., S TOKES , R. B., L OVATT -S MITH , P. & L OVE , M. A. 1997a. Late Jurassic collision in Northern Thailand and the significance of the Khorat Group. In: D EERADILOK , P., H INTHONG , C. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific, 2. Department of Mineral Resources, Bangkok, 412–423. R ACEY , A., H IGHTON , P. J. C., L EKUTHAI , T., A LDERSON , A. & P OLACHAN , S. 1997b. Mesozoic oils and source rocks from Peninsular Thailand. In: D EERADILOK , P., H INTHONG , C. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific, 2. Department of Mineral Resources, Bangkok, 511–524. R EGALI , M. S. P. & V IANA , C. F. 1989. Sedimentos do Neojurassico–Eocretaceo do Brasil: idade e correlacao com a escala internacional. Petrobras– SEDES, Rio de Janiero. S ATTAYARAK , N. 1983. Review of the continental Mesozoic stratigraphy of Thailand. In: N UTALAYA , P. (ed.) Proceedings of a workshop on stratigraphic correlation of Thailand and Malaysia I. Geological Society of Thailand and Geological Society of Malaysia, Bangkok, 127– 148. S ATTAYARAK , N. & S RIGULAWONG , S. 2008. The Western Khorat Plateau Geological Field Trip. Journal of the Geological Society of Thailand, Special Issue, 1, 1– 40. S ATTAYARAK , N., S RIGULAWONG , S. & P ATARAMETHA , M. 1991. Subsurface stratigraphy of the nonmarine Mesozoic Khorat Group, Northern Thailand. In: P OLACHAN , P. & T HANVARACHORN , P., P ISUTHA -A RNOND , V. ET AL . (eds) Seventh Regional Conference on Geology, Mineral Resources and Energy Resources of Southeast Asia. GEOSEA VII, 5 –8 November 1991, Bangkok, Thailand. Geological Society of Thailand, Bangkok, Thailand. S TOKES , R. B., L OVATT -S MITH , P. F. & S OUMPHONPHAKDY , K. 1996. Timing of the Shan Thai– Indochina collision: New evidence from the Pak Lay Foldbelt of the Lao PDR. In: H ALL , R. & B LUNDELL , D. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 225 –232. T HU , T. V. 1986. Some features of salt bearing Cretaceous red sediments in some areas of Indochina. In: Proceedings of first conference on the Geology of Indochina. Ho Chi Minch City, 5– 7 December 1986. GDG, Hanoi, 147– 153. T REVISAN , L. 1971. Dicheiropollis, a pollen type from Lower Cretaceous sediments of southern Tuscany (Italy). Pollen et Spores, 13, 561– 596. V AKHRAMEEV , V. A. 1987. Climates and distribution of some gymnosperms in Asia during the Jurassic and Cretaceous. Review of Palaeobotany and Palynology, 51, 205 –212.

Silhouette and palaeoecology of Mesozoic trees in Thailand MARC PHILIPPE1*, VE´RONIQUE DAVIERO-GOMEZ1 & VARAVUDH SUTEETHORN2 1

UMR5125 (PEPS) of the CNRS, Universite´ Lyon 1 (Campus de la Doua, Darwin A), F69622 Villeurbanne cedex, France

2

Fossil Research and Museums Bureau, Department of Mineral Resources, Rama VI Road, Bangkok, 10400, Thailand *Corresponding author (e-mail: [email protected]) Abstract: Trees are an obvious component of most landscapes. Artists’ views of Mesozoic landscapes regularly feature modern trees such as firs (Abies and Picea) and monkey-puzzle trees (Araucaria araucana). However, these reconstructions are highly hypothetical and, in reality, very little is known about the silhouettes of Mesozoic trees. In Mesozoic (Middle Jurassic to Early Cretaceous) strata of Thailand, large conifer logs with different types of architecture are evident, a rare opportunity for architectural studies. Various methods to estimate the original diameter and height of the trees are assessed. Among our material some trees have the classical Christmas-tree shape, whereas others are more oak-like in silhouette. All these trees lived in forest environments. Tree shape is strongly related to environment, and is still under-used as a palaeoecological tool. Reconstructing trees and vegetation has wide implications, from evaluating dinosaur herbivory to calculating elements of the carbon cycle.

Since the Middle Devonian forests have been greening the Earth (Stein et al. 2007). There are studies on early tree shapes (Daviero & Lecoustre 2000; Meyer-Berthaud & Decombeix 2007) but very little is known about the architecture of Mesozoic trees (Philippe & Daviero 2000), as well as the forests they formed (Pole 1999). During most of the Mesozoic conifers dominated canopies, whereas ginkgos, cycads and pteridosperms were mainly low- to middle-sized trees and shrubs. It is only during the Cenomanian (Late Cretaceous) that dicotyledonous trees really set roots into forests, and in terms of biomass they became a significant element there only by the Santonian (Coiffard et al. 2006). The representations of forests and trees that prevailed in the Mesozoic are strongly influenced by the countless reconstructions of dinosaurs in vegetated landscape, in which trees are merely a background for the reptiles. Artists’ views of Mesozoic landscapes commonly feature trees or even forests, but these reconstructions are hypothetical. They are not based on observation of fossil trees but on supposed architectural similarities between Mesozoic and extant gymnosperms. Modern conifers most familiar to the general public are Christmas trees (firs in general). Consequently, most reconstructions use silhouettes of fir (e.g. Abies and Picea), in some cases mixed with the highly recognizable silhouette of monkeypuzzle trees (Araucaria araucana K. Koch), probably the best known southern conifer species. Tropical and temperate southern conifers (e.g.

Podocarpus, Dacrydium or Agathis) are largely ignored, as are extinct families (e.g. Cheirolepidiaceae and Miroviaceae) that were significant components of the canopy vegetation by that time. Consequently, tree shape in these reconstructions should not be accepted uncritically; instead, tree shape should be based on scientific evaluation of the fossil record rather than on uniformitarianism. Pioneering studies on fossil logs from Patagonia (Calder 1953), Alexander Island (Antarctica; Jefferson 1982) and the Isle of Purbeck (UK; Francis 1983) gave the first indications of tree shape and forest density in the Mesozoic, but were subsequently questioned (Pole 1999). Other famous fossil forests, such as that of Curio Bay (New Zealand), or fossil log assemblages, such as that of Petrified Forest National Park (Arizona), were also investigated in an architectural perspective (Pole 1999; Ash & Creber 2000; Savidge & Ash 2006). Work relevant to the spacing and structure of Mesozoic trees has been described by Creber & Chaloner (1985, and references therein). Possible links between wood anatomy and tree architecture were suggested (Philippe 1992; Philippe & Daviero 2000), and an uncommon branching pattern was recognized amongst some Mesozoic conifers belonging to the genus Frenelopsis (Daviero et al. 2001). However, the collective evidence for tree profile and forest structure is rather scanty, and more studies are needed. Mesozoic forests and trees are of interest not only to palaeobotanists. The forest biome is the

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 85–96. DOI: 10.1144/SP315.7 0305-8719/09/$15.00 # The Geological Society of London 2009.

86

M. PHILIPPE ET AL.

place where the evolution of several early birdrelated lineages took place, it is the only source of food that could sustain large dinosaur herbivory, and it also represents an important trap for carbon (by far the most significant C sink of the terrestrial ecosystem), which could buffer natural oscillations in CO2 atmospheric concentration. Obtaining information on tree shape, forest structure, biomass and productivity is thus relevant to a wide range of disciplines. Investigations on the Mesozoic of Thailand offered us the opportunity to study two rich fossil log assemblages. Logs were measured and their size and branching pattern studied using the concepts of architectural analysis (Barthe´le´my & Caraglio 2007). This allowed us to determine for the first time the occurrence in the Mesozoic of large conifer trees with polyarchic silhouette, and also to propose hypotheses on the architectural features of the vegetation from which the logs were derived. It is also shown that tree architecture can be used as a palaeoecological indicator.

plant debris, and apart from trunks we noticed only woody axes with a diameter exceeding 6 cm, making it plausible that smaller plant debris was winnowed. Twenty- two logs were observed and measured. The second set of material comes from the Phu Phan Range, in NE Thailand (Isaan) and belongs to the Indochina block. The Phu Phan Range is a partly eroded anticline in which the Khorat Group crops out extensively (Mouret 1994). Fossil wood is a relatively common fossil in this area and is preserved as lignite or charcoal, or is silicified (Philippe et al. 2004). Many large silicified logs were observed at the top of the Phu Kradung Formation, for example at Lam Payang, Phu Pho and Phu Roi (Kalasin Province), and at nearby Ban Thung Chuak (Udon Thani Province). This silicified

Material Two sets of material have been used in this study. The first component derives from a locality in Southern Peninsular Thailand in Khlong Thom District, Krabi Province, about 20 km south of the town of Khlong Thom (Fig. 1), and belongs to the Shan-Thai block. As amber was recently reported from this locality (Philippe et al. 2005), the exact location will not be given here, but inquiries by colleagues are welcome. The stratigraphy of the nonmarine sediments in Southern Peninsular Thailand was reviewed by Teerarungsigul et al. (1999) and Meesook et al. (2005). These clastic red beds, with some lacustrine to marine calcareous intercalations (Dill et al. 2004), are known as the Trang Group, which is now subdivided into a basal Khlong Min Formation, overlain by the Lam Thap, Sam Chom and Phun Phin Formations in ascending order (Meesook et al. 2002). The amber bed belongs to the Khlong Min Formation, and its age is ascribed to the Jurassic (Buffetaut et al. 2005; Philippe et al. 2005; Cuny et al. 2009). Recent palynological investigations by Racey, however, identified the occurrence of Dicheiropollis etruscus in the amber layer, a pollen species that has never been found earlier than the Early Cretaceous (A. Racey, pers. comm.). This finding contradicts most other stratigraphical indices, but forces us to consider the age of the amber locality as equivocal. Fossil logs are included in calcareous sandstones underlying the amber level. These include large trunks that are preserved as lignite or, more locally, calcitized. The sediment is poor in other

Fig. 1. Location map. The amber locality, in Peninsular Thailand, belongs to the Shan-Thai block, the Phu Phan Range to the Indochina block.

TREE SILHOUETTES IN THAI MESOZOIC

wood-rich level could belong to the Waritchaphum Formation, which is supposed to be Kimmeridgian in age (Mouret 1994; Philippe et al. 2004). However, although it is clearly delineated on seismic profiles, this formation is difficult to identify in the field. The dating of the wood level is still controversial (Buffetaut & Suteethorn 1998; Philippe et al. 2004) and its age can be constrained only to the interval Late Jurassic – Barremian. Although the silicified wood level is probably slightly diachronous across the region, the sedimentology is similar (fine-grained fluviatile reddish to whitish sandstones), and the taxonomical composition of wood flora is uniform. For these reasons, the palaeovegetation of the various localities will be considered as uniform. Forty-four logs were observed and measured in the Phu Phan Range.

Methods Several fossil logs were sampled from both assemblages for xylological analysis. Thin-sections were prepared by traditional methods and samples were assigned to morphogenera on the basis of their xylological features (Bamford & Philippe 2001). Scanning electronic microscopy was performed on some specimens to resolve their identity. To reconstruct the shape and the architecture of trees, two sets of morphological data are necessary: (1) morphometric parameters such as the diameter and the length of trunks and branches; (2) morphological parameters such as the branching pattern (diffuse, rhythmic or continuous) and the geometry of branches (e.g. horizontal or vertical, inserted at an angle or perpendicular). Fossil logs are mostly parts of trunks on which all branches have been broken during transport. The architectural data were thus rather limited.

Tree diameter Most fossil logs have a flattened section (i.e. an elliptic rather than circular transversal section). To estimate trunk diameter before flattening, we hypothesize that Mesozoic trees had a rounded section as do most trees today. Elliptic sections amongst modern trees remain exceptional and usually result from accommodation to local conditions. To estimate the original diameter we tried successively three methods. First, we supposed that the trunk perimeter was preserved during the flattening process. The perimeter of an ellipse could be estimated by various ways, the first approximation, by Ramanujan (1914), being one of the most used: P ¼ p{3(a þ b)  [(a þ 3b)(3a þ b)]1=2 }

(1)

87

where P is the perimeter, a is the major axis radius and b the minor axis radius. If the ellipse is comparable with a flattened circle of diameter D, then P ¼ pD ¼ p{3(a þ b)  [(a þ 3b)(3a þ b)]1=2 }:

(2)

The original diameter D of the tree is D ¼ 3(a þ b)  [(a þ 3b)(3a þ b)]1=2 :

(3)

If we apply equation (3) to an observed fossil log in which a ¼ 27.5 cm and b ¼ 11 cm, the original diameter D was 40.29 cm. However, this first method could be inadequate, as observed ellipses (i.e. trunk transverse sections) mainly have high eccentricity, and the degree of flattening is not regular. Close examination usually reveals a sharp inflexion of growth rings near the middle of the ellipse, along its major axis. Microscopic investigation of the zones with sharp inflexion usually shows that tracheid rows slip one against the other, inducing non-linear deformation. A second method to estimate original diameter was used by Clifford (1996), who assumed that horizontal log width approximates original diameter and thus estimated D as 2a (D ¼ 55 cm with our example). We propose as a third method to estimate the ellipse perimeter P as twice the fossil log breadth (2a): P ¼ 4a

(4)

The original diameter D is then P ¼ 2 pR ¼ 4a

(5)

D ¼ 2R ¼ 4a=p  1:27a:

(6)

If we apply equation (6) to the fossil log mentioned above the original diameter D was 35.01 cm. With the first method (equation (3)) when log flattening increases p towards infinite, the D/a ratio ffiffiffi tends towards (3 2 3), which is not very different from 4/p (see equation (6)). On the other hand, if there is no flattening, D/a ¼ 2 (as in Clifford’s method). As equation (6) gives the most conservative estimation, we used that equation, being aware that it systematically underestimates original tree diameter and that, being constrained by sediment pressure, trunks probably do not expand much laterally during fossilization. Only fossil logs with complete preserved section were measured to estimate original diameters.

88

M. PHILIPPE ET AL.

Most ecologists and foresters use diameter at breast height (DBH, taken at 1.3 m above ground level). DBH cannot be calculated from a fossil log if its flaring base is not preserved, together with at least 1.3 m of the trunk. Fossil logs are mostly isolated trunk parts, and those with their enlarged base preserved are rare in general (three cases reported here). When the base is present, breadth is measured above flaring.

Tree height The relationship between the trunk diameter (DBH) and the tree height (H ) is an important subject in modern tree literature (Niklas & Spatz 2006). It is, however, more rarely discussed for fossil logs (Pole 1999). Palaeobotanists and paleoecologists usually apply Mosbrugger’s formula (Mosbrugger 1990; Mosbrugger et al. 1994), which, however, proved to give inaccurate estimations for many extant single trees (Pole 1999). This is probably because it is calculated on the basis of record-sized trees, and although it gives a reasonable height estimate for older trees it is much less adequate for younger ones. Another commonly used relationship between H and DBH is that given by Halle´ et al. (1978): H ¼ 100DBH. This relationship, however, was based on the observation of dicotyledonous angiosperm trees in equatorial lowland forest. Pole (1999) gave an interesting discussion of the problem of estimating tree height from DBH. He demonstrated that the height of modern Araucariaceae can be twice or even three times the value estimated from Mosbrugger’s curve. This is especially the case for those species that, like most extant Araucariaceae, are particularly modelconforming (i.e. trees that maintain a consistent architectural model without much adaptative reiteration or with only partial reiteration of lateral branches, resulting in a Christmas tree-like silhouette). Trees of this type have trunks much larger in diameter than branches, and this is typical of a ‘hierarchic’ architecture (Edelin 1991). Not all Mesozoic trees conformed to this model or were hierarchic. Logs with main branches similar in diameter to the trunk have been documented from the Jurassic onward (Philippe 1992). Amongst modern Podocarpaceae mature trees have this type of architecture, described as ‘polyarchic’ by Edelin (1991). The height v. diameter scattergram of such trees is different, and for old trees height can be as small as 13.06 diameters (Podocarpus falcatus (Thunb.) Mirb. in Knysna forest, South Africa) or 10.74 diameters (P. totara G. Benn. ex D. Don ‘Pouakani’ tree, New Zealand). Considering the tree height v. diameter scattergram, ‘champion trees’ of the Podocarpaceae

are much closer to Mosbrugger’s curve than to Pole’s ‘low curve’. Moreover, Pole (1999) noted that ecological conditions strongly influence the ratio between tree height and DBH. Trees growing in closed forest have slender trunks and branch higher than those growing in open landscapes such as savannas or grasslands. It is impossible to know the position of fossil logs within the original trunk. Only in rare cases is the flaring base with the typical conical shape preserved. Fossil logs are commonly slightly conical, with a diameter D1 at one tip broader than the diameter D2 at the other. If we approximate the original tree trunk to a cone with straight sides (i.e. not parabolic ones) and call L the fossil length, then the ‘thinning coefficient’ is (D1 2 D2)/L and the corresponding theoretical log reaches a zero diameter at a distance X from the base: X ¼ L  D1=(D1  D2):

(7)

Equation (7) can theoretically be used to determine a minimum height of the original tree from partial logs. Unfortunately, this method was found to give unrealistic values, probably because of abrasion of the outer part of the trunk before deposition, and because of the existence of branching. Eventually, we decided to use Pole’s ‘low curve’, noting that above 90 cm the larger the DBH the weaker the reliability of tree height prediction. As compared with the results of Ash & Creber (2000), who used Niklas’ allometric method (Niklas 1994), our estimates for tree heights are clearly conservative.

Branch Often broken during the sedimentation process, branches are rarely preserved on fossil logs. The probability of branch preservation is supposed to depend on the distance of transport, the dynamism of the transport, the decay preceding uprooting and mobilization, or on the branch diameter. When branches are broken post mortem, they leave a clear scar on the trunk. The size of the scar, albeit usually somewhat larger, would indicate the basal diameter of the branch. Scar distribution on the logs is regularly used (Francis 1983; Ash & Creber 2000; Savidge & Ash 2006) to determine branch arrangement and rhythmicity.

Results Tree size analysis Data are given in Tables 1 and 2. Only largest log breadth is indicated when there is a significant

TREE SILHOUETTES IN THAI MESOZOIC

89

Table 1. Log size at the amber locality (Peninsular Thailand) Breadth (cm)

Calculated diameter (cm)

Observed length (cm)

Calculated tree height (m)

52.84 33.10 65.57 29.92 8.28 9.55 89.13 49.66 20.37 26.10 53.48 30.56 15.28 58.57 62.39 27.37 29.28 45.84 45.84 29.92 20.37 28.65

300 400 250 100 31 100 8501 100 100 1501 75 135 254 318 340 170 92 230 510 110 130 200

30 22 35.5 20.5 10.4 10.5 43 29 15.5 18.5 30.5 21 11.5 32 34 19 20.5 28 28 20.5 15.5 19.5

83 52 103 47 13 15 140 78 32 41 84 48 24 92 98 43 46 72 72 47 32 45 1

Trunks displaying ramification.

Table 2. Log size in the Phu Phan Range (Isaan, Thailand) Width (cm) 55 52 70 31 51 35 34 26 37 35 150 115 125 62 38 81 15 50 40 14 46 32 1

Calculated diameter (cm) 35.01 33.10 44.56 19.74 32.47 22.28 21.65 16.55 23.55 22.28 95.49 73.21 79.58 39.47 24.19 51.57 9.55 31.83 25.46 8.91 29.28 20.7

Flaring base preserved. With a large branch. 3 Stump. 2

Observed Calculated tree Width Calculated length (cm) height (m) (cm) diameter (cm) 80 146 107 61 190 110 320 67 34 36 5501 450 180 582 60 94 – – – – – –

23 22 27.5 15 22 17 16.5 12.5 17.5 16.5 44.5 38 40 24.5 12.5 29.5 10.5 21.5 18.5 10.5 20.5 15.5

70 31 63 35 35 26 37 35 55 52 120 150 32 15 60 69 42 32 20 30 50 24

44.56 19.74 40.11 22.28 22.28 16.55 23.55 22.28 35.01 33.10 76.39 95.49 20.7 9.55 38.2 43.93 26.74 20.7 12.73 19.1 31.83 15.28

Observed Calculated tree length (cm) height (m) 107 61 190 110 300 79 24 36 82 136 1500 203 15 60 150 91 – –3 – – – –

27.5 15 26 16.5 16.5 17 17.5 17 23 22 39 44.5 15.5 10.5 24.5 27 19 15.5 10 14.5 21.5 11.5

90

M. PHILIPPE ET AL.

difference in diameter between both tips. Trunk diameter is calculated from equation (6). Minimum tree height is estimated from calculated diameter and Pole’s low curve. At the amber locality (Peninsular Thailand) the highest estimated height is 43 m. Several logs, however, indicate lower heights (between 30 and 35 m). None of these logs displays any basal flaring; tree heights are calculated from an underestimated diameter and a ‘low’ curve, which is considered to estimate tree heights conservatively. These criteria all suggest that the original trees reached at least 40 m tall. Among the Phu Phan Range material, the logs with basal flaring and the largest stump have remarkably similar diameters. Largest tree height can reasonably be estimated at 45 m.

Branch analysis Logs with preserved branches were rare in the assemblages. At the amber locality only two out of 22 specimens were branched (Fig. 2). These branched trunks have a Brachyoxylon type of wood (Philippe et al. 2005). The branches are interpreted as the bases of main branches. Both specimens are typical for trees with a polyarchic silhouette. The angle of branching is variable, and

Fig. 2. Sketch of the two ramified logs observed at the amber locality (from Philippe et al. 2005). Scale bar 1 m.

could be very different from the original angle because of flattening. The unramified logs are cylindrical to slightly conical straight trunk parts, ranging from 8 to 89 cm in diameter. Even pieces with small diameter or relatively long specimens (2.5, 3 or even 4 m long), do not show any longitudinal curvature. Two wood morphogenera, Agathoxylon and Brachyoxylon, were recognized among these logs with no branches. It should be noted that these morphogenera have little ‘biological’ value, and that the occurrence of two wood genera does not mean that a unique biological species was not the source for all the material (Philippe 1992). Whatever their position in the wood parataxonomy, unbranched specimens could be either branches or trunks. None displays the characteristic axial dissymmetry of horizontal branches, and 4 m long straight and unbranched orthotropous branches are not very realistic from an architectural point of view. As smaller axes were probably removed by winnowing during transport, we think it is much more realistic to consider that all observed axes are actually trunks or trunk parts. The absence of branching on most trunk parts is interpreted as a sign that trees developed in forest conditions, with a closed canopy, under which basal branches are naturally pruned. In the Phu Phan Range branched specimens are also rare (only three out of the 44 specimens studied). A typical specimen (Fig. 3) has hierarchic ramification (i.e. with branch diameter 7 –15 times smaller than trunk diameter) and displays three whorls of four perpendicular branches. From the basal to distal whorl branch scars are superimposed and decrease in diameter from 8 to 5 cm. Growth units are respectively 35 and 25 cm long. Although growth unit length is not a linear function of the distance to the tip of the tree, it is suspected that the specimen corresponds to a segment originally situated relatively close to the tree apex. Partial fractures allowed us to observe that branch traces were at an angle to the vertical. As it is impossible to orientate the specimen, we cannot decipher whether branch traces were sloping downward or upward, but modern branch traces are usually turned upward. The branching pattern observed in this specimen is strongly reminiscent of that described for modern Araucaria. Only specimens with complete axis sections are considered here; however, during field studies we observed several other specimens, which were only parts of the original trunk cylinder. These sometimes displayed branch scars (22 observations) or branches (two cases), always organized in whorls. Both scars and branches were always below 5 cm in diameter, and strongly hierarchic. Only two less clearly hierarchic branches were observed. In both cases, at a level with a whorl of branch scars, one

TREE SILHOUETTES IN THAI MESOZOIC

91

Fig. 3. A log part with branch bases in Ban Tung Chuak, Phu Phan Range. (Note the three whorls of branches.)

of the scars is much larger than the others (trunk diameter–scar diameter being 80 cm –13 cm in one case and 62 cm – 9 cm in the other). This is interpreted as evidence that one of the branches of a whorl sometimes persisted and grew larger than the others. Most of observed fossil logs have no apparent branching, and except for the three specimens mentioned above, no branch scar was found on any fossil logs exceeding 40 cm in diameter, even on very long specimens (up to 15 m long). Obviously, natural

branch shedding occurred, freeing long portions of trunks. As in the case of the amber locality, we interpret this as evidence that these trees grew in forest conditions, with a canopy high above ground level, and no sub-canopy branches. In this respect the Phu Phan trees strongly differ from modern Araucariaceae, in which plagiotropous branches are often reiterated (Edelin 1986). Two morphogenera, Agathoxylon (with one species) and Brachyoxylon (with three species) are represented among the Phu Phan Range wood

92

M. PHILIPPE ET AL.

assemblages (Philippe et al. 2004). All the evident wood morphospecies are very variable and, as in the case of the amber locality, it is not possible to exclude the possibility that all these wood types were formed by a single biological species. Whatever the number of biological species represented in the log assemblages, no significant architectural difference was observed, and we conclude that only one type of architecture occurred. This was strongly hierarchic, with rhythmic branching and whorls of four to six plagiotropous branches. In the tropics such trees, when mature, are usually umbrella-shaped (Edelin 1986), possibly for better light interception (Creber & Chaloner 1987). Plagiotropous branches were pruned naturally, and only exceptional branches persisted, perhaps giving the trees more irregular silhouettes, similar to those described for Araucaria araucana (Mol.) K. Koch by Grosfeld et al. (1999) or Philippe & Daviero (2000). Stumps and flaring bases were observed only among the Phu Phan Range assemblage. Flaring is always limited, and no buttress has been observed. In enlarging trunk zones undulated wood (tonewood) is frequently observed as sets of transverse ridges on the wood surface, similar to those illustrated by Ash & Creber (2000) for Araucarioxylon arizonicum Knowlton. Here, however, ridges were always associated with the trunk base and not with the base of plagiotropous lateral branches, as described for the US material.

Discussion Architectural analysis reveals two very different types of trees among the studied material. Although both types were forest trees with canopies well above 30 m, trees at the amber locality display a polyarchic silhouette, whereas those from the Phu Phan Range are hierarchized. It is tempting to attribute this difference to palaeoecological reasons. Conifer trees displaying these two types of architectures are known to sporadically grow together in some subtropical forests, as we observed, for example, in southern Queensland (Lamington National Park, with Podocarpus elatus Endl. and Araucaria cunninghamii Sweet). However, conifer forests with only one type of architecture tend to be the rule today. Thailand was within the tropical zone during the Mesozoic (Charusiri et al. 1997; however, see also Carter & Bristow 2003). Modern tropical zone conifer forests with only hierarchic architectures occupy much drier areas than those with only polyarchic architecture. The latter are encountered mostly in montane rainforest with substantial precipitation (typically above 2 m) well distributed

over the year; for example, those of Khao Yai National Park in Thailand. The former can be encountered in drier areas, with more marked seasonality; for example, those of Phu Kradung National Park in Thailand. Use of uniformitarianism for palaeoecological inferences on Mesozoic forests is risky as conifer forest ecology, since the rise and dominance of angiosperms in the Cretaceous, was most probably displaced and some conifer growth forms may have become extinct. It should be noted, however, that such would not be, in our two cases, contradictory with the palaeoclimatological inferences drawn from sedimentological features, which indicate a tropophilous (strongly seasonal) climate in Isaan (Philippe et al. 2004) and a more humid climate at the amber locality (Philippe et al. 2005). Observed log sizes can also be interpreted palaeoecologically. Large trees, reaching 1.5 m in diameter and an estimated height of 45 m, cannot develop under an arid climate. To use uniformitarianism again, an amount of 2000 mm of annual rain seems to be the threshold below which such large trees do not occur today in the tropical zone. If some Mesozoic conifer trees had been more resistant to drought than modern species, they would probably have lingered until now. Interestingly, we have not observed any fossil log in the Phu Kradung Formation of Isaan with the architecture described by Srisuk (2000). That worker, on the basis of a fossil log observed at Phu Tang Kwian, reconstructed a tree with a slender silhouette, continuous branching and persistent branches, fitting the silhouette proposed by Ash & Creber (2000) for Araucarioxylon arizonicum, the dominant conifer of Arizona Petrified Forest National Park. Later, however, the fossil log used for reconstruction by Ash & Creber (2000) was reassigned to a new genus, Arboramosa, on the basis of morphological features; those features seem to be of rare occurrence in the US National Park (Savidge & Ash 2006). It is not clear that Srisuk’s log displays all the diagnostic features for Arboramosa, and this specimen needs to be re-examined. Based on the results and discussion above, two silhouettes are here proposed (Figs 4 and 5). These silhouettes are not reconstructed from complete fossil evidence, as data are still limited, but are based on extant conifer trees in tropical to subtropical forests that have size and branching characteristics most similar to the fossil material. A silhouette for trees from the amber locality (Fig. 4) is based on Podocarpus neriifolia D. Don tree in Khao Yai National Park (Thailand). A mature Araucaria cunninghamii Sweet, observed in Lamington National Park (Queensland), served as a model for the reconstruction of trees from the

TREE SILHOUETTES IN THAI MESOZOIC

Fig. 4. Silhouette of Podocarpus neriifolia D. Don tree, drawn in Khao Yai National Park (Thailand), within a dense tropical forest. Trees with such a polyarchic and slender silhouette may be the source for the fossil logs observed at the amber locality.

Phu Phan Range (Fig. 5). In both cases no buttresses were drawn. These are absent in Phu Phan Range material, and although tree bases have not been observed at the amber locality we hypothesize that buttresses were also absent in the parent tree, as they are rarely developed by modern conifers. With closed-canopy forests, reaching 35–40 m in height at least, the biomass of phytocoenosis was important. As we have no access to tree density, as observed log assemblages are all allochthonous, we estimated forest biomass by allometric calculation, using the usual equation (Jenkins et al. 2003): forest biomass ¼ 0.8 (forest height)1.75.

93

Fig. 5. Silhouette of mature Araucaria cunninghamii Sweet tree, observed in Lamington National Park (Queensland, Australia), within a dense subtropical forest. Trees with such a hierarchic and umbrella-like silhouette may be the source for the fossil logs observed in the Phu Phan Range localities.

On this basis, a 40 m high forest has a biomass of about 510 t ha21 (dry weight). This is significantly less than the 579–617 t ha21 estimated by Pole (1999) for a sub-polar Mesozoic forest, and it could indicate that Thai Mesozoic forests were less dense than their New Zealand relatives. These biomass values are difficult to compare, however, as they were not obtained from the same type of calculation. Maximum diameters for logs in the Thai assemblages are larger than those observed in New Zealand and, turning to uniformitarianism again, 45 m tall conifer trees do not grow today in scattered

94

M. PHILIPPE ET AL.

stands with low density. To sum up, the biomass of Thai Mesozoic forests was probably in excess of 600 t ha21. Our results challenge the view of the seminal work on the Khorat Group (Ward & Bunnag 1964), according to which coal is missing in this unit as ‘plant-growth was limited at the time of sedimentation’. It would be interesting to relate the Thai Mesozoic vegetation to corresponding zoocoenoses. It has been estimated that large sauropods, which roamed at this time in the area (Buffetaut & Suteethorn 1998), needed about 200 kg of foliage per day (Weaver 1983; Farlow 1987). Conifer foliage, in the Mesozoic as today, has low palability and low digestibility (because of the thick cuticle and the chemicals it contains), but its caloric value (about 18–20 kJ kg21, dry weight) is not very different from that of angiosperm foliage (Bobkova & Tuzhilkina 2001; Hummel et al. 2008). Provided sauropods managed to bypass the drawbacks of conifer foliage, the biomass inferred here (600 t ha21) should be enough to sustain a population of large herbivores. However, with a canopy above 20 m high, most foliage biomass in the forests that yielded the studied fossil logs was probably beyond the reach of even the largest sauropods. We still lack too much information (tree density, annual productivity, leaf/wood ratio of the biomass) to propose sensible hypotheses about the way herbivorous reptiles exploited phytocoenoses to ensure their huge daily uptake. Forest density and structure is also probably relevant to the discussion of the origin and evolution of early birds.

Conclusions Our investigations of fossil log assemblages from the Mesozoic of Thailand demonstrated that: (1) large conifer trees with polyarchic silhouette already existed in the mid-Mesozoic; (2) tall forests with closed canopy existed during the midMesozoic in Thailand; (3) the architecture of fossil logs indicates a more humid and uniform climate in Peninsular Thailand at that time, as compared with a drier and more seasonal climate in Isaan; (4) these palaeoecological inferences conform to those obtained from sedimentology, and this suggests that fossil tree architecture could be more often used for palaeoecological purposes. We thank D. Barthe´le´my and an anonymous reviewer for valuable suggestions for the improvement of the manuscript. This work was supported by the TRF-CNRS special programme for Biodiversity Research and Training Programme (BRT/BIOTEC/NSTDA) grant BRT R_245007, and the French CNRS programme ‘Eclipse’. We thank E. Buffetaut for his constant support and commitment to the study of Thai Mesozoic vegetation.

G. Cuny helped greatly in the field and in other aspects of our research. Our colleagues A. Meesook and N. Teerarungsigul kindly assisted us in the field in southern Thailand. Several Thai palaeontologists helped with field studies, especially J. Butiman, U. Deesri, P. Intasoi, S. Khamha, S. Khansubha and K. Wongko. An anonymous reviewer improved a first version of this paper. This is publication UMR5125-09.002 of the CNRS.

References A SH , S. R. & C REBER , G. 2000. The Late Triassic Araucarioxylon arizonicum trees of the Petrified Forest National Park, Arizona, USA. Palaeontology, 43, 15–28. B AMFORD , M. & P HILIPPE , M. 2001. Gondwanan Jurassic– Early Cretaceous homoxylous woods: A nomenclatural revision of the genera with taxonomical notes. Review of Palaeobotany and Palynology, 113, 287–297. B ARTHE´ LE´ MY , D. & C ARAGLIO , Y. 2007. Plant architecture: a dynamic, multilevel and comprehensive approach to plant form, structure and ontogeny. Annals of Botany, 99, 375–407. B OBKOVA , K. S. & T UZHILKINA , V. V. 2001. Carbon concentrations and caloric value of organic matter in Northern Forest ecosystems. Russian Journal of Ecology, 32, 63–65. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of South-East Asia. Backhuys, Leiden, 83–90. B UFFETAUT , E., S UTEETHORN , V., T ONG , H. & K OSˇ IR , A. 2005. First dinosaur from the Shan-Thai block of South-East Asia: A Jurassic sauropod from the southern peninsula of Thailand. Journal of the Geological Society, London, 162, 481– 484. C ALDER , M. G. 1953. A coniferous petrified forest in Patagonia. Bulletin of the British Museum (Natural History) Geology, 2, 99– 138. C ARTER , A. & B RISTOW , C. S. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau Basin, eastern Thailand. Basin Research, 15, 271– 285. C HARUSIRI , P., K OSUWAN , S. & I MSAMUT , S. 1997. Tectonic evolution of Thailand: From Bunopas (1981)’s to a new scenario. In: D HEERADILOK , P., H INTHONG , C., C HAODUMRONG , P. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific, 1. Department of Mineral Resources, Bangkok, 414–420. C LIFFORD , H. T. 1996. Geometrical study of a cast of Leptophloeum australe (McCoy) Walton from Queensland. Memoirs of the Queensland Museum, 39, 227– 230. C OIFFARD , C., G OMEZ , B., K VACEK , J. & T HE´ VENARD , F. 2006. Early Angiosperm ecology, evidence from the Albian–Cenomanian of Europe. Annals of Botany, 98, 495–502. C REBER , G. T. & C HALONER , W. G. 1985. Tree growth in the Mesozoic and Early Tertiary and the reconstruction

TREE SILHOUETTES IN THAI MESOZOIC of palaeoclimates. Palaeogeography, Palaeoclimatology, Palaeoecology, 52, 35– 60. C REBER , G. T. & C HALONER , W. G. 1987. The contribution of growth-ring studies to the reconstruction of past climates. In: W ARD , R. G. W. (ed.) Application of Tree-Ring Studies. British Archaeological Reports International Series, 333, 37–67. C UNY , G., S RISUK , P., K HAMHA , S., S UTEETHORN , V. & T ONG , H. 2009. A new elasmobranch fauna from the middle Jurassic of southern Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 95–111. D AVIERO , V. & L ECOUSTRE , R. 2000. Computer simulation of sphenopsids architecture. II. Calamites multiramis Weiss, an example of Late Palaeozoic arborescent Equisetales. Review of Palaeobotany and Palynology, 109, 135– 148. D AVIERO , V., G OMEZ , B. & P HILIPPE , M. 2001. Uncommon branching pattern within conifers: Frenelopsis turolensis, a Spanish Lower Cretaceous Cheirolepidiaceae. Canadian Journal of Botany, 79, 1400–1408. D ILL , H. G., L UPPOLD , F. W., T ECHMER , A., C HAODUMRONG , P. & P HOONPHUN , S. 2004. Lithology, micropaleontology and chemical composition of calcareous rocks of Paleozoic through Cenozoic age (Surat Thani Province, central Peninsular Thailand): Implications concerning the environment of deposition and the economic potential of limestones. Journal of Asian Earth Sciences, 23, 63–89. E DELIN , C. 1986. Strate´gie de re´ite´ration et e´dification de la cime chez les conife`res. Colloque ‘L’arbre, 1986’. Naturalia monspeliensia, A5, 139–158. E DELIN , C. 1991. Que savons-nous des arbres sympodiaux: Le concept de plan d’organisation. Colloque ‘L’arbre, 1991’. Naturalia monspeliensia, A7, 127–154. F ARLOW , J. O. 1987. Speculations about the diet and digestive physiology of herbivorous dinosaurs. Paleobiology, 13, 60–72. F RANCIS , J. E. 1983. The dominant conifer of the Jurassic Purbeck Formation, England. Palaeontology, 26, 277–294. G ROSFELD , J., B ARTHE´ LE´ MY , D. & B RION , C. 1999. Architectural variations of Araucaria araucana (Molina) K. Koch (Araucariaceae) in its natural habitat. In: K URMANN , M. H. & H EMSLEY , A. R. (eds) The Evolution of Plant Architecture. Royal Botanic Garden, Kew, 109–122. H ALLE´ , F., O LDEMANN , R. A. A. & T OMLINSON , P. B. 1978. Tropical Trees and Forests. Springer, Heidelberg. H UMMEL , J., G EE , C. T., S U¨ DEKUM , K.-H., S ANDER , P. M., N OGGE , G. & C LAUSS , M. 2008. In vitro digestibility of fern and gymnosperm foliage: implications for sauropod feeding ecology and diet selection. Proceedings of the Royal Society, Series B, 275, 1015–1021. J EFFERSON , T. H. 1982. Fossil forests from the Lower Cretaceous of Alexander Island, Antarctica. Palaeontology, 25, 681–708. J ENKINS , J. C., C HOJNACKY , D. C., H EATH , L. S. & B IRDSEY , R. A. 2003. National-scale biomass

95

estimators for United States tree species. Forest Science, 49, 12– 36. M EESOOK , A., S UTEETHORN , V., C HAODUMRONG , P., T EERARUNGSIGUL , N., S ARDSUD , A. & W ONGPRAYOON , T. 2002. Mesozoic rocks of Thailand: a summary. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Resources, Bangkok, 82–94. M EESOOK , A., S AENGSRICJAN , W. & T EERARUNGSIGUL , N. 2005. Stratigraphy and faunal aspects of the marine Jurassic rocks in southern peninsular Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 558– 569. M EYER -B ERTHAUD , B. & D ECOMBEIX , A.-L. 2007. A tree without leaves. Nature, 446, 861–862. M OSBRUGGER , V. 1990. The Tree Habit in Land Plants— a Functional Comparison of Trunk Construction with a Brief Introduction into the Biomechanics of Trees. Springer, Berlin. M OSBRUGGER , V., G EE , C. T., B ELZ , G. & A SHRAF , A. R. 1994. Three-dimensional reconstruction of an in-situ Miocene peat forest from the Lower Rhine Embayment, northwestern Germany—new methods in palaeovegetation analysis. Palaeogeography, Palaeoclimatology, Palaeoecology, 110, 295– 317. M OURET , C. 1994. Geological history of North-Eastern Thailand since the Carboniferous. Relations with Indochina and Carboniferous to Early Cenozoic evolution model. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 132– 158. N IKLAS , K. J. 1994. Predicting the height of fossil plant remains: An allometric approach to an old problem. American Journal of Botany, 81, 1235– 1242. N IKLAS , K. J. & S PATZ , H.-C. 2006. Allometric theory and the mechanical stability of large trees: Proof and conjecture. American Journal of Botany, 93, 824– 828. P HILIPPE , M. 1992. Un de´terminisme architectural pour les bois de Conife`res a` ponctuation mixte (Protopinaceae Krau¨sel, 1917)? Canadian Journal of Botany, 70, 1834–1839. P HILIPPE , M. & D AVIERO , V. 2000. Observations de re´ite´rations verticales chez l’Araucaria araucana (Mol.) K. Koch et conse´quences pour la reconstitution des paysages du Me´sozoı¨que. Revue de Pale´obiologie, 19, 157 –163. P HILIPPE , M., S UTEETHORN , V., L UTAT , P., B UFFETAUT , E., C AVIN , L., C UNY , G. & B ARALE , G. 2004. Stratigraphical and palaeobiogeographical significance of fossil wood from the Mesozoic Khorat Group of Thailand. Geological Magazine, 141, 319–328. P HILIPPE , M., C UNY , G., S UTEETHORN , V. ET AL . 2005. A Jurassic amber locality in Thailand. Historical Biology, 17, 1– 6.

96

M. PHILIPPE ET AL.

P OLE , M. 1999. Structure of a near-polar latitude forest from the New Zealand Jurassic. Palaeogeography, Palaeoclimatology, Palaeoecology, 147, 121–139. R AMANUJAN , S. 1914. Modular equations and approximations to P. Quarterly Journal of Pure and Applied Mathematics, 45, 350–372. S AVIDGE , R. A. & A SH , S. R. 2006. Arboramosa semicircumtrachea, an unusual Late Triassic Tree in Petrified Forest National Park, Arizona, USA. In: P ARKER , W. G., A SH , S. R. & I RMIS , R. B. (eds) A Century of Research at Petrified Forest National Park: Geology and Paleontology. Museum of Northern Arizona Bulletin, 62, 65–81. S RISUK , P. 2000. The Late Jurassic Araucarioxylon sp. tree of North-eastern Thailand. Bulletin of the Srisuk’s House Museum, Series A, 3, 1– 12. S TEIN , W. E., M ANNOLINI , F., V AN A LLER H ERNICK , L., L ANDING , E. & B ERRY , C. M. 2007. Giant

cladoxylopsid trees resolve the enigma of the Earth’s earliest forest stumps at Gilboa. Nature, 446, 904–907. T EERARUNGSIGUL , N., R AKSASKULWONG , L. & K HANTAPRAB , C. 1999. Reconsideration of lithostratigraphy of non-marine Mesozoic rocks in Thung Yai–Khlong Thom area, Southern Thailand. In: K HANTRAPRAB , C. & S ARAPIROME , S. (eds) Proceedings of the Symposium on Mineral, Energy and Water Resources of Thailand: Towards the Year 2000. Department of Mineral Resources, Bangkok, 109–114. W ARD , D. E. & B UNNAG , D. 1964. Stratigraphy of the Mesozoic Khorat Group in North-Eastern Thailand. Report of investigation, Department of Mineral Resources, 6. W EAVER , J.-C. 1983. The improbable endotherm: The energetics of the sauropod dinosaur Brachiosaurus. Paleobiology, 9, 173–182.

A new elasmobranch fauna from the Middle Jurassic of southern Thailand GILLES CUNY1* PALADEJ SRISUK2, SUCHADA KHAMHA3, VARAVUDH SUTEETHORN4 & HAIYAN TONG5 1

Natural History Museum of Denmark, University of Copenhagen, Øster Voldgade 5 – 7, 1350 Copenhagen K, Denmark 2

Srisuk’s House Museum, 100/12 Moo 1, Tambon Khao Yoi, Changwat Phetchaburi 76140, Thailand

3

Palaeontological Research and Education Centre, Mahasarakham University, Tambon Khamriang, Amphoe Kantarawichai, Mahasarakham 44150, Thailand 4

Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand 5

16 cour du Lie´gat, 75013 Paris, France

*Corresponding author (e-mail: [email protected]) Abstract: We describe a new elasmobranch fauna from the lower part of the Khlong Min Formation in Thailand. The fauna includes Hybodus sp., Asteracanthus sp., Lonchidion reesunderwoodi sp. nov., Belemnobatis aominensis sp. nov., and possibly a second species of Belemnobatis. This fauna supports a Bathonian– Callovian age for the Khlong Min Formation, and suggests a close taxonomic relationship between the Middle Jurassic elasmobranch faunas of Europe and Thailand. The presence of a monolayered enameloid in Belemnobatis aominensis sp. nov. and other primitive batoids is interpreted as the retention of a primitive character for neoselachians, which would suggest a divergence time between the batoids and the rest of the neoselachian sharks as early as the Carboniferous –Permian boundary.

Thailand consists of two continental blocks or microcontinents. The eastern part, including the Khorat Plateau, belongs to the Indochina block, whereas the western part, including the southern peninsula, is part of a terrane called Shan-Thai or Sibumasu (Fig. 1; Metcalfe 1996). The study of the Mesozoic elasmobranch faunas from this country has hitherto been focused on the freshwater assemblages from the Khorat Plateau on the Indochina block, which are mainly Early Cretaceous in age (Cappetta et al. 1990, 2006; Cuny et al. 2006). These faunas show a high level of endemism, although the presence of the genera Lonchidion and Parvodus in the lowermost Cretaceous Sao Khua Formation suggests some relationships between the European and Thailand faunas (Cuny et al. 2006). In contrast, there are few reports on elasmobranchs from the Shan-Thai block (see Cuny et al. 2005, for a review). We describe here the most diverse elasmobranch assemblage found so far in the Jurassic of the southern peninsula of Thailand. The fossils described below are housed at the Sirindhorn Museum, Sahatsakhan, Kalasin Province (SM-TF numbers) and at the Srisuk’s House Museum, Khao Yoi, Phetchaburi Province (SHMAM numbers).

Geological setting The elasmobranch teeth described here come from an outcrop of the Lower Khlong Min Formation near Ao Min village, Thung Song district, Nakhon Si Thammarat Province (Figs 1 and 2). To protect the fossiliferous site, its exact location cannot be provided in this paper, according to Department of Mineral Resources’ policy. For scientific purposes, its global positioning system coordinates can, however, be obtained on request from V. Suteethorn. The Khlong Min Formation is the basal unit of the Mesozoic Thung Yai Group (previously called Trang Group; see Meesook et al. 2005) in the southern peninsula of Thailand; it consists mainly of non-marine mudstone, siltstone, sandstone and limestone (Meesook et al. 2005). Its various fossil contents, including palynoflora, charophytes, bivalves and vertebrates, suggest a late Middle to Late Jurassic age (Buffetaut et al. 2005; Meesook et al. 2005; Philippe et al. 2005). The lower part of the outcrop shows well-bedded calcareous siltstones and sandstones interbedded with thin-bedded pale grey limestone and limestone lenses (Meesook et al. 2005, fig. 8), which have yielded many brackish bivalves, primarily oysters

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 97–113. DOI: 10.1144/SP315.8 0305-8719/09/$15.00 # The Geological Society of London 2009.

98

G. CUNY ET AL.

fragments. Approximately 1500 fossils were recovered and are currently under study by one of us (S. Khamha).

Systematic palaeontology Class Chondrichthyes Huxley Subclass Elasmobranchii Bonaparte Order Hybodontiformes Maisey Family Hybodontidae Owen Subfamily Hybodontinae Owen sensu Maisey 1989 Genus Hybodus Agassiz Hybodus sp. (Fig. 3a–d) Material. SM-TF 9017, one incomplete tooth; SHM-AM 656– 739, SHM-AM 1078–1082, SHM-AM 1433–1987, SHM-AM 1989–1994, 638 tooth fragments.

Fig. 1. Sketch map of part of SE Asia showing the position of the Ao Min locality (black star) and the approximate limits of the main continental blocks (dashed lines). Black squares indicate the position of other localities mentioned in the text: 1, Ao Luk; 2, Pha Dang Zinc Mine. Modified from Buffetaut & Suteethorn (1998).

(Meesook et al. 2005; Philippe et al. 2005) as well as some teleosaurid crocodile remains, a ?pycnodont tooth and an Asteracanthus tooth (the last being mistakenly reported as coming from Mab Ching by Cuny et al. 2005). On top of the oyster-rich beds there is a series of dark grey siltstones and limestone. The elasmobranch remains described here, except for the most complete Asteracanthus tooth, were found in this series in two thin layers of rust-coloured siltstone, 5 m above the top of the uppermost oyster-rich bed. The amount of sediment screen-washed was c. 15 kg from each bed. The same samples have also yielded numerous other types of fish remains, including otoliths, teeth and scales of taxa such as pycnodonts, semionotiforms and perhaps Caturus. Reptile remains include possible pterosaur teeth, teeth and vertebrae of crocodiles, as well as many unidentified bone

Description. Most of the material consists of isolated cusps, some of them preserving part of the root preserved and one or rarely two accessory cusplets. The most complete specimen (SM-TF 9017; Fig. 3a –d) measures 4 mm mesiodistally and is 3.2 mm high. The base of the crown measures 1.3 mm labiolingually. The main cusp is flanked by two lateral cusplets on the preserved side of the crown. The first one is half the height of the main cusp and the second one is half the height of the first one. The main cusp is slightly bent distally. The cutting edge is well developed and is not interrupted between the cusp and cusplets. Cusp and cusplets are moderately compressed labiolingually, with convex labial and lingual faces. Welldeveloped ridges ornament both faces of cusp and cusplets. Some of them attain their apex, and they show some branching. They generally do not reach the crown base, which is smooth. Another fragmentary tooth with part of the root preserved (SHM-AM 1078) shows both lingual and labial ridges reaching the base of the crown. The number of ridges on the labial and lingual faces of the crown is roughly the same. The root is slightly projected lingually, and is deeper lingually than labially. Both root faces show rather small and randomly distributed foramina. Comparisons. Incomplete teeth are inadequate for a determination at a specific level. The teeth described here are, however, generally similar in size, morphology and ornamentation pattern to the 12 incomplete cusps and one incomplete crown recovered from Mab Ching, a site situated 4 km north of Ao Min, and slightly higher in the Khlong Min Formation (Buffetaut et al. 1994; Tong et al. 2002). Only one tooth from Mab Ching is morphologically different in that the main cusp is flanked by a lateral

JURASSIC ELASMOBRANCHS FROM THAILAND

99

Fig. 2. Photograph of Ao Min locality. White arrows indicate the two layers that were sampled.

cusplet, and there is a larger gap between them than can be observed on the teeth from Ao Min. Differences between the Ao Min and Mab Ching teeth are therefore minor, and could be easily explained through monognathic and/or dignathic heterodonty pattern. It is therefore probable that the teeth from Ao Min and those from Mab Ching belong to the same species. Another tooth of Hybodus was recovered from the upper part of the Khlong Min Formation, in a grey calcareous breccia –conglomerate that also yielded crocodile teeth, turtle plates referable to Siamochelys, and an euhelopodid sauropod vertebra. This site lies 20 km south of the town of Khlong Thom in Krabi Province, c. 70 km SW of Ao Min (Fig. 1; Buffetaut et al. 2005). This tooth could not be extracted from the matrix, and only the labial face can be observed. Its general shape is also similar to that of the teeth from Ao Min, with a high cusp flanked by two pairs of lateral cusplets, but it is much larger, measuring at least 10 mm mesiodistally and 5 mm in height. It also shows a different pattern of ornamentation. The ridges are more numerous and not as strong as in the teeth from Ao Min, and they do not reach the apex of

the cusps and cusplets. It is therefore unlikely that the tooth from the upper part of the Khlong Min Formation represents the same species as the one in the lower part of this Formation. Subfamily Acrodontinae Casier sensu Maisey 1989 Genus Asteracanthus Agassiz Asteracanthus sp. (Figs 3e –j and 4) Material. SM-TF 9018, one incomplete crown; SHM-AM 626, 1076, 1077, three crown fragments, SHM-AM 621, one cephalic spine; SHM-AM 622, one cephalic spine fragment. The SM material was collected 5 m below the SHM material (see Geological setting). Description. TF 9018 measures 15.5 mm mesiodistally, 4.5 mm labiolingually, and is 3 mm high (Figs 3e –g and 4). There is a low, dome-shaped main cusp. Only one extremity is preserved, and tapers asymmetrically. In occlusal view, the labial contour of the crown is straight, whereas the lingual one is convex. The preserved extremity and the labial face of the crown are ornamented by reticulate folds forming an irregularly pitted

100

G. CUNY ET AL.

Fig. 3. (a–d) Tooth of Hybodus sp. (SM-TF 9017) in (a) distal, (b) lingual, (c) apical and (d) labial views. (e–j) Asteracanthus sp. (e–g) Anterior tooth (SM-TF 9018) in (e) lingual, (f) apical and (g) labial views. (h– j) Cephalic spine (SHM-AM 621) in (h) lateral, (i) anterior and (j) basal views. (k –r) Lonchidion reesunderwoodi sp. nov., morphotype 1. (k –n) Tooth (SM-TF 9020) in (k) mesial or distal, (l) labial (m) apical and (n) lingual views. (o–r) Tooth (SM-TF 9021) in (o) labial, (p) apical, (q) lingual and (r) mesial or distal views. Scale bars: (a– d) 2 mm; (e– g, i) 5 mm; (h, j) 10 mm; (k, r) 200 mm; (l – q) 500 mm.

JURASSIC ELASMOBRANCHS FROM THAILAND

101

ridge. The rest of the crown is smooth. The lateral and mesial processes of the base have been worn away, and the base is projected posteriorly. Its basal face shows a well-developed anteroposterior groove (Fig. 3j).

Fig. 4. Anterior tooth of Asteracanthus sp. (SM-TF 9018) in (a) lingual, (b) apical and (c) labial views. Scale bars: 5 mm.

surface. The lingual face is ornamented by a dense network of irregular ridges. There is no longitudinal crest. The basal lingual face of the crown makes a smooth recess. The angle between the ornamented upper part of the crown and the smooth lower one is c. 608. As a result, the root insertion area is more reduced than the upper crown area. SHMAM 626, 1076 and 1077, three crown fragments, show no significant difference from SM-TF 9018. The complete cephalic spine SHM-AM 621 measures 28 mm anteroposteriorly and is 17 mm high. The base of the spine is 18 mm anteroposteriorly and 8 mm in maximum width. The crown is hook-shaped, compressed mesiolaterally, and 5 mm thick (Fig. 3h and i). The lower part of the posterior extremity of the crown bears a welldeveloped barb. At the base of the barb, there is a short lateral ridge. There is also a dorsal ridge that runs from the apex to the crown base. It is displaced in a lateral position at the apex and lies in a medial position at the base of the crown. This cephalic spine is therefore a left one (see Maisey 1982). On the mesial side (see Maisey 1982, p. 30, for a definition of mesial), there is a ridge originating from the base of the barb and reaching the dorsal ridge on the medial lower part of the crown. There are some irregular and anastomosed ridges in the basal anterior part of the crown, originating from the dorsal

Comparisons. Teeth of Asteracanthus were previously mentioned from two other sites on the Shan-Thai block (Cuny et al. 2005). A complete crown (SM-TF 9034) and three fragments (SM collection, unnumbered) were found near the town of Ao Luk (Krabi Province), c. 70 km west of Ao Min (Fig. 1), in the upper part of the Khlong Min Formation. In addition, one incomplete tooth (SM-TF 9035) was recovered from the Huai Hin Fon Formation at Pha Dang Zinc Mine (Mae Sot district, Tak Province) in the NW Thailand, some 930 km north of Ao Min (Fig. 1). The Huai Hin Fon Formation was dated as Tithonian on the basis of its ammonite content (Fontaine 1990). The complete crown from Ao Luk (SM-TF 9034) measures 27.1 mm mesiodistally, 14.8 mm labiolingually and 6.9 mm apicobasally (Cuny et al. 2005, fig. 4). The crown is flat without a cusp. In occlusal view, the labial contour is concave whereas the lingual one is slightly convex. The mesial and distal faces are nearly straight. The tooth is attritionally worn, and the enameloid has disappeared in the lingual part of the crown, revealing the underlying columnar dentine. The ornamentation of the enameloid comprises reticulate folds forming a pitted surface. The labial face is ornamented with numerous, irregular and branching ridges that reach the crown base. The tooth from Mae Sot (SM-TF 9035) measures 13.8 mm mesiodistally, 16.4 mm labiolingually and is 9.8 mm high (Cuny et al. 2005, fig. 1). The mesial or distal extremity was manually cut, probably to make a thin section, but the rest of the tooth cannot be found in the SM collection. In apical view, the preserved part of the crown is nearly quadrangular. There is no cusp, but the crown appears higher and more rounded in distal or mesial view than SM-TF 9034. The tooth is attritionally worn and most of the enameloid has disappeared, revealing the underlying columnar dentine. Enameloid is preserved only on the lower part of the labial and lingual face, where it is ornamented with a dense network of irregular and branching ridges that reach the base of the crown on the labial side. Both teeth from Ao Luk and Mae Sot show lingual recesses similar to the one seen in the tooth from Ao Min (SM-TF 9018). However, the angle between the lower smooth and upper ornamented parts of the crown is 858 in the tooth from Ao Luk and 1108 in the tooth from Mae Sot, whereas it is only 608 in the tooth from Ao Min. These recesses probably indicate an interlocking

102

G. CUNY ET AL.

system of the teeth similar to the one described in Asteracanthus cf. A. reticulatus by Rieppel (1981). SM-TF 9018 is interpreted to be an anterior one, whereas those from Ao Luk and Mae Sot are probably laterals (Rees & Underwood 2008), which makes comparison difficult. However, the teeth from Thailand appear close phylogenetically to the group A. magnus þ A. medius defined by Rees & Underwood (2008); that is, taxa possessing teeth with a weak reticulate ornamentation and a dentition with wide and rectangular lateral teeth and weakly arched anterior teeth. The tooth from Ao Luk (SM-TF 9034) appears similar to those of A. magnus from the Bathonian of England and France (Agassiz 1837–1843, plate 18, figs 12 and 13; Woodward 1889, plate 15, fig. 7; Rees & Underwood 2008). As anterior teeth of this species lack a well-developed longitudinal crest (Rees & Underwood 2008), it might be possible that the anterior tooth from Ao Min also belongs to the same species. On the other hand, the more domed crown of the tooth from Mae Sot is more reminiscent of A. medius than of A. magnus. However, the material at hand is insufficient to make a decisive determination as to whether or not all the teeth from Thailand belong to a single species. SHM-AM 621 shares with cephalic spines of Asteracanthus a crown extending farther posteriorly than the base, lack of lateral and mesial cusplet, and a dorsal ridge displaced laterally (Maisey 1982). It differs from cephalic spines of A. ornatissimus by a less developed ornamentation, as well as from the fact that the mesial ridge reaches the dorsal one medially (Maisey 1982). It also differs from those of A. biformatus by a shallower base (Kriwet 1995). Asteracanthus is reported from the Jurassic of Japan (Cappetta 1987; Goto et al. 1991; Goto 1994) on the basis of some anterior teeth (Goto 1994). The tooth figured by Goto et al. (1991) from the Hettangian Niranohama Formation is not well preserved, which does not allow meaningful comparison with the tooth from Ao Min. The tooth of Asteracanthus sp. described by Yabe (1902) from Tomizawa shares with the teeth from Thailand a finely reticulated surface and the lack of a longitudinal crest. However, its coronal contour is rhomboidal, suggesting that it is probably a first anterior tooth, whereas the more elongated tooth from Ao Min is probably a second anterior one, making comparisons between the Japanese and Thailand teeth rather difficult. However, both the tooth from Ao Min and the one from Tomizawa appear close to anterior ones of A. magnus (see above). The tooth described by Yabe (1902) was subsequently designated to the holotype of a new species, A. somaensis (see Goto 1994), which might well be a junior synonym of A. magnus.

Family Lonchidiidae Herman Genus Lonchidion Estes Lonchidion reesunderwoodi sp. nov. (Figs 3k –r and 5) Derivation of name. The specific name is given in honour of Jan Rees and Charlie J. Underwood, for their important contribution to the knowledge of the family Lonchidiidae. Material. SM-TF 9019 (holotype), one tooth; SM-TF 9020–9023 (paratypes), four teeth, SM-TF 9036 (paratypes), five teeth of morphotype 1; SM-TF 9037 (paratypes), 10 teeth of morphotype 2; SM-TF 9038 (paratypes), five teeth of morphotype 3; SHM-AM 628–655, SHM-AM 869–1070, 200 teeth. Type locality. A road cut on road 4124 near the village of Ao Min, Thung Song District, Nakhon Si Thammarat Province. Type stratum. Two thin (1–2 cm thick) layers of rust-coloured siltstone in the lower part of the Khlong Min Formation (upper Middle Jurassic). Diagnosis. Teeth with a sparse ornamentation, becoming more developed on larger teeth; main cusp flanked by up to two pairs of lateral cusplets; longitudinal crest well developed, sometimes crenulated. Three tooth morphotypes present: first one including teeth with a broad, triangular labial peg; second one including teeth with a narrower and more protruding labial peg bearing sometimes a labial accessory cusplet and a faint lingual protuberance; third one including teeth elongated mesiodistally, with higher cusps, narrow labial peg, and a lingual protuberance broader than the labial peg. Description. The largest tooth (SM-TF 9036) measures 4 mm mesiodistally and 1 mm labiolingually, whereas the smallest one (SM-TF 9037) is 1.1 mm mesiodistally and 0.5 mm labiolingually. One tooth (SM-TF 9020) with the root preserved measures 1.3 mm mesiodistally, 0.5 mm labiolingually and 0.8 mm apicobasally, where the root represents half the crown height. These teeth can be divided into three morphotypes. Teeth with a broad, triangular labial peg are here referred to morphotype 1 (Fig. 3k– r, SM-TF 9020 and 9021). The second morphotype encompasses teeth with a narrower and more protruding labial peg and a faint lingual one (Fig. 5a–d, SM-TF 9022). The third morphotype is characterized by teeth with a more mesiodistally elongated crown, narrow labial peg as well as a well-developed lingual peg that is broader than the labial one (Fig. 5e–l, SM-TF 9019 and 9023). The main cusp in morphotypes 1 and 2 is low with a triangular outline in apical view, but

JURASSIC ELASMOBRANCHS FROM THAILAND

103

Fig. 5. Lonchidion reesunderwoodi sp. nov. (a–d) Tooth of morphotype 2 (SM-TF 9022) in (a) labial, (b) apical, (c) lingual and (d) mesial or distal views. (e–l) Teeth of morphotype 3. (e–h) tooth (SM-TF 9023) in (e) mesial or distal, (f) labial, (g) apical and (h) lingual views. (i– l) Holotype (SM-TF 9019) in (i) labial, (j) apical, (k) mesial or distal and (l) lingual views. Scale bars: (a– c, f– j, l) 500 mm; (d, e, k) 200 mm.

becomes higher in morphotype 3. The main cusp is flanked by one or two pairs of low lateral cusplets, which can be high in teeth of morphotype 3. Teeth of morphotype 2 are symmetrical, whereas teeth of morphotypes 1 and 3 can be asymmetrical, with a main cusp displaced mesially(?) and the labial peg slightly bent in a mesial(?) direction. One tooth of morphotype 2 (SM-TF 9022) shows a faint accessory cusp at the extremity of the labial peg (Fig. 5a, b and d). In morphotype 3, the mesial and/or distal extremities of the crown may be bent lingually (e.g. SM-TF 9019). The crown is ornamented by a longitudinal crest, which can appear crenulated on some teeth. There is a ridge ascending the labial peg to reach the longitudinal crest at the top of the main cusp, and in morphotypes 1 and 3, there are up to three secondary ridges branching to

it. On some teeth of morphotype 1 (e.g. SM-TF 9021), a labial ridge can also descend from the top of the lateral cusplets. The largest tooth (SM-TF 9036), which is also of morphotype 1, shows a greater number of ridges, up to three on each cusp and cusplets, both labially and lingually. Some teeth of morphotype 3 (e.g. SM-TF 9019) show a faint horizontal, interrupted ridge at mid-height of the labial crown face. Some teeth of morphotypes 1 and 3 (e.g. SM-TF 9019 and 9021) show a lingual ridge descending the main cusp, which may bifurcate to form a wide inverted Y. None of the labial or lingual ornamentation reaches the crown base, which is smooth. Three teeth of morphotype 1 (SM-TF 9020, 9036) and three of morphotype 3 (SM-TF 9038) have their root preserved. The preserved roots in

104

G. CUNY ET AL.

morphotype 3 are partly obscured by matrix, but do not appear to be different from those of morphotype 1. The root is slightly inclined lingually. The labial face appears slightly concave in mesial or distal view, showing a recess with a row of small foramina on its upper part and a row of larger foramina below the recess. The lingual face is convex in mesial and distal view, and shows large, randomly distributed foramina. The basal face is flat, and shows a medial groove oriented labiolingually in one of the teeth (SM-TF 9020). Comparisons. It is difficult to reconstruct the heterodonty pattern of a shark that is known only from isolated teeth. Teeth of morphotype 3 possess the highest main cusp, and although they represent the most mesiodistally elongated teeth in the sample and are often asymmetrical, they are likely to represent anterolateral teeth. Teeth of morphotype 2 possess a very prominent labial peg, but are less high than teeth of morphotype 3 and show a reduction of the lingual protuberance. They could therefore be either mesial or distal to morphotype 3. Teeth of morphotype 1 are more massive and wider labiolingually than teeth of morphotypes 2 and 3, and with a broad labial peg. They are likely to be posterolateral teeth. Most of the teeth from Ao Min appear slender and narrow labiolingually, and have a low main cusp. The labial protuberance is strongly developed and narrow, especially in morphotypes 2 and 3. They thus appear more similar to the teeth of Lonchidion than those of Lissodus, except for the teeth of morphotype 1, which would be more Lissodus-like (Rees & Underwood 2002). However, teeth of morphotype 3 represent the largest teeth in our sample and are likely to be anterolateral teeth whereas teeth of morphotype 1, which are on average half the mesiodistal length of teeth of morphotype 3, are likely to be posterolateral ones. Lissodus is characterized by an enlargement of the lateral teeth (Rees & Underwood 2002; Milner & Kirkland 2006), and this appears not to be the case here. It is also possible that teeth of morphotype 1 belong to Lissodus, whereas teeth of morphotypes 2 and 3 belong to Lonchidion. However, there is a gradual change from one morphotype to another, and defining two different taxa is therefore difficult. Moreover, the root lacks a horizontal row of small circular foramina near the crown–root junction, which is characteristic of Lissodus (Rees & Underwood 2002). All the teeth from Thailand are thus attributed to the genus Lonchidion rather than to Lissodus. The teeth of Lonchidion striatum from the ?Hauterivian–Barremian of England (Patterson 1966) show a more developed ornamentation than the teeth from Ao Min, whereas the teeth of

Lonchidion anitae from the ?Aptian –Albian of Texas (Duffin 1985) possess more developed cusp and cusplets. The teeth of L. griffisi from the Campanian of Wyoming can be separated from those of L. reesunderwoodi sp. nov. in possessing up to five pairs of lateral cusplets and two accessory cusplets on the labial peg (Case 1987; Duffin 2001). Contrary to the teeth from Ao Min, the teeth of Lonchidion babulskii from the Campanian of New Jersey, L. delsatei from the Toarcian of France, L. khoratensis from the Lower Cretaceous of Thailand, L. humblei from the Carnian of Texas, L. microselachos from the Barremian– Aptian of Spain, and L. selachos from the Maastrichtian of Wyoming and Campanian of Texas are devoid of ornamentation (Estes 1964; Cappetta & Case 1975; Murry 1981; Duffin 1985, 2001; Welton & Farish 1993; Cuny et al. 2006; Heckert et al. 2007). The teeth from Ao Min share with those of Lonchidion marocensis from the ?Berriasian of Morocco and those of L. weltoni from the Cenomanian of Oregon a lingual protuberance at the base of the main cusp (Duffin 1985; Duffin & SigogneauRussell 1993). However, teeth of L. marocensis lack an accessory cusplet on the labial peg (Duffin & Sigogneau-Russell 1993), whereas those of L. weltoni lack a crenulate longitudinal crest (Duffin 1985). The teeth from Ao Min do not possess a crown as strongly angled as those of Lonchidion inflexum (Underwood & Rees 2002). The species L. triaktis from the Triassic of Russia is better attributed to the genus Lonchidion than to Lissodus based on the narrow crown of its teeth showing a strongly developed and narrow labial peg (Minikh 2001). Its teeth can be separated from those from Ao Min because they lack a lingual protuberance at the base of the main cusp. Teeth of Lonchidion crenulatum from the Berriasian –Valanginian of England (Patterson 1966; Duffin 1985; Underwood & Rees 2002) share with the teeth from Ao Min the presence of an occasional accessory cusplet on the labial peg and a crenulated longitudinal crest. However, their crowns are more ornamented than the ones from Thailand. The teeth from Ao Min are similar to those of Lonchidion breve from the Valanginian– Barremian of England (Patterson 1966) in sharing similar crown morphology and sparse ornamentation. However, the teeth of L. breve do not show an inverted Y-shaped ridge on the lingual face of the main cusp. They lack a faint accessory cusp at the extremity of the labial peg, a feature that can be seen in one tooth from Ao Min (Fig. 5a, b and d). Moreover, mesiodistally elongated teeth of L. breve are not the ones with the highest cusps (Patterson 1966, fig. 15C) although they do possess a rather high cusp when compared with

JURASSIC ELASMOBRANCHS FROM THAILAND

other teeth of the same species (Patterson 1966, figs 14– 16). This may indicate that anterior teeth similar to those of L. breve (Patterson 1966, fig. 16A) are missing from our sample from Thailand. These teeth in L. breve are similar to our morphotype 2, albeit with higher cusps. It should also be noted that elongated teeth of L. crenulatum possess a rather high main cusp (Patterson 1966, fig. 17C). L. reesunderwoodi sp. nov., L. breve and L. crenulatum might, therefore, share a similar heterodont pattern, and could be closely related to each other. Subcohort Neoselachii Compagno Superorder Batoidea Compagno Order Rhinobatiformes Compagno Incertae familiae Genus Belemnobatis Thiollie`re Belemnobatis aominensis sp. nov. (Figs 6, 7 and 8a–d) Derivation of name. The specific name is derived from Ao Min, the type locality. Material. SM-TF 9024 (holotype), one tooth; SM-TF 9025–9029 (paratypes), five teeth; SM-TF 9039 (paratypes), 49 teeth; SHM-AM 497 –620, SHM-AM 839 –850, SHM-AM 1071–1072, 71 teeth. Type locality. A road cut on road 4124 near the village of Ao Min, Thung Song District, Nakhon Si Thammarat Province. Type stratum. Two thin (1–2 cm thick) layers of rust-coloured siltstone in the lower part of the Khlong Min Formation (upper Middle Jurassic). Diagnosis. Belemnobatis with a dignathic heterodonty: upper teeth with a projected labial protuberance flanked by one or two pairs of bulges and a rectilinear to convex longitudinal crest; lower teeth with moderate labial protuberance and a rectilinear to lingually concave longitudinal crest; root with a basal median groove partially roofed lingually in some teeth; foramina on either side of the uvula reduced. Description. Teeth of this taxon show a rather homogeneous size and considerable morphological variations (Figs 6 and 8a–d). The largest tooth measures 1.1 mm mesiodistally and 0.9 mm labiolingually, whereas the smallest one measures 0.7 mm mesiodistally and 0.5 mm labiolingually. They are generally wider mesiodistally than long labiolingually, except for one tooth (SM-TF 9025; Fig. 6b), which measures 0.8 mm mesiodistally and 0.9 mm labiolingually. The crown is low and cruciform in apical view. It is flat to slightly convex in labial view without a well-defined main cusp. The lingual uvula is narrow and elongated, longer than wide, and tapering lingually or

105

showing a rather angular lingual extremity. It may reach the lingual extremity of the root. In mesial or distal view, the uvula often shows a concavity corresponding to a lingual wear facet. One tooth (SM-TF 9026; Fig. 6g) shows a vestigial lingual uvula, which might be pathological. The labial face of the crown is flat. A well-developed labial protuberance overhangs the root without being basally inclined. It commonly shows a circular wear facet at its labiobasal extremity, corresponding to the one observed on the lingual uvula. The teeth with a particularly protruding labial protuberance show one or two small bulges on each side of the protuberance in apical view. The longitudinal crest is well developed. It is rectilinear to concave lingually in teeth showing a moderate labial protuberance, whereas it is rectilinear to convex lingually in teeth showing a protruding labial protuberance. Many teeth are slightly asymmetrical, with a distal part more elongated than the mesial one, probably indicating a lateral, rather than anterior, position on the jaw. The root is moderately projected lingually and shows a deep median groove (holaulacorhize root vascularization), which is sometimes partly roofed lingually (hemiaulacorhize root vascularization). This groove widens labially, and shows a central foramen in its middle part. It separates two massive lobes with a triangular outline in basal view. It opens lingually in the continuation of the lingual uvula or is slightly displaced mesially or distally. In apical view, a notch corresponding to the lingual opening of the median canal may be developed, where it is almost nonexistent in some teeth. However, the root is damaged in many teeth, and it is likely that the lingual notch was secondarily abraded on some teeth. There is a very small marginolingual foramen on each side of the uvula, which is barely visible. The linguomesial and linguodistal margins of the root are rectilinear to concave. Enameloid microstructure. One tooth (SM-TF 9028) was etched for 70 s in 10% hydrochloric acid and the surface of the tooth was observed with a scanning electron microscope. The micrographs reveal only a single crystallite enameloid made of small (1–2 mm in length) rod-shaped crystallites of apatite with no preferential orientation (Fig. 7). Comparisons. Two types of teeth can be recognized among our sample. The first type is with a protruding labial protuberance flanked by one or two pairs of bulges and a lingually convex longitudinal crest (Fig. 6a–l, SM-TF 9024–9026). The second type is with a moderate labial protuberance and a lingually concave longitudinal crest (Figs 6m–p and 8a–d, SM-TF 9027 and 9029). The change

106

G. CUNY ET AL.

Fig. 6. Belemnobatis aominensis sp. nov. (a –d) Upper tooth (SM-TF 9025) in (a) labial, (b) apical, (c) lingual and (d) mesial or distal views. (e–h) Upper tooth (SM-TF 9026) in (e) labial, (f) mesial or distal, (g) apical and (h) lingual views. (i – l) Upper tooth, holotype (SM-TF 9024), in (i) mesial or distal, (j) lingual, (k) labial and (l) apical views. (m–p) Lower tooth (SM-TF 9027) in (m) mesial or distal, (n) apical (o) lingual and (p) labial views. Scale bars: 500 mm.

JURASSIC ELASMOBRANCHS FROM THAILAND

Fig. 7. Surface of the enameloid of a tooth of Belemnobatis aominensis sp. nov. (SM-TF 9028) etched 70 s in 10% hydrochloric acid.

of orientation of the longitudinal crest suggests that this corresponds to a dignathic heterodonty. In Belemnobatis morinicus, the teeth with a well-developed labial protuberance flanked by bulges correspond to the upper teeth (Cavin et al. 1995). Teeth of Belemnobatis aominensis sp. nov. are easily separated from those of Engaibatis by the lack of lateral cusplets (Arratia et al. 2002) and from those of Rhinobatos by the lack of marginal lingual uvulae (the genus Rhinobatos is here restricted to specimens possessing teeth with marginal uvulae, Cappetta & Case 1999; Cuny et al. 2004). They also differ from those of Rhombopterygia and Iansan by the presence of a well-developed protuberance on the labial apron (Cappetta 1980; Maisey 1991). In addition, teeth of Iansan possess lateral uvulae, and a rather high crown compared with those of B. aominensis sp. nov. (Maisey 1991; Brito & Se´ret 1996). The taxonomy of Jurassic and Early Cretaceous batoids, such as Belemnobatis, Spathobatis and Asterodermus, is currently in a state of flux (Kriwet & Klug 2004; Rees 2005). However, their low and mesiodistally expanded crown as well as the narrow lingual uvula suggest that the teeth from Thailand belong to the genus Belemnobatis, rather than Spathobatis (Cappetta 1987; Cavin et al. 1995; Underwood et al. 1999; Underwood & Rees 2002; Underwood 2004a; Underwood & Ward 2004b). The tooth tentatively attributed to Asterodermus by Leidner & Thies (1999, fig. 3G) is figured only in labial view, so meaningful comparisons with the teeth from Thailand are difficult. Moreover, Asterodermus is probably a junior synonym of Spathobatis (Underwood & Ward 2004a; Rees 2005; see, however, Kriwet & Klug 2004, for a different opinion).

107

Teeth of Belemnobatis aominensis sp. nov. differ from those of previously described Belemnobatis species. For example, the lower teeth of B. aominensis sp. nov. differ from those of B. sismondae from the Kimmeridgian of France (Cavin et al. 1995) in possessing a labial apron that is less developed, giving the crown a more slender appearance. At the level of the root, they also lack the large marginolingual foramina seen in the teeth of B. sismondae. Furthermore, a partial roofing of the basal median groove was not noted in the latter species (Cavin et al. 1995). Contrary to teeth of B. aominensis sp. nov., those of B. werneri from the Callovian of England (Thies 1983) show a more protruding and narrower labial protuberance as well as faint lateral lingual uvulae (Thies 1983), whereas those of B.? moorbergensis from the Toarcian of Germany possess a faint main cusp and a more angular labial protuberance (Thies 1983). Teeth of both B. morinicus from the Tithonian of France and B. variabilis from the lowermost Cretaceous Purbeck Group in England (Cavin et al. 1995; Underwood & Rees 2002) possess a higher crown than the teeth from Thailand, and show larger marginolingual foramina. Moreover, these two species do not show a partially roofed basal groove (Cavin et al. 1995; Underwood & Rees 2002). The lack of large marginolingual foramina allows us to differentiate B. aominensis sp. nov. teeth from those of Belemnobatis sp. from the Valanginian of Poland (Rees 2005) as well as from those of B. picteti (Cappetta 1975; Underwood 2004b). ‘Belemnobatis’ noviodunumensis from the Aalenian of France appears to differ from B. aominensis sp. nov., by the presence of a higher cusp (Delsate & Candoni 2001), but the lack of proper description and illustration of this species makes comparisons difficult. The upper teeth of Belemnobatis aominensis sp. nov. are very similar in overall morphology to those of B. stahli from the Bathonian of England (Underwood & Ward 2004a), but the lower teeth are different in lacking a protruding labial protuberance. On the contrary, the lower teeth of B. aominensis sp. nov. are similar to those of B. kermacki, also from the Bathonian of England. Belemnobatis stahli and B. kermacki were found in different environments (Underwood 2004a; Underwood & Ward 2004b), and they cannot represent the upper and lower dentition of the same species. One should therefore consider the possibility that the teeth referred to B. aominensis sp. nov. represent two different species as well. However, in contrast to what has been observed in England (Underwood 2004a; Underwood & Ward 2004b), all the teeth come from the same layers in Thailand and a similar dignathic heterodonty has been observed in B. morinicus, known from a complete specimen (Cavin et al. 1995). Moreover, the teeth of B.

108

G. CUNY ET AL.

Fig. 8. (a–d) Lower tooth (SM-TF 9029) of Belemnobatis aominensis sp. nov. in (a) labial, (b) apical, (c) lingual and (d) mesial or distal views. (e –h) Tooth (SM-TF 9030) of Belemnobatis sp. in (e) mesial or distal, (f) labial, (g) apical and (h) lingual views. (i, j) Hybodont dermal denticle (SM-TF 9031) in (i) anterior, and (j) lateral views. (k– m) Dermal denticle of Belemnobatis (SM-TF 9032) in (k) apical, (l) lateral and (m) anterior or posterior views. (n– q) Hybodont dermal denticle (SM-TF 9033) in (n) anterior, (o) apical, (p) lateral and (q) posterior views. Scale bars: 500 mm.

JURASSIC ELASMOBRANCHS FROM THAILAND

kermacki possess larger marginolingual foramina than those of B. stahli (Underwood & Ward 2004b), a difference that is not found between the lower and upper teeth from Thailand. It is therefore more parsimonious to consider that the teeth from Thailand represent a single new species. Belemnobatis sp. (Fig. 8e –h) Material. SM-TF 9030, one tooth. Description. SM-TF 9030 differs from the teeth of B. aominensis sp. nov. by the presence of a medial cusp and the lack of longitudinal crest. The crown also shows faint marginal lingual uvulae. It measures 1.2 mm mesiodistally and 1 mm labiolingually, and is therefore slightly larger than the teeth of B. aominensis sp. nov. Comparisons. The presence of a well-developed medial cusp in the tooth described above is reminiscent of Spathobatis. However, it possesses a lingual uvula that is more elongated than in the teeth of Spathobatis bugesiacus (including S. uppensis and S. mutterlosei) and those of S. halteri (Kriwet 1999; Underwood 2002). Teeth of S. rugosus and S. delsatei have a long uvula, but show teeth with a more robust crown (Underwood et al. 1999; Underwood & Ward 2004a). Based on its mesiodistally expanded crown and its long and narrow lingual uvula, the tooth from Thailand is therefore better attributed to the genus Belemnobatis than Spathobatis. It differs from teeth of B.? moorbergensis by a more developed main cusp and the lack of a longitudinal crest, although the latter could be the result of wear. Gynandric heterodonty is well documented among Myliobatiformes and Rajiformes (e.g. see Feduccia & Slaughter 1974; Nordell 1994), as well as in more primitive batoids such as Rhinobatos casieri from the Campanian of Texas and R. incertus from the Turonian –Coniacian of Texas (Welton & Farish 1993). One may therefore consider SM-TF 9030 as belonging to a male Belemnobatis aominensis sp. nov., because it differs mainly from the teeth of the latter by the presence of a more developed main cusp. However, to consider this hypothesis, one must explain the rarity of this kind of teeth in our sample: one cuspated tooth against 126 non-cuspated teeth. One possibility is sexual segregation between male and female outside the reproductive period. Another possibility would be a periodic shift in male dentition, like that observed in Dasyatis sabina, from a female-like molariform to a cuspate form during the reproductive season (Kajiura & Tricas 1996). However, such hypotheses are impossible to demonstrate based on the available material, and do not explain the presence of faint marginal lingual uvulae in SM-TF 9030. The tooth

109

is therefore better considered as belonging to Belemnobatis sp., although we cannot demonstrate the presence of a second species of Belemnobatis in our sample. Elasmobranchii Bonaparte (Fig. 8i –q) Material. SM-TF 9031– 9033, three dermal denticles; SM-TF 9040, 25 dermal denticles; SHM-AM 740 –823, SHM-AM 1083–1432, SHM-AM 1988, 406 dermal denticles. Description. Most of the material is represented by dermal denticles up to 1.4 mm in diameter. The crown is generally flat with a well-developed and pointed apex, which can project posteriorly beyond the base in some dermal denticles (SM-TF 9033; Fig. 8n– q). The crown is ornamented by ridges radiating from the apex to the base of the crown. They divide basally up to three times. However, this ornamentation has been worn away in many specimens. In some specimens, the crown is more erect, and may also be globular, without any difference in the ornamentation pattern (Fig. 8i and j, SM-TF 9031). The base of these dermal denticles is circular in outline and is larger than the crown base on all sides. The basal face is flat to slightly convex without a central foramen. A single dermal denticle (SM-TF 9032) appears different from the type described above. Its crown is reduced, forming a narrow crest orientated anteroposteriorly, with a convex outline in lateral view (Fig. 8k–m). In apical view, the base has a circular outline with a diameter of 0.9 mm. It shows an ornamentation made of 10 radiating ridges. The basal face is slightly convex. Comparisons. The first type of dermal denticles, although showing a variety of crown shapes, shares a common ornamentation pattern, and they are all likely to belong to the same elasmobranch species. Based on similar denticles found in Mab Ching, Cuny et al. (2005) tentatively attributed these dermal denticles to hybodont sharks, as only hybodont teeth had been recovered from the Khlong Min Formation at that time. The discovery of Belemnobatis adds to the list of elasmobranchs known from this formation. Dermal denticles of Jurassic rhinobatoids show variable crown morphology, but they seem to all possess a base with a radiating ornamentation (Leidner & Thies 1999). Such ornamentation is lacking in the first type of dermal denticles from Thailand, but is found in the second type. The latter has an unusually reduced crown, but it probably belongs to Belemnobatis. The first type of denticles is more likely to belong to Hybodontiformes, based on the heavy ornamentation of the crown. They were found together only with Hybodus teeth in Mab Ching, and with both

110

G. CUNY ET AL.

Lonchidion, Asteracanthus and Hybodus remains in Ao Min, whereas such dermal denticles have never been found previously in association with any of these genera. Their exact affinities remain therefore difficult to decipher.

Discussion Belemnobatis forms a clade of basal batoids together with Spathobatis, although their exact position on the batoid phylogenetic tree remains poorly understood (Brito & Se´ret 1996; Underwood 2006). Members of this clade were hitherto known only from Europe (Cavin et al. 1995; Underwood & Rees 2002; Underwood 2004a, b; Underwood & Ward 2004b; Rees 2005). Therefore, Belemnobatis from Thailand reported here greatly expands the geographical distribution of these early batoids. Another early batoid has been found in the Tithonian of Argentina, but it belongs to a different, as yet unnamed genus (Cione 1999). Interestingly, the rest of the elasmobranch fauna from Ao Min shows a similar taxonomic composition to European faunas. Lonchidion reesunderwoodi sp. nov. appears phylogenetically closer to L. breve than to any other Lonchidion species. Teeth of Asteracanthus from Thailand are similar to those of A. magnus and A. medius from Europe. Cuny et al. (2006) proposed a European origin for some of the Early Cretaceous hybodont sharks from Thailand, and the present study supports this contention. Belemnobatis aominensis sp. nov. appears phylogenetically closer to B. stahli and B. kermacki from the Bathonian of England than to any other species of Belemnobatis, and the Asteracanthus described here is interpreted to be close to A. magnus and A. medius, which are restricted to the Bathonian–Callovian interval (Rees & Underwood 2009). This strongly suggests a late Middle Jurassic age for the lower part of the Khlong Min Formation, which is in agreement with previous studies on this formation (Buffetaut et al. 2005; Meesook et al. 2005; Philippe et al. 2005). Lonchidion reesunderwoodi sp. nov., on the other hand, appears closely related to Early Cretaceous Lonchidion species, most notably L. breve, but this comparison is biased by the absence of Lonchidion species in Middle and Upper Jurassic outcrops. The study of the enameloid microstructure of Belemnobatis aominensis sp. nov. confirms the absence of a triple-layered enameloid in primitive batoids, as demonstrated by Thies (1983) on Belemnobatis? moorbergensis. Underwood (2006) considered this as a derived character of the Batoidea. Because two putative primitive batoids from the Early Jurassic, Jurobatos and Doliobatis, were

reported to possess a multilayered enameloid (Thies 1983; Delsate 2003), Underwood (2006) considered that the single crystallite enameloid of batoids is a reversal from the neoselachian multilayered enameloid. The latter was considered by Underwood (2006) as the primitive state for neoselachians because many Triassic non-batoid neoselachians possess a triple-layered enameloid (Cuny 1998; Cuny & Benton 1999; Cuny et al. 2001; Cuny & Risnes 2005). He also justified this polarization of the character by the fact that posterior teeth of Heterodontus possess a monolayered enameloid. This interpretation, however, is arguable and needs further discussion. First, the report of a multilayered enameloid in Doliobatis is based on a misinterpretation of its tooth histology. The tissue that was interpreted as a tangled–bundled enameloid (also called tangled–fibred enameloid) is in fact the dentine (see Delsate 2003, plate 3, fig. 4, where one can clearly see the absence of crystallites of apatite in the tissue). Concerning Jurobatos cappettai, a revision of this species by Thies (1993) concluded that this taxon was not a batoid, but a galean shark, a conclusion we fully agree with. Finally, posterior teeth of Heterodontus do not show a single-layered enameloid, but a multilayered one made of shiny layered enameloid and tangled–bundled enameloid (Reif 1973). Monolayered enameloid has therefore never been documented among non-batoid neoselachians, only the loss of the parallel-bundled enameloid (also called parallel-fibred enameloid: see Cuny et al. 2001) in their triple-layered enameloid. Because a single-crystallite enameloid is the primitive state observed in the two successive sister-groups of the Neoselachii, the Ctenacanthiformes and Hybodontiformes (Reif 1973; Gillis & Donoghue 2007), it appears more parsimonious to consider that the Batoidea has retained a primitive state of enameloid microstructure (Maisey et al. 2004; Rees & Cuny 2007) rather than to consider their monolayered enameloid as a derived character, which was reversed from the neoselachian state. Recent phylogenetic analysis, based both on molecular and morphological data, indicates that batoids are the sister-group of the rest of the neoselachians (Douady et al. 2003; Maisey et al. 2004; McEachran & Aschliman 2004; Winchell et al. 2004). They therefore diverged from the rest of the neoselachians before the latter developed a triple-layered enameloid. As the oldest undisputable neoselachian shark, Synechodus antiquus, has been found in the Lower Permian units of Russia (Ivanov 2005), this event can be dated at about the Carboniferous –Permian boundary. However, the oldest batoids have not been reported in the fossil record before the Early Jurassic (Underwood 2006). This stratigraphic gap might only be apparent, however, because Rees & Cuny (2007) have

JURASSIC ELASMOBRANCHS FROM THAILAND

recently proposed that some Triassic sharks of uncertain affinities, such as Doratodus and Vallisia, could indeed be primitive batoids. Doratodus is known as early as the Anisian (Cuny 1998). One may also wonder whether Cooleyella (Duffin & Ward 1983; Duffin et al. 1996) is not a primitive batoid as well, which would indicate a divergence date between batoid and non-batoid neoselachians as early as the Early Carboniferous (Duffin & Ward 1983). In that case, and contrary to Underwood’s (2006) assumption, the characteristic batoid tooth design appeared well after their divergence from the other neoselachians.

Conclusion The elasmobranch fauna from the lower part of the Khlong Min Formation in Thailand includes Hybdodus sp., Asteracanthus sp., Lonchidion reesunderwoodi sp. nov., Belemnobatis aominensis sp. nov., and possibly a second species of Belemnobatis. This fauna supports a Bathonian–Callovian age for the lower part of the Khlong Min Formation, and shows a strong similarity to the Middle Jurassic elasmobranch faunas of Europe. This paper marks the first report of the genus Belemnobatis outside Europe, and the first record of the genus Lonchidion from the Middle Jurassic. The presence of a monolayered enameloid in Belemnobatis aominensis sp. nov. and other primitive batoids is interpreted as the retention of a primitive character for neoselachians, which would suggest a divergence time between the batoids and the rest of the neoselachians as early as the Carboniferous– Permian boundary. This work was supported by the Danish Natural Science Research Council and the TRF-CNRS Special Programme for Biodiversity Research and Training Programme (BRT/ BIOTEC/NSTDA) Grant BRT R_245007, as well as by the Carlsberg Foundation, the Department of Mineral Resources in Bangkok, the University of Maha Sarakham, the Jurassic Foundation, the CNRS ECLIPSE Programme, and the Institut National des Sciences de l’Univers from the CNRS. We are indebted to K. Shimada and D. Ward, whose reviews greatly improved the initial version of the manuscript. We would also like to thank all the people who took part in fieldwork, including E. Buffetaut, P. Bunchalee, L. Cavin, S. Chitsing, J. Claude, U. Deesri, S. Khansubha, K. Lauprasert, J. Le Loeuff, M. Philippe, T. Saenyamoon, C. Souillat, S. Suteethorn and S. Trisivakul. One of us (G. C.) is indebted to I. Sasagawa for providing us with a copy of Yabe’s (1902) paper.

References A GASSIZ , L. 1837– 1843. Recherches sur les poissons fossiles. Tome 3 contenant l’histoire de l’ordre des placoides. Imprimerie de Petitpierre, Neufchaˆtel.

111

A RRATIA , G., K RIWET , J. & H EINRICH , W.-D. 2002. Selachians and actinopterygians from the Upper Jurassic of Tendaguru, Tanzania. Mitteilungen aus dem Museum fu¨r Naturkunde in Berlin, Geowissenschaftliche Reihe, 5, 207–230. B RITO , P. M. & S E´ RET , B. 1996. The new genus Iansan (Chondrichthyes, Rhinobatoidea) from the Early Cretaceous of Brazil and its phylogenetic relationships. In: A RRATIA , G. & V IOHL , G. (eds) Mesozoic Fishes—Systematics, Paleoecology. Dr Friedrich Pfeil, Mu¨nchen, 47–62. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 83– 90. B UFFETAUT , E., R AKSASKULWONG , L., S UTEETHORN , V. & T ONG , H. 1994. First post-Triassic temnospondyl amphibians from the Shan-Thai block: Intercentra from the Jurassic of peninsular Thailand. Geological Magazine, 131, 837–839. B UFFETAUT , E., S UTEETHORN , V., T ONG , H. & K OSˇ IR , A. 2005. First dinosaur from the Shan-Thai block of SE Asia: A Jurassic sauropod from the southern peninsula of Thailand. Journal of the Geological Society, London, 162, 481–484. C APPETTA , H. 1975. Se´laciens et holoce´phale du Gargasien de la re´gion de Gargas (Vaucluse). Ge´ologie Me´diterrane´enne, 2, 115– 134. C APPETTA , H. 1980. Les se´laciens du Cre´tace´ supe´rieur du Liban. II: batoı¨des. Palaeontographica A, 168, 149– 229. C APPETTA , H. 1987. Chondrichthyes II: Mesozoic and Cenozoic Elasmobranchii. Handbook of Paleoichthyology, 3B. Gustav Fischer, Stuttgart. C APPETTA , H., B UFFETAUT , E. & S UTEETHORN , V. 1990. A new hybodont from the Lower Cretaceous of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 1990, 659–666. C APPETTA , H. & C ASE , G. R. 1975. Contribution a` l’e´tude des se´laciens du groupe Monmouth (Campanien– Maestrichtien) du New-Jersey. Palaeontographica, Abteilung A, 151, 1– 46. C APPETTA , H. & C ASE , G. R. 1999. Additions aux faunes de se´laciens du Cre´tace´ du Texas (Albien supe´rieur–Campanien). Palaeo Ichthyologica, 9, 5– 111. C APPETTA , H., B UFFETAUT , E., C UNY , G. & S UTEETHORN , V. 2006. A new elasmobranch assemblage from the Lower Cretaceous of Thailand. Palaeontology, 49, 547–555. C ASE , G. R. 1987. A new selachian fauna from the Late Campanian of Wyoming (Teapot Sandstone Member, Mesaverde Formation, Big Horn Basin). Palaeontographica, Abteilung A, 197, 1– 37. C AVIN , L., C APPETTA , H. & S E´ RET , B. 1995. Re´vision de Belemnobatis morinicus (Sauvage, 1873) du Portlandien du Boulonnais (Pas-de-Calais, France). Comparaison avec quelques Rhinobatidae jurassiques. Geologica et Palaeontologica, 29, 245–267. C IONE , A. L. 1999. First report of a Jurassic ray outside of Europe. In: A RRATIA , G. & S CHULTZE , H.-P. (eds) Mesozoic Fishes 2—Systematics and Fossil Record. Dr Friedrich Pfeil, Mu¨nchen, 21–28.

112

G. CUNY ET AL.

C UNY , G. 1998. Primitive neoselachian sharks: A survey. Oryctos, 1, 3 –21. C UNY , G. & B ENTON , M. J. 1999. Early radiation of the Neoselachian sharks in Western Europe. Geobios, 32, 193– 204. C UNY , G. & R ISNES , S. 2005. The enameloid microstructure of the teeth of Synechodontiform sharks (Chondrichthyes: Neoselachii). PalArch, 3, 8 –19. C UNY , G., R IEPPEL , O. & S ANDER , P. M. 2001. The shark fauna from the Middle Triassic (Anisian) of NorthWestern Nevada. Zoological Journal of the Linnean Society, 133, 285 –301. C UNY , G., O UAJA , M., S RARFI , D., S CHMITZ , L., B UFFETAUT , E. & B ENTON , M. J. 2004. Fossil sharks from the Early Cretaceous of Tunisia. Revue de Pale´obiologie, Volume Spe´cial 9, 127– 142. C UNY , G., S UTEETHORN , V. & K AMHA , S. 2005. A review of the hybodont sharks from the Mesozoic of Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 588– 593. C UNY , G., S UTEETHORN , V., K AMHA , S., B UFFETAUT , E. & P HILIPPE , M. 2006. A new hybodont shark assemblage from the Lower Cretaceous of Thailand. Historical Biology, 18, 21–31. D ELSATE , D. 2003. Une nouvelle faune de poissons et requins toarciens du Sud du Luxembourg (Dudelange) et de l’Allemagne (Scho¨mberg). Bulletin de l’Acade´mie Lorraine des Sciences, 42, 13–49. D ELSATE , D. & C ANDONI , L. 2001. Description de nouveaux morphotypes dentaires de Batomorphii toarciens (Jurassique infe´rieur) du Bassin de Paris: Archaeobatidae nov. fam. Bulletin de la Socie´te´ Naturelle du Luxembourg, 102, 131– 143. D OUADY , C. J., D OSAY , M., S HIVJI , M. S. & S TANHOPE , M. J. 2003. Molecular phylogenetic evidence refuting the hypothesis of Batoidea (rays and skates) as derived sharks. Molecular Phylogenetics and Evolution, 26, 215 –221. D UFFIN , C. J. 1985. Revision of the hybodont selachian genus Lissodus Brough (1935). Palaeontographica, A, 188, 105– 152. D UFFIN , C. J. 2001. Synopsis of the selachian genus Lissodus Brough, 1935. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Abhandlungen, 221, 145– 218. D UFFIN , C. J. & S IGOGNEAU -R USSELL , D. 1993. Fossil shark teeth from the Early Cretaceous of Anoual, Morocco. In: H ERMAN , J. & VAN W AES , H. (eds) Elasmobranches et stratigraphie. Belgian Geological Survey, Professional Papers, 264, 175 –190. D UFFIN , C. J. & W ARD , D. J. 1983. Neoselachian sharks’ teeth from the Lower Carboniferous of Britain and the Lower Permian of the U.S.A. Palaeontology, 26, 93–110. D UFFIN , C. J., R ICHTER , M. & N EIS , P. A. 1996. Shark remains from the Late Carboniferous of the Amazon Basin, Brazil. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 1996, 232– 256. E STES , R. 1964. Fossil vertebrates from the late Cretaceous Lance Formation, Eastern Wyoming. University

of California, Publications in Geological Sciences, 49, 1 –180. F EDUCCIA , A. & S LAUGHTER , B. H. 1974. Sexual dimorphism in skates (Rajidae) and its possible role in differential niche utilization. Evolution, 28, 164–168. F ONTAINE , H. 1990. The Upper Palaeozoic and Mesozoic fossils of West Thailand. In: FONTAINE , H. (ed.) Ten years of CCOP research on the pre-Tertiary of East Asia. CCOP technical Secretariat, Bangkok, 353– 357. G ILLIS , J. A. & D ONOGHUE , P. C. J. 2007. The homology and phylogeny of chondrichthyan tooth enameloid. Journal of Morphology, 268, 33–49. G OTO , M. 1994. Palaeozoic and early Mesozoic fish faunas of the Japanese islands. Island Arc, 3, 247– 254. G OTO , M., K UGA , N. & H ACHIYA , K. 1991. On the hybodont elasmobranch teeth of three genera from the Mesozoic of Japan. Journal of the Geological Society of Japan, 97, 743–750. H ECKERT , A. B., I VANOV , I. & L UCAS , S. G. 2007. Dental morphology of the hybodontoid shark Lonchidion humblei Murry from the Upper Triassic Chinle Group, USA. New Mexico Museum of Natural History and Science Bulletin, 41, 45–48. I VANOV , A. 2005. Early Permian chondrichthyans of the Middle and South Urals. Revista Brasileira de Paleontologia, 8, 127 –138. K AJIURA , S. M. & T RICAS , T. C. 1996. Seasonal dynamics of dental sexual dimorphism in the Atlantic stingray Dasyatis sabina. Journal of Experimental Biology, 199, 2297– 2306. K RIWET , J. 1995. Beitrag zur Kenntnis des Fisch-fauna des Ober-Jura (unteres Kimmeridge) der Kohlengrube Guimarota bei Leiria, Mittel-Portugal: 1. Asteracanthus biformatus n. sp. (Chondrichthyes: Hybodontoidea). Berliner Geowissenschaftliche Abhandlungen, E16, 683–691. K RIWET , J. 1999. Neoselachier (Pisces, Elasmobranchii) aus der Unterkreide (unteres Barremium) von Galve und Alcaine (Spanien, Provinz Teruel). Palaeo Ichthyologica, 9, 113–142. K RIWET , J. & K LUG , S. 2004. Late Jurassic selachians (Chondrichthyes, Elasmobranchii) from southern Germany: Re-evaluation on taxonomy and diversity. Zitteliana, A44, 67– 95. L EIDNER , A. & T HIES , D. 1999. Placoid scales and oral teeth of Late Jurassic elasmobranchs from Europe. In: A RRATIA , G. & S CHULTZE , H.-P. (eds) Mesozoic Fishes 2—Systematics and Fossil Record. Dr Friedrich Pfeil, Mu¨nchen, 29– 40. M AISEY , J. G. 1982. The anatomy and interrelationships of Mesozoic hybodont sharks. American Museum Novitates, 2724, 1– 48. M AISEY , J. G. 1989. Hamiltonichthys mapesi, g. and sp. nov. (Chondrichthyes; Elasmobranchii), from the Upper Pennsylvanian of Kansas. American Museum Novitates, 2931, 1– 42. M AISEY , J. G. 1991. Rhinobatos Link, 1790. In: M AISEY , J. G. (ed.) Santana Fossils: An Illustrated Atlas. T.F.H. Publications, Neptune City, NJ, 110– 115. M AISEY , J. G., N AYLOR , G. J. P. & W ARD , D. J. 2004. Mesozoic elasmobranchs, neoselachian phylogeny and the rise of modern elasmobranch diversity. In: A RRATIA , G. & T INTORI , A. (eds) Mesozoic Fishes

JURASSIC ELASMOBRANCHS FROM THAILAND 3—Systematics, Paleoenvironment and Biodiversity. Dr Friedrich Pfeil, Mu¨nchen, 17– 56. M C E ACHRAN , J. D. & A SCHLIMAN , N. 2004. Phylogeny of Batoidea. In: C ARRIER , J. C., M USICK , J. & H EITHAUS , M. R. (eds) Biology of Sharks and their Relatives. CRC Press, Boca Raton, FL, 79– 113. M EESOOK , A., S AENGSRICHAN , W. & T EERARUNGSIGUL , N. 2005. Stratigraphy and faunal aspects of the marine Jurassic rocks in southern peninsular Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 558–569. M ETCALFE , I. 1996. Gondwanaland dispersion, Asian accretion and evolution of eastern Tethys. Australian Journal of Earth Sciences, 43, 605 –623. M ILNER , A. R. C. & K IRKLAND , J. I. 2006. Preliminary review of the Early Jurassic (Hettangian) freshwater Lake Dixie fish fauna in the Whitmore Point Member, Moenave Formation in southwest Utah. New Mexico Museum of Natural History and Science Bulletin, 37, 510–521. M INIKH , A. V. 2001. Sharks from the Triassic of European Russia. Transactions of the Scientific Research Geological Institute of the N. G. Chernyshevskii Saratov State University, New Series, 8, 46–54. M URRY , P. A. 1981. A new species of freshwater hybodont from the Dockum Group (Triassic) of Texas. Journal of Paleontology, 55, 603–607. N ORDELL , S. E. 1994. Observations of the mating behavior and dentition of the round stingray, Urolophus halleri. Environmental Biology of Fishes, 39, 219–229. P ATTERSON , C. 1966. British Wealden sharks. Bulletin of the British Museum (Natural History), Geology, 11, 283–350. P HILIPPE , M., C UNY , G. ET AL . 2005. A Jurassic amber deposit in southern Thailand. Historical Biology, 17, 1–6. R EES , J. 2005. Neoselachian shark and ray teeth from the Valanginian, Lower Cretaceous, of Wawal, Central Poland. Palaeontology, 48, 209– 221. R EES , J. & C UNY , G. 2007. On the enigmatic neoselachian Agaleus dorsetensis from the European Early Jurassic. GFF, 129, 1– 6. R EES , J. & U NDERWOOD , C. J. 2002. The status of the shark genus Lissodus Brough 1935, and the position of nominal Lissodus species within the Hybodontoidea (selachii). Journal of Vertebrate Paleontology, 22, 471–479. R EES , J. & U NDERWOOD , C. J. 2008. Hybodont sharks of the English Bathonian and Callovian (Middle Jurassic). Palaeontology, 51, 117 –147. R EIF , W. E. 1973. Morphologie und Ultrastruktur des Hai-‘Schmelzes’. Zoologica Scripta, 2, 231– 250. R IEPPEL , O. 1981. The hybodontiform sharks from the Middle Triassic of Mte. San Giorgio, Switzerland.

113

Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Abhandlungen, 161, 324 –353. T HIES , D. 1983. Jurazeitliche Neoselachier aus Deutschland und S-England. Courier Forschungsinstitut Senckenberg, 85, 1– 117. T HIES , D. 1993. New evidence of Annea and Jurobatos, two rare neoselachians (Pisces, Chondrichthyes) from the Jurassic of Europe. In: H ERMAN , J. & VAN W AES , H. (eds) Elasmobranches et stratigraphie. Belgian Geological Survey, Professional Paper, 264, 137– 146. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2002. Middle Jurassic turtles from southern Thailand. Geological Magazine, 139, 687– 697. U NDERWOOD , C. J. 2002. Sharks, rays and a chimaeroid from the Kimmeridgian (Late Jurasssic) of Ringstead, southern England. Palaeontology, 45, 297– 325. U NDERWOOD , C. J. 2004a. Environmental controls on the distribution of neoselachian sharks and rays within the British Bathonian (Middle Jurassic). Palaeogeography, Palaeoclimatology, Palaeoecology, 203, 107– 126. U NDERWOOD , C. J. 2004b. Barremian and Aptian (Cretaceous) sharks and rays from Speeton, Yorkshire, NE England. Proceedings of the Yorkshire Geological Society, 55, 107–118. U NDERWOOD , C. J. 2006. Diversification of the Neoselachii (Chondrichthyes) during the Jurassic and Cretaceous. Paleobiology, 32, 215–235. U NDERWOOD , C. J. & R EES , J. 2002. Selachian faunas from the lowermost Cretaceous Purbeck Group of Dorset, Southern England. Special Papers in Palaeontology, 68, 83–101. U NDERWOOD , C. J. & W ARD , D. J. 2004a. Neoselachian sharks and rays from the British Bathonian (Middle Jurassic). Palaeontology, 47, 447–501. U NDERWOOD , C. J. & W ARD , D. J. 2004b. Environmental distribution of Bathonian (Middle Jurassic) neoselachians in southern England. In: A RRATIA , G. & T INTORI , A. (eds) Mesozoic Fishes 3—Systematics, Paleoenvironments and Biodiversity. Dr Friedrich Pfeil, Mu¨nchen, 111– 122. U NDERWOOD , C. J., M ITCHELL , S. F. & V ELTKAMP , K. J. 1999. Shark and ray teeth from the Hauterivian (Lower Cretaceous) of north-east England. Palaeontology, 42, 287– 302. W ELTON , B. J. & F ARISH , R. F. 1993. The Collector’s Guide to Fossil Sharks and Rays from the Cretaceous of Texas, Before Time, Lewisville, Texas. W INCHELL , C. J., M ARTIN , A. P. & M ALLATT , J. 2004. Phylogeny of elasmobranchs based on LSU and SSU ribosomal RNA genes. Molecular Phylogenetics and Evolution, 31, 214–224. W OODWARD , A. S. 1889. Catalogue of the Fossil Fishes in the British Museum (Natural History). Part 1. British Museum (Natural History), London. Y ABE , H. 1902. Notes on some shark’s teeth from the Mesozoic Formation of Japan. Journal of the Geological Society of Japan, 9, 399– 404.

Morphometric and taphonomic study of a ray-finned fish assemblage (Lepidotes buddhabutrensis, Semionotidae) from the Late Jurassic – earliest Cretaceous of NE Thailand UTHUMPORN DEESRI1,5,6*, LIONEL CAVIN2, JULIEN CLAUDE3, VARAVUDH SUTEETHORN4 & PHISIT YUANGDETKLA1 1

Phu Kum Khao Dinosaur Research Centre, The Sirindhorn Museum, Sahat Sakhan District, Kalasin 46140, Thailand 2

Department of Geology and Palaeontology, Muse´um d’Histoire naturelle, CP6434, 1211 Gene`ve 6, Switzerland

3

2, place Euge`ne Bataillon, Universite´ de Montpellier 2, ISEM, UMR 5554 CNRS, 34095, Montpellier cedex 5, France

4

Bureau of Fossils Research and Geological Museum, Rama VI Road, Bangkok 10400, Thailand 5

Palaeontological Research and Education Centre, Mahasarakham University, Maha Sarakham 44150, Thailand 6

Department of Biology, Faculty of Science, Mahasarakham University, Maha Sarakham 44150, Thailand *Corresponding author (e-mail: [email protected])

Abstract: Most Mesozoic vertebrate species are represented by scarce and incomplete specimens, preventing statistical studies of morphometric features. Moreover, rich vertebrate assemblages are rarely excavated in conditions that allow taphonomical studies. Lepidotes buddhabutrensis is a common species found in the Late Jurassic–Early Cretaceous locality of Phu Nam Jun, Phu Kradung Formation, in NE Thailand. Individuals, collected during systematic excavation since 2002, show great variations in preservation states and body postures. In this paper we study the mode of variation of morphometric features of the fish population, the growth mode, and the relationship between morphology and size. We assess the range of variation in preservation and taphonomy, based on arbitrarily defined scales, to test if vertical variations occur in the sample of individuals within the site. We test possible favoured orientation of specimens within the assemblage. In contrast to preliminary field observations, statistical analyses show that all individuals belong to a single Gaussian population and that gross morphological shape variations are related only to size during fish growth. L. buddhabutrensis shows a positive allometric growth for the pectoral to dorsal, and pectoral to anal fin distances, and a negative allometric growth for the unpaired fins (dorsal and anal fins lengths). We detected no relationships between the vertical location of the fishes within the fossiliferous deposit and the body shape of the specimens, nor between the state of preservation and the taphonomy, but there are significant differences in the state of preservation according to the position of the fishes in the fossiliferous deposit. The occurrence of a single Gaussian population and the absence of morphological and preservational variations through the depositional column are evidence that the fish assemblage is probably the result of a single mass mortality event. The apparent diversity in morphology is probably due to variations in the mode of preservation. The fish appear to have been oriented by a current at the time of deposition at the top of the fossiliferous deposit only. Supplementary material: Primary measurements are available at http://www.geolsoc.org.uk/ SUP18347.

The Mesozoic sediments from the Khorat Plateau, NE Thailand, have yielded an abundant vertebrate assemblage ranging in age from the Late Triassic to the Early Cretaceous (Buffetaut & Suteethorn

1998). The bony fish record is rather rich, but mainly represented by isolated teeth, scales and tooth plates (Martin & Ingavat 1982; Martin et al. 1984; Cavin et al. 2009). Numerous isolated

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 115–124. DOI: 10.1144/SP315.9 0305-8719/09/$15.00 # The Geological Society of London 2009.

116

U. DEESRI ET AL.

scales and some teeth have been referred to semionotid fish. In 1998, fragments of articulated and well preserved ganoid fish were discovered at a locality named Phu Nam Jun in Tambon Lao Yai, Amphoe Kuchinarai, Kalasin Province (Fig. 1). Between 2002 and 2008, systematic excavations at this site produced more than 200 fish specimens, most referable to Lepidotes buddhabutrensis Cavin et al. 2003. This freshwater semionotiform fish, with no crushing dentition, shows derived characters in its buccal apparatus, very distinct from the type species, L. elvensis (de Blainville 1818), indicating that the Thai species may belong to a different genus. Its gross morphology, reminiscent of a cyprinid fish, and its lower jaw dentition with small horizontally oriented teeth suggest a vegetarian diet, possibly by scratching food on the substrate. So far no study has investigated whether the fish assemblage represents one or several populations, although field observations led Cavin et al. (2004) to suggest that two morphotypes, a deep one and a shallow one, possibly occur together. The aim of this paper is to test the intraspecific variation of the fossil fish assemblage of L. buddhabutrensis in the locality of Phu Nam Jun, and to examine some of the taphonomical processes involved in the formation of the fossiliferous deposit. Although not all specimens have yet been prepared and although many more specimens probably remain to be uncovered at the site, we have performed this preliminary study to test field observations and to assess the scientific potential for future studies.

Geological setting The Phu Nam Jun locality is situated in the upper part of the fluvial and lacustrine Phu Kradung Formation (Buffetaut et al. 1994), which is the lowest formation of the Khorat Group as currently defined. Post-Triassic Mesozoic deposits of the Khorat Group have long been regarded as Jurassic in age. However, palynological studies (Racey et al. 1994, 1996; Carter & Bristow 2003) suggest younger ages for most of the formations included in that group, although the Phu Kradung Formation has yielded inconclusive palynological evidence so far. According to age constraints provided by the overlying formations, the Phu Kradung Formation may be either Late Jurassic or earliest Cretaceous in age. Changes in lithology occur in the fossiliferous deposit of the Phu Nam Jun site, but we cannot follow them laterally. They may represent lateral variations within a single large depositional event or discrete depositional episodes. So far, sedimentological field observations do not allow us to make a decision on that issue. The single visible clear lithological change is a greenish sandy deposit at the bottom of the excavated area but its lateral extension and 3D structure are unknown. The upper part of the fossiliferous deposit consists of a maroon mudstone with sandstones lens, mica and concretions.

Materials and methods All specimens are housed in the Sirindhorn Museum, Sahat Sakhan District, Kalasin Province, Thailand. The specimens were prepared from their upper side with air-pens and scalpels in the Sahat Sakhan Dinosaur Research Centre. Sixty-one individuals were used in morphometric and meristic analyses and 92 individuals in studies dealing with preservation modes and body postures. The remaining specimens are yet to be prepared.

Measurements

Fig. 1. Map of NE Thailand showing the Phu Kradung outcrops in black and the Phu Nam Jun locality (open circle).

The position and orientation of each individual L. buddhabutrensis were recorded on a map prior to being removed in plaster jackets. Measurements and meristic features of the body were recorded following McCune’s method (McCune 1987; Fig. 2). The term ‘preservation mode’ corresponds here to the state of decay before fossilization while ‘taphonomic state’ corresponds here to the body posture and not the state of decay. These two parameters have been divided into arbitrarily defined qualitative scales, comprising four classes of preservation and four classes of taphonomic

MORPHOMETRIC STUDY OF THAI LEPIDOTES

117

Fig. 2. Morphometric methods following McCune (1987). (a) Numbers correspond to the following measurements: 1, HDL, head length; 2, HDD, head depth; 3, BL, body length; 4, DPTH, body depth; 5, SL, standard length; 6, PDL, pre-dorsal length; 7, PAL, pre-anal length; 8, DFPT, dorsal–pectoral length; 9, PTAL, pectoral– anal length; 10, PTPV, pectoral– pelvic length; 11, DFAN, dorsal–anal length; 12, DFPV, dorsal–pelvic length; 13, DFCD, dorsal– caudal length; 14, AFCD, anal–caudal length; 15, TDL, total dorsal fin length; 16, TAL, total anal fin length; 17, MAXCD, maximum caudal peduncle depth; 18, MINCD, minimum caudal peduncle depth; (b), Scale counts abbreviations.

states, defined as follow: Preservation (Fig. 3) State 1: very well preserved: the body of the fish is complete and articulated; the fins are preserved (although rarely completely). State 2: well preserved: the body of the fish is articulated and almost complete, sometimes a few scales are missing; the fins are incompletely

preserved but their position on the body can be accurately identified. State 3: not well preserved: the fish is partly articulated but incomplete; fragments of fins may be preserved but their position on the body cannot be identified with certainty. State 4: badly preserved: only fragments of the body are preserved; the outline of the fish is not always recognizable.

Fig. 3. Line drawings of examples of specimens of L. buddhabutrensis exemplifying preservation classes. 1, very well preserved; 2, well preserved; 3, not well preserved; 4, badly preserved. Numbers refer to field numbers. Specimens are not to the same scale.

118

U. DEESRI ET AL.

Fig. 4. Line drawings of examples of specimens of L. buddhabutrensis exemplifying taphonomic classes. 1, taphonomic state 1: fish lies flat on one side; 2, taphonomic state 2: fish lies flat on one side with folds in the squamation (arrows point to folds); 3, taphonomic state 3: fish is partly or totally dorsoventrally preserved; 4, taphonomic state 4: fish is bent or twisted. Numbers refer to field numbers. Specimens are not to the same scale.

Taphonomy (Fig. 4) State 1: The fish lies straight on one side. State 2: The fish lies straight on one side; the squamation shows folds in the abdominal area (probably caused by expulsion of gases during decay). State 3: The fish is partly or totally dorsoventrally preserved. State 4: The fish is bent or twisted. To assess the vertical variation in preservation and taphonomy, the fossiliferous deposit was divided in three sub-layers. The upper limit of the lowest layer is defined by a lithological change, from a greenish fine and loose sandstone below to a maroon mudstone above. This change may be related to a time gap in a progressive sediment deposition, or to a heterogeneity in a single rapid depositional event. As there is no visible change in the lithology in the upper part of the fossiliferous deposit, the limit between the uppermost two layers is arbitrarily defined as situated 60 cm above the lower limit. The upper layer is also about 60 cm thick.

Tests To test whether there is more than one statistical population, the normality Shapiro –Wilk test was

used. The relationship between morphometric and meristic parameters were investigated, and an allometric analysis using a linear regression of an estimation of general body size with all variables corrected for size was performed. Body size of the fish was estimated as the square root of the sum of the square of all variables. As all observations were needed to yield this body size estimator, missing values were estimated following a regression procedure: parameters of all possible regressions were computed from the available measurements. For each individual, missing values were estimated as the mean of all possible predictions from the different regressions. These estimations were performed to obtain a body size estimator only, but were not used for the other tests. The relationships between morphology (measurements and meristic) and position within the fossiliferous deposit, and between morphology and taphonomy (body posture) were analysed using ANOVA tests; the relationships between the location of individuals within the fossiliferous deposit and meristic features, taphonomy, states of preservation, and between preservation and taphonomy were tested using x2 tests. Tests and statistics were computed with the R language and environment (Ihaka & Gentleman 1996).

Table 1. Mean primary measurements Variable

HDL

HDD

BL

DPTH

SL

PDL

PAL

DFPT

PTAL

Average Sample size

12.55 55

6.8 55

29.67 60

15.4 53

40.1 43

25 46

32.6 37

18.7 50

23.87 47

Variable

PTPV

DFAN

DFPV

DFCD

AFCD

TDL

TAL

MINCD

MAXCD

Average Sample size

11.85 45

15.98 48

13.93 47

18.17 58

10.44 56

5 61

4.36 55

6.56 57

5.65 58

See Figure 2 for abbreviations.

18 42

32

Primary measurements are available as supplementary material. Mean measurements are shown in Table 1, meristic values and sample sizes for each taphonomic class are shown in Table 2, and sample sizes for preservation and taphonomy classes in each layer are shown in Table 3.

Normality of characters

23 46

Morphometric measurements All but three morphometric variables are significantly and positively correlated (Table 4), indicating that these parameters are isometrically directly related to the general body size of the fish. The three exceptions are dorsal and anal fin lengths, which are significantly negatively correlated with the distance between pectoral and pelvic fin, and the anal fin length, which shows a significant negative correlation with the distance between dorsal and pelvic fin. There is no relationship between morphometric and meristic parameters, indicating that the number of scales is not related to size and does not vary during growth, as in most actinopterygians.

14 48

11 42

23

28

20

21

51

23

Thirteen variables are normally distributed, except five: head depth (HDD, W ¼ 0.9128, n ¼ 55, p ¼ 0.0005); dorsal–pectoral length (DFPT, W ¼ 0.945, n ¼ 50, p ¼ 0.0195); pectoral– anal length (PTAL, W ¼ 0.9512, n ¼ 47, p ¼ 0.0483); dorsal –anal length (DFAN, W ¼ 0.9038, n ¼ 48, p ¼ 0.0008); maximum caudal peduncle depth (MAXCD, W ¼ 0.9526, n ¼ 58, p ¼ 0.025). After inspection of distribution histograms, the significance of Shapiro tests is better interpreted as the presence of a few outliers in the distribution rather than the existence of several modes. Further inspection showed that outliers are the result of taphonomical processes rather than of biological abnormality. As most variables are normally distributed, and that there is no evidence of bi- or multinormality, the preserved assemblage is interpreted as belonging to a single Gaussian population.

See Figure 2 for abbreviations.

25 44 16 39 26 37 46 49 Average Sample size

119

Results

5

13

2 1 4 3 2 3 2

DRS PLVSC VDSC DFSC VVSC ANFSC CDSC Variable

Table 2. Primary meristic values and sample sizes for each taphonomic class

1

Preservation

4

1

Taphonomy

Level

3

MORPHOMETRIC STUDY OF THAI LEPIDOTES

Allometric growth Significant positive allometries have been detected for the distance between dorsal and pectoral fins (DFPT; a ¼ 0.0007, F-test ¼ 4.19, df ¼ 1,49, p ¼ 0.04) and for the distance between pectoral and anal fins (PTAL; a ¼ 0.001, F-test ¼ 5.58, df ¼ 1,45, p ¼ 0.02). In contrast, dorsal fin length and anal fin length (TDL; a ¼ 20.0004, F-test ¼ 5.77, df ¼ 1,61, p ¼ 0.01, TAL;

120

U. DEESRI ET AL.

Table 3. Sample sizes for preservation and taphonomy classes in each layer Preservation state

Fish at level 1 Fish at level 2 Fish at level 3

1

2

3

4

15 (35.71%) 5 (15.62%) 3 (16.66%)

11 (26.19%) 9 (28.12%) 8 (44.44%)

9 (21.42%) 5 (15.62%) 6 (33.33%)

7 (16.66%) 13 (40.62%) 1 (5.55%)

Total 42 32 18

Taphonomic state

Fish at level 1 Fish at level 2 Fish at level 3

1

2

24 (57.14%) 18 (56.25%) 9 (50%)

13 (30.95%) 5 (15.62%) 5 (27.77%)

3 2 (4.76%) 2 (6.25%) 1 (5.55%)

4 3 (7.14%) 7 (21.87%) 3 (16.66%)

Total 42 32 18

a ¼ 20.0003, F-test ¼ 5.29, df ¼ 1,53, p ¼ 0.02, respectively) show negative allometries (Fig. 5).

Mode of preservation and location within the fossiliferous deposit

Morphology and position within the fossiliferous deposit

There is no significant relationship between the position of the fish in the fossiliferous deposit and the taphonomy (x2 ¼ 4.56, df ¼ 6, critical value ¼ 12.59, p ¼ 0.05), but there is a significant difference between the state of preservation and the position of the fish in the fossiliferous deposit (x2 ¼ 14.58, df ¼ 6, critical value ¼ 14.44, p ¼ 0.025).

ANOVA test shows that there is no difference in the general morphology of the fish between the three layers, suggesting that the morphology of the assemblage does not vary within the fossiliferous deposit. The x2 test did not detect any relationship between the meristic countings and the position within one of the three fish layers.

Taphonomy, preservation and morphology Taphonomy was found as a source of variation for four morphological measurements: body depth (DPTH, ANOVA: F ¼ 9.05, df ¼ 2,51, p ¼ 0.0004), dorsal–pectoral length (DFPT, ANOVA: F ¼ 3.31, df ¼ 2,48, p ¼ 0.044), dorsal–pelvic length (DFPV, ANOVA: F ¼ 6.28, df ¼ 2,44, p ¼ 0.003) and anal –caudal length (AFCD, ANOVA: F ¼ 3.92, df ¼ 3,52, p ¼ 0.01). Significant differences in preservation were observed for the anal –caudal length (AFCD, ANOVA: F ¼ 3.43, df ¼ 3,52, p ¼ 0.023) and for the dorsal fin length (TDL, ANOVA: F ¼ 2.82, df ¼ 3,59, p ¼ 0.046).

Mode of preservation and taphonomy There is no significant relationship between the preservation and the taphonomy as defined here (body posture) (total x2 13.81, df ¼ 9, critical value ¼ 16.91, p ¼ 0.05). This may indicate that the postures of the carcasses are not related to the type of preservation (i.e. a specimen lying on its flank is, on average, not better or worst preserved than a specimen lying on its back).

Fish orientation The rose diagrams show one main direction, NNW– SSE (Fig. 6), as well as a secondary orientation, which is perpendicular to the main one (ENE – WSW), for the uppermost layer (layer 1). Hydrodynamic processes oriented the fish carcasses parallel to the current, probably head against the current, or perpendicular to the current. Consequently, the main current was possibly from SSW toward NNW. Other layers show no clear main direction.

Discussion The average standard length of L. buddhabutrensis is 401 mm, the average head length is 125.5 mm and the average body depth is 154 mm. These values represent a rather large species, compared with other species of Lepidotes (L. microrhis: 92, 30 and 46 mm, respectively, for the holotype, Wenz 2003; L. xinjinensis: length of 320 mm, Su 1983; L. wenzae, standard length (SL) estimated at 330 mm, Brito & Gallo 2003) or Pliodetes nigeriensis (SL ¼ 220 mm, Wenz 1999) and with most species of Semionotus (McCune 1986), but it is rather similar in size to Lepidotes piauhyensis (SL ranging from 320 to 480 mm, Gallo 2005) and to Araripelepidotes (SL ranging from 170 to

MORPHOMETRIC STUDY OF THAI LEPIDOTES

121

Table 4. Coefficients of correlation between continuous variables Variables

HDL

HDD

BL

DPTH

SL

PDL

PAL

DFPT

PTAL

HDL HDD BL DPTH SL PDL PAL DFPT PTAL PTPV DFAN DFPV DFCD AFCD TDL TAL MINCD MAXCD

1.000 0.615 0.606 0.538 0.643 0.635 0.768 0.630 0.577 0.285 0.627 0.644 0.591 0.299 0.245 0.288 0.604 0.606

0.615* 1.000 0.189 0.544 0.299 0.250 0.499 0.333 0.175 0.164 0.408 0.361 0.337 0.255 20.012 0.323 0.366 0.460

0.606* 0.189‡ 1.000 0.324 0.872 0.745 0.883 0.613 0.814 0.673 0.402 0.444 0.655 0.342 0.324 0.136 0.343 0.321

0.538* 0.544* 0.324* 1.000 0.283 0.349 0.419 0.609 0.327 0.287 0.599 0.968 0.225 0.290 0.044 0.110 0.698 0.587

0.643* 0.299† 0.872* 0.283† 1.000 0.755 0.783 0.649 0.645 0.586 0.407 0.477 0.589 0.284 0.347 0.212 0.481 0.471

0.635* 0.250† 0.745* 0.349* 0.755* 1.000 0.658 0.805 0.527 0.413 0.190 0.513 0.360 0.564 0.216 0.181 0.561 0.548

0.768* 0.499* 0.883* 0.419* 0.783* 0.658* 1.000 0.580 0.915 0.681 0.442 0.463 0.551 0.227 0.313 0.182 0.447 0.442

0.630* 0.333* 0.613* 0.609* 0.649* 0.805* 0.580* 1.000 0.636 0.468 0.448 0.715 0.378 0.565 0.023 0.180 0.706 0.583

0.577* 0.175‡ 0.814* 0.327* 0.645* 0.527* 0.915* 0.636* 1.000 0.671 0.457 0.364 0.588 0.192 0.107 0.179 0.304 0.312

Variables

PTPV

DFAN

DFPV

DFCD

AFCD

TDL

TAL

MINCD

MAXCD

0.627* 0.644* 0.408* 0.361* 0.402* 0.444* 0.599* 0.968* 0.407* 0.477* ‡ 0.190 0.513* 0.442* 0.463* 0.448* 0.715* 0.457* 0.364* 0.392* 0.282† 1.000 0.479* 0.479 1.000 0.708 0.248 0.231 0.214 0.090 0.130 0.136 20.001 0.628 0.584 0.556 0.418

0.591* 0.337* 0.655* 0.225† 0.589* 0.360* 0.551* 0.378* 0.588* 0.487* 0.708* 0.248† 1.000 0.292 0.059 0.165 0.271 0.346

0.299† 0.245† 0.288† 0.255† 20.012 0.323* 0.342* 0.324* 0.136‡ † § 0.290 0.044 0.110‡ 0.248† 0.347* 0.212‡ 0.564* 0.216‡ 0.181‡ 0.313† 0.182‡ 0.227† 0.565* 0.023* 0.180‡ 0.192‡ 0.107‡ 0.179‡ 0.130‡ 20.130‡ 20.146‡ 0.231† 0.090‡ 0.136‡ ‡ ‡ 0.214 0.130 20.001 0.292† 0.059§ 0.165‡ 1.000 0.305† 0.640* 0.305 1.000 0.357* 0.640 0.357 1.000 0.455 0.183 0.325 0.328 0.111 0.327

0.604* 0.366* 0.343* 0.698* 0.481* 0.561* 0.447* 0.706† 0.304† 0.240† 0.628* 0.584* 0.271† 0.455* 0.183‡ 0.325* 1.000 0.897

0.606* 0.460* 0.321* 0.587* 0.471* 0.548* 0.442* 0.583† 0.312† 0.241† 0.556* 0.418* 0.346* 0.328* 0.111‡ 0.327* 0.897* 1.000

HDL 0.285* HDD 0.164‡ BL 0.673* DPTH 0.287† SL 0.586* PDL 0.413* PAL 0.681* DFPT 0.468* PTAL 0.671* PTPV 1.000 DFAN 0.392 DFPV 0.282 DFCD 0.487 AFCD 0.130 TDL 20.130 TAL 20.146 MINCD 0.240 MAXCD 0.241

*Very highly significant (0 , p , 0.001). † Highly significant (0.001 , p , 0.01). ‡ Significant (0.01 , p , 0.05). § Nearly significant (0.05 , p , 0.1). See Figure 2 for abbreviations.

450 mm, Maisey 1991), and much smaller than some species of Lepidotes such as L. maximus, which could reach 2 m in total length (Jain 1985). The negative correlation between the dorsal and anal fin sizes with the distance between pectoral and pelvic fin is partly corroborated by allometric growth of these fins, indicating that unpaired fins grow less rapidly than body size. This result shows either that the size proportions of the unpaired fins vary during growth, or that fin size, as measured on the carcasses, was influenced by external factors, such as decay. If the latter solution is

correct, it indicates that the fins of large fish were affected proportionally more than fins of the small fish, because they were proportionally more exposed to decay. The other allometric pattern observed is the proportionally longer distance between pectoral and dorsal fins and between pectoral and anal fins. This gives the larger specimens a larger belly, and may be the reason for the field observation that two morphotypes seemed to be present in the fish assemblage. Our morphometric examination shows, however, that shape variations correspond to a continuum rather than to two

122

U. DEESRI ET AL.

Fig. 5. Growth is isometric, except for the four allometric measurements indicated by arrows. þ, positive allometry; 2, negative allometry.

Fig. 6. Rose diagrams representing orientation of fish specimens in each layer.

distinct morphotypes. The preserved assemblages belong to a single Gaussian population, although five parameters (HDD, DFPT, PTAL, DFAN, MAXCD) are not normally distributed. We suggest that the non-normal distributions are caused by inclusion of outliers that underwent strong taphonomic deformations. The morphology and taphonomy of the fish carcasses do not vary within the fossiliferous deposit, in contrast to the state of preservation, which varies between the fossiliferous layers. This observation might support the hypothesis that fish carcasses dried in the open air before burial by showing that fish from the lowermost layer were less affected by taphonomic decay. This hypothesis has been previously suggested on the basis of scale disarticulation, which is more reminiscent of desiccation of the body after death than rapid embedding in sediment (Cavin et al. 2004). Supporting this hypothesis is the fact that very few specimens from the lowermost layer are poorly preserved (5.6% of state 4), whereas this percentage is higher in layers 1 and 2 (16.7% and 40.6%, respectively), although the proportion

of very well preserved specimens does not show the same pattern (35.7%, 15.6% and 16.7% from top, middle and bottom layers, respectively). This contradictory pattern shows either that the hypothesis is wrong (specimens from the bottom of the fossiliferous deposit are not better preserved that specimens from the top, which were supposedly more exposed to atmospheric factors causing decay before fossilization) or that the method we used is biased. The latter interpretation is possible as two strong biases are present: (1) the layers were arbitrarily defined and do not correspond to distinct events in terms of sedimentology; (2) mechanical preparation of specimens was not fully random because more specimens from the upper layer have already been prepared than specimens from the lower layers because they were the first ones to be removed from the site (42, 32 and 18 individuals, respectively) and because a choice was made to prepare the more informative specimens first. We suggest that similar analyses should be made in the future when the sample will be larger (ideally when all fishes have been prepared)

MORPHOMETRIC STUDY OF THAI LEPIDOTES

to remove the sampling bias. Another planned study (U.D.) is to assess the variability of the microstructural ornamentation of the ganoin layer at the surface of the scales, to check if the ornamentation is the same in different parts of the body of a single fish and between individuals of different sizes and states of preservation. The fact that all specimens belong to a single Gaussian population will make easier this next step of the study of the L. buddhabutrensis assemblage of Phu Nam Jun. The absence of relationships between preservation and taphonomy indicates that the mode of decay of the fish is not dependent on the body postures. A weak current might explain the orientation of fish in the layers. However, the current is clearly present in the uppermost layer only (Fig. 6). It is possible that the fish have been reoriented after death subsequent to a flooding event, or to a weak current toward the centre of the drying pond where the last water remained. Another explanation is an abrupt incoming of mud that would explain the sandy layer at the base of the deposit as well as the preferred orientation of the carcasses at the top of the assemblage. Such a big mud flow, embedding several hundreds of fish individuals at the same time, might indicate strongly contrasted dry and rainy seasons. We cannot exclude, however, that the possible preferable orientation is an artefact related to the small sample sizes.

Conclusion Statistical studies of morphometric and meristic features show that the prepared specimens of Lepidotes buddhabutrensis from Phu Nam Jun belong to a single Gaussian population rather than two as suggested by field observations of Cavin et al. (2004). The morphometric variables indicated that the general body size growth of fish is isometric, except for negative allometries detected on unpaired fin lengths and positive allometries detected for the distance from pectoral to dorsal and anal fins. Differences in the state of preservation show that variations in morphology are probably due to the mode of preservation. However, there is no difference between the morphology and the taphonomy of the fish carcasses within the fossiliferous deposit. A current is clearly present in the uppermost layer only. The authors would like to thank all members of the staff of Phu Kum Khao Dinosaur Research Centre for their support in field and laboratory work. This is publication ISEM 2008-007 of J.C. The work was supported by the TRF-CNRS and the grant programme under the Commission on Higher Education in Palaeontology to U.D., and the PHC bilateral programme supported by the Ministry of Foreign and European Affairs (MAEE, France) and

123

Ministry of Education ‘Evolution of aquatic vertebrates from the Mesozoic and Cenozoic from Thailand’. L.C.’s research was partly supported by the Swiss National Science foundation (grant 200021-113980). We thank R. Mutter (London) and P. Olsen (New York) for reviewing the manuscript.

References B RITO , P. M. & G ALLO , V. 2003. A new species of Lepidotes (Neopterygii: Semionotiformes: Semionotidae) from the Santana Formation, Lower Cretaceous of northeastern Brazil. Journal of Vertebrate Paleontology, 23, 47–53. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of Southeast Asia. Backhuys, Leiden, 83–90. B UFFETAUT , E., T ONG , H. & S UTEETHORN , V. 1994. First post-Triassic labyrinthodont amphibian in South East Asia: A temnospondyl intercentrum from the Jurassic of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 7, 385 –390. C ARTER , A. & B RISTOW , C. S. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau, Eastern Thailand. Basin Research, 15, 271 –285. C AVIN , L., S UTEETHORN , V., K HANSUBHA , S., B UFFETAUT , E. & T ONG , H. 2003. A new Semionotid (Actinopterygii, Neopterygii) from the Late Jurassic–Early Cretaceous of Thailand. Comptes Rendus Pale´vol, 2, 291– 297. C AVIN , L., S UTEETHORN , V. ET AL . 2004. A new fish locality from the continental Late Jurassic–Early Cretaceous of Northeastern Thailand. Revue de Pale´obiologie, 9, 161–167. C AVIN , L., D EESRI , U. & S UTEETHORN , V. 2009. The Jurassic and Cretaceous bony fish record (Actinopterygii, Dipnoi) from Thailand. In: B UFFETANT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 123 –137. D E B LAINVILLE , H. D. 1818. Sur les ichthyolites ou les poissons fossils. In: L EVRAULT , F. G. (ed.) Nouveau dictionnaire d’histoire naturelle, applique´ aux arts, a` l’e´conomie rurale et domestique, a` la Me´decine, etc. Deterville, Paris, 310– 395. G ALLO , V. 2005. Redescription of Lepidotes piauhyensis Roxo and Lo¨fgren, 1936 (Neopterygii, Semionotiformes, Semionotidae) from the Late Jurassic– Early Cretaceous of Brazil. Journal of Vertebrate Paleontology, 25, 757–769. I HAKA , R. & G ENTLEMAN , R. 1996. A language for data analysis and graphics. Journal of Computational and Graphical Statistics, 5, 299–314. J AIN , S. L. 1985. Some new observations on Lepidotes maximus (Holostei: Semionotiformes) from the German Upper Jurassic. Journal of the Palaeontological Society of India, 30, 18– 25.

124

U. DEESRI ET AL.

M AISEY , J. G. 1991. Araripelipidotes Silva Santos, 1985. In: M AISEY , J. G. (ed.) Santana fossils. An illustrated Atlas. TFH Publications, Neptune City, NJ, 118– 123. M ARTIN , M. & I NGAVAT , R. 1982. First record of an Upper Triassic Ceratodontid (Dipnoi, Ceratodontiformes) in Thailand and its palaeogeographical significance. Geobios, 15, 111– 114. M ARTIN , M., B UFFETAUT , E. & I NGAVAT , R. 1984. Fossil vertebrates and the Late Triassic age of the Lom Sak Formation of Central Thailand. Journal of the Geological Society of Thailand, 7, 19–24. M C C UNE , A. R. 1986. A revision of Semionotus (Pisces, Semionotidae) from the Triassic and Jurassic of Europe. Palaeontology, 29, 213–233. M C C UNE , A. R. 1987. Toward the phylogeny of a fossil species flock: Semionotid fishes from a lake deposit in the Early Jurassic Towaco Formation, Newark Basin. Peabody Museum of Natural History Yale University Bulletin, 43, 1– 108. R ACEY , A., G OODALL , J. G. S., L OVE , M. A., P OLCHAN , S. & J ONES , P. D. 1994. New age data for the Mesozoic Khorat Group of Northeast Thailand.

In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245–252. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLCHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. Part 1: Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5– 40. S U , D.-Z. 1983. Note on a new Lepidotes from the Cretaceous of Sichuan. Vertebrata PalAsiatica, 21, 177– 187. W ENZ , S. 1999. Pliodetes nigeriensis, gen. nov. et sp. nov., a new semionotid fish from the Lower Cretaceous of Gadoufaoua (Niger Republic): Phylogenetic comments. In: A RRATIA , G. & S CHULTZE , H.-P. (eds) Mesozoic Fishes 2—Systematics and Fossil Record. Dr Friedrich Pfeil, Mu¨nchen, 107– 120. W ENZ , S. 2003. Les Lepidotes (Actinopterygii, Semionotiformes) du Cre´tace´ infe´rieur (Barre´mien) de Las Hoyas (Province de Cuenca, Espagne). Geodiversitas, 25, 481– 499.

The Jurassic and Cretaceous bony fish record (Actinopterygii, Dipnoi) from Thailand LIONEL CAVIN1*, UTHUMPORN DEESRI2 & VARAVUDH SUTEETHORN3 1

Department of Geology and Palaeontology, Muse´um d’Histoire naturelle, CP 6434, 1211 Gene`ve 6, Switzerland

2

Phu Kum Khao Dinosaur Research Centre, The Sirindhorn Museum, Sahat Sakhan District, Kalasin 46140, Thailand

3

Bureau of Palaeontological Research and Museum, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected]) Abstract: This first overview of the bony fish record from the Jurassic and Cretaceous continental deposits of Thailand reveals a significant diversity, with 16 taxa in four formations (the Khlong Min, Phu Kradung, Sao Khua and Khok Kruat Fms). Four of these taxa have already been diagnosed and described, and a couple of others are sufficiently well preserved to be diagnosed in the future. The other taxa are represented at present by fragmentary and isolated remains. The highest diversity is observed among ‘semionotids’, which occur in the four formations. Sinamiids are represented by at least three taxa that occur only in the Sao Khua and the Khok Kruat Formations. Pycnodont fishes are known by rare and isolated dentitions and teeth in the Khlong Min and Sao Khua Formations, and lungfishes referred to Ferganoceratodus occur in the Khlong Min and the Phu Kradung Formations. The assemblages provide few palaeogeographical indications at present, except for evidence of relationships with China and Central Asia. However, it is expected than once the phylogenetic relationships of these taxa are resolved, we will be able to reconstruct precise palaeogeographical scenarios.

Bony fish remains represent an important part of the fossils discovered during palaeontological field work conducted in the continental Jurassic and Cretaceous deposits of Thailand for more than 25 years. They have been mentioned in several publications as a component of the vertebrate assemblages, but have rarely been properly described, except for a handful of taxa known from more complete material. No overview of the succession of the fish assemblages throughout the Mesozoic deposits of Thailand has yet been published. Here we describe the fossil record of bony fishes from the Middle– Late Jurassic of the Khlong Min Formation in the peninsula of Thailand to the Early Cretaceous Khok Kruat Formation in the Khorat Plateau. Triassic fishes are known in Thailand, but they are not reviewed here because they belong to assemblages separated from the Jurassic assemblages by a large stratigraphic gap. All specimens described here, unless specified otherwise, are housed in the Sirindhorn Museum, Sahat Sakhan, Kalasin Province.

Geological setting The bony fish remains described below were found in several localities of the Khorat Plateau in NE

Thailand and in one locality in southern Thailand, Mab Ching. The main tectonic units in Thailand comprise two continental blocks or microcontinents (Fig. 1): the eastern part (including the Khorat Plateau) belongs to the Indochina block, and the western part (including the southern peninsula) is part of a block called ‘Shan-Thai’ or ‘Sibumasu’. The stratigraphy of the non-marine sediments in southern peninsular Thailand has been recently reviewed (Teerarungsigul et al. 1999). These clastic red beds are known as the Trang Group, which is now subdivided into a basal Khlong Min Formation, overlain by the Lam Thap, Sam Chom and Phun Phin Formations (Meesook et al. 2002). The Mab Ching locality in the Khlong Min Formation has yielded hybodont sharks (Cuny et al. 2009), lungfishes (Martin et al. 1997), temnospondyls (Buffetaut et al. 1994), mesosuchian crocodiles and the turtle Siamochelys (Tong et al. 2002). This formation was dated to the Middle or Late Jurassic on the basis of charophytes and palynomorphs (Lei 1993; Buffetaut et al. 1994), an age confirmed by the turtles (Tong et al. 2002; see also Cuny et al. 2009). The occurrence of a euhelopodid sauropod in the Khlong Min Formation indicates a continental connection with mainland Asia (Buffetaut et al. 2006).

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 125–139. DOI: 10.1144/SP315.10 0305-8719/09/$15.00 # The Geological Society of London 2009.

126

L. CAVIN ET AL.

Fig. 1. Map of SE Asia, showing location of the geological formations yielding fossils described in the text.

The Khorat Plateau in NE Thailand is composed of non-marine sediments deposited during the Mesozoic. A recent study (Carter & Bristow 2003) restricted the Khorat Group to five formations: the Phu Kradung, Phra Wihan, Sao Khua, Phu Phan

and Khok Kruat Formations, in ascending order. The main lithologies of the rocks are reddish brown to light grey sandstones, conglomeratic sandstones, siltstones, claystones and conglomerates. The rocks are interpreted as having been deposited

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND

127

Institutional abbreviations MHNG: collection of the Natural History Museum of Geneva, Switzerland. TF: collection of the Sirindhorn Museum, Sahat Sakhan, Kalasin Province. SHM: collection of the Srisuk’s House Museum, Khao Yoi, Petchaburi Province.

Systematic palaeontology Osteichthyes Huxley Actinopterygii Cope Indeterminate

Phu Kradung Formation An isolated spine from the locality of Chong Chat in the Phu Kradung Formation shows a spherical articular proximal head and parallel ridges running along its base (TF 8022, Fig. 2). Three main edges make the spine triangular in section, with one border of the triangle concave in shape. There is a trace of enamel on the surface and no axis of symmetry, indicating that it may correspond to a spine bordering a paired fin. The spherical articular head differs from the spines of chondrichthyes and the specimen is provisionally referred to an indeterminate actinopterygian. ‘Palaeonisciformes’ Hay cf. Ptycholepis Agassiz Fig. 2. Isolated spine of an indeterminate actinopterygian (TF 8022), Chong Chat, Phu Kradung Formation. Scale bar: 15 mm.

by meandering and braided rivers in semi-arid conditions (Meesook 2000). In the Khorat Plateau, three of the five formations have yielded vertebrate body fossils. The Phu Kradung Formation provided no conclusive fossils for dating, but as it is conformably overlain by the Early Cretaceous Phra Wihan Formation, which contains palynomorphs at its base, the Phu Kradung Formation is regarded as Late Jurassic or Early Cretaceous in age (Racey et al. 1996). Dinosaur remains from the Phu Kradung Formation favour a Late Jurassic age (Buffetaut & Suteethorn 2007). The topmost formation of the Khorat Group, the Khok Kruat Formation, is Aptian in age based on palynomorphs (Racey et al. 1996) and vertebrates (Cappetta et al. 1990). The Sao Khua Formation overlies the Phra Wihan Formation and is separated from the Khok Kruat Formation by the Phu Phan Formation. Its age is thus constrained to the Early Cretaceous (probably Hauterivian to Barremian).

Phu Kradung Formation Pieces of an articulated fish from the locality of Khok Sanam were discovered in July 1997. The specimen comprises several slabs of fine sandstone, with the most complete part consisting of a fragment of a squamation (TF 8023, Fig. 3a) plus five pieces with a few articulated scales. Most of the scales have gone, leaving only dark imprints in the sediment. A few scales, however, are still preserved and show a typical ganoin covering. Thirteen rows of scales are preserved. Scales of the first 11 rows are curved, slightly sigmoidal, and elongated in shape. The short posterior border of each scale is fringed as the result of two or three grooves in the ganoin covering (visible as ridges on the imprints) extending along the posterior half of the scales. The ventralmost two pieces of scales of the fifth row display a peculiar ornamentation with curved and reticulated grooves. They are regarded here as fragments of a cloacal scale. The twelfth row is composed of very shallow and elongated scales, with a groove running along their whole length, and a tapering and slightly curved posterior extremity. This row is separated from the more anterior scales by a

128 L. CAVIN ET AL. Fig. 3. Cf. Ptycholepis, Khok Sanam, Phu Kradung Formation. (a) Portion of an articulated specimen (TF 8023). Scale bar: 30 mm. (b –d) Portion of an articulated specimen and isolated elements from a single specimen (TF 8024). (b) Collection of isolated scales. Scale bar: 15 mm. (c) Disarticulated scales on a slab. Scale bar: 30 mm. (d) Fragment of the skull. Scale bar: 30 mm.

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND

gap, but this is probably due to a taphonomic process. The scales of the last row have long axes at an angle of about 358 to the axes of the other scales. Although poorly preserved, the general shape appears fan-like, with a tapering proximal end, and a surface ornamented with smooth grooves. Because of their orientation and morphology, they are regarded as lepidotrichs. Another semi-articulated specimen discovered in February 2004 may be referred to the same taxon (TF 8024, Fig. 3b –d). The morphology of the scales is variable (Fig. 3b). Elongated scales with longitudinal ridges ending posteriorly as spines are preserved on pieces of sandstone. Fragments of bones ornamented with tubercles are present, but most of them are not identifiable. Isolated scales show that the basal bony plates are proportionally thick, generally with a budge in a non-central position. A fragment of the skull visible in internal view is present, but the bones are too fractured to allow identification (Fig. 3d). Scales with similar morphology have been discovered in the locality of Wang Din So, near Phitsanulok, in the Phu Kradung Formation (P. Srisuk, pers. comm.). Thick elongated ganoid scales with longitudinal ridges or grooves are not common among Holostei, but they occur in more basal actinopterygians. Ptycholepis, in particular, has scales with such ornamentation (Bu¨rgin 1992). A specimen of P. bollensis from the Early Jurassic of Holzmaden, housed in the Geneva Natural History Museum (MHNG V. 1600), shows that scales located in the posterior part of the body, near the dorsal and ventral borders of the body, are very shallow and elongated, and slightly sigmoid in shape. The cloacal scale, moreover, is proportionally larger and shows patterns of ganoid ridges reminiscent of some of the larger scales in the Thai specimen, although the pattern is symmetrical and more regular in the German species. If the specimens from Khok Sanam are correctly interpreted, an important difference from Ptycholepis bollensis and the Triassic species of Ptycholepis, as far as this part of the body is known (Bu¨rgin 1992), is the longer anal fin in the Thai species. Although other basal actinopterygians display ganoid scales with strong ornamentation, Colobodus (Mutter 2004) or Ameghinichthys antarcticus from the Tithonian of the Antarctic Peninsula (Arratia et al. 2004) for instance, they are never as shallow as in Ptycholepis. Consequently, we refer that material provisionally to cf. Ptycholepis. Ptycholepis is a genus known from the middle Triassic to the late Early Jurassic (Bu¨rgin 1992). If it belongs to that genus, the Thai taxon is probably the youngest one known so far. Other ‘palaeonisciforms’ are known in the Late Jurassic and in the Early Cretaceous, such as several species of Coccolepis and some

129

related genera (Hilton et al. 2004), but these are very different from the specimens described here. The occurrence of cf. Ptycholepis in the Phu Kradung Formation supports an ‘old’ age for that formation, Late Jurassic rather than Early Cretaceous, in accordance with dinosaur remains (Buffetaut & Suteethorn 2007). However, we should keep in mind that (1) the assignation of that material to cf. Ptycholepis is tentative, pending the discovery of more complete material, and (2) if the identification is confirmed, the occurrence of Ptycholepis in the Early Cretaceous is not impossible in freshwater environments, where basal fish taxa appear to persist longer. A ‘palaeonisciform’ has been reported from the Late Jurassic Long Binh Formation in Vietnam (Filleul & Vu Khuc 2001), but its state of preservation does not allow comparisons with material from Khok Sanam. Holostei sensu Grande 2005 Semionotiformes sensu Olsen & McCune 1991 ‘Semionotidae’ incertae sedis Pending a better diagnosis of the taxa described here, we refer here to the poorly defined and possibly not monophyletic family ‘Semionotidae’ (or ‘semionotids’) all the fish remains that share with Semionotus, Lepidotes and relatives similar gross morphology of scales and, for some of them, dentition. Ganoid scales (i.e. scales with a bony plate covered by ganoin) occur in other taxa than ‘Semionotidae’. We observed the general morphology of ganoid scales from taxa belonging to several clades of basal actinopterygians (Fig. 4). Different morphologies can be recognized in scales located in the lateral abdominal area among these clades (scales located near the dorsal and ventral margins, as well as scales from the caudal area have less characteristic shape). Most taxa have one or two pegs located on the dorsal margin of the scales (basal teleosts, aspidorhynchiformes, Dapedium, marcosemiids, ionoscopiforms) or have no pegs at all (ginglymodi, amiiforms) (personal observations). ‘Semionotidae’ (Lepidotes, Semionotus and some closely related genera) are the only taxa we checked that have pegs at both the dorsal and ventral edges of the anterior margin of the scales (Fig. 4, arrows). This feature has been observed in several species referred to the genus Lepidotes housed in the Natural History Museum in London (i.e. L. latifrons, L. elvensis, L. minor and L. semiserratus). Although a broader survey will be necessary to assess the distribution of this feature, we hypothesize here that the presence of two pegs on the anterior margin of ganoid scales is an apomorphy shared by Lepidotes, Semionotus and probably some closely related taxa. This character allows assignment of isolated scales to ‘Semionotidae’.

130

L. CAVIN ET AL.

Fig. 4. Schematic outlines of scales from the abdominal flanks of various basal actinopterygians mapped on a composite phylogeny. ‘Semionotidae’ (Lepidotes, Semionotus and some closely related genera) are the only ones to bear pegs at the dorsal and ventral edges of the anterior margin of the scales (arrows). Simensichthys group: Siemnsichthys macrocephalus, redrawn from Arratia (2000, fig. 10); Aspidorhynchiformes: Belonostomus genevensis, MHNG V. 1672; Dapedium: Dapedium sp., MHNG V. 1589; ‘Semionotidae’: Lepidotes mantelli, redrawn from Woodward (1916, plate VII, fig. 5); Macrosemiidae: Notagogus denticulatus, redrawn from Schultze (1996, fig. 5); Ginglymodi: Atractosteus africanus, redrawn from Cavin et al. (1996, plate I, fig. 4); Amiiformes: Sinamia zdanskyi, redrawn from Schultze (1996, fig. 5); Ionoscopiformes: Oligopleurus sp. (regarded tentatively as an ionoscopiformes here), MHNG V. 1797.

Khlong Min Formation The locality of Mab Ching in peninsular Thailand, on the Shan-Thai block, yielded smooth ganoid scales and some skull fragments assigned to a ‘semionotid’ fish. Some scales are large, more than 2.5 cm in length, and correspond to individuals that were over 1 m in length. A small isolated tooth, with a bulbous crown and a conical cap of acrodine, is referred with caution to that taxon (Fig. 5). No button-shaped tooth has been reported from the Khlong Min Formation so far.

Phu Kradung Formation The Phu Kradung Formation yielded a rich ‘semionotid’ fauna. Two species, Lepidotes buddhabutrensis Cavin et al. 2003 and Isanichthys palustris Cavin & Suteethorn 2006, have been described from the Phu Nam Jun locality on the

basis of articulated specimens, and will not be mentioned here, except for purposes of comparison. Scales of both taxa were sampled on articulated specimens and are shown in Figure 5. It is noteworthy that L. buddhabutrensis has been referred with caution to the genus Lepidotes, but a study in progress shows that it is not closely related to the type species, L. elvensis (de Blainville 1818), and it will probably be assigned to a new genus. Two other localities of the Phu Kradung Formation yield abundant ‘semionotid’ remains: Khok Sanam and Chong Chat. Apart from numerous isolated thick and smooth ganoid scales and some skull bones (among them a probable extrascapula), the locality of Khok Sanam yielded five fragment of a single Lepidotes-like individual. Each piece shows large diamond-shaped ganoid scales. The almost complete left opercle is preserved (TF 8025, Fig. 6a). The bone is 1.7 times deeper than long and its surface is marked with faint grooves diverging from the centre of ossification. The articular facet is large, it protrudes slightly from the anterior margin, and its external side is marked by a thickening of the bone. The dorsal margin of the bone is straight and horizontal. The posterior margin is regularly curved and the bone tapers to a point ventrally. The anterior margin is slightly convex. An area along the anterior border of the bone forms a plane, which marks an angle with the main body of the bone; this surface was probably covered by the preopercle. The shape of the opercle of this Lepidotes-like fish differs from that of L. buddhabutrensis, which is almost square in shape and has no ventral tapering (Fig. 6b). The individual from Khok Sanam shows a fragment of an elongated bone, slightly shifted, with a ‘L’ section reminiscent of the cleithrum. Another piece contains the left cleithrum, still embedded in matrix with scales and fragments of bones. The identification of these fragmentary specimens is tentative. Isolated scales from Khok Sanam, with similar gross morphology to scales attached to the opercle, show the bases of two broken anterior pegs, which are anteriorly directed (Fig. 5). Although most ‘semionotids’ have an opercle with a straight and long horizontal ventral margin, there are ‘semionotids’, such as L. lennieri (Wenz 1967), with an opercle with a ventral tapering extremity. The shape of the opercle of the specimen from Khok Sanam differs from that of L. buddhabutrensis, which is almost square in shape and has no ventral tapering (Fig. 6b). Chong Chat yielded numerous isolated ‘semionotid’ material and a sub-complete specimen (TF 8026), which is currently under description. It is now unclear if this specimen belongs to L. buddhabutrensis. It can be distinguished from I. palustris by its deeper body, with at least 30 rows

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND

131

Fig. 5. Diversity of ‘semionotid’ remains in the Jurassic and Cretaceous of Thailand. Only a selection of localities and specimens is shown. The succession between localities within each formation does not correspond to the stratigraphic succession. The two lines of scales for Khok Pha Suam correspond to the two recognized taxa known in that locality (scales with ridges and smooth scales). Arrows show both anterior pegs. All specimens are at the same scale, unless otherwise specified (scale bars: 5 mm).

of scales along a vertical series in the middle of the body, whereas I. palustris has only 20, among other characters. Dan Luang, Phu Dan Kaeng, Lam Payang (Phu Klang) and Wan Din So are other localities of the Phu Kradung Formation yielding smooth ganoid Lepidotes-like scales. It is worth pointing out that

no button-shaped crushing teeth are known from the Phu Kradung Formation so far.

Sao Khua Formation The localities of Phu Wiang, Phu Phan Thong, Phu Mai Paw in the Sao Khua Formation have yielded

132

L. CAVIN ET AL.

Fig. 6. (a) Left opercle and fragment of squamation of a ‘semionotid’ fish (TF 8025), Khok Sanam, Phu Kradung Formation. (b) Left opercle of Lepidotes buddhabutrensis, Phu Nam Jun, Phu Kradung Formation, K12-2 (holotype). Scale bar: 30 mm.

smooth ganoid scales and button-shaped crushing teeth referable to ‘semionotids’ (Fig. 5). The localities of Phu Kum Khao (K4) and Phu Phok, in the Sao Khua Formation, have both yielded a single button-shaped tooth referable to a ‘semionotid’ (the Phu Phok material includes numerous ganoid scales, but most of them probably belong to sinamiids; see below). Cuny et al. (2006) reported button-shaped and hook-shaped pharyngeal teeth in the Phu Phan Thong locality that are referable to Lepidotes. The button-shaped teeth are rounded or ovoid, with a smooth surface and sometimes a small pit at their apex (Fig. 5). We are not able to distinguish several taxa among the available ‘semionotid’ remains from the Sao Khua Formation at present.

Khok Kruat Formation The Khok Pha Suam locality has yielded the most abundant ‘semionotid’ material from the Khok Kruat Formation. Thousands of isolated scales have been collected. They can easily be separated into two sets (Fig. 5). One set is represented by smooth ganoid scales, with the lateral face showing a narrow band without ganoin that was covered by the anterior scale. The other set consists of slightly larger scales on average, and the ganoid surface is ornamented with parallel or slightly diverging grooves that sometimes mark the posterior edge of the ganoin with a denticulate pattern. These scales are deeper on average than those

from the other set, they have better-marked edges, and the anterior ganoin-free area that was covered by the preceding scale is proportionally larger. In well-preserved scales, the anterior margin bears dorsally and ventrally a pair of well-developed pegs. As the morphology of the scales of the two sets is clearly distinguishable, and as no scale with intermediate morphology between the two sets has been found, we regard them as belonging to two different taxa. Dorsal ridge scales are abundant at Khok Pha Suam and show a great variety of shape. These scales have been used to distinguish species, or populations, among semionotid assemblages of the Late Triassic of the USA (McCune 1987). However, this diagnostic feature cannot be used on the Khok Pha Suam material, as articulated specimens are not available. Although we cannot completely rule out the possibility that the scales of Khok Pha Suam do not belong to a ‘semionotid’, their general morphology, with a pair of welldeveloped anterior pegs, is in favour of an assignment to ‘semionotids’. No button-shaped crushing teeth are known from Khok Pha Suam at present. This is probably not a sampling artefact, as these teeth are extremely resistant and the collecting effort was considerable at this site. In the locality of Lam Pao Dam a large Lepidotes-like fish is known from button-shaped crushing teeth and thick ganoid scales. The latter show traces of two anterior pegs, and their surface is ornamented with parallel rows of small pits (Fig. 5). An indeterminate piece of bone from the skull shows a rather similar ornamentation, indicating that this pattern probably covered the whole body. This pattern is unique among ‘semionotids’ from Thailand and probably corresponds to a particular taxon. Isolated ‘semionotid’ remains constitute a large component of the fossil vertebrates from Jurassic and Cretaceous localities of Thailand. Figure 5 provides an overview of the main types of scales and teeth in the succession of formations. Although few taxa are sufficiently known to be diagnosed, seven forms can be distinguished on the basis of the scales –teeth associations found among the assemblages: a large semionotiform in the Khlong Min Formation, two or three taxa in the Phu Kradung Formation (two in Phu Nam Jun, L. buddhabutrensis and Isanichthys, and possibly one in Chong Chat represented by a sub-complete specimen), one with crushing teeth in the Sao Khua Formation, three in the Khok Kruat Formation (two forms without crushing teeth in Khok Pha Suam distinguishable on the basis of the scale morphology, and a large one with crushing teeth in Lam Pao Dam). A histological study of the micro-ornamentation of the surface of the

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND

ganoin conducted by one of us (U.D.) should allow the testing of this diversity by using another approach. Halecomorphi sensu Grande & Bemis 1998 Caturidae Owen cf. Caturus Agassiz Sao Khua Formation Cuny et al. (2006) reported from the Phu Phan Thong locality pointed teeth with two welldeveloped carinae, which may belong to cf. Caturus. Sinamiidae Berg

Sao Khua Formation The first sinamiid from SE Asia, Siamamia naga Cavin et al. (2007b), was reported in the locality of Phu Phok. Since then, remains of sinamiids have been recognized in the collection of the Sirindhorn Museum, coming from several other localities in the Sao Khua Formation. An isolated basioccipital similar in structure to the basioccipital from Siamamia naga is known from the locality of Phu Mai Paw (TF 8027, Fig. 7b). It shares with S. naga lateral sides of the ossification deeply excavated, an anterior process that was embedded in the cartilage flooring the brain cavity, a V-shaped

133

concavity and paired unfinished areas that were capped with cartilage on the dorsal side of the bone. However, differences from S. naga are observed, especially in the proportions of the ossification, which is shorter, wider and flatter in the Phu Mai Paw material. The isolated ganoid scales from Phu Mai Paw may be referred to the sinamiid because of an almost complete covering of the dorsal surface by a thin layer of ganoin (there was probably no overlapping between scales) and because of the absence of peg and socket structure. A small articulated sinamiid specimen comprising the head and about half of the squamation was discovered in Phu Phok (TF 8028, Fig. 8). The skull, nearly 25 mm in length, is preserved in three dimensions and weakly distorted, but the ossifications are fractured, making the recognition of sutures difficult. The parietal appears to be unpaired, which is an autapomorphy of the sinamiids. The recognizable skull bone pattern agrees with the sinamiid pattern in the shape of the frontal, dermopterotic and preopercle, and the shape of the teeth. The scales are also reminiscent of sinamiid scales in their complete and thin ganoin covering. There are, however, significant differences from other sinamiids, in particular from Siamamia naga. The main difference is the lower jaw, which appears to be short and very deep, with a short anterior

Fig. 7. Sinamiidae. (a –c) Basioccipitals in dorsal views. (a) Siamamia naga (TF 8008), Phu Phok, Sao Khua Formation; (b) sinamiid indet. (TF 8027), Phu Mai Paw, Sao Khua Formation; (c) sinamiid indet. (TF 8030), Khok Pha Suam, Khok Kruat Formation. (d, e) Left dermopterotics. (d) Siamamia naga (TF 8005), Phu Phok, Sao Khua Formation; (e) sinamiid indet. (TF 8032), Khok Pha Suam, Khok Kruat Formation; (f, g) Right dentaries. (f) Siamamia naga, Phu Phok, Sao Khua Formation; (g) sinamiid indet. (TF 8031), Khok Pha Suam, Khok Kruat Formation. (h, i) Centra. (h) Siamamia naga (TF 8015), Phu Phok, Sao Khua Formation; (i) sinamiid indet. (TF 8029), Khok Pha Suam, Khok Kruat Formation. Scale bar: 15 mm.

134

L. CAVIN ET AL.

Fig. 8. Subcomplete individual of an indeterminate Sinamiidae (TF 8028), Phu Phok, Sao Khua Formation. (a) lateral view; (b) dorsal view. Scale bar: 30 mm.

toothed area (seven cylindrical teeth with a sharp apex are visible on the right hemi-mandible), followed posteriorly by a deep coronoid process. The ventral margin of the lower jaw is gently curved and the articulation with the quadrate is probably anterior to the orbit. This specimen, which probably belongs to a new taxon, is currently under description.

Khok Kruat Formation Five centra, a posterior fragment of a basioccipital, fragments of jaws, a piece of a left dermopterotic and scales from the locality of Khok Pha Suam can be assigned with confidence to a sinamiid fish. The centra are ovoid in shape, short, with lateral faces ornamented with small grooves in small individuals and almost smooth in large ones. One small centrum is pierced by a small hole for the notochord (TF 8029, Fig. 7i). The fragment of basioccipital (TF 8030) has the lateral sides deeply excavated with a foramen in the bottom, and paired unfinished areas that were capped with cartilage dorsally. The shape in ventral view, with diverging ridges, is more reminiscent of the sinamiid from Phu Mai Paw than of S. naga from Phu Phok (compare Fig. 7a, b and c). Pieces of jaws comprise two fragments of dentaries, a fragment of a premaxilla and two fragments referred with caution to maxillae. The largest fragment of dentary is a right one still bearing three broken teeth (TF 8031, Fig. 7g). The

teeth are slightly anteroposteriorly compressed, with a cylindrical shaft that slightly curved inward at the top. In internal view a large V-shaped concavity for the Meckel cartilage is present. The ventral margin of the bone marks a thinning just anterior to the anterior extremity of the concavity, indicating that the symphyseal region was probably shallower than the posterior region of the mandible. The other fragment shows laterally five elongated openings for the mandibular sensory canal and dorsally the sockets for seven compressed teeth. The general shape of these ossifications is reminiscent of S. naga (in the shape of teeth, and the opening for the sensory canal) but in the Khok Pha Suam specimen the lower jaw looks shallower at the symphysis, and was probably shorter in length. A fragment of a premaxilla, probably the posterior part of a left one, is present. The sockets for three large teeth, slightly compressed anteroposteriorly, are visible. There are no marked differences from that part of the premaxilla in S. naga. The two small fragments referred to maxillae bear deep sockets for the teeth. One has empty sockets, subrectangular in shape, and the other still has some small pointed teeth covered laterally by a bony lamella. The former specimen is reminiscent of S. naga, whereas the latter is only tentatively referred to a sinamiid here. The left dermopterotic (TF 8032) is reminiscent of that of S. naga with its flat dorsal face (skull roof) and a ventral lamina. However, this lamina is more developed than in

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND

the Phu Phok species and the ornamentation on the dorsal face is coarser (compare Fig. 7d and e). As described above, the sinamiid remains from Khok Pha Suam show differences from S. naga from the Sao Khua Formation of Phu Phok, and this sinamiid probably belongs to another taxon. However, more diagnostic material is required to coin a new name. Pycnodontiformes Berg cf. Gyrodus Agassiz

135

Khlong Min Formation A small prearticular pycnodont dentition, preserved on a slab of sandstone, is reported from the locality of Mab Ching (TF 8033, Fig. 9a and b). The specimen, 9 mm in length, bears three main rows of teeth. The lateral-most row comprises nine large ovoid teeth decreasing in size anteriorly. The two large posterior-most teeth bear two rows of faint tubercles, whereas the anterior teeth are smooth. The medial row consists of 12 transversally elongated

Fig. 9. Pycnodontiformes. (a–b) Left prearticular dentition of cf. Gyrodus (TF 8033), Mab Ching, Khlong Min Formation. (a) Complete dentition, scale bar: 3 mm; (b) close-up view of the outlined area in (a), scale bar: 2 mm. (c– f) Vomers ((c) TF 8034; (d) TF 8035) and isolated teeth ((e) TF 8036; (f) TF 8037) of cf. Anomoeodus, Phu Phan Thong, Sao Khua Formation. (c–f) Scale bars: 3 mm.

136

L. CAVIN ET AL.

Table 1. Jurassic and Cretaceous bony fish taxa from Thailand

Actino. indet. Khok Kruat Fm Sao Khua Fm Phu Kradung Fm Khlong Min Fm

cf. Ptycholepis

Semio. indet. 1

L. I. buddhabutrensis palustris

Semio. indet. Semio. indet. 2 (button 3 (smooth teeth) scales)  











teeth (one or two supplementary teeth may have been present anteriorly). Their contour is irregular in shape, because of the large cusps on the occlusal face. On the posterior-most tooth five cusps similar in size are transversally aligned, forming crenulations (Fig. 9b), and on the more anterior teeth the internal-most cusp increases in size. The anterior teeth are smooth. The intermediate row, between the lateral and the medial rows, is formed by at least 12 teeth, which are irregular in size and shape. The posterior teeth have one to three cusps, and the anterior teeth are smooth. At the anterior tip of the dentition the bases of two small rounded teeth are visible, lateral to the lateral row of large teeth. It is unclear whether they are supplementary teeth or are part of a fourth row that could have developed during growth (the specimen is probably a juvenile individual). Crenulations on vomering and prearticular teeth are present in the pycnodonts Gyrodus, Mesturus, Micropycnodon and Tepexichthys, as described by Poyato-Ariza & Wenz (2002). According to those workers, Mesturus has only subcircular prearticular teeth, whereas the other three genera have oval prearticular teeth; three rows are present in Tepexichthys on the prearticular, four rows are present in Micropycnodon and Gyrodus, and more than four rows are present in Mesturus, although two are irregular; Gyrodus and Tepexichthys have nine teeth on the main prearticular row, as in the described specimen. According to these characters, defined on articulated specimens, the dentition from Mab Ching should be referred to Tepexichthys, a genus described in the Albian of Mexico. However, if we regard the two lateral teeth on the prearticular as evidence of a fourth row, the specimen can be referred to Gyrodus, a genus known in the Late Jurassic of Europe and Chile. We retain here the latter identification, but this should be regarded with caution, as is usual for identification of isolated dentitions. cf. Anomoeodus Forir

Sao Khua Formation Pycnodont remains have been recently discovered in the locality of Phu Phan Thong (Cuny et al.

2006, P. Srisuk, pers. comm.). They consist of two small incomplete vomerine dentitions, plus some isolated teeth. The vomerine dentitions (TF 8034 and TF 8035, Fig. 9c and d, respectively) have a median row of ovoid teeth, slightly curved with the concave margin facing anteriorly. A ridge runs along the transversal axis, delimiting anteriorly a groove. On the largest teeth, some cusplets are present along the anterior margin of that depression. Lateral to the median row is a paired row of teeth, squarish in shape on one specimen and ovoid in shape on the other (the latter specimen probably corresponds to a more anterior portion of the vomer dentition than the former). A concavity, with irregular margin on the largest teeth, is present on the squarish teeth. A paired lateral row of ovoid teeth is present on both specimens. Isolated teeth, from the same locality, are referred to cf. Anomoeodus (TF 8036 and TF 8037, Fig. 9e and f, respectively). The unworn teeth are circular or ovoid, with a central rounded cusp and radiating merging ridges. Some of these teeth are larger than teeth on the vomers, indicating the occurrence of rather large pycnodont fishes. P. Srisuk (pers. comm.) referred isolated teeth from Phu Phan Thong similar in shape to the above-described teeth to Gyrodus sp. P. Srisuk (pers. comm.) reported four ‘splenial’ (prearticular) and one vomerine dentitions from the Sao Khua Formation in a locality situated along the highway between Udon Thani and Nong Bua Lam Phu. The gross morphology of the vomerine dentition (SHM-PT5, referred to Stemmatodus sp. by P. Srisuk, pers. comm.) is reminiscent of the vomerine dentitions from Phu Phan Thong. A prearticular dentition (SHM-PT1, referred to Micropycnodon sp. by P. Srisuk, pers. comm.) displays four rows of teeth. The main row is composed of elongated and slightly curved ovoid teeth. The lateral-most row shows smaller drop-shaped teeth. Between these rows of elongated teeth occurs a row of small sub-circular teeth, and another row with similar teeth occurs along the medial margin of the bone. Identification of these isolated fragmentary remains is difficult. If we regard this material as belonging to a single taxon, the shape and

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND

Semio. Semio. indet. 5 (button Sinamiid indet. 4 cf. Siamamia new (ridged teeth & pits on scales) Caturus naga species 1 scales) 



Sinamiid new cf. cf. species 2 Gyrodus Anomoeodus

137

Ferganoceratodus martini

 





 

ornamentation of vomerine teeth showing grooves and crenulations, and the arrangement and shape of prearticular teeth are reminiscent of the genus Anomoeodus as defined by Kriwet (1999), a genus known in the Early and Late Cretaceous, with doubtful occurrences in the Early Tertiary (Kriwet 1999). Srisuk also referred prearticular dentitions from Phu Phan Thong to Anomoeodus sp., and mentioned the occurrence of Micropycnodon and Stemmatodus in the locality of Phu Noi, Sakon Nakhon, in the Sao Khua Formation (P. Srisuk pers. comm.). We prefer here to retain the occurrence in the Sao Khua Formation of cf. Anomoeodus only, pending further discoveries. Sarcopterygii Romer Dipnoi Mu¨ller Ferganoceratodus Kaznyshkin & Nessov Ferganoceratodus martini was described from the Phu Nam Jun locality in the Phu Kradung Formation on the basis of an articulated skull roof and associated jaws (TF 7712, Cavin et al. 2007a). A taxon close to or similar to F. martini has been reported in the Triassic Huai Hin Lat Formation (Martin & Ingavat 1982), as well as in the Middle Jurassic locality of Mab Ching and in Ban Khok Sanam, in the Phu Kradung Formation (Martin et al. 1997; Cavin et al. 2007a) on the basis of isolated tooth plates. An isolated left mandibular tooth plate has been found in 2007 in the locality of Chong Chat, Phu Kradung Formation (TF 3038).

Succession of the Jurassic and Cretaceous fish assemblage in Thailand Table 1 shows an overview of the bony fish assemblages from the Jurassic and Cretaceous of Thailand. The diversity, with at least 16 taxa, is higher than previously assumed, especially for ‘semionotids’. Fishes from the Khlong Min Formation are still poorly known. The bony fish fauna provides little evidence to confirm or invalidate dating of the Mab Ching locality or to test

 

palaeogeographical scenarios for the Middle–Late Jurassic. The occurrence of Ferganoceratodus is in agreement, however, with a terrestrial connection between southern Thailand and Central Asia at that time (Buffetaut et al. 2006). The Phu Kradung formation contains the most diverse bony fish fauna so far. Two or three ‘semionotids’ are present (depending on whether the taxon from Chong Chat is a new species or belongs to L. buddhabutrensis). At present, we are not able to assess the palaeogeographical affinities of these taxa. Dinosaurs from the Phu Kradung Formation show affinities with those from the Upper Shaximiao Formation of Sichuan and the Shishugou Formation of Xinjiang (Buffetaut et al. 2006; Buffetaut & Suteethorn 2007), and further studies should allow us to test whether ‘semionotids’ from the Phu Kradung Formation also show close relationships with ‘semionotids’ from the Late Jurassic of West and South China. The possible occurrence of a palaeonisciform in the Phu Kradung Formation is of little palaeogeographical interest until the taxon is better known, but it may indicate a refugium for relicts in SE Asia, as was suggested for the Jehol biota (Luo 1999). Five bony fish taxa are known in the Sao Khua Formation and two of them represent the oldest occurrences in SE Asia: a ‘semionotid’ with buttonshaped teeth and sinamiids. The osteology of the ‘semionotid’ with button-shaped teeth is not sufficiently known to assess whether it has affinities with other ‘semionotids’ with similar teeth or whether this pattern is due to convergence. The sinamiids, however, indicate close relationships with Chinese faunas, but we cannot decide at present whether the Thai taxa are closer to sinamiids from the Lycoptera fauna from North China, to sinamiids from the Mesoclupea fauna from south China, or to the sinamiids from Korea and Japan described by Yabumoto et al. (2006) (Cavin et al. 2007b). It is worth mentioning that the dinosaur assemblage from the Sao Khua Formation appears rather different from the Early Cretaceous dinosaur assemblage of China (Buffetaut et al. 2006). The Khok Kruat bony fish assemblage is different from assemblages

138

L. CAVIN ET AL.

of the preceding formations, showing a new diversification of the freshwater ‘semionotids’.

Conclusion The present study shows the potential of bony fish remains for biostratigraphic and palaeogeographical studies, with very distinct faunas in each formation (and probably differences within each formation); however, most of the work remains to be done. For instance, Cuny et al. (2006) observed a wide geographical distribution of the hybodont taxa from the Sao Khua Formation, whereas the Khok Kruat shark assemblage appears to be more endemic, although with a wide distribution within a South Asian continental province (Cappetta et al. 2006). We cannot test this hypothesis with the bony fish record for the time being. One important issue is to assess whether species from the two main lineages reported here, ‘semionotids’ and sinamiids, form clades endemic to SE Asia in the Late Jurassic and Early Cretaceous, or are representatives of taxa from surrounding areas (especially mainland Asia) that dispersed at several times towards SE Asia. Phylogenetic analyses based on more complete osteological material, including microstructure of the ornamentation of the ganoin of the scales, are now necessary. This paper is based on material collected by the Thai – French palaeontological team for more than 25 years. Numerous people are involved in this work and they cannot be named individually here. We want to thank all these workers, who have participated in the collection and preparation of the fossils. We thank Z. Johanson (London) for access to the fossil fish collection in the Natural History Museum, and A. Lo´pez-Arbarello (Mu¨nchen) and J. Kriwet (Berlin) for reviewing the manuscript. L.C.’s research was partly supported by the Swiss National Science Foundation (grant 200021-113980).

References A RRATIA , G. 2000. New teleostean fishes from the Jurassic of southern Germany and the systematic problems concerning the ‘pholidophoriforms’. Pala¨ontologische Zeitschrift, 74, 113–143. A RRATIA , G., S CASSO , R. A. & K IESSLING , W. 2004. Late Jurassic fishes from Longing Gap, Antarctic Peninsula. Journal of Vertebrate Paleontology, 24, 41–55. B UFFETAUT , E. & S UTEETHORN , V. 2007. A sinraptorid theropod (Dinosauria: Saurischia) from the Phu Kradung Formation of northeastern Thailand. Bulletin de la Socie´te´ Ge´ologique de France, 178, 497–502. B UFFETAUT , E., R AKSASKULWONG , L., S UTEETHORN , V. & T ONG , H. 1994. First post-Triassic temnospondyl amphibians from the Shan-Thai Block: Intercentra

from the Jurassic of peninsular Thailand. Geological Magazine, 131, 837 –839. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J., K OBAYASHI , Y., H UANG , D. & L EE , Y. (eds) Papers from the 2005 International Dinosaur Symposium. Geological Publishing House, Beijing, 19– 37. B U¨ RGIN , T. 1992. Basal ray-finned fishes (Osteichthyes; Actinopterygii) from the Middle Triassic of Monte San Giorgio (Canton Tessin, Switzerland). Me´moires Suisses de Pale´ontologie, 114, 1 –164. C APPETTA , H., B UFFETAUT , E. & S UTEETHORN , V. 1990. A new hybodont from the Lower Cretaceous of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 11, 659–666. C APPETTA , H., B UFFETAUT , E., C UNY , G. & S UTEETHORN , V. 2006. A new elasmobranch assemblage from the Lower Cretaceous of Thailand. Palaeontology, 49, 547 –555. C ARTER , A. & B RISTOW , C. 2003. Linking hinterland evolution and continental basin sedimentation using detrital zircon thermochronology: A study of the Khorat Plateau Basin, Eastern Thailand. Basin Research, 15, 271– 285. C AVIN , L. & S UTEETHORN , V. 2006. A new Semionotiform (Actinopterygii, Neopterygii) from Upper Jurassic– Lower Cretaceous deposits of North-East Thailand, with comments on the relationships of Semionotiforms. Palaeontology, 49, 339 –353. C AVIN , L., M ARTIN , M. & V ALENTIN , X. 1996. De´couverte d’Atractosteus africanus (Actinopterygii, Lepisosteidae) dans le Campanien infe´rieur de Ventabren (Bouches-du-Rhoˆne, France). Implications pale´obioge´ographiques. Revue de Pale´obiologie, 15, 1– 7. C AVIN , L., S UTEETHORN , V., K HANSUBHA , S., B UFFETAUT , E. & T ONG , H. 2003. A new Semionotid (Actinopterygii, Neopterygii) from the Late Jurassic– Early Cretaceous of Thailand. Comptes Rendus Pale´vol, 2, 291–297. C AVIN , L., S UTEETHORN , V., B UFFETAUT , E. & T ONG , H. 2007a. A new Thai Mesozoic lungfish (Sarcopterygii, Dipnoi) with an insight into post-Palaeozoic dipnoan evolution. Zoological Journal of the Linnean Society, 149, 141–177. C AVIN , L., S UTEETHORN , V., B UFFETAUT , E., C LAUDE , J., C UNY , G., L E L OEUFF , J. & T ONG , H. 2007b. The first sinamiid fish (Holostei, Halecomorpha) from south-east Asia (Early Cretaceous of Thailand). Journal of Vertebrate Paleontology, 27, 827– 837. C UNY , G., S UTEETHORN , V., K AMHA , S., B UFFETAUT , E. & P HILIPPE , M. 2006. A new hybodont shark assemblage from the Lower Cretaceous of Thailand. Historical Biology, 18, 21– 31. C UNY , G., S RISUK , P., K HAMHA , V., S UTEETHORN , V. & T ONG , H. 2009. A new elasmobranch fauna from the Middle Jurassic of Southern Thailand. In: B UFFETAUT , E., C UNY , G., LE L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geogical Society, London, Special Publications, 315, 95–111. DE B LAINVILLE , H. D. 1818. Sur les ichthyolites ou les poissons fossiles. In: L EVRAULT , F. G. (ed.)

JURASSIC AND CRETACEOUS BONY FISHES FROM THAILAND Nouveau dictionnaire d’histoire naturelle, applique´e aux arts, a` l’e´conomie rurale et domestique, a` la Me´decine, etc. Deterville, Paris, 310–395. F ILLEUL , A. & V U K HUC , D. 2001. A new fish fauna from the Jurassic of Vietnam. Journal of Asian Earth Sciences, 19, 641– 647. G RANDE , L. 2005. Phylogenetic study of gars and closely related species, based mostly on skeletal morphology. The resurrection of Holostei. In: P OYATO -A RIZA , F. J. (ed.) Fourth International Meeting on Mesozoic Fishes—Systematics, Homology and Nomenclature, Miraflores de la Sierra, Extended Abstracts. Universidad Auto´noma, Madrid, 119–121. G RANDE , L. & B EMIS , W. E. 1998. A comprehensive phylogenetic study of amiid fishes (Amiidae) based on comparative skeletal anatomy. An empirical search for interconnected patterns of natural history. Journal of Vertebrate Paleontology, 18, Supplement, 1–690. H ILTON , E. J., G RANDE , L. & B EMIS , W. 2004. Morphology of Coccolepis bucklandi AGASSIZ, 1843 (Actinopterygii, Coccolepidae) from the Solnhofen Lithographic Limestone deposits (Upper Jurassic, Germany). In: T INTORI , A. & A RRATIA , G. (eds) Mesozoic Fishes III. Systematics, Paleoenvironments and Biodiversity. Dr Friedrich Pfeil, Mu¨nchen, 209– 238. K RIWET , J. 1999. Pycnodont fishes (Neopterygii, Pycnodontiformes) from the Lower Cretaceous of Una (E-Spain) with comments on branchial teeth in pycnodontid fishes. In: A RRATIA , G. & S CHULTZE , H.-P. (eds) Mesozoic Fishes 2—Systematics and Fossil Record. Dr Friedrich Pfeil, Mu¨nchen, 215–238. L EI , Z. 1993. The discovery and significance of the Late Jurassic sporopollen assemblage in peninsular Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology. Chiang Mai University, Chiang Mai, 361–379. L UO , Z. 1999. A refugium for relicts. Nature, 400, 23– 24. M ARTIN , M. & I NGAVAT , R. 1982. First record of an Upper Triassic Ceratodontid (Dipnoi, Ceratodontiformes) in Thailand and its paleogeographical significance. Geobios, 15, 111 –114. M ARTIN , M., B UFFETAUT , E., T ONG , H. & S UTEETHORN , V. 1997. New Jurassic dipnoans from Thailand. Geological Society of Denmark, On Line Series, 1. World Wide Web Address: http:// www.2dgf.dk/Publikationer/DGF_On_Line/Vol_1/ newjur.htm. M C C UNE , A. R. 1987. Toward the phylogeny of a fossil species flock: semionotid fishes from a lake deposit in the Early Jurassic Towaco Formation, Newark Basin. Peabody Museum of Natural History, Yale University Bulletin, 43, 1– 108.

139

M EESOOK , A. 2000. Cretaceous environments of Northeastern Thailand. In: O KADA , H. & M ATEER , N. J. (eds) Cretaceous Environment of Asia. Elsevier, Amsterdam, 207–223. M EESOOK , A., S UTEETHORN , V., C HAODUMRONG , P., T EERARUNGSIGUL , N., S ARDSUD , A. & W ONGPRAYOON , T. 2002. Mesozoic rocks of Thailand: a summary. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Resources, Bangkok, 82–94. M UTTER , R. 2004. The ‘Perleidiform’ family Colobodontidae: A review. In: T INTORI , A. & A RRATIA , G. (eds) Mesozoic Fishes III. Systematics, Paleoenvironments and Biodiversity. Dr Friedrich Pfeil, Mu¨nchen, 197– 208. O LSEN , P. E. & M C C UNE , A. R. 1991. Morphology of the Semionotus elegans species group from the Early Jurassic part of the Newark Supergroup of eastern North America with comments on the family Semionotidae (Neopterygii). Journal of Vertebrate Paleontology, 11, 269 –292. P OYATO -A RIZA , F. J. & W ENZ , S. 2002. A new insight into pycnodontiform fishes. Geodiversitas, 24, 139– 248. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. Part 1: Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5– 40. S CHULZE , H.-P. 1996. The scales of Mesozoic actinopterygians. In: A RRATIA , G. & V IOHL , G. (eds) Mesozoic Fishes—Systematics and Paleoecology. Dr Friedrich Pfeil, Mu¨nchen, 83–93. T EERARUNGSIGUL , N., R AKSASKULWONG , L. & K HANTAPRAB , C. 1999. Reconsideration of lithostratigraphy of non-marine Mesozoic rocks in Thung Yai–Khlong Thom area, Southern Thailand. In: K HANTRAPRAB , C. & S ARAPIROME , S. (eds) Proceedings of the Symposium on Mineral, Energy and Water Resources of Thailand: Towards the Year 2000, Department of Mineral Resources, Bangkok, 109– 114. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2002. Middle Jurassic turtles from southern Thailand. Geological Magazine, 139, 687– 697. W ENZ , S. 1967. Comple´ments a` l’e´tude des poissons actinopte´rygiens du Jurassique franc¸ais. Cahiers de Pale´ontologie, Paris, 1– 276. W OODWARD , A. S. 1916. The fossil fishes of the English Wealden and Purbeck formations. Part 1. Monograph of the Palaeontographical Society, 69, 1– 48. Y ABUMOTO , Y., Y ANG , S.-Y. & K IM , T. W. 2006. Early Cretaceous freshwater fishes from Japan and Korea. Journal of the Paleontological Society of Korea, 22, 119– 132.

Turtle assemblages of the Khorat Group (Late Jurassic – Early Cretaceous) of NE Thailand and their palaeobiogeographical significance HAIYAN TONG1 *, JULIEN CLAUDE2, VARAVUDH SUTEETHORN3, WILAILUCK NAKSRI4 & ERIC BUFFETAUT1 1

16 cour du Lie´gat, 75013 Paris, France

2

ISE-M, UMR 5554 CNRS, Universite´ de Montpellier 2, 34095 Montpellier, France

3

Bureau of Fossil Research and Museum, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand 4

Sirindhorn Museum, Sahatsakhan, Kalasin 46140, Thailand

*Corresponding author (e-mail: [email protected]) Abstract: The turtle assemblages from the Khorat Group consist mainly of trionychoids. They include the primitive Trionychoidae Basilochelys and basal eucryptodiran turtles from the Phu Kradung Formation (?Late Jurassic); the adocid Isanemys srisuki, the carettochelyid Kizylkumemys sp. and undetermined Trionychoidea from the Sao Khua Formation (Early Cretaceous); and the carettochelyid Kizylkumemys khoratensis and the adocid Shachemys sp. from the Khok Kruat Formation (Aptian). Our study shows some faunal links between the turtle faunas from the Khorat Group and those from the peripheral regions of Asia during the time span of the Khorat Group. Thus the coastal regions of Asia, and more particularly SE Asia, may have been important places for the origin and early diversification of the trionychoids.

The Mesozoic turtle record of Thailand ranges in age from the Late Triassic to the Early Cretaceous. It includes Proganochelys ruchae from the Late Triassic Huai Hin Lat Formation of NE Thailand (de Broin et al. 1982; de Broin 1984) and Siamochelys peninsularis from the Middle Jurassic of the southern peninsula (Tong et al. 2002). However, most turtle remains come from the Late Jurassic to Early Cretaceous Khorat Group, in the northeastern part of the country (Tong et al. 2003a, b, 2004a, 2005, 2006a, b). In this paper, we present an updated review that focuses on the turtles from the Khorat Group, as they constitute a succession of turtle assemblages during a crucial period for the evolution of Testudines, when many modern groups appeared. The Thai turtle assemblages are compared with coeval assemblages from other regions of Asia, especially those from China and Japan, to assess the palaeogeographical distribution of turtles in Asia during the Late Jurassic and Early Cretaceous and to gain a better understanding of turtle evolution during that period. The Khorat Group is a succession of non-marine deposits occurring on the Khorat Plateau, a vast tabular area in the northeastern part of Thailand. It overlies uncomformably the Late Triassic Nam Phong Formation. According to the current

interpretations (Racey et al. 1996), the Khorat Group consists of five formations. They are, from bottom to top, the Phu Kradung, Phra Wihan, Sao Khua, Phu Pan and Khok Kruat Formations. Three of them (Phu Kradung, Sao Khua and Khok Kruat) have yielded turtle remains.

The Late Jurassic(?) Phu Kradung Formation The Phu Kradung Formation consists of sandstones, siltstones and mudstones of mainly fluvial origin (Racey et al. 1996). It is dated as either Late Jurassic or more probably basal Cretaceous on the basis of palynology (Racey et al. 1996) and detrital zircon thermochronology (Carter & Bristow 2003), whereas the evidence from fossil vertebrates supports a Late Jurassic age for this formation (Buffetaut et al. 2006; Buffetaut & Suteethorn 2007). The vertebrate fauna of the Phu Kradung Formation includes freshwater sharks, actinopterygian fishes, temnospondyl amphibians, turtles, crocodiles, and sauropod, theropod and ornithopod dinosaurs (Buffetaut et al. 2006). Turtle remains are abundant in some localities, but they are often very fragmentary and poorly preserved.

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 141–152. DOI: 10.1144/SP315.11 0305-8719/09/$15.00 # The Geological Society of London 2009.

142

H. TONG ET AL.

The recently discovered material of a large turtle, Basilochelys macrobios, found in Mukdahan Province, includes a braincase, two shells and elements of the appendicular skeleton (Tong et al. 2009). Basilochelys has a low shell without a keel on the carapace, with a shell length reaching 900 mm. Although it is primitive in some aspects, several derived features support its trionychoid affinity. These characters include sculpture on both skull roof and shell surface, a neural formula of 6 . 4 , 6 , 6 , 6 , 6, very narrow vertebral scutes, anteroposteriorly elongated eleventh and twelfth marginal scutes overlapping the suprapygal and costal plates, a plastron sutured to the carapace, with large anterior and posterior lobes, a long bridge, a wide entoplastron and narrow axillary and inguinal notches. However, the presence of a large foramen basisphenoidale, the foramen posterius canalis caroticus laterale exposed on the ventral surface of the skull, amphicoelous cervical vertebrae and long first thoracic ribs extending to the lateral end of the first costal are primitive characters reminiscent of basal eucryptodiran turtles such as xinjiangchelyids or macrobaenids. The phylogenetic analysis placed Basilochelys in a basal position among Trionychoidae and near the origin of Nanhsiungchelyidae and Adocidae. The remains of Basilochelys are common in the Phu Kradung Formation; they have been collected at most localities of that age on the Khorat Plateau. Other shell fragments collected in the Phu Kradung Formation suggest the occurrence of more than one morphotype of large trionychoids. Isolated and fragmentary shell plates, collected near Ban Khok Sanam, in Kalasin Province, indicate the presence of other taxa. Fragments of plastron with a peg-like process on the lateral plastral margin indicate a ligamentous attachment between the carapace and plastron, as in xinjiangchelyids and macrobaenids. The surface of the neurals and costals shows fine plications arranged in a radiating manner, which is reminiscent of some Xinjiangchelys species, such as X. latimarginalis from the Late Jurassic Shangshaximiao Formation of Sichuan, China (Young & Chow 1953; Ye 1994). Such ornamentation is, however, not limited to xinjiangchelyids, but also occurs in some macrobaenids–sinemydids, such as Wuguia efremovi (Danilov & Sukhanov 2006). These turtle remains are thus referable to xinjiangchelyids or macrobaenids–sinemydids. If the above-mentioned specimens belong to xinjiangchelyids, the turtle assemblage from the Phu Kradung Formation could be correlated with the Late Jurassic turtle fauna of China and Mongolia. Xinjiangchelyids are known from the Middle Jurassic to Early Cretaceous and were the dominant group of turtles during the Late Jurassic in Asia.

In western China, the Qigu Formation of the Shishugou Group in Xinjiang has yielded two species of Xinjiangchelys: X. latimarginalis (including X. junggarensis (Ye 1986)) and X. qiguensis (Peng & Brinkman 1993; Matzke et al. 2004a). Most turtles from the Shangshaximiao Formation and equivalent beds in Sichuan, and the Upper Lufeng Formation in Yunnan, which were previously referred to Plesiochelys (Young & Chow 1953; Ye 1963, 1973a, 1994; Ye & Fang 1982; Peng et al. 2005) can also be assigned to Xinjiangchelys or Xinjiangchelyidae. In Mongolia, several taxa of xinjiangchelyids (Annemys latiens, A. levensis, Shartegemys laticentralis and Undjulemys platensis) have been reported from the Late Jurassic Shar Teeg in the Transaltai ¨ nju¨u¨l in Central Mongolia (Sukhanov Gobi and O 2000; Sukhanov & Narmandakh 2006). The large trionychoid Basilochelys from the Phu Kradung Formation has no counterpart in the Late Jurassic of anywhere else in Asia. On the whole, the Jurassic trionychoid record is poor. An incomplete shell originally described as ‘Plesiochelys’ tatsuensis (Ye 1963) from the Late Jurassic of Dazu, Chongqing was later renamed as Yehguia and considered as an adocid (Nessov 1977b; Meylan & Gaffney 1989) or closely related to Adocidae and Nanhsiungchelyidae (Danilov & Parham 2006). Originally attributed to the Trionychidae (Young & Chow 1953), Sinaspideretes wimani, an incomplete shell from ?Late Jurassic or Early Cretaceous deposits along the Chengyu railway, Sichuan, was later tentatively placed in the Carettochelyidae (Meylan & Gaffney 1992). Considered as one of the basal-most members of Trionychoidae, Basilochelys represents an important evolutionary link between the basal Eucryptodira from the Jurassic and more advanced trionychoids from the Cretaceous. At the moment, the turtle assemblage from the Phu Kradung Formation does not provide much information about its age. The primitiveness of the large trionychoid Basilochelys with regard to the Cretaceous nanhsiungchelyids would support the idea that the Phu Kradung Formation is older than mid-Cretaceous, and can be either Late Jurassic or Early Cretaceous in age. What can be noticed is that the composition of the Phu Kradung turtle assemblage is unusual when compared with those from the Late Jurassic or Early Cretaceous of mainland Asia, in that it is mainly composed of trionychoids. The occurrence of basal eucryptodiran xinjiangchelyids or macrobaenids –sinemydids in the Phu Kradung Formation may provide a possible correlation with the turtle faunas located farther north in Asia, but better material from Thailand is needed for further comparative studies. Furthermore, the Phu Kradung turtle assemblage seems to be more closely related to that of southern

MESOZOIC TURTLES OF NE THAILAND

China than to those farther north, because of the occurrence in the Late Jurassic of Chongqing of Yehguia, a primitive trionychoid that is considered as a primitive relative of Adocidae þ Nanhsiungchelyidae (Danilov & Parham 2006). It should be noted also that some Chinese Late Jurassic turtle assemblages may be older than the Phu Kradung Formation, as the Shangshaximiao Formation and Qigu Formation were both placed in the lower part of the Late Jurassic (Dong 1992; Peng & Brinkman 1993; Peng et al. 2005).

The Early Cretaceous Sao Khua Formation The Early Cretaceous Sao Khua Formation consists of red sandstones, mudstones and conglomerates,

143

deposited in low-energy, meandering rivers and on extensive floodplains (Mouret et al. 1993; Racey et al. 1996). The age of the Sao Khua Formation is considered as ante-Aptian and probably not basal Cretaceous (Mouret et al. 1993; Buffetaut & Suteethorn 1999). The Sao Khua Formation has yielded the most abundant and most diversified vertebrate fauna of the Khorat Group, comprising sharks, bony fishes, turtles, crocodiles, dinosaurs and birds (Cavin et al. 2007). Turtle remains are frequently found in the Sao Khua Formation. The adocid turtle Isanemys srisuki Tong, Buffetaut and Suteethorn, 2006 (Figs 1 and 2) was originally based on more than 20 shells collected from two localities: Phu Kum Khao, in Kalasin Province and Phu Wat 1, in Khon Kaen Province (Tong et al. 2006a). Additional isolated shell plates of I. srisuki have

Fig. 1. The adocid Isanemys srisuki (K4-658 (holotype, right (a, c) left (b, d)) and K4-659) from the Early Cretaceous Sao Khua Formation; Phu Kum Khao locality, Kalasin Province, NE Thailand. (a, c) dorsal view; (b, d) ventral view. Scale bar: 50 mm. (After Tong et al. 2006a.)

144

H. TONG ET AL.

Fig. 2. Reconstruction of Isanemys srisuki from the Early Cretaceous Sao Khua Formation, NE Thailand (modified from Tong et al. 2006a).

been recovered from Ban Na Krai, in Kalasin Province, and Phu Phok, in Sakon Nakhon Province. Isanemys presents various adocid characters, including the punctuate ornamentation of the shell surface, an incomplete neural series with the posterior costals meeting on the midline, two suprapygals, the first one very small and the second much larger, and the plastron shorter than the carapace, with a truncated anterior lobe and a straight anterior margin. The shell of Isanemys is primitive relative to that of other adocids, including Adocus from the Late Cretaceous of North America (Meylan & Gaffney 1989), Adocoides from the Late Cretaceous of Mongolia (Narmandakh 1985; Sukhanov 2000; Sukhanov & Narmandakh 2006), Ferganemys from the midCretaceous of Kirghizstan and Uzbekistan (Nessov & Khozatskii 1977; Sukhanov 2000) and an unnamed trionychoid from the Neocomian of Japan (Hirayama 2000; Hirayama et al. 2000). The primitive characters of Isanemys include the neural formula of 4 . 6 . 6 . 6 . 6 . 5(7), a longer and not medially expanded pectoral scute and a larger cervical scute. Isanemys is thus considered as the sister taxon of all other known adocids. The carettochelyid turtle, Kizylkumemys sp. (Fig. 3c– f) is represented by abundant, but very fragmentary shell plates from Phu Wat 2, in Khon

Kaen Province (Tong et al. 2004a). Additional material of K. sp. has been collected at Phu Mai Paw and Khok Kong, Kalasin Province, and Phu Phan Thong, Nong Bua Lamphu Province. The strongly keeled neural plates closely resemble K. schultzi from the Late Cretaceous of Uzbekistan and Mongolia (Nessov 1977a, 1981; Sukhanov 2000), but the fragmentary nature of the specimens prevents detailed comparisons. An isolated nuchal plate collected from the Phu Paeng locality, Kalasin Province, represents a third taxon of trionychoids. It is much wider than long and presents strong vermiculated ridges on the surface (Tong et al. 2003b). It is not easy to compare the Sao Khua turtle assemblage with those from China and Central Asia as the Early Cretaceous beds of these regions have yielded mainly macrobaenid –sinemydid turtles. In China, the formations roughly equivalent to the Sao Khua Formation on the basis of their geological age that have yielded turtle remains include the Yixian and Jiufotang Formations in Liaoning, the Tugulu Group in the Junggar Basin, Xinjiang, the Mengyin Formation in Shandong, and the Luohandong Formation in Inner Mongolia. Based on radiometric dating, the Yixian and Jiufotang Formations of Liaoning, in NE China, are considered as ranging from Barremian to Aptian in age (Ji et al. 2004). The Yixian Formation

MESOZOIC TURTLES OF NE THAILAND

145

Fig. 3. The carettochelyid Kizylkumemys from the Khorat Plateau, NE Thailand. (a, b) Partial carapace of K. khoratensis (NRRU A1861, holotype) from the late Early Cretaceous Khok Kruat Formation (Ban Saphan Hin, Nakhon Ratchasima Province). Scale bar: 20 mm (after Tong et al. 2005). (c, f) Neurals of Kizylkumemys sp. from the Early Cretaceous Sao Khua Formation (Phu Wat 2, Khon Kaen Province), in lateral (c, e) and dorsal (d, f) views. Scale bar: 10 mm. (After Tong et al. 2006b.)

has yielded abundant specimens of the macrobaenid Ordosemys liaoxiensis (Ji 1995; Li & Liu 1999; Tong et al. 2004b). A sinemydid, Manchurochelys manchoukuoensis, has been reported from equivalent beds in the same area (Endo & Shikama 1942). Turtle remains have also been reported from the overlying Jiufotang Formation (Liu 2003), but they have not been described. In eastern China, the Mengyin Formation, in Shandong, has yielded Sinemys lens, Scutemys tecta and Sinochelys applanata. This turtle fauna is dominated by far by the sinemydid S. lens (Wiman 1930; Brinkman & Peng 1993b). Sinochelys and Scutemys were placed later in the Sinochelyidae (Chkhikvadze 1987). The Mengyin Formation is regarded as either Early Cretaceous (Wiman 1930), or Late Jurassic (Dong 1992) in age. In Inner Mongolia, the Luohandong Formation contains the macrobaenid Ordosemys leios and the sinemydid Sinemys gamera (Brinkman & Peng 1993a, b; Brinkman & Wu 1999). Further to the west, in Xinjiang, two formations of the Tugulu Group have yielded turtles along the southern margin of the Junggar Basin: Wuguia hutubeiensis from the Hutubei Formation (Matzke & Maisch 2004; Matzke et al. 2004b) and Dracochelys wimani (a synonym of Wuguia efremovi according to Danilov & Sukhanov (2006)) from the overlying Lianmouxin Formation (Maisch et al. 2003). From the northwestern part of the Junggar basin, Ye (1973b) reported a dozen turtles from the Early Cretaceous Tugulu Group in the Wuerho area. Most of these turtles were assigned to Sinemys wuerhoensis (Ye 1973b) and one skull was described as

Dracochelys bicuspis (Gaffney & Ye 1992). Additional material of D. bicuspis has recently been described from the same horizon in Xinjiang (Brinkman 2001). The revision of the type series of S. wuerhoensis revealed that the specimens actually belong to three taxa: most of them belong to the macrobaenid Ordosemys (O. brinkmania); the type specimen is referred to Xinjiangchelys sp. and one isolated skull is attributed to Trionychia (Danilov & Parham 2007). The Tugulu Group is considered as Early Cretaceous in age, with the Hutubei Formation of Hauterivian –Barremian age and the Lianmuxin Formation of Aptian –Albian age (Eberth et al. 2001). In Russia, the Neocomian Gusinoje Ozero Group in the Baikal Lake region contains mainly specimens of the macrobaenid Kirgizemys dmitrievi (Nessov 1984; Danilov et al. 2006). To sum up, the Early Cretaceous turtle assemblages from China and Central Asia predominantly contain macrobaenids –sinemydids. In contrast to the turtle assemblage from the Sao Khua Formation, the trionychoids are fairly rare. Beside the Trionychia skull from Wuerho mentioned above, other trionychoid records include Aspideretes maortuensis, a nearly complete trionychid shell from Dashukou, Maortu, Inner Mongolia, and Sinamyda fuchienensis, a carapace from Hekou, Fujian, but the Early Cretaceous age of both specimens is uncertain (Ye 1965, 1974, 1994; Chkhikvadze 2000). Some fragmentary trionychid specimens have also been reported from the Early Cretaceous On Gong Formation of Inner Mongolia (Young & Chow 1953).

146

H. TONG ET AL.

The difference in composition between the turtle assemblage from the Sao Khua Formation and those from China and Central Asia suggests that SE Asia was more isolated from mainland Asia during the Early Cretaceous than before; a conclusion also supported to some extent by the composition of the dinosaur fauna from the Sao Khua Formation (Buffetaut et al. 2006) and by the analysis of whole vertebrate assemblage (Fernandez et al. 2009). On the other hand, the Sao Khua turtle fauna is more comparable with that from the Neocomian Kuwajima Formation of the Tetori Group in Central Japan. The lacustrine deposits of the Kuwajima Formation have yielded trionychoids, possible testudinoids and sinemydids (Hirayama 2000; Hirayama et al. 2000). The same trionychoid turtles, as well as a xinjiangchelyid (Hirayama 2006) are also reported from the equivalent Okurodani Formation. Based on shell morphology, these trionychoids are members of Adocidae, as is Isanemys from the Sao Khua Formation. The Neocomian age of this horizon is well supported by fission-track dating of the overlying tuff from the Okurodani Formation and palaeontological evidence (Hirayama et al. 2000).

The late Early Cretaceous Khok Kruat Formation The Khok Kruat Formation is the geologically youngest formation of the Khorat Group. It consists mainly of red siltstones, sandstones and conglomerates, indicative of a predominantly fluvial depositional environment (Racey et al. 1996). The age of the Khok Kruat Formation is well constrained as Aptian– Albian by the occurrence of the freshwater hybodont shark Thaiodus ruchae (Cappetta et al. 1990). It is considered as Aptian based on palynology (Racey et al. 1996; Racey & Goodall 2009). The vertebrate fauna of the Khok Kruat Formation includes hybodont sharks, actinopterygian fishes, turtles, crocodiles, and sauropod, theropod and ornithopod dinosaurs (Buffetaut et al. 2005). Only trionychoids have been hitherto recorded from the Khok Kruat Formation. The carettochelyid Kizylkumemys khoratensis Tong, Suteethorn, Claude, Buffetaut & Jintasakul 2005 is represented by shell material from the Ban Saphan Hin and Ban Khok Kruat localities, in Nakhon Ratchasima Province, and the Khok Pa Suam locality in Ubon Ratchathani Province (Tong et al. 2005). K. khoratensis presents the typical morphology of the vertebral scutes of Anosteirinae, with the very large first vertebral scute extending from the nuchal to the third costal plate and the second vertebral bounded laterally by the first vertebral scute. The presence of a small central scute on the second to

fourth neurals and included in the second vertebral scutes, and the sulcus between the first and the second vertebral scutes crossing the second neural are characteristic of Kizylkumemys. The carapace of K. khoratensis reaches 350 mm in length. It differs from the type species of Kizylkumemys, K. schultzi Nessov, 1977 from the Cenomanian – Turonian of Karakalpakia, Uzbekistan (Nessov 1977a), mainly by the absence of the dorsal keel (Fig. 3a and b). Although the high keel on the carapace of K. schultzi presents some sexual dimorphism, both males and females had a strong keel on the carapace. The male probably had a higher dorsal keel than the female (Nessov 1986). Adocids are represented by shell material recovered from Ban Sapan Hin, in Nakhon Rachasima Province (Tong et al. 2005) and Khok Pa Suam, in Ubon Ratchatani Province (Fig. 4). The tiny pores on the shell surface, the large plastron with wide posterior lobe and narrow inguinal notch, and the presence of large inframarginal scutes are reminiscent of Shachemys. Shachemys is a freshwater turtle characterized by a hinge between the epiplastra and ento-hyoplastra, and by the absence of all or nearly all neural plates. The Thai Shachemys is too incomplete for specific assignment. Three species of Shachemys have been so far described: S. laosiana from the late Early Cretaceous of Laos (Lapparent de Broin 2004), and S. baibolatica and S. ancestralis from the Late Cretaceous of Central Asia (Kuznetzov 1976; Nessov & Krasovskaya 1984; Nessov 1986; Danilov et al. 2007). Although the Thai Shachemys and S. laosiana are similar in age, the size of Shachemys from Thailand (shell length 400 mm) is clearly larger than that of S. laosiana (165–200 mm), but similar to that of S. baibolatica. In addition, a few fragmentary shell plates collected near Khorat are referable to Trionychidae. Comparison with approximately coeval turtle faunas from other countries in Asia may begin with the ‘Gre`s supe´rieur’ Formation in Laos, dated as Aptian –Albian by the occurrence of the fresh water bivalve Trigonioidea. That formation is considered as an equivalent of the Khok Khruat Formation on the eastern fringe of the Khorat Plateau (Buffetaut 1991). The turtle assemblage from the red floodplain deposits of that formation in Savannakhet province contains mainly trionychoids, which include the adocid Shachemys laosiana, Carettochelyidae, Trionychidae, and an uncertain basal eucryptodira referred to Aff. Xinjiangchelys sp. (de Lapparent de Broin 2004). This turtle fauna closely resembles that from the Khok Kruat Formation in the trionychoid-dominated composition, and in the presence of Shachemys and carettochelyids, although the assignment to Xinjiangchelys is doubtful.

MESOZOIC TURTLES OF NE THAILAND

147

Fig. 4. The adocid Shachemys sp. from the late Early Cretaceous Khok Kruat Formation (Ban Saphan Hin, Nakhon Ratchasima Province). (a, b) Fragment of carapace with right seventh and eighth peripherals and lateral part of the fifth and sixth costal plates (NRRU A1254) in dorsal (a) and ventral (b) views; (c) right hypoplastron (NRRU A1864) in ventral view. Scale bar: 40 mm. (d) Detail of ornamentation on plastron (NRRU A1864). (After Tong et al. 2006b.)

The Khok Kruat turtle fauna is comparable with that from the Kitadani Formation of the Tetori Group, in central Japan. The Kitadani Formation is considered as Barremian or Aptian in age and contains mainly trionychoids (the adocid Adocus sp., the nanhsiungchelyid Basilemys sp. and Trionychidae), and a few Sinemydidae and Testudinoidea (Hirayama 2002). Some resemblances can also be noted between the turtle assemblage from the Khok Kruat Formation and those from Central Asia. The

turtles from the Alamyshyk Formation (Albian) of Kylodzhun, in the Fergana Depression (Kirghizstan) consist of the adocid Ferganemys, the trionychid Trionyx and the macrobaenid Kirgizemys. The lower and middle parts of the Khodzhakul Formation (Upper Albian (Nessov 1984) or Lower Cenomanian (Averianov & Archibald 2005)), western Sultanuvais Ridge, in Uzbekistan, have yielded macrobaenids, adocids, nanhsiungchelyids, carettochelyids and testudinoids (Nessov 1984). As mentioned above, the adocid Ferganemys is

148

H. TONG ET AL.

more advanced than Isanemys from the Sao Khua Formation (Tong et al. 2006a). On the other hand, the turtle assemblages from the late Early Cretaceous of China and Mongolia are different from that from the Khok Kruat Formation. There are few late Early Cretaceous deposits in China that have yielded turtle remains. In Gansu, several turtle taxa (Osteopygis, Peishanemys, Heishanemys, Tsaotanemys and Yumenemys) were collected by the Sino-Swedish expedition from the Early Cretaceous beds in the Jiayuguan area in 1929–1930 (Bohlin 1953). According to Dong (1992), these specimens come from the Chijinbao Formation, which belongs to the Jehol Biota. The Chijinbao Formation is the lowermost of three formations of the Xinminbao Group (Ma et al. 1982). Based on the invertebrates and vertebrates collected in the neighbouring Mazhongshan area, the age of the Xinminbao Group is considered as ranging from the Late Barremian to the Albian (Tang et al. 2001). Among the turtle remains from Jiayuguan described by Bohlin (1953), the so-called Osteopygis was recently referred to the macrobaenid Kirgizemys (Danilov et al. 2006), Tsaotanemys is considered as a primitive testudinoid (Lindholmemydidae) (Hirayama et al. 2000), and Peishanemys (probably including Heishanemys) is placed in the Sinochelyidae (Chkhikvadze 1987). Peishanemys is also reported from Aptian –Albian deposits in Mongolia (Nessov & Verzilin 1981; Nessov 1987) and from the Qingshan Formation of Shandong, eastern China (Chow 1954). Based on isotopic dating, the age of the Qingshan Formation is Aptian (Dong 1992). In Mongolia, the Aptian – Albian turtle assemblage from Ho¨o¨vo¨r is dominated by the macrobaenid Kirgizemys (Hangaiemys) boburensis (Sukhanov 2000). To sum up, the turtle assemblage from the Khok Kruat Formation in the Khorat Plateau (including the ‘Gre`s supe´rieurs’ Formation in Laos) resembles those from Japan and Central Asia in being dominated by trionychoids and in the occurrence of the carettochelyid Kizylkumemys. The Chinese and Mongolian turtle assemblages are different, being dominated by macrobaenids. Two macrobaenid genera had a wide geographical distribution during that period: Kirgizemys (including Hangaiemys according to Danilov et al. (2006)), represented by five species and some Kirgizemys sp. records, is known from Barremian to Albian deposits in Mongolia, Kirghizstan, northern China, South Korea and Russia (Sukhanov 2000; Averianov et al. 2006; Danilov et al. 2006); and Ordosemys, roughly coeval with Kirgizemys, represented by four species, is recorded from northern and NE China and Mongolia (Brinkman & Peng 1993a; Tong et al. 2004b; Danilov & Parham 2007).

Conclusion The turtle assemblages from the Khorat Group consist mainly of trionychoids, which include the primitive Trionychoidae Basilochelys from the Late Jurassic(?) Phu Kradung Formation; the adocid Isanemys srisuki, the carettochelyid Kizylkumemys sp., and undetermined Trionychoidea from the Sao Khua Formation; and the carettochelyid Kizylkumemys khoratensis and the adocid Shachemys sp. from the Khok Kruat Formation. Basal eucryptodiran xinjiangchelyids or macrobaenids – sinemydids are represented in the Phu Kradung Formation by a few fragmentary remains. The turtle fauna from the Sao Khua Formation appears to be more similar to that from the Khok Kruat Formation than to that of the Phu Kradung Formation. Adocids and carettochelyids are recorded from both Sao Khua and Khok Kruat Formations; however, they are represented by different taxa. On the basis of the above comparison with the Late Jurassic to Early Cretaceous turtle faunas from other areas of Asia, the turtle assemblage from the Phu Kradung Formation can be correlated with either the Late Jurassic or the Early Cretaceous turtle assemblages of China, because of the presence of basal eucryptodiran xinjiangchelyids or macrobaenids –sinemydids, and it shows apparently more similarities to the turtle fauna from southern China than to those from more northerly areas. The turtles from the Sao Khua Formation show similarities to those from the Neocomian Kuwajima Formation in Japan, with the occurrence of adocids, but are different from those of China, which are dominated by macrobaenids –sinemydids. This may suggest that SE Asia became more isolated than before from mainland Asia during the Early Cretaceous. The differences between the Sao Khua dinosaur assemblage and the roughly coeval assemblages in China, notably those from the Jehol Group of NE China, have already been noted, and several hypotheses have been put forward, including differences in taphonomic conditions, and the existence of geographical or environmental barriers (Buffetaut et al. 2006; Fernandez et al. 2009). One possible bias may be that there are very few Early Cretaceous vertebrate localities in southern China and turtles of that age are particularly poorly known from that region. The close resemblance between the turtle fauna of the Khok Kruat Formation and the ‘Gre`s supe´rieur’ Formation in Laos supports that they belonged to the same palaeobiogeographical province during the late Early Cretaceous. They closely resemble the faunas from the Kitadani Formation of Japan and from Central Asia, but are very different from those from China and Mongolia. On the whole, within a palaeogeographical context, there are some faunal links

MESOZOIC TURTLES OF NE THAILAND

between SE Asia and other peripheral regions during the time span of the Khorat Group. In conclusion, our study supports the hypothesis of Hirayama et al. (2000), according to which during the Late Jurassic and Early Cretaceous there were two main palaeobiogeographical areas in Asia that controlled turtle distribution: the palaeogeographically coastal area (including SE Asia (Thailand and Laos), Japan and Central Asia (Uzbekistan, Kirghizstan and Kazakhstan)), which was dominated by trionychoid turtles, and the palaeogeographically inland area (China, Mongolia and the Lake Baikal region of Russia), which was dominated by basal eucryptodiran turtles such as xinjiangchelyids and macrobaenids –sinemydids. The coastal regions of Asia, and more particularly SE Asia, may have been an important area for the origin and early diversification of the trionychoids. This work was supported by the Department of Mineral Resources, Thailand, the ECLIPSE Programme of CNRS, France, the TRF-CNRS Special Programme for Biodiversity Research and Training (BRT/BIOTEC/ NSTDA), Grant BRT R-24507, and the PHC ‘Evolution et biodiversite´ des verte´bre´s aquatiques Me´sozoiques et Ce´nozoiques en Thaı¨lande’ (16610UG). We thank I. Danilov (St Petersburg) and R. Hirayama (Tokyo) for reviewing the manuscript. This is publication ISEM 2008-011 of J.C.

References A VERIANOV , A. O. & A RCHIBALD , J. D. 2005. Mammals from the mid-Cretaceous Khodzhakul Formation, Kyzylkum Desert, Uzbekistan. Cretaceous Research, 26, 593–608. A VERIANOV , A. O., V ORONKEVICH , A. V., L ESHCHINSKIY , S. V. & F AYNGERTZ , A. V. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of West Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology, 4, 359–395. B OHLIN , B. 1953. Fossil reptiles from Mongolia and Kansu. Statens Etnografiska Museum, Stockholm. B RINKMAN , D. 2001. New material of Dracochelys (Eucryptodira: Sinemydidae) from the Junggar Basin, Xinjiang, People’s Republic of China. Canadian Journal of Earth Sciences, 38, 1645–1651. B RINKMAN , D. & P ENG , J.-H. 1993a. Ordosemys leios n. gen., n. sp., a new turtle from the Early Cretaceous of the Ordos Basin, Inner Mongolia. Canadian Journal of Earth Sciences, 30, 2128–2138. B RINKMAN , D. & P ENG , J.-H. 1993b. New material of Sinemys (Testudines, Sinemydidae) from the Early Cretaceous of China. Canadian Journal of Earth Sciences, 30, 2139– 2152. B RINKMAN , D. & W U , X.-C. 1999. The skull of Ordosemys, an Early Cretaceous turtle from the Inner Mongolia, People’s Republic of China, and the interrelationships of Eucryptodira (Chelonia, Cryptodira). Paludicola, 2, 134– 147.

149

B UFFETAUT , E. 1991. On the age of the Cretaceous dinosaur-bearing beds of southern Laos. Newsletter on Stratigraphy, 24, 59– 73. B UFFETAUT , E. & S UTEETHORN , V. 1999. The dinosaur fauna of the Sao Khua Formation of Thailand and the beginning of the Cretaceous radiation of dinosaurs in Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 150, 13–23. B UFFETAUT , E. & S UTEETHORN , V. 2007. A sinraptorid theropod (Dinosauria: Saurischia) from the Phu Kradung Formation of northeastern Thailand. Bulletin de la Socie´te´ Ge´ologique de France, 178, 497–502. B UFFETAUT , E., S UTEETHORN , V., L E L OEUFF , J., K HANSUBHA , S., T ONG , H. & W ONGKO , K. 2005. The dinosaur fauna from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 575– 581. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J., K OBAYASHI , Y., H UANG , D. & L EE , Y. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19– 37. C APPETTA , H., B UFFETAUT , E. & S UTEETHORN , V. 1990. A new hybodont shark from the Lower Cretaceous of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 11, 659– 666. C ARTER , A. & B RISTOW , C. S. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau Basin, eastern Thailand. Basin Research, 15, 271–285. C AVIN , L., S UTEETHORN , V., B UFFETAUT , E., C LAUDE , J., C UNY , G., L E L OEUFF , J. & T ONG , H. 2007. The first sinamiid fish (Holostei, Halecomorpha) from southeast Asia (Early Cretaceous of Thailand). Journal of Vertebrate Paleontology, 27, 827 –837. C HKHIKVADZE , V. M. 1987. Sur la classification et les caracte`res de certaines tortues fossiles d’Asie rares et peu e´tudie´es. Studia Palaeocheloniologica, 2, 55–86. C HKHIKVADZE , V. M. 2000. On the systematic position of some extinct trionychids of North America and Asia. Trudy Tbilisskogo gosudarstvennogo pedagogicheskogo universiteta, 7, 199–213. C HOW , M.-C. 1954. Cretaceous turtles from Laiyang, Shantung. Acta Palaeontologica Sinica, 2, 395– 410. D ANILOV , I. G. & P ARHAM , J. F. 2006. A redescription of Plesiochelys tatsuensis from the Late Jurassic of China, with comments on the antiquity of the crown clade Cryptodira. Journal of Vertebrate Paleontology, 26, 573 –580. D ANILOV , I. G. & P ARHAM , J. F. 2007. The type series of ‘Sinemys’ wuerhoensis, a problematic turtle from the Lower Cretaceous of China, includes at least three taxa. Palaeontology, 50, 431– 444. D ANILOV , I. G. & S UKHANOV , V. B. 2006. A basal eucryptodiran turtle ‘Sinemys’ efremovi (¼ Wuguia efremovi) from the Early Cretaceous of China. Acta Palaeontologica Polonica, 51, 105–110.

150

H. TONG ET AL.

D ANILOV , I. G., A VERIANOV , A. O., S KUTCHAS , P. P. & R EZVYI , A. S. 2006. Kirgizemys (Testudines, ‘Macrobaenidae’): New material from the Lower Cretaceous of Buryatia (Russia) and taxonomic revision. Fossil Turtles Research, Vol. 1, Russian Journal of Herpetology, 13(Supplement), 46–62. D ANILOV , I. G., S YROMYATNIKOVA , E. V. & S UKHANOV , V. B. 2007. Turtles of the genus Shachemys from the Upper Cretaceous of Asia. In: R OZANOV , A. Y., L OPATIN , A. V. & P ARKHAEV , P. Y. (eds) Modern Paleontology: Classical and New Methods—Third all-Russian School 2006. Palaeontological Institute, Moscow, 59–72. DE B ROIN , F. 1984. Proganochelys ruchae n. sp., Che´lonien du Trias supe´rieur de Thaı¨lande. Studia Geologica Salamanticensia, Studia Palaeocheloniologica I, Special Volume, 1, 87– 97. DE B ROIN , F., I NGAVAT , R., S ATTAYARAK , N. & J ANVIER , P. 1982. Triassic turtle remains from northeastern Thailand. Journal of Vertebrate Paleontology, 2, 41– 46. DE L APPARENT DE B ROIN , F. 2004. A new Shachemydinae (Chelonii, Cryptodira) from the Lower Cretaceous of Laos: Preliminary data. Comptes Rendus Pale´vol, 3, 387– 396. D ONG , Z.-M. 1992. Dinosaurian Faunas of China. China Ocean Press, Springer, Berlin. E BERTH , D. A., B RINKMAN , D. B., C HEN , P.-J., Y UAN , F.-T., W U , X.-C., L I , G. & X IAN -S HEN , C. 2001. Sequence stratigraphy, paleoclimate patterns, and vertebrate fossil preservation in Jurassic–Cretaceous strata of the Juggar Basin, Xijiang Autonomous Region, People’s Republic of China. Canadian Journal of Earth Sciences, 38, 1627–1644. E NDO , R. & S HIKAMA , T. 1942. Mesozoic fauna in the Jehol mountainland, Manchoukuo. Bulletin of Central National Museum of Natural History, Manchoukuo, 3, 1– 23. F ERNANDEZ , V., C LAUDE , J., E SCARGUEL , G., B UFFETAUT , E. & S UTEETHORN , V. 2009. Biogeographical affinities of Jurassic and Cretaceous continental vertebrate assemblages from SE Asia. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 281– 296. G AFFNEY , E. S. & Y E , X. 1992. Dracochelys, a new Cryptodiran turtle from the Early Cretaceous of China. American Museum Novitates, 3048, 1–13. H IRAYAMA , R. 2000. Fossil turtles from the Kiwajima Formation (Tetori Group; Early Cretaceous) of KasekiKabe, Shiramine-mura, Ishikawa Prefecture, Central Japan. In: MATSUOKA , H. (ed.) Fossils of the Kuwajima ‘Kaseki-Kabe’ (Fossil bluff). Scientific report on a Neocomian (Early Cretaceous) fossil assemblage from the Kuwajima Formation, Tetori Group, Shiramine, Ishikawa, Japan. Shiramine-mura Education Board, Ishikawa, 75– 92. H IRAYAMA , R. 2002. Preliminary report of the fossil turtles from the Kitadani Formation (Early Cretaceous) of the Tetori Group of Katsuyama, Fukui Prefecture, Central Japan. Memoir of the Fukui Prefectural Dinosaur Museum, 1, 29–40.

H IRAYAMA , R. 2006. Review of fossil turtles of Japan. Fossils, 80, 47– 59. H IRAYAMA , R., B RINKMAN , D. & D ANILOV , I. 2000. Distribution and biogeography of non-marine Cretaceous turtles. Russian Journal of Herpetology, 7, 181–198. J I , Q., C HEN , W., W ANG , W.-L. ET AL . 2004. Mesozoic Jehol Biota of Western Liaoning, China. Geological Publishing House, Beijing. J I , S.-A. 1995. Reptilia. In: R EN , D., L U , L.-W., G UO , Z.-L. & J I , S.-A. (eds) Faunae and Stratigraphy of Jurassic–Cretaceous in Beijing and Adjacent Areas. Seismic Press, Beijing, 140–158. K UZNETZOV , V. V. 1976. A freshwater turtle from the Senonian of northeastern Cisaralia. Paleontological Journal, 1976, 498–500. L I , J.-L. & L IU , J. 1999. The skull of Manchurochelys laoxiensis (Testudines: Sinemydidae) from the Yixian Formation of Beipiao, Liaoning and phylogenetic position of this taxon. In: C HEN , P.-J. & J IN , F. (eds) Palaeoworld II, Jehol Biota. Press of University of Sciences and Technology of China, Hebei, 281–295. L IU , J. 2003. Turtles. In: Z AHNG , M.-M., C HEN , P.-J., W ANG , Y.-Q., W ANG , Y. & M IAO , D.-S. (eds) The Jehol Biota. Shanghai Scintific and Technical Publishers, Shanghai, 86–87. M A , Q.-H., L IN , Q.-B., Y E , C.-H. & S HENG , Y.-B. 1982. Division and correlation of the Xinminbao Group and the Chijinbao Formation in western Jiuquan Basin. Journal of Stratigraphy, 6, 112–120. M AISCH , M. W., M ATZKE , A. T. & S UN , G. 2003. A new sinemydid turtle (Reptilia: Testudines) from the Lower Cretaceous of the Juggar Basin (NW China). Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 2003, 705–722. M ATZKE , A. T. & M AISCH , M. W. 2004. New information and specimens of Wuguia hutubeiensis (Reptilia: Testudines) from the Lower Cretaceous Tugulu Group of the southern Juggar Basin (NW China). Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 2004, 473– 495. M ATZKE , A. T., M AISCH , M. W., G E , S. & P FRETZSCHNER , H.-U. 2004a. A new xinjiangchelyid turtle (Testudines, Eucryptodira) from the Jurassic Qigu Formation of the southern Junggar Basin, Xinjiang, north-west China. Palaeontology, 47, 1211– 1266. M ATZKE , A. T., M AISCH , M. W., P FRETZSCHNER , H.-U., S UN , G. & S TO¨ HR , H. 2004b. A new basal sinemydid turtle (Reptilia: Testudines) from the lower Cretaceous Tugulu Group of the Junggar Basin (NW China). Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 2004, 151– 167. M EYLAN , P. A. & G AFFNEY , E. S. 1989. The skeletal morphology of the Cretaceous Cryptodiran turtle, Adocus, and the relationships of the Trionychoidea. American Museum Novitates, 2941, 1 –60. M EYLAN , P. A. & G AFFNEY , E. S. 1992. Sinaspideretes is not the oldest trionychid turtle. Journal of Vertebrate Paleontology, 12, 257– 259. M OURET , C., H EGGEMANN , H., G OUADAIN , J. & K RISADASIMA , S. 1993. Geological history of the siliciclastic Mesozoic strata of the Khorat Group in the Phu Phan Range area, northeastern Thailand.

MESOZOIC TURTLES OF NE THAILAND In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology. Chiang Mai University, Chiang Mai, 23– 49. N ARMANDAKH , P. 1985. A new species of Adocus from the Late Cretaceous of Mongolia. Paleontologicheskii Zhurnal, 1985, 85–93. N ESSOV , L. A. 1977a. A new genus of pitted-shelled turtle from the Upper Cretaceous of Karakalpakia. Paleontological Journal, 1, 96– 107. N ESSOV , L. A. 1977b. Skull morphology of Early Cretaceous turtle belonging to the family Adocidae. Trudy Zoologicheskogo Instituta AN SSSR, 74, 74–79. N ESSOV , L. A. 1981. On turtles of the family Dermademydidae from the Cretaceous of the Amur river basin, and some other rare discoveries of the remains of ancient turtles of Asia. In: B ORKIN , L. J. (ed.) Herpetological Investigations in Siberia and the Far East. Zoological Institute of the Russian Academy of Sciences, Leningrad, 69– 73. N ESSOV , L. A. 1984. Data on Late Mesozoic turtles from the USSR. Studia Geologica Salamanticensia, Volumen Especial 1, Studia Palaeocheloniologica, 1, 215– 224. N ESSOV , L. A. 1986. Some Late Mesozoic and Paleocene turtles of Soviet Middle Asia. Studia Geologica Salamanticensia, Volumen Especial 2, Studia Palaeocheloniologica, 2, 7– 22. N ESSOV , L. A. 1987. On some Mesozoic turtles of the Soviet Union, Mongolia and China, with comments on systematics. Studia Geologica Salamanticensia, Volumen Especial 2, Studia Palaeocheloniologica, 2, 87– 102. N ESSOV , L. A. & K HOZATSKII , L. I. 1977. Freshwater turtle from the Early Cretaceous of Fergana. Ezhegodnik Vsesoyuznogo Paleontologicheskogo Obshchestva, 20, 248–262. N ESSOV , L. A. & K RASOVSKAYA , T. B. 1984. Transformations in the composition of turtle assemblages in the Cretaceous of Middle Asia. Vestnik Leningradeskogo Universiteta, 3, 15–25. N ESSOV , L. A. & V ERZILIN , N. N. 1981. Remains of turtles from Aptian–Albian deposits of the Transaltai Gobi in Mongolia and conditions of their burial. In: BARSBOLD , R. ET AL . (eds) Fossil Vertebrates of Mongolia. Trudy Sovmestnoi Sovetsko-Mongol’skoi Paleotologicheskoi Ekspeditsii, Moscow, 15, 12– 26. P ENG , J.-H. & B RINKMAN , D. 1993. New material of Xinjiangchelys (Reptilia : Testudines) from the Late Jurassic Qigu Formation (Shishugou Group) of the Pingfengshan locality, Junggar Basin, Xinjiang. Canadian Journal of Earth Sciences, 30, 2013–2026. P ENG , G.-Z., Y E , Y., G AO , Y.-H., S HU , C.-K. & J IANG , S. (eds) 2005. Jurassic Dinosaur Fauna in Zhigong. Shichuan People’s Press, Chengdu. R ACEY , A. & G OODALL , J. G. S. 2009. Palynology and stratigraphy of the Mesozoic Khorat Group red bed sequences from Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 67–81. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat

151

Group, NE Thailand. 1. Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5 –40. S UKHANOV , V. B. 2000. Mesozoic turtles of Middle and Central Asia. In: B ENTON , M. J., S HISHKIN , M. A., U NWIN , D. M. & K UROCHKIN , E. N. (eds) The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge, 309–367. S UKHANOV , V. B. & N ARMANDAKH , P. 2006. New taxa of Mesozoic turtles from Mongolia. Russian Journal of Herpetology, 13(Supplement), 119–127. T ANG , F., L UO , Z.-X., Z HOU , Z.-H., Y OU , H.-L., G EORGI , J.-A., T ANG , Z.-L. & W ANG , X.-Z. 2001. Biostratigraphy and palaeoenvironment of the dinosaur-bearing sediments in Lower Cretaceous of Mazongshan area, Gansu Province, China. Cretaceous Research, 22, 115–129. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2002. Middle Jurassic turtles from southern Thailand. Geological Magazine, 139, 687–697. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2003a. Mesozoic turtles of Thailand. Mahasarakham University Journal, 22(Special Issue 2003), 41–48. T ONG , H., B UFFETAUT , E., S UTEETHORN , V. & S RISUK , P. 2003b. Turtle remains from the Sao Khua Formation (Early Cretaceous) of the Khorat Plateau, northeastern Thailand. Abstract. In: DANILOV , I. G., CHEREPANOV , G. O., HIRAYAMA , R. & PARHAM , J. F. (eds) Symposium on Turtle Origin, Evolution and Systematics. Zoological Institute of the Russian Academy of Sciences, Saint Petersburg, 30. T ONG , H., B UFFETAUT , E., S UTEETHORN , V. & S RISUK , P. 2004a. First carettochelyid turtle from the Lower Cretaceous of Thailand (abstract). In: DOSTAL , O., GREGORAVA , R. & IVANOV , M. (eds) 2nd EAVP Meeting. Moravian Museum, Brno, 44. T ONG , H., J I , S.-A. & J I , Q. 2004b. Ordosemys (Testudines: Cryptodira) from the Yixian Formation of Liaoning Province, Northeastern China: New specimens and systematic revision. American Museum Novitates, 3438, 1 –20. T ONG , H., S UTEETHORN , V., C LAUDE , J., B UFFETAUT , E. & J INTASAKUL , P. 2005. The turtle fauna from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 610–614. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2006a. Isanemys, a new adocid turtle from the Sao Khua Formation (Early Cretaceous) of the Khorat Plateau, northeastern Thailand. Russian Journal of Herpetology, 13(Supplement), 128–137. T ONG , H., C LAUDE , J., B UFFETAUT , E., S UTEETHORN , V., N AKSRI , W. & S UWANNA , C. 2006b. Fossil turtles of Thailand: an updated review. In: L U¨ , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y.-N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 183– 194. T ONG , H., C LAUDE , J. ET AL . 2009. Basilochelys macrobios n. gen. and n. sp., a large cryptodiran turtle from the Phu Kradung Formation (latest Jurassic–earliest Cretaceous) of the Khorat Plateau, NE Thailand.

152

H. TONG ET AL.

In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 151–170. W IMAN , C. 1930. Fossile Schildkro¨ken aus China. Palaeontologia Sinica, Series C, 6, 5 –53. Y E , X.-K. 1963. Fossil turtles of China. Palaeontologia Sinica, 150, 1– 112. Y E , X.-K. 1965. New materials of fossil turtles of Inner Mongolia. Vertebrata Palasiatica, 9, 47– 48. Y E , X.-K. 1973a. Discovery of Plesiochelys from Upper Lufeng Series, Oshan, Yunnan and its stratigraphical significance. Vertebrata Palasiatica, 11, 160–164. Y E , X.-K. 1973b. Fossil turtles from Wuerho. Reports of the palaeontological expedition to the Xinjiang 2.

The Pterosaurian fauna from Wuerho. Academia Sinica, Institute of Vertebrate Palaeontology and Palaeoanthropology, Memoirs, 11, 8 –11. Y E , X.-K. 1974. A new fossil Trionyx from Fuchien. Vertebrata Palasiatica, 12, 190–192. Y E , X.-K. 1986. A Jurassic turtle from Junggar, Xinhjiang. Vertebrata Palasiatica, 24, 171–181. Y E , X.-K. 1994. Fossil and Recent Turtles of China. Science Press, Beijing. Y E , X.-K. & F ANG , Q.-R. 1982. A new species of fossil Plesiochelys from Jingyuan, Sichuan. Acta Herpetologica Sinica, 1, 63– 68. Y OUNG , C. C. & C HOW , M.-C. 1953. New fossil reptiles from Szechuan, China. Acta Scientica Sinica, 2, 216–243.

Basilochelys macrobios n. gen. and n. sp., a large cryptodiran turtle from the Phu Kradung Formation (latest Jurassic –earliest Cretaceous) of the Khorat Plateau, NE Thailand HAIYAN TONG1 *, JULIEN CLAUDE2, WILAILUCK NAKSRI3, VARAVUDH SUTEETHORN4, ERIC BUFFETAUT5, SASIDHORN KHANSUBHA4, KAMONRAK WONGKO4 & PHISIT YUANGDETKLA4 1

16 cour du Lie´gat, 75013 Paris, France

2

ISE-M, UMR 5554 CNRS, Universite´ de Montpellier 2, 34095 Montpellier, France 3

Sirindhorn Museum, Sahatsakhan, Kalasin 46140, Thailand

4

Bureau of Fossil Research and Museum, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand

5

CNRS (UMR 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure), 24 rue Lhomond, 75005 Paris, France *Corresponding author (e-mail: [email protected]) Abstract: A large cryptodiran turtle, Basilochelys macrobios n. gen. n. sp. is described from the latest Jurassic–earliest Cretaceous Phu Kradung Formation of NE Thailand, on the basis of skull, shell and other postcranial elements. Basilochelys presents a combination of primitive and derived characters. The derived characters include sculptured skull roof and shell surface; deeply embedded canalis caroticus internus; foramen posterius canalis carotici interni completely surrounded by pterygoid; neural formula of 6 . 4 , 6 , 6 , 6 , 6; anteroposteriorly expanded eleventh and twelfth marginal scutes extending onto the suprapygal and costal plates; narrow vertebral scutes; plastron sutured to the carapace, with large and wide anterior and posterior lobes, long and narrow bridge, very narrow axillary and inguinal notch; wide entoplastron; humeropectoral sulcus located on the posterior part of the entoplastron; anal notch absent. This taxon is placed in Trionychoidae and considered as the most basal member of that group.

The latest Jurassic to mid-Cretaceous non-marine beds of the Khorat Group, in NE Thailand, are rich in vertebrate remains, including turtles. Adocid and carettochelyid turtles have been described from the Early Cretaceous Sao Khua Formation (Tong et al. 2003, 2004a, 2006a) and the mid-Cretaceous Khok Kruat Formation (Tong et al. 2005); however, the turtles from the underlying Phu Kradung Formation are still poorly known (Tong et al. 2006b). Although turtle remains have been collected from several localities of that formation for several years, they were too fragmentary for an accurate systematic study. Recently, complete shells and a partial skull of a large cryptodiran turtle have been discovered in the Phu Kradung Formation at the Kham Phok locality, Mukdahan Province. The purpose of the present paper is to describe these new turtle specimens and to discuss their systematic position. The specimens are housed in the Sirindhorn Museum, Sahat Sakhan, Kalasin Province, Thailand.

Geological setting and taphonomy Two nearly complete large turtle shells were discovered in summer 2004 near the village of Kham Phok, Mukdahan Province (Fig. 1). When discovered, both shells were lying upside down in a siltstone level with sandy to micro-conglomerate lenses. A braincase, shell fragments and other postcranial elements were collected within a few square metres around one shell (MD8-2) and probably belong to a single individual. The pelvic girdle was found inside both shells during preparation, and an incomplete femur and a cervical vertebra were found inside MD8-2. This suggests that the soft parts were at least partially decomposed before burial and that these turtles were not much transported after decomposition. Abundant shark teeth and one shark spine have been found inside the shell of MD8-1. A theropod tibia was found beneath MD8-1 (Buffetaut & Suteethorn 2007). As both turtles were found upside down and are of rather large size, we

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 153–173. DOI: 10.1144/SP315.12 0305-8719/09/$15.00 # The Geological Society of London 2009.

154

H. TONG ET AL.

Fig. 1. Map showing location of Kham Phok locality, and outcrops of the Phu Kradung Formation (after Buffetaut & Suteethorn 2007).

suppose that getting into that position may have caused their death. Although the shells were found articulated, cracks (from 1 to 5 mm) are present in both. These cracks are not filled by calcite but probably result from the pressure of overlying sediment. Other material studied here includes abundant shell fragments from Dan Luang and Huai Pai, and a partial carapace from the bed of Huai Sai river, in Mukdahan Province. All these specimens come from the upper levels of the Phu Kradung Formation in the eastern part of NE Thailand. The Phu Kradung Formation consists of fluvial sandstones, siltstones and mudstones and is dated as either Late Jurassic or more probably Early Cretaceous on the basis of palynology (Racey et al. 1996) and detrital zircon thermochronology (Carter & Bristow 2003), whereas evidence from fossil vertebrates supports a Late Jurassic age (Buffetaut & Suteethorn 2007). Beside turtles, the vertebrates from the Phu Kradung Formation include freshwater sharks, actinopterygian fishes, lungfishes, temnospondyl amphibians, crocodilians and dinosaurs (Buffetaut et al. 2006).

Systematic palaeontology Megaorder Cryptodira Cope Parvorder Eucryptodira Gaffney Epifamily Trionychoidae Fitzinger ( fide Meylan & Gaffney 1989) Genus Basilochelys new genus Etymology. Basileus: Greek, king; chelys: Greek, turtle. In honour of His Majesty the King of Thailand.

Diagnosis. Cryptodiran turtle of large size, with shell length reaching 900 mm. Characterized by the combination of the following features: sculptured skull roof and shell surface; skull with large vomer separating completely the palatines, long pterygoid midline suture between the vomer and the basisphenoid, foramen posterius canalis caroticus laterale exposed on ventral surface, large foramen caroticum basisphenoidale, deeply embedded canalis caroticus internus, and foramen posterius canalis carotici interni completely surrounded by pterygoid. Shell low, with smooth carapace margin and without keel; neural formula 6 . 4 , 6 , 6 , 6 , 6. . .; narrow vertebral scutes, second marginal scute boot-shaped, contacting the first vertebral scute; eleventh and twelfth marginals anteroposteriorly expanded, extending onto the costal and suprapygal plates. Plastron sutured to the carapace, with large and wide anterior and posterior lobes, long and narrow bridge, very narrow axillary and inguinal notch; wider than long entoplastron; humeropectoral sulcus located on the posterior part of the entoplastron; anal notch absent. Type species. Basilochelys macrobios sp. nov. Etymology. Macrobios: Greek, long life. In honour of His Majesty King Rama IX’s eightieth birthday. Holotype. A nearly complete carapace articulated with a partial plastron, pelvic girdle and a cervical vertebra (MD8-2, collection of the Sirindhorn Museum, Phu Kum Khao, Sahatsakhan, Kalasin Province, Thailand).

NEW GIANT MESOZOIC TURTLE FROM THAILAND

155

Table 1. Shell measurements of Basilochelys macrobios n. gen. n. sp. from the Late Jurassic– Early Cretaceous of Phu Kradung Formation, NE Thailand (in mm)

Carapace Length Width Plastron Length Width Width of anterior lobe Length of bridge Length of posterior lobe Width of posterior lobe

MD8-1

MD8-2

MD4-1

900 790

9001 7201

8201 –

7201 6001 4501 295 230 350

– – – 260 265 330

– – – – – –

1

Estimated.

Hypodigm. Kham Phok locality: an incomplete skull (MD8-3); a complete carapace articulated with partial plastron, and pelvic girdle (MD8-1); an isolated cervical (MD8-4); a humerus (MD8-5), an incomplete femur (MD8-6); a distal phalanx (MD8-7), and other limb bone fragments and shell fragments. Huai Sai locality: an incomplete carapace (MD4-1). Dan Luang locality: a partial shell with anterior lobe of plastron, nuchal, first costal and two peripherals (MD3-1), and numerous fragments of plates. Huai Pai locality: an incomplete anterior lobe of plastron (MD5-1), an isolated entoplastron (unnumbered, collection of Buddhabutr Temple, Mukdahan Province, Thailand) and other shell fragments (collection of the Sirindhorn Museum, Phu Kum Khao, Sahatsakhan, Kalasin Province, Thailand). Type locality. Kham Phok, Mukdahan Province, Khorat Plateau, NE Thailand. Horizon. Phu Kradung Formation, Jurassic –basal Cretaceous.

terminal

Species diagnosis. As for genus, only species. Measurements. See Table 1.

Description and comparisons Skull (MD8-3; Fig. 2) Preservation. Only the braincase is preserved, including both pterygoids, epipterygoids, prootics, opisthotics, basisphenoid and basioccipital; and incomplete quadrates, palatines, vomer, parietals, supraoccipital and exoccipitals. General appearance. Most of the skull roof is missing; the shape and size of the temporal emargination cannot be determined. However, some free margin of the skull roof is preserved on the right

side, indicating that the temporal emargination was probably fairly large, with the foramen stapedio-temporale exposed in dorsal view, unlike the fully covered skull roof of Nanhsiungchelys (Yeh 1966). However, judging by the width of the preserved posterior part of the parietal, the temporal emargination of the Kham Phok skull is less developed than that of Adocus sp. (Meylan & Gaffney 1989) and Zangerlia neimongolensis (Brinkman & Peng 1996). The morphology of the ventral surface of the skull is somewhat similar to that of Adocus sp. The pterygoid, basisphenoid and basioccipital form a flat surface; and laterally the pterygoid and the quadrate together form a large and deep anteroposteriorly directed, triangular depression (see pterygoid). Parietal. The parietal is the only preserved skull roof element and only the posteromedial part of both parietals is preserved, with a damaged and worn surface. However, a small portion of the dorsal surface of the parietal is better preserved, showing that the skull roof was sculptured. In lateral view, the posterior portion of the processus inferior parietalis is preserved. It contributes to the medial margin of the foramen nervi trigemini, and contacts the pterygoid anteroventrally and the epipterygoid ventrally anterior to the foramen nervi trigemini. The parietal contacts the prootic ventrally, and the supraoccipital posteriorly, posterior to the foramen nevi trigemini. In anterior view, the parietal sends a lateral process that underlies the processus trochlearis oticum, but it does not contribute to the processus on the dorsal surface. This is different from Adocus sp. in which the parietal forms one-third of the processus trochlearis oticum. Vomer. Only the posterior end of the vomer is preserved. The vomer is a single bone that lies

156

H. TONG ET AL.

Fig. 2. Skull of Basilochelys macrobios n. gen. n. sp. (MD8-3) from the Late Jurassic– Early Cretaceous Phu Kradung Formation, Kham Phok locality, NE Thailand. Above, dorsal view; middle, ventral view; below, right lateral view. Scale bar: 50 mm. bo, basioccipital; bs, basisphenoid; ept, epipterygoid; fcb, foramen caroticum basisphenoidale; fcl, foramen posterius canalis caroticus laterale; fnt, foramen nervi trigemini; fpcci, foramen posterius calalis carotici interni; fst, foramen stapedio-temporale; op, opisthotic; pa, parietal; pal, palatine; pr, prootic; pt, pterygoid; pto, processus trochlearis oticum; qu, quadrate; vo, vomer.

between the palatines. The posterior part of the vomer is similar to that of Adocus sp. in that it is relatively wide with the lateral margins converging anteriorly. In addition to the palatine contact, the vomer contacts the pterygoid posteriorly. There is a pair of tubercles on the ventral surface of the vomer, close to the vomer–pterygoid suture.

Palatine. Most of both palatines are missing; only their posterior ends are preserved. The posteromedial corner of the triturating surface, preserved on the right side, is formed by the palatine. The medial limit of the triturating surface is close to the midline, which would indicate a rather wide triturating surface, at least as wide as in Adocus sp.,

NEW GIANT MESOZOIC TURTLE FROM THAILAND

rather than a narrow one as in Dracochelys (Gaffney & Ye 1992). Although the foramen posterius palatinus is not preserved on either side, it is unlikely that it was a large foramen as seen in Sinemys (Brinkman & Peng 1993b) and Dracochelys. The palatine contacts the pterygoid posteriorly and the vomer medially. Pterygoid. Both pterygoids are nearly complete and well preserved, lacking only their anterolateral part. In ventral view, the pterygoid, together with the basisphenoid and the basioccipital, forms a flat palatal surface. The anterolateral margin of the pterygoid is damaged; the processus pterygoideus externus is not preserved. However, the lateral margin of the right pterygoid presents a slight notch, which indicates that the base of the process is there, anterior to this notch, but its shape and size cannot be determined. Posterolateral to the flat surface, a deep concavity is formed by the pterygoid medially and the quadrate laterally. The concavity is narrow anteriorly and becomes wider and deeper posteriorly. It is limited laterally by the quadrate process of the pterygoid and the quadrate and medially by a distinct anteroposteriorly directed ridge on the pterygoid, which extends posteriorly by a thin and nearly vertical flange on the pterygoid. The concavity is open posteriorly. This structure is very similar to that of Adocus sp. described by Meylan & Gaffney (1989). The foramen caroticum basisphenoidale is a long oval opening lying on the anterior portion of the basisphenoid–pterygoid suture. It measures 5 mm in length and 1.5 mm in width on the right side and 5.5 by 1.5 mm on the left. The canalis carotici interni is deeply embedded; the depth of the canal at the foramen caroticum basisphenoidale is about 2 mm. The macrobaenids –sinemydids, such as Dracochelys, Ordosemys (Brinkman & Wu 1999; Tong et al. 2004b) and Sinemys, differ from this new material in that the canal is covered by a thin bone. Anterolateral to the foramen caroticum basisphenoidale, there is a pair of posteriorly facing foramina, which are interpreted here as the foramen posterius canalis caroticus laterale (Sukhanov 2000), for the palatine artery (Gaffney 1979). The posterior part of the canalis caroticus lateralis is in an open, wide and shallow depression. The position and morphology of the foramen caroticum basisphenoidale in MD8-3 is similar to that of Adocus, whereas the canalis caroticus lateralis is not exposed on the ventral surface in Adocus. The ventrally exposed foramen posterius canalis caroticus laterale, the canalis caroticus lateralis and foramen caroticum basisphenoidale are present in Kallokibotion (Gaffney & Meylan 1992),

157

xinjiangchelyids, sinemydids– macrobaenids, Mongolochelys (Khosatzky 1997) and Chubutemys (Gaffney et al. 2007). However, the condition in xinjiangchelyids, Kallokibotion and Mongolochelys is more primitive than in the Kham Phok skull in that the canalis caroticus internus is not covered by bone ventrally. Meiolania platyceps is considered also to have a canalis caroticus lateralis that is not completely enclosed in bone, but the canal is confluent with the intrapterygoid slit, an autapomorphic structure of Meiolaniidae (Gaffney 1983). The open morphology of the arterial area of the Kham Phok skull seems to be more similar to that of Hangaiemys (Sukhanov 2000) and Ordosemys (Brinkman & Wu 1999) than to that of Sinemys, Kirgizemys and Judithemys (Brinkman & Peng 1993b; Brinkman & Wu 1999; Parham & Hutchison 2003; Tong et al. 2004b; Danilov et al. 2006). In these forms, the arterial areas are more restricted by bony margins. In Dracochelys (Gaffney & Ye 1992) and Chubutemys, the foramen posterius canalis caroticus laterale and foramen caroticum basisphenoidale are more distant, which approaches the condition of Kallokibotion (Gaffney & Meylan 1992; Gaffney 1996). Zangerlia neimongolensis has a large depression that contains the canalis caroticus lateralis and canalis caroticus internus (Brinkman & Peng 1996), which seems comparable with the condition in the Kham Phok skull, but the poor preservation of this area prevents any detailed comparison. The canalis caroticus internus extends posteriorly within the pterygoid. The foramen posterius canalis carotici interni is formed entirely by the pterygoid and lies near the posterior margin of the pterygoid, under a pterygoid flange, as in Adocus. On the ventral surface, the pterygoid contacts the vomer anteriorly and the palatine anterolaterally. Between the vomer and the basisphenoid, the left and right pterygoids meet on the midline by a long suture. Posteromedially, the pterygoid contacts the basisphenoid and posterolaterally the quadrate. The pterygoid– basisphenoid suture is clearly visible in its anterior portion, whereas its posterior part is in an area of broken bone. In lateral view, the pterygoid is visible on both sides. The anterior contact with the palatine is not visible. Together with the parietal, the prootic, the quadrate and the epipterygoid, the pterygoid forms the foramen nervi trigemini, which is a large and oval foramen. The pterygoid contacts the parietal anterodorsally, the epipterygoid dorsally and the quadrate posteriorly. Epipterygoid. Both epipterygoids are complete. The epipterygoid is an elongate bone, forming most of the anterior margin of the foramen nervi trigemini. Anteriorly, ventrally and posterolaterally, it

158

H. TONG ET AL.

is surrounded by the pterygoid; dorsally, it contacts the parietal. Quadrate. Incomplete quadrates are preserved on both sides, whereas the cavum tympani and the processus articularis are missing on both sides. Ventrally, the quadrate sends an anteromedial process to meet the pterygoid lateral to the pterygoid concavity (see pterygoid). It contacts the pterygoid anteromedially and medially. Anterodorsally, the quadrate contributes to the lateral part of the processus trochlearis oticum (see prootic). It contacts the pterygoid anteroventrally, the prootic anteromedially, and the opisthotic posteromedially. Basisphenoid. The basisphenoid is complete; only its ventral surface is exposed. The basisphenoid has a triangular and flat ventral exposure. The large foramen caroticum basisphenoidale lies on the basisphenoid–pterygoid suture (see pterygoid). A pair of small and shallow concavities can be seen on the posterior part of the basisphenoid; a similar structure is present in some sinemydids– macrobaenids, such as Sinemys, Ordosemys, Kirgizemys and Judithemys (Parham & Hutchison 2003). The basisphenoid contacts the pterygoid anteriorly and laterally, and the basioccipital posteriorly. Basioccipital. The basioccipital is nearly complete; however, the occipital condyle is missing. The basioccipital has a rectangular and slightly concave ventral surface. The tuberculum basioccipitale is a huge and dorsoventrally flattened process. The occipital condyle is not preserved. The basioccipital contacts the basisphenoid anteriorly by a slightly anteriorly convex suture, the pterygoid laterally and the exoccipital dorsally.

Exoccipital. Both exoccipitals are damaged at their posterior end, lacking the occipital condyle. One foramen nervi hypoglossi is visible on the posterolateral surface of the exoccipital. The exoccipital contacts the basioccipital ventrally, the opisthotic laterally, the pterygoid ventrolaterally and the supraoccipital dorsally. Prootic. The right prootic is complete; the left one is damaged, missing the lateral end. The prootic has a wide dorsal exposure, differing from the narrow prootic of Ferganemys (Nessov 1977), Adocus and a primitive adocid skull from the Late Cretaceous of Kizylkum, Uzbekistan (Danilov & Parham 2005). The prootic forms most of the processus trochlearis oticum; the quadrate contributes to its lateral portion. There is no parietal contribution to the dorsal surface of the process (see parietal). The processus trochlearis oticum is a thick, blunt and forward directed ridge. Its medial end is not well defined, but continues smoothly to the lateral surface of the braincase, whereas its lateral portion is more developed, forming a large swelling at the lateral end. The whole process, however, is not as developed as in Adocus sp., but more similar to that of Zangerlia neimongolensis (Brinkman & Peng 1996). Anteriorly, the prootic contributes to the dorsal margin of the foramen nervi trigemini (see pterygoid). The contacts with the parietal medially, the quadrate laterally and the pterygoid ventrally are clearly visible, whereas the posterior contact with the opisthotic is not clearly discernible. There are three depressions on the dorsal surface of the prootic and opisthotic; the foramen stapediotemporale is located in the most anterior one, on or near the prootic– opisthotic suture. Although the depression is fairly large, the foramen itself is much smaller.

Fig. 3. Shell of Basilochelys macrobios n. gen. n. sp. (MD8-1) from the Late Jurassic–Early Cretaceous Phu Kradung Formation, Kham Phok locality, NE Thailand. (a) Dorsal view; (b) ventral view; (c) carapace in ventral view, plastron removed. Scale bar: 200 mm.

NEW GIANT MESOZOIC TURTLE FROM THAILAND

159

Fig. 4. Shell of Basilochelys macrobios n. gen. n. sp. (Type, MD8-2) from the Late Jurassic– Early Cretaceous Phu Kradung Formation, Kham Phok locality, NE Thailand. Above, dorsal view; below, ventral view. Scale bar: 200 mm.

Opisthotic. Both opisthotics are preserved but damaged. The sutures of the opisthotic are not clearly visible, and the foramen jugulare posterius and fenestra postotica are covered by matrix.

Shell (Figs 3 – 6, 8) Shell surface ornamentation. A well-preserved shell surface is visible on the carapace from Huai Sai (MD4-1) and some isolated plates from Huai Pai and Dan Luang. In these specimens, the shell surface is covered with a distinct sculpture consisting of raised irregular ridges and tubercles, which tend to arrange themselves into a parallel pattern. The ridges are anteroposteriorly directed on the

middle of the carapace, on the neurals and at least the medial part of the costals, whereas on the anterior part of the carapace they are anterolaterally directed (Fig. 5). This pattern of ornamentation is reminiscent of that of Anomalochelys, a nanhsiungchelyid from the Late Cretaceous of Japan (Hirayama et al. 2001). It differs from the pitted pattern of adocids and ‘pock-mark’ sculpturing of Basilemys and Nanhsiungchelys (Meylan & Gaffney 1989). On the isolated entoplastron from Huai Pai (Fig. 6c), there are light posterolaterally directed ridges radiating from the midline. In the specimens from Kham Phok, the shell surface of MD8-1 and MD8-2 is worn and covered by a layer of ferruginous matrix that is difficult to remove.

160

H. TONG ET AL.

Fig. 5. Partial carapace of Basilochelys macrobios n. gen. n. sp. (MD4-1) from the Late Jurassic–Early Cretaceous Phu Kradung Formation, Huai Sai locality, NE Thailand. Scale bar: 200 mm.

However, the whole shell surface is rough. The vermiculated ornamentation, although not well preserved, is discernible on the anterior part of the carapace; the posterior peripherals have a rather smooth surface. Carapace (Figs 3 –5). The shell of MD8-1 is slightly dorsoventrally crushed. The carapace has a rounded outline with the length only slightly greater than the width. There is a wide and shallow ‘notch’ on the right posterolateral margin of this specimen that we interpret as an anomaly. MD8-2 is more elongated than MD8-1 (see Table 1) and slightly deformed by dorsolateral crushing. The shells are low with a smooth margin

throughout and without keels or knobs on the carapace. There is no shallow central depression along the midline of the carapace. Such a depression is observed in macrobaenids –sinemydids (Sinemys (Brinkman & Peng 1993b), Ordosemys (Brinkman & Peng 1993a; Tong et al. 2004b; Danilov & Parham 2007), Wuguia (Matzke et al. 2004b; Danilov & Sukhanov 2006)), xinjiangchelyids (Peng & Brinkman 1993; Matzke et al. 2004a), and Siamochelys from the Middle Jurassic of southern Thailand (Tong et al. 2002). MD8-1 and MD4-1 have a shallow nuchal emargination, whereas the anterior rim of the carapace of MD8-2 is damaged. The anterior part of the lateral margin of the shell is thickened, rounded and

NEW GIANT MESOZOIC TURTLE FROM THAILAND

161

Fig. 6. Basilochelys macrobios n. gen. n. sp. from the Late Jurassic– Early Cretaceous Phu Kradung Formation. (a, b) partial plastron (MD5-1) from Dan Luang (scale bar: 50 mm); (c) entoplastron from Huai Pai (scale bar: 50 mm); (d) cervical vertebra (MD8-4) from Kham Phok (scale bar: 20 mm).

Fig. 7. Pelvis of Basilochelys macrobios n. gen. n. sp. (MD8-1) from the Late Jurassic– Early Cretaceous Phu Kradung Formation, Kham Phok locality, NE Thailand. Dorsal view. Scale bar: 50 mm.

upturned, forming a wide and shallow gutter along the margin, which is clearly visible in MD8-1, MD8-2 and MD4-1. The posterior peripherals, from the seventh to the eleventh, as well as the pygal, are flared and have a thin and rather sharp free margin, except the posterior peripherals on the right side of MD8-1, which have a rather rounded margin because of the anomaly. The nuchal is certainly complete in MD8-1, but sutures are not visible. It is incomplete in MD8-2 and MD4-1. The nuchal of MD8-2 is wide and very short, being much wider than long and shorter than the first neural. The nuchal of MD4-1 is wide too, but longer than the first neural. The neural series is complete in MD8-2, but only the outlines of the first to sixth neurals are distinguishable; the sutures of the posterior neurals, from the posterior part of the sixth neural to the suprapygal, are not visible. In MD8-1, neither inter-neural sutures, nor neural–costal sutures are visible, so the neurals and costals may have been completely fused together. An incomplete neural series, from the first to fifth neural, is preserved in

162

H. TONG ET AL.

Fig. 8. Reconstruction of carapace and plastron of Basilochelys macrobios n. gen. n. sp. from the Late Jurassic–Early Cretaceous Phu Kradung Formation, NE Thailand, based on MD4-1, MD8-2, MD8-1 and MD5-1.

MD4-1, the fifth neural being incomplete. The neural morphology is similar in MD8-2 and MD4-1: the first neural is six-sided with short posterolateral sides. The second neural is four-sided and clearly shorter than the first. The third to sixth neurals are six-sided with short anterolateral sides. All these neurals are longer than wide, the third and the fourth neurals are the longest, and clearly longer than the first neural in both MD4-1 and MD8-2. The length of the neurals decreases from the third to the sixth. This neural pattern of 6 . 4 , 6 , 6 , 6 , 6 is shared by adocids and nanhsiungchelyids (Meylan & Gaffney 1989; Hirayama et al. 2001). Suprapygal(s) are preserved in MD8-1 and MD8-2, but sutures are not visible in MD8-1. In MD8-2, at least one large posterior suprapygal is distinguishable. The pygal plate, clearly visible in MD8-2, is slightly wider than long, with the anterior and posterior rim convex backward and the lateral borders slightly convergent forward. Eight costal plates are present in MD8-2. The first costal is short, and as long as the second costal plate. The morphology of the first costal plate of MD4-1 and MD3-1 is similar to that of MD8-2. Other costal plates have parallel anterior and posterior sides. The first costal plate reaches the anterior margin of the third peripheral, and the eighth costal plate contacts the ninth and tenth peripheral plates.

Eleven pairs of peripherals are present in MD8-2. The first peripheral, preserved on the right side, is small and as long as wide, with the anterior free margin longer than the posterior rim. The size of the peripherals and their mediolateral width increases from the front to the back. The eighth to eleventh peripherals are mediolaterally expanded. On the inner side of the carapace, the first costal plate bears a long and sharp transverse and anteriorly concave ridge, extending from the first dorsal rib head to the lateral margin of the plate. A strong, long stick-shaped first thoracic rib extends along this ridge laterally to the third peripheral plate. The contact is, however, not clearly preserved. A similar structure is present in Ordosemys leios (Brinkman & Peng 1993a). In Xinjiangchelys latimarginalis the first thoracic rib is much shorter, extending only halfway across the first costal (Peng & Brinkman 1993). All dorsal rib heads are large and flattened. The scute sulci are clearly visible in MD4-1 (Fig. 5), whereas in MD8-1 they are discernible only on the posterior end of the shell. In MD8-2, they are preserved near the margin, on the anterior-right side and along the left and posterior margin (Fig. 4). The right half of the cervical scute is preserved on MD4-1. When it is reconstructed, the cervical scute is large and wider than long, with the posterior sulcus strongly convex anteriorly.

NEW GIANT MESOZOIC TURTLE FROM THAILAND

The right half of the first to third vertebral scutes is preserved in MD4-1. The first vertebral is roughly bell-shaped, with the anterior sulcus strongly convex anteriorly and posterolaterally convergent lateral sulci. When reconstructed, it is about as long as wide. The sulcus between the first and the second vertebral scutes crosses the first neural plate. The second and the third vertebral scutes are much longer than wide, as in some nanhsiungchelyids, such as Hanbogdemys (Sukhanov 2000) and Basilemys (Mlynarsky 1976). The first marginal scute, visible in MD4-1 and MD8-2, is four-sided, trapezoidal in shape, with the anterior margin much longer than the posterior. The second marginal is complete in MD8-2 and nearly so in MD4-1. It is ‘boot-shaped’ and is anteroposteriorly longer than mediolaterally wide. The sulcus between the first and second marginals and that between the second marginal and the first pleural scute form an acute angle, extending onto the nuchal plate and barely contacting the first vertebral scute. This arrangement is clearly visible in MD8-2 and MD4-1. The third marginal scute, complete on the right side of MD8-2, is much longer than wide. The left seventh to twelfth marginals, as well as the right tenth to twelfth ones are preserved in MD8-2. The anterior marginals, from the first to the third, and the lateral ones, from the eighth to the tenth, are restricted to the peripheral plates. The seventh marginal extends onto the costal plates; and the eleventh and twelfth marginals are strongly expanded anteroposteriorly, extending onto the suprapygal and costal plates, as in most adocids and nanhsiungchelyids. The posterior marginal scutes are clearly larger than the anterior ones. The eleventh and twelfth marginals are visible in MD8-1; they are similar to those of MD8-2. Plastron (Figs 3, 4 and 6). The plastron is preserved in MD8-1 and MD8-2, both lacking most of the anterior lobe. An incomplete anterior lobe, lacking the left part, is preserved in MD3-1. The plastron is strongly sutured to the carapace. The entire plastron is strongly built, with large anterior and posterior lobes and a long bridge. The anterior lobe (MD3-1) is wide and long. The right epiplastron is nearly complete, with a long midline suture to its mate that represents about 40% of the length of the entoplastron. The entoplastron (Fig. 6c) is complete, being a five-sided plate, wider than long, with a pointed anterior end and a nearly straight and transversal posterior margin. On the inner side, a pair of rounded tubercles is present on the anterior part of this plate. The axillary buttress, preserved in MD8-1, MD8-2 and MD3-1, is long. It reaches at least the third peripheral plate, but does not extend to the costal plate. The bridge is long and narrow. It is slightly longer than

163

the posterior lobe. The axillary and inguinal notches are very narrow. The posterior lobe, complete in both MD8-1 and MD8-2, is wide, with a rounded posterior end. The anal notch is absent. The xiphiplastron is slightly shorter than the hypoplastron. The hypoplastron–xiphiplastron suture is slightly convex anteriorly. The general morphology of the plastron of the Phu Kradung turtles is reminiscent of adocids (Adocus, Adocoides) and nanhsiungchelyids (Zangerlia, Basilemys, Hanbogdemys, Nanhsiungchelys). The plastron of xinjiangchelyids and macrobaenids –sinemydids differs significantly from that of the Phu Kradung turtles in the loose carapace –plastron attachment, the reduced anterior and posterior lobes, the short, wide and fan-shaped bridge, and the small and much narrower entoplastron. The sulci on the plastron are only partly visible in MD8-2 (Fig. 4). They are not preserved in MD3-1. The anal scute appears to be restricted to the xiphiplastron, not reaching the hypoplastron– xiphiplastron suture. On the isolated entoplastron from Huai Pai, the humeropectoral sulcus is preserved, on the posterior part of the entoplastron. Vertebrae. An isolated sixth cervical vertebra is preserved inside MD8-2, exposed in left lateral view. The cervical is long and low as in Zangerlia neimongolensis (Brinkman & Peng 1996), unlike the short and high sixth cervical of Dracochelys (Brinkman 2001). The amphicoelous centrum bears a large ventral keel extending along the full length of the centrum, with a straight ventral margin. The anterior articular surface is single whereas the posterior one is double. The postzygaphophysis rises gently from the middle of the centrum posterolaterally. The transverse process is damaged; it is located under the prezygapophysis. An isolated anterior cervical vertebra (MD8-4) has been collected near MD8-2 (Fig. 6d). This cervical is deformed, being compressed laterally. As preserved, it measures 39 mm in total length. It is narrow, high and long, being twice as long as wide. The centrum bears a thin and deep ventral keel. It is weakly amphicoelous. The anterior articular surface is slightly damaged. It is single, roughly rectangular in shape and wider than high. The posterior articular surface is single, narrow, higher than wide and appears more concave than the anterior one. The large transverse process is well separated from the prezygapophysis and located at the anterior end of the vertebra, below the prezygapophysis. There is a distinct ridge under the transverse process extending from the lateral side of the centrum to the under-surface of the transverse process. The neural arch has widely

164

H. TONG ET AL.

separated prezygapophyses and postzygaphophyses and a low neural spine. The first thoracic vertebra is preserved in MD8-1 and MD8-2, articulated to the shell. The anterior articular surface of the first thoracic vertebra is concave and single, facing anteroventrally. The amphicoelous cervical centrum is a primitive feature of cryptodires (Gaffney 1990, 1996). It is present in plesiochelyids (Thalassemys (Rieppel 1980)) and Xinjiangchelys (Peng & Brinkman 1993; Matzke et al. 2004a) among Eucryptodira. In general shape, the long and slender cervicals from Kham Phok resemble those of Xinjiangchelys qiguensis, Zangerlia neimongolensis (Brinkman & Peng 1996), Basilemys sp. (Brinkman 1998) and Adocus (Meylan & Gaffney 1989), but differ from the short cervicals of Dracochelys (Brinkman 2001), Macrobaena (Tatarinov 1959) and Ordosemys (Brinkman & Peng 1993a). The cervicals of X. qiguensis differ from our specimens in that the transverse process is located more posteriorly (Matzke et al. 2004a). A double articular surface between the sixth and seventh cervicals occurs in Adocus (Meylan & Gaffney 1989) and the nanhsiungchelyids Zangerlia neimongolensis and Basilemys sp. (Brinkman 1998). They differ from the cervicals from Kham Phok in the opisthocoelous second to seventh cervicals. In addition, Z. neimongolensis has a double central articulation between the eighth cervical and the first thoracic vertebrae, whereas the first thoracic vertebra from Kham Phok has a single anterior central articulation. Adocus has a single but concave posterior central articulating surface of the eighth cervical. A complete series of 10 dorsal vertebrae is preserved in MD8-1 and MD8-2, articulated to the shell. They are long, slender and high, with a rounded ventral surface.

Pectoral and pelvic girdle The complete ilium, pubis and ischium are preserved in MD8-1 (Fig. 7); the pubis and ischium remain articulated to each other, but turned 908 to the left, and both ilia are crushed. A disarticulated pelvis is preserved in MD8-2. The ilium is short and stout, with an enlarged distal end to form the iliac blade. There is no evidence of a thelial process. The pubis and ischium, exposed in dorsal view in MD8-1, form a longer than wide ventral plate. The pubis is narrow and long. The epipubis is not ossified. The pectineal process is absent or very weak, not visible in dorsal view. This is different from Adocus, in which a short pectineal process is present at the mid-length of the pubis. The ischium has a very strong and long metischial process, which is directed backwards and slightly outwards. Both

the pubic and ischiac symphyses are long. The thyroid fenestra is not completely divided; a short space separates the pubis from the ischium on the midline, as in Adocus (Meylan & Gaffney 1989). The general morphology of the pelvis is most similar to that of Adocus sp. (Meylan & Gaffney 1989), but very different from that of Lissemys and other trionychids (Meylan 1987). The absence of the pectineal process is unusual among turtles. The long metischial process is found also in Xinjiangchelyis (Peng & Brinkman 1993; Matzke et al. 2004a), macrobaenids (Tatarinov 1959; Brinkman & Peng 1993a; Parham & Hutchison 2003), Adocus (Meylan & Gaffney 1989), Basilemys (Meylan & Gaffney 1989), and an undescribed new genus and species of trionychoid from the Early Cretaceous of Japan (Hirayama 2000, 2002). The pubis of xinjianchelyids is different from that of the Kham Phok specimens because it is wider than long and bears a well-developed and long pectineal process (Peng & Brinkman 1993; Matzke et al. 2004a).

Limb bones A right humerus (MD8-5) was found near MD8-2 and an incomplete femur (MD8-6) was found inside MD8-2. The humerus is nearly complete, but the medial process and the head are a little damaged. The humerus is 160 mm long, which represents only 18% of the shell length of MD8-2. The shoulder is not developed. The shaft is slightly curved ventrally. The posterior half of the humerus is strongly expanded. The condyles are well developed and face ventrally. The ectepicondyle foramen is crushed. The femur (MD8-6) has both its proximal and distal ends damaged. Its length is similar to or slightly larger than that of the humerus. It has a slightly curved shaft. One isolated nearly complete distal phalanx (MD8-7), 51 mm long, is preserved. It is laterally compressed, with a higher than wide articular surface.

Discussion The large turtles from the Phu Kradung Formation described above clearly belong to the Cryptodira, as the pelvis is not sutured to the shell and the processus trochlearis oticum is present. They belong to Eucryptodira because the carotid artery is enclosed by the pterygoid posteriorly and the mesoplastra are absent (Gaffney & Meylan 1988; Gaffney 1996). Among Eucryptodira, the amphicoelous cervical vertebrae of the Phu Kradung turtles is a primitive

NEW GIANT MESOZOIC TURTLE FROM THAILAND

feature that is reminiscent of the early Eucryptodira, such as the xinjiangchelyid Xinjiangchelys (Peng & Brinkman 1993) and the plesiochelyid Thalassemys (Rieppel 1980). The Phu Kradung turtles are excluded from xinjiangchelyids and plesiochelyids because the posterior portion of the canalis caroticus internus is covered by bone, a derived character found in Centrocryptodira. Several characters are shared by the Phu Kradung turtles and the Early Cretaceous turtles that have been placed in the families Sinemydidae and Macrobaenidae. These characters include the ventrally exposed foramen posterius canalis caroticus laterale, the unossified posterior portion of the canalis caroticus lateralis, a large foramen caroticum basisphenoidale, and the presence of a pair of pits on the ventral surface of the basisphenoid. In addition, the Phu Kradung turtles have a long first thoracic rib that extends along the full width of the first costal plate, a character observed also in the macrobaenid Ordosemys leios from the Early Cretaceous of Inner Mongolia, China (Brinkman & Peng 1993a). The Phu Kradung turtles are separated from sinemydids–macrobaenids by several derived characters. The deeply embedded canalis caroticus internus is characteristic of Polycryptodira. The plastron sutured to the carapace, the broad epiplastra, the large and broad entoplastron, the large plastron with wide anterior and posterior lobes, and the long and narrow bridge are shared with Trionychoidea and Testudinoidea. In addition, the first thoracic vertebra of the Phu Kradung turtles has an anteroventrally facing anterior central articulation. This character is part of the neck retraction mechanism characteristic of advanced cryptodires. Although measuring this character is delicate, it suggests that these large turtles were probably capable of retracting the head inside the shell. The plastral buttresses not reaching the costal plates exclude the Phu Kradung turtles from Testudinoidea. Among Trionychoidea, the very ventral position of the foramen posterior canalis carotici interni, which is completely enclosed by the pterygoids in the Kham Phok skull, is characteristic of Trionychoidae (Meylan & Gaffney 1989). Furthermore, the skull and shells of the Phu Kradung turtles share several derived characters with both Adocidae and Nanhsiungchelyidae. Several previous studies considered Adocidae and Nanhsiungchelyidae as a monophyletic group (Brinkman & Peng 1996; Brinkman 1998; Joyce 2007). According to Brinkman & Peng (1996), the clade Adocidae – Nanhsiungchelyidae is supported by four unambiguous characters. With the exception of the first character (scute sulci on skull roof), which is not preserved in the Kham Phok skull, the other three characters are all present in the Phu Kradung turtles.

165

The neural formula 6 . 4 , 6 , 6 , 6 , 6 occurs in nearly all Adocidae and Nanhsiungchelyidae, with a single exception in adocids: Isanemys from the Early Cretaceous of Thailand (Tong et al. 2006a). This character is used as a synaphomorphy to unite these two families (Gaffney & Meylan 1988; Brinkman & Peng 1996; Hirayama et al. 2001; Joyce 2007), or considered as an independent occurrence (Meylan & Gaffney 1989; Tong et al. 2006a). In Nanhsiungchelyidae, the neural series is complete, reaching the suprapygal; whereas in Adocidae, there is a diminution of the number of neurals, the posterior costals meeting on the midline. Only the shape of the first six neurals could be determined in the large turtles from the Phu Kradung Formation (see reconstruction, Fig. 8). The pectoral scute reaching the entoplastron is present in all nanhsiungchelyids and advanced adocids, but absent in primitive adocids, including Isanemys, Ferganemys, the undescribed trionychoid from the Neocomian of Japan (Hirayama et al. 2000), and Yehguia tatsuensis, a primitive trionychoid closely related to adocids and nanhsiungchelyids (Danilov & Parham 2006). Thus this character could be also considered as occurring independently in Adocidae and Nanhsiungchelyidae. The advanced adocids, such as Adocus and Adocoides, have the lateral and posterior marginals mediolaterally expanded, overlapping the costal plates, whereas in most adocids, including primitive adocids, and most nanhsiungchelyids, only the eleventh and twelfth marginals are anteroposteriorly expanded, being clearly longer than wide, extending beyond the peripheral plates. In adocids, this occurs in Ferganemys, Adocus, Adocoides, the undescribed trionychoid from Japan, Shachemys baibolatica and S. ancestralis (Nessov 1986; Danilov et al. 2007). In Shachemys laosiana, the twelfth marginals are elongated and narrow as in the Phu Kradung turtles, overlapping the suprapygal, whereas the eleventh marginals are included in the peripheral plates (de Lapparent de Broin 2004). In nanhsiungchelyids, this character is observed in Hanbogdemys, Bulganemys (Sukhanov 2000; Sukhanov & Narmandakh 2006), Basilemys variolosa (Langston 1956), but not in Zangerlia testudimorpha (Mlynarski 1972). The eleventh and twelfth marginals are short, and wider than long in sinemydids –macrobaenids. They are square in some xinjiangchelyids. In these forms, either the eleventh and twelfth marginals are included in the pygal and peripherals or only the twelfth marginals extend onto the posterior suprapygal. Testudinoids also have short eleventh and twelfth marginals. The twelfth pair of marginals extending onto the pygal in this group is due to the short pygal plate, but not due to the elongation of the marginals.

166

Table 2. Distribution of characters from the Appendix 20

30

40

50

60

70

80

0000000000 0000010000 0011001000 0011011100 0011011110 0011011110 0111011111 ???????111 0011011111 0011111111 ?111111111 0011011111 0111011111 1111011111 0111011111 1111011111 1111011111 1111011111 ?????11111 11?1??1??? 1111011111 0111011??? 1111011111 1111011111 1111011111

00?0000000 0000001000 00?0000000 0000?00000 1001101000 1011101000 a000001000 0000?11??0 01?0000000 0000011110 00?0?111?0 0000?11100 11?0?11100 1110001100 1100001100 ?110001110 1110001111 1110001110 0100?11??1 ??????1110 ???0?01110 11????0001 1110001110 1110001110 11100011a0

?000000000 100000aaa0 0000000??0 0000000??0 000010a??0 0000001100 0000001??0 ??000?1??0 1000000001 10000?1??1 00000?1101 00000?1?01 00000?11?1 100000a001 000000a001 0010011111 1011110111 1011111111 ?00?0????0 1?00??1011 10100?11?1 1?0?1?00?1 0100010001 0100010001 a000001a01

0000000000 001fe0100a 00?0000?00 00?0000?00 00?0000?00 001e00a000 0010001?00 10?0000001 0001100000 1??220??00 0??2200001 1?11200000 1111101001 1111101111 1111111111 1112211111 1012211111 1012211111 11????1?10 1112?111?? 1112211111 ???2?????? 1113111111 1113111111 111dc11111

000001?00? 00a0000ca0 000???0010 000???0010 00???0011c 0000000c10 ?????00010 ??0?0?0110 000001?11? ?????00110 ??????0110 111??00110 111??00110 11000002a0 0110000210 ?1?0101212 011010021? 011?100??? ???01?1211 ????111210 ????111211 ???0?11111 0111100210 0111100210 a1a00aac1a

00??000000 0abaa00100 1010000000 0030000100 10b0000100 00b0000110 0a001001?0 0000101000 00??101001 001010101? 00111110?1 00001110?? 00001110?1 0000111011 00b01a1011 1121100010 112?111010 112?1a101? 1??1100010 1a11100010 10a1100010 1011100??0 01c1100010 0a2?110010 0aa010a110

01000000?0 001110e00a 0100000000 0011?000?0 001c100000 0a1a100000 001a10000a 0010100000 001?0000?1 001?1??00a 00101??001 00101?0001 0010110000 0010110001 0010110001 0012101000 00111????0 00121????1 00121?10?0 111210?010 11121020a0 1112102010 001211?100 001111?100 aa1111e000

0000 0010 0000 0010 00a0 00a0 0100 0101 0?0? 1?00 1??0 0100 0100 1100 1000 0001 a?0? a??? 000? 0001 0001 0001 0011 1000 00a0

Taxon/character matrix; a ¼ (01); b ¼ (02); c ¼ (12); d ¼ (13); e ¼ (012); f ¼ (023).

H. TONG ET AL.

Proganochelys Pleurodira Kayentachelys Kallokibotion Pleurosternidae Baenidae Plesiochelyidae Xingjiangchelys Meiolaniidae Sinemys Dracochelys Ordosemys Judithemys Chelydridae Chelonioidea Adocus Carettochelyidae Trionychidae Basilochelys Basilemys Zangerlia Nanhsiungchelys Dermatemys Staurotypus Testudinoidea

10

NEW GIANT MESOZOIC TURTLE FROM THAILAND

167

Fig. 9. Strict consensus cladogram produced from the data matrix in Table 2.

Two additional characters can be added to the above-mentioned characters of the Adocidae and Nanhsiungchelyidae: both the shell and the skull roof surface are sculptured and the vertebral scutes are narrow. A sculptured shell surface is found in most Trionychoidea, including Trionychidae, Carettochelyidae, Adocidae, Nanhsiungchelyidae and Dermatemydidae. The pattern of the ornamentation seen on the Huai Sai carapace is most similar to that of Anomalochelys, a

nanhsiungchelyid from the Late Cretaceous of Japan (Hirayama et al. 2001). The sculpturing of both shell and skull roof surface occurs in Nanhsiungchelys (Yeh 1966) and Carettochelys, whereas in Zangerlia neimongolensis only the shell is sculptured (Brinkman & Peng 1996). The skull roof of Basilemys sp. is also sculptured, but it does not match the sculpturing of the carapace (Brinkman 1998). Both adocids and nanhsiungchelyids have narrow vertebrals. The very narrow

168

H. TONG ET AL.

vertebrals, with the second to the fourth vertebrals much longer than wide as seen in the Phu Kradung turtles, are present in some nanhsiungchelyids, such as Hanbogdemys orientalis and Basilemys variolosa, and some testudinoids. This is also the case in carettochelyid, when the scutes are present on the shell. As shown in the description and comparisons, the general morphology of the Kham Phok skull presents strong similarities to that of Adocus from the Late Cretaceous of North America (Meylan & Gaffney 1989). The skull of Adocus is more derived than that from Kham Phok in that the palatine artery is fully embedded in bone; thus the foramen posterius canalis caroticus laterale is no longer visible on the ventral surface, although a foramen caroticum basisphenoidale is still present. Another derived feature that separates Adocus from the Phu Kradung turtles is the significant contribution of the parietal to the processus trochlearis oticum. For these two characters, the condition of the Phu Kradung turtles is closer to that of nanhsiungchelyids. Zangerlia neimongolensis also has the canalis caroticus lateralis and canalis caroticus internus exposed in a large depression, but the poor preservation of that region in Zangerlia prevents a detailed comparison (Brinkman & Peng 1996). The significant contribution of the parietal to the processus trochlearis oticum, considered as a synapomorphy of Trionychoidae by Meylan & Gaffney (1989) is observed in the adocid Adocus, Ferganemys and a primitive adocid skull from the Late Cretaceous of Uzbekistan (Danilov & Parham 2005), but absent in the nanhsiungchelyid Basilemys sp. and also in Nanhsiungchelys according to Brinkman (1998). We performed a cladistic analysis based on previously defined data matrices (Meylan 1987; Meylan & Gaffney 1989; Rougier et al. 1995; Brinkman & Peng 1996; Gaffney 1996; Shaffer et al. 1997; Brinkman & Wu 1999; Hirayama et al. 2000, 2001; Joyce 2007) using the parsimony algorithm of PAUP 4.0 (Swofford 1996). We considered Hanbogdemys and Nanhsiungchelys as belonging to a monophyletic entity, following Brinkman & Nicholls (1993). The different states of 74 characters were scored for 25 taxa (Table 2 and Appendix). The analysis resulted in a single tree of 177 steps (Fig. 9) with a consistency index of 0.47 and retention index of 0.74, suggesting that homoplasy is rampant in the dataset. On the 74 characters, 42 were observed in Basilochelys. Our analysis agrees with the cladogram obtained by Gaffney (1996) for interfamilial relationships except for the position of Kayentachelys and Kallokibotion, and with that of Meylan & Gaffney (1989) for the relationships within Trionychoidea, in placing Adocus as the sister

group of Nanhsiunchelyidae þ Trionychia. Basilochelys is in a early position within the monophyletic Trionychoidae and closely related to Adocus and nanhsiungchelyids. The high homoplasy rate in the tree implies that the relationships must be considered with caution. There is no exclusive apomorphy for uniting Trionychoidea, Trionychoidae or Nanhsiungchelyidae. These groups are defined more by character combinations than by a series of unambiguous autapomorphies. In contrast, several characters (e.g. 34, 35) occur only in Trionychoidae (Trionychia, Adocus and Nanhsiungchelyidae) and in some macrobaenid –sinemydid turtles, which suggests that the Trionychoidae may have an earlier position within eucryptodiran turtles. This may explain both the early occurrence of Trionychoidae in the fossil record, in contrast to Kinosternoidae, Chelydroidea or Testudinoidea; and the non-monophyletic nature of Kinosternoidea þ Trionychoidae based on recent molecular phylogenies (Shaffer et al. 1997; Fujita et al. 2004), although the tree that we have obtained does not support this scenario. The association of primitive and derived characters in Basilochelys would also support a more primitive position for Trionychoidae. Further discoveries could therefore challenge the current scenario.

Conclusion The large turtles from the terminal Jurassic–earliest Cretaceous Phu Kradung Formation of the Khorat Group, NE Thailand, described herein represent a new genus and new species of Eucryptodira, Basilochelys macrobios n. gen. n. sp. This taxon is placed in Trionychoidae and considered as the most basal member of that group. The combination of primitive and derived characters of Basilochelys suggests that the group Trionychoidae may have originated from xinjiangchelyids and their close relatives. Siamochelys from the Middle Jurassic of the southern peninsula of Thailand may represent the sister taxon of Trionychoidae, because of its ligamentous carapace – plastron attachment, sculptured shell surface and wide entoplastron. These new discoveries add significantly to the still poorly known turtle fauna from the Late Jurassic–Early Cretaceous of SE Asia and provide important information about the origin and early evolution of modern cryptodiran turtles. This work was supported by the Department of Mineral Resources, Thailand, the ECLIPSE Programme of CNRS, France, the TRF-CNRS Special Programme for Biodiversity Research and Training (BRT/BIOTEC/ NSTDA), Grant BRT R-24507; and the PHC ‘Evolution et biodiversite´ des verte´bre´s aquatiques Me´sozoiques et Ce´nozoiques en Thaı¨lande’ (16610UG). We thank all the

NEW GIANT MESOZOIC TURTLE FROM THAILAND people who took part in the field work and prepared the specimens; and P. Meylan (Saint Petersburg, Florida) and D. Brinkman (Drumheller, Alberta) for reviewing the paper. This is publication ISEM 2007-166 of J.C.

17.

Appendix: Definition of characters and their states used in the analysis of relationships of Basilochelys

19.

18.

20. 21.

Character state definition: Skull: 1. 2. 3. 4. 5. 6. 7. 8.

9.

10.

11.

12.

13.

14.

15. 16.

Nasals: present ¼ 0; absent ¼ 1 (Gaffney 1996: 1). Prefrontals: do not meet in midline ¼ 0; meet in midline ¼ 1 (Gaffney 1996: 2). Prefrontal –vomer contact: absent ¼ 0; present ¼ 1 (Gaffney 1996: 3). Vertical flange on processus pterygoideus externus: absent ¼ 0; present ¼ 1 (Gaffney 1996: 4). Foramen palatinum posterius: relatively small ¼ 0; relatively large ¼ 1 (Gaffney 1996: 5). Interpterygoid vacuity: open ¼ 0; closed ¼ 1 (Gaffney 1996: 6). Processus trochlearis oticum: absent ¼ 0; present ¼ 1 (Gaffney 1996: 7). Middle ear with ossified floor formed by posteromedial process of pterygoid: pterygoid process absent ¼ 0; pterygoid process present ¼ 1 (Gaffney 1996: 8). Canalis caroticus internus at least partially formed by pterygoid: not formed by pterygoid to any extent ¼ 0; partially or entirely formed by pterygoid ¼ 1 (Gaffney 1996: 9). Canalis caroticus internus formed entirely by pterygoid posteriorly (distally): formed partially or not by pterygoid ¼ 0; formed entirely by pterygoid ¼ 1 (Gaffney 1996: 10). Canalis caroticus internus and canalis caroticus lateralis completely embedded in bone: both canals open ventrally ¼ 0; both canals embedded in bone ¼ 1 (Gaffney 1996: 11). Thickness of pterygoid floor of canalis caroticus internus: thin or absent ¼ 0; thick ¼ 1 (Gaffney 1996: 12). Canalis caroticus lateralis versus canalis caroticus internus: canalis caroticus lateralis equal to or larger than canalis caroticus internus ¼ 0; canalis caroticus lateralis smaller than canalis caroticus internus ¼ 1 (Gaffney 1996:13; some scores according to Shaffer et al. 1997) Foramen posterius canalis carotici interni: not formed by basisphenoid and pterygoid ¼ 0; formed by basisphenoid and pterygoid, midway along basisphenoid– pterygoid suture ¼ 1 (Gaffney 1996: 14). Fenestra perilymphatica: relatively large ¼ 0; relatively small ¼ 1 (Gaffney 1996: 15). Paired pits on ventral surface of basisphenoid: pits absent ¼ 0; pits present ¼ 1 (Gaffney 1996: 16).

22. 23.

24. 25.

26.

27.

169

Posterior temporal emargination: not developed ¼ 0; at least partially developed ¼ 1 (Gaffney 1996: 17). Parietal–squamosal contact: present ¼ 0; absent ¼ 1 (Gaffney 1996: 18). Postorbital– squamosal contact: present ¼ 0; absent ¼ 1 (Gaffney 1996: 19). Skull roof surface sculptured: no ¼ 0; yes ¼ 1 (Brinkman & Peng 1996: 15). Incisura columellae auris: open ¼ 0; closed ¼ 1 (Meylan & Gaffney 1989: 13). Foramen stapedio-temporale: large ¼ 0; absent or small ¼ 1 (Shaffer et al. 1997: 92). Contribution of parietal to processus trochlearis oticum: little or none ¼ 0; large ¼ 1 (Meylan & Gaffney 1989: 44) Fused premaxillae: no ¼ 0; yes ¼ 1 (Meylan 1987: 44). Ventral contact between left and right pterygoid: present ¼ 0; absent ¼ 1 (Meylan & Gaffney 1989: 12). Palatine contribution to the side wall of braincase: small or absent ¼ 0; large ¼ 1 (Shaffer et al. 1997: 68). Lower cheek emargination: absent or shallow, with processus pterygoidei externus concealed from lateral ¼ 0; deep, with processus pterygoidei externus exposed from lateral ¼ 1 (Meylan & Gaffney 1989: 42).

Lower jaw: 28. 29.

Coronoid tall and located in the middle of the mandible: no ¼ 0; yes ¼ 1 (Meylan & Gaffney 1989: 43). Retroarticular process: absent ¼ 0; present ¼ 1 (Meylan & Gaffney 1989: 18).

Axial skeleton: 30. Central articulations of cervical vertebrae: unformed (platycoelous or amphicoelous) ¼ 0; formed (concavo-convex) ¼ 1 (Gaffney 1996: 20). 31. Transverse processes of cervical vertebrae: on middle of centrum ¼ 0; on anterior of centrum ¼ 1 (Gaffney 1996: 21). 32. Posterior cervicals with strong ventral process: absent ¼ 0; present ¼ 1 (Gaffney 1996: 22; and see Joyce 2007, for codings in Trionychia). 33. Cervical ribs: present ¼ 0; absent ¼ 1 (modified from Gaffney 1996: 23, according Joyce 2007). 34. Fourth cervical central articulation (unordered): amphicoelous ¼ 0; biconvex ¼ 1; opisthocoelous ¼ 2; procoelous ¼ 3 (Gaffney 1996: 24). 35. Eighth cervical central articulation (unordered): amphicoelous ¼ 0; procoelous 1; biconvex or opisthocoelous ¼ 2 (modified from Gaffney 1996: 25). 36. Double (i.e. transversely paired) central articulations between the seventh and eighth cervicals: absent ¼ 0; present ¼ 1 (Gaffney 1996: 26).

170 37.

38. 39.

40.

41.

42.

43.

H. TONG ET AL. Cervical vertebra having a distinct double transverse process (i.e. diapophysis and parapophysis): present in at least some cervicals ¼ 0; absent ¼ 1 (Gaffney 1996: 27). Neural spine on eighth cervical: high ¼ 0; low ¼ 1 (Gaffney 1996: 28). Anterior articulation of first thoracic centrum: faces anteriorly or slightly anteroventrally ¼ 0; faces strongly anteroventrally ¼ 1 (Gaffney 1996: 33). First thoracic rib: extends to peripherals or nearly so and lies behind the tip of the axillary buttress ¼ 0; extends less than halfway across first costal ¼ 1 (Gaffney 1996: 32). A biconcave caudal near base of tail: absent ¼ 0; present ¼ 1 (Gaffney 1996: 29; Sinemys unknown as suggested by Joyce for this character). Caudal central articulations: all centra amphicoelous or opisthocolous ¼ 0; at least first two caudals procoelous ¼ 1 (Gaffney 1996: 30). Chevrons: well developed and present on nearly all caudals ¼ 0; small to absent (if present, only on a few anterior caudals) ¼ 1 (Gaffney 1996: 31; score according to Joyce 2007).

Girdles: 44. 45.

Thelial process: absent ¼ 0; present ¼ 1 (Meylan & Gaffney 1989: 37). Thyroid fenestra: subdivided by midline contact of pubis and ischium ¼ 0; confluent ¼ 1 (Shaffer et al. 1997: 114; testudinoids coded as non-confluent).

Appendicular skeleton: 46.

Phalangeal formula: most digits with three phalanges ¼ 0; most digits with two phalanges ¼ 1 (Meylan & Gaffney 1989: 40).

Carapace: 47. Neural formula 6 . 4 , 6 , 6 , 6 , 6: no ¼ 0; yes ¼ 1 (Meylan & Gaffney 1989: 20). 48. 2nd to 4th vertebral scutes: much broader than pleural scute ¼ 0; as broad as pleural and wider than long ¼ 1; narrower than pleurals and longer than wide ¼ 2 (modified from Hirayama et al. 2000: 59; ordered). 49. Supramarginal scutes: present ¼ 0; absent ¼ 1 (Gaffney 1996: 36). 50. Marginal scutes reaching costal plates: no ¼ 0; posteriorly only ¼ 1; posteriorly and laterally ¼ 2 (Meylan & Gaffney 1989: 47; ordered). 51. Shell ornamentation with pits, ridges or vermiculation: absent ¼ 0; present ¼ 1. 52. Posterior costals meeting on the midline: no ¼ 0; yes ¼ 1 (Meylan & Gaffney 1989: 21). 53. Suprapygal: nearly equally divided into two ¼ 0; first suprapygal much smaller than second ¼ 1; first suprapygal absent ¼ 2; second suprapygal

54.

much smaller than first or absent ¼ 3 (Hirayama et al. 2000: 61). Second suprapygal: contacts peripheral 11 ¼ 0; contacts peripheral 10 and 11 ¼ 1 (Hirayama et al. 2001: 23).

Plastron: 55. 56. 57. 58. 59. 60. 61.

62.

63. 64.

65. 66.

67.

68. 69. 70.

71. 72.

Mesoplastra: present ¼ 0; absent ¼ 1 (Gaffney 1996: 34). Length of bridge: long ¼ 0; short ¼ 1 (Brinkman & Wu 1999: 52). Attachment of carapace and plastron: sutured ¼ 0; ligamentous ¼ 1 (Gaffney 1996: 35). Axillary buttress reach costal bone: absent ¼ 0; present ¼ 1 (Meylan & Gaffney 1989: 24). Dorsal process on epiplastron: present ¼ 0; absent ¼ 1 (Gaffney 1996: 37). Epiplastron: broad ¼ 0; narrow ¼ 1 (Gaffney 1996: 39). Thick epiplastron with broad dorsal extension of gular scute: absent ¼ 0; present ¼ 1 (Hirayama et al. 2001: 35). Epiplastron projected beyond the anterior margin of carapace: no ¼ 0; yes ¼ 1 (Hirayama et al. 2001: 36). Entoplastron separating epiplastra: yes ¼ 0; no ¼ 1 (Gaffney 1996: 38). Entoplastron: narrow (W  L, like Ordosemys) ¼ 0; broad (W ¼ L or W , L, like testudinoids) ¼ 1; very broad (W  L, like nanhsiungchelyids and Adocus) ¼ 2 (modified from Hirayama et al. 2000: 72, with one more character state) (ordered). Entoplastron: size of posterior entoplastral process: present ¼ 0; absent ¼ 1 (Rougier et al. 1995: 46). Loss of plastral scute set 2 (gulars or extragulars): full set of scales 1 and 2 (gular, extragular see Hutchison & Bramble 1981; equal to intergular, gular of older terminology) ¼ 0; one set of scales absent (scale set 2 of Hutchison & Bramble 1981) ¼ 1 (Gaffney 1996: 40). Entoplastron: plastral scute set 4 (pectorals) well behind entoplastron ¼ 0; plastral set 4 reaching the posterior end of entoplastron ¼ 1; plastral set 4 reaching half length of the entoplastron ¼ 2 (modified from Meylan & Gaffney 1989: 33). Plastral scute set 4: present ¼ 0; absent ¼ 1 (Meylan & Gaffney 1989: 32). 6th pair of marginal scutes greatly expanded ventromedially:no ¼ 0;yes ¼ 1 (Hirayama et al. 2001:30). Plastral fontanelle between hyoplastra and hypoplastra: absent ¼ 0; retained in adult ¼ 1 (Hirayama et al. 2000: 63). Xiphiplastron: broad ¼ 0; narrow ¼ 1 (Hirayama et al. 2000: 75). Femoro-anal sulcus reaching hypoplastral– xiphiplastral suture: no ¼ 0; yes ¼ 1 (Hirayama et al. 2000: 68).

NEW GIANT MESOZOIC TURTLE FROM THAILAND 73. 74.

Anal notch: absent ¼ 0; present ¼ 1 (Hirayama et al. 2000: 76). Midline plastral sulcus sinuous: no ¼ 0; yes ¼ 1 (Meylan & Gaffney 1989: 30).

References B RINKMAN , D. 1998. The skull and neck of the Cretaceous turtle Basilemys (Trionychoidea, Nanhsiungchelyidae), and the interrelationships of the genus. Paludicola, 1, 150 –157. B RINKMAN , D. 2001. New material of Dracochelys (Eucryptodira:Sinemydidae) from the Junggar Basin, Xinjiang, People’s Republic of China. Canadian Journal of Earth Sciences, 38, 1645–1651. B RINKMAN , D. & N ICHOLLS , E. L. 1993. New specimen of Basilemys praeclara Hay and its bearing on the relationships of the Nanhsiungchelyidae (Reptilia: Testudines). Journal of Paleontology, 67, 1027–1031. B RINKMAN , D. & P ENG , J.-H. 1993a. Ordosemys leios n. gen., n. sp., a new turtle from the Early Cretaceous of the Ordos Basin, Inner Mongolia. Canadian Journal of Earth Sciences, 30, 2128–2138. B RINKMAN , D. & P ENG , J.-H. 1993b. New material of Sinemys (Testudines, Sinemydidae) from the Early Cretaceous of China. Canadian Journal of Earth Sciences, 30, 2139– 2152. B RINKMAN , D. & P ENG , J.-H. 1996. A new species of Zangerlia (Testudines: Nanhsiungchelyidae) from the Upper Cretaceous redbeds at Bayan Mandahu, Inner Mongolia, and the relationships of the genus. Canadian Journal of Earth Sciences, 33, 526 –540. B RINKMAN , D. & W U , X.-C. 1999. The skull of Ordosemys, an Early Cretaceous turtle from the Inner Mongolia, People’s Republic of China, and the interrelationships of Eucryptodira (Chelonia, Cryptodira). Paludicola, 2, 134– 147. B UFFETAUT , E. & S UTEETHORN , V. 2007. A sinraptorid theropod (Dinosauria: Saurischia) from the Phu Kradung Formation of northeastern Thailand. Bulletin de la Socie´te´ Ge´ologique de France, 178, 497–502. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J., K OBAYASHI , Y., H UANG , D. & L EE , Y. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19–37. C ARTER , A. & B RISTOW , C. S. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau Basin, eastern Thailand. Basin Research, 15, 271– 285. D ANILOV , I. & P ARHAM , J. F. 2005. A reassessment of the referral of an isolated skull from the Late Cretaceous of Uzbekistan to the stem-testudinoid turtle genus Lindholmemys. Journal of Vertebrate Paleontology, 25, 784–791. D ANILOV , I. & P ARHAM , J. F. 2006. A redescription of Plesiochelys tatsuensis from the Late Jurassic of China, with comments on the antiquity of the crown clade Cryptodira. Journal of Vertebrate Paleontology, 26, 573–580.

171

D ANILOV , I. G. & P ARHAM , J. F. 2007. The type series of ‘Sinemys’ wuerhoensis, a problematic turtle from the Lower Cretaceous of China, includes at least three taxa. Palaeontology, 50, 431– 444. D ANILOV , I. G. & S UKHANOV , V. B. 2006. A basal eucryptodiran turtle ‘Sinemys’ efremovi (¼Wuguia efremovi) from the Early Cretaceous of China. Acta Palaeontologica Polonica, 51, 105–110. D ANILOV , I. G., A VERIANOV , A. O., S KUTCHAS , P. P. & R EZVYI , A. S. 2006. Kirgizemys (Testudines, ‘Macrobaenidae’): New material from the Lower Cretaceous of Buryatia (Russia) and taxonomic revision. Russian Journal of Herpetology, 13(Supplement), 46–62. D ANILOV , I., S YROMYATNIKOVA , E. V. & S UKHANOV , V. B. 2007. Turtles of the genus Shachemys from the Upper Cretaceous of Asia. In: R OZANOV , A. Y., L OPATIN , A. V. & P ARKHAEV , P. Y. (eds) Modern Paleontology: Classical and New Methods—2007. Third all-Russian School 2006. Palaeontological Institute, Moscow, 59– 72. DE L APPARENT DE B ROIN , F. 2004. A new Shachemydinae (Chelonii, Cryptodira) from the Lower Cretaceous of Laos: Preliminary data. Comptes Rendus Pale´vol, 3, 387– 396. F UJITA , M. K., E NGSTROM , T. N., S TARKEY , D. E. & S HAFFER , H. B. 2004. Turtle phylogeny: Insights from a novel nuclear intron. Molecular Phylogenetics and Evolution, 31, 1031–1040. G AFFNEY , E. S. 1979. Comparative cranial morphology of recent and fossil turtles. Bulletin of the American Museum of Natural History, 164, 67–376. G AFFNEY , E. S. 1983. The cranial morphology of the extinct horned turtle, Meiolania platyceps, from the Pleistocene of Lord Howe Island, Australia. Bulletin of the American Museum of Natural History, 175, 361– 479. G AFFNEY , E. S. 1990. The comparative osteology of the Triassic turtle Proganochelys. Bulletin of the American Museum of Natural History, 194, 1 –263. G AFFNEY , E. S. 1996. The postcranial morphology of Meiolania platyceps and a review of the Meiolaniidae. Bulletin of the American Museum of Natural History, 229, 1– 166. G AFFNEY , E. S. & M EYLAN , P. A. 1988. A phylogeny of turtles. In: B ENTON , M. J. (ed.) The Phylogeny and Classification of the Tetrapods. Clarendon Press, Oxford, 157–219. G AFFNEY , E. S. & M EYLAN , P. A. 1992. The Transylvanian turtle, Kallokibotion, a primitive Cryptodire of Cretaceous age. American Museum Novitates, 3040, 1– 37. G AFFNEY , E. S. & Y E , X.-K. 1992. Dracochelys, a new Cryptodiran turtle from the Early Cretaceous of China. American Museum Novitates, 3048, 1– 13. G AFFNEY , E. S., R ICH , T. H., V ICKERS -R ICH , P., C ONSTANTINE , A., V ACCA , P. & K OOL , L. 2007. Chubutemys, a new Eucryptodiran turtle from the Early Cretaceous of Argentina, and the relationships of the Meiolaniidae. American Museum Novitates, 3599, 1 –35. H IRAYAMA , R. 2000. Fossil turtles from the Kiwajima Formation (Tetori Group; Early Cretaceous) of KasekiKabe, Shiramine-mura, Ishikawa Prefecture, Central

172

H. TONG ET AL.

Japan. In: MATSUOKA , H. (ed.) Fossils of the Kuwajima ‘Kaseki-Kabe’ (Fossil bluff). Scientific report on a Neocomian (Early Cretaceous) fossil assemblage from the Kuwajima Formation, Tetori Group, Shiramine, Ishikawa, Japan. Shiramine-mura Education Board, Ishikawa, 75– 92. H IRAYAMA , R. 2002. Preliminary report of the fossil turtles from the Kitadani Formation (Early Cretaceous) of the Tetori Group of Katsuyama, Fukui Prefecture, Central Japan. Memoir of the Fukui Prefectural Dinosaur Museum, 1, 29–40. H IRAYAMA , R., B RINKMAN , D. & D ANILOV , I. 2000. Distribution and biogeography of non-marine Cretaceous turtles. Russian Journal of Herpetology, 7, 181– 198. H IRAYAMA , R., S AKURAI , K., C HITOKU , T., K AWAKAMI , G. & K ITO , N. 2001. Anomalochelys angulata, an unusual land turtle of Family Nanhsiungchelyidae (Superfamily Tronychoidea; Order Testudines) from the Upper Cretaceous of Hokkaido, North Japan. Russian Journal of Herpetology, 8, 127– 138. H UTCHISON , J. H. & B RAMBLE , D. M. 1981. Homology of the plastral scales of the Kinosternidae and related turtles. Herpetelogica, 37, 73– 85. J OYCE , W. G. 2007. Phylogenetic relationships of Mesozoic turtles. Bulletin of the Peabody Museum of Natural History, 48, 3– 102. K HOSATZKY , L. I. 1997. Big turtle of the Late Cretaceous of Mongolia. Russian Journal of Herpetology, 4, 148– 154. L ANGSTON , W. J R . 1956. The shell of Basilemys varialosa (Cope). Bulletin and Annual Report of the National Museum of Canada, 142, 155–163. M ATZKE , A. T., M AISCH , M. W., G E , S. & P FRETZSCHNER , H.-U. 2004a. A new xinjiangchelyid turtle (Testudines, Eucryptodira) from the Jurassic Qigu Formation of the southern Junggar Basin, Xinjiang, north-west China. Palaeontology, 47, 1211–1266. M ATZKE , A. T., M AISCH , M. W., P FRETZSCHNER , H.-U., S UN , G. & S TO¨ HR , H. 2004b. A new basal sinemydid turtle (Reptilia: Testudines) from the lower Cretaceous Tugulu Group of the Junggar Basin (NW China). Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 2004, 151 –167. M EYLAN , P. A. 1987. The phylogenetic relationships of soft-shelled turtles (Family Trionychidae). Bulletin of the American Museum of Natural History, 186, 1 –101. M EYLAN , P. A. & G AFFNEY , E. S. 1989. The skeletal morphology of the Cretaceous Cryptodiran turtle, Adocus, and the relationships of the Trionychoidea. American Museum Novitates, 2941, 1–60. M LYNARSKI , M. 1972. Zangerlia testudinimorpha n. gen. n. sp., a primitive land tortoise from the Upper Cretaceous of Mongolia. Palaeontologia Polonica, 27, 85–92. M LYNARSKY , M. 1976. Testudines. Fischer, Stuttgart. N ESSOV , L. A. 1977. Skull morphology of Early Cretaceous turtle belonging to the family Adocidae. Trudy Zoologicheskogo Instituta AN SSSR, 74, 74– 79. N ESSOV , L. A. 1986. Some Late Mesozoic and Paleocene turtles of Soviet Middle Asia. Studia Geologica Salamanticensia, Volumen Especial 2, Studia Palaeocheloniologica, 2, 7 –22.

P ARHAM , J. F. & H UTCHISON , J. H. 2003. A new eucryptodiran turtle from the Late Cretaceous of North America (Dinosaur Provincial Park, Alberta, Canada). Journal of Vertebrate Paleontology, 23, 783–798. P ENG , J.-H. & B RINKMAN , D. 1993. New material of Xinjiangchelys (Reptilia:Testudines) from the Late Jurassic Qigu Formation (Shishugou Group) of the Pingfengshan locality, Junggar Basin, Xinjiang. Canadian Journal of Earth Sciences, 30, 2013– 2026. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. 1. Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5 –40. R IEPPEL , O. 1980. The skull of the Upper Jurassic cryptodire turtle Thalassemys, with a reconsideration of the chelonian braincase. Palaeontographica, 171, 105–140. R OUGIER , G. W., F UENTE , D. L. M. S. & A RCUCCI , A. B. 1995. Late Triassic turtles from South America. Science, 268, 855– 858. S HAFFER , H. B., M EYLAN , P. A. & M C K NIGHT , M. L. 1997. Tests of turtle phylogeny: Molecular, morphological, and paleontological approches. Systematic Biology, 46, 235– 268. S UKHANOV , V. B. 2000. Mesozoic turtles of Middle and Central Aisa. In: B ENTON , M. J., S HISHKIN , M. A., U NWIN , D. M. & K UROCHKIN , E. N. (eds) The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge, 309– 367. S UKHANOV , V. B. & N ARMANDAKH , P. 2006. New taxa of Mesozoic turtles from Mongolia. Russian Journal of Herpetology, 13(Supplement), 119 –127. S WOFFORD , D. L. 1996. PAUP*: Phylogenetic Analysis using Parsimony (*and other Methods), Version 4.0. Sinauer, Sunderland, MA. T ATARINOV , L. P. 1959. New turtle of the family Baenidae from the Lower Eocene of Mongolia. Paleontologicheskii Zhurnal, 1, 100– 113. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2002. Middle Jurassic turtles from southern Thailand. Geological Magazine, 139, 687– 697. T ONG , H., B UFFETAUT , E., S UTEETHORN , V. & S RISUK , P. 2003. Turtle remains from the Sao Khua Formation (Early Cretaceous) of the Khorat Plateau, northeastern Thailand. (abstract). In: D ANILOV , I. G., C HEREPANOV , G. O., H IRAYAMA , R. & P ARHAM , J. F. (eds) Symposium on Turtle Origin, Evolution and Systematics. Zoological Institute of the Russian Academy of Sciences, Saint Petersburg, 30. T ONG , H., B UFFETAUT , E., S UTEETHORN , V. & S RISUK , P. 2004a. First carettochelyid turtle from the Lower Cretaceous of Thailand (abstract). In: DOSTAL , O., GREGORAVA , R. & IVANOV , M. (eds) 2nd EAVP Meeting. Moravian Museum, Brno, 44. T ONG , H., J I , S.-A. & J I , Q. 2004b. Ordosemys (Testudines: Cryptodira) from the Yixian Formation of Liaoning Province, Northeastern China: New specimens and systematic revision. American Museum Novitates, 3438, 1– 20.

NEW GIANT MESOZOIC TURTLE FROM THAILAND T ONG , H., S UTEETHORN , V., C LAUDE , J., B UFFETAUT , E. & J INTASAKUL , P. 2005. The turtle fauna from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 610–614. T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2006a. Isanemys, a new adocid turtle from the Sao Khua Formation (Early Cretaceous) of the Khorat Plateau,

173

northeastern Thailand. Russian Journal of Herpetology, 13(Supplement), 128–137. T ONG , H., C LAUDE , J., B UFFETAUT , E., S UTEETHORN , V., N AKSRI , W. & S UWANNA , C. 2006b. Fossil turtles of Thailand: An updated review. In: L U¨ , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y.-N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 183– 194. Y EH , X.-K. 1966. A new Cretaceous turtle of Nanhsiung, northern Kwangtung. Vertebrata Palasiatica, 10, 191– 200.

Khoratosuchus jintasakuli gen. et sp. nov., an advanced neosuchian crocodyliform from the Early Cretaceous (Aptian – Albian) of NE Thailand K. LAUPRASERT1,2*, G. CUNY3, K. THIRAKHUPT4 & V. SUTEETHORN5 1

Department of Biology, Faculty of Science, Mahasarakham University, Khamrieng, Kantharawichai, Mahasarakham, 44150 Thailand

2

Palaeontological Research and Education Centre, Mahasarakham University, Khamrieng, Kantharawichai, Mahasarakham, 44150 Thailand

3

Natural History Museum of Denmark, University of Copenhagen, Øster Voldgade 5 – 7, 1350 Copenhagen K, Denmark 4

Department of Biology, Faculty of Science, Chulalongkorn University, Bangkok 10300, Thailand 5

Bureau of Fossil Research and Museums, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected])

Abstract: A new slender-snouted neosuchian crocodyliform, Khoratosuchus jintasakuli gen. et sp. nov., is described from the late Early Cretaceous Khok Kruat Formation of NE Thailand. This discovery represents the youngest and most advanced Mesozoic crocodyliform known in Thailand on the basis of the following cranial features: the secondary choanae are relatively posterior and almost enclosed by the pterygoid; the lateral margin of the maxilla is relatively straight without lateral constrictions; the dorsal surface of the skull lacks ridges and fossae; maxillary teeth are homodontous; the anterior end of the jugal and prefrontal terminate at the same level. The specimen bears resemblances to Chinese and European derived neosuchians and suggests a close relationship between the late Early Cretaceous neosuchians of China, Europe and SE Asia.

In the past 26 years, the Khorat Group of Thailand has yielded rich continental vertebrate faunas, including various dinosaurs, pterosaurs, crocodyliforms, turtles, actinopterygian fishes, and hybodont sharks (Buffetaut & Ingavat 1980, 1983; Buffetaut & Suteethorn 1998, 1999; Buffetaut et al. 2003; Cavin et al. 2003, 2004; Cuny et al. 2003, 2006; Tong et al. 2003, 2005; Lauprasert et al. 2007). However, only three taxa of crocodyliforms from the Phu Kradung and Sao Khua Formations have been described: Sunosuchus thailandicus Buffetaut & Ingavat 1980, Goniopholis phuwiangensis Buffetaut & Ingavat 1983, and Siamosuchus phuphokensis Lauprasert et al. 2007. In summer 2004, a Thai–French expedition had an opportunity to visit the collection of the Museum of Petrified Wood and Mineral Resources, Nakhon Ratchasima Province. Many vertebrate fossils from the Ban Saphan Hin locality were housed in the museum. One of the specimens is an interesting crocodyliform skull, described

here as the fourth crocodyliform taxon from the Khorat Group.

Geological setting Khoratosuchus jintasakuli gen. et sp. nov. was found in a sandstone quarry at Ban Saphan Hin, Nakhon Ratchasima Province (Fig. 1) during the fossil excavation of P. Jintasakul and his colleagues in summer 2004. The specimen was embedded in reddish brown mudstone containing pebbles, silcretes and calcretes. The Ban Saphan Hin locality is currently considered as belonging to the Khok Kruat Formation, the uppermost formation of the Khorat Group of NE Thailand (Meesook 2000; Carter & Bristow 2003), which consists of reddish brown, fine- to medium-grained sandstones, siltstones, mudstones and conglomerates, indicative of a meandering river depositional environment (Racey et al. 1996; Meesook 2000; Tong et al. 2005).

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 175–187. DOI: 10.1144/SP315.13 0305-8719/09/$15.00 # The Geological Society of London 2009.

176

K. LAUPRASERT ET AL.

Fig. 1. Geographical position of the Ban Saphan Hin locality (filled star), Nakhon Ratchasima Province, Thailand. PWMR, Museum of Petrified Wood and Mineral Resources.

The Khok Kruat Formation is currently dated as late Early Cretaceous (Aptian –Albian) by the occurrence of the freshwater hybodont shark Thaiodus ruchae, which is also known from the Aptian– Albian Takena Formation of the Lhasa block of Tibet (Cappetta et al. 1990) and by palynomorphs suggesting an Aptian age (Racey et al. 1994, 1996). Besides crocodyliforms, this formation has also yielded remains of various vertebrates, including hybodont sharks (Cuny et al. 2003), semionotiform fishes, turtle (Tong et al. 2005), diverse dinosaurs and pterosaurs (Buffetaut et al. 2003, 2005).

Material and methods Part of a crocodyliform skull, NRRU-A 1803 (NRRU stands for Nakhon Ratchasima Rajabhat University), lacking the premaxillae, external naris, quadratojugals, and ectopterygoids, was discovered at Ban Saphan Hin by P. Jintasakul, Director of the Museum of Petrified Wood and Mineral Resources of Nakhon Ratchasima, where the specimen is housed. Preparation was made at the Sirindhorn Museum, Kalasin Province by pneumatic air-pen and 10% formic acid solution bath. The sutures and surface of the specimen are well preserved in both dorsal and palatal views. No postcranial remains of this specimen were found.

Systematic palaeontology Crocodylomorpha Walker Crocodyliformes Hay Mesoeucrocodylia Whetstone & Whybrow Neosuchia Benton & Clark Khoratosuchus gen. nov. Etymology. ‘Khorat’ referring to an informal name of Nakhon Ratchasima Province, where the holotype was collected and Greek souchos, for crocodile. Type species. Khoratosuchus jintasakuli sp. nov. Diagnosis. A crocodyliform with an elongate and slender snout distinguished from other neosuchians in having the following autapomorphy: anterior ends of jugal and prefrontal terminating at the same level. It can also be separated from other neosuchians by the possession of the combination of the following apomorphies: secondary choana almost completely surrounded by pterygoids with palatines extending posteriorly and forming the anterior margins of the choana; maxilla relatively straight without lateral constrictions; homodont maxillary teeth. Khoratosuchus jintasakuli sp. nov. Holotype. NNRU-A 1803; partial skull represented by part of maxillae, part of nasals, lacrimals,

A NEW CRETACEOUS NEOSUCHIAN FROM THAILAND

prefrontals, frontal, postorbitals, squamosals, parietal, quadrates, occiput, palatines and part of pterygoids. Locality and horizon. Ban Saphan Hin, Amphoe Muang, Nakhon Ratchasima Province, Thailand; Khok Kruat Formation, late Early Cretaceous (Albian –Aptian). Etymology. ‘jintasakuli’ to honour P. Jintasakul, Director of the Museum of Petrified Wood and Mineral Resources of Nakhon Ratchasima, who allowed us to study the specimen. Diagnosis. Same as for the genus.

Description Skull Form and proportions. NNRU-A 1803 is a partial cranium with a total length of 191 mm (Fig.2a –d). Because the premaxillae and cranial part of the maxillae are not preserved, it is rather difficult to determine the original total length of the specimen. From the cranial end, the straight lateral margin of the rostrum widens gradually caudally. There is no evidence of contact between premaxillae and maxillae at the preserved tip of the specimen. It seems unlikely that the specimen was a short-snouted crocodyliform. A median suture on the caudodorsal surface of the frontal can be observed, indicating that the skull belongs to a young adult crocodyliform. In dorsal aspect, the surface of this specimen is moderately sculptured, particularly from the caudal margin of the skull to the rostral margin of the orbits. In front of the orbits, the rostrum is slightly sculptured on its surface. Its maximum width at the level of the cranial margins of the orbits is 75 mm. The cranial table of the specimen is nearly rectangular. The lateral part of the left postorbital, the left squamosal and the caudal part of the parietal are crushed. It seems that the left side of the specimen was compressed and rather extended laterally. However, its shape can be reconstructed by comparison with the right side. In occipital view, all bones, except the quadrates, are well preserved. Ventrally, the matrix was removed and the specimen shows clear sutural contacts as well as the secondary choana and maxillary teeth. The measurements of NNRU-A 1803 are given in Table 1. Maxilla. Large parts of the maxillae are preserved, although their cranial portions are missing. As preserved, they are 115 mm in length. Dorsally, the right maxilla is narrower than the left one, which is relatively flattened. The natural width of the maxilla is 11 mm, measured from the right side.

177

The medial surface of the maxilla touches the lateral surface of the nasal. This suture gradually leans caudolaterally and is separated from the cranial margin of the lacrimal at the point of the maximum width of the nasal bones. The suture exhibits a strong caudolateral bend and contacts the lateral margin of the lacrimal, and the cranial and lateral margins of the jugal in caudal view. The lateral margin of the maxilla is rather straight without any depression. There are several moderate grooves on its lateral surface. The ventral surface of the maxilla is flat. Medially, the palatine processes of the maxillae meet each other and run caudally to the cranial margin of the palatine process of palatines. The sutural contact gradually leans caudolaterally to the caudolateral edge of the maxilla. This suture also forms the cranial and craniolateral margins of the palatine and that of the suborbital fenestra. Thirteen maxillary alveoli are preserved on the right side of the specimen whereas 15 maxillary alveoli can be counted on the left side, but the last three maxillary alveoli are merged together. The alveoli shape is almost circular. Nasal. The nasal bones are preserved without their cranial parts. The lateral margin of the nasal exhibits a straight suture with the medial margin of the maxilla. This suture begins to curve caudomedially at the point of the maximum width of the nasal towards the medial margins of the lacrimal and prefrontal. Caudomedially, the terminal parts of the nasal unite together and form a wedge, which separates the nasal from the cranial margin of the frontal. Medially, the nasal bones meet each other in the sagittal plane. The maximum width of the nasal is 28 mm. Dorsally, the surfaces of the nasal bones are relatively smooth without ornamentation compared with the lacrimal and prefrontal. Lacrimal. The lacrimals are long and wide (i.e. 57 mm long and 17 mm wide). The cranial tip tapers cranially, separating the maxilla laterally from the nasal medially. From the caudal end to three-fifths of its length, the lacrimal is bordered by the prefrontal medially and jugal laterally. Its caudal margin shows a distinct concavity forming a notch in the cranial wall of the orbit. The left lacrimal has a slightly convex dorsal surface. Ventrally, the opening of the lacrimal duct is not visible, because of matrix covering. Prefrontal. Both prefrontals are 40 mm long and 10 mm wide. The prefrontal margins taper cranially towards the medial margin of the lacrimal. The caudolateral margin of the prefrontal forms the craniomedial wall of the orbit. One-third of its medial margin is articulated with the nasal rostrally and

178

K. LAUPRASERT ET AL.

Fig. 2. Holotype of Khoratosuchus jintasakuli, gen. et sp. nov., a nearly complete cranium (NNRU-A 1803), in (a) dorsal view; (b) palatal view; (c) occipital view; (d) lateral view. Scale bars represent 5 cm.

Table 1. Measurement of the cranium of Khoratosuchus jintasakuli gen. et sp. nov. 10

20

30

40

50

60

70

80

90

100

0000000100 0000?10000 0000000000 0?00000000 0000000000 000?100100 1001000000 000?000000 00000000?1 000?000000 0 ?00??0?0?? ?????00000 0??00?0000 0?????0000 ?0?0000000 ?00??0?1?? ??01?????0 ????0?0?0? 00000000?1 000?0?0000 0 100?000000 1?00000000 0000000000 0000?00010 001000?0?0 00000?00?0 0000000000 000?000010 00000000?0 1000000000 0

Ingroups Notosuchus Baurusuchus Libycosuchus Sebecus Pelagosaurus Teleosauridae Metriorhynchidae Mahajangasuchus Peirosauridae Hsisosuchus Trematochampsa Uruguaysuchus Malawisuchus Comahuesuchus Simosuchus Araripesuchus Alligatorium Theriosuchus Phoiidosaurus Calsoyasuchus Eutretauranosuchus “Goniopholis”lucasii Sunosuchus junggarensis Sunosuchus miaoi Siamosuchus phuphokensis Goniopholis simus G. baryglyphaeus Khoratosuchus jintasakulii Rugosuchus nonganensis Bernissartia

?12?001011 0200001001 ?10?001011 1200?01011 161?011011 061?011011 061?011011 1????????? 132?111021 122??????1 132?0010?1 031?0??01? 032?00?011 04??0??011 1320101010 1521011021 152101?13? 152101?031 161??11011 162?010031 132??1?0?1 132?01?031 132?010031 132??????1 132001?03? 132?011031 132?01?031 132??1???1 132?11?031 132?111031

0100101000 ??00?01000 ??00?01000 0000?01010 0011020111 0011000110 0011?20111 ?????????? 1?002?10?0 1?00?10000 00??201??0 ??00??1??1 0100?01100 1?00?????0 1?00?10000 0100201000 0000?00010 0100100010 ??11?01111 0210?0???0 0010201010 ??10?01010 1?10?01010 ?????010?0 0?00?01010 0010?01010 0010101010 00??1010?0 0010?01010 ??00?01010

1?010?0100 ??010?0100 0?010?0100 1?011?0?00 0110000010 0110000010 0110000010 ?????????? 01011?0100 0101?00100 01011?0100 ???????1?1 1001?011?0 10???????? 11010?0100 0001100100 00011?00?1 1101110001 01011?0000 0001?10000 01011?00?0 01011?0000 0101110000 00011?0000 0?0110000? 01011?0000 0101110000 0101100000 11011?0000 0?011?0000

1001?00000 ???1?1??21 ?001??00?1 00110100?1 0011000020 00010000?0 00010000?0 ?????????? 001101002? 0011000020 001?????21 0?11?00021 0011?00000 0001????2? 0012?10011 0?01000011 ?10???00?? 010100002? 1001100011 ???2?200?? 00020200?1 0001?000?1 0001120021 0001?200?? 0011????11 0011?000?1 0011?????1 0011?000?? 1011?00021 ?001??0011

021000110? 021000?10? 021010?1?? 021000?1?0 1110010011 11100110?1 11100110?1 ?????????? 02?00001?? 0100000??2 02?0?001?? ?????????? ?1??00???0 ?????????? ?1110000?0 02?000?100 0??000???0 0?1000??00 12101?111? ?????????? 021010?1?? 02101011?0 021010?102 0????????? 00?010?1?? 00??1??1?0 00?010?1?? ???010?1?? ?2101001?? 02?010?1??

?00??4?10? 02?114?101 0?0??4?001 0201041001 1111?20001 1111?2?001 1111?20011 ????14110? ?1110???0? ?01111?00? ?01??4???? ?000????0? ?10103100? 1?0?0???0? 010013110? 000104?001 0?0??3?010 0101?3?010 1201?20??? 1000?????? 110003?000 1200?3?01? 1200?31001 ??00????0? 12???????? 120013100? 120??3?01? ?201?????? ?20?02?01? 120??3?010

100?0??1?0 10010????? ?1000????0 11000????0 2100001010 2100011010 2100011010 ???010?111 1?1??????? 1?10000?0? 1????????? ?1001??1?? 101?1????? 11100????? 01000????1 1110101110 1?20100110 1120100110 2?0???11?0 ?12??????? 1120101??0 112010?1?0 11201??11? ?????????? ?12????110 112??????? 112??????0 ??0??????? 110??????? ?12010?1?0

0?0??????? 0????????? 0?0??????? ?00??????? 0000000101 000?000100 000?????1? 1000211??? ?????????? ????0?0000 100011???? ?????????? 100??????1 ?????????? 000?111?00 10001?0??? ????0?0010 011?000010 ??0??0???1 ?????????? ??0?00???? ?00?000?0? 1000000?01 ????00???? 1000000?00 ?????0???? 1000000?00 ?????????? ????01??10 0002011000

0??0?00001 0?1??00111 ??1??00??? 0010?001?1 0000001001 0000001001 000000100? 111111001? ?????11?1? 0?00000??0 1111?1100? ?11??0???? ??0??0???2 ??00??00?? 0?20?00002 0100111001 0000011001 0000011001 ??0??0???0 ??1??????1 0000011001 0000011001 ?????????? ?10??00??? ?00???1??1 1?0??11001 1?0??11??1 ??????1??1 1?0??0?100 0000011001

201?100000 2???1000?0 20????00?0 2001100000 1001000000 100?000000 1001000000 ?0??000??0 ????1000?0 100?000000 ??0?0000?0 111?2100?0 221?1100?0 20??0?00?0 21102100?0 10101000?0 20????0?00 1001000000 1000000000 1??1??1110 1001001110 1001002000 10?1??1010 ??011?1010 1?0?0?2??0 1?010?2?01 1?0?0?2??1 1?010?0?00 1000010000 1001000000

1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2

179

(Continued)

A NEW CRETACEOUS NEOSUCHIAN FROM THAILAND

Outgroups Protosuchus Hemiprotosuchus Orthosuchus

180

K. LAUPRASERT ET AL.

Distribution of the character-states for 101 characters among 34 Mesoeucrocodylia taxa and 3 outgroups considered in the analyses.

100?????2? ?2???????? 120??4???? ??0??1???1 ??00111010 0?01?????2 10??0?0??0 1001001121 ?21?00?1?? ?200?????? ?????????? ?????????? ??0??????1 2?010??000 1001001121 ?21?1??1?? 12000??01? ?1???????? ?????????? 0?00?110?1 ?211010000 100?001121 021?00?1?? ?200?????? ?????????? ?????????? ??00?????1 100?0?0?00 0101??00?0 01011?00?0 0?001?10?? 01011?00?0 1210?01010 1???101010 1?10101010 01??101010 132011??3? 1????????1 1120011011 1320111031

30

Neosuchia –Eusuchia transitional taxa Susisuchus Hylaeochampsa Iharkutosuchus Allodaposuchus

Table 1. Continued

10

20

40

50

60

70

80

90

100

? 3 3 3

two-thirds with the frontal caudally. In ventral view, the lateral portion of the right prefrontal pillar is exposed from the matrix. It is located medially at the cranial margin of the orbit and extends to the dorsal surface of the palate. Frontal. The dorsal surface of the frontal is slightly concave, particularly in the interorbital region. Its surface is also marked with deep pits. The lateral margins of the frontal taper rostrally and form a ‘W’-shaped sutural contact with the prefrontal and nasal. The frontal contacts the prefrontal craniolaterally with a concave suture and forms the medial wall of the orbit laterally. Caudolaterally, the frontal is articulated with the postorbital by a zigzagged suture and forms the craniomedial edge of the supratemporal fenestra. Its caudal margin extends into one-fourth of the supratemporal fenestra. The median suture on the caudodorsal surface is visible and disappears on the interorbital surface, indicating that this specimen probably was a young adult individual. In ventral view, two strong ridges form a craniocaudal groove, 5 mm wide, along the midline. Its cranial end is impossible to determine, because it is covered by matrix. Parietal. The left portion of the parietal is crushed at the left caudal margin. In dorsal view, the surface of the parietal is heavily sculptured, in a manner similar to the frontal. The cranial margin of the parietal contacts the frontal with a concave suture. Its lateral margin forms two-thirds of the depth of the medial and caudomedial portions of the supratemporal fenestra. Caudally, the parietal sends a lateral process to contact the squamosal. Its caudal margin is slightly concave at the midline of the skull and contacts the craniodorsal margin of the supraoccipital. Postorbital. The right postorbital is completely preserved, whereas the left one presents only its cranial portion and descending process. Its craniolateral corner makes the cranial table sub-rectangular in outline. The cranial margin of the postorbital forms the caudal edge of the orbit. In medial view, the postorbital sends a process to contact the frontal and caudally forms the cranial wall of the supratemporal fenestra. The curvature of its mediocaudal margin forms the craniolateral wall of the supratemporal fenestra. In caudal view, the postorbital contacts the squamosal with an oblique suture, which runs caudally from a craniolateral to caudomedial direction. In dorsal view, the surface of the right postorbital is slightly abraded. Consequently, it is not possible to observe its ornamentation. In ventral view, the robust and thick descending process of the postorbital is inset from the dorsal surface. Its diameter is c. 11 mm.

A NEW CRETACEOUS NEOSUCHIAN FROM THAILAND

Squamosal. The left squamosal is crushed and distorted whereas the right one is abraded on the craniodorsal surface. It is a triradiate bone that consists of cranial, medial and caudolateral processes. In dorsal view, the squamosal sends a cranial process, which is rather straight and slightly bent craniomedially, to contact the postorbital. In medial view, the cranial and medial processes form the caudomedial wall of the supratemporal fenestra. The medial process is bounded by the lateral margin of the parietal. In caudal view, its ventral margin is widely connected with the craniodorsal surface of the exoccipital. Laterally, the caudolateral process extends downward and overlaps with the otoccipital process caudally. The sculpture on the dorsal surface of the squamosal is similar to that of the parietal. Quadrate. Both quadrates show only the tips of their craniomedial ends that form the caudoventral part of the supratemporal fenestrae. Dorsally, the quadrate contacts the ventral surface of the parietal medially and squamosal laterally. Jugal. The right jugal is nearly complete, lacking only its caudal end. It is more complete than the left one. In lateral view, the jugal contacts the medial margin of the maxilla rostrally by an oblique suture, which runs rostrally from a caudoventral to craniodorsal direction. The jugal arch of the jugal is about twice as broad as the infratemporal bar. In dorsal view, the jugal is located laterally to the lateral margin of the cranial table. The jugal tapers cranially and wedges between the lacrimal and maxilla. The cranial tip of the jugal is at the same level as that of the prefrontal. Medially, the jugal contacts two-thirds of the lacrimal caudally and forms the lateral margin of the orbit. At the caudal end of the orbit, the jugal displays an unsculptured ascending process, which is inset from the lateral surface and leans lateromedially to form the ventral part of the postorbital bar. It is relatively flat mediolaterally and broad craniocaudally. A foramen forms a moderately elongate opening, 7 mm long, at the caudal end of the ascending process. The suture between the ectopterygoid and jugal is not visible. The ventral surface of the jugal is smooth. Supraoccipital. The supraoccipital is broadly rectangular in shape. Dorsally, it can be observed as a small semi-circular element on the skull roof that contacts the caudal margin of the parietal. In occipital view, its surface is almost completely abraded except on its lateral and ventral margins. The supraoccipital is bounded by the otoccipital laterally and ventrally whereas the dorsolateral margin of the supraoccipital is bordered by the parietal.

181

Exoccipital. The paired wings of the exoccipital slightly extend caudolaterally. In occipital view, the dorsal margin of the exoccipital contacts the ventral margin of the cranial table. The occipital surface, from dorsal to ventral, gradually protrudes caudally at an angle of c. 308 to the vertical axis. The ventral edge of the exoccipital forms a transversal ridge at the point beyond the dorsal margin of the occipital condyle. Its surface becomes nearly vertical at the level of the occipital condyle. The lateral margin of the right process is rounded whereas the left process is crushed. In lateral view, the exoccipital extends laterally and reaches the lateral margin of the squamosal. At its lateral tip, the dorsal surface of the otoccipital is overlapped by the ventrolateral surface of the squamosal. Medially, its dorsal surface is bordered by the ventral surfaces of the supraoccipital and parietal. Medially, the otoccipitals contact each other for 8 mm along the midline of the skull, separating the supraoccipital from the dorsal margin of the foramen magnum. Below this suture, the medial margin extends widely, is concave and forms the dorsal margin of the foramen magnum. The wedge-shaped process on the lateroventral margin of the foramen magnum is not evident. The suture between the exoccipital and basioccipital cannot be observed because the surface is worn. Basioccipital. The basioccipital surface is nearly vertical. It entirely forms the occipital condyle, which has a sub-circular articular facet. In occipital view, the bone is bounded by the foramen magnum dorsally and the exoccipital dorsolaterally. Its lateral margin is rather thick and robust. The ventral part of the occipital condyle presents a well-developed medial crest reaching the dorsal margin of the median Eustachian foramen. This crest separates two strong lateral depressions. The median Eustachian foramen can be observed in the caudoventral part of the basioccipital margin. It looks like a broad inverted ‘v’. The lateral eustachian tubes are visible laterally to the medial eustachian foramen at the ventrolateral margin of the basioccipital. They are dorsoventrally compressed and about 2 mm in lateromedial diameter. Foramina for cranial nerve are invisible, because the surface is covered by the matrix. Basisphenoid. In occipital view, the basisphenoid is exposed at the ventral margin of the basioccipital because the posterior margin of the pterygoid is lacking. Its dorsal margin forms the ventral wall of the medial eustachian foramen and of the lateral eustachian tubes. Ventrally, the caudal margin of the basisphenoid is concave cranially. The basisphenoid rostrum is not preserved.

182

K. LAUPRASERT ET AL.

Supratemporal fenestra. The supratemporal fenestrae are oval in outline. Their dorsal surface is placed at the same level as that of the cranial table. The medial wall of the fenestrae is formed by the contact of the lateral surface of the parietal, the craniodorsal surface of the quadrate, and the dorsolateral surface of the laterosphenoid. In dorsal view, the supratemporal fenestrae are bordered craniolaterally by the postorbital, craniomedially by the frontal, medially and caudomedially by the parietal, and caudolaterally by the squamosal. The size of their dorsal opening is nearly equal to that of the orbit. Infratemporal fenestra. Because the quadratojugal and the caudal portion of the jugal are lacking, a reconstruction of the infratemporal fenestra is not possible. However, the remains of the jugal indicate that the cranial part of the infratemporal fenestra was narrow dorsoventrally and elongate in shape. Secondary choana. The secondary choana is oval in shape, 29 mm long and 14 mm wide. It is bordered by the palatines cranially, for about 6 mm, whereas the lateral and caudal margins are formed by the pterygoids. There is no septum dividing the secondary choana. Pterygoid. In dorsal view, the cranial process of the pterygoid is a very thin bar. It extends craniodorsally to overlap the dorsal surface of the palatine ventrally. It also forms the caudal part of the dorsal roof of the narial passage. Laterally, the pterygoid flanges are broken. Caudally, the pterygoid forms the base of the braincase for half of its total length. In ventral view, only the medial surface of the pterygoid is preserved, which is about 39 mm in length and 25 mm in width, as the pterygoid wings are lacking. The pterygoid contacts the caudal margin of the palatine cranially whereas its posterior margin is broken, exposing the ventral part of the basisphenoid. Rostromedially, the pterygoid is concave caudally and forms the lateral and caudal walls of the secondary choana for about two-thirds of its length. Laterosphenoid. The laterosphenoids are partly preserved on both sides. Their dorsal parts widely contact the frontal and parietal medially and the postorbital laterally. In lateral view, the laterosphenoid forms the craniolateral part of the braincase. Caudally, the caudoventral margin of the laterosphenoid forms the rostral and rostrodorsal margins of a moderately large foramen, which corresponds to the passage for cranial nerve V. This foramen measures about 9 mm rostrocaudally and 6 mm dorsoventrally. The caudal part of the laterosphenoid also shows sutural connections with the

prootic caudoventrally and quadrate caudodorsally. The laterosphenoid bridge is not preserved in this specimen. Palatine. The palatine is elongate in shape, and 99 mm in length. Its ventral surface is smooth and rather convex at the middle of its length. Rostrally, it contacts the maxilla by a suture that extends caudolaterally to the rostromedial margin of the palatal fenestra. The caudal margin of the palatine contacts the pterygoid caudolaterally and the secondary choana caudomedially. In dorsal view, one-third of its length is overlain by the ventral surface of the pterygoid caudally. Cranially, at the end of the overlap, the palatine contacts the descending process of the prefrontal in its caudal portion. Suborbital fenestra. In ventral view, the suborbital fenestra tapers cranially. Its cranial and lateral margins are formed by the palatal process of the maxilla whereas its medial margin is bordered by the palatal process of the palatine. The caudal margins of these openings are lacking on both sides, because the ectopterygoids are not preserved. The width of the cranial margin of a suborbital fenestra is approximately half the width of its caudal portion. Dentition. Most of the cranial region of the maxillae and the whole premaxillae are missing. Therefore only 15 alveoli can be counted in each maxilla from the caudal end of the maxilla forward. Only the bases of the crowns of some teeth are preserved. They are all circular in cross-section and show a size similar to that of the alveoli. The mesiodistal diameter of each alveolus is about 4 mm, and they are separated from each other by the same distance. The tooth row is straight.

Comparison and discussion A phylogenetic analysis was carried out, based on 101 characters, to resolve the position of Khoratosuchus. Character 1 and characters 3–100 were taken from the data matrix of Lauprasert et al. (2007), and character 2 was modified from Sereno et al. (2001) and Wu et al. (2001). One additional character was included in this analysis (character 101). All characters were treated as unordered. The data matrix contains 31 taxa of Mesoeucrocodylia and three outgroup taxa, which consist of Protosuchus, Hemiprotosuchus, and Orthosuchus. Pholidosaurus and Rugosuchus were added to this study to determine the affinities of long-snouted crocodyliform and advanced neosuchians, respectively. The data matrix was run using PAUP, Version 4.0b10 for 32-bit Microsoft Windows, using a random,

A NEW CRETACEOUS NEOSUCHIAN FROM THAILAND

183

Fig. 3. (a) Strict consensus tree of 66 equally most parsimonious trees (length ¼ 316 steps, consistency index (CI) ¼ 0.418, and retention index (RI) ¼ 0.625) derived from the analysis of 102 characters in 31 taxa of Mesoeucrocodylia and three outgroup taxa, which consist of Protosuchus, Hemiprotosuchus and Orthosuchus. (b) Strict consensus tree after added an advanced Neosuchia, Susisuchus, and three more Eusuchia taxa, Hylaeochampsa, Iharkutosuchus and Allodaposuchus. The result shows 9855 equally most parsimonious trees (length ¼ 326 steps, consistency index (CI) ¼ 0.405, and retention index (RI) ¼ 0.616).

184

K. LAUPRASERT ET AL.

stepwise addition, heuristic search algorithm. There are 99 parsimony-informative characters and four parsimony-uninformative characters. The strict consensus tree (Fig. 3a) shows that Khoratosuchus jintasakuli, Atoposauridae, Goniopholididae, Rugosuchus, Bernissartia, the Eusuchia and the non-eusuchian longirostrine forms (Clark 1994; Wu et al. 1997, 2001), illustrated by Pelagosaurus, Teleosauridae, Metriorhynchidae and Pholidosaurus, form an unresolved monophyletic group, representing the Neosuchia (Benton & Clark 1988). All members of this clade share an unambiguous diagnostic character, which is the dorsal part of the postorbital possessing anterior and lateral edges (28[0]). Although the maxilla of Khoratosuchus jintasakuli is slender and long, it does not appear to be closely related to the non-eusuchian longirostrine forms. On the contrary, K. jintasakuli can be rejected from the non-eusuchian longirostrine forms on the basis of the following characters. The rostrum of Khoratosuchus is wider than high, longer than the remainder of the skull and broadening caudally (2[3], 3[2]) whereas that of the longirostrine forms is nearly tubular in shape and is also much longer than the remainder of the skull and parallel-sided (2[6], 3[1]). In addition, the secondary choana of the Thai specimen is bordered by the pterygoids except its most anterior margin, which is bordered by the caudal margin of the palatines (101[2]). In non-eusuchian longirostrine forms, the secondary choana is entirely bordered by the palatines or half by the palatines and half by the pterygoids (101[1]). The position of the secondary choana of K. jintasakuli is an important diagnostic character that precludes this species from belonging to the families Atoposauridae or Goniopholididae (Wu et al. 1997), which share this feature with the non-eusuchian longirostrine forms. It also suggests that Khoratosuchus is phylogenetically closer to the derived neosuchians and the Eusuchia than to the family Goniopholididae (sensu Wu et al. 2001). Additionally, the postorbital bar of K. jintasakuli is displaced medially, which is an unusual character in primitive crocodiles (E. Buffetaut, pers. comm.), suggesting that the Thai neosuchian is an advanced form. Currently, the Eusuchia are diagnosed as a monophyletic group based on a single synapomorphy (Clark 1986; Ortega et al. 2000; Wu et al. 2001), which is that the internal choana is entirely situated within the pterygoids. The craniomedial tip of the internal choana of the Thai neosuchian is bounded by the palatines. This character indicates that K. jintasakuli is not a member of the Eusuchia. The new crocodyliform from Thailand is comparable with derived neosuchians such as

Bernissartia fagesii Dollo 1883 from the Early Cretaceous of Europe (Buscalioni & Sanz 1990, fig. 2); Shamosuchus spp. from the Late Cretaceous of Mongolia (Efimov 1988; figs 17 and 19); and an unnamed taxon (¼ ‘Glen Rose form’ of Clark in Benton & Clark 1988) described from two skulls from the Early Cretaceous of North America (Langston 1973, fig. 6E), on the basis of the position of the secondary choana, which is mostly bounded by the pterygoids. K. jintasakuli was compared with Shamosuchus spp. and the ‘Glen Rose form’ to understand the phylogenetic position of the Thai new crocodyliform, although these two taxa were not included in the ingroups of our phylogenetic analysis, as their descriptions are inadequate. The skull of K. jintasakuli also resembles that of a Chinese crocodyliform, Rugosuchus nonganensis from the late Early Cretaceous of the Nenjiang Formation, Jilin Province, NE China, which shows comparable cranial features (Wu et al. 2001, figs 3A, B and 4). Both of them have an elongate and slender skull, and the lateral margins of their maxillae are relatively straight. However, Khoratosuchus can be separated from Rugosuchus by the following cranial characters: (1) the maxillary teeth were homodont based on the sub-circular shape of the base of the preserved crowns; (2) the dorsal surfaces of the frontal and parietal do not bear a median ridge; (3) the dorsal surface of the maxillae lacks a series of fossae; (4) the suborbital fenestrae are larger whereas the interfenestral region of the palatines is much smaller; (5) the secondary choana is almost entirely bounded by the pterygoids; (6) the supraoccipital has no vertical ridge along the midline of its occipital surface. In addition to the secondary choana of K. jintasakuli being more caudal in position than seen in Bernissartia fagesii, and Shamosuchus spp., the following cranial characters can distinguish Khoratosuchus from those taxa: (1) the skull of K. jintasakuli is more slender than that of these two taxa, which are relatively broader and stouter; (2) the dorsal surface of the frontal and parietal lacks a median ridge along the midline whereas B. fagesii has a median ridge on the concave frontal surface; the frontal and parietal of Shamosuchus spp. are often partially ridged along the dorsal midline; (3) the lateral margin of the maxilla is relatively straight without any strong constrictions as commonly seen in B. fagesii and Shamosuchus spp. The following additional characters distinguish K. jintasakuli from B. fagesii: the interorbital region is broader than the interfenestral region; the occipital surface of the supraoccipital is smooth; the ventral border of the orbit is concave and does not rise as a rim-like edge. The jugal of K. jintasakuli lacks a longitudinal ridge on the lateral surface and the interfenestral region of the palatines

A NEW CRETACEOUS NEOSUCHIAN FROM THAILAND

is longer and much narrower than that of Shamosuchus spp. These features also exclude the Thai specimen from Shamosuchus spp. Only the posteroventral view of the skull of the ‘Glen Rose Form’ is available for comparison (Langston 1973, fig. 6E). Except for the position of the internal choana, which can be compared, the other features, as mentioned above, are unknown for the ‘Glen Rose Form’. According to the above comparisons, it is clear that the new Thai neosuchian crocodyliform cannot be referred to any previously known taxa. A new taxon, Khoratosuchus jintasakuli, is therefore erected on the basis of the combination of the following cranial features: the relatively straight and elongate lateral margin of the maxilla; the absence of a median ridge along the flat dorsal surface of the frontal and parietal; the secondary choana almost entirely situated within the pterygoids; the relatively flat, wider than long snout. This combination of characters also suggests that K. jintasakuli may be closer to the Eusuchia than the Chinese derived neosuchian, Rugosuchus. To determine the phylogenetic position of Khoratosuchus and better understand its palaeobiogeographical significance, some Neosuchia –Eusuchia transitional taxa, that is, Hylaeochampsa, Iharkutosuchus, Allodaposuchus and an advanced Neosuchia from Gondwana, Susisuchus (Salisbury et al. 2006; ´´ si et al. 2007), were used to replace the Eusuchia O taxa in the previous data matrix. Unfortunately, the strict consensus tree of the latter analysis not only left unresolved the position of Khoratosuchus, but also created polytomies at several levels in the tree (Fig. 3b). Many internal nodes in the cladogram collapsed. Susisuchus, for instance, forms a polytomy clade with the Eusuchia (Hylaeochampsa, Iharkutosuchus and Allodaposuchus). Therefore, the results from both phylogenetic analyses, at the moment, cannot be used to determine the biogeographical occurrence of the neosuchian–eusuchian transition. The addition of Khoratosuchus to the eusuchian phylogenetic analysis of Brochu (2004) or Salisbury et al. (2006) in a further analysis will probably allow us to understand the phylogenetic relationships of this Thai basal neosuchian.

Conclusions Khoratosuchus jintasakuli is the first crocodyliform from the Aptian– Albian Khok Kruat Formation to be described from a skull and also represents the most advanced form among all the crocodyliforms hitherto reported from the Khorat Group. Previous studies have already described three taxa of Thai neosuchians; that is, ‘Sunosuchus’ thailandicus (Buffetaut & Ingavat 1980) from the Phu Kradung Formation, and ‘Goniopholis’ phuwiangensis

185

(Buffetaut & Ingavat 1983) and Siamosuchus phuphokensis (Lauprasert et al. 2007) from the Sao Khua Formation. The discovery of K. jintasakuli not only increases the diversity of the neosuchian crocodilians from the the Khorat Group, but also represents the most advanced form of noneusuchian crocodilian known from SE Asia. The authors would like to thank Dr Mueller-To¨we and Dr S. Salisbury for their suggestions as reviewers, and Dr P. Jintasakul (Museum of Petrified Wood and Mineral Resources of Nakhon Ratchasima) for providing access to valuable specimens during this work. We thank Dr P. Lauprasert (Mahasarakham University, Thailand), Dr ´´ si J. Claude (University of Montpellier, France), Dr A. O (Eo¨tvo¨s Lora´nd University, Budapest), and all the staff of Sirindhorn Museum for their useful suggestions, comments and encouragement. This work has been supported by the Royal Golden Jubilee PhD Program (PHD/0069/ 2546), Thailand Research Fund, and the CNRS-TRF special programme for Biodiversity Research and Training Programme (BRT/BIOTEC/NSTDA) Grant BRT R_245007. G.C. acknowledges Danish Research Council and Carlsberg Foundation grants to support his work in Thailand. K.L. acknowledges Franco-Thai Cooperation Programme in Higher Education and Research Year 2007.

Additional characters of the phylogenetic analysis Characters 1 and 3 –100 were taken from the data matrix of Lauprasert et al. (2007) and character 2 was modified from Sereno et al. (2001). Character 101 was added. All characters were treated as unordered. Coding for character states: 0 (ancestral), 1, 2 and 3 (derived), ‘?’ (unknown state). 2. (Modified from Sereno et al. 2001; Wu et al. 2001). Snout wider than high, shorter than remainder of skull and parallel-sided (0), snout wider than high, shorter than remainder of the skull and broadening caudally (1), snout wider than high, longer than remainder of the skull and parallel-sided (2), snout wider than high, longer than remainder of the skull and broadening caudally (3), snout very much wider than high, longer than remainder of skull and parallel-sided (4), snout much wider than high, wider than long and broadening caudally (5), or snout nearly tubular, much longer than remainder of the skull and parallel-sided (6). 101. (Modified from Wu et al. 2001). Internal choana bordered by maxillae only (0), or either by palatines or half palatine and half pterygoid (1), or by almost all of pterygoids (2), or by pterygoids only (3).

References B ENTON , M. J. & C LARK , J. M. 1988. Archosaur phylogeny and the relationships of the Crocodilia. In: B ENTON , M. J. (ed.) The Phylogeny and Classification of the Tetrapods, Vol. 1. Clarendon Press, Oxford, 295– 338.

186

K. LAUPRASERT ET AL.

B ROCHU , C. A. 2004. Alligatorine phylogeny and the status of Allognathosuchus Mook, 1921. Journal of Vertebrate Paleontology, 24, 857–873. B UFFETAUT , E. & I NGAVAT , R. 1980. A new crocodilian from the Jurassic of Thailand, Sunosuchus thailandicus n. sp. (Mesosuchia, Goniopholididae), and the palaeogeographical history of South-East Asia in the Mesozoic. Geobios, 13, 879–889. B UFFETAUT , E. & I NGAVAT , R. 1983. Goniopholis phuwiangensis nov. sp., a new mesosuchian crocodile from the Jurassic of northeastern Thailand. Geobios, 16, 79– 91. B UFFETAUT , E. & S UTEETHORN , V. 1998. Early Cretaceous dinosaurs from Thailand and their bearing on the early evolution and biogeographical history of some groups of Cretaceous dinosaurs. New Mexico Museum of Natural History and Science Bulletin, 14, 205– 210. B UFFETAUT , E. & S UTEETHORN , V. 1999. The dinosaur fauna of the Sao Khua Formation of Thailand and the beginning of the Cretaceous radiation of dinosaurs in Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 150, 13–23. B UFFETAUT , E., S UTEETHORN , V. ET AL . 2003. Dinosaurs in Thailand. Mahasarakham University Journal, Special Issue, 22, 69– 82. B UFFETAUT , E., S UTEETHORN , V., L E L OEUFF , J., K HANSUBHA , S., T ONG , H. & W ONGKO , K. 2005. The dinosaur fauna from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 575–581. B USCALIONI , A. D. & S ANZ , J. L. 1990. The small crocodile Bernissartia fagesii from the Lower Cretaceous of Galve (Teruel, Spain). Bulletin de l’Institut Royal des Sciences Naturelles de Belgique, 60, 129–150. C APPETTA , H., B UFFETAUT , E. & S UTEETHORN , V. 1990. A new hybodont shark from the Lower Cretaceous of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 11, 659 –666. C ARTER , A. & B RISTOW , C. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau Basin, eastern Thailand. Basin Research, 15, 271–285. C AVIN , L., S UTEETHORN , V., K HANSUBHA , S., B UFFETAUT , E. & T ONG , H. 2003. A new semionotid (Actinopterygii, Halecostomi) from the Late Jurassic– Early Cretaceous of Thailand. Comptes Rendus Pale´vol, 2, 291– 297. C AVIN , L., B UFFETAUT , E. ET AL . 2004. A new fish locality from the continental Late Jurassic–Early Cretaceous of north-eastern Thailand. Revue de Pale´obiologie,Volume Spe´cial, 9, 161–167. C LARK , J. M. 1986. Phylogenetic relationships of the Crocodylomorph Archosaurs. PhD thesis, University of Chicago. C LARK , J. M. 1994. Patterns of evolution in Mesozoic Crocodyliformes. In: F RASER , N. C. & S UES , H. D. (eds) In the Shadow of the Dinosaurs: Early Mesozoic

Tetrapods. Cambridge University Press, Cambridge, 84–97. C UNY , G., S UTEETHORN , V., B UFFETAUT , E. & P HILIPPE , M. 2003. Hybodont sharks from the Mesozoic Khorat Group of Thailand. Mahasarakham University Journal, Special Issue, 22, 49–68. C UNY , G., S UTEETHORN , V., K AMHA , S., B UFFETAUT , E. & P HILIPPE , M. 2006. A new hybodont shark assemblage from the Lower Cretaceous of Thailand. Historical Biology, 18, 21– 31. D OLLO , M. L. 1883. Premie`re note sur les crocodiliens de Bernissart. Bulletin de l’Institut Royal des Sciences Naturelles de Belgique, 2, 309– 338. E FIMOV , M. B. 1988. On the fossil crocodiles of Mongolia and the USSR. Trudy Sovmestnoi SovetskoMongol’skoi Paleontologicheskoi ekspeditsii, 34, 81–90 [in Russian]. L ANGSTON , W. 1973. The crocodilian skull in historical perspective. In: G ANS , C. & P ARSONS , T. (eds) Biology of the Reptilia, Vol. 4. Academic Press, London, 263–284. L AUPRASERT , K., C UNY , G., B UFFETAUT , E., T HIRAKHUPT , K. & S UTEETHORN , V. 2007. Siamosuchus phuphokensis, a new goniopholidid from the Early Cretaceous (ante-Aptian) of northeastern Thailand. Bulletin de la Socie´te´ Ge´ologique de France, 178, 201 –216. M EESOOK , A. 2000. Cretaceous environments of Northeastern Thailand. In: O KADA , H. & M ATEER , N. J. (eds) Cretaceous Environment of Asia. Elsevier, Amsterdam, 207– 223. O RTEGA , F., G ASPARINI , Z., B USCALIONI , A. D. & C ALVO , J. O. 2000. A new species of Araripesuchus (Crocodylomorpha, Mesoeucrocodylia) from the Lower Cretaceous of Patagonia (Argentina). Journal of Vertebrate Paleontology, 20, 57– 76. ´´ SI , A., C LARK , J. M. & W EISHAMPEL , D. B. 2007. O First report on a new basal eusuchian crocodyliform with multicuspid teeth from the Upper Cretaceous (Santonian) of Hungary. Neues Jahrbuch fur Geologie und Palaontologie, Abhandlungen, 243, 169–177. R ACEY , A., G OODALL , J. G. S., L OVE , M. A., P OLACHAN , S. & J ONES , P. D. 1994. New age data for the Mesozoic Khorat Group of northeastern Thailand. In: A NGSUWATHANA , P. (ed.) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245– 252. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. Journal of Petroleum Geology, 19, 5 –40. S ALISBURY , S. W., M OLNAR , R. E., F REY , E. & W ILLIS , P. M. A. 2006. The origin of modern crocodyliforms: New evidence from the Cretaceous of Australia. Proceedings of the Royal Society of London, Series B, 273, 2439–2448. S ERENO , P. C., L ARSSON , H. C. E., S IDOR , C. A. & G ADO , B. 2001. The Giant Crocodyliform Sarcosuchus from the Cretaceous of Africa. Science, 294, 1516– 1519.

A NEW CRETACEOUS NEOSUCHIAN FROM THAILAND T ONG , H., B UFFETAUT , E. & S UTEETHORN , V. 2003. Mesozoic turtles of Thailand. Mahasarakham University Journal, Special Issue, 22, 41–48. T ONG , H., S UTEETHORN , V., C LAUDE , J., B UFFETAUT , E. & J INTASAKUL , P. 2005. The turtle faunas from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 610 –614.

187

W U , X. C., S UES , H. D. & D ONG , Z. M. 1997. Sichuanosuchus shuhanensis, a new ?Early Cretaceous protosuchian (Archosauria: Crocodyliformes) from Sichuan (China), and the monophyly of Protosuchia. Journal of Vertebrate Paleontology, 17, 89–103. W U , X. C., W U , Z. W. & R USSELL , A. P. 2001. Cranial anatomy of a new crocodyliform (Archosauria: Crocodylomorpha) from the Lower Cretaceous of Song-Liao Plain, northeastern China. Canadian Journal of Earth Science, 38, 1653–1663.

A new skeleton of Phuwiangosaurus sirindhornae (Dinosauria, Sauropoda) from NE Thailand S. SUTEETHORN1 *, J. LE LOEUFF2, E. BUFFETAUT3, V. SUTEETHORN4, C. TALUBMOOK5 & C. CHONGLAKMANI6 1

Institut des Sciences de l’Evolution de Montpellier, Universite´ Montpellier 2, 34095 Montpellier Cedex 5, France 2

Muse´e des Dinosaures, 11260 Espe´raza, France

3

CNRS, UMR 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure, 24 rue Lhomond, 75005 Paris, France 4

Bureau of Fossils Research and Geological Museum, Department of Mineral Resources, Bangkok, Thailand

5

Faculty of Science, Maha Sarakham University, Tambon Kamriang, Kantarawichai District, Mahasarakham 44150, Thailand 6

School of Geotechnology, Institute of Engineering, Suranaree University of Technology, 111 University Avenue, Muang District, Nakhon Ratchasima 30000, Thailand *Corresponding author (e-mail: [email protected]) Abstract: A new skeleton of a sauropod dinosaur has been discovered in the Early Cretaceous Sao Khua Formation at Ban Na Khrai in Changwat Kalasin (NE Thailand). All sauropod bones from Ban Na Khrai share all their characteristics with the type specimen of Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994. The 60% complete skeleton is very well preserved and includes cranial elements (a tooth, a frontal, a postorbital, a squamosal, both quadrates, and the braincase), whereas the type specimen is only 10% complete and consists of postcranial bones only. The material from Ban Na Khrai belongs to a single subadult individual of Phuwiangosaurus, as attested by the unfused neurocentral sutures of the vertebrae, which are firmly fused and larger in size in the holotypic specimen. Supplementary material: Catalogue numbers of Ban Na Khrai specimens are available at http://www.geolsoc.org.uk/SUP18348.

Thailand has an especially good record of sauropods, with Triassic, Jurassic and Cretaceous fossils from many localities across the country. The first sauropod identified in Thailand was Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994 from the Sao Khua Formation (Early Cretaceous). The holotype consists of postcranial elements only and the species was not assigned to a precise sauropod family by Martin et al. (1994). According to cladistic analyses, based on the post-cranial skeleton only, P. sirindhornae represents one of the most basal titanosaurs (Upchurch 1998; Upchurch et al. 2004; Curry Rogers 2005). Slender teeth and jaw elements were subsequently discovered associated with P. sirindhornae skeletons at Phu Kum Khao (Suteethorn et al. 1995; K4 or ‘Wat Sak Kawan’ locality; see below for details), and were tentatively assigned to this form. They closely resemble those

of Nemegtosaurus mongoliensis Nowinski 1971, a Late Cretaceous sauropod from Mongolia, thus Buffetaut & Suteethorn (1999) suggested that P. sirindhornae belongs to the family Nemegtosauridae. However, more evidence is needed to confirm that this cranial material does belong to P. sirindhornae and that this form belongs to the Nemegtosauridae, as it is clear that two different sauropod species are present at Phu Kum Khao (Buffetaut et al. 2002; Buffetaut & Suteethorn 2004). The discovery of a partly articulated skeleton at Ban Na Khrai, representing about 60% of a complete skeleton and including a braincase and cranial elements, provides the first opportunity to confirm the link between cranial and post-cranial elements suggested by Buffetaut & Suteethorn (1999). The objective of this paper is to briefly describe the new sauropod skeleton from Ban Na Khrai and to compare it with the holotype of P. sirindhornae.

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 189–215. DOI: 10.1144/SP315.14 0305-8719/09/$15.00 # The Geological Society of London 2009.

190

S. SUTEETHORN ET AL.

Abbreviations and Thai vocabulary CNRS, Centre National de la Recherche Scientifique, France; DMR, Department of Mineral Resources, Bangkok, Thailand; K1 –K16, localities in Changwat Kalasin; PW1 –PW5, localities in Phu Wiang area, Changwat Khon Kaen; SM, Sirindhorn Museum, Changwat Kalasin, Thailand (the specimens housed at the Sirindhorn Museum are identified by the acronym SM). In the present paper, the appellations of the localities are in Thai; they consist of specific Thai names and prefixes, including ‘Amphoe’, ‘Changwat’, ‘Wat’, ‘Phu’ and ‘Ban’. The meanings of these words are as follows: ‘Amphoe’ means district (e.g. Amphoe Phu Wiang refers to the district of Phu Wiang); ‘Changwat’ means province (e.g. Changwat Kalasin refers to Kalasin province); ‘Wat’ means temple or monastery (e.g. Wat Sak Kawan refers to the temple named Sak Kawan); ‘Phu’ means mountain or hill (e.g. Phu Kum Khao refers to Kum Khao hill); ‘Ban’ means village (e.g. Ban Na Khrai refers to Na Khrai village).

Historical review of sauropod discoveries in the Sao Khua Formation of Thailand The first dinosaur bone from Thailand was found in 1976 by S. Yaemniyom, a geologist from the DMR in the course of a uranium survey at Phu Wiang (Amphoe Phu Wiang, Changwat Khon Kaen). It is a fragment of a sauropod femur from the Sao

Khua Formation (Ingavat & Taquet 1978; Ingavat et al. 1978). Since then, a collaboration led by the Department of Mineral Resources (Thailand) and the Centre National de la Recherche Scientifique (France) has led to the discovery of many other sauropod remains (see Buffetaut et al. 2002, for a recent summary). The Sao Khua Formation has yielded more dinosaur remains than any other formation on the Khorat Plateau, and most of them belong to sauropods. Sauropod remains occur both as more or less complete articulated skeletons in Phu Wiang 1 (PW1), Phu Kum Khao (K4) and Ban Na Khrai (K11) and as accumulations of disarticulated bones (Phu Wiang 5 (PW5), Phu Pha Ngo (K1), Phu Peng (K16)) (Fig. 1).

Changwat Khon Kaen In 1981, an incomplete sauropod femur was discovered on Phu Wiang (PW1A). The year after, a partly articulated sauropod skeleton associated with theropod teeth was found at PW1 (Fig. 1), a few metres from the first locality (Buffetaut & Suteethorn 1989). This partial skeleton became the type of Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994. In 1989, seven cervical vertebrae of a sauropod were found in connection at another site on Phu Wiang (PW2). An attempt was made to preserve them in situ, but unfortunately an efficient protection proved impossible and the specimen was almost completely destroyed before it could be studied (Martin et al. 1999).

Fig. 1. (a) Fossil sites that have yielded Phuwiangosaurus sirindhornae plotted on a map of the Cretaceous formations (black areas) of Thailand. (b) Map of Changwat Kalasin, showing the localities that have yielded material of P. sirindhornae. PW1, Phu Wiang 1; K1, Phu Pha Ngo; K2, Ban Nong Mek; K4, Phu Kum Khao; K5, pond at Phu Pha Ngo; K11, Ban Na Khrai; K16, Phu Peng.

NEW PHUWIANGOSAURUS SKELETON

In 1990, many small bones were found in a forest on Phu Wiang (PW4) by schoolboys, and given to the Department of Mineral Resources by their teacher. These small bones were the first remains of juvenile sauropods to be found in Thailand (Martin et al. 1993, 1999). Then in 1991 and 1992, more sauropod remains were discovered, including many bones of juveniles (PW5, PW5A). The juvenile sauropods belonged to the same species as the adult remains they were associated with, P. sirindhornae (Martin 1994).

Changwat Kalasin In 1980, a sauropod humerus was discovered at Wat Sak Kawan in Amphoe Sahat Sakhan. The monks used to collect archaeological objects and pottery, and they thought that this bone was petrified wood (Martin et al. 1993, and see below). The first important finds in the Kalasin region were made at Phu Pha Ngo and Phu Kum Khao localities (K1 and K4, respectively). In January 1991, 200 km west of Phu Wiang, the Thai– French team discovered about 40 bones and many other bone fragments at K1 locality near the temple of Wat Pha Sawan, which is located at the foot of Phu Pha Ngo in Amphoe Kuchi Narai. Most of the bones were damaged and broken into small pieces. These bones had been placed there by local workers during the erection of the temple and were thought to be elephant bones. The Thai– French team excavated 60 m2 from March to April 1991, and found only 10 more bones, including bone fragments, as well as two crocodile teeth, fish scales, and turtle shell elements. The material from K1 can be assigned to at least three individuals of P. sirindhornae, on the basis of identifiable specimens, such as three right pubes (Martin et al. 1999). In the same year, 10 km south of Phu Pha Ngo, the Thai– French team found a few vertebrae and ribs at Ban Nong Mek locality (K2). Martin et al. (1999) referred these specimens to P. sirindhornae. In 2005 the villagers at Ban Nong Mek found more bones, which are kept at Wat Nong Mek. These specimens, including a femur, sacral and caudal vertebrae, belong to P. sirindhornae. In September 1994, geologists from the Department of Mineral Resources discovered a remarkable locality at Phu Kum Khao where fossil bones had first been found by P. V. Sahaskhunthe, the head of the monks in Wat Sak Kawan, many years before (see above). The bone-bed is located at the base of Phu Kum Khao. The excavation started on 19 November 1994. More than 600 specimens were discovered in one year. They consist of wellpreserved, partly articulated remains of several individuals; jaw elements were found for the first time in association with skeletons of P. sirindhornae

191

(Suteethorn et al. 1995; Martin et al. 1999) and a second undescribed sauropod species was also recognized (Buffetaut et al. 2002). The ‘Phu Kum Khao’ locality (K4) was mentioned as the ‘Wat Sakawan’ or ‘Wat Sak Kawan’ locality according to the transliteration system used, both names referring to the temple name ‘Sak KaWan’ where the fossils were found and placed (Suteethorn et al. 1995; Buffetaut & Suteethorn 1999; Martin et al. 1999). In 1998–1999, the ‘Phu Kum Khao Dinosaur Research Centre’ was established, 200 m from the excavation site (K4), and the name of this site was changed to ‘Phu Kum Khao’, which is the name of the mountain where the excavation site is located (Buffetaut et al. 2002, 2003; Buffetaut & Suteethorn 2004). Since 2002, the Department of Mineral Resources has built a palaeontological museum near the Phu Kum Khao Dinosaur Centre, which houses most of the palaeontological collections of Thailand. The museum opened in 2007 under the name of ‘Sirindhorn Museum’, in honour of Her Royal Highness Princess Maha Chakri Sirindhorn, Princess of Thailand. The material described in the present paper is housed at the Sirindhorn Museum. Several years after the excavation at K1 locality (Fig. 1), S. Suwantri and villagers at Ban Na Khrai (300 m from K1 on the same hillside and in the same bone-bed layer of the Sao Khua Formation) discovered some bones while digging a pond. At that time a team from the DMR surveyed the site (K5) and found only a few bones. Later, the villagers enlarged the pond to the north of K5 site. They found many bones on the opposite side of the pond 50 m from K5. This new site is referred to as K11. Students brought one of the large bones, a part of a left femur (SM K11-0151), to Wat Pha Sawan. Then in 1998, the Thai–French expedition team went to Wat Pha Sawan and surveyed this site. They found more than 10 bones (one femur, one fibula, one ischium, one rib, and several vertebrae). A 17 day excavation was then begun on 13 March 1998 under the direction of one of us (V. S.), yielding about 100 bones. Remaining bones were collected on 14–17 May 1998, and all the remains were brought to the Dinosaur Research Centre in Phu Kum Khao. This material is described for the first time in the present paper. Finally, on 15 February 2001, villagers in Ban Na Sombun found some bone fragments on an outcrop of the Sao Khua Formation at the bottom of Phu Peng (K16) and brought them to Wat Phu Peng (Amphoe Sahat Sakhan, Changwat Kalasin). An excavation was undertaken over 10 m2 and more than 25 bones were recovered. Most of them belong to one or several adult sauropods and some to a juvenile or small sauropod (Buffetaut et al. 2002). These bones show some differences from P. sirindhornae and will be described later.

192

S. SUTEETHORN ET AL.

Geological setting The specimens from K11 were collected over a 25 m2 excavation area (Fig. 2). The bone-bed layer is 2 m below the ground level. The fossilbearing layer consists of brownish red sandstone, siltstone and claystone. The layer is interbedded

with lime nodule conglomerate. The sediments are characteristic of the Sao Khua Formation of the Khorat Group, which is composed of various cycles of reddish brown silty claystones interbedded with siltstones, fine- to medium-grained sandstones and conglomerates, and caliches, calcrete nodules, and thin-bedded and nodular silcretes (Meesook

Fig. 2. Excavation map of Ban Na Khrai Phuwiangosaurus sirindhornae skeleton (drawing by A. Kumchu). Scale bar: 1 m.

NEW PHUWIANGOSAURUS SKELETON

et al. 2002). Based on palynological data, Racey & Goodall (2009) gave an Early Cretaceous (Barremian –Aptian) age for the Sao Khua Formation. The red siltstones are floodplain deposits; the area can be interpreted as a wide floodplain environment with low-energy meandering rivers and semiarid climate with two distinct seasons. This is confirmed by the co-occurrence of articulated dinosaur skeletons or parts of skeletons together with remains of aquatic animals, such as fish scales and fish vertebrae (Buffetaut & Suteethorn 1998; Martin et al. 1999).

Taphonomy Of 167 labelled specimens from K11, 129 are elements of the axial skeleton that fit in size. Most of them are unfused elements, such as centra, neural arches, and ribs. There are only five cervical vertebrae with fused neural arch and centrum, although their ribs are sometimes unfused. The presence of 11 right cervical ribs, 11 right dorsal ribs and five sacral centra show that this sauropod possessed at least 13 cervical (including atlas and axis), 12 dorsal and five sacral vertebrae (see details below). The K11 specimen has 34 caudal vertebrae (out of a total of more than 54 present in an articulated Phuwiangosaurus skeleton from K4, Phu Kum Khao). There is an important variation in the shape of sauropod vertebrae and it can be demonstrated that there is no overlapping in the vertebrae of K11. Furthermore, the vertebral and pelvic remains were found partly in articulation. There are 11 right dorsal ribs that were still in situ, in the same position as when the sauropod carcass was buried (Fig. 2). The posterior dorsal and sacral centra were connected and the right pelvic girdle (pubis, ischium and ilium) lay below the sacral vertebrae. Most cervical vertebrae and eight skull bones were discovered in the same area but

193

they were not articulated as the sacral vertebrae were. The tail elements were dispersed all over the 25 m2 of the excavation, with two concentrations east and west of the sacrum. In K11 there are two fibulae (one right and one left), two femora (one right and one left), and two ischia (one right and one left) but no extra paired bone (i.e. a third femur, for example). Moreover, all left and right paired bones have the same length, and in all likelihood belonged to the same individual. The preserved parts of the left ribs also fit very well in size and shape with the wellpreserved right ribs. The skeleton from K11 is about 60% complete (167 labelled specimens out of a total evaluated at 284 pieces) (Fig. 3). Based on the absence of extra paired bones, the identical measurements of paired bones, the partial articulation of the dorsal and sacral regions, the limited scattering of the cervical and caudal regions, and the consistency in size of all elements, all the bones most probably belong to a single individual that was lying on its right side. Scavenging by a theropod dinosaur is attested by the discovery of one theropod tooth (SM K11-0168), which was probably broken when the meat-eater was defleshing the corpse, and also by the presence of tooth marks on specimen SM K11-0092 (Fig. 4). Buffetaut & Suteethorn (1989, p. 80) reported a similar example of theropod teeth associated with a sauropod skeleton at PW1. They suggested ‘the sauropod carcass was exposed for some time on a floodplain surface before it was eventually buried under fine sediment, presumably during a flood. The theropod teeth found among the sauropod bones can safely be interpreted as those of one or several carnosaurs which fed on the carcass (probably contributing to its partial disarticulation) before it was buried. Whether the sauropod was killed by carnosaurs before it was eaten by them, or whether they merely acted as scavengers on the carcass of a sauropod which had died

Fig. 3. Reconstruction of Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994. Shaded bones represent the material from Ban Na Khrai locality.

194

S. SUTEETHORN ET AL.

Fig. 4. (a) Theropod tooth (SM K11-0168) found associated with Ban Na Khrai skeleton and (b) cervical rib (SM K11-0092) of Phuwiangosaurus sirindhornae with tooth marks. Scale bar: 5 cm.

from other causes, cannot be determined on the basis of the available data.’ Scavenging and running water possibly were the primary causes of the dispersion of the skeleton from K11. These phenomena could have led to the disappearance of some parts of the skeleton before its burial.

Systematic palaeontology Sauropodomorpha Huene Sauropoda Marsh Titanosauria Bonaparte & Coria Phuwiangosaurus Martin, Buffetaut & Suteethorn

NEW PHUWIANGOSAURUS SKELETON

Type species. Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994 Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn Holotype. SM PW1-0001 to SM PW1-0022; partly articulated skeleton including three cervical vertebrae, three dorsal vertebrae, scapulae, left humerus, left ulna, ilia, pubes, ischia, femora, left fibula, coracoid, chevron and rib. Referred material. Most of the abundant sauropod bones found in the Phu Wiang and Kalasin areas, including the baby and juvenile material described by Martin (1994) and Martin et al. (1999). All sauropod specimens from K11 (SM K11-0001 to SM K11-0167) as follows. Cranial elements including SM K11-0001, right frontal; SM K11-0002, right postorbital; SM K11-0003, right squamosal; SM K11-0004, left quadrate; SM K11-0005, right quadrate; SM K11-0006, co-ossified braincase consisting of supraoccipital, exoccipital –opisthotic, prootic, basioccipital and basisphenoid–parasphenoid; SM K11-0007, laterosphenoid–orbitosphenoid; SM K11-0008, an isolated tooth. Postcranial skeleton with partly articulated caudal dorsal and sacral centra lacking two or three cervical vertebrae, cranial dorsal vertebrae, most of cranial and middle caudal vertebrae, forelimbs and distal hind limbs. The catalogue numbers of specimens are available as supplementary material. Type locality. Phu Wiang Site PW1, Sao Khua Formation, northeastern Thailand. Age. Early Cretaceous, Barremian–Aptian (Racey & Goodall 2009). Emended diagnosis. Middle-sized sauropod (15– 20 m long); postorbital with short and acute caudal process; hook-shaped squamosal with vertical flat rostral process and L-shaped caudal process; robust quadrate condyle with deep quadrate fossa that faces caudolaterally and a prominent bulge on medial margin; exoccipital with a foot-like structure that participates to the occipital condyle; heartshaped occipital condyle; paroccipital process extended caudolaterally with rounded distal end; triangular basal tubula flattened with lateral rugose surface; basipterygoid process directed rostroventrally; anterior cervical vertebrae with a very low and wide neural arch; diapophyses and parapophyses very developed lateroventrally; large zygapophyses situated low and far from each other, firmly diverging laterally from the centrum; neural spine of the posterior cervical vertebrae widely bifurcated with no median spine; cervical vertebrae with a well-developed system of laminae and cavities; centra of the dorsal vertebrae

195

opisthocoelous with deep pleurocoels; posterior dorsal vertebrae with unforked neural spine; neural spine elongated craniocaudally; long diapophyses directed more dorsally than laterally, nearly reaching the level of the spine; hyposphene–hypantrum system present; elongated scapula with lateral ridge on the proximal extremity at right angle with the shaft, and slight distal expansion; humerus similarly expanded at both ends; anterior blade of the ilium well developed; pubic peduncle of the ilium straight, long and directed at right angle to the direction of the blade; ischiatic peduncle of the ilium faintly marked; pubis with very open angle between the axis of the shaft and the ischiatic border; well-marked curvature of the caudal border of the shaft of the ischium; femur flattened anteroposteriorly with the head situated slightly above the level of the great trochanter; fourth trochanter crest-shaped, located medially above the midlength of the shaft; very large lateral epicondyle at the distal end of the femur; slight lateral bending of the shaft of the fibula.

Description We describe below in some detail the skeletal elements from Ban Na Khrai that provide new information about the osteology of Phuwiangosaurus sirindhornae; namely, the skull bones and the vertebral column. Other elements, which do not differ significantly from those already described for Phuwiangosaurus sirindhornae on the basis of the type and referred specimens (Martin et al. 1999), such as girdles and limb bones, are alluded to more briefly in the ‘comparison’ section. The nomenclature for vertebral laminae used herein follows Wilson (1999) and Apesteguı´a (2005). There are 167 catalogue entries for Ban Na Khrai specimens, including one theropod tooth. Each isolated specimen has received a separate number, for example one disarticulated cervical vertebra consists of four specimens: one centrum, one neural arch and two cervical ribs, and each of them received a separate number. The total number of catalogue entries is different from the field number (156), as some more bones were found during preparation of the plaster jackets. (See online supplementary data.)

Description of the skull elements from Ban Na Khrai Frontal. The right frontal SM K11-0001 is well preserved and shows a straight suture line for the left frontal (Fig. 5). The bone is flat and rectangular in shape. In dorsal view its rostral edge is concave and acute rostrolaterally. Its lateral margin is

196

S. SUTEETHORN ET AL.

slightly oblique caudolaterally with its greatest breadth at the caudal end and forms the dorsal rim of the orbit. Its lateral surface is smooth along the frontal and postorbital. The frontal –nasal contact is obscured by matrix. The prefrontal –frontal contact is an embayment, indicating that the caudal part of the prefrontal is embraced by the frontal as in Nemegtosaurus (Nowinski 1971, p. 66; Wilson 2005, p. 298, fig. 7). The caudal part of the frontal contacts the parietal and postorbital. It seem that the frontal does not participate in the supratemporal fossa (see ‘postorbital’ below). In ventral view the suture with the orbitosphenoid is located in the middle of the frontal. A strong ridge extends rostrolaterally from the rostral part of the frontal– orbitosphenoid suture, separating the orbit from the nasal cavity. The frontal– laterosphenoid suture slopes caudolaterally and interdigitates with the frontal –postorbital suture. A small convexity is present on the dorsal surface of the frontal caudomedially, corresponding to the rostral part of the endocranial cavity. This condition differs from the doming on the frontal of Rapetosaurus (Curry Rogers & Forster 2004, fig. 13), which is more prominent and situated more rostrally than in K11-68A. Postorbital. The right postorbital SM K11-0002 is a triradiate, T-shaped bone in lateral view, with

an elongate ventral process and relatively short rostral and caudal processes (Fig. 6). The lateral surface of the postorbital that forms the margin of the orbit and infratemporal openings is smooth along its length. It differs from that of Nemegtosaurus, which is heavily ornamented (Wilson 2005, fig. 7). The rostral process of the postorbital is flattened dorsoventrally and expands transversely towards its distal end. The distal end of the rostral process contacts the frontal rostrally, the parietal medially and the laterosphenoid ventrally. This interdigitating process separates the orbit from the supratemporal fenestra. Medially, the rostral process has a shallow depression, forming the lateral margin of the supratemporal opening. It appears that the postorbital contacted both ends of the parietal, enclosing the supratemporal opening without the frontal and squamosal. The caudal process of the postorbital is extremely short and tapers distally to a pointed end, where it contacts the rostral process of the squamosal on its lateral surface. The postorbital –squamosal contact is flat vertically and oriented rostrocaudally (see ‘squamosal’ below). The caudal process of the postorbital separates the supratemporal fenestra from the infratemporal fenestra. The ventral process of the postorbital is long and thin, compressed rostrocaudally and widened transversely towards its proximal end. The ventral

Fig. 5. Right frontal (SM K11-0001) of Phuwiangosaurus sirindhornae in (a) ventral view and (b) dorsal view. enc, endocranial cavity; ls, laterosphenoid facet; ns, nasal facet; nsc, nasal cavity; obs, orbitosphenoid facet; orb, orbit; p, parietal facet; po, postorbital facet; prf, prefrontal facet. Scale bar: 5cm.

NEW PHUWIANGOSAURUS SKELETON

197

Fig. 6. Right postorbital (SM K11-0002) of Phuwiangosaurus sirindhornae in (a) dorsal view, (b) medial view, (c) caudal view and (d) lateral view; (e) cross-section of the ventral process, showing socket for the jugal contact. cdp, caudal process; f, frontal facet; j, jugal facet; ls, laterosphenoid facet; orb, orbit; p, parietal facet; rtp, rostral process; sq, squamosal facet; stf, supratemporal fenestra; vtp, ventral process. Scale bar: 5 cm.

process is bent toward the orbit and its rostral surface is concave transversely and faces rostrolateral. The postorbital–jugal contact is a socket on the distal portion of the ventral process (Fig. 6). The socket opening faces caudolaterally and occupies one-third of the length of the ventral process. As the lacrimal is not preserved in K11; its contact with the postorbital is unknown. Squamosal. The right squamosal SM K11-0003 is a triradiate element in lateral view, with a hookshaped rostral process and broad ventral and caudal processes (Fig. 7). The squamosal shaft is plate-like, flat and thin transversely. The smooth rostral margin of the squamosal forms the margin of the infratemporal opening. The rostral process of the squamosal is hookshaped and projects rostrally. The rostral process bears two facets on its medial surface. A small triangular facet is located on the most rostral part of the medial surface. It is oriented vertically in sagittal plane, contacting the caudal process of the postorbital. This contact differs from the transverse orientation of the postorbital –squamosal contact in Nemegtosaurus as noted by Wilson (2005, p. 297). The postorbital –squamosal contact separates the squamosal from the supratemporal opening. The second facet is a large one occupying the dorsal part of the squamosal. It is separated from the squamosal shaft by a ridge that extends from the

rostral facet, contacting the caudal process of the parietal. The ventral process of the squamosal expands distally. The distal end of the ventral process bears a faint articulation, which contacts the quadratojugal. Although the quadratojugal is not preserved in K11, the presence of the quadrate–quadratojugal articulation along the lateral margin of the quadrate fossa (see ‘quadrate’ below) confirms that the quadratojugal reaches the squamosal. SM K11-0003 presents a ‘spur’ on the caudal surface of the ventral process (Fig. 7), a character considered as an autapomophy of Nemegtosaurus by Wilson (2005, p. 297). The caudal process of the squamosal is separated from the caudal margin of the squamosal shaft by a notch on the lateral surface. Medially, a horizontal ridge originates from the squamosal– quadratojugal contact to the mid-length of the squamosal shaft and bends backwards, forming the lower floor of the caudal process. The caudal process is L-shaped in caudal view, receiving the paroccipital process. Below the ridge where it bends backward, there is a shallow depression or socket that received the dorsal head of the quadrate. This ridge separates the paroccipital process from the quadrate. Quadrate. Both left and right quadrates (SM K11-0004, -0005, respectively) are preserved (Fig. 8). The dorsal head of the quadrate is long

198

S. SUTEETHORN ET AL.

Fig. 7. Right squamosal (SM K11-0003) of Phuwiangosaurus sirindhornae in (a) medial view, (b) caudal view and (c) lateral view. cdp, caudal process; itf, infratemporal fenestra; p, parietal facet; po, postorbital facet; pop, paroccipital process facet; q, quadrate facet; qj, quadratojugal facet; rtp, rostral process; vtp, ventral process. Scale bar: 5 cm.

and thin dorsoventrally. The dorsal head bends backward, perpendicular to the long axis of the quadrate condyle, and forms the articulation at its distal end to attach the squamosal socket (see ‘squamosal’ above). The quadrate fossa is large, deep and has a triangular shape in caudal view. The quadrate fossa faces caudolaterally relative to the pterygoid flange which was oriented rostrocaudally and look like those of Nemegtosaurus and Quaesitosaurus (Wilson 2005, figs 9 and 18). A sutural scar is present along the lateral surface of the quadrate from the ventral condyle to the dorsal articulation. This suture contacts the squamosal dorsolaterally and the quadratojugal ventrolaterally. The pterygoid flange of the quadrate is platelike, flattened transversely and blunt distally. It is rectangular in lateral view, oriented vertically and rostrocaudally. Medially, the middle portion of the pterygoid flange is depressed and presents a rugose surface. This depression received the palate (pterygoid), which can be seen as a medial bulge in caudal view. The quadrate condyle is robust and oriented ventrally. In ventral view the articular condyle is kidney-shaped, with a convex caudal margin and slightly concave rostral margin. In caudal view the quadrate condyle is sloped so its facet faces ventrolaterally. The quadrate condyle of K11 is similar to that of Nemegtosaurus (Wilson 2005, fig. 11). The shape of the quadrate is similar to that of Malawisaurus dixeyi (Gomani 2005, fig. 5), but

the quadrate fossa of Malawisaurus is oval-shaped and wider transversely. Braincase. The braincase is well preserved and nearly complete, lacking the left laterosphenoid – orbitosphenoid, and the distal end of the parasphenoid rostrum (Figs 9 and 10). The right laterosphenoid –orbitosphenoid unit (SM K11-0007) is preserved separately from the main body of braincase (SM K11-0006), which is co-ossified and consists of the following elements: supraoccipital, exoccipital opisthotic, prootic, basioccipital, basisphenoid and parasphenoid. These elements can be distinguished by the sutures, except the exoccipital –opisthotic, basisphenoid–parasphenoid and laterosphenoid –orbitosphenoid sutures, which cannot be traced. Supraoccipital. The supraoccipital is the most dorsal part of the braincase and forms the dorsal margin of the foramen magnum (Fig. 9). In occipital view the supraoccipital is triangular in shape. The sutures that connect the supraoccipital with the exoccipital –opisthotic complex ventrally and the prootic rostrally are very clear. The parietal contact is marked by a rostromedial groove on the dorsal surface of the supraoccipital. A vertical median ridge, which is high and prominent dorsally, extended from the dorsal margin of the foramen magnum across the occipital surface of the supraoccipital. As in Nemegtosaurus, this supraoccipital ridge can be seen as a small,

NEW PHUWIANGOSAURUS SKELETON

199

Fig. 8. Quadrates of Phuwiangosaurus sirindhornae. (a, c, e, g) Left quadrate (SM K11-0004); (b, d, f) right quadrate (SM K11-0005). (a, b) lateral view; (c, d) caudal view; (e, f) medial view; (g) ventral view. dsp, dorsal process; pt, pterygoid facet; ptf, pterygoid flange; qc, quadrate condyle; qf, quadrate fossa; qj, quadratojugal facet; sq, squamosal facet. Scale bar: 5 cm.

200

S. SUTEETHORN ET AL.

Fig. 9. Braincase (SM K11-0006) of Phuwiangosaurus sirindhornae in (a) caudal view, (b) caudolateral view, (c) rostral view and (d) left lateral view. bo, basioccipital; bpt, basipterygoid process; bt, basal tubera; cpr, crista prootica; ds, dorsum sella; enc, endocranial cavity; eo, exoccipital; fm, foramen magnum; fo, fenestra ovalis; ic, internal carotid artery; msf, medien subcondylar foramen; oc, occipital condyle; p, parietal facet; pif, pituitary fossa; pop, paroccipital process; pr, prootic; ps, parasphenoid rostrum; so, supraoccipital; V, foramen for trigeminal nerve; VI, foramen for abducens nerve; IX– XI, foramen for glossopharyngeal, vagus, and spinal accessory nerves; XII, foramen for hypoglossal nerve. Scale bar: 5cm.

triangular process in dorsal view (Wilson 2005, p. 300, fig. 7). Exoccipital–opisthotic. The exoccipital– opisthotic suture cannot be determined. The exoccipital– opisthotic complex is united with the supraoccipital

dorsally, prootic rostrally, basioccipital ventrally, and squamosal laterally (Fig. 9). The ventral footlike structure of the exoccipital forms the lateral and ventral margin of the foramen magnum, which is enclosed by the supraoccipital dorsally. This structure extends backward, participating in the

NEW PHUWIANGOSAURUS SKELETON

201

Fig. 10. Right laterosphenoid– orbitosphenoid (SM K11-0007) of Phuwiangosaurus sirindhornae in (a) craniolateral view and (b) dorsal view. can, crista antotica; enc, endocranial cavity; f, frontal facet; obs, orbithosphenoid facet; po, postorbital facet; pr, prootic facet; I, foramen for olfactory nerve; II, foramen for optic nerve; III, foramen for oculomoto nerve; IV, foramen for trochlear nerve. Scale bar: 5 cm.

dorsolateral part of the occipital condyle. The caudal surface of the exoccipital shows the opening for the exit of the hypoglossal nerve (cranial nerve XII) lateral to the occipital condyle. In lateral view, the exoccipital–opisthotic, basioccipital, basisphenoid, and prootic form the otic region (Fig. 9). The otic region shows two distinctive fenestrae, although covered by matrix; a rostral one, which is rounded and oval-shaped; and a caudal one, which is more expanded

dorsoventrally. The rostral one is the fenestra ovalis for the reception of the footplate of the stapes and the caudal one is the metotic foramen through which the glossopharyngeal, vagus and spinal accessory nerves (cranial nerves IX –XI), and the caudal branch of the jugular vein are transmitted (see Chatterjee & Zheng 2005). The opisthotic forms part of the paroccipital process, which is flattened transversely and expands caudolaterally. The concavity on the

202

S. SUTEETHORN ET AL.

dorsal margin of the paroccipital process forms the ventral margin of the post-temporal opening. The distal end of the paroccipital process is rounded, flattened transversely and expands slightly caudoventrally. Prootic. The prootic is a plate-like element (Fig. 9). It faces rostrolaterally and forms the lateral wall of the temporal region. The prootic contacts the supraoccipital caudodorsally, the exoccipital–opisthotic caudally, the basioccipital and the basisphenoid caudoventrally, and the laterosphenoid rostrally. The rostral surface presents two foramina nearly parallel to each other, which are referred to the trigeminal nerve (cranial nerve V). The foramina are directed ventrolaterally and enclosed rostrally by the laterosphenoid. The dorsal foramen marks the passage for the maxillary branch of the trigeminal nerve, and the ventral foramen marks the passage for the mandibular branch (Chatterjee & Zheng 2005). The caudal edge of the prootic is sharp and forms the rostral margin of the otic region (see ‘exoccipital– opisthotic’ above). This edge extends caudally along the opisthotic to the mid-length of the paroccipital process. The prootic expands ventrally, forming a plate-like element, the crista prootica. Basioccipital. The basioccipital forms most of the occipital condyle except the dorsolateral part made by the exoccipitals (see ‘exoccipital– opistotic’ above). The occipital condyle projects and faces caudoventrally when the occipital surface of the supraoccipital is oriented vertically (Fig. 9). The articular surface of the occipital condyle is heart-shaped in occipital view, slightly concave dorsally and hemispherical in lateral view. The width of the occipital condyle is twice that of the foramen magnum. The basal tubera are formed by the basisphenoid rostrally and the basioccipital caudally. The platelike basal tubera are triangular in shape and project ventrally. The right basal tubera is larger and more prominent than the left one. Although the right basal tubera seems to have an articulation on its lateral margin, it does not contact the quadrate. This condition differs from Nemegtosaurus and Quaesitosaurus, where the basal tubera articulate with the quadrate (Wilson 2005, figs 9 and 18). The lateral margins of the basal tubera are slightly rugose, particularly on their basisphenoid portion. This irregular surface extends ventrally on the midline down to the posterior face of the basipterygoid processes and may represent an extension for muscle attachment (Curry Rogers & Forster 2004). Four small openings are oriented vertically in sequence between the basal tubera. This is

different from Rapetosaurus, Nemegtosaurus and Quaesitosaurus, which have a unique small foramen between the basal tubera (Curry Rogers & Forster 2004, figs 23 and 24; Wilson 2005, fig. 11), and from Suuwassea, where the small foramen is referred to the median subcondylar foramen (Harris & Dodson 2004). Basisphenoid–parasphenoid. The basisphenoid –parasphenoid suture cannot be determined. The basisphenoid forms a rostral portion of the basal tubera and the basipterygoid processes (Fig. 9). The basipterygoid processes direct rostroventrally when the occipital surface is vertically oriented and diverge ventrolaterally at approximately a 308 angle from one another. The basipterygoid processes are subtriangular in cross-section with a prominent ridge along the rostral margin. The length of the basipterygoid process is 3.8 times its basal width. The distal ends of these processes are gently rounded. A smooth, slightly concave, triangular surface occupies the area between the bases of basipterygoid processes and the parasphenoid rostrum. The internal carotid artery opens at the lateral base of the basipterygoid process. The parasphenoid is represented by the parasphenoid rostrum. The parasphenoid rostrum is flattened transversely and expanded rostrodorsally from the bases of basipterygoid processes. Its distal end is incomplete. At the base of the rostrum, a vertical wall, the dorsum sella, extends from the basisphenoid on either sides of the large circular opening (pituitary fossa) to articulate with the orbitosphenoid. The dorsum sella is pierced by the opening for the abducens nerve (cranial nerve VI) and is similar to that of Camarasaurus lentus and Rapetosaurus (Curry Rogers & Forster 2004; Chatterjee & Zheng 2005). Laterosphenoid–orbitosphenoid. The laterosphenoid –orbitosphenoid suture cannot be determined. The laterosphenoid– orbitosphenoid unit forms the rostral and lateral wall of the endocranial cavity and the medial wall of the orbit (Fig. 10). The orbitosphenoid is a plate-like element on the rostral portion of the unit. A median suture separates the orbitosphenoids from one another. A large V-shaped opening of the olfactory nerve (cranial nerve I) passes through the orbitosphenoids rostrally, between the contacts of the frontals and orbithosphenoids. Farther ventrally, a large foramen for the optic nerve (cranial nerve II) exits through the orbitosphenoid just rostral to that of the oculomotor nerve (cranial nerve III). Slightly above the oculomotor foramen, a small foramen for the trochlear nerve (cranial nerve VI) oriented transversely through the laterosphenoid –orbitosphenoid.

NEW PHUWIANGOSAURUS SKELETON

The laterosphenoid expands outward and upward, forming a wing-like structure, the crista antotica. The crista antotica interdigitates with the frontal and postorbital, separating the temporal region of the skull from the orbital region (see ‘frontal’ above). Caudally, a rugose surface of the laterosphenoid contacts the prootic, enclosing the parallel foramina for cranial nerve V. Tooth. A single peg-like tooth SM K11-0008 is preserved (Fig. 11). This tooth was half broken. The upper part of the tooth crown lost some enamel, especially at the wear facet. The lower part of the crown with root is embedded in matrix. The tooth crown is narrow and tapers toward a blunt apex. There are two wear facets (apical and mesial–distal). The apical wear facet is diagonal and set on the distal end of the crown. The mesial–distal facet is nearly vertical. The ridges on both mesial and distal edges are not prominent. The enamel surface is finely wrinkled throughout the crown. The tooth crown is slightly D-shaped in crosssection and the apical wear facet on the flattened side (lingual side) of the crown suggests that this crown is an upper tooth, by comparison with the teeth of Nemegtosaurus mongoliensis (see details given by Nowinski 1971, p. 71; Wilson 2005, p. 305). In this case, the tooth is lingually curved.

Fig. 11. Tooth (SM K11-0008) of Phuwiangosaurus sirindhornae in (a) lingual view and (b) distal view. awf, apical wear facet; wf, wear facet. Scale bar: 1 cm.

203

Description of the axial skeleton from Ban Na Khrai Cervical vertebrae. Ten cervical vertebrae have been recovered from K11 (Fig. 12). Their original positions were determined using the relative sizes of the neural arches and/or articulation of the centra (Table 1). The K11 sauropod had probably 13 cervical vertebrae. Eight of the 10 preserved cervical are referable to the cervical C2–C9 and the other two, which are represented by their centra only, are more posterior ones. C5– C9 have fused centra and neural arches. Eleven right cervical ribs are preserved separately from the vertebrae. A comparison with the fused left ribs of C5 and C7 suggests that these ribs belonged to C3–C13. The last cervical rib (C13) presents a distinctive character from the preceding one and from dorsal ribs, the rib shaft being reduced to a short and small, pencil-like process. The axis (C2, SM K11-0009) preserved only the neural arch and the prezygapophyses are incomplete. The neural arch is low and wide. The diapophysis is directed laterally. The transversal width of the diapophysis is equal to that of the postzygapophyses. A shallow depression is present on the lateral surface of the neural spine dorsal to the diapophyses. In lateral view the prespinal lamina is straight and directed dorsoposteriorly to its greatest height. The postzygapophyses overhang the posterior end of the neural arch. The postzygapophyseal facets are oval, horizontal and flat. The spinopostzygapophyseal lamina diverges and is directed ventroposteriorly to support the postzygapophyses dorsally with a posterior extension, the epipophysis (¼‘torus dorsalis’, Ksepka & Norell 2006). The prominent epipophysis projects posteriorly beyond the postzygapophyseal facets. All the cervical centra are strongly opisthocoelous (Fig. 12). The vertebrae are relatively low and elongate. The centra lengthen from C3 to C8. C8 is the longest centrum, C9 becoming shorter. The articulations of the anterior centra are ovalshaped, flat dorsoventrally, and became more spherical in the succeeding vertebrae. There are two pleurocoels on the lateral surface of the centrum. A shallow depression is present on the ventral surface of the centra between the parapophyses. The ventral keel is not present in cervical centra. The elongated neural spines are extended anteroposteriorly so that the prezygapophyses overhang the anterior end of the centra whereas the postzygapophyses overhang their posterior ends. The neural spine begins to divide in C7 and the distance between the two neural spines becomes larger in the succeeding vertebrae with no median spine.

204 S. SUTEETHORN ET AL. Fig. 12. Cervical vertebrae of Phuwiangosaurus sirindhornae. (a, i) Axis (SM K11-0009); (b, c, j, k) cervical 4 (SM K11-0012 and SM K11-0011); (d, l) cervical 5 (SM K11-0013); (e, m) cervical 6 (SM K11-0014); (f, n) cervical 7 (SM K11-0015); (g, o) cervical 8 (SM K11-0016); (h, p, q) cervical 9 (SM K1-0017). (a –h) anterior view; (i– p) lateral view; (q) dorsal view. aap, anterior accessory process; acdl, anterior centrodiapophyseal lamina; cap, capitulum; cprl, centroprezygapophyseal lamina; cpol, centropostzygapophyseal lamina; di, diapophysis; epp, epipophyses; ns, neural spine; pa, parapophysis; pc, pleurocoel; pcdl, posterior centrodiapophyseal lamina; podl, postzygodiapophyseal lamina; poz, postzygapophysis; prdl, prezygodiapophyseal lamina; prz, prezygapophysis; r, rib; spol, spinopostzygapophyseal lamina; sprl, spinoprezygapophyseal lamina; tu, tuberculum. Scale bar: 10 cm.

NEW PHUWIANGOSAURUS SKELETON

205

Table 1. Measurements of cervical vertebrae of Ban Na Khrai Phuwiangosaurus sirindhornae C

1

2

3

4

5

6

7

8

9

10

11

12

13

GLC HA WA HC WC LPRZ LPOZ LNS TH

– – – – – – – – –

– – – – – – – 115 60*

– – – – – – – * *

200 30 45 40 60 55 66 240 *

215 35 70 45 80 80 80 290 130

240 40 70 60 90 80 90 308 140

235 50 90 70 110 100 100 315 160

270 56 89 70 100 90 100 340 180

220 65 95 90 125 120 155 270 200

180* – – 100 145 – – – *

– – – – – – – – –

– – – – – – – * *

200 85 140 95* 145 – – * *

Measurements in millimetres. C, cervical; GLC, greatest length of centrum; HA, height of articulation; WA, width of articulation; HC, height of centrum at posterior; WC, width of centrum at posterior; LPRZ, length between prezygapophyses; LPOZ, length between postzygapophyses; LNS, length of neural spine; TH, total height. *Incomplete.

The dorsal heads of the bifid spines are blunt and thick. The prezygapophyseal facets are oblique relative to the sagittal plane and face anterodorsally in lateral view. The facets of the prezygapophyses and postzygapophyses are flat and oval in outline, wider transversally. The diapophyses are compressed dorsoventrally and expand anteroposteriorly, supported by the prezygodiapophyseal, posterior centrodiapophyseal and postzygodiapophyseal laminae. The prezygodiapophyseal lamina extends anteriorly beyond the prezygapophyseal facet, forming the accessory anterior process. The prominent plate-like epipophysis extends posteriorly and overhangs the postzygapophyseal facet. The cervical ribs are longer than the posterior end of the centrum. The capitulum process presents a strong depression on its dorsal surface. A vertical lamina is present between the tuberculum and capitulum processes. The tuberculum –capitulum angle of the rib is acute, so that the rib is hanging lateroventrally to the centrum. The caudal process of the rib is slightly V-shaped in cross-section. The ventral surface of the rib presents a shallow depression and a faint ridge lateroventrally. Dorsal vertebrae. All dorsal centra and neural arches are unfused (Fig. 13). The K11 sauropod had twelve dorsal vertebrae (Table 2). Eleven of them are preserved. Five of the 11 dorsal (D8– D12) were found articulated with four sacral centra (S1–S4). Three others are anterior dorsals D1, D3 and D4, considering the position of the parapophysis that is located on the centrum and moved dorsally at the level of neural canal in D4, and considering the length of the centrum relative to its width (McIntosh 1990). Twelve dorsal neural spines are preserved. Six anterior dorsal neural arches (D1–D6) were found articulated; the other

six posterior dorsal neural arches (D7–D12) were disarticulated. Eleven right dorsal ribs were preserved in situ. All centra are opisthocoelous. The articulations of the centra are semispherical in the anterior vertebrae and became less prominent, more flattened anteroposteriorly in the succeeding elements. The pleurocoel is a simple large pit. It is circular in D1 and D3, and becomes spindle-shaped in D4–D7; it gradually extends dorsally and becomes dropshaped in D8–D12. A ventral keel is present on the ventral surface of the centra D3– D7. The neural arches D1–D5 are relatively low and wide transversally, and the neural spines are divided by a U-shaped cleft. The posterior neural arches are short anteroposteriorly and uniform in height. The first undivided neural spine is D6. The undivided neural spines are short related to the neural arches. The spine is directed dorsally in D12 and slightly inclined dorsoposteriorly in the preceding vertebrae. These spines were supported anteriorly by the prespinal and spinoprezygapophyseal laminae, and posteriorly by the postspinal, spinodiapophyseal and spinopostzygapophyseal laminae. The prespinal and postspinal laminae are present as roughened ridges for interspinous ligaments. In dorsal view, the distal end of the spine D6–D12 expands laterally where the spinopostzygapophyseal lamina joins with the spinodiapophyseal lamina, forming the triangular process (Fig. 13). The prezygapophyses of the anterior dorsals D1 –D4 are situated at the base of the diapophyses and above the level of the diapophyses. Their facets are large, flat horizontally and diverge from each other. The prezygapophyses of the succeeding dorsals D5–D12 move downward and inward and are close to one another, just below the base of the diapophyses. The prezygapophyses and

206

S. SUTEETHORN ET AL.

Fig. 13. Dorsal vertebrae of Phuwiangosaurus sirindhornae. (a, d, e) Dorsal 12 (SM K11-0042); (b, f, h, i) dorsal 11 (SM K11-0040 and SM K11-0039); (c, g) dorsal 10 (SM K11-0038). (a–c, h) Anterior view; (d) dorsal view; (e–g, i) right lateral view. acpl, anterior centroparapophyseal lamina; cprl, centroprezygapophyseal lamina; di, diapophysis; hp, hyposphene; hpa, hypantrum; mcpo, medial centropostzygapophyseal lamina; ns, neural spine; pa, parapophysis; pc, pleurocoel; pcdl, posterior centrodiapophyseal lamina; pcpl, posterior centroparapophyseal lamina; podl, postzygodiapophyseal lamina; poz, postzygapophysis; ppdl, paradiapophyseal lamina; prdl, prezygodiapophyseal lamina; prpl, prezygoparapophyseal lamina; prsl, prespinal; prz, prezygapophysis; spdl, spinodiapophyseal lamina; spol, spinopostzygapophyseal lamina; sprl, spinoprezygapophyseal lamina. Scale bar: 10 cm.

Table 2. Measurements of dorsal vertebrae of Ban Na Khrai Phuwiangosaurus sirindhornae D GLC WC LNS HNS

1

2

3

4

5

6

7

8

9

10

11

12

185 135 – –

– – – –

105 140 – –

160 90 – –

160 105 * *

130 110 * *

160 115 160 270

150 120 150 260

140 120 150 280

145 135 160 260

135 120 150 270

130 130 130 270

Measurements in millimetres. D, dorsal; GLC, greatest length of centrum; WC, width of centrum at posterior; LNS, length of neural spine; HNS, height of neural spine. *Incomplete.

NEW PHUWIANGOSAURUS SKELETON

postzygapophyses are oriented obliquely in D7 and gradually inclined to horizontal in D12. The hyposphene–hypantrum system is present on D7–D12. It cannot be observed on the articulated specimen and the posterior part of D6 is incomplete. The hypantrum is limited by the closely placed prezygapophyses and centroprezygapophyseal laminae. The hyposphenes of D7–D10 are developed at the connection of the postzygapophyses and are buttressed ventrally by the paired medial centropostzygapophyseal laminae. Those of D11 and D12 are buttressed by a single medial centropostzygapophyseal lamina (Apesteguı´a 2005). The diapophyses are flattened dorsoventrally and directed laterally in the anterior dorsals. They became rounded and directed dorsolaterally in the posterior dorsals. The diapophyses of the posterior dorsals are buttressed by five laminae, the prezygodiapophyseal, spinodiapophyseal, postzygodiapophyseal, paradiapophyseal and posterior centrodiapophyseal laminae (Fig. 13). The parapophyses of D7 are situated at the level of the prezygapophyses, anteroventral to the diapophyses, which migrate to a more ventral position in the succeeding dorsals. The articular facets of the parapophyses are subcircular to oval-shaped from D7 to D12; they are elongated dorsoventrally and smaller in the last vertebrae. The parapophyses of the posterior dorsal are buttressed by four laminae, the prezygoparapophyseal, paradiapophyseal, anterior and posterior centroparapophyseal laminae. Eleven right dorsal ribs were preserved in situ. The proximal head of the anterior ribs is large and flat. It are smaller and rounder in the posterior ribs. The capitulum processes project at right angles to the long axis of the rib in D2–D7; then they are slightly curved along their axis in D8 – D12. The tuberculum processes are short and blunt with a depression on the lateral surface of the processes. Sacral vertebrae. Four articulated sacral centra and four co-ossified neural arches are considered as the sacrals S1 –S4 (Table 3, Fig. 14). An unfused centrum and neural spine present characteristics of Table 3. Measurements of sacral vertebrae of Ban Na Khrai Phuwiangosaurus sirindhornae S GLC WC HNS

1

2

3

4

5

110 170 275

125 135 280

115 125 280

125 100 *

90 135 140

Measurements in millimetres. S, sacral; GLC, greatest length of centrum; WC, width of centrum at posterior; HNS, height of neural spine. *Incomplete.

207

both sacrals and caudals, and we regard them as a caudosacral or last sacral S5. Six sacral ribs are preserved, including four right ribs and two left (Fig. 15). The numbers of preserved centra and neural arches and the shapes of sacral ribs, especially the distal part that contact the ilium, suggest that K11 sauropod had five sacral vertebrae, as noted by Martin et al. (1999, p. 65) for Phuwiangosaurus sirindhornae. All sacral centra are amphiplatyan. A small pneumatic fossa is present on the lateral surface of S1 –S4 just anterior to the parapophyses and vanishes in S5. The anterior articulations of the centra are roughened; trapezoid-shaped in S2 and S3; rectangular in S4; slightly oval, wider dorsoventrally in S5. The posterior articulation of S5 is flat and circular, which is similar to that of the caudal centrum. The co-ossified neural spines of S1– S4 are compressed laterally. The heights of the spines are nearly the same as those of the posterior dorsals (Table 2). The neural spines of S1 –S4 are welldeveloped laminae, very thin transversely and blunt at the distal end. That of S5 thickens transversely through the spine without the blunt end. The diapophyses of S1 and S2 are directed laterally. The parapophyses move down to the centrum, inserting both on the centrum and neural arch. The sacral ribs are triradiate elements in S1 –S4; S5 sacral rib is rectangular (Fig. 15). The proximal parts that contact the vertebrae are plate-like and thin transversely. The distal parts are robust and oriented horizontally in S1–S3, vertically in S4 and S5. The distal ends of S1 –S5 connect to each other, forming the sacrocostal yoke that contacts the acetabulum. The attachment scar for the sacrocostal yoke can be seen on the medial side of the ilium, and is restricted to the distal part of the acetabulum. Caudal vertebrae. A few poorly preserved proximal caudal vertebrae were recovered. Most of the preserved centra and neural arches are unfused (Table 4, Fig. 16). The amphicoelous centra are short anteroposteriorly relative to their height. The anterior and posterior articulations of the centra are circular in shape. The lateral surface of the centrum is strongly concave anteroposteriorly. The chevron facet is not present on the most proximal caudal CA1 and CA2. The neural arch of CA1 is flattened anteroposteriorly and its spine is flattened transversely. The spine of CA1 presents roughened ridges on its anterior and posterior surfaces. The neural spines of proximal caudal are slightly curved posteriorly. The prezygapophyseal facets are nearly vertical, corresponding to the postzygapophyses. The transverse processes are situated at the level of the neural canal and directed lateroposteriorly.

208

S. SUTEETHORN ET AL.

Fig. 14. Sacral vertebrae of Phuwiangosaurus sirindhornae. (a) Fused neural spines of sacral 1– 4 (SM K11-0044, SM K11-0046, SM K11-0048 and SM K11-0050); (b, g, k) neural spine and centrum of sacro-caudal/sacral 5 (SM K11-0052 and SM K11-0051); (c) sacral 1 (SM K11-0043); (d, h) sacral 2 (SM K11-0045); (e, i) sacral 3 (SM K11-0047); (f, j) sacral 4 (SM K11-0049). (a–g) Right lateral view; (h– k) anterior view. di, diapophysis; ns, neural spine; pc, pleurocoel; prz, prezygapophysis. Scale bar: 10 cm.

The middle caudal centra are amphiplatyan, nearly rectangular in lateral view. The ventral surfaces are flat and present a faint longitudinal ridge. The chevron facets are marked posteriorly and triangular in shape. The neural spines are very thin transversely and extended posteriorly; the postzygapophyses overhang the posterior margin of its centrum. The distal caudal centra become amphiplatyan and biconvex in the most distal caudals (Fig. 17). Their centra are elongated anteroposteriorly. In anterior view the articular face of the centrum is sub-rectangular in outline. Sixteen chevrons are preserved. The proximal facets are not confluent. The chevron blades are flat transversally and curved caudoventrally. The chevron blades are shorter in the posterior elements.

The most posterior chevron blades are L-shaped in lateral view.

Comparison with the holotype of Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994 The holotype of Phuwiangosaurus sirindhornae from Phu Wiang (PW1) is a partly articulated skeleton with 21 elements consisting of cervical vertebrae, dorsal vertebrae, pelvic girdle, left scapula, coracoid, left humerus, left ulna, both femora, left fibula, chevrons and ribs. Two more bones were found later: a posterior dorsal vertebra and an anterior caudal vertebra (Suteethorn et al. 2009).

Measurements in millimetres. CA, caudal; GLC, greatest length of centrum; WC, width of centrum at posterior; HNS, height of neural spine. *Incomplete.

GLC 90 80 85 90 * – – – – – – – – – – – – – – – – – – – 70 80 80 90 85 105 60 83 85 90 95 90 65 70 65 75 – – * 30 50 40 47 50 45 40 * 33 33 30 WC 140 155 * 140 * – – – – – – – – – – – – – – – – – – – 90 77 70 60 55 57 55 55 55 55 52 43 45 45 40 40 – – 35 25 25 22 22 17 15 8 8 8 9 8 HNS 150 145 * 145 * – – – – – – – – – – – – – – – – – – – 100 70 50 70 75 65 60 65 68 65 62 50 50 45 45 45 – – 30 30 25 22 20 13 20 12 12 11 11 11

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 3 2 1

Martin et al. (1999) emended the diagnosis of the type as follows: (1) middle-sized sauropod (15– 20 m long); (2) anterior cervical vertebrae with a very low and wide neural arch; (3) diapophyses and parapophyses very developed lateroventrally; (4) large zygapophyses situated low and far from each other, firmly diverging laterally from the centrum; (5) neural spine of the posterior cervical vertebrae widely bifurcated with no median spine; (6) cervical vertebrae with a well-developed system of laminae and cavities; (7) centra of the dorsal vertebrae opisthocoelous with deep pleurocoels; (8) posterior dorsal vertebrae with unforked neural spine; (9) neural spine elongated craniocaudally; (10) long diapophyses directed more

209

CA

Fig. 15. Sacral ribs of Phuwiangosaurus sirindhornae in posterior view. (a) Sacral 3 left; (b) sacral 4 left; (c) sacral 1 right; (d) sacral 2 right; (e) sacral 4 right; (f) sacral 5 right. c, centrum process; il, ilium process; scy, sacrocostal yoke. Scale bar: 10 cm.

Table 4. Measurements of caudal vertebrae of Ban Na Khrai Phuwiangosaurus sirindhornae

NEW PHUWIANGOSAURUS SKELETON

210

S. SUTEETHORN ET AL.

Fig. 16. Proximal caudal vertebrae of Phuwiangosaurus sirindhornae. (a –d) Proximal-most caudal vertebra (SM K11-0054 and SM K11-0053); (e–f) proximal caudal vertebra (K11-64). (a, b, e) Anterior view; (c, d, f) left lateral view. nc, neural canal; poz, postzygapophysis; prz, prezygapophysis; tp, transverse process. Scale bar: 10 cm.

dorsally than laterally, nearly reaching the level of the spine; (11) hyposphene –hypantrum system present; (12) elongated scapula with lateral ridge of the proximal extremity at right angle with the shaft, and slight distal expansion; (13) humerus similarly expanded at both ends; (14) anterior blade of the ilium well developed; (15) pubic peduncle of the ilium straight, long and directed at right angles to the direction of the blade; (16) ischiatic peduncle of the ilium faintly marked; (17) pubis with very open angle between the axis of the shaft and the ischiatic border; (18) well-marked curvature of the caudal border of the shaft of the ischium; (19) femur flattened anteroposteriorly with the head situated slightly above the level of the great trochanter; (20) fourth trochanter crest-shaped, located medially above the mid-length of the shaft; (21) very large lateral epicondyle at the distal end of

the femur; (22) slight lateral bending of the shaft of the fibula. Among these characters, important apomorphies of the taxon are: (3) diapophyses and parapophyses very developed lateroventrally; (5) neural spine of the posterior cervical vertebrae widely bifurcated with no median spine; (8) posterior dorsal vertebrae with unforked neural spine; (15) pubic peduncle of the ilium straight, long and directed at right angles to the direction of the blade; (16) ischiatic peduncle of the ilium faintly marked; (18) well-marked curvature of the caudal border of the shaft of the ischium; (21) very large lateral epicondyle at the distal end of the femur. Upchurch et al. (2004), however, retained only two diagnostic characters for P. sirindhornae: cranial cervicals with low, wide neural arches and transverse widths of the proximal and distal ends of the humerus subequal. Those

Fig. 17. Distal and distal-most caudal vertebrae of Phuwiangosaurus sirindhornae (SM K11-0075, SM K11-0082 to SM K11-0089). Scale bar: 5cm.

NEW PHUWIANGOSAURUS SKELETON

workers mentioned that the other ‘diagnostic’ characters listed by Martin et al. (1994, 1999) have a wider phylogenetic distribution among sauropods. The holotype skeleton is only 10% complete whereas the skeleton from K11 is about 60% complete and much better preserved. However, the skeleton from K11 is a small individual compared with the type; for example, the femur of the type specimen (SM PW1-0016) is 25% longer than that of the K11 specimen (SM K11-0152).

Skull PW1 and K11 have both yielded a single sauropod tooth. The specimen from PW1 is a half tooth preserved in longitudinal section and embedded in matrix. It is long and slender. The tooth from K11 is better preserved. Its shape is similar to that of the tooth from PW1.

Axial skeleton The cervical vertebrae of the sauropod from K11 are very well preserved. SM K11-0013 and SM K11-0014 are very similar to each other, SM K11-0014 being larger than SM K11-0013. These cervical vertebrae share their characteristics with SM PW1-0001 (Fig. 18). The centra are strongly opisthocoelous and elongated craniocaudally. The hemispherical articular surface of the centrum of SM K11-0013 is similar to that of SM PW1-0001, which is more flattened dorsoventrally than in SM K11-0014. The neural arches are low and wide. Diapophyses and parapophyses are very developed

211

lateroventrally. The prezygapophyses and postzygapophyses of the K11 vertebrae are more complete than in the type and show additional developed spines; one is below the prezygapophyseal facet (the accessory anterior process) and another is above the postzygapophyseal facet (the epipophysis). These elements cannot be observed on the holotype because of breaks but can be observed on the juvenile specimen SM PW5A-0042 (Martin 1994). The shape of the middle cervical vertebra SM K11-0017 is similar to that of the type SM PW1-0002 (Fig. 19). The centrum is elongated and strongly opisthocoelous with a hemispherical cranial articular surface. The neural spine is relatively high. The neural spine of SM K11-0017 is divided in the same way as in SM PW1-0002, but the latter is situated lower than the postzygapophyses. The postzygapophyseal facet of SM PW1-0002 is larger than in SM K11-0017 and more oblique dorsally. The position of the neural spine and the orientation of postzygapophyses in SM PW1-0002 suggest that it is located more caudally in the vertebral column than SM K11-0017. The centrum SM K11-0020 and neural arch SM K11-0021 share several characteristics with SM PW1-0003, which can be considered as a posterior cervical or an anterior dorsal. The anterior articular surface of the centrum is hemispherical. The centrum is compressed dorsoventrally in lateral view. On the right lateral surface the parapophyses are situated cranially in the ventral half of the lateral surface of the centrum. The neural spine is divided, with a U-shaped cleft in anterior view. The blade of the neural spine is swollen, thin

Fig. 18. Comparison of cervical vertebrae of Phuwiangosaurus sirindhornae between the type and Ban Na Khrai specimens. (a, b) Cervical 5 of the type specimen (SM PW1-0001); (c, d) cervical 5 of Ban Na Khrai specimen (SM K11-0013). (a, c) Anterior view; (b, d) left lateral view. Scale bar: 10 cm.

212

S. SUTEETHORN ET AL.

Fig. 19. Comparison of cervical vertebrae of Phuwiangosaurus sirindhornae between the type and Ban Na Khrai specimens. (a, b) Cervical 10 (?) of the type specimen (SM PW1-0002); (c, d) cervical 9 of Ban Na Khrai specimen (SM K11-0017). (a, c) Right lateral view; (b, d) anterior view. Scale bar: 10 cm.

cranially and becomes thicker posteriorly with a posterolateral curve. The spine is higher than the postzygapophyses. The neural spine of SM K11-0021 is close to that of SM PW1-0003. SM PW1-0003 is characterized by a very tall neural spine and a deep bifurcation of this spine, which differs from SM K11-0021 and the more posterior spines from K11. This difference probably results from the different position in the vertebral column and from a differential compression (dorsoventral in SM K11-0021 and transversal in SM PW1-0003). The posterior dorsal neural arches SM K11-0038 and SM K11-0042 present the same characters as the neural arches SM PW1-0006 and SM PW1-0023 (topotype) (Suteethorn et al. 2009), respectively (Fig. 20). The spine is unforked. The prezygapophyses are flat and close to each other. Parapophyses are at the same level as the prezygapophyses. The long diapophyses are flattened craniocaudally. The dorsal vertebrae from K11 are

well preserved but all neural arches and centra are unfused, which is probably linked to the fact that the sauropod from K11 is a young individual.

Pelvic girdle and appendicular skeleton The pelvic girdle and the appendicular skeleton of the specimen from K11 present the same shape as the homologous bones from the type of Phuwiangosaurus sirindhornae (PW1). The right ilium SM K11-0147 is well preserved. The preacetabular process tapers and projects laterally. The pubic peduncle of the ilium is straight, long and directed at right angles to the direction of the blade. The ischiatic peduncle of the ilium is faintly marked. SM K11-0147 shares these characteristics with SM PW1-0011 but it is smaller. The right pubis SM K11-0148 is well preserved. The iliac peduncle is broadly expanded into a large blade. The ischiatic peduncle is elongated and tapers

NEW PHUWIANGOSAURUS SKELETON

213

Fig. 20. Comparison of dorsal vertebrae of Phuwiangosaurus sirindhornae between the topotype and Ban Na Khrai specimens. (a– c) Dorsal 12 of the topotype specimen (SM PW1-0023); (d – f) dorsal 12 of Ban Na Khrai specimen (SM K11-0042). (a, d) Anterior view; (b, e) lateral view; (c, f) posterior view. Scale bar: 10 cm.

caudally. The obturator foramen is completely enclosed in the bone. SM K11-0148 has exactly the same characteristics as the right pubis SM PW1-0013. Both ischia, SM K11-0149 (left) and SM K11-0150 (right), are well preserved. The pubic peduncle of the left ischium is incomplete. They have the same characteristics as SM PW1-0014 (left) and SM PW1-0015 (right), with a well-marked curvature of the caudal border of the shaft. The right femur SM K11-0151 is damaged whereas the left one (SM K11-0152) is very well preserved. The head is situated above the level of the greater trochanter. The shaft of the femur is flattened craniocaudally. The fourth trochanter, directed medially, is situated on the medial edge of the shaft above mid-length of the femur. Two developed epicondyles are separated by a wide groove. The medial condyle is larger transversally than the lateral one and very prominent. Both femora from K11 share these characteristics with SM PW1-0016 (left) and SM PW1-0017 (right). Both fibulae, SM K11-0153 (left) and SM K11-0154 (right), are well preserved. In proximal view the proximal end of the fibula is expanded craniocaudally. The medial edge presents a sigmoid

curvature, concave cranially and convex caudally. The cranial ridge is thin and expanded to the middle of the fibula, forming a shaft. The distal end is triangular in distal view, flat caudally and points cranially. Both fibulae from K11 present the same characteristics as SM PW1-0018.

Discussion and conclusion In all characters observable in both specimens, the skeleton from K11 matches the type specimen of Phuwiangosaurus sirindhornae. We thus conclude that this sauropod does belong to the species P. sirindhornae. The difference in size is linked to a difference in age. Klein & Sander (2009) have suggested that there is a general correlation between bone size (body size) and ontogenetic stage (age) in P. sirindhornae. Based on the unfused vertebrae and the structure of the bone tissue, those workers interpreted the P. sirindhornae skeleton from K11 as a young or subadult (ontogenetic stage 7 out of 13 of Klein & Sander) whereas the type is nearly a fully grown adult (ontogenetic stage 12 out of 13). We can also conclude that the skull elements from K11 do belong to P. sirindhornae

214

S. SUTEETHORN ET AL.

as all the described bones belong to a single individual. The skeleton of P. sirindhornae from K11 is very well preserved and the bones represent almost 60% of this sauropod skeleton. The cranial elements of P. sirindhornae show close resemblances to those of the Late Cretaceous Nemegtosaurus (Nowinski 1971) from Mongolia. They share a postorbital– squamosal contact vertically oriented, a squamosal with spur, quadrate condyle facing ventrolaterally, and peg-like teeth with high-angled apical wear facet. Unfortunately, the postcranial skeleton of Nemegtosaurus is unknown. The postcranial skeleton of P. sirindhornae represents that of a basal titanosaur, more derived than Brachiosaurus but more primitive than Malawisaurus, following Upchurch et al. (2004) and Curry Rogers (2005). In the near future a phylogenetic analysis of P. sirindhornae will thus be complemented with new information drawn from the skeleton from K11 and information from the nearly complete skeleton of the derived titanosaur Ampelosaurus atacis (Le Loeuff 1995) from southern France. This should shed light on the early stages of the evolutionary history of the group of Cretaceous Asian sauropods to which Phuwiangosaurus belongs. This study was supported by the University of Mahasarakham (Thailand), the Department of Mineral Resources (Bangkok, Thailand), the Muse´e des Dinosaures (Espe´raza, France), the ECLIPSE Programme of CNRS and a joint project of the Centre National de la Recherche Scientifique and the Thailand Research Fund. The authors would like to thank L. Cavin and J. Claude for their help and comments. Thoughtful review comments made by J. A. Wilson and L. Salgado greatly improved the manuscript.

References A PESTEGUI´ A , S. 2005. Evolution of the hyposphene– hypantrum complex within sauropoda. In: T IDWELL , V. & C ARPENTER , K. (eds) Thunder-lizards: the Sauropodomorph Dinosaurs. Indiana University Press, Bloomington, 248– 267. B UFFETAUT , E. & S UTEETHORN , V. 1989. A sauropod skeleton associated with theropod teeth in the Upper Jurassic of Thailand: Remarks on the taphonomic and palaeoecological significance of such associations. Palaeogeography, Palaeoclimatology, Palaeoecology, 73, 77– 83. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 83–90. B UFFETAUT , E. & S UTEETHORN , V. 1999. The dinosaur fauna of the Sao Khua Formation of Thailand and the beginning of the Cretaceous radiation of dinosaurs in Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 150, 13–23.

B UFFETAUT , E. & S UTEETHORN , V. 2004. Comparative odontology of sauropod dinosaurs from Thailand. Revue de Pale´obiologie, Volume Spe´cial, 9, 151–159. B UFFETAUT , E., S UTEETHORN , V., L E L OEUFF , J., C UNY , G., T ONG , H. & K HANSUBHA , S. 2002. A review of the sauropod dinosaurs of Thailand. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Resources, Bangkok, 95– 101. B UFFETAUT , E., S UTEETHORN , V., C UNY , G., K HANSUBHA , S., T ONG , H., L E L OEUFF , J. & C AVIN , L. 2003. Dinosaurs in Thailand. Mahasarakham University Journal, Special Issue, 22, 69–82. C HATTERJEE , S. & Z HENG , Z. 2005. Neuroanatomy and dentition of Camarasaurus lentus. In: T IDWELL , V. & C ARPENTER , K. (eds) Thunder-lizards: the Sauropodomorph Dinosaurs. Indiana University Press, Bloomington, 199–211. C URRY R OGERS , K. A. 2005. Titanosauria: A phylogenetic overview. In: C URRY R OGERS , K. A. & W ILSON , J. A. (eds) The Sauropods: Evolution and Paleobiology. University of California Press, Berkeley, 50– 103. C URRY R OGERS , K. A. & F ORSTER , C. A. 2004. The skull of Rapetosaurus krausei (Sauropoda: Titanosauria) from the Late Cretaceous of Madagascar. Journal of Vertebrate Paleontology, 24, 121– 144. G OMANI , E. M. 2005. Sauropod dinosaurs from the Early Cretaceous of Malawi, Africa. Palaeontologia Electronica, 8, 27A. World Wide Web Address: http:// palaeo-electronica.org/paleo/2005_1/gomani27/ issue1_05.htm H ARRIS , J. D. & D ODSON , P. 2004. A new diplodocoid sauropod dinosaur from the Upper Jurassic Morrison Formation of Montana, USA. Acta Palaeontologica Polonica, 49, 197–210. I NGAVAT , R. & T AQUET , P. 1978. First discovery of dinosaur remain in Thailand. Journal of the Geological Society of Thailand, 3, 1– 6. I NGAVAT , R., J ANVIER , P. & T AQUET , P. 1978. De´couverte en Thaı¨lande d’une portion de fe´mur de dinosaure sauropode. Comptes-Rendus Sommaires de la Socie´te´ Ge´ologique de France, 3, 140– 141. K LEIN , N., S ANDER , M. & S UTEETHORN , V. (2009). Bone histology and its implications for the life history and growth of the Early Cretaceous titanosaur Phuwiangosaurus sirindhornae. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 213– 224. K SEPKA , D. T. & N ORELL , M. A. 2006. Erketu ellisoni, a long-necked sauropod from Bor Guve´ (Dornogov Aimag, Mongolia). American Museum Novitates, 3508, 1– 16. L E L OEUFF , J. 1995. Ampelosaurus atacis (nov. gen., nov. sp.), a new titanosaurid (Dinosauria, Sauropoda) from the Late Cretaceous of the Upper Aude Valley (France). Comptes Rendus de l’Acade´mie des Sciences, 321, 693 –699. M ARTIN , V. 1994. Baby sauropods from the Sao Khua Formation (Lower Cretaceous) in Northeastern Thailand. Gaia, 10, 147– 153.

NEW PHUWIANGOSAURUS SKELETON M ARTIN , V., B UFFETAUT , E. & S UTEETHORN , V. 1993. Jurassic sauropod dinosaurs of Thailand: a preliminary report. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology. Chiang Mai University, Chiang Mai, 415– 425. M ARTIN , V., B UFFETAUT , E. & S UTEETHORN , V. 1994. A new genus of sauropod dinosaur from the Sao Khua Formation (Late Jurassic or Early Cretaceous) of Northeastern Thailand. Comptes Rendus de l’Acade´mie des Sciences, 319, 1085– 1092. M ARTIN , V., S UTEETHORN , V. & B UFFETAUT , E. 1999. Description of the type and referred material of Phuwiangosaurus sirindhornae Martin, Buffetaut and Suteethorn, 1994, a sauropod from the Lower Cretaceous of Thailand. Oryctos, 2, 39–91. M C I NTOSH , J. S. 1990. Sauropoda. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria. University of California Press, Los Angeles, 345–401. M EESOOK , A., S UTEETHORN , V., C HAODUMRONG , P., T EERARUNGSIGUL , N., S ARDSUD , A. & W ONGPRAYOON , T. 2002. Mesozoic rocks of Thailand: a summary. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on the Geology of Thailand. Department of Mineral Resources, Bangkok, 83–94. N OWINSKI , A. 1971. Nemegtosaurus mongoliensis n. gen. n. sp. (Sauropoda) from the Uppermost Cretaceous of Mongolia. Palaeontologia Polonica, 25, 57–81. R ACEY , A. & G OODALL , J. G. S. 2009. Palynology and stratigraphy of the Mesozoic Khorat Group red bed

215

sequences from Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 67– 81. S UTEETHORN , S., LE L OEUFF , J., B UFFETAUT , E. & S UTEETHORN , V. 2009. Description of topotypes of the sauropod Phuwiangosaurus sirindhornae (Phu Wiang, Changwat Khon Kaen). Neues Jahrbuch fu¨r Geologie und Pala¨ontologie (accepted). S UTEETHORN , V., M ARTIN , V., B UFFETAUT , E., T RIAMWICHANON , S. & C HAIMANEE , Y. 1995. A new dinosaur locality in the Lower Cretaceous of northeastern Thailand. Comptes Rendus de l’Acade´mie des Sciences, Se´rie IIa, 321, 1041–1047. U PCHURCH , P. 1998. The phylogenetic relationships of sauropod dinosaurs. Zoological Journal of the Linnean Society, 124, 43– 103. U PCHURCH , P., B ARRETT , P. M. & D ODSON , P. 2004. Sauropoda. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria, 2nd edn. University of California Press, Los Angeles, 259– 322. W ILSON , J. A. 1999. A nomenclature for vertebral laminae in sauropods and other saurischian dinosaurs. Journal of Vertebrate Paleontology, 19, 639–653. W ILSON , J. A. 2005. Redescription of the Mongolian sauropod Nemegtosaurus mongoliensis Nowinski (Dinosauria: Saurischia) and comments on Late Cretaceous sauropod diversity. Journal of Systematic Palaeontology, 3, 283– 318.

Bone histology and its implications for the life history and growth of the Early Cretaceous titanosaur Phuwiangosaurus sirindhornae NICOLE KLEIN1 *, MARTIN SANDER1 & VARAVUDH SUTEETHORN2 1

Steinmann Institute, Paleontology, University of Bonn, Nussalle 8, 53115 Bonn, Germany 2

Geological Survey Division, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected]) Abstract: Bone histology is the most comprehensive way of obtaining data on growth and life history for dinosaurs. Humeri and femora of the basal titanosaur Phuwiangosaurus sirindhornae from the Early Cretaceous of Thailand were sampled by core drilling. The sample represents growth series with humeri ranging in size from 71.0 to 110.0 cm and femora ranging in size from 38.5 to 112.0 cm. The bone tissue is continuously growing laminar fibro-lamellar bone typical for virtually all sauropods. Several ontogenetic stages can be distinguished, and a general growth pattern is deduced on the basis of different-sized individuals. Humeri differ from femora in generally showing more remodelling by secondary osteons.

In extant animals growth and other life history data can be easily obtained by field or laboratory work (e.g. mark –release –recapture studies, observation). For extinct animals these options do not exist. The most comprehensive way to obtain information about past events in the life of an extinct terrestrial vertebrate is from the growth record preserved in its teeth and bones. Palaeohistology, the study of the microstructure of fossilized bone, gives insights into the growth of an extinct animal, its life history and its general growth pattern, including individual age, age at sexual maturity, the minimal longevity of an individual, and an estimation of growth rates. The study of dinosaur bone histology has become a standard method in recent years. A general overview of the state of the art has been given by Chinsamy-Turan (2005) and Erickson (2005). Studies dealing exclusively with sauropods have been made by Curry (1999), Sander (1999, 2000), Sander & Tu¨ckmantel (2003), Sander et al. (2004, 2006), and Klein & Sander (2008). In most tetrapods, growth in linear dimensions decreases gradually during ontogeny, most of all after sexual maturity is reached, and finally stops completely some time during life history. In modern mammals the attainment of sexual maturity coincides roughly with this final growth stop, or growth continues only for a short period afterwards. In birds maximal size is reached well before sexual maturity (Erickson et al. 2007). Reptiles have a completely different life history and growth pattern. After sexual maturity is reached, they show continued growth, sometimes for

several years (Castanet et al. 1993; Castanet 1994; Sander & Klein 2005; Klein & Sander 2007). The purpose of this study is to describe the long bone histology of Phuwiangosaurus sirindhornae and its ontogenetic variation. This is of importance because P. sirindhornae is the first titanosaur to be studied by bone histology and thus the most highly derived sauropod for which such data have become available. Given that titanosaurs differ in a number of features of the locomotory apparatus from less derived sauropods, suggesting a different style of locomotion (Curry-Rogers 2005; Wilson 2005), it was of interest to see if this is reflected in their long bone histology.

Bone histology, ontogeny and life history studies The bone tissue of all vertebrates can be divided into three main types, a classification based on the organization of the collagenous fibres and bone apatite crystallites integrated with them (FrancillonVieillot et al. 1990). Lamellar bone is highly organized and represents relatively slow growth. Fibrous or woven bone is generally deposited very fast and thus is not well organized. Parallel-fibred bone is in organization and rate of deposition intermediate between the two others. Additionally to the bone matrices, a bone tissue is characterized by the organization of its vascular canals. The possibilities of vascular canal organization are numerous (Francillon-Vieillot et al. 1990).

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 217–228. DOI: 10.1144/SP315.15 0305-8719/09/$15.00 # The Geological Society of London 2009.

218

N. KLEIN ET AL.

In most living reptiles, lamellar–zonal bone forms the bulk of bone; this means that the primary bone tissue consists of lamellar bone, and that the bone tissue is regularly interrupted by growth marks. The vascularization consists of mainly longitudinally arranged vascular canals. During ontogeny, when growth rate decreases, vascularization descreases simultaneously and disappears finally when growth stops completely. The bone tissue is then avascular. Bird, mammal and also dinosaur bone primarily consists of fibro-lamellar bone. In dinosaur long bones, the vascular system is most commonly dominated by a laminar organization, mainly consisting of circumferential vascular canals, and less often of longitudinal, radial or reticular vascular canals (Francillon-Viellot et al. 1990; Chinsamy-Turan 2005; Erickson 2005). In some dinosaurs the bone tissue changes in the outer cortex from fibrolamellar bone to lamellar–zonal bone. This lamellar–zonal bone then gradually becomes avascular and finally forms an external fundamental system (EFS, Horner et al. 2001; outer circumferential layer, OCL; Chinsamy-Turan 2005), indicating an extreme slow-down in growth. Annual growth marks or growth cycles (lines of arrested growth (LAGs), zones, annuli) can be used for skeletochronology (Castanet et al. 1993; Castanet 1994). They are not restricted to a specific bone tissue and are common in many dinosaurs. However, in sauropods they are rare and only irregularly developed except in an EFS. Thus, skeletochronology cannot be applied to most sauropod samples, and the trigger for the development of growth marks in sauropods is not yet clear (Chinsamy & Hillenius 2004; Padian & Horner 2004). One hypothesis is that growth of young sauropods was too fast for growth marks to develop. A similar situation can be observed in modern birds (Chinsamy et al. 1995; Starck & Chinsamy 2002) and some mammals (Sander & Andrassy 2006). Ontogenetic variations in bone histology as well as life histories deduced from bone histology have been studied in various dinosaur taxa (e.g. Chinsamy 1993; Varricchio 1993; Horner et al. 2000; Chinsamy-Turan 2005; Erickson 2005; Sander & Klein 2005) and in sauropods (Curry 1999; Sander 1999, 2000; Sander & Tu¨ckmantel 2003; Sander et al. 2006). Complete ontogenetic series, from early juveniles to senescent adults, were described for several sauropod taxa by Klein & Sander (2008), who provided a detailed description and definition of successive ontogenetic bone tissue types and histological ontogenetic stages (HOS). In this study, we apply these ontogenetic bone tissue types and histological ontogenetic stages to Phuwiangosaurus.

Phuwiangosaurus sirindhornae Phuwiangosaurus was a medium-sized sauropod dinosaur of around 15 –20 m body length (Martin et al. 1999). Although it seems relatively well supported that Phuwiangosaurus is a basal titanosaur (Upchurch et al. 2004; Curry-Rogers 2005), its exact systematic position is not yet clear. Phuwiangosaurus is found in several localities in northern Thailand (Khon Kaen Province), with large numbers of specimens at different ontogenetic stages (Martin 1994; Martin et al. 1994, 1999). The bones are mainly found in floodplain deposits of low-energy meandering river systems of the Sao Khua Formation. These beds have been dated as Early Cretaceous by palynology (Racey et al. 1994). Based on sedimentological and palynological data, the climate in northern Thailand during the Early Cretaceous was semiarid with two distinct seasons (Martin et al. 1999). Sauropod trackways found in South Korea provide evidence for gregarious behaviour in some sauropods, not only among the same age classes but also for different-sized individuals, from early juveniles to adults (Lockley 1994). This is difficult to envisage because of the great size and weight difference between juvenile and adult sauropods. However, the finds of juvenile and adult bones of Phuwiangosaurus in the same sites in northern Thailand seem to support this behaviour (Martin 1994). Limb bones of varying size classes are common in these localities, providing a good database for a bone histological study of Phuwiangosaurus. Bones are usually found in bone-beds but partially articulated or associated skeletons are also common (Martin et al. 1999; see Table 1 this study). Morphological studies have referred the material to a single species, Phuwiangosaurus sirindhornae (Martin 1994; Martin et al. 1999), but it seems possible that more than one species is represented (E. Buffetaut, pers. commun.).

Method of sampling As has been shown repeatedly in histological studies, limb bones are best suited for studies of the growth record because of their appositional growth. In their midshaft areas most of the growth record of the living individual may be preserved (Sander 1999, 2000; Chinsamy-Turan 2005; Erickson 2005; Klein & Sander 2007, 2008). Most samples were obtained by core drilling of long bones, but some by sectioning of fragmentary bones. The coring method has been described in detail elsewhere (Sander 1999, 2000; Klein & Sander 2007). The cores were obtained using a diamond-studded coring bit mounted in an electrical

PHUWIANGOSAURUS BONE HISTOLOGY

219

Table 1. The Phuwiangosaurus sirindhornae material sampled and its histology Bone

Bone length (cm)

Specimen number

Comments or corresponding femur length

Femur

38.5

K16-20

Individual B

Femur

39.0

K16-33

Individual B

Femur

41.0

PW5 A-1

Individual C

Femur

42.0

PW5 A-2

Individual C

Femur

58.0

Femur I

Femur

62.0

Femur II

Femur

93.0

K4-366

Femur

100.0

K11-1

Femur

 103.0

No number

Femur

 105.0

K4-69

Femur

112.0

K21

Femur

Not measurable

PW4-6

Femur

Not measurable

PW4-9

Femur

Not measurable

PW4-10

Femur

Not measurable

PW4-12

Femur

Not measurable

PW4-18

Femur

Not measurable

PW4-20

Humerus

71.0

K4-428

89.0 cm

Humerus

73.0

K4-162

91.3 cm

Individual D

Bone tissue type

Type B bone tissue grading into type C bone tissue Type B bone tissue grading into type C bone tissue Type B bone tissue, type C bone tissue, and type D bone tissue Type B bone tissue, type C bone tissue, and type D bone tissue Type E bone tissue grading into type F bone tissue Type E bone tissue grading into type F bone tissue Type E bone tissue grading into type F bone tissue Type C bone tissue grading into type D bone tissue Type E bone tissue grading into type F bone tissue Type F bone tissue grading into type G bone tissue Type E bone tissue grading into type F bone tissue Type B bone tissue grading into type C bone tissue Possible type A bone tissue Possibly type A bone tissue Type B bone tissue grading into type C bone tissue Type A bone tissue grading into type B bone tissue Type B bone tissue grading into type C bone tissue Type E bone tissue grading into type F bone tissue Type E bone tissue grading into type F bone tissue

HOS

HOS-5 HOS-5 HOS-5.5

HOS-5.5

HOS-10 HOS-10 HOS-10.5 HOS-7 HOS-10 HOS-12.5 HOS-11 HOS-4 HOS-1 HOS-1 HOS-5 HOS-3 HOS-4 HOS-10.5 HOS-10

(Continued)

220

N. KLEIN ET AL.

Table 1. Continued Bone

Bone length (cm)

Specimen number

Comments or corresponding femur length

Humerus

77.0

K1-28

96.3 cm

Humerus

102.0

PW1-8

Humerus

110.0

KD2-1

125.0 cm individual A 137.5 cm

Tibia

. 32.0

Tibia I

Tibia

. 34.0

Tibia II

Pubis

Not measurable

PW1-13

Individual A

Pubis

62.0

K11-123

Individual D

Pubis

Not measurable

K1-15

Individual A

Rib

Not measurable

K11-75

Rib

Not measurable

K11-14

Bone tissue type

HOS

Type E bone tissue grading into type F bone tissue Type F bone tissue

HOS-11

Type E bone tissue grading into type F bone tissue Type C and type D bone tissue Type D bone tissue grading into type E bone tissue Completely remodelled by secondary osteons Similar to type D bone tissue Almost completely remodelled by secondary osteons Similar to type E bone tissue Similar to type C bone tissue grading into type D bone tissue

HOS-11.5

HOS-12

HOS-7 HOS-8.5 Not applicable Not applicable Not applicable Not applicable Not applicable

The classification of the bone tissue types to histological ontogenetic stages (HOS) follows Klein & Sander (2008).

drill with adjustable speed (ideally, torque of the drill should be low and increase with speed). The drill was mounted in a portable drill press. The internal diameter of the coring bit used was 13 mm. To reproduce results and make comparison between taxa possible, cores were drilled at a standardized location. In humeri usually the posterior bone side and in femora the anterior one was sampled in the midshaft area. The Phuwiangosaurus material is very well preserved and well mineralized. Thus, in some samples it was possible to drill from one bone side through the medullary cavity to the opposite bone side, thus sampling the cortex twice. After recovery from the hole, the cores were embedded in synthetic resin and cut in half perpendicular to the long axis of the bone. Thus the plane of section represents a segment of the cross-section of the bone at midshaft and records the appositional growth in this region. One half of the sample was processed into a standard petrographic thin section and the other half into a polished section (Sander 2000; Klein & Sander 2007). The thin sections were then studied under transmitted light, both normal and polarized. The

polished section were studied in incident light using bright field illumination. However, the polished sections from the Phuwiangosaurus sample were not very informative and did not show, for example, the polish lines described by Sander (2000) in sauropods and by Klein & Sander (2007) in a prosauropod. Thus, the polished sections were used only to augment the result from the thin sections.

Material The Phuwiangosaurus sirindhornae material sampled for this study is stored in the palaeontological collection of the Department of Mineral Resources (PC.DMR) of the Government of Thailand, Khon Kaen Province, Kalasin, Thailand. The thin sections PW 4-6, PW 4-9, PW 4-10, PW 4-12, PW 4-18 and PW 4-20 were cut by V. Martin (Martin 1994) but were not published by her. These sections were made available to us by V. Suteethorn & E. Buffetaut in 2005. The sections of Martin originate from femur fragments of juvenile

PHUWIANGOSAURUS BONE HISTOLOGY

Phuwiangosaurus specimens (Martin 1994). The other samples were obtained by one of us (N.K.) on a trip to Thailand in 2005. A total of 31 bones of Phuwiangosaurus were sampled for our bone histological studies. Eleven femora represent a size range from 38.5 to 112.0 cm. The femur sample also includes the eight incomplete femora sampled by Martin. The thin sections were of limited value because of the unknown dimension of these bones, but bone histology indicates that they indeed represent juveniles (Table 1). The humerus sample consists of five humeri, ranging from 71.0 to 110.0 cm in length. The largest humerus (110.0 cm) indicates a larger maximum body size than is represented by the maximum femur length (112.0 cm) because the ratio of humerus to femur length as measured in the holotype is 80% (Martin et al. 1999). Thus the largest humerus in the sample belonged to an individual of around 137.5 cm femur length. The remaining seven sampled bones belong to tibiae, pubes and ribs. However, the focus of this study is on the histology of the humeri and femora. Some individuals (Table 1) were sampled from more than one bone to evaluate differences in histology beween the different bones of the same skeleton.

Results Description of bone tissue types and histological ontogenetic stages The cortex of the humeri and femora shows typical laminar fibro-lamellar bone. In many bones, the medullary region is crushed, and the originally cancellous structure made up of bony trabeculae is destroyed and pressed together. A similar thing happened to the remodelling zone, which consists of large erosion cavities. However, if preserved, the medullary region is not an open cavity but its centre is characterized by large erosion cavities. Towards the inner cortex, the erosion cavities become more rectangular. All the erosion cavities are connected by bony trabeculae of lamellar bone, given the whole medullary cavity a cancellous or spongy appearance. Superficial to the medullary cavity is the remodelling zone of the inner cortex, likewise consisting of erosion cavities, which, however, are much smaller, more round to oval in shape, and with their long axis oriented parallel to the bone surface. In the smaller bones, the erosion cavities are medium sized to large and arranged in a layer attached to the medullary cavity. In larger bones, the erosion cavities become smaller and more scattered, and the tissue between has started to develop secondary osteons.

221

The primary cortex, although exclusively made of laminar fibro-lamellar bone, consists of different bone tissue types, the presence and sequence of which can be used to assign the individuals sampled to a histological ontogenetic stage (HOS) and a biological ontogenetic stage (Klein & Sander 2008). All these bone tissue types can be easily distinguished on the base of the density and organization of the vascular system, the development of the primary osteons, the relative amount of woven (fast-growing bone tissue), parallel-fibred (intermediate-growing bone tissue) or lamellar bone (slow-growing bone tissue) within the fibro-lamellar complex, the density of secondary osteons, and the deposition of growth marks (Klein & Sander 2008). These changes in bone microstructure are all linked to a gradually decreasing rate of linear growth from very young to fully grown individuals. If annual growth marks are expressed, these ontogenetic stages can be roughly linked with an individual age in years. The thin sections of the fragmentary Phuwiangosaurus femora sampled by V. Martin show type A to C bone tissues (Table 1). The incomplete femora were studied first hand in 2005 and can be unequivocally identified as femora, but sampling location was sometimes not in the midshaft area of the bone, which make comparison difficult. The smallest two femora (K16-20, K16-33, Figs 1a and 3b –d) pertain to one individual, and at 38.5 cm and 39.0 cm, respectively, show the remains of type B bone tissue in the remodelling zone followed by a thick layer of type C bone tissue (HOS-5). In the femora (PW5 A-1, PW5 A-2, Figs 1a and 3b – d) of a slightly larger individual the thick layer of type C bone tissue grades towards the outer cortex into type D bone tissue (HOS-5.5). These bones show neither secondary osteons nor any growth marks. Interestingly, one of the largest femora (K11-1, 100.0 cm) also shows type D bone tissue through most of its cortical section. Only in the outer cortex does the type D bone tissue grade into type E bone tissue (HOS-7). Secondary osteons are still few in number in this sample, but it shows a growth mark, an annulus, developed in its outer cortex. Two femora, 58.0 cm and 62.0 cm long (femur I, femur II), are strongly altered diagenetically, resulting in a remarkable expansion of the cortex (Figs 1f and 3a). Crystal growth originating in the circumferential vascular canals has spread the successive bone laminae apart, commonly to an extent that the former vascular canals are twice as wide as the bone tissue in-between. In these femora, the primary bone tissue is difficult to evaluate. The most one can say is that the bone tissue type in the outer cortex is type E bone tissue, but no exact HOS can be assigned. The primary cortices of a humerus

222

N. KLEIN ET AL.

Fig. 1. (a) Thin section of left femur PW5 A-1 (41.0 cm). The main part of the cortex consists of laminar type C bone tissue grading in the outermost cortex into type D bone tissue. In the inner cortex remains of non-laminar type B bone tissue are preserved. The histological ontogenetic stage of this bone is HOS-5. The sample shows no real growth marks but cracks that were not accompanied by a change in histology (e.g. vascularity). (b) Thin section of right femur K4-366 (93.0 cm). (c) Thin section of left humerus K4-428 (71.0 cm). The primary cortex of these two bones is

PHUWIANGOSAURUS BONE HISTOLOGY

73.0 cm in length (K4-162), a femur 93.0 cm in length (K4-366, Figs 1b and 3e, f) and a femur 103.0 cm in length (no number) consists of type E bone tissue. These bones have some growth marks developed in the outer third of the cortex, and are all of HOS-10 to HOS-10.5. Type E bone tissue grading into type F bone tissue is developed in two humeri with a length of 71.0 cm (K4-428, Fig. 1c) and 77.0 cm (K1-28), respectively, and one femur (K21, 112.0 cm, Figs 1e and 3g, h). Their HOS is HOS-10.5 to HOS-11. Humeri PW1-8 (102.0 cm) and KD2-1 (110.0 cm, Fig. 1d) show type E to type F bone tissue with a number of growth marks deposited in the outer cortex. Both show strong remodelling by secondary osteons. These two humeri are of HOS-11.5 to HOS-12. In none of the Phuwiangosaurus bones is an EFS developed, which is contrary to the definition of the type F bone tissue in diplodocoid and basal marcronarian sauropods of Klein & Sander (2008). A 105 cm long femur (K4-69) is rather atypical. The sample, consisting of a core drilled completely through the bone, shows a large uncrushed medullary cavity but thin cortices at the anterior and the posterior bone side. On both sides, the primary cortex is almost completely remodelled by secondary osteons (type G bone tissue, HOS-12.5). However, at 4.0 cm, the thickness of the bone as recorded in the sample is very low, and the cortices are very thin for such a large bone. Possibly, the outer part of both cortices has been removed by weathering or preparation, the section thus preserving only the inner, remodelled part of the cortex and thus suggesting an older ontogenetic stage than is actually true.

Correlation between body size and histological ontogenetic stage Based on the infilling of primary osteons by lamellar bone and the relative amount of parallel-fibred bone within the fibrolamellar complex, the bone tissue in the larger bones belongs to older individuals. This is true despite the lack of strong remodelling by secondary osteons in femora and closely spaced LAGs. Because growth marks in Phuwiangosaurus are restricted to the outer cortex, they are not useful in establishing a

223

correlation between the HOS and actual age in years in Phuwiangosaurus. A plot of body size v. HOS (Fig. 2) indicates that there is a good correspondence between these two parameters in the sample. An outlier is represented by the femur 100.0 cm in length (K11-1), which shows a rather young HOS (HOS-7) compared with the other individuals of this size. Two explanations are conceivable. One is that the 100.0 cm femur (K11-1) does not belong to the same taxon as the other bones. Another possibility is that a sexual dimorphism is represented by the Phuwiangosaurus sample. The 100.0 cm femur (K11-1) would then represent the faster growing sex that attained a larger final size. Individual variation as a cause is excluded because the size difference is beyond the range of individual variation seen in the remainder of this sample and in other sauropods (Klein & Sander 2008). According to an unpublished MSc thesis (S. Suteethorn pers. comm.), femur K11-1 belongs to a partly articulated skeleton (bones K11-1 to K11-160) that undoubtedly shares all the characters of the type specimen of P. sirindhornae (PW1-1 to PW1-21). A humerus of the type specimen was also sampled for the current study (PW1-8). However, the current sample is not large enough to properly differentiate between sexual dimorphism in P. sirindhornae and the presence of two different taxa.

Reconstruction of life history Phuwiangosaurus hatchlings grew very fast up to about one-third to one-fourth (40.0 cm femur length) of their maximum adult size (presumably at least 140.0 cm femur length). Growth continued at a high, although reduced rate in the juveniles until sexual maturity was reached. The onset of sexual maturity is difficult to constrain because of the size gap in the sample but must have occurred between a femur length of 45.0 and 90.0 cm. A femur length of around 70.0– 80.0 cm appears likely, which would be half of the maximum adult size. Growth continued for several years after sexual maturity was reached, with a growth rate slow enough for growth marks to develop. Because the largest bones sampled (the femur 112.0 cm in

Fig. 1. (Continued ) built mainly of type E bone tissue. The histological ontogenetic stage is HOS-10.5, which indicates a still growing bone, but which clearly grades into type F bone tissue. (d) Thin section of left humerus KD2-1 (110.0 cm). (e) Thin section of left femur K21 (112.0 cm). The primary cortex in these two bones is made primarily of type F bone tissue. The histological ontogenetic stage is HOS-11 to HOS-12, indicating a clear decrease in growth rate, finally resulting in a complete growth stop. (f) Thin section of right femur II (62.0 cm) showing massive diagenetic overprint. The medullary region is in the middle of the image, whereas at the top and bottom the inner and middle cortex is visible. (b –e) show the entire thin section obtained from the core, and the outer bone surface is at the top of the image. Scale bars in (b– f), 3 mm.

224

N. KLEIN ET AL.

Fig. 2. Body size as represented by femur length plotted against histological ontogenetic stage (HOS) in the Phuwiangosaurus sample. Humeri are plotted for their corresponding femur size. Except for three problematic femora (I, II, K4-69) and femur K11-1, there is a close correspondence between body size and ontogenetic stage.

length and especially the humerus 110.0 cm in length) did not show a growth stop, the sampled individuals probably do not represent the maximum size of Phuwiangosaurus. Because growth marks occur only in the outer cortex in Phuwiangosaurus, a correlation of the ontogenetic stages with actual years is not possible. Based on other bone histological studies on sauropodomorphs (Sander 1999, 2000; Sander & Klein 2005; Klein & Sander 2007), sexual maturity was presumably reached between 10 and 15 years, or perhaps even earlier.

Histological differences between humeri and femora In Phuwiangosaurus, the humeri generally show more remodelling by secondary osteons than the femora corresponding in size and HOS. In the humeri, secondary osteons are dense in the inner cortex, and scattered osteons extend to the outer cortex. Even in the largest femora, secondary osteons are mainly restricted to the remodelling zone. Except for the ontogenetic variation, no other differences in the primary bone tissue are observed between femora and humeri. However, the trigger and function of bone remodelling by secondary osteons is not yet well understood (Currey 2002). The different amount of secondary osteons in Phuwiangosaurus humeri and femora could be an indication of different strains and forces during limb growth and locomation, or may simply reflect the slower apposition rate of the cortical bone in the humerus compared

with the femur because of the smaller mid-shaft diameter of the humerus.

Discussion The size gap A complete understanding of ontogenetic changes in bone histology of Phuwiangosaurus is hampered by a conspicuous size gap in the current sample. Whereas femora around 40.0 cm in length are well represented, as are large ones, medium-sized femora are not present. The two medium-sized femora (femur I, 58.0 cm; femur II, 62.0 cm) discussed above are from an unknown locality and of uncertain taxonomic affinity, in addition to being strongly altered diagenetically. The humeri do not cover this size class either, because the smallest humerus (K4-428, 71.0 cm) corresponds to a femur length of 89.0 cm. This is close to the smallest of the large femora in the sample, which has a length of 93.0 cm (K4-366). Bone histology reflects this size gap in the lack of subadult and young, still fast-growing adults (HOS-6 to HOS-9).

HOS and biological ontogenetic stages The application of biological ontogenetic stages such as embryo, hatchling, juvenile, subadult and adult is in common use for extinct taxa. However, their use is problematic because they are difficult to verify in fossils, and because they are inconsistently used in modern animals (Klein & Sander 2008).

PHUWIANGOSAURUS BONE HISTOLOGY

225

Fig. 3. (a) Detail of the diagenetically expanded and modified bone tissue in femur II (62.0 cm). The expansion, best seen in the middle of the image, was brought about by radial crystal growth initiated in the vascular canals. (b) Overview of type C bone tissue in left femur PW5 A-1 (41.0 cm) with laminar and reticular vascular organization. (Note the high density of wide and large vascular canals.) (c, d) Detail of type C bone tissue in left femur PW5 A-1 (41.0 cm). (c) Normal light; (d) polarized light. (Note the decreased size of vascular canals and their regular laminar arrangement.) (e, f) Detail of type E bone tissue in right femur K4-366 (93.0 cm). (e) Normal light; (f) polarized light. (Note the narrow vascular canals and the strongly laminar organization.) (g, h) Detail of type F bone tissue in right femur with no specimen number (c. 103.0 cm). (g) Normal light; (h) polarized light. (Note the thin vascular canals and the clearly increased amount of parallel-fibred and lamellar bone tissue.) Scale bars: 0.5 mm.

Because of its heuristic value, we nevertheless hypothesize a correspondence between bone tissue types detected in this study and biological ontogenetic stages: Type B bone tissue, type C bone tissue, and type D bone tissue are interpreted as typical bone tissue types of hatchling, juvenile, and subadult individuals. Adult (i.e. sexually mature) individuals are represented by three bone tissue types: a stage that records a clearly decreased growth rate but with growth still continuing slowly (type E bone tissue), a stage that records growth having stopped (type F bone tissue), and a final stage in which the primary bone tissue is completely remodelled by secondary osteons (type G bone tissue) and that represents a senescent individual.

The missing HOS-13 Barring the possibly incomplete femur (K4-69, 105.0 cm), a cortex entirely made up by type G bone tissue (HOS 13) is missing from our sample

of Phuwiangosaurus. In particular, none of the humeri, which generally show stronger remodelling by secondary osteons than the femora, are of this HOS, not even the largest humerus (KD2-1, 110.0 cm) which corresponds to a femur 137.5 cm in length. Two hypotheses can be advanced to explain these observations. It is possible that because of the lack of an old enough individual, HOS-13 is not represented in the current sample. This would also imply that the maximum size of Phuwiangosaurus is larger than currently known from the fossil record. This is also supported by the observation that the largest bone (humerus KD2-1, 110.0 cm) does not show a strong slowdown in growth (i.e. an EFS) in its outer cortex. A second hypothesis is that HOS-13 was not expressed in the life history of Phuwiangosaurus. However, this seems rather implausible because the intense remodelling in the humeri indicates that this process would have continued until all primary

226

N. KLEIN ET AL.

bone tissue was replaced by secondary osteons. This complete replacement characteristic for HOS-13 is rather typical for all sauropods known from sufficiently large samples (Klein & Sander 2008), and there is no compelling reason why Phuwiangosaurus should have been any different. We thus favour the first hypothesis and use it to predict that larger individuals of Phuwiangosaurus will be found in the future (see also below).

Phuwiangosaurus, remodelling will affect relatively more of the primary bone than in faster growing diplodocoid and basal macronarian sauropods. If this hypothesis were true, it would support the notion that none of the sampled material of Phuwiangosaurus represents fully grown individuals.

Conclusion Comparisons with other sauropods Phuwiangosaurus bone histology and life history parameters derived from it are similar to those of all other large sauropods that have been studied in detail (Sander 2000; Sander & Tu¨ckmantel 2003; Sander et al. 2004; Klein & Sander 2008). Only the dwarfed Europasaurus (Sander et al. 2006) differs in the development of regularly spaced growth marks early in ontogeny. Phuwiangosaurus differs from the prosauropod Plateosaurus (Sander & Klein 2005; Klein & Sander 2007) in showing a close correlation between body size and ontogenetic stage. This is also seen in all other sauropods including Europasaurus (Klein & Sander 2008) and in the only other well-studied prosauropod, Massospondylus (Chinsamy 1993). One of the differences in the bone microstructure of Phuwiangosaurus from that of other sauropods is the degree of remodelling, which is much stronger in the humeri than in the femora. This is not seen in the other sauropods studied by us so far (Sander 2000; Sander & Tu¨ckmantel 2003; Sander et al. 2004; Klein & Sander 2008), even those that have shorter humeri than Phuwiangosaurus. Some difference in histology is to be expected because smaller bones such as the humerus require lower local bone apposition rates than the largest bones (i.e. the femur). However, these are generally not observed, with the exception of Phuwiangosaurus. The significance of this is unclear at present. Another conspicuous difference in Phuwiangosaurus bone histology from that of other sauropods is the higher amount of parallel-fibred bone in the fibro-lamellar complex of type D bone tissue and type E bone tissue. This can also be observed in the titanosaur Ampelosaurus atacis (Klein et al. 2006; Klein & Sander 2008). This suggests a generally lower growth rate for Phuwiangosaurus and Ampelosaurus compared with diplodocoids and basal macronarians (Klein & Sander 2008). The co-occurence of intense remodelling and the high amount of parallel-fibred bone in Phuwiangosaurus may indicate that remodelling progresses outwards at a constant rate independent of the appositition rate of the primary tissue. If apposition rate is comparatively low as in

We studied Phuwiangosaurus bone histology from growth series of humeri and femora in sections obtained by cutting or core drilling from the midshaft region. Phuwiangosaurus bone histology is similar to that of other large sauropods in being dominated by laminar fibro-lamellar bone. As in other large sauropods, there is a general correspondence between bone size (body size) and histological ontogenetic stage (ontogenetic stage, age). There are indications of two different taxa or a sexual dimorphism in the current sample. The sample may not include very large and old individuals because HOS-13, a fully remodelled cortex and an external fundamental system are missing. The life history of the titanosaur Phuwiangosaurus is similar to that of other sauropods. Phuwiangosaurus differs from other sauropods in that the humerus is much more remodelled than the femur and in a generally lower growth rate. Bone histology is based on destructive methods, therefore the main premise for this kind of study is the open minds of colleagues. Thus, we are very grateful to E. Buffetaut (LMC) for establishing contacts with the Thai colleagues. N.K. is very grateful to the staff at the Sahat Sakhan Dinosaur Museum in Kalasin, Thailand, for their help, support and hospitality. We also would like to thank O. Du¨lfer and G. Oleschinski (both Steinmann Institute, Paleontology, University of Bonn), who cut the thin sections and photographed them for study. The manuscript has benefited from reviews by K. Curry-Rogers and an anonymous reviewer. Our research was funded by the Deutsche Forschungsgemeinschaft (DFG) through grants SA 469/7 and SA 469/16. This paper is contribution 41 of the DFG Research Unit 533 ‘Biology of the Sauropod Dinosaurs’.

References C ASTANET , J. 1994. Age estimation and longevity in reptiles. Gerontology, 40, 174 –192. C ASTANET , J., F RANCILLON -V IEILLOT , H., M EUNIER , F. J. & DE R ICQLES , A. 1993. Bone and individual aging. In: H ALL , B. K. (ed.) Bone. Volume 7: Bone Growth—B. CRC Press, Boca Raton, FL, 245– 283. C HINSAMY , A. 1993. Bone histology and growth trajectory of the prosauropod dinosaur Massospondylus carinatus (Owen). Modern Geology, 18, 319–329.

PHUWIANGOSAURUS BONE HISTOLOGY C HINSAMY , A. & H ILLENIUS , W. J. 2004. Physiology of nonavian dinosaurs. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria. 2nd edn. University of California Press, Berkeley, 643–659. C HINSAMY , A., C HIAPPE , L. M. & D ODSON , P. 1995. Mesozoic avian bone microstructure: Physiological implications. Paleobiology, 21, 561– 574. C HINSAMY -T URAN , A. 2005. The Microstructure of Dinosaur Bone. The Johns Hopkins University Press, Baltimore, MD. C URREY , J. D. 2002. Bones: Structure and Mechanics. Princeton University Press, Princeton, NJ. C URRY , K. 1999. Ontogenetic histology of Apatosaurus (Dinosauria: Sauropoda). New insights on growth rates and longevity. Journal of Vertebrate Paleontology, 19, 654– 665. C URRY -R OGERS , K. 2005. Titanosauria: A phylogenetic overview. In: C URRY -R OGERS , K. & W ILSON , J. (eds) The Sauropods. Evolution and Paleobiology. University of California Press, Berkeley, 50– 103. E RICKSON , G. M. 2005. Assessing dinosaur growth patterns: A microscopic revolution. Trends in Ecology and Evolution, 20, 677– 684. E RICKSON , G. M., C URRY R OGERS , K., V ARRICCHIO , D. J., N ORELL , M. & X U , X. 2007. Growth patterns in brooding dinosaurs reveals the timing of sexual maturity in non-avian dinosaurs and genesis of the avian condition. Biological Letters, 2007, 558–561. F RANCILLON -V IEILLOT , H., C ASTANET , J., G ERAUDE , J., M EUNIER , F. J., S IRE , J. Y., Z YLBERBERG , L. & DE R ICQLES , A. 1990. Microstructure and mineralization of vertebrate skeletal tissues. In: C ARTER , J. E. (ed.) Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends. Van Nostrand Reinhold, New York, 471–530. H ORNER , J. R., DE R ICQLES , A. & P ADIAN , K. 2000. Long bone histology of the hadrosaurid dinosaur Maiasaura peeblesorum: Growth dynamics and physiology based on an ontogenetic series of skeletal elements. Journal of Vertebrate Paleontology, 20, 115– 129. H ORNER , J. R., P ADIAN , K. & DE R ICQLES , A. 2001. Comparative osteohistology of some embryonic and perinatal archosaurs: Developmental and behavioral implications for dinosaurs. Paleobiology, 27, 39– 58. K LEIN , N. & S ANDER , P. M. 2007. Bone histology and growth of the prosauropod dinosaur Plateosaurus engelhardti (von Meyer 1837) from the Norian bonebeds of Trossingen (Germany) and Frick (Switzerland). Special Papers in Palaeontology, 77, 169–206. K LEIN , N. & S ANDER , P. M. 2008. Ontogenetic stages in the long bone histology of sauropod dinosaurs. Paleobiology, 34, 247–263. K LEIN , N., S ANDER , P. M. & L E L OEUFF , J. 2006. An unusual bone histology and growth pattern in Ampelosaurus atacis, a titanosaurid sauropod from South France. Journal of Vertebrate Paleontology, 26, Supplement Issue 3, 85A. L OCKLEY , M. G. 1994. Dinosaur ontogeny and population structure: Interpretations and speculations based on fossil footprints. In: C ARPENTER , K., H IRSCH , K. F. & H ORNER , J. R. (eds) Dinosaur Eggs and Babies. Cambridge University Press, Cambridge, 347– 365.

227

M ARTIN , V. 1994. Baby sauropods from the Sao Khua Formation (Lower Cretaceous) in Northeastern Thailand. Gaia, 10, 147–153. M ARTIN , V., B UFFETAUT , E. & S UTEETHORN , V. 1994. A new genus of sauropod dinosaur from the Sao Khua Formation (Late Jurassic to Early Cretaceous) of northeastern Thailand. Comptes Rendus de l’Acade´mie des Sciences, Se´rie II, 319, 1085– 1092. M ARTIN , V., S UTEETHORN , V. & B UFFETAUT , E. 1999. Description of the type and referred material of Phuwiangosaurus sirindhornae Martin, Buffetaut and Suteethorn 1994, a sauropod from the Lower Cretaceous of Thailand. Oryctos, 2, 39–91. P ADIAN , K. & H ORNER , J. R. 2004. Dinosaur physiology. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria. 2nd edn. University of California Press, Berkeley, 660–671. R ACEY , A., G OODALL , J. G. S., L OVE , M. A., P OLACHAN , S. & J ONES , P. D. 1994. New age data for the Mesozoic Khorat Group of Northeastern Thailand. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245– 252. S ANDER , P. M. 1999. Life history of Tendaguru sauropods as inferred from long bone histology. Mitteilungen Museum fu¨r Naturkunde Berlin, Geowissenschaftliche Reihe, 2, 103– 112. S ANDER , P. M. 2000. Long bone histology of the Tendaguru sauropods: Implications for growth and biology. Paleobiology, 26, 466– 488. S ANDER , P. M. & A NDRASSY , P. 2006. Lines of arrested growth and long bone histology in Pleistocene large mammals from Germany: Implications for dinosaur physiology. Palaeontographica A, 277, 143– 159. S ANDER , P. M. & K LEIN , N. 2005. Unexpected developmental plasticity in the life history of an early dinosaur. Science, 310, 1800– 1802. S ANDER , P. M. & T U¨ CKMANTEL , C. 2003. Bone lamina thickness, bone apposition rates, and age estimates in sauropod humeri and femora. Pala¨ontologische Zeitschrift, 77, 161– 172. S ANDER , P. M., K LEIN , N., B UFFETAUT , E., C UNY , G., S UTEETHORN , V. & L E L OEUFF , J. 2004. Adaptive radiation in sauropod dinosaurs: Bone histology indicates rapid evolution of giant body size through acceleration. Organisms, Diversity and Evolution, 4, 165– 173. S ANDER , P. M., M ATEUS , O., L AVEN , T. & K NO¨ TSCHKE , N. 2006. Bone histology indicates insular dwarfism in a new Late Jurassic sauropod dinosaur. Nature, 441, 739–741. S TARCK , J. M. & C HINSAMY , A. 2002. Bone microstructure and developmental plasticity in birds and other dinosaurs. Journal of Morphology, 254, 232– 246. U PCHURCH , P., B ARRETT , P. M. & D ODSON , P. 2004. Sauropoda. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria.

228

N. KLEIN ET AL.

2nd edn. University of California Press, Berkeley, 259– 322. V ARRICCHIO , D. J. 1993. Bone microstructure of the Upper Cretaceous theropod dinosaur Troodon formosus. Journal of Vertebrate Paleontology, 13, 99–104.

W ILSON , J. A. 2005. Overview of sauropod phylogeny and evolution. In: C URRY -R OGERS , K. & W ILSON , J. A. (eds) The Sauropods: Evolution and Paleobiology. University of California Press, Berkeley, CA, 50–103.

An early ‘ostrich dinosaur’ (Theropoda: Ornithomimosauria) from the Early Cretaceous Sao Khua Formation of NE Thailand ERIC BUFFETAUT1*, VARAVUDH SUTEETHORN2 & HAIYAN TONG1 1

CNRS (UMR 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure), 24 rue Lhomond 75231 Paris Cedex 05, France

2

Bureau of Fossil Research and Museum, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected]) Abstract: Postcranial remains of a small theropod dinosaur, including vertebrae, incomplete pubes, tibiae, an incomplete fibula, metatarsals and phalanges, from the Early Cretaceous Sao Khua Formation of Phu Wiang, Khon Kaen Province, NE Thailand, are described as a new taxon of ornithomimosaur, Kinnareemimus khonkaenensis, gen. et sp. nov. This early ‘ostrich dinosaur’ is characterized by a fairly advanced metatarsus, in which metatarsal III, although still visible proximally between metatarsals II and IV in cranial view, is markedly ‘pinched’ more distally and becomes triangular in cross-section. The condition of its metatarsus shows that Kinnareemimus khonkaenensis is more derived than the geologically younger primitive ornithomimosaurs Harpymimus and Garudimimus, but less derived than Archaeornithomimus. Its occurrence in the Early Cretaceous of Thailand suggests that advanced ornithomimosaurs may have originated in Asia.

The fossil vertebrate locality of Phu Wiang 5, in red clays of the Sao Khua Formation in the Phu Wiang hills near the town of Phu Wiang (Phu Wiang District, Khon Kaen Province, NE Thailand), has yielded a curious dinosaur assemblage consisting mainly of disarticulated bones of large and small individuals of the sauropod Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994 (Martin 1994; Martin et al. 1994, 1999), and a small ornithomimosaur theropod (Buffetaut et al. 1995; Buffetaut & Suteethorn 1998, 1999). The latter represents a new taxon of Ornithomimosauria that, although geologically ancient, shows some derived characters, and provides new evidence about the acquisition of some peculiar characters of the hind limb in the so-called ‘ostrich dinosaurs’.

Geographical and geological setting The Phu Wiang hills are a vast synclinal structure in the western part of the Khorat Plateau of NE Thailand, located just NE of the town of Phu Wiang, and about 100 km NNE of the provincial capital Khon Kaen. Several of the non-marine formations of the Mesozoic Khorat Group crop out in the Phu Wiang syncline, with the Aptian Khok Kruat Formation occupying the centre, followed towards the periphery (and therefore of increasing geological age) by the Phu Phan, Sao Khua and Phra Wihan Formations. Dinosaur footprints are known from the Phra Wihan Formation at Phu Wiang, but the most productive formation in terms

of vertebrate remains is the Sao Khua Formation. The first dinosaur bone identified in Thailand (Ingavat et al. 1978) came from the Sao Khua Formation at Phu Wiang, which also yielded the type skeletons of the sauropod Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn 1994, and the theropod Siamotyrannus isanensis, as well as abundant remains of fishes, turtles and crocodilians. A number of vertebrate localities are known in the Sao Khua Formation at Phu Wiang (distinguished by numbers; see map published by Martin 1994) and have different fossil contents, although all are part of the same general assemblage, the differences being largely due to taphonomic conditions. The Sao Khua Formation is one of the most fossiliferous formations of the Khorat Group. Its age is not well constrained, although it can firmly be placed in the Early Cretaceous because it overlies the Phra Wihan Formation, which has yielded Early Cretaceous palynomorphs (Racey et al. 1996). It is overlain by the Phu Phan Formation, which has yielded very few fossils (except for dinosaur footprints and rare sauropod bones) and is itself overlain by the Khok Kruat Formation, which is referred to the Aptian on the basis of its vertebrates and palynomorphs (Cappetta et al. 1990; Racey et al. 1996). The Sao Khua Formation can thus be considered as later than earliest Cretaceous and earlier than Aptian. A Valanginian–Hauterivian age was suggested by Buffetaut & Suteethorn (1999), but the Sao Khua Formation may be somewhat younger (Barremian?).

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 229–243. DOI: 10.1144/SP315.16 0305-8719/09/$15.00 # The Geological Society of London 2009.

230

E. BUFFETAUT ET AL.

The depositional environment of the Sao Khua Formation was discussed at some length by Mouret et al. (1993) and Racey et al. (1996); it is considered as having been deposited on an extensive floodplain of a low-energy meandering river system. At Phu Wiang 5, the fossil-bearing layer, exposed in a small creek, consists of red clays in which dinosaur bones were found scattered. None of the theropod bones described below were found in articulation, but they almost all come from a small excavation (Phu Wiang 5A), not more than 2 m2, where they were mingled with bones of very young specimens of Phuwiangosaurus (described by Martin 1994; Martin et al. 1999). There is every reason to believe that all the ornithomimosaur bones from that locality are from a few individuals of a single taxon. Other vertebrate fossils found at Phu Wiang 5 include a tooth of the spinosaurid theropod Siamosaurus, crocodile teeth, turtle plates and bivalves (Martin et al. 1999).

third metatarsal (PW5A-103); the proximal end of a left third metatarsal (PW5A-104); the proximal end of a right third metatarsal (PW5A-131); a complete left second metatarsal (PW5A-101); a nearlycomplete left second metatarsal (PW5A-105); the proximal end of a left fourth metatarsal (PW5A102); a complete right fourth metatarsal (PW5A106); two proximal ends of fourth metatarsals (PW5A-108, 109); seven more or less complete pedal phalanges (PW5A-115, 116, 117, 118, 119, 120, 121); an incomplete pedal ungual phalanx (PW5A-122); a complete right tibia (PW5A-110); a complete left tibia (PW5A-111); the proximal end of a left fibula (PW5A-112); the proximal end of a left pubis (PW5A-114); the proximal end of a right pubis (PW5A-113); a dorsal centrum (PW5A-123); an incomplete caudal vertebra (PW5A-124); an anterior caudal centrum (PW5A130); a middle caudal centrum (PW5A-125); four distal caudal centra (PW5A-126, 127, 128, 129).

Description of the material Systematic description Dinosauria Owen 1842 Theropoda Marsh 1881 Coelurosauria von Huene 1914 Ornithomimosauria Barsbold 1976 Kinnareemimus gen. nov. Etymology. From Kinnaree, graceful beings of Thai mythology, with the body of a woman and the legs of a bird, said to inhabit the depths of the legendary Himmapan Forest, by allusion to the bird-like feet of this dinosaur. Diagnosis. As for type species. Kinnareemimus khonkaenensis sp. nov. Etymology. From Khon Kaen Province in NE Thailand, where the type material was found. Diagnosis. An ornithomimid dinosaur in which metatarsal III is visible in cranial view between the proximal ends of metatarsals II and IV (unlike the condition in Archaeornithomimus and more derived ornithomimids), but becomes rod-like distally and expands again, with a triangular crosssection, closer to the distal end (unlike the condition in Garudimimus and Harpymimus). Locality and horizon. Locality Phu Wiang 5, Phu Wiang district, Khon Kaen Province, Thailand; Sao Khua Formation, Early Cretaceous. Holotype. An incomplete left third metatarsal, with the distal end and part of the shaft (PW5A-100). Referred material. The distal end of a right third metatarsal (PW5A-107); the middle part of a left

Vertebrae (Fig. 1). One comparatively large, weakly amphicoelous, hourglass-shaped centrum (PW5A123) may be from a posterior dorsal vertebra, as it lacks chevron facets. The neural arch is missing, having detached at the level of the unfused neurocentral suture, suggestive of an immature individual. Several caudal vertebrae are present. One of the largest (PW5A-124) shows the neural arch completely fused to the centrum, without any clear trace of a suture, which may suggest an adult individual. It is incomplete, only the posterior half being preserved. The centrum is hollow, a condition previously reported for Archaeornithomimus asiaticus (Gilmore 1933). There are no transverse processes, but a longitudinal ridge is present on each lateral face of the centrum. The neural spine appears to have been very low, and the postzygapophyses are broken. The specimen generally resembles the 15th to 17th caudals of Gallimimus bullatus (Osmo´lska et al. 1972). Another similarsized caudal vertebra (PW5A-130) is reduced to its amphicoelous centrum, with an open neurocentral suture. The hourglass-shaped centrum, with chevron facets at the posterior end, is relatively short and high, and its anterior articular surface is wider than the posterior facet. It may be a first caudal. A small specimen (PW5A-128), which also shows longitudinal ridges, has a low neural arch firmly fused to the centrum. The remaining caudal vertebrae (PW5A-126, 127) are small and reduced to their centra, with no sign of fusion along the neurocentral centure. The amphicoelous centra are elongate, hourglass-shaped, with chevron facets at both the anterior and posterior end, linked by paired longitudinal ridges. A larger

EARLY OSTRICH DINOSAUR FROM THAILAND

231

Fig. 1. Kinnareemimus khonkaenensis gen. et sp. nov. Vertebrae in dorsal (a, c, e, g, i, k, m, o) and lateral (b, d, f, h, j, l, n, p) views; (a, b) PW5A-123; (c, d) PW5A-130; (e, f) PW5A-124; (g, h) PW5A-125; (i, j) PW5A-126; (k, l) PW5A-127; (m, n) PW5A-128; (o, p) PW5A-129. Scale bar: 50 mm.

caudal vertebra (PW5A-125) shows chevrons only at the posterior end and a single ventral ridge. All the vertebral elements from Phu Wiang 5 are generally similar to the vertebrae described in other ornithomimosaurs (Barsbold & Osmo´lska 1990). Pubis (Fig. 2). The only pelvic elements present are the proximal ends of a right (PW5A-114) and a left

(PW5A-113) pubis, broken at the beginning of the shaft. A well-marked obturator notch is present. The sutural area for the ilium is large and crescentshaped. A smooth area corresponding to the rim of the acetabulum is short with a trapezoidal outline. The contact area for the ischium is small and smooth, without rugosities. By comparison with the pubes of other ornithomimosaurs, this bone

Fig. 2. Kinnareemimus khonkaenensis gen. et sp. nov. Proximal ends of pubes. Left pubis PW5A-113 in lateral (a) and proximal (b) views; right pubis PW5A-114 in lateral view (c). Scale bar: 10 mm.

232

E. BUFFETAUT ET AL.

appears very slender, but this may be due to immaturity of the specimens. Tibia (Fig. 3). The tibia is represented by two complete specimens, one from the right side (PW5A111) and one from the left (PW5A-110). Although they are nearly the same length, the left tibia is more robust than the right one and probably belongs to a slightly older individual, although sexual dimorphism may also be involved. The distal articular area shows a well-defined cnemial crest which is thickened cranially. In proximal view, the lateral margin of the cnemial crest is markedly concave, forming a distinct notch between the fibular condyle and the cnemial crest, whereas the medial margin is convex. The condyles for articulation with the femur are separated caudally by a distinct incisura. On the cranial surface of the shaft, the cnemial crest decreases quickly in height, being continued along the shaft by a faint ridge, which is more pronounced on the more robust left tibia; a similar condition was described by Sullivan (1997) in a tibia from the Kirtland Formation of New Mexico referred to Ornithomimus antiquus. The fibular crest is prominent, extending to the proximal third of the length of the shaft, and is more strongly developed on the left tibia, in accordance with general

greater robustness. The shaft is slender and rather straight, with only a faint concavity of the medial margin and convexity of the lateral margin. At mid-shaft, the cross-section is D-shaped. The distal end is expanded mediolaterally and compressed craniocaudally. The cranial face shows a flat surface for reception of the ascending process of the astragalus, which is bounded medially by the raised medial rim. The distal articular surface is approximately triangular in outline, the medial margin being distinctly longer than the lateral one. The caudal face bears a blunt ridge that is offset medially. Its lateral rim is raised into a distinct ridge, corresponding to the articulation with the fibula. The tibiae from Phu Wiang 5 are generally similar to those of Late Cretaceous ornithomimids. In Archaeornithomimus asiaticus, however, the proximal articular region shows a proximal protrusion that is not present in Kinnareemimus khonkaenensis. Comparisons with more basal forms are limited because the tibia is unknown in Pelecanimimus polyodon and Shenzhousaurus orientalis, and very incompletely preserved in the holotype of Harpymimus okladnikovi. Garudimimus brevipes has a more robust tibia than Kinnareemimus khonkaenensis, which may be the result of differences in individual age, but otherwise few morphological differences can be seen.

Fig. 3. Kinnareemimus khonkaenensis gen. et sp. nov. tibiae and fibula. Left tibia PW5A-110 in cranial (a), medial (c), caudal (e), lateral (g) and proximal (i) views. Right tibia PW5A-111 in cranial (b), medial (d), caudal (f), lateral (h) and proximal (j) views. Proximal end of left fibula PW5A-112 in medial view (k). Scale bar: 50 mm.

EARLY OSTRICH DINOSAUR FROM THAILAND

In the material from Phu Wiang 5, the astragalus and calcaneum are not attached to the tibia, reflecting the juvenile status of the specimens. Fibula (Fig. 3). The fibula is represented only by a left proximal end (PW5A-112), which is expanded craniocaudally and comma-shaped in proximal view, with the sharp end in caudal position. The lateral face is convex and the medial face concave, showing the inception of a broad and deep groove that apparently extended for some distance along the shaft. A similar groove is present in Gallimimus bullatus (Osmo´lska et al. 1972). Metatarsus. Several more or less complete metatarsal bones, belonging to several individuals, have been found at Phu Wiang 5 and allow a fairly accurate reconstruction of the metatarsus, on which the diagnosis of Kinnareemimus khonkaenensis is based. Metatarsal II (Fig. 4) is represented by two complete (PW5A-101) or nearly complete (PW5A-105) specimens from the left side and two proximal ends from the right side. The proximal articular region is more expanded craniocaudally than mediolaterally. The proximal articular surface is approximately triangular in outline, with a rounded anterolateral corner that overhangs the shaft. The caudal angle is elongate and overhangs the posterior face of the

233

shaft, forming a well-marked rounded process that markedly projects caudally. This process does not project to the same extent in advanced ornithomimids such as Struthiomimus altus (Osborn 1916), Gallimimus bullatus (Osmo´lska et al. 1972) or Sinornithomimus dongi (Kobayashi & Lu¨ 2003). In less derived forms such as Archaeornithomimus asiaticus (Smith & Galton 1990), Garudimimus brevipes (Kobayashi & Barsbold 2005b) and Harpymimus okladnikovi (Kobayashi & Barsbold 2005a), the process is more developed, and therefore resembles the condition in Kinnareemimus khonkaenensis. A first digit is present in Garudimimus brevipes (Barsbold 1981), and its presence is uncertain in Harpymimus okladnikovi (Kobayashi & Barsbold 2005a). The shaft of metatarsal II is relatively straight in its proximal half, with a Dshaped cross-section, the lateral face being convex and the medial face flat and smooth (this flat area may have served for the contact with a first metatarsal, of which no remains have been found). More distally, the condition becomes reversed, with a flat lateral face (to accommodate the flat surface of metatarsal III) and a convex medial face. In its distal third, the shaft curves medially, so that the distal articular end is directed somewhat laterally rather than distally. The lateral condyle is much larger and more bulbous than the medial one, which is very narrow mediolaterally. The condyles

Fig. 4. Kinnareemimus khonkaenensis gen. et sp. nov. Left metatarsals II PW5A-101 (a –c) and PW5A-105 (d– f) in lateral (a, b), cranial (c, d) and medial (e, f) views. Scale bar: 50 mm.

234

E. BUFFETAUT ET AL.

Fig. 5. Kinnareemimus khonkaenensis gen. et sp. nov. metatarsals III. Proximal end of metatarsal III PW5A-104 in (a) cranial and (b) medial views. Middle part of shaft of right metatarsal III PW5A-103 in (c) cranial and (d) medial views. Distal end of left metatarsal III PW5A-107 in cranial (e) and caudal (f) views. Distal end of left metatarsal III PW5A-100 in cranial (g) and caudal (h) views. Scale bar: 50 mm.

are separated by a deep caudal groove. There are deep fossae both medially and laterally. Metatarsal III (Fig. 5) is represented by two proximal ends, one from the left side (PW5A-104) and one from the right (PW5A-131), the middle part of a shaft from the right side (PW5A-103), a distal segment from the left side with the articular end and a part of the shaft (PW5A-100, the holotype of Kinnareemimus khonkaenensis), and a left distal articular end (PW5A-107). On the basis of these various elements, which complement one another, a fairly accurate reconstruction of metatarsal III can be obtained. In proximal view, the proximal articular surface is much compressed mediolaterally and roughly triangular in outline, with the apex

located cranially. Both the lateral and medial faces are flat. Unlike the condition in advanced ornithomimids, such as Ornithomimus velox (Marsh 1890, 1896), Struthiomimus altus (Lambe 1902; Osborn 1916), Gallimimus bullatus (Osmo´lska et al. 1972), Anserimimus planinychus (Barsbold 1988) or Sinornithomimus dongi (Kobayashi & Lu¨ 2003), and even Archaeornithomimus asiaticus (Smith & Galton 1990), the craniocaudal extent of the proximal end of metatarsal III in Kinnareemimus khonkaenensis was such that it was still visible in cranial view, between the proximal ends of metatarsals II and IV, as in Garudimimus brevipes and Harpymimus okladnikovi. More distally, the shaft becomes extremely narrow both mediolaterally

EARLY OSTRICH DINOSAUR FROM THAILAND

and craniocaudally, forming a thin bony rod with a rectangular cross-section, a condition similar to that seen in Archaeornithomimus asiaticus and more advanced ornithomimids (the extreme slenderness of the bony rod explaining why specimens are often broken or incomplete at that level). In both Garudimimus brevipes and Harpymimus okladnikovi, on the other hand, the corresponding part of the shaft, although compressed mediolaterally, never becomes so thin and rod-like. Yet more distally, the shaft of metatarsal III in Kinnareemimus khonkaenensis becomes triangular in crosssection, with a flat cranial face and oblique lateral and medial faces, which meet at a sharp angle caudally. The bone simultaneously become much broader cranially, thus overlapping to some extent metatarsals II and III. This is the usual condition

235

in advanced ornithomimids (including Archaeornithomimus asiaticus), which differs from that seen in Garudimimus brevipes and Harpymimus okladnikovi, in which the cross-section of metatarsal III at this level is more D-shaped than triangular. The ‘medial expansion’ of the shaft (where it slightly overlaps metatarsal II: see Kobayashi & Barsbold 2005b, fig. 20) is apparently closer to the distal end in Kinnareemimus khonkaenensis than in Garudimimus brevipes and Harpymimus okladnikovi, and in this Kinnareemimus approximates more closely the arctometatarsalian condition seen in derived ornithomimids. In the more distal part, close to the articulation, the shaft of metatarsal III in Kinnareemimus khonkaenensis becomes D-shaped in cross-section and slightly narrower cranially. The articular head shows two subequal

Fig. 6. Kinnareemimus khonkaenensis gen. et sp. nov. Right metatarsal IV PW5A-106 in cranial (a), medial (b) and caudal (c) views. Left proximal parts of metatarsals PW5A-102 (d, e) and PW5A-108 (f, g) in medial (d, f) and lateral (e, g) views. Scale bar: 50 mm.

236

E. BUFFETAUT ET AL.

condyles separated by a shallow groove caudally. There are deep ligamentous fossae on both sides. Metatarsal IV (Fig. 6) is represented by a complete right specimen (PW5A-106, in which part of the distal half was poorly preserved and was reconstructed), a proximal fragment comprising the articular region and part of the shaft from the left side (PW5A-102), and two left proximal ends with small portions of the shaft (PW5A-108, 109). This bone is longer than metatarsal II and more slender. The proximal articular surface is somewhat teardrop-shaped in outline, with a rounded cranial margin and a pointed caudal end. The medial face of the bone, in its proximal part, is flat, to accommodate metatarsal III. This flat facet extends from the cranial to the caudal margin, showing that there was no cranial contact with metatarsal II and that the proximal part of metatarsal III was visible in cranial view. The cranial edge of the articular head overhangs the shaft. Caudally, there is a process that projects less markedly than the similar process on metatarsal II. A distinct concavity on the lateral face of this process probably corresponds to a contact area for metatarsal V (of which no remains have been found). The proximal half of

the shaft is straight, with a cross-section that is rounded cranially and angled caudally. More distally, the medial face of the shaft becomes flatter, where it could have accommodated the flat oblique lateral surface of metatarsal III. In its distal third, the shaft curves laterally, so that the distal articular region is oriented somewhat laterally, at an angle to the axis of the metatarsus. The distal articulation is ginglymoid, with a medial condyle that is more developed than the narrow lateral condyle. The condyles are separated caudally by a deep and narrow groove. There is a deep fossa on the medial side, and a much smaller one on the lateral side. Phalanges (Fig. 7). Eight more or less complete pedal phalanges are referable to Kinnareemimus khonkaenensis. They include a right proximal phalanx of digit II (PW5A-116) closely resembling a specimen referred by Gilmore (1920, fig. 73) to Ornithomimus affinis, two right second phalanges of digit III (PW5A-113, -119) also resembling a specimen illustrated by Gilmore (1920, fig. 74), a right proximal phalanx of digit III (PW5A-115), which is remarkably long and slender, two possible

Fig. 7. Kinnareemimus khonkaenensis gen. et sp. nov. Phalanges in dorsal (a, c, e, g, i, k, m, o) and lateral (b, d, f, h, j, l, n, p) views; (a, b) PW5A-116; (c, d) PW5A-115; (e, f) PW5A-120; (g, h) PW5A-119; (i, j) PW5A-118; (k, l) PW5A-123; (m, n) PW5A-121; (o, p) PW5A-122. Scale bar: 50 mm.

EARLY OSTRICH DINOSAUR FROM THAILAND

237

right second phalanges of digit III (PW5A-117), and a right third phalanx of digit IV (PW5A-121) similar to the corresponding element in Gallimimus bullatus (Osmo´lska et al. 1972). All these elements resemble the corresponding bones of other ornithomimosaurs but appear slightly more slender and elongate than the phalanges of Harpymimus okladnikovi and Garudimimus brevipes (Kobayashi & Barsbold 2005a, b), and even Gallimimus bullatus (Osmo´lska et al. 1972). A single ungual phalanx (PW5A-122), lacking it tip, is present. It shows a moderate curvature, has a groove on the medial side, lacks a flexor tubercle on the ventral face, and closely resembles the ungual phalanx of the third digit of Struthiomimus altus, as illustrated by Osborn (1916). The ventral surface is bounded medially by a ridge that ends posteriorly in a small spur (not visible on the damaged lateral margin). All these characters occur in other ornithomimids (Osmo´lska et al. 1972).

Relationships of Kinnareemimus khonkaenensis Although many elements of its skeleton remain unknown, Kinnareemimus khonkaenensis can be distinguished from other ornithomimosaurs by the structure of its metatarsus (Fig. 8). Ornithomimosaurs as a group exhibit various degrees of constriction of metatarsal III (Makovicky et al. 2004); but the so-called ‘arctometatarsalian’ condition (Holtz 1994, 2001) is not developed to the same extent in all members of the group. In basal forms in which the metatarsus is known (which excludes Pelecanimimus polyodon and Shenzhousaurus orientalis), metatarsal III is only moderately compressed mediolaterally and its proximal end is still visible between those of metatarsals II and IV in cranial view; it does not show a triangular cross-section in its distal half. Both Garudimimus brevipes and Harpymimus okladnikovi show this type of metatarsal III (the reconstruction of the tarsus of Garudimimus brevipes in Currie 2000, fig. 22.6 J, which shows metatarsal III hidden proximally by metatarsals II and IV, is incorrect: see Kobayashi & Barsbold 2005b), the former being slightly more advanced in the direction of the arctometatarsalian condition than the latter (Kobayashi & Barsbold 2005b). In contrast, more advanced forms (Ornithomimidae sensu Makovicky et al. 2004), including Archaeornithomimus asiaticus and more derived forms, have the proximal part of metatarsal III hidden from cranial view by metatarsals II and IV, which meet cranially in their proximal region; in addition, metatarsal III, which is very thin in its proximal half, becomes broader more distally, with a triangular cross-section, and spreads medially and

Fig. 8. Kinnareemimus khonkaenensis gen. et sp. nov. Reconstruction of the metatarsus in proximal (a) and cranial (b) views.

238

E. BUFFETAUT ET AL.

laterally over metatarsals II and IV. Kinnareemimus khonkaenensis shows an intermediate condition between that of the more basal forms and that of the Ornithomimidae: its metatarsal III is still visible, as a thin sliver of bone, between metatarsals II and IV in cranial view, but more distally becomes a thin rod of bone before it expands again, with a triangular cross-section, in its distal half. Although the metatarsal III of Kinnareemimus khonkaenensis still shows a relatively basal condition in its proximal end, it is extremely similar to the metatarsal III of advanced ornithomimids more distally. Nevertheless, a fully arctometatarsalian condition, with the proximal part of metatarsus III hidden from cranial view by metatarsals II and IV, is considered as one of the main synapomorphies of Ornithomimidae within Ornithomimosauria (Makovicky et al. 2004). From this point of view, Kinnareemimus khonkaenensis cannot be placed among the Ornithomimidae, but can be considered as the sister-group of that clade, occupying an intermediate position between Garudimimus brevipes and Archaeornithomimus asiaticus, currently the most basal ornithomimid. Garudimimus brevipes and Harpymimus okladnikovi have been placed in families of their own (Garudimimidae Barsbold 1981 and Harpymimidae Barsbold & Perle 1984), which have been accepted by various workers (Barsbold & Osmo´lska 1990; Kobayashi & Barsbold 2005a, b), although they are not recognized by Makovicky et al. 2004). Because its metatarsus is more advanced than that of both Garudimimus brevipes and Harpymimus okladnikovi, there seems to be no reason for referring Kinnareemimus khonkaenensis to either of these families. Pending the discovery of more complete material, it seems preferable not to erect a new family for the Thai form. As noted above, the age of the Sao Khua Formation is not perfectly well constrained, but it clearly belongs to the Early Cretaceous and is older than Aptian. This makes Kinnareemimus khonkaenensis one of the geologically oldest known ornithomimosaurs. The purported Late Triassic ornithomimosaur Shuvosaurus inexpectatus from Texas (Chatterjee 1993), is not an ornithomimosaur according to Rauhut (1997, 2003); the recent description of Effigia okeeffeae, a Late Triassic suchian from New Mexico showing convergently acquired ornithomimosaur-like features, shows that Shuvosaurus is a member of a Late Triassic non-dinosaurian group of ornithomimosaur mimics (Nesbitt & Norell 2006; Nesbitt 2007). Elaphrosaurus bambergi, from the Late Jurassic of Tanzania (Janensch 1925), once considered as an ornithomimosaur (Nopcsa 1928; Russell 1972; Galton 1982) is now placed among the Ceratosauria (Rauhut 2003; Tykoski & Rowe 2004). Other reports of pre-Cretaceous ornithomimosaurs

include two phalanges from the Kimmeridgian of southern England (Brokenshire & Clarke 1993), which do resemble the corresponding elements of ornithomimosaurs, but are too fragmentary for a more accurate assessment. Among Early Cretaceous ornithomimosaurs, Kinnareemimus khonkaenensis is either coeval with or slightly older than the two most basal forms known from that group, Pelecanimimus polyodon from the Barremian of Spain (Pe´rez-Moreno et al. 1994) and Shenzhousaurus orientalis from the Barremian of NE China (Ji et al. 2003). Both taxa show various primitive characters, including the presence of teeth. Unfortunately, Kinnareemimus khonkaenensis cannot be compared with these forms, because of the lack of skeletal elements in common. In particular, the tarsus is unknown in both Pelecanimimus polyodon and Shenzhousaurus orientalis, and nothing is known of the jaws of Kinnareemimus khonkaenensis. It cannot be excluded that Pelecanimimus polyodon or Shenzhousaurus orientalis had a Kinnareemimus-like tarsus. However, in the only other known toothed ornithomimosaur, Garudimimus brevipes, metatarsal III is less derived than in Kinnareemimus khonkaenensis. It may be mentioned that Kinnareemimus khonkaenensis is older than the purported ornithomimosaur Timimus hermani from the late Aptian – early Albian Otway Group of Victoria, Australia (Rich & Vickers-Rich 1994). Whatever the exact systematic position of this taxon, it postdates the earliest well-established ornithomimosaurs from Asia (and Europe) and therefore cannot be used to support a Gondwanan origin of ornithomimosaurs. Both of the primitive ornithomimosaurs that can be compared with Kinnareemimus khonkaenensis are from Mongolia and probably of later geological age than the Thai form. Harpymimus okladnikovi is from the Shinekhudag Svita, which, although regarded as Hauterivian –Barremian (Shuvalov 2000) or Aptian –Albian (Currie 2000) by some workers, has been referred to the late Albian on the basis of magnetostratigraphy and palynology by Hicks et al. (1999), an age estimate confirmed by Nichols et al. (2006) on the basis of palynomorphs. Garudimimus brevipes is from the Bayn Shireh Svita, which is considered as Cenomanian – Turonian (Currie 2000) or Cenomanian –Santonian (Hicks et al. 1999; Shuvalov 2000). Kinnareemimus khonkaenensis is thus in all likelihood geologically older than Harpymimus okladnikovi and certainly older than Garudimimus brevipes. In the structure of its metatarsus (which is the only feature that allows significant comparisons), the Thai form is clearly more advanced than the geologically younger Mongolian taxa. This suggests more or less separate evolution of the Central Asian and SE Asian ornithomimosaurs during the Early

EARLY OSTRICH DINOSAUR FROM THAILAND

Cretaceous, with the Mongolian forms (or at least some of them) remaining more basal in foot structure than those from SE Asia. In this connection, it is worth mentioning that the dinosaur fauna from the Sao Khua Formation shows some peculiar features (such as the apparent absence of ornithischians) that may point to isolation of SE Asia at the time of its deposition (Buffetaut et al. 2006), although the geographical and/or environmental reasons for such an isolation remain obscure (the presence of mountains chains resulting from Mesozoic tectonic activity in Asia may be an explanation). The dinosaur assemblage from the Sao Khua Formation appears to be rather different from the more or less coeval assemblages of the ‘Jehol Biota’ of NE China, for instance, and resemblances with Mongolian faunas are also limited. It should be mentioned, however, that an ornithomimosaur pes reminiscent of Kinnareemimus khonkaenensis has been described from the Middle Grey Unit of the Xinminbao Group of the Mazongshan area, Gansu, China, by Shapiro et al. (2003); metarsals II and III of this form are generally similar to those of the Thai taxon and suggest that the metatarsus was close to the arctometatarsalian condition (however, the proximal end of metatarsal III is missing). According to Tang et al. (2001), the Middle Grey Unit of the Xinminbao Group may be Albian. The specimen from Mazongshan thus indicates that an ornithomimosaur more advanced than the roughly coeval Harpymimus okladnikovi was present in what is now NW China at the end of the Early Cretaceous. Although this distribution pattern may suggest that ornithomimosaurs with an advanced type of metatarsus appeared in SE Asia somewhat earlier than in other parts of the world, the fact that metatarsus structure is unknown in important Early Cretaceous taxa such as Pelecanimimus polyodon and Shenzhousaurus orientalis makes this assumption difficult to test on the basis of the available fossils. Archaeornithomimus asiaticus is currently the most basal ornithomimid, showing a fully arctometatarsalian condition. It is geologically much younger than Kinnareemimus khonkaenensis, coming as it does from the Iren Dabasu Formation of Inner Mongolia, which was once thought to be Cenomanian, but may in fact be as young as Senonian, possibly Campanian (Currie & Eberth 1993) or even early Maastrichtian (Van Itterbeeck et al. 2005). The question of when the arctometatarsalian condition was achieved in ornithomimosaurs currently remains unanswered because of insufficient evidence concerning that group during the early stages of the Late Cretaceous. According to Alifanov & Averianov (2006), the ornithomimosaur bones from the Santonian of Kansai, Tajikistan, do not differ from those of other late Cretaceous

239

ornithomimids; however, but this material does not include metatarsals. Similarly, the ornithomimosaur material reported from the Turonian of Uzbekistan (Nessov 1995; Averianov 2006) and the Cenomanian of Kyrgyzstan (Averianov 2006) provides no information about the condition of the metatarsus. However, the fact that Kinnareemimus khonkaenensis was very close to the arctometatarsalian condition suggests that it may have been attained much earlier (i.e. in the Early Cretaceous) than could be expected on the basis of the fossil record as it was known before the Thai form was discovered. This reasoning could be reversed, however: the fact that Kinnareemimus khonkaenensis antedates the less derived (as far as the metatarsus is concerned) Harpymimus okladnikovi and Garudimimus brevipes could be considered as being in accordance with Holtz’s opinion that these non-arctometatarsalian ornithimosaurs ‘represent reversal to a more primitive state’ (Holtz 2001, p. 113). In that case, the condition in Kinnareemimus khonkaenensis could be interpreted as an incipient stage in this reversal from a fully arctometatarsalian condition; although this is not how Holtz (2001) interpreted the (at the time still unnamed) Thai form on the basis of the preliminary description given by Buffetaut & Suteethorn (1998). However, all recent studies of ornithomimosaur phylogeny (Ji et al. 2003; Kobayashi & Lu¨ 2003; Makovicky et al. 2004; Kobayashi & Barsbold 2005a, b) place Harpymimus and Garudimimus in a basal position relative to the fully arctometatarsalian Ornithomimidae, except that of Kobayashi & Barsbold (2006, fig. 8B), in which Archaeornithomimus is more basal than Garudimimus. It therefore seems more likely that Kinnareemimus in fact occupies an intermediate position between these basal forms and the more derived ones, including Archaeornithomimus. An attempt has been made to assess the cursorial abilities of Kinnareemimus khonkaenensis by comparison with other ornithomimosaurs, using the length ratio of the tibia to metatarsal III, under the assumption that forms with a longer metatarsus were more advanced in this respect than those with a shorter metatarsus. The length of metatarsus III in Kinnareemimus khonkaenensis had to be estimated because no complete specimen is available. By articulating together metatarsal II PW5A-101 and incomplete metatarsal III PW5A-100, which are of similar robustness and may be from the same individual, a total length of 145 mm was estimated for PW5A-100. This was compared with the length of tibia PWA-1, which may be from the same individual, although that cannot be demonstrated. Because of these various assumptions, the tibia/ metatarsal III ratio thus obtained for Kinnareemimus khonkaenensis must be considered as only an

240

E. BUFFETAUT ET AL.

approximation. A comparison with the ratios obtained for some other ornithomimosaurs is shown below: length of metatarsal III/length of tibia: Kinnareemimus khonkaenensis (PW5A-100 and PW5A-101) 0.56 Garudimimus brevipes (GIN 100, data from Kobayashi & Barsbold 2005b) 0.59 Struthiomimus altus (AMNH 5339, data from Osborn 1916) 0.68 Gallimimus bullatus (GIN 100/10, data from Osmo´lska et al. 1972) 0.72 This comparison shows that Kinnareemimus khonkaenensis had a relatively short tarsus, shorter than that of ornithomimids, and even than that of a relatively primitive taxon such as Garudimimus brevipes, which in metatarsus structure is less advanced than in the Thai form. This may be explained by the greater geological age of Kinnareemimus khonkaenensis and suggests that metatarsus elongation and the acquisition of an arctometatarsalian condition were not directly linked and did not evolve at the same rate in all ornithomimosaur lineages. Russell (1972) noted that Archaeornithomimus asiaticus has relatively stouter metatarsals than more advanced ornithomimids, as shown by the ratio between the midshaft circumference of metatarsal IV and its length. From this point of view, Kinnareemimus khonkaenensis has remarkably slender metatarsals, as shown by the following comparison: circumference of metatarsal IV at midlength of shaft/Length of metatarsal IV: Archaeornithomimus asiaticus (AMNH 6565, after Russell 1972) 25.9% Struthiomimus altus (AMNH 5375, after Russell 1972) 20.6% Kinnareemimus khonkaenensis (PW5-106) 18.5% This shows an interesting contrast to the metatarsal III/tibia length ratios given above, as it suggests that in the slenderness of its metatarsals Kinnareemimus khonkaenensis was more derived than some Late Cretaceous ornithomimids (if slenderness of the metatarsus is considered as a derived feature). However, the slenderness of metatarsal IV in Kinnareemimus khonkaenensis may be linked to the juvenile character of most of the ornithomimosaur material from Phu Wiang.

Conclusions Kinnareemimus khonkaenensis is one of the geologically oldest well-attested representatives of the Ornithomimosauria. Its occurrence in SE Asia

may be significant from a biogeographical point of view, as an addition to the growing list of Early Cretaceous ornithomimosaurs from Asia, which includes Harpymimus okladnikovi from the Aptian –Albian of Mongolia, Shenzhousaurus orientalis from the Barremian of China, and a possible ornithomimosaur vertebra from the late Barremian to Aptian Sebayashi Formation of Japan (Hasegawa et al. 1999). However, ornithomimosaurs were clearly not restricted to Asia in the Early Cretaceous, as attested by the presence of Pelecanimimus polyodon in the Barremian of Spain and possible remains from the Jurassic of England. Their presence in North America in the Early Cretaceous is questionable. Very fragmentary remains from the Aptian Arundel Formation of Maryland were described as Ornithomimus affinis by Gilmore (1920), and placed in the genus Archaeornithomimus by Russell (1972), but Smith & Galton (1990, p. 264) concluded that most of the material could not be identified beyond theropod incertae sedis, and two incomplete metatarsals could only be referred to ‘a coelurosaurian grade’. It should be noted, however, that, as already stated by Gilmore (1920, 1921), the pedal phalanges described as Ornithomimus affinis are remarkably similar to those of ornithomimosaurs (including Kinnareemimus). Of special importance is a fragmentary metatarsal III (USNM 5684; Gilmore 1920, fig. 72; 1921, plate 113), which apparently becomes triangular in cross-section a short distance proximally to the distal head, suggesting a form at least as advanced as Kinnareemimus in this respect. However, the proximal part of the specimen is not preserved. Although some additional specimens have been reported (Weishampel & Young 1996), more material is clearly necessary to better assess the systematic position of this material, but it does seem to suggest that early ornithomimosaurs may have been present in North America in the Early Cretaceous. Makovicky et al. (2004) and Weishampel (2006) considered the material from the Arundel Formation as belonging to indeterminate ornithomimosaurs. These Early Cretaceous records of ornithomimosaurs from Europe and possibly North America, together with the enigmatic Timimus hermani from the Aptian –Albian of Australia, show that an Asian origin for ornithomimosaurs cannot be taken for granted. Nevertheless, the advanced features of the tarsus of Kinnareemimus khonkaenensis, indicative of a trend toward the condition seen in Late Cretaceous ornithomimids, may suggest that derived ornithomimosaurs evolved in Asia, although it should be admitted that there is no evidence about metatarsus structure in important non-Asian Early Cretaceous forms such as Pelecanimimus polyodon.

EARLY OSTRICH DINOSAUR FROM THAILAND

What the condition of the metatarsus in Kinnareemimus khonkaenensis does suggest is that the acquisition of the arctometatarsalian condition in ornithomimosaurs was not a simple linear process. Although it is one of the earliest well-attested ornithomimosaurs, Kinnareemimus khonkaenensis is more advanced in this respect than some geologically younger forms such as Harpymimus okladnikovi and Garudimimus brevipes. Our field work at Phu Wiang was funded by the Department of Mineral Resources (Bangkok) and the Centre National de la Recherche Scientifique (CNRS, Paris). Subsequent research on the fossil vertebrates from the Sao Khua Formation has been supported by the ECLIPSE Programme of CNRS, a Thailand Research Fund–CNRS joint project, and a ‘Partenariat Hubert Curien’ French–Thai project. We thank E. S. Gaffney (New York), A. Milner (London), and R. Barsbold (Ulaan Baatar) for access to specimens in their care. This paper is dedicated to the memory of the late Halszka Osmo´lska (Warsaw), a leading expert on ornithomimosaurs, who helped us in many ways in our study of the Thai material.

References A LIFANOV , V. R. & A VERIANOV , A. O. 2006. On the finding of ornithomimid dinosaurs (Saurischia, Ornithomimosauria) in the Upper Cretaceous Beds of Tajikistan. Paleontological Journal, 1, 103–108. A VERIANOV , A. O. 2006. On an ornithomimid dinosaur (Saurischia, Ornithomimosauria) from the Cenomanian of Fergana. Paleontological Journal, 3, 323– 327. B ARSBOLD , R. 1976. On the evolution and systematics of the late Mesozoic dinosaurs. The Joint Soviet– Mongolian Paleontological Expedition, Transactions, 3, 68– 75. B ARSBOLD , R. 1981. Toothless carnivorous dinosaurs of Mongolia. The Joint Soviet– Mongolian Paleontological Expedition, Transactions, 15, 28–39. B ARSBOLD , R. 1988. A new Late Cretaceous ornithomimid from the MPR. Paleontologicheskii Zhurnal, 1, 122–124. B ARSBOLD , R. & O SMO´ LSKA , H. 1990. Ornithomimosauria. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria. University of California Press, Berkeley, 225– 244. B ARSBOLD , R. & P ERLE , A. 1984. On the first new find of a primitive ornithomimosaur from the Cretaceous of the MPR. Paleontologicheskii Zhurnal, 2, 121– 123. B ROKENSHIRE , A. J. & C LARKE , J. B. 1993. Important recently collected dinosaurian remains from the Lower Kimmeridge Clay at Weymouth. Proceedings of the Dorset Natural History and Archaeological Society, 115, 177–178. B UFFETAUT , E. & S UTEETHORN , V. 1998. Early Cretaceous dinosaurs from Thailand and their bearing on the early evolution and biogeographical history of some groups of Cretaceous dinosaurs. New Mexico Museum of Natural History and Science Bulletin, 14, 205–210.

241

B UFFETAUT , E. & S UTEETHORN , V. 1999. The dinosaur fauna of the Sao Khua Formation of Thailand and the beginning of the Cretaceous radiation of dinosaurs in Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 150, 13–23. B UFFETAUT , E., S UTEETHORN , V., M ARTIN , V., T ONG , H., C HAIMANEE , Y. & T RIAMWICHANON , S. 1995. New dinosaur discoveries in Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 157– 161. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J., K OBAYASHI , Y., H UANG , D. & L EE , Y. N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19– 37. C APPETTA , H., B UFFETAUT , E. & S UTEETHORN , V. 1990. A new hybodont shark from the Lower Cretaceous of Thailand. Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Monatshefte, 11, 659– 666. C HATTERJEE , S. 1993. Shuvosaurus—a new theropod. National Geographic Research and Exploration, 9, 274– 285. C URRIE , P. J. 2000. Theropods from the Cretaceous of Mongolia. In: B ENTON , M. J., S HISHKIN , M. A., U NWIN , D. M. & K UROCHKIN , E. N. (eds) The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge, 434–455. C URRIE , P. J. & E BERTH , D. A. 1993. Palaeontology, sedimentology and palaeoecology of the Iren Dabasu Formation (Upper Cretaceous), Inner Mongolia, People’s Republic of China. Cretaceous Research, 14, 127 –144. G ALTON , P. M. 1982. Elaphrosaurus, an ornithomimid dinosaur from the Upper Jurassic of North America and Africa. Pala¨ontologische Zeitschrift, 56, 265–275. G ILMORE , C. W. 1920. Osteology of the carnivorous Dinosauria in the United States National Museum, with special reference to the genera Antrodemus (Allosaurus) and Ceratosaurus. United States National Museum Bulletin, 110, 1 –159. G ILMORE , C. W. 1921. The fauna of the Arundel Formation of Maryland. Proceedings of the US National Museum, 59, 581– 594. G ILMORE , C. W. 1933. On the dinosaurian fauna of the Iren Dabasu Formation. Bulletin of the American Museum of Natural History, 67, 23–78. H ASEGAWA , Y., M ANABE , M., K ASE , T., N AKAJIMA , S. & T AKAKUWA , Y. 1999. An ornithomimid vertebra from the Early Cretaceous Sebayashi Formation, Sanchu Terrane, Gunma Prefecture, Japan. Bulletin of Gunma Museum of Natural History, 3, 1–6. H ICKS , J. F., B RINKMAN , D. L., N ICHOLS , D. J. & W ATABE , M. 1999. Paleomagnetic and palynologic analyses of Albian to Santonian strata at Bayn Shireh, Burkhant, and Khuren Dukh, eastern Gobi Desert, Mongolia. Cretaceous Research, 20, 829–850. H OLTZ , T. R. 1994. The arctometatarsalian pes, an unusual structure of the metatarsus of Cretaceous Theropoda (Dinosauria: Saurischia). Journal of Vertebrate Paleontology, 14, 480– 519.

242

E. BUFFETAUT ET AL.

H OLTZ , T. R. 2001. Arctometatarsalia revisited: The problem of homoplasy in reconstructing theropod phylogeny. In: G AUTHIER , J. & G ALL , L. F. (eds) New Perspectives on the Origin and Early Evolution of Birds. Peabody Museum of Natural History, New Haven, CT, 99– 121. I NGAVAT , R., J ANVIER , P. & T AQUET , P. 1978. De´couverte en Thaı¨lande d’une portion de fe´mur de dinosaure sauropode (Saurischia, Reptilia). Comptes Rendus Sommaires de la Socie´te´ Ge´ologique de France, 3, 140– 141. J ANENSCH , W. 1925. Die Coelurosaurier und Theropoden der Tendagura-Schichten Deutsch-Ostafrikas. Palaeonto graphica, Supplement VII, 1. Reihe, 1, 1 –99. J I , Q., N ORELL , M. A., M AKOVICKY , P. J., G AO , K., J I , S. & Y UAN , C. 2003. An early ostrich dinosaur and implications for ornithomimosaur phylogeny. American Museum Novitates, 3420, 1–19. K OBAYASHI , Y. & B ARSBOLD , R. 2005a. Anatomy of Harpymimus okladnikovi Barsbold and Perle 1984 (Dinosauria: Theropoda) of Mongolia. In: C ARPENTER , K. (ed.) The Carnivorous Dinosaurs. Indiana University Press, Bloomington, 97– 126. K OBAYASHI , Y. & B ARSBOLD , R. 2005b. Reexamination of a primitive ornithomimosaur, Garudimimus brevipes Barsbold, 1981 (Dinosauria: Theropoda), from the Late Cretaceous of Mongolia. Canadian Journal of Earth Sciences, 42, 1501–1521. K OBAYASHI , Y. & B ARSBOLD , R. 2006. Ornithomimids from the Nemegt Formation of Mongolia. Journal of the Paleontological Society of Korea, 22, 195– 207. K OBAYASHI , Y. & L U¨ , J. 2003. A new ornithomimid dinosaur with gregarious habits from the Late Cretaceous of China. Acta Palaeontologica Polonica, 48, 235– 259. L AMBE , L. M. 1902. New genera and species from the Belly River Series (Mid-Cretaceous). Contributions to Canadian Palaeontology, 3, 23– 81. M AKOVICKY , P. J., K OBAYASHI , Y. & C URRIE , P. J. 2004. Ornithomimosauria. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria. 2nd edn. University of California Press, Berkeley, 137– 150. M ARSH , O. C. 1881. Principal characters of American Jurassic dinosaurs. Part V. American Journal of Science, 21, 417–423. M ARSH , O. C. 1890. Description of new dinosaurian reptiles. American Journal of Science, 39, 81–86. M ARSH , O. C. 1896. The dinosaurs of North America. US Geological Survey Annual Report, 16, 133– 244. M ARTIN , V. 1994. Baby sauropods from the Sao Khua Formation (Lower Cretaceous) in northeastern Thailand. Gaia, 10, 147–153. M ARTIN , V., B UFFETAUT , E. & S UTEETHORN , V. 1994. A new genus of sauropod dinosaur from the Sao Khua Formation (Late Jurassic or Early Cretaceous) of northeastern Thailand. Comptes Rendus de l’Acade´mie des Sciences, Se´rie II, 319, 1085–1092. M ARTIN , V., S UTEETHORN , V. & B UFFETAUT , E. 1999. Description of the type and referred material of Phuwiangosaurus sirindhornae Martin, Buffetaut & Suteethorn, 1994, a sauropod from the Lower Cretaceous of Thailand. Oryctos, 2, 39–91.

M OURET , C., H EGGEMANN , H., G OUADAIN , J. & K RISADASIMA , S. 1993. Geological history of the siliciclastic strata of the Khorat Group in the Phu Phan Range area, northeastern Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology, 1. Department of Geological Sciences, Chiang Mai University, 23–49. N ESBITT , S. J. 2007. The anatomy of Effigia okeeffeae (Archosauria, Suchia), theropod-like convergence, and the distribution of related taxa. Bulletin of the American Museum of Natural History, 302, 1 –84. N ESBITT , S. J. & N ORELL , M. A. 2006. Extreme convergence in the body plans of an early suchian (Archosauria) and ornithomimid dinosaurs (Theropoda). Proceedings of the Royal Society of London, Series B, 273, 1045–1048. N ESSOV , L. A. 1995. Dinosaurs of Northern Eurasia: New data about assemblages, ecology and palaeobiogeography. University of Saint-Petersburg, Institute of Earth Crust. N ICHOLS , D. J., M ATSUKAWA , M. & I TO , M. 2006. Palynology and age of some Cretaceous nonmarine deposits in Mongolia and China. Cretaceous Research, 27, 241–251. N OPCSA , F. 1928. The genera of reptiles. Palaeobiologica, 1, 163–188. O SBORN , H. F. 1916. Skeletal adaptations of Ornitholestes, Struthiomimus and Tyrannosaurus. Bulletin of the American Museum of Natural History, 35, 733–771. O SMO´ LSKA , H., R ONIEWICZ , E. & B ARSBOLD , R. 1972. A new dinosaur, Gallimimus bullatus n.gen., n.sp. (Ornithomimidae) from the Upper Cretaceous of Mongolia. Palaeontologia Polonica, 27, 103– 143. O WEN , R. 1842. Report on British fossil reptiles. Part II. Report of the British Association for the Advancement of Science, 11, 60– 204. P E´ REZ -M ORENO , B. P., S ANZ , J. L., B USCALIONI , A. D., M ORATALLA , J. J., O RTEGA , F. & R ASSKIN G UTMAN , D. 1994. A unique multitoothed ornithomimosaur dinosaur from the Lower Cretaceous of Spain. Nature, 370, 363– 367. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. Part 1: Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5 –40. R AUHUT , O. W. M. 1997. Zur Scha¨delanatomie von Shuvosaurus inexpectatus (Dinosauria, Theropoda). Terra Nostra, 7, 17–21. R AUHUT , O. W. M. 2003. The interrelationships and evolution of basal theropod dinosaurs. Special Papers in Palaeontology, 69, 1 –213. R ICH , T. H. & V ICKERS -R ICH , P. 1994. Neoceratopsians and ornithomimosaurs: Dinosaurs of Gondwana origin? National Geographic Research and Exploration, 10, 129–131. R USSELL , D. A. 1972. Ostrich dinosaurs from the Late Cretaceous of western Canada. Canadian Journal of Earth Sciences, 9, 375– 402. S HAPIRO , M. D., Y OU , H., S HUBIN , N. H., L UO , Z. & D OWNS , J. P. 2003. A large ornithomimid pes from

EARLY OSTRICH DINOSAUR FROM THAILAND the Lower Cretaceous of the Mazongshan area, northern Gansu Province, People’s Republic of China. Journal of Vertebrate Paleontology, 23, 695– 698. S HUVALOV , V. F. 2000. The Cretaceous stratigraphy and palaeobiogeography of Mongolia. In: B ENTON , M. J., S HISHKIN , M. A., U NWIN , D. M. & K UROCHKIN , E. N. (eds) The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge, 256– 278. S MITH , D. & G ALTON , P. 1990. Osteology of Archaeornithomimus asiaticus (Upper Cretaceous, Iren Dabasu Formation, People’s Republic of China). Journal of Vertebrate Paleontology, 10, 255–265. S ULLIVAN , R. M. 1997. A juvenile Ornithomimus antiquus (Dinosauria: Theropoda: Ornithomimosauria), from the Upper Cretaceous Kirtland Formation (De-Na-Zin Member), San Juan Basin, New Mexico. In: A NDERSON , O., K UES , B. & L UCAS , S. (eds) Mesozoic Geology and Paleontology of the Four Corners Area. New Mexico Geological Society Guidebook, 48th field conference, New Mexico Geological Society, Socorro, 249–254. T ANG , F., L UO , Z., Z HOU , Z., Y OU , H., G EORGI , J. A., T ANG , Z. & W ANG , X. 2001. Biostratigraphy and palaeoenvironment of the dinosaur-bearing sediments

243

in Lower Cretaceous of Mazongshan area, Gansu Province, China. Cretaceous Research, 22, 115 –129. T YKOSKI , R. S. & R OWE , T. 2004. Ceratosauria. In: W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. (eds) The Dinosauria. 2nd edn. University of California Press, Berkeley, 47– 70. V AN I TTERBEECK , J., H ORNE , D. J., B ULTYNCK , P. & V ANDENBERGHE , N. 2005. Stratigraphy and palaeoenvironment of the Upper Cretaceous dinosaurbearing Iren Dabasu Formation (Inner Mongolia, People’s Republic of China). Cretaceous Research, 26, 699 –725. ¨ rliche System der SaurVON H UENE , F. 1914. Das natu ischia. Centralblatt fu¨r Mineralogie, Geologie und Pala¨ontologie, B, 1914, 154–158. W EISHAMPEL , D. B. 2006. Another look at the dinosaurs of the East Coast of North America. In: T ORCIDA F ERNA´ NDEZ -B ALDOR , F. & H UERTA H URTADO , P. (eds) Actas de las III Jornadas Internacionales sobre Paleontologia de Dinosaurios y su Entorno. Colectivo Arqueolo´gico– Paleontolo´gico de Salas, Salas de Los Infantes, 129– 168. W EISHAMPEL , D. B. & Y OUNG , L. 1996. Dinosaurs of the East Coast. Johns Hopkins University Press, Baltimore, MD.

Mesozoic vertebrate footprints of Thailand and Laos JEAN LE LOEUFF1*, TIDA SAENYAMOON2, CHRISTEL SOUILLAT1, VARAVUDH SUTEETHORN2 & ERIC BUFFETAUT3 1

Muse´e des Dinosaures, 11260 Espe´raza, France

2

Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand

3

CNRS, UMR 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure, 24 rue Lhomond, 75231 Paris Cedex 05, France *Corresponding author (e-mail: [email protected]) Abstract: Vertebrate footprints have been discovered in recent years from seven Mesozoic formations of Thailand and Laos dating from the Late Triassic (Kuchinari Group) and the Early Cretaceous (Khorat Group). The sites are reviewed here in chronological order. The ichnological record reflects fairly well the broad picture of the evolution of continental vertebrates in Asia known from the skeletal record. Norian basal archosaurs are replaced by Rhaetian dinosaurs although both footprint morphotypes look different from the contemporaneous European and North American forms. Two successive ornithopod radiations can be observed in the Early Cretaceous, with primitive small tetradactyl Hypsilophodon-like dinosaurs in the Earliest Cretaceous followed by advanced iguanodontoids with tridactyl fleshy footprints in the Aptian. Late Early Cretaceous dinosaur footprints from NE Thailand, however, do not validate previous hypotheses on the geographical distribution of Cretaceous ornithopod tracks in Asia. The ichnological record also reveals a hitherto unsuspected high diversity of theropods in the early Cretaceous with many different morphotypes.

The first fossil vertebrate footprints from SE Asia were reported by Buffetaut et al. (1985a), who described an assemblage of theropod footprints from the Phu Phan Formation in Phu Luang Wild Life Sanctuary (Loei Province, NE Thailand: see below). The same workers (Buffetaut et al. 1985b) drew attention to the fact that a much earlier albeit doubtful mention of fossil footprints from Thailand was made by the French explorer Mouhot (1863, 1864, 1868), who travelled in Siam between 1858 and 1861 and mentioned footprints of ‘antediluvian’ animals seen by him at Phrabat (or Phra Phuttabat, Saraburi Province) in central Thailand. Le Loeuff et al. (2006) have established that the first mention of these footprints was made by Pallegoix (1854), a French priest of the Missions Etrange`res de Paris, Titular Bishop of Mallus, who was the Vicar Apostolic of Siam between 1841 and 1862. Pallegoix, who was aware of Hitchcock’s discovery of fossilized ‘bird’ footprints in Connecticut (Hitchcock 1836), claimed that he had discovered many animal footprints (tigers, elephants, birds, deers) in the stones around the famous temple of Phrabat, dedicated to a Buddha footprint, while visiting it in March 1849. Mouhot went to Phrabat in November 1858 and saw the same footprints, reaching the conclusion that they ‘had been made by antediluvian animals’. He noted that according to a local legend, ‘all these beings belonged to Buddha’s cortege

when he went on the mountain’ (Mouhot 1863, 1864, 1868). The reports by Pallegoix and Mouhot have been mentioned by a few researchers, such as the French geographer Reclus (1883), who remarked (p. 811) that these possible footprints had yet to be examined by geologists. Reclus’ wish was realized only in 2005, when the senior author of the present paper examined the outcrops around the Phrabat temple. It appeared that both Pallegoix’s and Mouhot’s mentions refer to embedded corals and chert nodules in marine middle Permian limestones of the Khao Khad Formation mistakenly considered by them as fossil footprints (Le Loeuff et al. 2006). Thus the oldest mentions of fossil footprints in Thailand do not refer to genuine footprints but to erosional forms in Permian limestones. Since 1985, however, several new sites have been recognized and investigated in different continental formations of NE Thailand (Buffetaut & Suteethorn 1993; Polahan & Daorerk 1993; Buffetaut et al. 1997) and Laos (Allain et al. 1997). Most Thai footprints localities are located in NE Thailand (Fig. 1), on the Khorat Plateau, a 300 km wide structure consisting of non-marine Mesozoic deposits. Vertebrate footprints are now known from six Mesozoic continental formations of NE Thailand (namely, the Huai Hin Lat, Nam Phong, Phra Wihan, Sao Khua, Phu Phan and Khok Kruat

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 245–254. DOI: 10.1144/SP315.17 0305-8719/09/$15.00 # The Geological Society of London 2009.

246

J. LE LOEUFF ET AL.

Fig. 1. Location map of the localities. 1, Tad Huai Nam Yai (Huai Hin Lat Fm); 2, Tha Song Khon (Nam Phong Fm); 3, Hin Lat Pa Chad; 4, Phu Faek; 5, Khao Yai; 6, Phu Kao; 7, Phu Hin Rong Kla (Phra Wihan Fm); 8, Nong Sung (Sao Khua Fm); 9, Phu Luang (Phu Phan Fm); 10, Huai Dam Chum (Khok Kruat Fm); 11, Muong Phalane (Gre`s supe´rieurs Fm). Scale bar ¼ 100 km.

Formations) as well as from the Gre`s supe´rieurs Formation of Laos, which is considered as an equivalent of the Thai Khok Kruat Formation. These sites are dated from the Norian (Huai Hin Lat Formation) to the Late Early Cretaceous (Khok Kruat Formation). Some of these formations have not yielded skeletal material and the study of these ichnocoenoses is the only way to approach the biodiversity of the corresponding time intervals. Since 1999 new studies of the footprint localities were undertaken and many new sites have been discovered (Le Loeuff et al. 2002, 2003, 2005, 2006, 2007; Lockley et al. 2006, 2009; Matsukawa et al. 2006), and the first comprehensive review of vertebrate footprint sites in SE Asia is presented here. As for skeletal remains, two periods of the Mesozoic are well illustrated in the SE Asian ichnological record: the Late Triassic, with the deposits of the Kuchinari Group (Huai Hin Lat and Nam Phong Formations), and the Early Cretaceous, with the continental rocks of the Khorat Group (Phra Wihan, Sao Khua, Phu Phan and Khok Kruat Formations). There are thus so far no Jurassic vertebrate footprints in SE Asia.

Huai Hin Lat Formation The age of the Huai Hin Lat Formation is well constrained by data from palynomorphs and the vertebrate fossils. Racey et al. (1996) suggested a Carnian to Norian age on the basis of palynomorphs (see also Racey & Goodall 2009). According to Buffetaut & Suteethorn (1998), the vertebrates closely resemble those from the Norian of Germany (Stubensandstein), with common genera of fishes, amphibians, turtles and phytosaurs. Fossil footprints have been known from a long time in a remote place called Tad Huai Nam Yai, near Nam Nao in Phetchabun Province, on the west side of the Khorat Plateau. Three long vertebrate trackways are exposed on a slab of sandstone affected by a steep slope, which prevented us from mapping the entire site. The lower part of this slab was mapped in October 2003 (Fig. 2). The pes are very elongated with a large metapodial impression representing half the entire length. There are three elongate digits (digits III and IV are the longest; sometimes digit II is as long as them) and a lateral expansion suggesting a reduced digit V. A few footprints

MESOZOIC VERTEBRATE FOOTPRINTS

247

show what might be a reduced digit I. The pedal digits show a slight curvature with an internal convexity. The trackway is remarkably wide, with a width of 31 cm between the inner margins of the pes tracks and a width of 62 cm (i.e. four times footprint width) between the outer margins of the pes tracks. The pes tracks show no rotation at all. There is an alternation of long and short paces, indicating an asymmetrical gait. Another remarkable feature of the Nam Nao trackways is that they do not show any tail drag mark, suggesting that the trackmakers held their tails off the ground. A detailed description of this material will be published elsewhere. The trackways, however, show a mosaic of characters unknown in other Triassic trackways, showing that the Nam Nao trackmakers were wide-gauge plantigrade animals (high pace angulation, large track width, large metatarsal impression) with tetra- or pentadactyl mesaxonic pes and tetradactyl ectaxonic(?) manus, using an asymmetrical gait. The systematic affinities of the trackmakers are not yet clear but they could be some primitive archosauromorph reptiles, or more precisely phytosaurs, which are known from skeletal material in the Huai Hin Lat Formation. An undescribed partial skull kept in the collections of the Sirindhorn Museum in Sahatsakhan (Kalasin Province) belongs to a 4 –5 m long longirostrine phytosaur related to Mystriosuchus. This estimate is in agreement with the inferred size of the Nam Nao trackmaker (gleno-acetabular length estimated at 120 cm).

Nam Phong Formation

Fig. 2. Archosaur trackway in Tad Huai Nam Yai (Huai Hin Lat Formation, Norian). Scale bar: 50 cm.

A new dinosaur footprint site was found in 2007 at Tha Song Khon close to the town of Phu Kradung (Province of Loei) on the bank of the Nam Phong River (Mae Nam Phong). The fossil site belongs to the Nam Phong Formation, a late Norian to Rhaetian unit that overlies the Huai Hin Lat Formation and the upper limit of which might even be in the earliest Jurassic (Racey & Goodall 2009). A preliminary account of the discovery was given by Le Loeuff et al. (2007). The footprint site is a slab of red argillites covered by mudcracks and ripple marks. Footprints occur as concave hyporeliefs. The site shows the trackway of a large tetradactyl biped (footprint length 50 cm) with a stride length of 270 cm (Fig. 3). Six footprints are preserved. In some footprints one-third of the length is made by the metatarsal impression. Digit III is the longest, and digits II and IV are subequal in length. The tracks are remarkable in that they show a welldeveloped hallux impression, which is directed medially 90–1008 from the middle toe axe. The length of digit I varies from 10 to 15 cm. These tracks resemble Gigandipus from the Early Jurassic

248

J. LE LOEUFF ET AL.

of Connecticut (Lull 1953), which seems to be a semiplantigrade footprint of large crouching theropod, a Eubrontes trackmaker in his ‘regular’ digitigrade gait (Gierlinski et al. 2005). However, the Thai specimen shows more developed digit I than Gigandipus. We interpret the trackmaker as a large theropod dinosaur (hip height 240 cm) with digit I oriented medially or posteromedially to digit II. It can be noted, however, that the footprints are also surrounded by a large rim, suggesting that the trackmaker progressed in a soft muddy substrate, which made the preservation of both metatarsal impressions and hallux easier. Skeletal remains from the Nam Phong Formation include so far the large sauropod Isanosaurus attavipachi Buffetaut et al. 2000 and a prosauropod (Buffetaut et al. 1995). Thus this new site is the first evidence of the presence of a large theropod dinosaur (its length can be estimated at around 6 m) in the latest Triassic of SE Asia.

Phra Wihan Formation The Phra Wihan Formation overlies the Late Jurassic to Early Cretaceous Phu Kradung Formation (which itself is separated by a hiatus from the underlying Nam Phong Formation) and is overlain by the Sao Khua Formation, which is also referred to the Early Cretaceous. The Phra Wihan deposits consist of sandstones and siltstones with claystone intercalations. Racey et al. (1994) described a rich palynological assemblage from the Phra Wihan Formation and suggested a Berriasian to Early Barremian age for this unit, an attribution confirmed by Racey & Goodall (2009). This formation is the main track-bearing unit of Thailand with five footprint sites: Hin Lat Pa Chad (Khon Kaen Province), Phu Faek (Kalasin Province), Khao Yai National Park (Prachin Buri Province), Phu Kao (Nong Bua Lam Phu Province) and Phu Hin Rong Kla National Park (Loei Province).

Hin Lat Pa Chad (Khon Kaen Province)

Fig. 3. Theropod footprints in Tha Song Khon (Nam Phong Formation, Rhaetian). Scale bar: 50 cm.

Hin Lat Pa Chad is located in a river bed in the hills of Phu Wiang, a few kilometres from the rich dinosaur bones localities of Phu Wiang. Footprints occur as concave hyporeliefs on the upper surface of a sandstone layer; several trackways can be recognized. At least one trackway was left by a small theropod. Others were made by small quadrupedal ornithopods described by Le Loeuff et al. (2002). Lockley et al. (2009) referred this peculiar track to the new ichnogenus Neoanomoepus (Early Cretaceous of Canada), suggesting that these footprints might reflect a world-wide radiation of small ornithopods.

MESOZOIC VERTEBRATE FOOTPRINTS

Phu Faek (Kalasin Province) The footprint site at Phu Faek was discovered in 1996 by two schoolgirls in a dry river bed. It has been described by Buffetaut et al. (1997) and Le Loeuff et al. (2002), who reported several theropod trackways as well as two isolated sauropod footprints. There are at least two kinds of theropods represented at Phu Faek, a large one with 30–35 cm long footprints and several smaller individuals. Matsukawa et al. (2006) have challenged the sauropod interpretation of the large isolated tracks. We have recently inspected again the Phu Faek site and have still no better explanation for these two large footprints than the possibility that they are sauropod footprints.

Khao Yai (Prachin Buri Province) The footprints of the Khao Yai National Park were reported by Polahan & Daorerk (1993). They were redescribed by Lockley et al. (2002, 2006) on the basis of a cast kept at Chulalongkorn University in Bangkok. We have not seen this outcrop (an isolated slab of reddish brown sandstone on the bank of the river Sai Yai, approximately on the boundary between Kabin Buri and Prachantakham districts, in the northeastern part of Khao Yai National Park) but an excellent cast is available for study at the Museum of Petrified Wood and Mineral Resources in Nakhon Ratchasima. Sedimentological data suggest that these footprints belong to the Phra Wihan Formation rather than to the Sao Khua or Phu Phan Formation as originally reported. The footprints are preserved as convex epireliefs. Eleven complete or partial tracks are exposed. Lockley et al. (2006) have given a precise

249

description of this material dubbed Siamopodus khaoyaiensis. Unfortunately, their description is based on a very poor cast preserved at the Geological Museum of Chulalongkorn University, designated by Lockley et al. (2006) as the holotype, on which most of the characters of the tracks cannot be recognized (see Fig. 4). Thus the definition of Siamopodus will have to be revised. Further field work of the banks of the river Sai Yai would in this respect be especially useful.

Phu Kao (Nong Bua Lam Phu Province) The Phu Kao locality is also situated in a seasonally dry river bed. We visited the site in February 2000 and could recognize five trackways and 25 footprints. All footprints seem to have been made by rather small tridactyl dinosaurs (footprint length between 19 and 25 cm).

Phu Hin Rong Kla (Loei Province) One more footprint site was reported to us in 2006 by the rangers of the Phu Hin Rong Kla National Park in Loei Province. Nineteen large tridactyl footprints belonging to at least three trackways can be seen. From our preliminary observations two types of footprints occur at this site, some showing a wider divarication than others. Footprints from the Phra Wihan Formation thus include several types of theropod footprints, as well as primitive ornithopod and possible sauropod footprints. The most remarkable are small footprints of a slender quadrupedal ornithopod described from Hin Lat Pa Chad (Le Loeuff et al. 2002; Lockley et al. 2009). As we have pointed out (Le Loeuff et al. 2002), the presence of ornithopod footprints

Fig. 4. Siamopodus khaoyaiensis drawn from a cast kept at the Museum of Petrified Wood and Mineral Resources (a, b) and from a cast kept at Chulalongkorn University, Bangkok (c, redrawn from Lockley et al. 2006). Scale bar: 10 cm.

250

J. LE LOEUFF ET AL.

in the Phra Wihan Formation suggests that its fauna is closer to the dinosaurs of the overlying Phu Kradung Formation (from which a small ornithopod femur was reported by Buffetaut et al. 2003, 2006) than to those of the overlying Sao Khua Formation, which has never yielded a single ornithopod bone, although it is the richest bone-bearing formation of SE Asia.

Sao Khua Formation The Sao Khua Formation is by far the richest geological formation in Thailand in terms of number of dinosaur bones found (Buffetaut et al. 2009). This unit, however, had so far never yielded vertebrate footprints. The first discovery of fossil footprints in the Sao Khua Formation is due to L. Cavin during field work in July 2007 in the Province of Mukdahan. While Cavin and geologists of the Department of Mineral Resources were examining displaced slabs of rock in a road cut south of Nong Sung, they observed several tridactyl footprints. We visited the site in December 2007 and could indeed recognize many isolated footprints on several blocks of sandstone. The footprints are preserved as convex hyporeliefs (i.e. natural casts) on highly bioturbated surfaces. All footprints are tridactyl, ranging in length from 6 cm to 34 cm. In all likelihood they can be referred to several kinds of theropod dinosaurs.

Phu Phan Formation The Cretaceous Phu Phan Formation consists of conglomeratic sandstones; it is underlain by the Early Cretaceous Sao Khua Formation and overlain by the Aptian –Albian Khok Kruat Formation. It appears to be younger than early Barremian according to Racey & Goodall (2009). A single footprint site is known in Phu Luang Wildlife Sanctuary, a flat-topped mountain SW of the city of Loei (Loei Province, NE Thailand). This site was described by Buffetaut et al. (1985a, b). Fifteen footprints on a sandstone slab were referred to large theropods (footprint length 35 cm).

Khok Kruat and Gre`s supe´rieurs Formations The Khok Kruat Formation is considered as Aptian– Albian in age on the basis of its vertebrate fauna (Buffetaut et al. 2009) and palynology (Racey & Goodall 2009); it is considered as a lateral equivalent of the upper part of the Gre`s supe´rieurs Formation of Laos. Two footprint sites have been reported from these units: Huai Dam Chum in NE Thailand

Fig. 5. General map of the Huai Dam Chum tracksite (Nakhon Phanom Province). Scale bar: 1 m.

MESOZOIC VERTEBRATE FOOTPRINTS

(Nakhon Phanom Province) and Muong Phalane in Laos (Savannakhet Province).

Huai Dam Chum (Nakhon Phanom Province) The Huai Dam Chum locality is situated in the province of Nakhon Phanom, on the road between Nakhon Phanom and Bang Pang, NW of the confluence of the Mae Nam Songkhram and Mekong rivers, about 18 km from the small town of Than Uthen. It was discovered by N. Sattayarak from the Department of Mineral Resources in a sandstone

Fig. 6. Theropod footprints from Huai Dam Chum (Khok Kruat Formation). Scale bar: 50 cm.

Fig. 7. Ornithopod trackway (cf. Caririchnium) from Huai Dam Chum (Khok Kruat Formation). Scale bar: 50 cm.

251

252

J. LE LOEUFF ET AL.

Fig. 8. A tridactyl ornithopod track from Muong Phalane (Gre`s supe´rieurs Formation, Laos). Scale bar: 25 cm.

quarry where large blocks are extracted to consolidate the banks of the Mekong river. In February 2003, we were able to observe several large displaced slabs in the quarry showing small theropod footprints, but we could not locate the track layers in situ (Le Loeuff et al. 2003). A few months later, an in situ slab was located, and it was cleaned in February 2004. It revealed an impressive assemblage of more than 200 footprints (Fig. 5) of small theropods, ornithopods and crocodiles. Most of the footprints are theropod footprints, which were suggested by Matsukawa et al. (2006) to be similar to Asianopodus from China and Japan. Although the footprints show some similarities they are also consistently smaller than Asianopodus. It seems that the parallel theropod trackways (Fig. 6) were not left by a single group but by several small packs of theropods, as evidenced by the varying calculated speeds. Two successive footprints are didactyl and could represent either deinonychosaur footprints or some unknown small theropod (Le Loeuff et al. 2003); however, we cannot rule out the possibility that these two footprints are poorly preserved tridactyl footprints (see Milan 2006; Milan & Bromley 2006). A single ornithopod trackway of four footprints has been discovered on the main slab (Fig. 7); the footprints show three large rounded digits, very similar to footprints from Japan described by Matsukawa et al. (2006). These footprints were left by advanced ornithopods with fleshy digits and a digitigrade stance very distinct from the older Neoanomoepus tracks from the Phra Wihan Formation. The Japanese footprints were referred to the ichnogenus Caririchnium, to which we provisionally refer the footprints from Tha Uthen. This remarkable site was acquired in 2007 by the Department of Mineral Resources. Further

excavations and the building of a roof were realized in 2008.

Muong Phalane (Savannakhet Province, Laos) The tracks from Muong Phalane in Laos (Allain et al. 1997) reveal a diverse dinosaur assemblage including sauropods, large theropods and a large iguanodontid ornithopod. We inspected this outcrop in February 2005 and made some observations very different from those of Allain et al. (1997). According to us, only one sauropod trackway is present on the right bank of the river (Le Loeuff et al. 2005; Matsukawa et al. 2006). Matsukawa et al. (2006) have described this trackway as wide gauge. A second trackway was misinterpreted by Allain et al. (1997) as that of a sauropod moving towards the north but it was clearly left by a bipedal tridactyl animal, possibly an ornithopod moving towards the south. Evidence is presented here showing (Fig. 8) a tridactyl footprint erroneously interpreted as a sauropod manus print.

Conclusions Eleven Mesozoic localities with vertebrate footprints are now reported from SE Asia (10 from Thailand, one from Laos). Their study complements our knowledge of fossil vertebrates between the Late Triassic and the Late Early Cretaceous. Ichnological data are in good agreement with osteological data for Cretaceous localities, as they mirror and better constrain the timing of the main faunal changes. Not surprisingly, the Norian footprints from the Huai Hin Lat Formation were left by nondinosaurian archosaurs, whereas dinosaur footprints

MESOZOIC VERTEBRATE FOOTPRINTS

appear in the overlying Late Norian or Rhaetian Nam Phong Formation. It can be noted that these ichnites apparently represent new ichnotaxa, very different from the European or North American contemporaneous tracks. All remaining localities are early Cretaceous in age. It is worth noting that most of these footprint localities are dominated by theropods, which, of course, does not reflect the population patterns deduced from the study of dinosaur bones. The diversity of the theropod morphotypes, however, does suggest a higher taxonomic diversity than suggested by the skeletal record. The footprints from the Phra Wihan Formation seem to correspond well to the dinosaurs of the underlying Phu Kradung Formation, with small primitive ornithischians (Neoanomoepus) that are absent in the skeletal record from the younger Sao Khua Formation. The footprints from the Khok Kruat Formation reflect very well the sudden appearance of larger advanced ornithopods, which is also clear from the skeletal record and occurs after the deposition of the Sao Khua Formation. Finding new footprint sites in the Phu Phan Formation would help to better define the timing of their arrival. Their occurrence in SE Asia does not fit the pattern evoked by Matsukawa et al. (2006) of an ornithopod-rich assemblage limited to temperate latitudes (e.g. Japan, Korea, NE China and Mongolia), as opposed to a low-latitude saurischian-dominated assemblage. We suspect that the geographical pattern of Matsukawa et al. (2006) reflects instead temporal differences between the localities. Advanced ornithopod tracks seem to appear in the Late Early Cretaceous everywhere in Eastern Asia. The palaeobiogeographical causes of this global phenomenon have yet to be investigated in detail. The footprints from the Khok Kruat Formation also demonstrate the presence of several small theropods. Further field work in the Khorat plateau of Thailand, Savannakhet Province of Laos and neighbouring central Vietnam should reveal new dinosaur footprint sites in the near future. This work was funded by the Centre National de la Recherche Scientifique (Paris), the Thailand Research Fund, the Department of Mineral Resources (Bangkok), the French Ministry for Foreign Affairs and the Muse´e des Dinosaures (Espe´raza, France). We thank all participants of the Thai–French palaeontological expeditions, especially those who joined the ichnological team (S. Khansubha, K. Wongko, S. Suteethorn) and the geologists of the Department of Mineral Resources, who have discovered most of the footprint sites. Special thanks to L. Cavin (Geneva), G. Cuny (Copenhagen), H. Tong (Paris), J. Claude (Montpellier) and H. Fontaine (Clamart) for sharing information about the Mesozoic of Thailand. The senior author thanks the curators of the Museum of Petrified Wood and Mineral Resources

253

(Nakhon Ratchasima) and the Chulalongkorn University Geological Museum (Bangkok) for authorizing access to their ichnological collections, and J. Milan (Copenhagen) and G. Gierlinski (Warsaw) for their helpful comments.

References A LLAIN , R., T AQUET , P. ET AL . 1997. Pistes de dinosaures dans les niveaux du Cre´tace´ infe´rieur de Muong Phalane, Province de Savannakhet (Laos). Comptes Rendus de l’Acade´mie des Sciences, Se´rie II, 325, 815– 821. B UFFETAUT , E. & S UTEETHORN , V. 1993. The dinosaurs of Thailand. Journal of Southeast Asian Earth Sciences, 8, 77–82. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 83– 90. B UFFETAUT , E., I NGAVAT , R., S ATTAYARAK , N. & S UTEETHORN , V. 1985a. Les premie`res empreintes de pas de Dinosaures du Sud-Est asiatique: pistes de Carnosaures du Cre´tace´ infe´rieur de Thaı¨lande. Comptes Rendus de l’Acade´mie des Sciences, Se´rie II, 309, 643– 648. B UFFETAUT , E., I NGAVAT , R., S ATTAYARAK , N. & S UTEETHORN , V. 1985b. Early Cretaceous dinosaur footprints from Phu Luang (Loei Province, Northeastern Thailand) and their Significance. In: T HANVARACHORN , P., H OKJAREON , S. & Y OUNGME , W. (eds) Proceedings of the Conference on Geology and Mineral Resources Development of the Northeast Thailand. Khon Kaen University, Khon Kaen, 71–76. B UFFETAUT , E., M ARTIN , V., S ATTAYARAK , N. & S UTEETHORN , V. 1995. The oldest known dinosaur from Southeast Asia: A prosauropod from the Nam Phong Formation (Late Triassic) of Northeastern Thailand. Geological Magazine, 132, 739– 742. B UFFETAUT , E., S UTEETHORN , V., T ONG , H., C HAIMANEE , Y. & K HANSUBHA , S. 1997. New dinosaur discoveries in the Jurassic and Cretaceous of Thailand. In: D HEERADILOK , P., H INTHONG , C., C HAODUMRONG , P. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific. Department of Mineral Resources, Bangkok, 177– 187. B UFFETAUT , E., S UTEETHORN , V., C UNY , G., T ONG , H., L E L OEUFF , J., K HANSUBHA , S. & J ONGAUTCHARIYAKUL , S. 2000. The earliest known sauropod dinosaur. Nature, 407, 72–74. B UFFETAUT , E., S UTEETHORN , V., C UNY , G., K HANSUBHA , S., T ONG , H., L E L OEUFF , J. & C AVIN , L. 2003. Dinosaurs in Thailand. Maha Sarakham University Journal, Special Issue, 22, 69–82. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y. N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19– 37.

254

J. LE LOEUFF ET AL.

B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2009. An early ‘ostrich dinosaur’ (Theropoda: Ornithomimosauria) from the Early Cretaceous Sao Khua Formation of NE Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 225–239. G IERLINSKI , G., L OCKLEY , M. G. & M ILNER , A. 2005. Traces of Early Jurassic crouching dinosaurs. Tracking Dinosaur Origins. The Triassic/Jurassic Terrestrial Transition. Abstract Volume. Dixie State College, St. George, Utah, 4. H ITCHCOCK , E. 1836. Ornithichnology. Description of the foot mark of birds (ornitichnites) on New Red Sandstone in Massachusetts. American Journal of Science, 29, 307–340. L E L OEUFF , J., K HANSUBHA , S., B UFFETAUT , E., S UTEETHORN , V., T ONG , H. & S OUILLAT , C. 2002. Dinosaur footprints from the Phra Wihan Formation (Early Cretaceous of Thailand). Comptes Rendus Pale´vol, 1, 287– 292. L E L OEUFF , J., S UTEETHORN , V. ET AL . 2003. The first dinosaur footprints from the Khok Kruat Formation (Aptian of Northeastern Thailand). Mahasarakham University Journal, Special Issue, 22, 83– 92. L E L OEUFF , J., S AENYAMOON , T., S UTEETHORN , V., K HANSUBHA , S. & B UFFETAUT , E. 2005. Vertebrate footprints of Southeast Asia (Thailand and Laos): A review. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 582 –587. L E L OEUFF , J., S UTEETHORN , V. & B UFFETAUT , E. 2006. The oldest mentions of fossil vertebrate footprints in Thailand: A reassessment of Bishop Pallegoix and Henri Mouhot’s writings. Ichnos, 13, 81– 86. L E L OEUFF , J., S AENYAMOON , T., S UTEETHORN , V., S OUILLAT , C. & B UFFETAUT , E. 2007. Triassic trackways from Thailand. In: T ANTIWANIT , W. (ed.) Proceedings of the International Conference on Geology of Thailand: Towards Sustainable Development and Sufficiency Economy. Department of Mineral Resources, Bangkok, 362–363. L OCKLEY , M., S ATO , Y. & M ATSUKAWA , M. 2002. A new dinosaurian ichnogenus from the Cretaceous of Thailand. In: M ANTAJIT , N. (ed.) Proceedings of the Symposium on Geology of Thailand. Department of Mineral Resources, Bangkok, 117– 119. L OCKLEY , M., M ATSUKAWA , M., S ATO , Y., P OLAHAN , M. & D AORERK , V. 2006. A distinctive new theropod dinosaur track from the Cretaceous of Thailand: Implications for theropod track diversity. Cretaceous Research, 27, 139–145. L OCKLEY , M. G., M C C REA , R. T. & M ATSUKAWA , M. 2009. Ichnological evidence for small quadrupedal ornithischians from the basal Cretaceous of SE Asia and North America: Implications for a global radiation. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 251–265.

L ULL , R. S. 1953. Triassic Life of the Connecticut Valley. Revised edn. Bulletin of the Connecticut State Geological and Natural History Survey, 81. M ATSUKAWA , M., L OCKLEY , M. & L I , J. 2006. Cretaceous terrestrial biota of East Asia, with special reference to dinosaur-dominated ichnofaunas: Towards a synthesis. Cretaceous Research, 27, 3 –21. M ILAN , J. 2006. Variations in the morphology of Emu (Dromaius novaehollandiae) tracks reflecting differences in walking pattern and substrate consistency: Ichnotaxonomic implications. Palaeontology, 49, 405–420. M ILAN , J. & B ROMLEY , R. G. 2006. True tracks, undertracks and eroded tracks, experimental work with tetrapod tracks in laboratory and field. Palaeogeography, Palaeoclimatology, Palaeoecology, 231, 253–264. M OUHOT , H. 1863. Voyage dans les Royaumes de Siam, de Cambodge, de Laos et autres parties centrales de l’Indo-Chine. Le Tour du Monde, Nouveau Journal des Voyages, 8, 219– 353. M OUHOT , H. 1864. Travels in the central parts of IndoChina (Siam), Cambodia and Laos, during the years 1858, 1859 and 1860, Vols I and II. John Murray, London. M OUHOT , H. 1868. Voyage dans les Royaumes de Siam, de Cambodge, de Laos et autres parties centrales de l’Indo-Chine. Hachette, Paris. P ALLEGOIX , J. B. 1854. Description du Royaume Thai ou Siam comprenant la topographie, histoire naturelle, mœurs et coutumes, le´gislation, commerce, industrie, langue, litte´rature, religion, annales des Thai et pre´cis historique de la mission. Imprimerie Vialat, Lagny. P OLAHAN , M. & D AORERK , V. 1993. Report on additional discovery of dinosaur’s footprints in Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia. Chiang Mai University, Chiang Mai, 225 –230. R ACEY , A. & G OODALL , J. G. S. 2009. Palynology and stratigraphy of the Mesozoic Khorat Group red bed sequences from Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 67–81. R ACEY , A., G OODALL , J. G. S., L OVE , M. A., P OLACHAN , S. & J ONES , P. D. 1994. New age data for the Mesozoic Khorat Group of Northeast Thailand. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , W., W ONGSOMSAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245– 252. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group North Eastern Thailand: Part 1, Stratigraphy and Sedimentary Evolution. Journal of Petroleum Geology, 19, 5–40. R ECLUS , E. 1883. Nouvelle Ge´ographie Universelle. La Terre et les Hommes. VIII L’Inde et l’Indo-Chine. Hachette, Paris.

Ichnological evidence for small quadrupedal ornithischians from the basal Cretaceous of SE Asia and North America: implications for a global radiation MARTIN G. LOCKLEY1*, RICHARD T. MC CREA2 & MASAKI MATSUKAWA3 1

Dinosaur Tracks Museum, University of Colorado at Denver, Campus Box 172, PO Box 173364, Denver, CO 80217-3364, USA 2

Peace Region Palaeontology Research Centre, Box 1540, Tumbler Ridge, B.C. V0C 2W0, Canada

3

Department of Environmental Sciences, Tokyo Gakugei University, Koganei, Tokyo, 184-8501, Japan *Corresponding author (e-mail: [email protected]) Abstract: Tracks of small quadrupedal ornithischians with five manual and four pedal digits have been recorded from sedimentary rocks near the Late Jurassic–Early Cretaceous (Tithonian– Berriasian) boundary in NE Thailand and British Columbia. These are compared with larger tracks of gracile, quadrupedal ornithopods from the earliest Cretaceous of Spain and smaller tracks of a quadruped of unknown age from Zimbabwe. The Thai and Canadian tracks are similar to the Early Jurassic (Liassic) ichnogenus Anomoepus and the small ornithopod tracks from the Late Jurassic of Spain. They are the only known post-Liassic ornithischian tracks in which up to five discrete manus digit impressions are clearly visible. Based on strong heteropody (manus much smaller than pes) in all cases we infer an ornithopod trackmaker rather than another ornithischian. The scattered, but widespread earliest Cretaceous occurrence of this ichnotaxon, herein assigned to Neoanomoepus perigrinatus ichnogen. and ichnosp. nov., on the basis of type material from Canada, suggests that these hitherto unknown earliest Cretaceous ichnofaunas may represent a radiation of small basal ornithopods (pes length less than 15 cm), appearing before the widespread radiation of large ornithopods (pes length up to 60 cm or more) later in the Neocomian (Valanginian– Barremian), Aptian– Albian and Late Cretaceous. The primitive condition of the trackmaker is indicated by the pedal and manual morphology, which consists of four and five digits respectively that are not enclosed by well-developed fleshy padding or integument. In contrast, all larger Cretaceous ornithopod tracks, mostly from post-Berriasian strata, have only three pedal digits enclosed in fleshy pads and a manus in which all functional digits are reduced and enclosed by substantial flesh.

Trackways of small ornithopod dinosaurs are still comparatively rare. The most famous, and firstdescribed, example of a trackway with what appears to be unequivocal ornithopod or basal ornithischian characteristics (five-toed manus and four-toed pes) is the well-known Early Jurassic ichnogenus Anomoepus (Hitchcock 1848; Lull 1953; Olsen & Rainforth 2003). This ichnogenus (Fig. 1) is well known from the early Jurassic of the eastern USA and may be abundant in southern Africa (Ellenberger 1972) but was, until recently, relatively poorly known from other regions such as the western USA (Lockley & Hunt 1995; Lockley & Gierlinski 2006), Europe (Avanzini et al. 2001; Gierlinski 1991; Gierlinski et al. 2004; Lockley & Meyer 2000) and Australia (Thulborn 1994). Anomoepus is characterized, in well-preserved examples, by a four-toed pes, sometimes with metatarsal

impressions, and a five-toed manus. In many cases, however, Anomoepus is hard to identify with confidence unless both manus and pes tracks are found (Lockley & Gierlinski 2006). In many cases the trackmaker was progressing bipedally, and hallux (digit I of the pes) is inconspicuous and may not be impressed. Moreover, if not well preserved it may be difficult to distinguish Anomoepus tracks from a number of other tridactyl or tetradactyl dinosaur footprints. However, there may be some clues to the ornithischian affinity of incomplete Anomoepus or Anomoepus-like ichnites in the general configuration of the trackway. As indicated by Lockley (1999), ornithopods typically, indeed consistently, have trackways that are relatively wide in comparison with those of theropods and consistently take shorter steps, with shorter (wider) pes tracks that toe inward.

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 255–269. DOI: 10.1144/SP315.18 0305-8719/09/$15.00 # The Geological Society of London 2009.

256

M. G. LOCKLEY ET AL.

Fig. 1. Type Anomoepus from the Early Jurassic of New England (after Olsen & Rainforth 2003, fig 19.5). (a) An idealized composite of walking pes and sitting manus and pes with metatarsal impressions; (b, c) composite trackways showing quadrupedal and bipedal impressions, respectively. rm, right manus; rp, right pes; lm, left manus; lp, left pes; t, tail.

Despite this distinctive combination of characters there are few known examples of post-Early Jurassic tracks that have been assigned unequivocally to Anomoepus. Moreover, with the possible exception of the enigmatic, presumed heterodontosaurid track Delatorrichnus from the Middle Jurassic of Argentina (Casamiquela 1964), there is only one reported example of a Late Jurassic ornithopod track that reveals quadrupedal progression. This is a hitherto undescribed specimen from the Late Jurassic of Asturias (Garcı´a Ramos et al. 2006; Lockley et al. 2009). Dinehichnus is the only other formally named Late Jurassic track attributed to an

ornithopod (Lockley et al. 1998). It is not until the Cretaceous that we find abundant ichnological evidence of quadrupedal and bipedal ornithopods in many regions including Asia (You & Azuma 1995; Lockley & Matsukawa 1998). However, most of the larger quadrupedal forms had a hoof-like manus without clear separation of the digit traces. Other than the Cretaceous material described herein, the only other footprints of small quadrupedal ornithopods, with distinct digit impressions, hitherto named from the Early Cretaceous are Hypsiloichnus (Stanford et al. 2004). All other tracks of quadrupedal Cretaceous ornithopods

TRACKS OF SMALL ORNITHISCHIANS

indicate much larger trackmakers whose hand and foot morphology bears little or no resemblance to that of the Anomoepus trackmaker. Trackways of large ornithopod dinosaurs are fairly well known in the Cretaceous. Most, however, are large with three well-padded pes impressions indicating the derived condition typical of iguanodontids or hadrosaurs (e.g. Iguanodontipus, Sarjeant et al. 1998). Some indicate quadrupedal progression, but in such cases the manus is revealed to be a relatively small sub-circular to oval hoof-like impression that rarely reveals any differentiation of digit traces. Only recently have reports emerged of Cretaceous trackways that evidently represent small quadrupedal ornithopods that exhibit the primitive, Anomoepus-like, condition of four pedal and five manual digits (Fig. 2). These discoveries have been made in such widely divergent localities as Thailand (Buffetaut & Suteethorn 1993; Le Loeuff et al. 2002; Matsukawa et al. 2006), Zimbabwe (LinghamSoliar & Broderick 2000) and Canada (this study). Other distinctive ornithopod morphotypes have

257

been recorded from Spain (Perez-Lorente et al. 1997). The primary purpose of this paper therefore is to describe these Cretaceous examples in detail and discuss their ichnotaxonomic status. A secondary objective is, where possible, to place the tracks in their biostratigraphic, palaeobiogeographical and palaeoecological context and discuss the extent to which they may shed light on the timing of the Early Cretaceous ornithopod radiation.

Track material and geological context Thailand specimens Tracks that closely resemble the Canadian specimens are known from the Phra Wihan Formation of NE Thailand at a site known as Hin Lat Pa Chad (Buffetaut & Suteethorn 1993; Le Loeuff et al. 2002; Matsukawa et al. 2006). The tracks occur at a single isolated locality in a jungle region. The outcrop consists of a single bedding

Fig. 2. Locality map showing the type locality for Neoanomoepus periginatus ichnogen. et ichnosp. nov. in British Columbia, and the occurrence of other probable Neoanomoepus or Neoanomoepus-like ichnites from Thailand, Zimbabwe, Spain and Canada.

258

M. G. LOCKLEY ET AL.

Fig. 3. Tracks from the Hin Lat Pra Chad site in Thailand (after Matsukawa et al. 2006). (a) Preliminary map produced in local field guide with numbered trackway segments 1– 10. Trackway 1 may be a continuation of 4, and 6 may be a continuation of 7. (b) Detailed map of northern section showing trackways 1, 2, 4, 7, 8 and 10 (with detail of 10), current ripple vector, and source of track replicas (University of Colorado at Denver 214.54– 58) and Phu Kum Kao Dinosaur Museum (PKKDM) replica. (c) Composite of ornithischian manus (M)– pes (P) set from trackway 1, showing 4 pes and 5 manus digit impressions. (d) Detail of trackway 7 also showing 4 pes digit impressions. (e) Detail of trackway 4. (f) Detail of large track (trackway 8). All track detail scale bars 10 cm.

plane of about 25 m  10 m with 10 trackways mostly trending from north to south (Fig. 3). As noted by Buffetaut & Suteethorn (1993, p. 77), some trackways ‘sometimes show a small, apparently tridactyl manus impression lateral to the pes impression . . . [which] . . . together with the general shape of the footprints suggest that they may have been left by ornithischians’. We agree with this interpretation generally, but note that the pes tracks sometimes show faint impressions of the hallux and are therefore tetradactyl, and that at least one manus track is also tetradactyl, and may be interpreted as pentadactyl. Le Loeuff et al. (2002, p. 291) considered the manus tracks ‘reminiscent’ of Anomoepus intermedius and A. scambus, but they stopped short of applying these names to the illustrated material (Le Loeuff et al. 2002, fig. 4). According to Olsen & Rainforth (2003), A. intermedius is a synonym of A. scambus, which is the only recognizable ichnospecies in the ichnogenus Anomoepus. However, Lockley & Gierlinski (2006) disputed the claim that the ichnogenus is monospecific.

The Thai material was re-examined by Matsukawa et al. (2006), who illustrated two trackways (Fig. 3) with both manus and pes footprints demonstrating that in some cases the former are pentadactyl and the latter tetradactyl. In that paper, Matsukawa et al. (2006) reproduced a previously unpublished map of the site in which 10 trackways were numbered. Comparison of these two maps (Fig. 3a and b), which use the same numbering scheme, reveals that trackways 3, 5, 6 and 9 are not included because they were covered by sediment at the time they were studied by Matsukawa et al. (2006). Moreover trackways 3, 5 and 6 are probably continuations of other trackways such as 1, 4 or 7. Nevertheless, these trackways (1, 4 and 7) all have associated manus impressions (Fig. 3c–e), whereas other trackways (e.g. 2 and 3) indicate biped progression on a functionally tridactyl pes, probably attributable to ornithopods. The possibility that some tridactyl tracks are theropodan cannot be discounted. One large tetradactyl track, with curved digit impressions, may be attributable to a crocdodilian (Fig. 3f).

TRACKS OF SMALL ORNITHISCHIANS

The Phra Wihan Formation is considered as Neocomian in age, or may be Berriasian. The probable age for the Phra Wihan Formation given by Buffetaut & Suteethorn (1993) is Late Jurassic. However, in a later paper (Buffetaut et al. 1997), they gave a revised age of early Cretaceous (Berriasian –Barremian) based on palynomorphs (Racey et al. 1994, 1996). This age estimate, followed by Le Loeuff et al. (2002) and Matsukawa et al. (2006), is further supported by Racey & Goodall (2009).

Canadian specimens Well-preserved and diagnostic small ornithischian ichnites are herein reported from the coal-bearing Mist Mountain Formation of the Elk Valley Coal region in SE British Columbia, Canada. About two dozen clear tracks are preserved as natural casts on the underside of a sandstone slab of about 3 m2 (Figs 4–6). In total almost 100 tracks may be discerned by careful observation of the specimen, mould and replica under controlled lighting conditions. Several trackway segments are discernible, including at least three with three or more consecutive manus–pes sets (Fig. 5, Table 1). The precise stratigraphic location from which the tracks originate is not known. However, the approximate location, within the Fording River Coal mine, is known, and all the strata in the mine are Tithonian to Berriasian. The tracks are thought to originate from the Berriasian part of the section. The predominant lithology in this region is a coal-bearing sequence consisting of grey mudstones and siltstones with intercalated buff-coloured sandstone and coals (Bustin & Smith 1993). Tracks are

259

abundant and include a variety of tridactyl tracks attributable to bipedal theropods, bipedal ornithopods and possibly water- or shore-birds (McCrea & Buckley 2005a). A few isolated sauropod tracks and one trackway have also been recorded (McCrea et al. 2005). As many as 80 track-bearing specimens have been set aside as outdoor displays near the mine entrance and more than a dozen specimens have been replicated for scientific study (see reference to specimens below). Plant fossils are also abundant, and locally moulds of freshwater unionid or unionid-like clams can be found. The discovery of this important slab appears to confirm the report of Currie (1989, p. 294), who suggested that a trackway of a tridactyl biped, with inward pes rotation, from the Mist Mountain Formation, British Columbia, could be ‘identified as Anomoepus’. However, this trackway, in the Tyrrell Museum (TMP 85.105.1), reveals no diagnostic manus tracks. Nevertheless, Currie’s identification was prescient given the quadrupedal trackway with manus traces described herein. Institutional abbreviations for replicas used in the systematic descriptions (below) are FGM for Fraser –Fort George Regional Museum, British Columbia; and CU for the University of Colorado at Denver, Dinosaur Tracks Museum.

Trackways from Spain Although there are a number of reports of trackways of quadrupedal ornithopods from the Early Cretaceous of Spain, and other parts of Europe, North America and Asia, almost all examples other than those cited above refer to large tracks with well-rounded manus impressions in which

Fig. 4. Photograph of large Neoanomoepus track-bearing slab (field number A48) from Mist Mountain Formation, British Columbia. It should be noted that the tracks are preserved as natural casts. (Compare with Fig. 5.)

260

M. G. LOCKLEY ET AL.

Fig. 5. Map of large track-bearing slab (field number A48) from Mist Mountain Formation, British Columbia. (Compare with Fig. 4, but note reversal to show non-natural cast aspect.) Trackway A is holotype for Neoanomoepus (see Figs 6 and 8). Details of trackways A– C are shown in Figure 8. Replica of complete slab in Fraser– Fort George Regional Museum (FGM 002.01.20).

Fig. 6. Photograph of the manus–pes set R1 from trackway A, the holotype of Neoanomoepus (FGM 002.01.20 and CU 199.22) (compare with trackway A of Fig. 5).

TRACKS OF SMALL ORNITHISCHIANS

261

Table 1. Measurements for standard trackway parameters for three Neoanomoepus trackways from the Mist Mountain Formation Track number AR1 AL2 AR2 BR1 BL1 BR2 BL2 BR3? BL3? BR4? CR1 CL1 CR2 CL2 Mean

Pes L:W

Manus L:W

Step

Stride

Pace Ang.

(19.0):9.8 12.0:10.5 10.0:13.7 19.2:10.7 (14.2):11.7 (14.5):13.5 13.0:11.0 14.8:11.4 12.0:11.0 11.5:10.5 (16.0): – 13.9:10.5 16.0:9.8 16.1:11.4

4.2:6.4 4.0:4.4 4.1:4.8 3.5:5.0 (4.5):(7.0) 3.0:5.5 3.0:5.5 2.5:(3.5) 4.7:6.8 5.1:6.6 – 3.8:6.1 2.5:5.4 –

– 29.4 27.3 – 36.1 25 32.5 26.3 28.5 31.6 – 32.3 25.5 19.5

– 55.0 – – 52.5 44.0 54.3 48.7 57.2 – – 47.0 37.0 –

– 142 – – 107 90 132 135 143 – – 109 115 –

–

–

31.1

49.5

122

L, length; W, width; Pace Ang., pace angulation for pes. Step measurements given for pes in row corresponding to completed step. Stride for pes given in row between two left or right footprints forming stride. Tracks with prefix A are from holotype trackway: those with B and C prefixes are paratypes (L, left; R, right). Brackets refer to measurements of tracks that are poorly, or incompletely, preserved.

individual manual digit traces are not differentiated or discernible. One exception of possible interest is the Tithonian– Berriasian ornithopod trackway reported by Perez-Lorente et al. (1997) from Las Cerradicas, Spain (see Lockley & Meyer 2000, fig. 8.4; Lockley & Wright 2001). This example is selected because it was, previously, the smallest and earliest described trackway of a quadrupedal ornithopod from the Cretaceous. One of us (M.G.L.) had the opportunity to observe these tracks and obtain accurate scale tracings (Fig. 7), which compare favourably with the largest trackways of Dinehichnus from the Late Cretaceous of NE Arizona. (Fig. 7). We present only a brief comparative analysis below, as the site is currently under investigation by others (see Acknowledgements). We conclude that in several respects the Cerradicas tracks are intermediate in character between Dinehichnus tracks and larger ornithopod tracks such as Iguanodontipus (Sarjeant et al. 1998). The ichnotaxonomic interpretation of these tracks is problematic for several reasons. The pes tracks are elongate and segmented with sharp claw traces, which make them morphologically convergent with theropod tracks, and similar to the recently described ichnogenus Asianopodus (Matsukawa et al. 2005). Furthermore, the Cerradicas pes tracks, which average about 22 –23 cm in length, are considerably larger than the examples described from Canada and Thailand; however, they retain some of the primitive characteristics of Anomoepus, such as discrete phalangeal pads and relatively sharp claw traces. However, there is no trace of a hallux

on the pes. Likewise, the manus is not robust and rounded or oval as in most large iguanodontid and hadrosaurid tracks, but is instead gracile and elongate with a rhomboidal shape, lacking clearly differentiated digit impressions, although blunt digit tip traces are inferred.

Trackways from Zimbabwe Lingham-Soliar & Broderick (2000) reported an enigmatic dinosaurian trackway from the Mesozoic Dande Sandstone Formation, which is broadly dated as ?Early Jurassic to MidCretaceous. This ichnite might be considered similar to Delatorrichnus (G. Gierlinski, pers. comm.). The trackway consists of 10 small pes tracks (about 5 cm long) and corresponding small manus tracks (2–3 cm long) associated with eight of the 10 pes tracks (Fig. 8). In this regard the tracks are similar in size to the smallest of the Thailand trackways. A second trackway consists of five or six small pes tracks, also about 5 cm long, with no associated manus. Lingham-Soliar & Broderick (2000) suggested that the tracks might be similar to the Late Triassic ichnogenus Atreipus because of the tridactyl manus and pes. As stated above, there are very few trackways attributable to small quadrupedal dinosaurs, or dinosaur relatives. Indeed, prior to the discovery of the previously unnamed Thailand specimens (Le Loeuff et al. 2002; Matsukawa et al. 2006) and the Canadian tracks herein assigned to Neoanomoepus, the only small quadrupedal tracks that

262

M. G. LOCKLEY ET AL.

Fig. 7. (a) Trackway of quadrupedal ornithopod trackway from Cerradicas, Spain, based on CU Denver tracing T 1189, compared with (b) large and small bipedal Dinehichnus trackways, from the Late Jurassic of NE Arizona (Lockley et al. 1998).

had been confidently attributed to ornithischian trackmakers were Anomoepus, Delatorrichnus and Hypsiloichnus. The Late Triassic ichnogenus Atreipus (Olsen & Baird 1986) is also considered dinosaurian (ornithischian) by some researchers, but may also be attributed to a non-dinosaurian archosaur (Thulborn 1990, and citations therein). For these reasons a Late Triassic or Early Jurassic age for the Zimbabwe material was considered probable. However, the age of the Dande Sandstone Formation is poorly known, and could be Cretaceous in

part. Possible trackmakers such as the heterodontosaurids survived until the earliest Cretaceous.

Systematic ichnology General observations Anomoepus (Hitchcock 1848), from the Lower Jurassic of the Connecticut Valley region, is one of the best-preserved examples of a distinctive dinosaur track. It forms the basis of the family

TRACKS OF SMALL ORNITHISCHIANS

263

Fig. 8. From left to right: comparison between ornithischian trackways from Zimbabwe, Thailand and Canada. All drawn to same scale. Detail of holotype trackway of Neoanomoepus (far right) consists of a trackway segment of three manus–pes sets (1– 3; preserved as replica FGM 002.01.20a and also as replica CU 199.22). Manus– pes sets numbered 4– 7 include representative tracks from trackways B and C (compare with Fig. 5).

Anomoepodidae, which Lull (1953) attributed to an ornithopodan trackmaker. This ichnofamily originally contained only one ichnogenus (Anomoepus), which was traditionally considered to contain several ichnospecies (Lull 1953). However, Olsen & Rainforth (2003) considered that all ichnospecies within North America may be accommodated in A. scambus (however, for alternative views, see Lockley (2005) and Lockley & Gierlinski (2006)). The ichnogenus Moyenosauripus (Ellenberger 1974) has also been considered a synonym of Anomoepus by some workers (Olsen & Galton 1984; Olsen & Rainforth 2003), but a distinct ichnogenus by others (Gierlinski 1991; Lockley & Gierlinski 2006). Anomoepus ichnospecies vary in size from about 5 to 15 cm (length of standing pes: sensu Lull 1953); that is, the length of the pes with hallux but without metatarsus impressions. However, metatarsus impressions are relatively common in Anomoepus. Olsen & Rainforth (2003, fig 19.18) identified one specimen that has a foot length of 19 cm (see Lockley 2005, fig. 2; Lockley & Gierlinski 2006). Among the features that distinguish the best-preserved material are the presence of a

anteriomedially directed hallux and a distinctive pattern of double creasing between the digital pad impressions of the pes (see Lull 1953, figs 60, 62 – 67; Olsen & Rainforth 2003, figs 19.4 –19.5). Some specimens also reveal distinctive skin impressions consisting of fine circular tubercles that are very regular in size and shape, as seen, for example, on specimen 48/1 in the Hitchcock collection. The manus impression, according to Lull (1953), also exhibits five digit impressions with distinct digital pad traces. The step is relatively short and the axis of digit III of the pes rotates inwards. Left and right tracks are clearly distinguishable, with a moderately wide trackway (about twice as wide as footprint width) and pes pace angulation values of between 1308 and 1508 (Fig. 1). Although some specimens in the Hitchcock collection show the fine clear detail indicated by Lull (1953) in his line drawings, most do not. For example, there are many trackways of Anomoepus that indicate bipedal progression, and even where manus impressions are preserved it is hard to see all five digit impressions as clearly as Lull’s drawings indicate. These deficiencies are rectified to some extent by the work of Olsen & Rainforth

264

M. G. LOCKLEY ET AL.

(2003), whose illustrations are a great improvement on Lull’s artwork. One also does not see the pes hallux impression (digit I) in many tracks, and many also do not show metatarsal impressions. Such variability is to be expected in a large sample and can be attributed to variable preservation and/or variability in mode of progression. When we compare these classic Liassic Anomoepus tracks with those found in the basal Cretaceous sequences of British Columbia and Thailand there are clearly some distinctive similarities. The tracks fall in the same general size range (between about 5 and 12 cm pes length) with short steps, inward rotation of pes digit III and pace angulation values averaging about 1408. However, the tracks are also different in a number of respects. Most notably, despite good preservation in multiple examples, the tracks appear to lack discrete digital pad impressions on either the pes or the manus, in most cases. This could be attributed to preservation, but given that the much smaller manus tracks show all five digit impressions, a phenomenon rarely seen in the Hitchcock collections, Lull’s artwork notwithstanding, we interpret this as a primary feature. It has been noted that many dinosaur track types reveal a trend of increased fleshiness with time (Lockley 1999, 2000; Lockley et al. 2000). Thus, it is most parsimonious to infer that the lack of discrete pad impressions separated by double creases, as in Anomoepus, is a primary feature of morphological, and hence ichnotaxonomic significance. It also appears that in Cretaceous material from Canada and Thailand the hallux is directed more anteriomedially and may also be more anteriorly situated than in the type material of Anomoepus. Also, despite their depth, which is generally greater than Early Jurassic Anomoepus, the Cretaceous tracks also consistently lack full metatarsus impressions in most cases. This suggests that the Cretaceous forms may have been more digitigrade than basal ornithischians. It is for these reasons that we propose a new ichnogenus: Neoanomoepus, accommodated in the ichnofamily Anomoepodidae. Another feature of the Cretaceous trackways is that they appear to represent quadrupedal progression in a majority of cases despite the small size of the sample in comparison with the Jurassic material. Although it is debatable whether this should be considered a significant factor in formal ichnotaxonomy, on balance, the distinction between biped and quadrupeds is generally taken to be of prime importance in ichnotaxonomic descriptions. Regardless of opinion on this issue it is worth pointing out that quadrupedal progression appears to be more common among Cretaceous ornithopods than among their Jurassic ancestors, and in this sense the tendency towards becoming

facultative quadrupeds absolutely diagnostic.

is

significant

if

not

Systematic descriptions Ichnofamily Anomoepodidae Lull 1953 Emended Gierlinski 1991 Ichnogenus Neoanomoepus (Figs 4–6, 8)

ichnogen.

nov.

Diagnosis. Small trackway of a quadruped with tetradactyl pes larger than pentadactyl manus. Pes axis inwardly rotated and pes digit I short and anteromedially directed. Manus outwardly rotated and situated lateral to pes digit III or IV. Step short and trackway irregular. Ichnospecies Neoanomoepus perigrinatus ichnosp. nov. (Figs 4–6, 8, Table 1) Description. Small trackway of a quadruped with tetradactyl, slightly elongate pes (mean length 13.2 cm excluding short heel or metatarsal trace; mean width 11.1 cm), which is much larger than the pentadactyl manus (mean length 3.6 cm; mean width 6.0 cm). Pes axis inwardly rotated between 10 and 258 from trackway mid-line and pes digit I short and anteromedially directed. Manus outwardly rotated and situated lateral to pes digit III or II. Step and stride short (mean step 31.1 cm and mean stride 49.5 cm for three trackways on type slab; see Table 1). Trackway, trackway width and pace angulation irregular. Mean pace angulation 1228 (range 90 –1438, n ¼ 8). Trackway width greater for manus (mean 27.2, range 19.5–34.8 cm) than for pes (mean 24.5 cm, range 19.3–34.2 cm, n ¼ 6). Etymology. Meaning travelled widely.

‘new

Anomoepus’

that

Type material. Holotype- and paratype-bearing original slab still in field on ‘crown’ land, but mould and replica of complete slab in Fraser– Fort George Regional Museum (FGM 002.01.20). Trackway A is designated as the holotype trackway (FGM 002.01.20a), with an additional replica of holotype trackway A (CU 199.22) in the University of Colorado at Denver, Dinosaur Tracks Museum. Trackway B is designated as a paratype trackway (FGM 002.01.20b), with additional replica of trackway B (CU 199.23) in the University of Colorado at Denver, Dinosaur Tracks Museum. Type locality. Elk Valley Coal region, SE British Columbia, Canada. Type horizon and age. Mist Mountain Formation, Berriasian.

TRACKS OF SMALL ORNITHISCHIANS

Distribution and probable affinity of Neoanomoepus Various workers have attributed Anomoepus to a basal ornithischian trackmaker or more specifically to an ornithopod. However, there are no known Early Jurassic ornithopod trackmakers with splayed manus digits that would fit the Anomoepus manus footprint. Compelling ichnological evidence of ornithischian affinity in either the Jurassic or the Cretaceous generally depends on the occurrence of quadrupedal trackways, because tracks made by bipeds are generally less diagnostic. A distinction can be made between ornithischian tracks that display marked heteropody (pes much larger than manus), which are usually attributed to basal ornithischians and ornithopods, and those with less pronounced heteropody (pes and manus more equal in size), which are usually attributed to thyreophorans or ceratopsians. Based on this general distinction, Neoanomoepus seems to be of ornithopod affinity because basal ornithischians are not recorded in the Cretaceous, whereas ornithopod are present and diverse. At present very few tracks of quadrupedal ornithischians are known from the Jurassic. Anomoepus, rare Delatorrichnus and an unnamed track from Asturias, Spain, are the only quadrupedal trackmakers with strong heteropody so far reported. Thus we infer that there is a significant gap in the distribution of Anomoepodidae between the early Jurassic and the early Cretaceous. Jurassic ornithischian tracks that demonstrate less heteropody such as an unnamed trackway from the Early Jurassic of France (Le Loeuff et al. 1999) or Deltapodus (Whyte & Romano 2001) from the Middle Jurassic of England, and a similar unnamed form from the Late Jurassic of Spain (Garcı´a-Ramos et al. 2004; Gierlinski & Sabath 2009), are still poorly known and mostly without accepted ichnotaxonomies. Canadian type Neoanomoepus occurs in strata that are well dated as Berriasian. Likewise, similar trackways from Thailand, which we herein refer to as Neoanomoepus sp., have also been assigned a Berriasian age. Such correlations may be coincidental but they may also indicate an early Cretaceous ornithischian radiation. As noted above, this inference arises from the scarcity of small Anomoepuslike quadrupedal ornithischian traces in the Middle and Late Jurassic, followed by a pronounced Early Cretaceous radiation of ornithopods, many of which, based on tracks, were quadrupedal. Thus, the appearance of Neoanomoepus in the earliest Cretaceous suggests that the Early Cretaceous radiation of ornithopods, and ornithischians in general, involved small as well as large species. Recent discoveries in British Columbia indicate that ‘Anomoepodidae-like’ tracks (Fig. 9) also occur in

265

the Gorman Creek Formation of Valanginian age stratigraphically above the Mist Mountain Formation (McCrea & Buckley 2005b, 2006). This supports the inference of diverse ornithopod ichnofaunas becoming more prevalent in the early Cretaceous. For a comprehensive review of available ichnological evidence it is necessary to compare Neoanomoepus with other purported ornithopod or ornithischian tracks made by small bipeds, or quadrupeds that exhibit extreme heteropody. The only tracks that fall in this category are Early–Middle Jurassic Delatorriichnus tracks (Casamiquela 1964; Gierlinski et al. 2004) and Dineichnus footprints (Lockley et al. 1998) from the Late Jurassic. The latter ichnogenus has been attributed to a dryomorph trackmaker (Fig. 7). Hypsiloichnus from the Early Cretaceous, as the name implies, is inferred to be of hypsilophodontid origin (Stanford et al. 2004). There are significant morphological differences in all these cases. Delatorriichnus possesses a tridactyl manual print reminiscent of heterodontosaurid morphology. Dinehichnus has no trace of a functional pes digit I, and no manus impressions, whereas Neoanomoepus has a short pedal digit I and small pentadactyl manual impressions. Hypsiloichnus differs from both these ichnogenera in having a more elongate pes, longer pedal digit I and very small manual impressions (Fig. 10); however, it is not known from a trackway sequence. These differences help justify the different ichnogenus designations, and suggest that Neoanomoepus was not made by either the trackmaker of Delatorriichnus, Dineichnus or Hypsiloichnus. The unnamed trackmaker from the Cerradicas site in Spain has a pedal morphology and inward rotation similar to large Dinehichnus (Lockley et al. 1998), which in turn is similar to Asianopodus (Matsukawa et al. 2005). The elongate foot, sharp claw traces and phalangeal pad segmentation of the Cerradicas tracks means that they could easily be mistaken for a theropod if it were not for the presence of manus tracks. However, the inward rotation of the pes and rounded symmetric heel pad, which are also features of Dinehichnus tracks are typical of ornithopods. As indicated above, the inner pes digit (I) of the trackmakers of Neoanomoepus and Hypsiloichnus (Fig. 10) was long enough to make contact with the substrate, although this was not the case with the maker of Dinehichnus tracks. A well-developed hallux (digit I) is typical of the primitive condition seen in basal ornithischian, basal thyreophorans and basal ornithopods, as well as in hypsilophodontids (Weishampel et al. 2004), which may have included the maker of anomoepodid tracks (Olsen & Rainforth 2003). In contrast, the lack of a pes hallux in more derived large ornithopods

266

M. G. LOCKLEY ET AL.

Fig. 9. ‘Anomoepodidae tracks’ from the Gorman Creek Formation (Valanginian) British Columbia. It should be noted that these tracks, unlike those from the Mist Mountain Formation, show metatarsal traces and one shows phalangeal pad impressions on pes digit III.

(iguanodontids and hadrosaurs) and their tracks and in some other ornithischians (ankylosaurs) is well known (Weishampel et al. 2004). The track record supports this by showing that small forms inferred to be ornithopods retained the primitive digit I. In contrast, traces of this digit are never found in large ornithopod tracks, although they are common in many other ornithischian tracks. This shows that the track and bone records consistently provide evidence of the reduction of pes digit I in some clades of large ornithopods and among some other ornithischian tracks. The track record is also helpful in showing us the significant reduction of manual digits I and V among ornithopods. However, we know of no examples of track evidence for such significant reductions

among other ornithischians. In the Early Cretaceous, the manus tracks of Neoanomoepus, the Cerradicas specimens (Perez-Lorente et al. 1997) and various iguanodontid tracks such as Iguanodontipus (Sarjeant et al. 1998), Caririchnium (Leonardi 1984; Lockley 1987; Lockley & Wright 2001), Amblydactylus (Currie & Sarjeant 1979) and Hadrosauropodus (Lockley et al. 2004) show clear evidence of the reduction of digits I –V to produce pronounced heteropody in comparison with other ornithischian tracks. Associated with this small manus we see an amalgamation of digits II, III and IV into an integument whose distal end made hooflike impressions. In short, as the trackmakers became larger their feet became more fleshy. Such morphological shifts are size-related and involve a

TRACKS OF SMALL ORNITHISCHIANS

Fig. 10. A Hypsiloichnus manus and pes set: modified after Stanford et al. (2004).

shortening of distal phalangeal elements while the proximal elements (metatarsals and metacarpals) lengthen in compensation. This trend of reduction in number of digits, coupled with reduction of distal elements, and a shortening and broadening of the foot is not unique to ornithopods. It has been noted in the saurischian clade (Lockley 1999, 2001; Lockley et al. 1997) and even parallels the evolution of the foot in Tertiary ungulates, notably the equids. Thus, the trend in Early Cretaceous ornithopods, whether small, intermediate or large sized, is towards greater, functional digitigrady. Available ichnological information for presumed ornithopods generally supports this trend. Thus, Early Jurassic Anomoepus shows the highest incidence of metatarsal impression (pes plantigrady) and pentadactyl manus traces. Younger Late Jurassic and Early Cretaceous tracks show a number of trends towards digit reduction and greater digitigrady. In smaller Cretaceous ichnotaxa such as Neoanomoepus and Hypsiloichnus such

267

trends are not pronounced. Traces of a functional digit I indicate minimal digit reduction and attest to a primitive, more-plantigrade condition similar to Anomoepus, although without strong evidence of plantigrady in metatarsal posture. The other ‘derived’ trend is more pronounced and involves the conspicuous reduction of lateral digits I and V in both the pes and manus. Digit I reduction in the pes is confirmed by body fossils (Weishampel et al. 2004). Although, based on skeletal evidence, manus digit reduction is less pronounced, manus traces are nevertheless highly diagnostic because of the evidence of extreme digitigrady. This is most pronounced in intermediate- and large-sized species, which developed a small hoof-like manus. The similarity between Early Jurassic Anomoepus and Early Cretaceous Neoanomoepus tracks suggests iterative radiations of small ornithopods on at least two occasions, although during the early Cretaceous there was a major radiation of large ornithopods. We infer that these evolutionary events left a discernible ichnological record of trackways attributable to both quadrupeds and bipeds that are best attributed to ornithopods rather than to other ornithischians. Despite the convergence suggested by the names and descriptions, the tracks from these two periods (near the Triassic –Jurassic and Jurassic–Cretaceous boundaries; at about 208 and 145 Ma, respectively) can be differentiated on minor morphological grounds. These occurrences suggest two acme zones representing significant radiations some 60–65 Ma apart. In this regard the track record is consistent with the skeleton record of an earliest Cretaceous radiation of ornithopods involving both small clades such as the hypsilophodontids and large clades such as the iguanodontids (Weishampel et al. 2004). This convergence of the skeletal and ichnological records inspires confidence in the utility of both for recording macro-evolutionary trends. We thank V. Suteethorn, Geological Survey Division, Department of Mineral Resources, Bangkok, Thailand for help with access to the Hin Lat Pa Chad site and the trackway replicas from that location. The opportunity to make preliminary observations at the Las Cerradicas tracksite was provided to one of us (M.G.L.) by the staff of Dinopolis (Teruel). The site is currently being studied by F. Pe´rez-Lorente (La Rioja) and J. I. Canudo (Zaragoza), and for this reason we limit our interpretations to confirming the published record through direct observation of the tracks illustrated in Figure 7 herein. The opportunity to research the Mist Mountain Formation tracks was made possible by a Jurassic Foundation Grant to R.T.M. and support from S. George Pemberton. We thank G. Gierlinski (Polish Geological Institute, Warsaw) and J. Le Loeuff (Muse´e des Dinosaurs, Espe´raza, France) for their helpful review of the manuscript.

268

M. G. LOCKLEY ET AL.

References A VANZINI , M., G IERLINSKI , G. & L EONARDI , G. 2001. First report of sitting Anomoepus tracks in European Lower Jurassic (Lavini de Marco site, Northern Italy). Rivista Italian di Paleontologia e Stratigrafia, 107, 131– 136. B UFFETAUT , E. & S UTEETHORN , V. 1993. The dinosaurs of Thailand. Journal of Southeast Asian Earth Sciences, 8, 77– 82. B UFFETAUT , E., S UTEETHORN , V., T ONG , H., C HAIMANEE , Y. & K HANSUBHA , S. 1997. New dinosaur discoveries in the Jurassic and Cretaceous of northeastern Thailand. In: D HEERADILOK , P., H INTHONG , C. ET AL . (eds) Proceedings of the International Conference on the Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific. Department of Mineral Resources, Bangkok, 177– 187. B USTIN , R. M. & S MITH , G. G. 1993. Coal deposits in the Front Ranges and foothills of the Canadian Rocky Mountains, southern Canadian Cordillera. International Journal of Coal Geology, 23, 1 –27. C ASAMIQUELA , R. M. 1964. Estudios ichnologicos. Colegio Industrial Pio IX, Buenos Aires. C URRIE , P. J. 1989. Dinosaur footprints of western Canada. In: G ILLETTE , D. D & L OCKLEY , M. G. (eds) Dinosaur Tracks and Traces. Cambridge University Press, Cambridge, 293–300. C URRIE , P. & S ARJEANT , W. A. S. 1979. Lower Cretaceous dinosaur footprints from the Peace River canyon, British Columbia, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology, 28, 103– 115. E LLENBERGER , P. 1972. Contribution a` la classification des pistes des verte´bre´s du Trias. Les types du Stormberg d’Afrique du Sud (1e`re partie). Palaeovertebrata, Me´moire Extraordinaire, Montpelier, 152p. E LLENBERGER , P. 1974. Contribution a` la classification des pistes de verte´bre´s du Trias; les types du Stormberg d’Afrique du Sud, 2e partie. Palaeovertebrata, Me´moire Extraordinaire, Montpelier, 170p. G ARCI´ A -R AMOS , J. C., P INUELA , L. & L IRES , J. 2004. Guı´a del Jura´sico de Asturias. Rutas por los yacimientos de huellas de dinosaurios. Zinco Comunicacio´n, Gijo´n. G ARCI´ A -R AMOS , J. C., P INUELA , L. & L IRES , J. 2006. Atlas del Jura´sico de Asturias. Nobel, Oviedo. G IERLINSKI , G. 1991. New dinosaur ichnotaxa from the Early Jurassic of the Holy Cross Mountains, Poland. Palaeogeography, Palaeoclimatology, Palaeoecology, 85, 137–148. G IERLINSKI , G. & S ABATH , K. 2009. Stegosaurian footprints in the Morrison Formation of Utah and their implications for interpreting other ornithischian tracks. Oryctos, 8, 29–46. G IERLINSKI , G., N IEDZWIEDZKI , G. & P IENKOPWSKI , G. 2004. Tetrapod track assemblage in the Hettangian of Soltykow, Poland, and its paleoenvironmental background. Ichnos, 11, 195–213. H ITCHCOCK , E. 1848. An attempt to discriminate and describe the animals that made the fossil footmarks of the United States, and especially of New England. Transactions of the American Academy of Arts and Sciences, 3, 129 –256.

L E L OEUFF , J., L OCKLEY , M. G., M EYER , C. A. & P ETIT , J.-P. 1999. Discovery of a thyreophoran trackway in the Hettangian of central France. Comptes Rendus de l’Academie des Sciences, Se´rie II, 328, 215–219. L E L OEUFF , J., K HANSUBHA , S., B UFFETAUT , E., S UTEETHORN , V., T ONG , H. & S OUILLAT , C. 2002. Dinosaur footprints from the Phra Wihan Formation (Early Cretaceous of Thailand). Comptes Rendus Pale´vol, 1, 287–292. L EONARDI , G. 1984. Le impronte fossili di dinosauri. In: B ONAPARTE , J. F., C OLBERT , E. H. ET AL . (eds) Sulle orme dei dinosauri. Errizo, Venice, 165– 186. L INGHAM -S OLIAR , T. & B RODERICK , T. 2000. An enigmatic Early Mesozoic dinosaur trackway from Zimbabwe. Ichnos, 7, 135– 148. L OCKLEY , M. G. 1987. Dinosaur footprints from the Dakota Group of Eastern Colorado. Mountain Geologist, 24, 107–122. L OCKLEY , M. G. 1999. The Eternal Trail: A Tracker Looks at Evolution. Perseus Books, Boulder, CO. L OCKLEY , M. G. 2000. Philosophical perspectives on theropod track morphology: Blending qualities and quantities in the science of ichnology. Gaia, 5, 279–300. L OCKLEY , M. G. 2001. Trackways—dinosaur locomotion. In: B RIGGS , D. E. G. & C ROWTHER , P. (eds) Paleobiology: A Synthesis. Blackwell, Oxford, 412–416. L OCKLEY , M. G. 2005. Book review: The Great Rift Valleys of Pangea in Eastern North America. Ichnos, 12, 79–86. L OCKLEY , M. G. 2007. The morphodynamics of dinosaurs, other archosaurs and their trackways: holistic insights into relationships between feet, limbs and the whole body. In: B ROMLEY , R. & M ELCHOR , R. (eds) Ichnology at the crossroads: a multidimensional approach to the science of organism-substrate interactions. Society of Economic Paleontologists and Mineralogists, Special Publication, 88, 27–51. L OCKLEY , M. G & G IERLINSKI , G. 2006. Diverse vertebrate ichnofaunas containing Anomoepus and other unusual trace fossils from the Lower Jurassic of the western United States: Implications for paleoecology and palichnostratigraphy. New Mexico Museum of Natural History and Science, Bulletin, 37, 175–191. L OCKLEY , M. G. & H UNT , A. P. 1995. Dinosaur Tracks and Other Fossil Footprints of the Western United States. Columbia University Press, New York. L OCKLEY , M. G. & M ATSUKAWA , M. 1998. Lower Cretaceous vertebrate tracksites of East Asia. In: L UCAS , S. G., K IRKLAND , J. I. & E STEP , J. W. (eds) Lower and Middle Cretaceous Terrestrial Ecosystems. New Mexico Museum of Natural History and Science Bulletin, 14, 135 –142. L OCKLEY , M. G. & M EYER , C. A. 2000. Dinosaur Tracks and other Fossil Footprints of Europe. Columbia University Press, New York. L OCKLEY , M. G., D OS S ANTOS , V. F., M EYER , C. & H UNT , A. P. 1998. A new dinosaur tracksite in the Morrison Formation, Boundary Butte, Southeastern Utah. In: C ARPENTER , K., C HURE , D. & K IRKLAND ,

TRACKS OF SMALL ORNITHISCHIANS J. (eds) The Upper Jurassic Morrison Formation: An Interdisciplinary Study. Modern Geology, 23, 317–330. L OCKLEY , M. G. & W RIGHT , J. L. 2001. The trackways of large quadrupedal ornithopods from the Cretaceous: A review. In: C ARPENTER , K. & T ANKE , D. (eds) Mesozoic Vertebrate Life. New Research Inspired by the Paleontology of Philip J. Currie. Indiana University Press, Bloomington, 428– 442. L OCKLEY , M. G., G ARCIA -R AMOS , J. C., L IRES , J., P INUELA , L. & A VANZINI , M. 2009. A review of vertebrate track assemblages from the Late Jurassic of Asturias, Spain with comparative notes on coeval ichnofaunas from the western USA: Implications for faunal diversity in association with siliciclastic facies assemblages. Oryctos, 8, 53–70. L OCKLEY , M. G., M EYER , C. A. & M ORATALLA , J. J. 2000. Therangospodus: Trackway evidence for the widespread distribution of a Late Jurassic theropod with well-padded feet. Gaia, 15, 339– 353. L OCKLEY , M. G., N ADON , G. & C URRIE , P. J. 2004. A diverse dinosaur-bird footprint assemblage from the Lance Formation, Upper Cretaceous, eastern Wyoming: Implications for ichnotaxonomy. Ichnos, 11, 229– 249. L ULL , R. S. 1953. Triassic life of the Connecticut Valley. Connecticut Geological and Natural History Survey Bulletin, 81, 1– 336. M ATSUKAWA , M., S HIBATA , K., K UKIHARA , R., K OARAI , K. & L OCKLEY , M. G. 2005. Review of Japanese dinosaur track localities: Implications for ichnotaxonomy, paleogeography and stratigraphic correlation. Ichnos, 12, 201–222. M ATSUKAWA , M., L OCKLEY , M. G. & L I , J. 2006. Cretaceous terrestrial biotas of east and Southeast Asia, with special reference to dinosaur dominated ichnofaunas: towards a synthesis. Cretaceous Research, 27, 3–21. M C C REA , R. T. & B UCKLEY , L. G. 2005a. A dinosaur track site from the Elk Valley coal district. Marella, 20, 19–20. M C C REA , R. T. & B UCKLEY , L. G. 2005b. Preliminary palaeontological survey of vertebrate tracks and other fossils from Kakwa Provincial Park (NTS 93I/ 01). Unpublished report for B.C. Parks, 11 November 2005. M C C REA , R. T. & B UCKLEY , L. G. 2006. Research on the fossil vertebrate tracksite in Kakwa Provincial Park (August 5– 30, 2006). Unpublished report for B.C. Parks, 7 December 2006. M C C REA , R. T., C URRIE , P. J. & P EMBERTON , S. G. 2005. Canada’s largest dinosaurs: Ichnological evidence of the northernmost record of sauropods in North America. Journal of Vertebrate Paleontology, 25, 91A. O LSEN , P. E. & B AIRD , D. 1986. The ichnogenus Atreipus and its significance for Triassic biostratigraphy. In: P ADIAN , K. (ed.) The Beginning of the Age of Dinosaurs. Cambridge University Press, Cambridge, 61– 87. O LSEN , P. & G ALTON , P. 1984. A review of the reptile and amphibian assemblages from the Stormberg of

269

southern Africa, with special emphasis on the footprints and the age of the Stormberg. Palaeontologia Africana, 25, 87–110. O LSEN , P. E & R AINFORTH , E. 2003. The Early Jurassic ornithischian dinosaurian ichnogenus Anomoepus. In: L ETOURNEAU , P. M. & O LSEN , P. E. (eds) The Great Rift Valleys of Pangea in Eastern North America. Columbia University Press, New York, 314– 368. P EREZ -L ORENTE , F., C UENCA -B ESCOS , C., A URELL , M., C ANUDO , J. I., S ORIA , A. I. & R UIZ O MENACA , J. I. 1997. Las Cerradicas tracksite (Berriasian, Galve, Spain): Growing evidence for quadrupedal ornithopods. Ichnos, 5, 109– 120. R ACEY , A. & G OODALL , J. G. S. 2009. Palynology and stratigraphy of the Mesozoic Khorat Group red bed sequences from Thailand. In: B UFFETAUT , E., C UNY , G., L E L OEUFF , J. & S UTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. Geological Society, London, Special Publications, 315, 67– 81. R ACEY , A., G OODALL , J. G. S., L OVE , M. A., P OLACHAN , S. & J ONES , P. D. 1994. New age data for the Mesozoic Khorat Group of Northeast Thailand. In: A NGSUWATHANA , P., W ONGWANICH , T., T ANSATHIEN , S., W ONGSOMAK , S. & T ULYATID , J. (eds) Proceedings of the International Symposium on Stratigraphic Correlation of Southeast Asia. Department of Mineral Resources, Bangkok, 245– 252. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. Part 1. Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5– 40. S ARJEANT , W. A. S., D ELAIR , J. B. & L OCKLEY , M. G. 1998. The footprints of Iguanodon: A history and taxonomic study. Ichnos, 6, 183–202. S TANFORD , R., W EEMS , R. E. & L OCKLEY , M. G. 2004. A new dinosaur ichnotaxon from the Lower Cretaceous Patuxent Formation of Maryland and Virginia. Ichnos, 11, 251 –259. T HULBORN , R. A. 1990. Dinosaur Tracks. Chapman & Hall, London. T HULBORN , R. A. 1994. Ornithopod dinosaur tracks from the Lower Jurassic of Queensland. Alcheringa, 18, 247– 258. W EISHAMPEL , D. B., D ODSON , P. & O SMO´ LSKA , H. 2004. The Dinosauria. 2nd edn. University of California Press, Berkeley. W HYTE , M. A. & R OMANO , M. 2001. Probable stegosaurian dinosaur tracks from the Saltwick Formation (Middle Jurassic) of Yorkshire, England. Proceedings of the Geologists’ Association, 112, 45–54. Y OU , H. & A ZUMA , Y. 1995. Early Cretaceous dinosaur footprints from Luanping, Hebei Province, China. In: S UN , A. & W ANG , Y. (eds) Proceedings of the Sixth Symposium of Mesozoic Terrestrial Ecosystems and Biota. China Ocean Press, Beijing, 151–156.

Oxygen isotope composition of continental vertebrate apatites from Mesozoic formations of Thailand; environmental and ecological significance ROMAIN AMIOT1,2*, ERIC BUFFETAUT3, CHRISTOPHE LE´CUYER2,4, VINCENT FERNANDEZ5, FRANC¸OIS FOUREL2, FRANC¸OIS MARTINEAU2 & VARAVUDH SUTEETHORN6 1

Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, #142 XiZhiMenWai DaJie, Beijing 100044, China

2

UMR CNRS 5125, Pale´oenvironnements et Pale´obiosphe`re, Universite´ Lyon1, 2 rue Raphae¨l Dubois, 69622 Villeurbanne Cedex, France 3

CNRS (UMR 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure), 24 rue Lhomond, 75231 Paris Cedex 05, France

4

Institut Universitaire de France, 103 bld Saint-Michel, 75005 Paris, France 5

European Synchrotron Radiation Facility, BP220, 6 rue Jules Horowitz, 38043 Grenoble Cedex, France

6

Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected])

Abstract: Phosphatic remains (tooth enamel, turtle shell fragments and fish scales) of continental vertebrates (freshwater fish, crocodilians, turtles, and theropod and sauropod dinosaurs) recovered from eight localities of NE Thailand ranging in age from the Late Jurassic to the late Early Cretaceous have been analysed for their oxygen isotopic compositions (d18Op). From these preliminary data, local meteoric water d18Ow values estimated using d18Op values of crocodilians and turtles range from 24.1 + 2‰ at the end of the Jurassic to 28.3 + 2‰ during the Early Cretaceous, suggesting a transition from dry to wetter climates with increasing amount of seasonal precipitation from several hundred millimetres per year to several thousand millimetres. Measurable offsets in d18Op values observed between dinosaur taxa (the spinosaurid theropod Siamosaurus, other theropods and nemegtosaurid sauropods) are interpreted in terms of differences in water strategies, and suggest that Siamosaurus had habits similar to those of semi-aquatic vertebrates such as crocodilians or freshwater turtles.

Mesozoic non-marine formations cropping out on the Khorat Plateau, in NE Thailand, constitute the Khorat Group, a 3200 m thick succession of clastic sediments ranging from the Late Jurassic (Phu Kradung Formation) to Albian –Aptian (Khok Kruat Formation). Detrital material probably originated from the erosion of the Qinling orogenic belt (north of the Sichuan basin, China) and was transported by large braided river systems before being deposited in low-energy, meandering fluvial channels and on extensive floodplains (Mouret et al. 1993; Heggemann 1994; Racey et al. 1996; Carter & Bristow 2003). During the deposition of the Khorat Group, the Khorat Plateau (situated on the Indochina block) was estimated to be 500 – 1300 km north of its present position, within South

China relatively close to the Sichuan foreland basin (Fig. 1), and was displaced to its present position during the Tertiary extrusion of SE Asia caused by the collision of India with Asia (Leloup et al. 1995; Sato et al. 1999; Carter & Bristow 2003). Dry climates prevailed during the deposition of the Khorat Group according to sedimentological and paleobotanical studies, except for the upper formations, which were probably deposited under wetter conditions (Hahn 1982; Mouret et al. 1993; Heggemann 1994; Philippe et al. 2004). However, quantitative estimates of these climatic variations have not been provided so far. The Late Jurassic and Early Cretaceous Phu Kradung, Sao Khua and Khok Kruat Formations of the Khorat Group have yielded rich vertebrate

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 271–283. DOI: 10.1144/SP315.19 0305-8719/09/$15.00 # The Geological Society of London 2009.

272

R. AMIOT ET AL.

Fig. 1. General map of SE Asia showing the present (continuous lines) and restored Cretaceous (dashed lines) locations of the Khorat Basin. The pre-extrusion location places the Khorat Basin much closer to the Qinling orogenic belt and the Sichuan foreland basin, considered as the source of Khorat sediments. Fossil sites: 1, Khok Pha Suam; 2, Khok Kong; 3, Phu Phok; 4, Dan Luang; 5, Phu Nam Jun; 6, Ban Khok Sanam; 7, Phu Wiang 1; and 8, Chong Chad. Map modified from Carter & Bristow (2003).

d18O OF MESOZOIC VERTEBRATES FROM THAILAND

faunas including lungfishes, elasmobranch and semionotid fishes, amphibians, crocodilians, turtles, pterosaurs, and theropod, sauropod, ornithopod, stegosaurid and psittacosaurid dinosaurs, as well as birds. The significance of these faunas in terms of vertebrate evolution and biogeographical history of SE Asia has been illustrated by numerous studies (e.g. Buffetaut & Suteethorn 1998, 1999; Buffetaut et al. 2000, 2006). Oxygen isotopic compositions of vertebrate phosphate in bioapatite such as tooth enamel (d18Op) may allow a better knowledge of the palaeoclimatic or palaeoenvironmental conditions in which the Khorat Group was formed. The d18Op value of vertebrate apatitic tissues (bones, teeth, fish scales) is a function of the d18Obw value of the animal’s body water as well as of its body temperature (Kolodny et al. 1983; Longinelli 1984; Luz et al. 1984). The d18Obw value is related to the d18Ow value of ingested water and to the animal’s ecology and physiology. For most continental vertebrates, the main source of ingested oxygen is drinking or plant water, which is meteoric water or derived from it (D’Angela & Longinelli 1990; Cormie et al. 1994; Kohn et al. 1996; Straight et al. 2004). As the d18Ow value of meteoric water depends on climatic parameters such as air temperature, hygrometry and amount of precipitation (Dansgaard 1964; von Grafenstein et al. 1996; Fricke & O’Neil 1999), vertebrates thus indirectly record in their phosphatic tissues the climatic conditions of their living environment. It is noteworthy that the d18Ow value of surface waters can differ from that of precipitation as a result of local processes such as evaporation or mixing with other water sources, thus complicating the interpretations in terms of climatic reconstructions. Physiological adaptations to specific habitat use (such as aquatic, semi-aquatic or terrestrial) affect the d18Obw value by controlling the magnitude of body input and output oxygen fluxes, some of them being associated with oxygen isotopic fractionations (Luz & Kolodny 1985; Bryant & Froelich 1995; Kohn 1996). From living and fossil communities of mammals and reptiles, it has been observed that differences in the range of d18Op values or in some cases differences in mean d18Op values between coexisting aquatic or semi-aquatic vertebrates and terrestrial ones are related to their habitat use (Fricke & Rogers 2000; Clementz & Koch 2001; Clementz et al. 2003; Amiot et al. 2006). Ecological specificities such as plant-water use among herbivorous communities also affect the d18Op value of vertebrates. Indeed, large differences in d18Op value have been observed between coexisting herbivorous mammals that drink surface waters and those that only rely on water in plants, usually enriched by several per mil relative to surface waters (Kohn et al. 1996).

273

The d18Ow values of ingested water can be estimated from d18Op values of fish, crocodilians, turtles and dinosaurs using PO4 –water fractionation equations established for extant species (e.g. Kolodny et al. 1983; Barrick et al. 1999; Amiot et al. 2007), the applicability of these equations having been tested with Mesozoic faunas (Barrick et al. 1999; Amiot et al. 2004, 2007; Billon-Bruyat et al. 2005). This study aims to interpret preliminary results obtained from the analysis of oxygen isotopic compositions of reptile and fish phosphatic tissues in terms of environmental conditions as well as to investigate physiological and ecological features of some dinosaurs from the Khorat Group.

Material and methods Sample collection Twenty-eight vertebrate samples were analysed for their oxygen isotope compositions. Samples include dinosaur (theropod and sauropod) and crocodilian tooth enamel, turtle bony plates and ganoine coating of fish scales. These fossil remains were collected as isolated specimens in eight localities belonging to the Khorat Group and ranging from the Late Jurassic to the late Early Cretaceous (Fig. 1; Table 1). The global palaeolatitude of the Khorat Plateau during the deposition of these Mesozoic formations was about 258N (24.38N for the Khok Pha Suam locality; Amiot et al. 2004), estimated using the method described by Besse & Courtillot (1988). The Dan Luang, Phu Nam Jun, Ban Khok Sanam and Chong Chad localities are in the Phu Kradung Formation, which is probably Late Jurassic or possibly earliest Cretaceous in age (Racey et al. 1996; Buffetaut & Suteethorn 1998). The Phu Phok, Khok Kong and Phu Wiang 1 localities belong to the Sao Khua Formation, the age of which is still uncertain, although it is clearly Early Cretaceous and ante-Aptian (see Buffetaut & Suteethorn 1999). The Khok Pha Suam locality is in the Khok Kruat Formation, considered as Aptian –Albian in age on the basis of the occurrence of the freshwater hybodont shark Thaiodus (Cappetta et al. 1990). As reptile teeth are continuously replaced and take several months to grow (Erickson 1996a, b), they can record seasonal variations in ingested surface water d18Ow values. To retrieve mean annual values of local waters, enamel was sampled from the base to the apex of each tooth and several teeth from each locality were analysed. According to Clementz & Koch (2001), d18O values of at least 16 individuals per taxon may accurately reflect seasonal variability of their environment, but fossil reptile teeth from Thailand are too scarce to allow the gathering of such a

Sample number

Nature

Taxon

Locality

Formation

Age

n

d18O (‰SMOW)

Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur bulk tooth Crocodilian tooth enamel Crocodilian tooth enamel Turtle osteoscute Fish scale Fish scale Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Crocodilian tooth enamel Crocodilian tooth enamel Fish scale Dinosaur tooth enamel Dinosaur tooth enamel Crocodilian tooth enamel Crocodilian tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Dinosaur tooth enamel Turtle osteoscute Turtle osteoscute Dinosaur tooth enamel Crocodilian tooth enamel Crocodilian tooth enamel Fish scale Fish scale Fish scale

Theropoda Theropoda Theropoda Nemegtosauridae Crocodilia Crocodilia Chelonia Lepidotes Lepidotes Theropoda Theropoda Siamosaurus Siamosaurus Siamosaurus Crocodilia Crocodilia Lepidotes Theropoda Siamosaurus Crocodilia Crocodilia Theropoda Theropoda Theropoda Nemegtosauridae Siamosaurus Siamosaurus Chelonia Chelonia Theropoda Crocodilia Crocodilia Lepidotes Lepidotes Lepidotes

Khok Pha Suam Khok Pha Suam Khok Pha Suam Khok Pha Suam Khok Pha Suam Khok Pha Suam Khok Pha Suam Khok Pha Suam Khok Pha Suam Phu Phok Phu Phok Phu Phok Phu Phok Phu Phok Phu Phok Phu Phok Phu Phok Khok Kong Khok Kong Khok Kong Khok Kong Phu Wiang 1 Phu Wiang 1 Phu Wiang 1 Phu Wiang 1 Phu Wiang 1 Phu Wiang 1 Phu Wiang 1 Phu Wiang 1 Dan Luang Dan Luang Dan Luang Phu Nam Jun Ban Khok Sanam Chong Chad

Khok Kruat Khok Kruat Khok Kruat Khok Kruat Khok Kruat Khok Kruat Khok Kruat Khok Kruat Khok Kruat Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Sao Khua Phu Kradung Phu Kradung Phu Kradung Phu Kradung Phu Kradung Phu Kradung

Aptian – Albian Aptian – Albian Aptian – Albian Aptian – Albian Aptian – Albian Aptian – Albian Aptian – Albian Aptian – Albian Aptian – Albian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Valanginian – Hauterivian Late Jurassic – earliest Cretaceous Late Jurassic – earliest Cretaceous Late Jurassic – earliest Cretaceous Late Jurassic – earliest Cretaceous Late Jurassic – earliest Cretaceous Late Jurassic – earliest Cretaceous

1 1 1 1 1 1 1 1 16 1 1 1 1 1 1 1 13 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 35 14 7

15.5 14.9 17.3 20.2 14.7 14.0 12.8 16.9 17.7 19.2 18.7 16.0 14.6 17.3 13.7 15.8 15.4 21.1 16.1 14.7 16.8 26.0 24.3 26.3 19.8 13.7 13.3 14.5 14.7 19.7 17.0 19.5 14.8 11.5 19.4

Sample identification, location and stratigraphic age are reported along with number of specimens per analysis (n). *, published values (Amiot et al. 2006).

R. AMIOT ET AL.

TH001* TH005* TH008* TH006* TH002* TH007* TH003* TH004 KPS-KKfm PP-07 PP-08 PP-04 PP-05 PP-06 PP-02 PP-03 PP-SKfm KK-04 KK-03 KK-01 KK-02 PW-03 PW-04 PW-05 PW-06 PW-01 PW-02 PW-07 PW-08 DL-01 DL-02 DL-03 PNJ-PKfm KS-PKfm CC-PKfm

274

Table 1. Oxygen isotope composition of phosphate from Late Jurassic – Early Cretaceous dinosaurs, fresh water crocodilians, turtles and fish

d18O OF MESOZOIC VERTEBRATES FROM THAILAND

sample collection for geochemical analysis. For each sample, the most mineralized apatitic part was used for geochemical analysis. Reptile tooth enamel was favoured against dentine, dense bone layers were selected from turtle shells, and ganoine (an enamel-like apatitic tissue) covering the surface of fish scales was sampled.

275

compositions are quoted in the standard d notation relative to V-SMOW. Silver phosphate precipitated from standard NBS120c (natural Miocene phosphorite from Florida) was repeatedly analysed (d18O ¼ 21.70 + 0.10‰; n ¼ 12) along with the silver phosphate samples derived from the fossil vertebrate remains.

Analytical techniques Measurements of oxygen isotope compositions of apatite consist in isolating PO3 4 using acid dissolution and anion-exchange resin, according to a protocol derived from the original method published by Crowson et al. (1991) and slightly modified by Le´cuyer et al. (1993). Silver phosphate was quantitatively precipitated in a thermostatic bath set at a temperature of 70 8C. After filtration, washing with double-deionized water, and drying at 50 8C, 15 mg of Ag3PO4 was mixed with 0.8 mg of pure powder graphite. 18O/16O ratios were measured by reducing silver phosphates to CO2 using graphite reagent (O’Neil et al. 1994; Lecuyer et al. 1998). Samples were weighed into tin reaction capsules and loaded into quartz tubes and degassed for 30 min at 80 8C under vacuum. Each sample was heated at 1100 8C for 1 min to promote the redox reaction. The CO2 produced was directly trapped in liquid nitrogen to avoid any kind of isotopic reaction with quartz at high temperature. CO2 was then analysed with a GV IsoprimeTM mass spectrometer at the Laboratory UMR CNRS 5125 ‘PEPS’, University Claude Bernard Lyon 1. Isotopic

Results Oxygen isotope compositions of vertebrate phosphates are given in Table 1. The whole dataset including seven published d18Op values (Amiot et al. 2004) ranges from 11.5 to 26.3‰ V-SMOW. Mean d18Op values for each taxonomic group (theropod and sauropod dinosaurs, Siamosaurus, crocodilians, turtles and fish) are shown in Figure 2a. Estimates of the mean values of ingested surface waters, calculated using the oxygen isotope fractionation equations established between crocodilian phosphate and water (Amiot et al. 2007), turtle phosphate and water (Barrick et al. 1999), and fish phosphate and water (Kolodny et al. 1983), are given in Table 2 and displayed in Figure 2b. d18Ow values estimated using d18Op values from crocodilians and turtles decrease from 24.1 + 2.0‰ in the upper part of the Phu Kradung Formation (Dan Luang locality) to values ranging from 26.2 + 2.0‰ to 27 + 2.0‰ in the Sao Khua Formation and from 27.3 + 2.0‰ to 29.4 + 2.0‰ in the Khok Kruat Formation. Assuming that typical subtropical temperatures of

Fig. 2. (a) Mean d18Op values of vertebrate faunas from the eight localities of the Khorat Group. Measurable offsets can be observed between coexisting dinosaurs (theropods, sauropods and the spinosaurid Siamosaurus), freshwater crocodilians, turtles and fish as a consequence of ecological and physiological differences. (b) d18Ow values of ingested waters estimated using crocodilian, turtle and fish d18Op values and related fractionations equations. An uncertainty of +2‰ is used for all estimated water d18Ow values although it only corresponds to the largest calculated uncertainty.

276

Table 2. Mean oxygen isotopic composition of crocodilians, turtles and fish along with oxygen isotopic composition of environmental waters estimated for the Khorat Group localities Locality

d18Op (‰SMOW) (turtles)

n

Mean

SD

n

Mean

SD

n

Mean

SD

Khok Kruat

2

14.4

0.5

1

12.8

–

2

17.3

0.5

27.3

Sao Khua Sao Khua Sao Khua Phu Kradung Phu Kradung Phu Kradung

2 2

14.8 15.8

1.5 1.5

1

15.4

–

27 26.2

1

14.8

–

Phu Kradung

2 2

18.2

14.6

d18Op (‰SMOW) (fish at 20 – 25 8C)

d18Ow (‰SMOW) (from crocodilians)

d18Ow (‰SMOW) (from turtles)

29.4

d18Ow (‰SMOW) (from fish) 20 8C

25 8C

24.0

22.9

25.9

24.8

26.5

25.4

27.6

0.2 24.1

0.8 1

11.5

–

29.8

28.7

1

19.4

–

21.9

20.8

The following fractionation equations were used: crocodilians: d18Ow ¼ 0.82  d18Op 219.13 (Amiot et al. 2007); turtles: d18Ow ¼ 1.01  d18Op 2 22.3 (Barrick et al. 1999); fish: T(8C) ¼ 113.3 2 4.38 (d18Op 2 d18Ow) (Kolodny et al. 1983).

R. AMIOT ET AL.

Khok Pha Suam Phu Phok Khok Kong Phu Wiang 1 Phu Nam Jun Dan Luang Ban Khok Sanam Chong Chad

d18Op (‰SMOW) (crocodilians)

Formation

d18O OF MESOZOIC VERTEBRATES FROM THAILAND

about 20–25 8C occurred during the deposition of the Khorat Group (see discussion below), d18Ow values estimated using fish d18Op values show significant fluctuations within the Phu Kradung Formation (21.3 + 2.0‰ at Chong Chad, 29.2 + 2.0‰ at Khok Sanam and 25.9 + 2.0‰ at Phu Nam Jun).

Discussion Preservation of the original oxygen isotope compositions of vertebrate apatites? Secondary precipitation of apatite and isotopic exchange during microbially mediated reactions may scramble the primary isotopic signal (Blake et al. 1997; Zazzo et al. 2004a). However, apatite crystals that make up tooth enamel or fish scale ganoine are large and densely packed, and isotopic exchange under inorganic conditions has little effect on the oxygen isotope composition of phosphates even at geological time scales (Kolodny et al. 1983; Le´cuyer et al. 1999). Turtle shell bone should be more susceptible to diagenesis because hydroxylapatite crystals of bones are smaller and less densely intergrown than enamel (Kolodny et al. 1996), although several case studies have shown that the original oxygen isotope composition is preserved in Mesozoic turtle remains (Barrick et al. 1999; Amiot et al. 2004, 2006; Billon-Bruyat et al. 2005). Although no method is available to demonstrate definitely whether the oxygen isotope composition of fossil vertebrate phosphate was affected by diagenetic processes, several ways to assess the preservation state of the primary isotopic record have been proposed (e.g. Kolodny et al. 1996; Fricke & Rogers 2000; Le´cuyer et al. 2003a; Puce´at et al. 2004; Zazzo et al. 2004b). Here, the main argument supporting the preservation of the original oxygen isotope composition is the systematic offset observed between dinosaurs and ectothermic reptiles (turtles and crocodilians), which probably resulted from differences in ecology and physiology. The present dataset illustrates systematic offsets between semi-aquatic animals (turtles and crocodilians) and terrestrial ones (dinosaurs, except Siamosaurus), the latter having d18Op values 1.5–10‰ more positive than the values of crocodilians and turtles. If early diagenetic processes had occurred, they would have homogenized d18Op values of all vertebrate remains whatever the physiology and ecology of the corresponding taxa (Le´cuyer et al. 2003a). However, as only d18Op values of scale ganoine from fish were measured for the localities of Chong Chad, Khok Sanam and Phu Nam Jun, the isotopic preservation state of these fossil scales

277

remains uncertain, and the obtained values must be considered with caution.

Ecological implications In low-latitude environments, ectothermic (here crocodilians and turtles) and endothermic (dinosaurs) animals should tend to have closely similar body temperatures, leading to very small differences in oxygen isotopic fractionation between their phosphate and body water. However, variable differences in d18Op values are observed between crocodilians, turtles and the various groups of dinosaurs, and such offsets may be related to differences in physiology, ecology and origin of water intake. The most obvious difference is between semiaquatic reptiles (crocodilians and turtles) and terrestrial dinosaurs, the latter having more positive d18Op values, as a result of more important transcutaneous water evaporation, than crocodilians and turtles (which spend most of their life in water). Indeed, differences ranging from 1.6 to 3.1‰ between coexisting dinosaurs and both crocodilians and turtles have been predicted using model equations and measured on Cretaceous reptile faunas from low palaeolatitudes, assuming they had a common source of water intake (Amiot et al. 2006). However, such differences are observed here only between theropod dinosaurs (excluding Siamosaurus) and semi-aquatic reptiles from the Dan Luang and Khok Pha Suam localities. In all other localities, and between the sauropods and semiaquatic reptiles from Khok Pha Suam, differences are larger, ranging from 4.2 to 11.0‰. Such differences suggest that these dinosaurs used different water sources from those used by crocodilians and turtles. Assuming that these large isotopic offsets reflect different sources of ingested water, the following preliminary interpretations can be made. (1) Water ingested by sauropod dinosaurs was water from plant leaves, generally enriched by several per mil relative to local water as a consequence of leaf evapotranspiration and inner physiological processes (e.g. Yakir et al. 1990; Flanagan & Ehleringer 1991). This may at least partly explain the 5.2–6.6‰ differences observed between sauropods and semi-aquatic reptiles. (2) The main source of water used by theropod dinosaurs and more generally carnivorous animals is assumed to be surface waters such as streams, rivers and ponds. Some theropod dinosaurs from Thailand may have preferentially drunk from small temporary and evaporating ponds, rather than from main river streams, leading to the highly positive d18Op values observed in the localities of Phu Phok, Khok Kong and Phu Wiang 1.

278

R. AMIOT ET AL.

Siamosaurus suteethorni is a dinosaur having anomalous isotopic values compared with other dinosaurs (Table 1; Fig. 2a). It is a large spinosaurid theropod originally identified from isolated teeth (Buffetaut & Ingavat 1986), the systematic position of which was later confirmed by the discovery of a partial skeleton (Buffetaut et al. 2005). Some anatomical features of spinosaurid dinosaurs, such as their elongate snout bearing rather conical teeth, suggest that these dinosaurs ate preferentially, but not exclusively, fish (Taquet 1984; Charig & Milner 1986). This hypothesis has been supported by the discovery of partially digested fish scales in the stomach region of a nearly complete skeleton of Baryonyx walkeri, a spinosaurid dinosaur recovered from the Early Cretaceous of the Isle of Wight (Charig & Milner 1997). Also, some workers have emphasized that the massive forelimbs of Baryonyx certainly possessed sufficient strength and adequate musculature for the quadrupedal posture required when fishing either on the edge of the water or in it (Charig & Milner 1997). On the other hand, despite the close resemblance between the jaws and teeth of spinosaurs and those of some crocodilians, and the likely ichthyophagy (fish diet) of spinosaurids, no evidence clearly supporting aquatic or semi-aquatic habits for these dinosaurs has been put forward. In fact, some anatomical features indicate that spinosaurs were terrestrial, largely bipedal, dinosaurs, like other theropods. d18Op values of Siamosaurus from the three localities of Phu Phok, Khok Kong and Phu Wiang1 are 3– 12‰ lower than d18Op values of other dinosaurs and, given the uncertainties attached to the values, are indistinguishable from those obtained for crocodilians or turtles. Two hypotheses may explain such low values. The first is that these spinosaurid dinosaurs were almost exclusively ingesting water that was 18O-depleted relative to the water where coexisting crocodilians and turtles were drinking; however, the existence of 18Odepleted sources different from the rivers where crocodilians and turtles were living is highly unlikely. The second, and more likely, possibility is that Siamosaurus had somewhat semi-aquatic living habits similar to those of crocodilians and turtles. Because they spend much time in water, semiaquatic animals have a higher body water turnover and less transcutaneous water evaporation than terrestrial forms. The body water (from which oxygen is used for apatite synthesis) of semi-aquatic animals is thus less 18O-enriched than that of terrestrial animals, resulting in lower d18Op values (Kohn 1996; Amiot et al. 2007). Although it has not yet been demonstrated that some dinosaurs had adapted to the ecological niche of a semi-aquatic predator, ichnological evidence of swimming

theropods (Ezquerra et al. 2007) and evidence of fish as the preferential diet of spinosaurid dinosaurs (Charig & Milner 1997) suggest that such a hypothesis is not inconceivable. The hypothesis of physical adaptations to the semi-aquatic life of this group has to be tested with systematic oxygen isotope measurements of spinosaurid dinosaurs sampled from Africa, Europe and South America.

Climatic implications It is now widely accepted that global climates during the Mesozoic, although being globally warmer than today, underwent important thermal fluctuations with significant cooling episodes, notably during the early Cretaceous, when ‘icehouse’ intervals have been recognized (Frakes & Francis 1988; Frakes 1999; Price 1999). During the deposition of the Khorat Group (from the late Jurassic –earliest Cretaceous to the Aptian –Albian), global palaeotemperatures estimated for latitudes of about 258N roughly varied within a 20–25 8C range (Frakes et al. 1994; Frakes 1999; Le´cuyer et al. 2003b; Puce´at et al. 2003; Steuber et al. 2005). It is also noteworthy that such mean palaeotemperatures are similar to those experienced today at subtropical latitudes (Table 3; IAEA– WMO 2004). We therefore hypothesize that this temperature range occurred in NE Thailand during the Early Cretaceous. As d18Omw values of meteoric waters at tropical and subtropical latitudes are influenced more by precipitation amounts than by air temperatures (Dansgaard 1964), the mean air temperature – mean d18Omw relationships are not applicable in this case, and d18Ow values of surface waters estimated from d18Op values of vertebrates may be interpreted in terms of humidity (amount of precipitations) or aridity. For example, in subtropical climates with high amounts of seasonal precipitation (several thousands of millimetres per year), such as the monsoon climates of SE Asia, rain waters are characterized by negative mean d18Ow values of 28‰ to 26‰, whereas dry climates, experienced for example in the Middle East, are characterized by a very small amount of precipitation (less than 100 mm a21) having mean d18Ow values of 22‰ to 21‰ (Table 3). Intermediate situations, such as in Karachi (Pakistan), are characterized by intermediate d18Ow values and amounts of precipitation (i.e. 23.9‰ and 200 mm a21). In NE Thailand, palaeoclimate markers for the time span of the deposition of the Khorat Group are very scarce and are mostly aridity indicators. Sedimentological and palaeobotanical data such as silicified wood or leaf cuticles showing xerophytic features suggest that rather dry climates prevailed during the deposition of the Phu Kradung Formation, and that climate became wetter during the

23.0 20.2 25.0 25.7 25.8 26.3 28.0 2 6.0 27.1 25.8 28.1 23.9 21.1 22.2 Hong Kong (King’s Park) Fuzhou Haikou Luang-Prabang Karachi Bahrain Jeddah 4500400 5884700 5975800 4893000 4178000 4115001 4047700

China China China Lao PDR Pakistan Bahrain Saudi Arabia

228190 0000 268030 0000 208010 1200 198520 4800 248540 000 268160 1200 218300 0000

2252.4 1340.8 1654.5 1228.6 199.8 76.2 71.6

MAAT (8C) Mean precipitation (mm) Mean d18Ow (‰ SMOW) Latitude (N) Country Name Station number

Table 3. Climatic data (mean annual air temperatures (MAAT) and total annual precipitation) and mean annual d18Ow values of meteoric water for selected climatic stations of the IAEA– WMO (IAEA –WMO 2004)

d18O OF MESOZOIC VERTEBRATES FROM THAILAND

279

deposition of the overlying formations of the Khorat Group (Mouret et al. 1993; Racey et al. 1996; Philippe et al. 2004). Low d18Ow values estimated for the Sao Khua (between 27 + 2‰ and 26.2 + 2‰) and Khok Kruat (between 29.4 + 2‰ and 27.3 + 2‰) Formations match those at similar latitudes today (between 20 and 258N) in SE Asia (Table 3). This suggests that similar wet–dry climates with high amounts of seasonal precipitations of a few thousands of millimetres occurred during this period in the region of the present Khorat Plateau. Assuming mean air palaeotemperatures in the typical range of 20 –25 8C occurring today in subtropical areas, d18Ow values estimated from fish match those estimated from crocodilians at the Phu Phok locality (Sao Khua Formation), but at the Khok Pha Suam site (Khok Kruat Formation), they are 4–6‰ higher (Fig. 2b). This difference may reflect peculiar conditions experienced by fish during apatite mineralization such as a temporary isolation of their living water followed by its intense evaporation. During periods of isolation of water bodies, tetrapods can move to other water bodies, whereas fish are trapped and record environmental changes in their phosphatic tissues. Higher d18Ow values are estimated for the uppermost part of the Phu Kradung Formation (24.1 + 2‰ at the Dan Luang locality). Such values are found today at similar latitudes in drier environments characterized by smaller amounts of precipitation (several hundred millimetres per year, as in Karachi; Table 3) and might reflect similar conditions during the deposition of the uppermost part of the Phu Kradung Formation. Oxygen isotope ratios of waters estimated using fish d18Op values vary from high values (21.9‰ to 20.8‰) at the Chong Chad locality to low values (26.5‰ to 25.4‰) at Phu Nam Jun and even lower values at Ban Khok Sanam (29.8‰ to 28.7‰). These values are difficult to interpret, as only fish remains were analysed from these localities. Keeping in mind that a diagenetic origin for these values cannot be excluded, the high d18Ow value obtained from Chong Chad fish may reflect extensive water evaporation of an isolated body of water. In contrast, the low d18Ow values documented at Ban Khok Sanam could result from wetter conditions with high amounts of precipitation. However, the contribution of 18O-depleted water inputs originating from high-altitude rainfall is another hypothesis that may explain the observed low d18Ow values. Numerous studies have shown that orography affects the d18Ow values of rainwater, the latter becoming more negative with increasing altitude (e.g. Fontes & Olivry 1976; Bortolami et al. 1978; Siegenthaler & Oeschger 1980; Gonfiantini et al. 2001). As the sediments of

280

R. AMIOT ET AL.

the Khorat Group probably originated from the erosion of the Qinling belt, which underwent active orogenies during the Palaeozoic and Early Mesozoic (Triassic) (e.g. Mattauer et al. 1985; Ratschbacher et al. 2006, and references therein), it is possible that surface waters ingested by Cretaceous reptiles and fishes were 18O-depleted waters coming from the Qinling area or partly mixed with local meteoric waters. However, the palaeoelevations of the Qinling Mountains during the deposition of Khorat sediments have not yet been estimated. Moreover, the Khorat basin was about 1000 km SW of the Qinling belt during the late Jurassic –early Cretaceous, and the waters flowing from the Qinling area to the Khorat basin may have been mixed on many occasions with local meteoric waters along their path, thus losing their original values. Therefore, the possible high-altitude origin of the low values obtained for Khorat waters remains purely hypothetical. Further investigations may clarify whether these values have a diagenetic or environmental origin, and the possible high-altitude origin of estimated waters could be tested by isotopic studies of contemporaneous faunas that lived closer to the Qinling belt, such as the vertebrate faunas from the Late Jurassic of Zigong (China), located in the Sichuan basin.

Conclusion Oxygen isotope compositions of continental vertebrate remains from eight localities of the Khorat Group have been used to investigate environmental conditions that prevailed in NE Thailand during the Late Jurassic – Early Cretaceous. Ecological aspects of dinosaur faunas have also been inferred. When compared with various present-day tropical and subtropical climates, estimated d18Ow values of past meteoric water for the Khorat Group suggest a transition from dry tropical climates with low amounts of precipitation (of a few hundreds of millimetres per year) for the Phu Kradung Formation to wet–dry tropical climates for the Sao Khua and Khok Kruat Formations, characterized by high amounts of seasonal precipitation of several thousand millimetres per year, similar to present-day monsoon climates. However, a possible highaltitude origin of the low d18Ow values observed in some localities cannot be excluded. Significant offsets in d18Op values observed between theropods, sauropods and the spinosaurid Siamosaurus are interpreted in terms of differences in ingested water sources (river, pond or plant water), and also suggest that Siamosaurus had semi-aquatic living habits similar to those of crocodilians or turtles. As new fossil remains are regularly found during the continuing excavation campaigns on the

Khorat Plateau, a more extensive isotopic study of these faunas will allow a refinement of these preliminary interpretations. The authors would like to thank G. Cuny and L. Cavin for their constructive comments, and are also grateful to the two reviewers H. Fricke and A. Zazzo, whose comments helped to greatly improve the manuscript. This study was supported by the ECLIPSE and ECLIPSE2 programmes of the Centre National de la Recherche Scientifique, by the Department of Mineral Resources (Bangkok) and by a joint project of CNRS and Thailand Research Fund.

References A MIOT , R., L E´ CUYER , C., B UFFETAUT , E., F LUTEAU , F., L EGENDRE , S. & M ARTINEAU , F. 2004. Latitudinal temperature gradient during the Cretaceous Upper Campanian– Middle Maastrichtian: d18O record of continental vertebrates. Earth and Planetary Science Letters, 226, 255– 272. A MIOT , R., L E´ CUYER , C., B UFFETAUT , E., E SCARGUEL , G., F LUTEAU , F. & M ARTINEAU , F. 2006. Oxygen isotopes from biogenic apatites suggest widespread endothermy in Cretaceous dinosaurs. Earth and Planetary Science Letters, 246, 41– 54. A MIOT , R., L E´ CUYER , C. ET AL . 2007. Oxygen isotope fractionation between crocodilian phosphate and water. Palaeogeography, Palaeoclimatology, Palaeoecology, 243, 412–420. B ARRICK , R. E., F ISCHER , A. G. & S HOWERS , W. J. 1999. Oxygen isotopes from turtle bone: Application for terrestrial paleoclimates? Palaios, 14, 186 –191. B ESSE , J. & C OURTILLOT , V. 1988. Paleogeographic maps of the continents bordering the Indian Ocean since the Early Jurassic. Journal of Geophysical Research, B, Solid Earth and Planets, 93, 791– 808. B ILLON -B RUYAT , J.-P., L E´ CUYER , C., M ARTINEAU , F. & M AZIN , J.-M. 2005. Oxygen isotope compositions of Late Jurassic vertebrate remains from lithographic limestones of western Europe: Implications for the ecology of fish, turtles and crocodilians. Palaeogeography, Palaeoclimatology, Palaeoecology, 216, 359–375. B LAKE , R. E., O’N EIL , J. R. & G ARCIA , G. A. 1997. Oxygen isotope systematics of biologically mediated reactions of phosphate; I, Microbial degradation of organophosphorus compounds. Geochimica et Cosmochimica Acta, 61, 4411–4422. B ORTOLAMI , G. C., R ICCI , B., S USELLA , G. F. & Z UPPI , G. M. 1978. Isotope hydrology of the Corsaglia Maritime Alps, Piedmont, Italy. Isotope Hydrology, 1, 327–350. B RYANT , D. J. & F ROELICH , P. N. 1995. A model of oxygen isotope fractionation in body water of large mammals. Geochimica et Cosmochimica Acta, 59, 4523– 4537. B UFFETAUT , E. & I NGAVAT , R. 1986. Unusual theropod dinosaur teeth from the Upper Jurassic of Phu Wiang, Northeastern Thailand. Revue de Pale´obiologie, 5, 217–220. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D.

d18O OF MESOZOIC VERTEBRATES FROM THAILAND (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 83–90. B UFFETAUT , E. & S UTEETHORN , V. 1999. The dinosaur fauna of the Sao Khua Formation of Thailand and the beginning of the Cretaceous radiation of dinosaurs in Asia. Palaeogeography, Palaeoclimatology, Palaeoecology, 150, 13– 23. B UFFETAUT , E., S UTEETHORN , V., C UNY , G., T ONG , H., L E L OEUFF , J., K HANSUBHA , S. & J ONGAUTCHARIYAKUL , S. 2000. The earliest known sauropod dinosaur. Nature, 407, 72–74. B UFFETAUT , E., S UTEETHORN , V., L E L OEUFF , J., K HANSUBHA , S., T ONG , H. & W ONGKO , K. 2005. The dinosaur fauna from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., S RISUK , K., Y OUNGME , W. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 575–581. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J., K OBAYASHI , Y., H UANG , D. & L EE , Y. N. (eds) Papers from the 2005 Heyuan International Dinosaur Symposium. Geological Publishing House, Beijing, 19–37. C APPETTA , H., B UFFETAUT , E. & S UTEETHORN , V. 1990. A new hybodont shark from the Lower Cretaceous of Thailand. Neues Jahrbuch fu¨r Geologie und Palaeontologie, Abhandlungen, 1990, 659–666. C ARTER , A. & B RISTOW , C. S. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau Basin, eastern Thailand. Basin Research, 15, 271– 285. C HARIG , A. J. & M ILNER , A. C. 1986. Baryonyx, a remarkable new theropod dinosaur. Nature, 324, 359–361. C HARIG , A. J. & M ILNER , A. C. 1997. Baryonyx walkeri, a fish-eating dinosaur from the Wealden of Surrey. Bulletin of the Natural History Museum, Geology Series, 53, 11–70. C LEMENTZ , M. T. & K OCH , P. L. 2001. Differentiating aquatic mammal habitat and foraging ecology with stable isotope in tooth enamel. Oecologia, 129, 461–472. C LEMENTZ , M. T., H OPPE , K. A. & K OCH , P. L. 2003. A paleoecological paradox: The habitat and dietary preferences of the extinct tethythere Desmostylus, inferred from stable isotope analysis. Paleobiology, 29, 506–519. C ORMIE , A. B., L UZ , B. & S CHWARCZ , H. P. 1994. Relationship between the hydrogen and oxygen isotopes of deer bone and their use in the estimation of relative humidity. Geochimica et Cosmochimica Acta, 58, 3439– 3449. C ROWSON , R. A., S HOWERS , W. J., W RIGHT , E. K. & H OERING , T. C. 1991. A method for preparation of phosphate samples for oxygen isotope analysis. Analytical Chemistry, 63, 2397–2400. D’A NGELA , D. & L ONGINELLI , A. 1990. Oxygen isotopes in living mammal’s bone phosphate: Further results. Chemical Geology, 86, 75–82.

281

D ANSGAARD , W. 1964. Stable isotopes in precipitation. Tellus, 16, 436 –468. E RICKSON , G. M. 1996a. Daily deposition of dentine in juvenile alligator and assessment of tooth replacement rates using incremental line counts. Journal of Morphology, 228, 189–194. E RICKSON , G. M. 1996b. Incremental lines of von Ebner in dinosaurs and the assessment of tooth replacement rates using growth line counts. Evolution, 93, 14623–14627. E ZQUERRA , R., D OUBLET , S., C OSTEUR , L., G ALTON , P. M. & P E´ REZ -L ORENTE , F. 2007. Were nonavian theropod dinosaurs able to swim? Supportive evidence from an Early Cretaceous trackway, Cameros Basin (La Rioja, Spain). Geology, 35, 507– 510. F LANAGAN , L. B. & E HLERINGER , J. R. 1991. Stable isotopic composition of stem and leaf water: Applications to the study of plant water use. Functional Ecology, 5, 270– 277. F ONTES , J.-C. & O LIVRY , J.-C. 1976. Gradient isotopique entre 0 et 4000 m dans les pre´cipitations du Mont Cameroun. Re´union Annuelle des Sciences de la Terre, 4, 171. F RAKES , L. A. 1999. Estimating the global thermal state from Cretaceous sea surface and continental temperature data. In: B ARRERA , E. & J OHNSON , C. C. (eds) Evolution of the Cretaceous ocean-climate system. Geological Society of America, Special Papers, 332, 49–57. F RAKES , L. A. & F RANCIS , J. E. 1988. A guide to Phanerozoic cold polar climates from high-latitude ice-rafting in the Cretaceous. Nature, 333, 547– 549. F RAKES , L. A., P ROBST , J.-L. & L UDWIG , W. 1994. Latitudinal distribution of paleotemperature on land and sea from early Cretaceous to middle Miocene. Comptes Rendus de l’Acade´mie des Sciences, Se´rie IIA, 318, 1209– 1218. F RICKE , H. C. & O’N EIL , J. R. 1999. The correlation between 18O/16O ratios of meteoric water and surface temperature; its use in investigating terrestrial climate change over geologic time. Earth and Planetary Science Letters, 170, 181– 196. F RICKE , H. C. & R OGERS , R. R. 2000. Multiple taxon– multiple locality approch to providing oxygen isotope evidence for warm-blooded theropod dinosaurs. Geology, 28, 799–802. G ONFIANTINI , R., R OCHE , M.-A., O LIVRY , J.-C., F ONTES , J.-C. & Z UPPI , G. M. 2001. The altitude effect on the isotopic composition of tropical rains. Chemical Geology, 181, 147– 167. H AHN , L. 1982. Stratigraphy and marine ingressions of the Mesozoic Khorat group in northeastern Thailand. Geologisches Jahrbuch, 43, 7 –35. H EGGEMANN , H. 1994. Sedimenta¨re Entwicklung de Khorat-Gruppe (Ober Trias bis Pala¨ogen) in NE- und N-Thailand. Go¨ttinger Arbeiten zur Geologie und Pala¨ontologie, 63, 1 –146. IAEA–WMO 2004. Global Network of Isotopes in Precipitation. The GNIP Database. World Wide Web Address: http://isohis.iaea.org K OHN , M. J. 1996. Predicting animal d18O: Accounting for diet and physiological adaptation. Geochimica et Cosmochimica Acta, 60, 4811–4829.

282

R. AMIOT ET AL.

K OHN , M. J., S CHOENINGER , M. J. & V ALLEY , J. W. 1996. Herbivore tooth oxygen isotope compositions; effects of diet and physiology. Geochimica et Cosmochimica Acta, 60, 3889–3896. K OLODNY , Y., L UZ , B. & N AVON , O. 1983. Oxygen isotope variations in phosphate of biogenic apatites; I, Fish bone apatite; rechecking the rules of the game. Earth and Planetary Science Letters, 64, 398– 404. K OLODNY , Y., L UZ , B., S ANDER , M. & C LEMENS , W. A. 1996. Dinosaur bones: Fossils or pseudomorphs? The pitfalls of physiology reconstruction from apatitic fossils. Palaeogeography, Palaeoclimatology, Palaeoecology, 126, 161– 171. L E´ CUYER , C., G RANDJEAN , P., O’N EIL , J. R., C APPETTA , H. & M ARTINEAU , F. 1993. Thermal excursions in the ocean at Cretaceous– Tertiary boundary (northern Morocco): d18O record of phosphatic fish debris. Palaeogeography, Palaeoclimatology, Palaeoecology, 105, 235–243. L E´ CUYER , C., G RANDJEAN , P., B ARRAT , J.-A., N OLVAK , J., E MIG , C., P ARIS , F. & R OBARDET , M. 1998. d18O and REE contents of phosphatic brachiopods; a comparison between modern and lower Paleozoic populations. Geochimica et Cosmochimica Acta, 62, 2429– 2436. L E´ CUYER , C., G RANDJEAN , P. & S HEPPARD , S. M. F. 1999. Oxygen isotope exchange between dissolved phosphate and water at temperatures ,1358C: Inorganic versus biological fractionations. Geochimica et Cosmochimica Acta, 63, 855–862. L E´ CUYER , C., B OGEY , C. ET AL . 2003a. Stable isotope composition and rare earth element content of vertebrate remains from the Late Cretaceous of northern Spain (Lano); did the environmental record survive? Palaeogeography, Palaeoclimatology, Palaeoecology, 193, 457– 471. L E´ CUYER , C., P ICARD , S., G ARCIA , J. P., S HEPPARD , S. M. F., G RANDJEAN , P. & D ROMART , G. 2003b. Thermal evolution of Tethyan surface waters during the Middle–Late Jurassic: Evidence from d18O values of marine fish teeth. Paleoceanography, 18, 1076–1091. L ELOUP , P. H., L ACASSIN , R. ET AL . 1995. The Ailao Shan–Red River shear zone (Yunnan, China), Tertiary transform boundary of Indochina. Tectonophysics, 251, 3– 10. L ONGINELLI , A. 1984. Oxygen isotopes in mammal bone phosphate; a new tool for paleohydrological and paleoclimatological research? Geochimica et Cosmochimica Acta, 48, 385–390. L UZ , B. & K OLODNY , Y. 1985. Oxygen isotope variations in phosphate of biogenic apatites; IV, Mammal teeth and bones. Earth and Planetary Science Letters, 75, 29–26. L UZ , B., K OLODNY , Y. & H OROWITZ , M. 1984. Fractionation of oxygen isotopes between mammalian bone-phosphate and environmental drinking water. Geochimica et Cosmochimica Acta, 48, 1689– 1693. M ATTAUER , M., M ATTE , P. ET AL . 1985. Tectonics of the Qinling Belt: Build-up and evolution of eastern Asia. Nature, 317, 496–500. M OURET , C., H EGGEMANN , H., G OUADAIN , J. & K RASIDASIMA , S. 1993. Geological history of the siliciclastic Mesozoic strata of the Khorat Group

in the Phu Phan range area, northeastern Thailand. In: T HANASUTHIPITAK , T. (ed.) Proceedings of the International Symposium on Biostratigraphy of Mainland Southeast Asia: Facies and Paleontology. Chiang Mai University, Chiang Mai, 23– 49. O’N EIL , J. R., R OE , L. J., R EINHARD , E. & B LAKE , R. E. 1994. A rapid and precise method of oxygen isotope analysis of biogenic phosphate. Israel Journal of Earth-Sciences, 43, 203– 212. P HILIPPE , M., S UTEETHORN , V., L UTAT , P., B UFFETAUT , E., C AVIN , L., C UNY , G. & B ARALE , G. 2004. Stratigraphical and palaeobiogeographical significance of fossil wood from the Mesozoic Khorat Group of Thailand. Geological Magazine, 141, 319 –328. P RICE , G. D. 1999. The evidence and implications of polar ice during the Mesozoic. Earth-Science Reviews, 48, 183–210. P UCE´ AT , E., L E´ CUYER , C., S HEPPARD , S. M. F., D ROMART , g., R EBOULET , S. & G RANDJEAN , P. 2003. Thermal evolution of Cretaceous Tethyan marine waters inferred from oxygen isotope composition of fish tooth enamels. Paleoceanography, 18, 7– 12. P UCE´ AT , E., R EYNARD , B. & L E´ CUYER , C. 2004. Can crystallinity be used to determine the degree of chemical alteration of biogenic apatites? Chemical Geology, 205, 83– 97. R ACEY , A., L OVE , M. A., C ANHAM , A. C., G OODALL , J. G. S., P OLACHAN , S. & J ONES , P. D. 1996. Stratigraphy and reservoir poential of the Mesozoic Khorat Group, NE Thailand. Part 1: Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 5– 40. R ATSCHBACHER , L., F RANZ , L. ET AL . 2006. The SinoKorean–Yangtze suture, the Huwan detachment, and the Paleozoic–Tertiary exhumation of (ultra)highpressure rocks along the Tongbai–Xinxian– Dabie Mountains. In: H ACKER , B. R., M C C LELLAND , W. C. & L IOU , J. G. (eds) Ultrahigh-pressure Metamorphism: Deep Continental Subduction. Geological Society of America, Special Papers, 403, 45–75. S ATO , K., L IU , Y., Z HU , Z., Y ANG , Z. & O TOFUJII , Y. 1999. Palaeomagnetic study of middle Cretaceous rocks from Yunlong, western Yunnan, China: Evidence of southward displacement of Indochina. Earth and Planetary Science Letters, 165, 1– 15. S IEGENTHALER , U. & O ESCHGER , H. 1980. Correlation of d18O in precipitation with temperature and altitude. Nature, 285, 314– 317. S TEUBER , T., R AUCH , M., M ASSE , J. P., G RAAF , J. & M ALKOCˇ , M. 2005. Low-latitude seasonality of Cretaceous temperatures in warm and cold episodes. Nature, 437, 1341–1344. S TRAIGHT , W. H., B ARRICK , R. E. & E BERTH , D. A. 2004. Reflections of surface water, seasonality and climate in stable oxygen isotopes from tyrannosaurid tooth enamel. Palaeogeography, Palaeoclimatology, Palaeoecology, 206, 239–256. T AQUET , P. 1984. Une curieuse spe´cialisation du craˆne de certains dinosaures carnivores du Cre´tace´: Le museau long et e´troit des Spinosauride´s. Comptes Rendus de l’Acade´mie des Sciences, Se´rie IIA, 299, 217– 222.

d18O OF MESOZOIC VERTEBRATES FROM THAILAND G RAFENSTEIN , U., E RLENKEUSER , H., M UELLER , J., T RIMBORN , P. & A LEFS , J. 1996. A 200 year midEuropean air temperature record preserved in lake sediments; an extension of the d18OP –air temperature relation into the past. Geochimica et Cosmochimica Acta, 60, 4025– 4036. Y AKIR , D., D E N IRO , M. J. & G AT , J. R. 1990. Natural deuterium and oxygen-18 enrichment in leaf water of cotton plants grown under wet and dry conditions: Evidence for water compartmentation and its dynamics. Plant, Cell and Environment, 13, 49– 56. VON

283

Z AZZO , A., L E´ CUYER , C. & M ARIOTTI , A. 2004a. Experimentally-controlled carbon and oxygen isotope exchange between bioapatites and water under inorganic and microbially-mediated conditions. Geochimica et Cosmochimica Acta, 68, 1 –12. Z AZZO , A., L E´ CUYER , C., S HEPPARD , S. M. F., G RANDJEAN , P. & M ARIOTTI , A. 2004b. Diagenesis and the reconstruction of paleoenvironments: A method to restore original d18O values of carbonate and phosphate from fossil tooth enamel. Geochimica et Cosmochimica Acta, 68, 2245– 2258.

Biogeographical affinities of Jurassic and Cretaceous continental vertebrate assemblages from SE Asia VINCENT FERNANDEZ1,2*, JULIEN CLAUDE3, GILLES ESCARGUEL1, ERIC BUFFETAUT4 & VARAVUDH SUTEETHORN5 1

UMR CNRS 5125 ‘Pale´oenvironnements & Pale´obiosphe`re’, Universite´ Lyon 1, Campus de la Doua, F-69622, Villeurbanne, France 2

Present address: European Synchrotron Radiation Facility, BP220, 6 rue Jules Horowitz, 38043 Grenoble Cedex, France 3

UMR CNRS 5554 ‘Institut des Sciences de l’Evolution’, Universite´ Montpellier 2, Place E. Bataillon, CC 064, F-34095, Montpellier Cedex 5, France

4

UMR CNRS 8538, Laboratoire de Ge´ologie de l’Ecole Normale Supe´rieure, 24 rue Lhomond, 75231 Paris Cedex 05, France 5

Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand *Corresponding author (e-mail: [email protected])

Abstract: Over the last 25 years, rich vertebrate assemblages have been discovered in three distinct formations of the Late Jurassic and Early Cretaceous of Thailand. This work aims to compare the taxonomic assemblages of SE Asia within their palaeogeographical context in Asia. Occurrences of 477 taxa in 94 Regional Faunal Assemblages (RFA) have provided the raw material for producing a dissimilarity matrix based on the Raup & Crick index. These distances have been investigated statistically to infer relationships between the diverse faunal assemblages in space and time. Our results show that the Thai formations are more similar to each other than to any other formations, suggesting a strong provincialism. The relationship of SE Asian RFAs with other Asian RFAs is more influenced by the presence of freshwater or near-shore taxa than by strictly terrestrial ones. Our analysis shows that the faunal interchange between RFAs was rather low from the Late Jurassic to the end of the Early Cretaceous. However, faunal dispersals dramatically decreased during the mid-Early Cretaceous in Asia. The faunas show an overall stronger provincialism during the mid-Early Cretaceous, indicating the role of possible geographical barriers. This event is characterized by the absence of ornithischian dinosaurs in the Sao Khua Formation although they are present in the under- and overlying formations. Taxonomic diversity and exchanges between faunal assemblages recovered rapidly as early as the Aptian in Asia, but the fauna of SE Asia still retained a strong biogeographical signature. Supplementary material: List of regional faunal assemblages is available at http:// www.geolsoc.org.uk/SUP18349.

For the last 25 years palaeontological field work on the Khorat Plateau of NE Thailand has yielded a large number of fossil vertebrates, providing a detailed picture of the faunal succession through the five non-marine formations of the Khorat Group, which spans at least the Early Cretaceous (and possibly the latest Jurassic). Such a relatively continuous record is known from very few places in the world in terms of stratigraphic completeness. The extraordinary richness of the Asian fossil record of non-marine Mesozoic vertebrates, especially in China, Mongolia and central Asia, allows comparison with Thai faunas and reveals some interesting

similarities: for instance, the discovery of a psittacosaurid in Thailand (Buffetaut & Suteethorn 1992) has shown that this early ceratopsian was not restricted to northern and central Asia. However, other taxa from Thailand indicate a high degree of endemism, as exemplified by hybodont sharks (Cuny et al. 2005). Unexpected absences have also been noted, such as the apparent lack of ornithischian dinosaurs in the Early Cretaceous Sao Khua Formation, whereas this group flourished in other parts of Asia at that time. The coexistence in the Khorat Group of endemic and widespread taxa introduces difficulties in understanding the

From: BUFFETAUT , E., CUNY , G., LE LOEUFF , J. & SUTEETHORN , V. (eds) Late Palaeozoic and Mesozoic Ecosystems in SE Asia. The Geological Society, London, Special Publications, 315, 285–300. DOI: 10.1144/SP315.20 0305-8719/09/$15.00 # The Geological Society of London 2009.

286

V. FERNANDEZ ET AL.

palaeogeographical relationships of the Thai faunas, and thus in determining the geological age of the formations of the Khorat Group. At the top of the Khorat Group, an Aptian age is well supported by biostratigraphical evidence for the Khok Kruat Formation; but the age of the older formations is much more uncertain (Racey et al. 1996). The main obstacle to more precise dating is the lack of marine intercalations in these fully continental formations (Meesook 2000). The marginal location of Thailand, on the southeastern margin of Laurasia, since the collision of the Shan-Thai and Indochina blocks with mainland Asia more than 200 Ma ago (Metcalfe 1998), may explain some of the peculiar features of the SE Asian faunal assemblages. In addition, the complex topography of SE Asia, induced first by the Mesozoic Indosinian orogeny, may have contributed to the isolation of faunas during the Jurassic and Cretaceous, and then to the local diversification of several groups of land vertebrates. In this context, biogeographical correlation between Thai assemblages and other Asian faunas could provide better age constraints for the formations of the Khorat Group. Studies of Mesozoic land vertebrate assemblages in Asia, and especially in China, have permitted the succession of faunas to be divided into different palaeobiogeographical provinces: for instance, the concept of an Early Cretaceous Psittacosaurus –pterosaur faunal complex was established and it is still well entrenched in the literature (e.g. Dong 1973, 1979, 1992, 1993, 1995; Zhen et al. 1985). Considering the lack of temporal precision of this complex, Jerzykiewicz & Russell (1991) suggested biochronological units based on more global land-vertebrate faunas. For Late Jurassic to Late Cretaceous times, they proposed Mongolian land-vertebrate ages (MOLVAs) based on formations and fossil vertebrate assemblages from Mongolia. Those MOLVAs were then recognized by Lucas (2001) in China. The latter suggested the use of the term ‘land-vertebrate faunachrons’ (LVFs) for time intervals recognized on the basis of vertebrate faunas. Although the power of LVFs for long-distance stratigraphic correlation can be contested, the Thai assemblages have not been compared with any kind of biochronological or biogeographical general scheme: comparisons were always made considering a specific clade of the entire assemblage or in an empirical way, based on similarity at the family level (see, e.g. Buffetaut & Suteethorn 1998; Buffetaut et al. 2006). In this paper, we have two aims: (1) to place Thai assemblages in the global Asian context, using automatic classification considering all vertebrate taxa found in non-marine Late Jurassic and Cretaceous

formations; and (2) to investigate whether this kind of comparison could improve the time scale of biogeographical models.

Temporal and spatial settings The Khorat Group is now considered as comprising five continental formations ranging from the Late Jurassic–Early Cretaceous to the Aptian –Albian (Fig. 1; Racey et al. 1996; Carter & Bristow 2003). Triassic and Late Cretaceous sediments are separated from the group by a hiatus and/or unconformities. This temporal range was the basis of the study. Because of the absence of marine intercalations and of volcanic lava flows or ash layers, the age of these formations is still uncertain. Because our purpose was to compare the Thai fossil assemblages with other Asian assemblages, we took into account all the assemblages dated from the Oxfordian to the Santonian. Hence, the dataset used in this study consists of the lists of vertebrate genera occurrence in Asian continental formations during this time span. Because a significant number of species determinations are still uncertain or highly debated by palaeontologists, analyses have been performed at the genus level. Generic data, however, have been interpreted as providing a clear biodiversity signal (Sepkoski 1996, 1997) and are considered robust enough to have been used in several recent global-scale biodiversity studies (Foote 2000a, b, 2001; Kirchner & Weil 2000a, b; Brayard et al. 2006). Furthermore, because they cannot be clearly associated with a

Fig. 1. Evolution of the chronostratigraphy of the Khorat Group (after Racey et al. 1996).

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS

genus, information from ichnofossils and eggshells was not considered here. On the basis of these criteria, the Phra Wihan Formation and the Phu Phan Formation were not used in this study (Fig. 1). Thus, we studied only the Phu Kradung Formation, the Sao Khua Formation and the Khok Kruat Formation in the Khorat Plateau. These formations have yielded an important amount of fossils in several localities. However, we estimated the fossil assemblage of a formation (i.e. a time span) as the compilation of all genera found in the localities referring to this formation. Indeed, our experience shows that any marked differences between localities are merely the result of different sedimentological conditions rather than of the geographical distance separating them. Hence the standard units analysed here were fossil assemblages from three geological formations from the Khorat Group, thus at the regional scale of the Khorat Plateau. For comparative purposes, we have defined the other Asian assemblages according to similar criteria, as follows. (1) The area representing the extension of the assemblage should be of a similar size to the Khorat Plateau. In most cases, administrative provinces of the various countries considered in this study have provided clearly delimited spatial areas of similar size, although this choice can be subjective. (2) The assemblage should be representative of a specific environment and of a time span (the narrower the time span, the better the resolution for the palaeobiogeographical study). Hence the best way to constrain the resolution was to consider all the localities from the same geological formation as a single unit representing the palaeo-assemblage for a given place and a given moment. Each administrative region was, thus, characterized by one or more regional faunal assemblages (RFA) considering the number of formations (i.e. number of time sequences we can distinguish) present in this area. The detailed list of those assemblages is available on request from the corresponding author. The dataset included 94 RFAs covering Asia from its southeastern part (mostly Thailand and Laos) to Central Asia (the list of RFAs is available as supplementary material). The compiled dataset consisted of all taxa published before May 2006. A total of 477 genera represented by 776 occurrences were included as presence or absence data (zero indicates absence and one indicates presence). Although abundance data are sometimes found to be more descriptive in palaeogeographical studies than incidence data (Johnson & McCormick 1999), they are very difficult to collect from a large area, such as that covered by the present study.

287

Dataset construction The construction of the 94 RFAs was undertaken using the following protocol. First, data were taken mostly from reviews (Lillegraven et al. 1979; Dong 1992; Lucas 2001; Benton et al. 2002; Weishampel et al. 2004) and specialized papers (a full list is available on request from the corresponding author) related to the considered time span and area. Complementary information was extracted from the online Paleobiology Database (http:// paleodb.org). Taxonomic data were homogenized from updated systematic revisions. When no taxonomic revisions were available, the original author’s point of view was followed. Second, all the local faunas from each region were pooled into taxonomically standardized regional lists of genera. Third, all regional lists were brought together, leading to the construction of a presence –absence matrix (available on request from the corresponding author). In this matrix, indeterminate taxa at the supra-generic level were excluded.

Processing method Biogeographical comparisons made in this study were based on the computation of overall taxonomic similarity between regional taxonomic assemblages. The starting assumption of this ‘numerical biogeography’ was that each compared assemblage was characterized by a reasonably comprehensive and unbiased list of taxa. Nevertheless, few tools have yet been developed to control such quantitative estimates of biogeographical relatedness for the quality of the fossil record under analysis. As the sampling effort and underlying real biological diversity are largely unknown, but are likely to vary within and between the studied geographical areas, conventional incidence or abundance-based diversity measurements (e.g. taxonomic richness, Shannon’s and Simpson’s indices, etc.) were useless for this purpose. Based on these considerations, we first used taxonomic distinctness analysis to check the studied assemblages for overall comparability. Then we used the Raup & Crick (1979) index of taxonomic similarity coupled with two distinct modes of graphical display to extract and to visualize the patterns of biogeographical relatedness contained in the available data. Of the 94 original RFAs only 49 were suitable for the analysis, containing 81 different taxa found at least in two different RFAs.

Taxonomic distinctness analysis Taxonomic distinctness analysis is a robust method of diversity analysis taking into account the taxonomic (hierarchical) structure of the studied

288

V. FERNANDEZ ET AL.

assemblages (Warwick & Clarke 1995, 1998, 2001; Clarke & Warwick 1998, 1999, 2001). Based on an incidence (presence or absence) matrix of taxonomic occurrence, two complementary indices were defined: the average taxonomic distinctness (AvTD) and the variability in taxonomic distinctness (VarTD). Both indices show highly robust statistical sampling properties, including a lack of dependence, in mean value, on sample size, sampling effort and taxonomic identification skills of different workers, a very appealing characteristic in the context of this study. The computation of these two indices first relies on the construction of a Linnean taxonomic tree as a reasonable proxy of the underlying phylogenetic history. For a given assemblage made of n distinct taxa, AvTD is the average taxonomic path length measured for each of the n  (n 2 1)/2 possible pairs of taxa in the taxonomic tree. Thus, one can consider AvTD as a taxonomic disparity index reflecting the level of phylogenetic heterogeneity of the assemblage: the higher the AvTD, the more the assemblage is made of several different phylogenetic groups; conversely, the lower the AvTD, the more the assemblage is dominated by a reduced number of higher rank taxonomic groups. For instance, a taxonomic assemblage of 10 genera from the same family has the same generic richness, but a lower taxonomic disparity than an assemblage with 10 species from 10 distinct families. Thus, unless we consider highly specialized, taxonomically impoverished assemblages corresponding to rather rare and atypical environmental conditions (see Warwick & Clarke 2001), taxonomic fossil assemblages (taphocoenosis) characterized by low AvTD values are likely to poorly represent their underlying life assemblages (biocoenosis). VarTD is the variance associated with AvTD. A low VarTD value indicates that the n taxa of the assemblage tend to be ‘taxonomically equidistant’, whereas a high VarTD value illustrates a heterogeneous distribution of the pairwise taxonomic distances. Complementary to AvTD, VarTD can be viewed as a confidence index of the randomness of the fossil assemblage when compared with the underlying life assemblage: a low VarTD value, especially when associated with a low AvTD value, is likely to indicate a sampling or preservation bias, indicating that the analysed taphocoenosis is not a random sample of its parent biocoenosis (see Fig. 2 for examples). Therefore, and regardless of the origin of the bias, this taxonomic assemblage carries peculiar biogeographical information and must be taken with caution for the study. For both indices, we applied a Monte-Carlo procedure (random resampling without replacement). It estimated the confidence funnels associated with the

null hypothesis that a given observed assemblage is made of n taxa randomly sorted from the global pool of taxa recorded in the analysed dataset (see Clarke & Warwick 1998, for methodological details). In this study, we used a partially resolved taxonomic tree made of four hierarchical Linnaean levels: genus, family, order and class. A simple linear weighting scheme was adopted, with a taxonomic path length of one when contrasting two genera from the same family; two for two genera from distinct families, but from the same order; three for two genera from distinct orders, but the same class; and four for two genera from distinct classes. The confidence funnels associated with AvTD and VarTD were estimated from 100 000 random samples. All the computations have been performed using the TDA.pro software (Escarguel & Legendre 2006).

Cluster analysis of taxonomic similarity We analysed the biogeographical matrix of generic occurrence using Raup & Crick’s taxonomic similarity coefficient (RC; Raup & Crick 1979). This index is the confidence level associated with a unilateral randomization test estimating the probability that the observed number of taxa shared by two assemblages is only due to chance. More formally stated, the Raup & Crick coefficient is the 1 2 p value associated with the significance test involving the following null and alternative hypotheses. H0: the species observed in the two regions are distributed between them by random sorting from a common pool of species made up of all the taxa recorded in the biogeographical matrix. This hypothesis of independent random sprinkling of each taxon implies that the observed number of species common to both regions is only due to chance. H1: the similarity observed between the two regions is higher than would be expected as the consequence of the random sorting from a common pool of taxa. Hence, a couple of regions characterized by a very high RC value (say, RC . 0.95) show a significant similarity between their studied taxonomic assemblages (they non-randomly share too many taxa); conversely, a couple of regions characterized by a very low RC value (say, RC , 0.05) show a significant difference between their studied taxonomic assemblages (they non-randomly share too few taxa). For each pair of regions, the associated null hypothesis was estimated by generating 499 successive random samplings from the common pool of taxa without taking into account the observed probabilities of taxa occurrence (following remark 2 of Harper 1981).

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS

289

Fig. 2. Examples of representation of the composition of assemblages in an AvTD v. VarTD graph. The upper-right assemblage is composed of few groups and each group is not well represented. The upper-left assemblage is dominated by one group and contains few members of other groups. The bottom-right assemblage is composed of a few groups that are equally represented. The bottom-left assemblage is represented by an exclusive group and few or no other groups.

Once computed, the resulting matrix of similarity (S) was converted into a Euclidean matrix of dissimilarity (D) using the transformation pffiffiffiffiffiffiffiffiffiffiffiffiffiffi D ¼ (1  S) (Gower & Legendre 1986, theorem 6). Then, D was clustered using the NeighbourJoining (NJ) method of tree reconstruction (Saitou & Nei 1987; program NEIGHBOR from the PHYLIP v. 3.5 package, Felsenstein 1993). The NJ algorithm is a widely used distance-based heuristic method of phylogenetic inference (Felsenstein 2004). From a given dissimilarity matrix, it allows the computation of the shortest total length additive unrooted tree with the branch lengths estimated by unweighted least squares. We predicted that an endemic area will have few exchanges

with others, if the regional faunal assemblages form clusters that have a geographical identity; in contrast, if no geographical identity is found in the clustering pattern, the cluster could correspond to a stratigraphic assemblage based on evolutionary grounds. Finally, environment and taphonomy can also cause clustering. We should therefore keep in mind that ecological, geographical and stratigraphic signals can interfere. Because of the statistical nature of Raup & Crick’s similarity index, the tree representation of the observed dissimilarity matrix D must be very cautiously considered. Indeed, D is made of square-rooted p values that are not additive (and even not a priori metric) quantities. Thus, we

290

V. FERNANDEZ ET AL.

performed a quality analysis of the resulting NJ-tree by computing the topological criteria proposed by Gue´noche & Garreta (2001). These criteria are based on the comparison of the topology of the quadruples (i.e. trees made of only four RFAs) implied by the observed dissimilarity matrix D and by the resulting NJ-tree T. Two overall indices are computed: (1) the overall rate of well-designed quadruples (Rq), defined as the percentage of quadruples having the same topology according to D and T; (2) the arboricity coefficient (Arb), defined as the percentage of quadruples of D for which the median sum involved by Buneman’s quadruplet inequality is closer to the largest one than to the smallest one (see Gue´noche & Garreta 2001, for details). These two criteria estimate the overall topological congruence of T and D: the higher Rq and Arb, the more T actually reflects the structural information contained in D. In addition, the percentage of well-designed quadruples was computed for each RFA, as well as the rate of elementary quadruples (Re) of D supporting each internal edge of T. ‘Individual’ Rq values allowed us to identify RFAs that were potentially ill-placed within the NJ-tree (low Rq values). Re values are reliability estimates of the bipartitions induced by each edge of T; they play the same role that bootstrap supports, whose computation is not straightforward in the case of the RC index. Thus, these ‘individual’ indices provided information on the confidence associated with the internal and external edges of the tree representation: a low Rq value indicates that its associated RFA is positioned in the tree with difficulty (e.g. because of a high degree of endemism of its components), or that this RFA clusters with another one by default (e.g. because they share different taxa with different assemblages); Re provides similar interpretation but for an internal edge separating two sets of RFAs.

time span. Because the Phu Kradung Formation is not clearly defined as a Cretaceous formation, the first unit should encompass Late Jurassic RFAs. The second unit consisted of the fauna of the Sao Khua Formation (tha003) and all the RFAs ranging in age from the early Valanginian to the late Barremian. The youngest fauna of the Khorat Group is the Khok Kruat Formation (tha001) and is clearly assigned an Aptian age, so this third unit is the best constrained from a temporal point of view. As RFA could show differences in composition because of factors that result not only from the real occurrence of taxa (taphonomical, collecting biases), we focused our attention on the geographical distribution of localities that showed a rather high similarity rather than to look at those that were different. The objective was therefore not to find a connection of some kind between all assemblages but to focus on geographical and stratigraphical causes of resemblance, considering especially time intervals that correspond to deposition of SE Asian formations. High values of the RC index indicate a non-random similarity. We decided to connect assemblages with an RC index higher than 0.5 (meaning a 50% probability that the similarity is a non-random effect). Thus, we also emphasized values higher than 0.9 and 0.95, considered as the most well-supported connections, and indicated values between 0.8 and 0.9 with different symbols. If only some close localities are connected whereas localities are disconnected from more distant ones, this will illustrate the

Representation of similarities on palaeogeographical maps during different time intervals To understand the evolution of biogeographical patterns in time and space, we opted for a graphical representation of RFA similarities at different time intervals and on palaeogeographical maps. Contrary to the clustering method, the dataset was divided into three temporal units, each of these units including one of the Thai faunas. Because of the temporal incertitude about the formations of the Khorat Group, these time intervals included a large number of RFAs considered to be possibly contemporaneous with the Thai ones (based on Fig. 1 and Racey et al. 1996). The first unit included the fauna from the Phu Kradung Formation (tha002), and covers the latest Jurassic to late Berriasian

Fig. 3. Average taxonomic distinctness of the analysed RFAs as a function of the taxonomic richness. Dotted lines represent the 95% confidence interval associated with the null hypothesis that a given observed assemblage is made of n taxa randomly sorted from the global pool of taxa recorded in the analysed dataset.

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS

291

presence of provincialisms, and possibly highlight the geographical position of barriers. In contrast, if a large proportion of localities are connected without respect to biogeographical distances, this indicates that our data do not contain a significant biogeographical pattern or homogeneity in the composition of fauna throughout Asia. The results were plotted on a palaeogeographical reconstruction from Schettino & Scotese (2001).

Results Taxonomic distinctness analysis The taxonomic distinctness analysis results are given in Figures 3 and 4. The average taxonomic distinctness (Fig. 3) shows an important portion of values that are outside the funnel zone. The low values represent mainly assemblages dominated by a specific group. The variability in taxonomic distinctness (Fig. 4) shows also some of the assemblages outside the funnel zone, mostly with high values showing that too much dispersion is present in the assemblage. The combined result of VarTD and AvTD of each assemblage is shown in Figure 5. The assemblages involved in this study showed mainly VarTD values in the 95% confidence range. This signifies that assemblages are composed of several different groups of taxa, and

Fig. 4. Variability in taxonomic distinctness of the analysed RFAs as a function of the taxonomic richness. Dotted lines represent the 95% confidence interval associated with the null hypothesis that a given observed assemblage is made of n taxa randomly sorted from the global pool of taxa recorded in the analysed dataset.

are not represented by a specific group. Conversely, AvTD is mostly divided into values contained in the 95% confidence interval and low values (AvTD , AvTD  2s). These low values indicate

Fig. 5. Average taxonomic distinctness v. variability in taxonomic distinctness for each RFA. Only RFAs outside the 95% confidence area (Rand.) are labelled, because these assemblages should be carefully considered, as they represent peculiar faunal composition.

292

V. FERNANDEZ ET AL.

Fig. 6. Neighbour-Joining topology (branch lengths are arbitrary) associated with the Raup & Crick similarity matrix computed from the first generic dataset (monogeneric RFAs and unique occurrences were discarded).

assemblages dominated by a group such as dinosaurs or fish. The effect of these values on the cluster analysis is discussed below.

Cluster analysis of taxonomic similarity The resulting tree shows three main clusters corresponding to different time spans: Late Jurassic,

early Early Cretaceous and mid-Cretaceous (Fig. 6, nodes a, b and c, respectively). Unfortunately, this study includes very few formations of Late Jurassic age, mostly because of insufficient knowledge of this period in Asia. Among this set of 49 RFAs, four groups were defined. Cluster ‘a’ is mainly composed of Late Jurassic RFAs from Central China. It represents typical

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS

late Jurassic faunas; in particular, those with the euhelopodid sauropods Mamenchisaurus and Omeisaurus, the goniopholidid crocodile Sunosuchus and the chelonians Xinjiangchelys, Sinaspiderestes and Plesiochelys. Cluster ‘b’ represents middle Cretaceous RFAs (late Albian to Turonian). This cluster is characterized at its base by eastern RFAs and the four SE Asian RFAs (tha001, tha002, tha003 and lao001), whereas a subcluster contains localities scattered from the eastern to the western parts of the range of the study. This group is mainly based on chelonians, especially adocids, nanhsiungchelyids, anosteirids, lindhomemydids and trionychids. Cluster ‘d’ consists of RFAs ranging from Barremian to Albian in age. This area is the widest of the tree, with fossil assemblages from Central Asia and a majority of the RFAs, from the Gobi Desert. The

293

ceratopsian Psittacosaurus is the taxon present in a majority of the RFAs in this group. Cluster ‘e’ is dominated by Early Cretaceous East Asian assemblages in which freshwater taxa play an important part. Because of this, a large part of this group includes assemblages with a low AvTD, outside the 95% confidence ‘funnel’. The representation of AvTD v. VarTD values does not greatly affect the topology of the tree. Only node e is constrained by low AvTD values, caused by their dominance by freshwater assemblages and more specifically by fish taxa. The problematic assemblages from a representativity point of view have been removed: these were assemblages with a small number of taxa or containing genera with a single occurrence in the entire database. Figure 7 shows the validity of the position of each RFA in the tree using the criteria of Gue´noche &

Fig. 7. Comparison of the topology of the quadruples implied by the observed Raup & Crick dissimilarity matrix D and the resulting NJ-tree: percentage of ill-designed quadruples containing each RFA and rate of elementary quadruples supporting internal edges.

294

V. FERNANDEZ ET AL.

Garreta (2001). It appears that three of the four basal nodes are significantly well supported (nodes a, d and e have a rate higher than 85%). The percentage of well-designed quadruples containing each RFA is globally high, with values ranging from 61% to 90.8% and 76% of the values higher than 80%. This means that the topology reflects the dissimilarity in this range. Thus the topology of the tree was considered suitable for study. SE Asian RFAs display rather low values (61% for tha001 and 74% for tha002); therefore, it may be necessary to consider other solutions in these cases.

Representation of similarities on palaeogeographical maps during different time intervals Unit 1: Oxfordian – Berriasian (Fig. 8) The resulting network suggests a good connection between the Late Jurassic RFAs from the mainland. There are few connections between Late Jurassic assemblages and Early Cretaceous ones. The Early Cretaceous assemblages have RC index values too low for them to be connected on the

Fig. 8. Representation of the Raup & Crick similarity matrix involving RFAs contemporaneous with the Phu Kradung fauna on a palaeomap. Thresholds are applied underlying well-supported connections.

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS

295

Fig. 9. Representation of the Raup & Crick similarity matrix involving RFAs contemporaneous with the Sao Khua fauna on a palaeomap. Thresholds are applied underlying well-supported connections.

palaeogeographical map. The assemblage of the Phu Kradung Formation is also not connected.

Unit 2: Valanginian– Barremian interval (Fig. 9) In this group of RFAs, only one connection is supported by the 95% threshold, between the two assemblages of the Jehol Group (Yixian Formation chi050 and Jiufotang Formation chi043). Otherwise, only very few RFAs, which were very close geographically, have similarity indicated by a .50% RC index.

Unit 3: Aptian – Albian interval (Fig. 10) The resulting network forms a cluster indicating good connections between all the RFAs from the northern part supported by the 80% threshold. SE Asian assemblages are isolated from this cluster.

Discussion The different methods used in this study permit an understanding of the pattern of faunal evolution from the Late Jurassic to the Aptian– Albian, although not all RFAs could be included in the analyses (several assemblages are composed of taxa

296

V. FERNANDEZ ET AL.

Fig. 10. Representation of the Raup & Crick similarity matrix involving RFAs contemporaneous with the Khok Kruat fauna on a palaeomap. Thresholds are applied underlying well-supported connections.

specific to their region and hence have too low a generic richness). Looking at resemblances between localities of similar age, we can infer that the peripheral and mainland assemblages were more similar during the late Jurassic –Berriasian and during the Aptian– Albian intervals (Figs 8 and 10) than during the Valanginian –Barremian interval (Fig. 9) when very few RFAs showed a significant similarity. This suggests that an Early Cretaceous event isolated all the various parts of Asia from each other. The first phase highlighted by this study is the transition between the Late Jurassic and the Early

Cretaceous. Both methods converge to suggest a relatively radical turnover for the fauna at the boundary between these two periods. The tree shows a cluster containing most of the Jurassic assemblages (Fig. 6) and the representation of the RC index on the palaeogeographical map indicates connections between Jurassic assemblages but not with Early Cretaceous ones (Fig. 8). During the early Early Cretaceous, the situation seems to have changed to one in which faunal interchange was more difficult, at least for terrestrial taxa. The scenario, however, was not the same for terrestrial and for aquatic faunas, as aquaticdominated assemblages are connected in the tree

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS

(Fig. 6, node e). However, the reason for this clustering is based on rather low RC values and is not indicated on the palaeogeographical representation (Fig. 9). Within these relatively isolated peripheral assemblages, only one strong connection is provided by a group of Japanese assemblages (the cluster is supported at 99% in Fig. 7). Clustering between mainland assemblages is random in the tree, and when they are linked to a peripheral assemblages this is likely to be because of co-occurrence of a limited number of terrestrial taxa. Furthermore, the edge supporting this cluster in the tree (Fig. 6, node g) is supported only at a level of 49% (Fig. 7). In contrast, similarities between peripheral RFAs are because of co-occurrence of aquatic faunas. This suggests that dispersal pathways in the peripheral parts of Asia were suitable for aquatic vertebrates but more difficult for terrestrial taxa during the Early Cretaceous interval. The Aptian –Albian situation shows a change in dispersal patterns, as demonstrated by the multiple connections between RFAs suggesting good pathways for faunal exchange in all Asia (Fig. 10). The late Early Cretaceous assemblages are mostly clustered together in the tree (Fig. 6, node d). This group is statistically well supported with an Re index of 87%. From an empirical point of view, the occurrence of the genus Psittacosaurus almost everywhere in Asia supports this result (Lucas 2006). The situation of the Thai assemblages is peculiar in this evolutionary pattern, and is strongly imprinted with endemism. Thailand constitutes a cluster separated from the others (Fig. 6, node h) as the edge supporting the other group with which it is clustered has an Re of 52%. Furthermore, on the palaeogeographical representation, the Thai assemblages are never linked to any other. The endemic situation of the Khorat Plateau is clearly represented by this quantitative analysis. However, Thailand is not completely isolated from the main continent, as there are some taxa similar to those in other regions: the presence of the anosteirid Kyzylkumemys and the adocid Shachemys in the Khok Kruat Formation and in the early Late Cretaceous of Central Asia suggests a connection between those two provinces at least from the Aptian onward. These turtles explain the position of the Khorat Group with other peripheral RFAs in the tree (Fig. 6, node b). The occurrence of the genus Psittacosaurus in the same formation suggests faunal interchange with NE China, where this genus was widespread. Nevertheless, this connection was perhaps just incipient or limited, as the endemic imprint still links the Khok Kruat assemblage to that from the Sao Khua Formation. The changing biogeographical patterns outlined above show SE Asia occupying a peculiar position

297

in Asian faunal evolution during the Cretaceous. At the boundary between the Late Jurassic and the Early Cretaceous, the Phu Kradung assemblage, at the base of the Khorat Group, is one of the latest occurrences of specific Jurassic faunal elements in Asia. At that time, or just before, Indochina was not isolated, permitting the dispersal of northern or central Chinese taxa such as the crocodilian Sunosuchus or euhelopodid dinosaurs. The Early Cretaceous assemblage of the Sao Khua Formation indicates a different biogeographical pattern, as it does not show close links with other RFAs. The only possible link is based on a hybodont shark, Heteroptychodus, present also in an unnamed formation of the Matsuo Group, in Japan. However, it is too weak to be shown in this representation. This suggests that Indochina was partly isolated from the Asian mainland. Only euryhaline taxa were thus able to disperse to other provinces along coastlines. The end of this isolation is highlighted by the occurrence of the genus Psittacosaurus in the Aptian Khok Kruat Formation. The history of the relationship between SE Asian and other Asian faunas is not easy to unravel because of correlation problems linked to discontinuities in the continental fossil record. Whether the Khorat Group really provides a continuous sedimentary record for the period spanning the latest Jurassic to mid-Cretaceous interval is uncertain, and in any case the three faunal assemblages studied in this paper are separated by two formations (the Phra Wihan and Phu Phan Formations) that have yielded very few body fossils (although they contain fairly abundant dinosaur footprints; Le Loeuff et al. 2005). This probably exaggerates the impression of rapid isolation of the Indochina block faunas at the time when the Sao Khua Formation was deposited. Buffetaut et al. (2006) compared Chinese and Thai dinosaur assemblages on an empirical basis. They proposed Chinese counterparts for some of the Thai dinosaur assemblages. For the Phu Kradung assemblage, they recognized similarities to Late Jurassic Chinese assemblages, but not to those indicated by the present study. They considered the Upper Shaximiao Formation (chi089) and the Shishugou Formation (chi101) assemblages as possible counterparts. Thai assemblages are positioned between the midCretaceous Central Asian cluster (Fig. 6, node f) and the Jurassic Eastern Asian one (node a). This reflects the similarity of the fauna from the base of the Khorat Group to Jurassic assemblages and of those from later parts of the group to midCretaceous assemblages. Buffetaut et al. (2006) considered the assemblage from the Khok Kruat Formation, which is fairly well dated as Aptian, as relatively similar to that from the upper part of the late Early Cretaceous

298

V. FERNANDEZ ET AL.

Xinminbao Group of Gansu, NW China (chi000). We did not notice a clear relationship in our analysis between these two groups. This may be because especially relevant components of the Khok Kruat fauna, such as early hadrosauroids (Buffetaut et al. 2005), have not yet been described in detail and therefore were not taken into consideration in the present analysis. Buffetaut et al. (2006) also discussed the case of the abundant fauna from the Sao Khua Formation and noted that it is difficult to find a counterpart among the Early Cretaceous dinosaur faunas of China. The closest relationships seemed to be with the poorly known assemblage from the Napai Formation of Guangxi (chi021). It is difficult to compare the results of the present analysis with those of Buffetaut et al. (2006) because they are not based on taxa of the same systematic rank. Nevertheless, both approaches indicate some isolation of the fauna of the Indochina block during the deposition of the Sao Khua Formation. Cuny et al. (2003, 2006) concluded that the hybodont shark fauna from the Sao Khua Formation appears to be much less endemic, at least at the generic level, than its dinosaur assemblage. On the basis of Maisey’s (1989) work, Cuny et al. (2005) supposed that similarities between hybodont sharks from different freshwater systems are linked to their euryhaline abilities, which allowed them to travel in coastal marine waters at least for short distances. Such a model could explain the fact that land vertebrates could not disperse in and out of SE Asia at that time, but freshwater taxa seem to have been able to do so. The biostratigraphical units defined by Jerzykiewicz & Russell (1991) in Mongolia and then recognized in China by Lucas (2001) are not reflected in our study. However, it should be kept in mind that biostratigraphical and biogeographical units can be compared only to a certain extent, as they do not necessarily coincide. Our work certainly suggests that caution should be exercised when trying to recognize LVFs established in Mongolia and northern China in SE Asia, presumably because of biogeographical differences between roughly coeval assemblages. The three units spanning the Early Cretaceous are respectively the Ningjiagouan, the Tsagantsabian and the Khukhtekian LVFs. The vertebrate fauna of the Mengyin Formation (chi075) is the basis of the Ningjiagouan LVF. The Phu Kradung assemblage might be associated with this faunachron considering its indexed fossils, notably a euhelopodid sauropod and an indeterminate stegosaur. However, this conclusion is not derived from our quantitative analyses, according to which the Mengyin assemblage is connected with Early Cretaceous assemblages (chi061 and chi057 in Fig. 6), although those edges are not well supported (0.35

and 0.58, respectively), permitting other positions for this assemblage. The two other faunachrons extend from the Barremian to the late Albian. Following the current definition by Lucas (2006, p. 10) the Tsagantsabian faunachron ranges from the early Barremian to the mid-Aptian and includes chi021, chi050, chi057, chi069, chi098, chi103, kir003, mon040 and mon041 in our dataset. The Khukhtekian faunachron ranges from mid-Aptian to late Albian and is composed of the Ximinbao Group (chi000), chi055, chi76, mon011, mon019, rus004 and rus006. As is shown in our tree (Fig. 6), we did not clearly recognize those faunachrons on the basis of a quantitative analysis. Only the Psittacosaurus biochron (Lucas 2006) is recognized in our tree; that is, the biogeographical unit that includes all the RFAs where the genus Psittacosaurus occurred (node d in Fig. 6). However, as mentioned by Lucas (2006), this unit corresponds to a relatively long time span (about 20 Ma), so that it does not provide a very precise basis for correlation.

Conclusions Our results indicate that the biogeographical history of SE Asia during the Early Cretaceous was complex. During the latest Jurassic, faunal interchange between China and Indochina was possible. We cannot conclude that this applies to the entire Asian continent because of insufficient information about other parts of Asia for that period. The situation seems to have changed drastically in the middle part of the Early Cretaceous, when SE Asia apparently became isolated from the rest of Asia, possibly by mountain ranges. This hypothesis is in agreement with the tectonic situation of Asia at that time, with collision between all the microblocks from the old Gondwana (Metcalfe 2006). One of the most striking peculiarities of the vertebrate assemblage from the Sao Khua Formation is the apparent absence of ornithischian dinosaurs, which were present in SE Asia both before and after that time interval. Only some freshwater taxa could disperse between SE Asia and the rest of the continent. The situation changed again during the Aptian, when faunal interchange with other parts of Asia again became possible. Whether this temporary isolation of SE Asia also affected other parts of Asia at that time cannot be determined because of an insufficient fossil record. The authors would like to thank H. Tong, G. Cuny, L. Cavin, K. Lauprasert and J. Le Loeuff for their corrections and advice on the database. This study was supported by the ECLIPSE 2 programme of the Centre National de la Recherche Scientifique, by the Department of Mineral Resources (Bangkok) and by a CNRS– Thailand Research

KHORAT GROUP VERTEBRATE FAUNAL DYNAMICS Fund joint project. This is publication isem 2007-167 (J.C.) and contribution UMR5125-08-003 (G.E.). Finally, comments from S. Lucas (New Mexico Museum of Natural History and Science) and M. Benton (University of Bristol) greatly improved the first version of this paper.

References B ENTON , M. J., S HISHKIN , M. A., U NWIN , D. M. & K UROCHKIN , E. N. (eds) 2002. The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge. B RAYARD , A., B UCHER , H., E SCARGUEL , G., F LUTEAU , F., B OURQUIN , S. & & G ALFETTI , T. 2006. The Early Triassic ammonoid recovery: Paleoclimatic significance of diversity gradients. Palaeogeography, Palaeoclimatology, Palaeoecology, 239, 374–395. B UFFETAUT , E. & S UTEETHORN , V. 1992. A new species of the ornithischian dinosaur Psittacosaurus from the Early Cretaceous of Thailand. Palaeontology, 35, 801–812. B UFFETAUT , E. & S UTEETHORN , V. 1998. The biogeographical significance of the Mesozoic vertebrates from Thailand. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 83–90. B UFFETAUT , E., S UTEETHORN , V., L E L OEUFF , J., K HANSUBHA , S., T ONG , H. & W ONGKO , K. 2005. The dinosaur fauna from the Khok Kruat Formation (Early Cretaceous) of Thailand. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 575–581. B UFFETAUT , E., S UTEETHORN , V. & T ONG , H. 2006. Dinosaur assemblages from Thailand: A comparison with Chinese faunas. In: L U¨ , J. C., K OBAYASHI , Y., H UANG , D. & L EE , Y.-N. (eds) Papers from the 2005 Heyuan International Symposium. Geological Publishing House, Beijing, 19–37. C ARTER , A. & B RISTOW , C. S. 2003. Linking hinterland evolution and continental basin sedimentation by using detrital zircon thermochronology: A study of the Khorat Plateau Basin, Eastern Thailand. Basin Research, 15, 271– 285. C LARKE , K. & W ARWICK , R. 1998. A taxonomic distinctness index and its statistical properties. Journal of Applied Ecology, 35, 523–531. C LARKE , K. & W ARWICK , R. 1999. The taxonomic distinctness measure of biodiversity weighting of step lengths between hierarchical levels. Marine Ecology Progress Series, 184, 21–29. C LARKE , K. & W ARWICK , R. 2001. A further biodiversity index applicable to species list: Variation in taxonomic distinctness. Marine Ecology Progress Series, 216, 265–278. C UNY , G., S UTEETHORN , V., B UFFETAUT , E. & P HILIPPE , M. 2003. Hybodont sharks from the Mesozoic Khorat Group of Thailand. Mahasarakham University Journal, Special Issue, 22, 49–68. C UNY , G., S UTEETHORN , V. & K AMHA , S. 2005. A review of the hybodont sharks from the Mesozoic

299

of Thailand. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 588– 593. C UNY , G., S UTEETHORN , V., K AMHA , S., B UFFETAUT , E. & P HILIPPE , M. 2006. A new hybodont shark assemblage from the Lower Cretaceous of Thailand. Historical Biology, 18, 21– 31. D ONG , Z. 1973. Dinosaurs from Wuerho. Memoirs of the Institute of Vertebrate Paleontology and Paleoanthropology, 11, 45–52 [in Chinese]. D ONG , Z. 1979. The Cretaceous dinosaur fossils in southern China. In: Institute of Vertebrate Paleontology and Paleoanthropology, Nanjing Institute of Geology and Paleontology, Chinese Academy of Sciences (eds) Mesozoic and Cenozoic Redbeds in Southern China. Science Press, Beijing, 342–350 [in Chinese]. D ONG , Z. 1992. Dinosaurian Faunas of China. China Ocean Press, Beijing; Springer, Berlin. D ONG , Z. 1993. Early Cretaceous dinosaur faunas in China: An introduction. Canadian Journal of Earth Sciences, 30, 2096–2100. D ONG , Z. 1995. The dinosaur complexes of China and their biochronology. In: S UN , A. & W ANG , Y. (eds) Sixth Symposium on Mesozoic Terrestrial Ecosystems and Biota, Short Papers. China Ocean Press, Beijing, 91–96. E SCARGUEL , G. & L EGENDRE , S. 2006. New methods for analysing deep-time metacommunity dynamics and their application to the Paleogene mammals from the Quercy and Limagne area (Massif Central, France). Strata, Series 1, 13, 245– 273. F ELSENSTEIN , J. 1993. PHYLIP (PHYLogeny Inference Package) version 3.5. Distributed by the author, Department of Genetics, University of Washington, Seattle. F ELSENSTEIN , J. 2004. Inferring Phylogenies. Sinauer, Sunderland, MA. F OOTE , M. 2000a. Origination and extinction components of taxonomic diversity: General problems. Paleobiology, Supplement, 26, 74– 102. F OOTE , M. 2000b. Origination and extinction components of taxonomic diversity: Paleozoic and post-Paleozoic dynamics. Paleobiology, 26, 578–605. F OOTE , M. 2001. Inferring temporal patterns of preservation, origination, and extinction from taxonomic survivorship analysis. Paleobiology, 27, 602– 603. G OWER , J. & L EGENDRE , P. 1986. Metric and Euclidean properties of dissimilarity coefficients. Journal of Classification, 3, 5– 48. G UE´ NOCHE , A. & G ARRETA , H. 2001. Can we have confidence in a tree representation? Proceedings of JOBIM’2000. Lecture Notes in Computer Sciences, 2066, 43–53. H ARPER , C. W. J. 1981. Tests of significance for pairs of localities with taxa in common: A supplement to Raup and Crick (1979). Journal of Paleontology, 55, 684– 685. J ERZYKIEWICS , T. & R USSELL , D. 1991. Late Mesozoic stratigraphy and vertebrates of Gobi Basin. Cretaceous Research, 12, 345–377.

300

V. FERNANDEZ ET AL.

J OHNSON , K. G. & M C C ORMICK , T. 1999. The quantitative description of biotic change using palaeontological databases. In: H ARPER , D. A. T (ed.) Numerical Palaeobiology. Wiley, Chichester, 227– 248. K IRCHNER , J. W. & W EIL , A. 2000a. Correlations in fossil extinction and origination rates through geological time. Proceedings of the Royal Society of London, Series B, 267, 1301– 1309. K IRCHNER , J. W. & W EIL , A. 2000b. Delayed biological recovery from the extinctions throughout the fossil record. Nature, 404, 177– 180. L E L OEUFF , J., S AENYAMOON , T., S UTEETHORN , V., K HANSUBHA , S. & B UFFETAUT , E. 2005. Vertebrate footprints of South East Asia (Thailand and Laos): A review. In: W ANNAKAO , L., Y OUNGME , W., S RISUK , K. & L ERTSIRIVORAKUL , R. (eds) Proceedings of the International Conference on Geology, Geotechnology and Mineral Resources of Indochina (GEOINDO 2005). Khon Kaen University, Khon Kaen, 582 –587. L ILLEGRAVEN , J. A., K IELAN -J AWOROWSKA , Z. & C LEMENS , W. A. 1979. Mesozoic Mammals: The First Two-thirds of Mammalian History. University of California Press, Berkeley. L UCAS , S. G. 2001. Chinese Fossil Vertebrates. Columbia University Press, New York. L UCAS , S. G. 2006. The Psittacosaurus biochron, Early Cretaceous of Asia. Cretaceous Research, 27, 189– 198. M AISEY , J. G. 1989. Hamiltonichthys mapesi, g. and sp. nov. (Chondrichthyes; Elasmobranchii), from the Upper Pennsylvanian of Kansas. American Museum Novitates, 2931, 1 –42. M EESOOK , A. 2000. Cretaceous environments of Northeastern Thailand. In: O KADA , H. & M ATEER , N. J. (eds) Cretaceous Environment of Asia. Elsevier, Amsterdam, 207–223. M ETCALFE , I. 1998. Palaeozoic and Mesozoic geological evolution of the SE Asian region, multidisciplinary constraints and implications for biogeography. In: H ALL , R. & H OLLOWAY , J. D. (eds) Biogeography and Geological Evolution of SE Asia. Backhuys, Leiden, 25– 41. M ETCALFE , I. 2006. Palaeozoic and Mesozoic tectonic evolution and palaeogeograpy of the East Asian

crustal fragments: The Korean Peninsula in context. Gondwana Research, 9, 24– 46. R ACEY , A., L OVE , M., C ANHAM , A., G OODALL , J., P OLACHAN , S. & J ONES , P. 1996. Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand. Part 1: Stratigraphy and sedimentary evolution. Journal of Petroleum Geology, 19, 25–41. R AUP , D. M. & C RICK , R. E. 1979. Measurement of faunal similarity in palaeontology. Journal of Paleontology, 53, 1213–1227. S AITOU , N. & N EI , M. 1987. The Neighbor-Joining method: A new method for reconstructing phylogenetic trees. Molecular Biology and Evolution, 4, 406–425. S CHETTINO , A. & S COTESE , C. R. 2001. New Internet software aids paleomagnetic analysis and plate tectonic reconstructions. EOS Transactions, American Geophysical Union, 82, 45. World Wide Web Address: http://www.serg.unicam.it/Reconstructions.htm S EPKOSKI , J. J., J R 1996. Patterns of Phanerozoic extinctions: A perspective from global database. In: W ALLISER , O. H. (ed.) Global Events and Event Stratigraphy. Springer, Berlin, 35–52. S EPKOSKI , J. J., J R 1997. Biodiversity: Past, present and future. Journal of Paleontology, 7, 533– 539. W ARWICK , R. & C LARKE , K. 1995. New biodiversity measures reveal a decrease in taxonomic distinctness with increasing stress. Marine Ecology Progress Series, 129, 301 –305. W ARWICK , R. & C LARKE , K. 1998. Taxonomic distinctness and environmental assessment. Journal of Applied Ecology, 35, 532–543. W ARWICK , R. & C LARKE , K. 2001. Practical measures of marine biodiversity based on relatedness of species. Oceanography and Marine Biology, 39, 207–231. W EISHAMPEL , D., D ODSON , P. & O SMO´ LSKA , H. 2004. The Dinosauria. 2nd edn. University of California Press, Berkeley. Z HEN , S., Z HEN , B., M ATEER , N. J. & L UCAS , S. G. 1985. The Mesozoic reptiles of China. Bulletin of the Geological Institutions of the University of Uppsala, New Series, 11, 133– 150.

Index Figures are shown in italic font, tables in bold Acrodontinae 99, 101– 102 Actinopterygii 127, 130 adocid 143–144, 147, 148 Adocus [turtle] 166, 167 comparison, Basilochelys 155–158, 163–165 amber locality 86, 88 Ameghinichthys [fish] 129 ammonites 48, 52 amphibian remains 2– 3 anhydrite 70 Anomoeodus [fish] 136 Anomoepodidae 264, 265, 266 Anomoepus [ichnofossil] 256, 258, 261, 262–264 comparison 264, 265– 267 Ao Min, fish locality 98, 99 apatite 63 Appendicisporites [palynomorph] 76, 77 Araucaria [tree] 85, 88, 92, 93 Archaeornithomimus [theropod] 237–240 archosaur trackway 247 Argoland 12, 13 Aruacariacites [palynomorph] 74 Asteracanthus [elasmobranch] 99–101 comparison 101–102 Asterodermus [elasmobranch] 107 Baenidae 166, 167 Ban Khok Sanam locality 272, 274, 276 Ban Na Khrai, sauropod site 189, 190, 192, 195 –214 Ban Na Krai, turtle locality 144 Ban Nong Mek, sauropod site 191 Ban Saphan Hin, crocodyliform locality 175, 176, 177 Barremian Event 44 basalt, Jurassic 53, 63 Basilochelys gen nov [turtle] 154– 168 Basilochelys macrobios sp nov 142, 148, 154–155 character definition 166, 168– 170 comparisons 155– 164 pelvis 161, 164 shell measurements 155 basins of South East Asia 42 Batoidea 105–109 Belemnobatis aominensis sp nov [elasmobranch] 105– 109, 110, 111 comparison 105, 107, 109 bioapatite composition 273– 277 biodiversity 286 biogeographical provinces, Gondwana-Tethys 11– 12, 15–17 biogeography of vertebrate assemblages 285– 298 biomass, forest 93, 94 birds 2 bivalves 48 blocks, South East Asia 43, 45– 46 bone histology, Phuwiangosaurus 217– 226 methods 218, 220– 221 tissue type 219– 220, 221 –223

Buddha’s cortege 245 burial temperature 63 calc-alkaline volcanism 61 calcrete 77 Callialisporites [palynomorph] 72, 75, 77, 79, 80, 81 Carettochelyidae 166, 167 Cathaysia Divide 10–12 Cathaysialand 8, 9, 15, 20 fauna 12–14 flora 16 Ceno-Tethys 8, 12, 19, 20 Centre National de la Recherche Scientifique (France) 190 Champon Formation 50 Chelomoidea 166, 167 Chelonia 274, 293 Chelydridae 166, 167 chert 18, 20 Chondrichthyes 98–99 Chong Chad, oxygen isotope analysis 272, 274, 275, 276, 277 Chong Chat, fish locality 127, 130, 131, 137 Chuiella, distribution of 15 Cicatricosisporites [palynomorph] 75, 76, 77, 79 Cimmerian continent 8, 12, 15, 16, 18, 20 Cimmerian Event 44, 46, 47, 51, 64 cladistic analysis actinopterygians 130 crocodyliforms 183, 184, 185 sauropods 189 turtles 167, 168 climate 3, 81, 92, 94 Khorat Group 271, 276, 278– 280 oxygen isotope data 278– 280 cluster analysis 289, 292– 294, 297 coal 47, 48 Coelurosauria 230– 241 collision 12, 43, 50, 51, 52 timing of 17–18 Colobodus [bony fish] 129 colonial influence in Thailand 25–29 Concavissimisporites [palynomorph] 75, 78 conifer 85, 92, 93, 94 conodont 15, 16 consensus tree, crocodyliforms 183, 184, 185 convergence 7 coral 38, 52 Corollina [palynomorph] 75 in correlation 71, 72, 77, 78, 79, 80, 81 Counillon, J. B. H. 25– 31 expedition 27 Counillon’s dicynodonts 33–34 crocodile remains 2– 3, 98, 99, 125 oxygen isotope composition 273–278 Crocodyliformes 182– 185 systematic palaeontology 176– 182

302 Cryptodira 154– 168 Cupet, P 26, 27, 28 Cyathidites [palynomorph] 72, 74, 77, 78, 79, 80 Dan Luang locality 272, 274, 275, 276 Dasyatis [elasmobranch] 109 Deltoidospora [palynomorph] 74, 76, 77, 80 Department of Mineral Resources (Thailand) 190, 191 depositional environment, Khorat Group 44, 46, 63 Cretaceous 72 Jurassic 50, 77 Triassic 78 Deprat Affair 30 diagenesis Khorat Group 63 diamictite 10, 14 diamonds, alluvial 13, 14 Dicheiropollis [palynomorph] 73, 77 stratigraphical range 80 Dicheiropollis etruscus [palynomorph] 86 Dicynodon [reptile] 16, 31 skull 34– 38 dicynodont 1, 16, 30, 31 dicynodont fauna 33– 39 age 38–39 morphology 34– 38 diet, dinosaur 86, 278 Dinehichnus [ichnofossil] 262 dinosaur 1, 2, 3, 86 correlation 72, 78, 127, 137 oxygen isotope composition 273–278 Dinosauria histology 217– 226 Sauropoda 189–214 Therapoda 229– 240 Dracochelys [turtle] 157, 166 dropstone 14 ecological conditions and tree shape 88, 92 ecology and oxygen isotope composition 273, 277– 278, 280 elasmobranch fauna, Mid Jurassic 97 systematic palaeontology 98– 110 Elasmobranchii Bonoparte 109– 110 enameloid microstructure 105, 107, 110 endemism 38, 97, 285, 289, 297, 298 Engaibatis [elasmobranch] 107 erosion, rate of 63 Eucryptodira 154– 168 Eurasia plate 8 Eusuchia 180, 183, 184– 185 evaporites 41, 50 Exesipollenites [palynomorph] 71, 72, 76 faunachron 286, 298 faunal assemblage analysis 289 faunal provinces 15– 17 Ferganoceratodus [fish] 137 fish 2 –3 Actinopterygii 125– 138 Semionotidae 115– 123

INDEX fish, elasmobranch 97–111 morphometric study 118– 123 orientation 120, 122 oxygen isotope composition 273– 278 fission track analysis 63 floral provinces 15–17 footprints, vertebrate 229, 245– 253 forest canopy 85, 88, 91, 92, 93, 94 fossil localities 190, 272 Ao Min (fish) 98, 99 Ban Na Khrai (sauropod) 189, 190, 192, 195– 214 Ban Na Krai (turtle) 144 Ban Khok Sanam 272, 274, 276 Ban Nong Mek (sauropod) 191 Ban Saphan Hin (crocodyliform) 175, 176, 177 Chong Chad, oxygen isotope analysis 272, 274, 275, 276, 277 Chong Chat (fish) 127, 130, 131, 137 Dan Luang 272, 274, 275, 276 Hin Lat Pa Chad (footprints) 248, 257, 258 Huai Dam Chum (trackways) 250, 251 Kalasin (sauropods) 191 Kham Phok (turtle) 153, 154, 155, 156 Khao Yai, (footprints) 249 Khok Kong 144 oxygen isotope 272, 274, 275, 276, 277 Khok Pha Suam (fish) 131, 272 oxygen isotope 274, 275, 276 Khok Sanam (fish) 131, 127, 275, 277 Khon Kaen (sauropods) 190– 191 Lam Pao Dam (fish) 131, 132 Mab Chin (fish) 131 Mab Ching (fish) 125, 130, 135 Muong Phalane (trackways) 252 Phu Dam Kaeng (fish) 131 Phu Faek (footprints) 249 Phu Hin Rong Kla (footprints) 249– 250 Phu Kao (footprints) 249 Phu Kum Khao 132, 143, 189, 190, 191 Phu Mai Paw 131, 133, 134, 144 Phu Nam Jun (fish) 116, 131, 130 oxygen isotope 272, 274, 275, 276, 277 Phu Noi (fish) 137 Phu Phan Thong 131, 132, 136, 144 Phu Phan Wiang 131, 190 Phu Phok 132, 133, 134, 144 oxygen isotope 272, 274, 277 Phu Wat (turtle) 143 Phu Wiang (dinosaurs) 195, 208, 229, 230 oxygen isotope 274, 275, 276, 277 Tad Huai Nam Yai (footprints) 246, 247 Tha Song Khon (footprints) 247, 248 Wan Din So (fish) 129, 131 Wat Sak Kawan (sauropod) 189, 191 freemasonry 26, 30 French influence in Laos 25–29 French Telegraph Company 25 Gagau Group 48, 51 ganoine analysis 275, 277 Garudimimus [theropod] 237–240 geochemical analysis, phosphatic tissue 275 geochemistry, Khorat Group 60–62, 63

INDEX Glen Rose form, crocodyliform 184, 185 Gondwana 11–12, 15–17, 298 terrane dispersion 7– 20 granites 50, 51, 52, 63 Granodiscus [palynomorph] 74 Gre`s Supe´rieurs 47, 48, 49, 79, 146, 148 footprints 246, 250 –252 Gyrodus [fish] 135– 136 habitat use 273 Halecomorphi 133 Himalayan Orogeny 44, 61 Hin Lat Pa Chad, footprint site 248, 257, 258 Hoffet, J. H. 1, 30 Holostei 129–133 Huai Dam Chum, trackway site 250, 251 Huai Hin Fon Formation 101 Huai Hin Lat Formation 44, 46, 49, 60, 61, 64 footprints 246–247 Hybodus [elasmobranch] 98, 100 comparison 98–99 ice-house, Cretaceous 278 ichnology, systematic palaeontology 262–264 Indochina Block 41, 43, 46, 97 Indochina terrane, vertebrates 2 Indochina, red beds 47– 48 Indochina– Sibumasu suture 61 Indosinian Block 41, 46 Indosinian Orogeny 44, 51, 52, 64 Indosinias Moyennes 47, 48, 49 inversion, Mid-Cretaceous 46, 50 Isanemys srisuki [turtle] 143–144, 148 Isanichthys [fish] 130 Jintasakul, P 175, 176 Kalasin, sauropod site 191 K– Ar age 63 Kham Phok, turtle locality 153, 154, 155, 156 Khao Yai, footprint site 249 Khlong Min Formation 97, 99, 101, 130 fossil fish 135– 136, 136– 137 Khok Kong locality 144 oxygen isotope 272, 274, 275, 276, 277 Khok Kruat Formation 1, 43, 44, 46, 48, 49 age 52, 63, 125, 127 crocodyliform skulls 175–185 faunal assemblages 53, 286, 287, 290, 296, 297 fish 132– 135, 136– 137 footprints 246, 250 –252, 253 location 126 oxygen isotope composition 271, 273–277, 279–280 palaeocurrents 58 palynology 72 provenance 62, 64 turtle fauna 146–148 Khok Pha Suam, fish locality 131, 272 oxygen isotope 274, 275, 276

303

Khok Sanam, fish locality 131, 127, 275, 277 Khon Kaen, sauropod site 190–191 Khorat Basin 43, 51– 52, 272 Khorat Group 1, 2, 3, 41–64 climate 271, 276, 278– 280 correlation 49, 81 crocodyliforms 175– 185 faunal assemblages 285, 286, 297 palaeogeography 52– 64 palynology 70–72, 80–81 provenance 60, 71–73 stratigraphy 44, 47– 48, 71 theropod site 229 turtle assemblages 141 –149 Khorat Plateau 1 Khoratosuchus gen nov [crocodyliform] 175–185 Khoratosuchus jintasakuli sp nov 176– 177 comparison 182– 185 cranium dimensions 178, 179– 180 description 177–182 phylogenetic characteristics 185 systematic palaeontology 176– 182 Kinnareemimus gen nov [theropod] 230 Kinnareemimus khonkaenensis sp nov 230– 237 comparison 240– 241 limb bones 232– 237 relationship 237– 240 vertebrae 230 –231 Kizylkumemys [turtle] 144, 145, 146, 148 Kuchinari Group 44, 64 Lam Pao Dam, fish locality 131, 132 land-vertebrate faunachrons (LVF) 286, 298 latitude 61, 273, 278 Le Chiens Aboient 30 Lepidotes [ray-finned fish] 129, 130, 132 and oxygen isotope composition 274 Lepidotes buddhabutrensis 130 primary measurements 118, 119 taphonomic study 115– 123 Leptolepidites [palynomorph] 74, 77 Lhasa Block 19, 20, 51 lignite 48, 86 log assemblages 86, 87–92 size 89 Lonchidiidae 102–105 Lonchidion reesunderwoodi sp nov [elasmobranch] 100, 102–104, 110, 111 comparison 104– 105 Lower Nam Phong Formation 44, 46, 47, 49, 53, 82 Luang Prabang 25, 26, 28, 29, 33 Luang Prabang Formation 29–31 lungfish 125 LVF see land–vertebrate faunachrons Lycoptera [fish] 137 Lystrosaurus 33– 36, 38–39 Mab Chin, fish locality 131 Mab Ching, fish locality 125, 130, 135 Macrobaenidae 164– 165 Mae Moei Group 52

304 Maha Sarakham Formation 44, 46, 48, 49 palynology 70–72 mammal-like reptile 3, 29– 30 Mansuy, H 30 marine facies, Jurassic 48, 50 marine, Cenomanian 79 Massie, M 26, 27, 28 McCarthy, J 26, 27 Meiolaniidae 166, 167 Mesoeucrocodylia 176–185 Meso-Tethys 8, 12, 16, 18, 19, 20 Mid-Cretaceous Event 44 Mist Mountain Formation (British Colombia) 259– 260, 261, 263, 264, 265 molasse 51 monkey-puzzle tree 85, 88, 92, 93 Mouhot, H 26, 245 Muong Phalane, trackway site 252 Muse´um National d’Histoire 33, 34 Museum of Petrified Wood and Mineral Resources 175, 176, 249 Nam Phong Formation 42, 44, 47, 54, 61, 64 age 52, 69 footprint sites 247– 248, 253 palynology 78, 81 Nanhsiungchelyidae 165, 166, 167, 168 Natural History Museum, Geneva 127 Natural History Museum, London 129 Neighbour Joining method (NJ) 289, 290, 292, 293 Neoanomoepus ichnogen nov 264 locality map 257 measurements 261 trackways 259, 260, 263 Neoanomoepus perigrinatus ichnosp nov 264 affinity 265– 267 Neosuchia 176–185 NJ see Neighbour Joining method ontogeny, Phuwiangosaurus 217 –218, 221–223, 224, 226 Ornithomimosauria 3, 229–241 ornithopod track 248, 251, 252, 255–267 Osteichthyes 127 ostrich dinosaur 3, 229– 241 oxygen isotope composition, bioapatite 271–280 analytical method and results 273 –277 climate and ecology 277 –280 palaeobotany, J. B. H. Counillon’s work 29 palaeoclimate see climate palaeocological indicator 86 palaeocurrents, Khorat Group 53– 58, 62 palaeogeographic interpretation and biogeography 290, 291, 294 –298 maps 294– 296 palaeogeography 14–20 Khorat Group 52–64 palaeohistology see bone histology palaeolatitude 61, 278 and oxygen isotope composition 273

INDEX palaeomagnetic data, Khorat Group 60– 61 Palaeonisciformes 127–129 Palaeontology Museum of Savannakhet 34 palaeontology, research history 1– 3 palaeontology, systematic Actinopterygii 127–137 Crocodyliformes 176 –182 Elasmobranchii 98– 110 ichnology 262– 264 Sauropoda 194– 214 Trionychoidae 154–168 Palaeo-Pacific 12 Palaeo-Tethys 8, 9, 10, 12, 13 palaeogeography 14–18 Pallegoix, J B 245 palynology, Khorat Group 70–72, 80– 81 palynomorph 48, 50, 52, 73– 76 distribution 72 Pangaea 15, 16, 17, 18, 20 Pavie Mission 25–29, 30 Pavie, A 25, 26, 30 pelvis 161, 164, 212 –213 Perinopollenites [palynomorph] 76 petroleum exploration 2 Pha Nok Khao Formation 44 Philippine plate 8 phosphatic tissues 273 Phra Wihan Formation 42, 44, 46, 47, 48, 49 age 63, 69– 70, 81, 127 footprints 229, 248–250, 253, 257, 263 palaeocurrents 56 palynology 77, 80 provenance 62 Phrabat temple site 245 Phu Dam Kaeng, fish locality 131 Phu Faek, footprint site 249 Phu Hin Rong Kla, footprint site 249– 250 Phu Kao, footprint site 249 Phu Kradung Formation 42, 44, 46, 47, 48, 49 age 52, 63, 69–70, 81, 82, 127 faunal assemblages 53, 286, 287, 290, 297, 298 fish 116, 127–131, 132, 136 –137 footprint site 250, 253 location 126 log assemblages 86 oxygen isotope composition 271, 273–277, 279, 280 palynology 78, 80–81 provenance 55, 61, 62, 64 turtle assemblages 141– 143, 153– 168 Phu Kum Khao Dinosaur Research Centre 191 Phu Kum Khao locality 132, 143, 189, 190, 191 Phu Mai Paw locality 131, 133, 134, 144 Phu Nam Jun, fish locality 116, 131, 130 oxygen isotope composition 272, 274, 275, 276, 277 Phu Noi, fish locality 137 Phu Phan Formation 43, 44, 46, 47, 49, 253 age 63, 127 palaeocurrents 57 palynology 72 provenance 61, 62 Phu Phan Thong locality 131, 132, 137, 144 Phu Phan Wiang locality 131, 190 Phu Phok locality 132, 133, 134, 144 oxygen isotope composition 272, 274, 277

INDEX Phu Tok Formation 44, 46, 49, 61 theropod footprints 245 Phu Wat, turtle locality 143 Phu Wiang, dinosaur site 195, 208, 229, 230 oxygen isotope 274, 275, 276, 277 Phuwiangosaurus sirindhornae [sauropod] 192, 193, 194 bone histology 217–226 comparison with holotype 208–213 comparison with other sauropods 226 pelvis 212–213 skull 195–203 tooth 203, 211 vertebrae 203–208, 209– 213 dimensions 205, 206, 207, 209 phytosaur remains 2– 3 phytosaur tracks 247 plant remains 1 –2, 48 see also tree plates 46 Plesiochelyidae 166, 167 plesiosaur 2 Pleurodira 166, 167 Podicarpus [elasmobranch] 85, 92, 93 probability funnel 288, 290, 293 provenance, Khorat Group 60 provincialism 286, 291 Psittacosaurus 293, 297, 298 Psittacosaurus–pterosaur faunal complex 286 pterosaur remains 2– 3 Ptycholepis [fish] 127–129 purple beds, fossils in 33, 34, 35, 37, 38–39 Pycnodontiformes 135– 136 Qiantang Block 51 Qinling Orogenic belt 62, 64 radiometric dating 50, 63 Raup and Crick’s taxonomic similarity coefficient (RC) 288– 290 RC values 288– 290, 292– 296, 297 Reclus, E 245 red beds 1, 3, 33 Khorat Group 41– 64, 69–82 Red River Fault 60 and red beds 59 regional faunal assemblages (RFA) 287, 289– 297 Repelin, J 33 reptile remains 29– 30, 48, 98 RFA see regional faunal assemblages Rhinobatiformes 105– 109 Rhinobatos [elasmobranch] 107 Rhombopterygia [elasmobranch] 107 rift basins 46 rifting, Gondwana 8 –10, 12– 14 River Mekong 26, 28 Royal Geographical Society 26 Sao Khua Formation 42, 44, 46, 47, 48, 56 age 69–70, 81, 127 faunal assemblages 53, 97 affinities 285– 287, 290, 295, 297, 298 fish 131– 132, 133– 134, 136– 138

305

footprints 250, 253 location 126 oxygen isotope composition 271, 273– 277, 279– 280 palynology 77, 80 sauropods 189, 190, 191, 193, 218 theropods 229, 230 turtle fauna 143– 146 Saraburi Group 44 Sarcopterygii 137 sauropod 3, 94, 99, 125 Sauropoda 189– 214 systematic palaeontology 194– 214 Savannakhet Basin 50, 53 sea-level change 63– 64 sedimentation rate, Khorat Group 53 seismic data 64 Semionotidae 129– 133 Service ge´ologique de l’Indochine 1 sexual dimorphism 37, 109, 146 Shachemys [turtle] 146, 147 Shan Thai terrane 10–12, 97, 125, 130 Shan-Thai Block 2, 41 shark 2–3, 125 shark teeth 78, 153 Siamamia [fish] 133 Siamopodus khaoyaiensis [footprint cast] 249 Siamosaurus, oxygen isotope values 274, 278, 280 Sibumasu Block 41, 43, 51 Sibumasu terrane 8, 9, 10–12, 13, 15, 18 vertebrates 2 Sibumasu terrane see Shan Thai terrane Sichuan Orogenic Basin 62 Sinamiidae 133–135 Sinemydidae 164–165 Singa Formetion 14 Sirindhorn Museum 97, 116, 125, 127, 153, 189, 191, 247 skull crocodyliform 177–182 dicynodont 34–38 sauropod 195– 203 turtle 154– 168 Song Da Suture 46, 51 Song Ma Suture 15, 51 Spathobatis [elasmobranch] 107, 109 Srisuk’s House Museum 127 subduction 18 Sukhothai Island Arc 8, 9, 13, 15 Surveying and exploring in Siam 26 sutures of South East Asia 9, 10, 43, 45– 46 systematic palaeontology see palaeontology, systematic Tad Huai Nam Yai, footprint site 246, 247 taphocoenosis 288 taphonomic study, fish 115 methods 116– 118 morphometric study 118–123 preservation state 117, 120 results 119– 123 taphonomy, turtles 148, 153– 154 taphonomy, sauropods 193–194 Taquet, P 33 taxonomic analysis 287–292 Tebak Formation 48

306 tectonic model, Khorat basin 51– 52 tectonic plates, South East Asia 8 teeth 78, 131, 153 elasmobranch 98– 110 and oxygen isotope composition 273, 275 sauropod 203, 211 theropod 193, 194 Tembeling Group 51 temnospondyls 125 temperature 278 Terrain Rouge 47, 48, 49, 79 terranes 45–46 South East Asia 9, 10, 13 terranes, origin of 7– 14 Testudinoidea 165, 166, 167, 168 Tethys, palaeogeography 16– 20 Tha Song Khon, footprint site 247, 248 Thai vocabulary 190, 191 thermal sag basin 43, 51, 52 theropod footprint 248, 249 Theropoda 229– 241 limb bones 233–237 pelvis 231–233 tooth 193, 194 vertebrae 230–231 thin-skinned thrust 46 Thung Yai Group 97 Titanosauria 194– 214 bone histology 217–226 tooth see teeth trackway, Cretaceous 256–267 trackway, Jurassic 255– 256, 262, 265, 267 trackways 218 British Columbian specimens 259, 260, 263, 264, 265, 266 Spanish specimens 259, 261, 262, 263, 265 Thai specimens 257–259, 263 Zimbabwian specimens 261– 262, 263 Trang Group 48, 86, 97, 125 palynology 79–80 tree shape (Mesozoic) 85–94 tree, size analysis 86–92 branch 88, 90– 92 diameter 87–88 height 88 tridactyl tracks 258 trilobites and Deprat Affair 30 Trionychoidae 3 limb bones 164

INDEX pelvis 164 shell 155, 158, 159–164 skull 156 systematic palaeontology 154–168 turtle assemblages, Khorat Group 141–149, 153–171 Cretaceous 143– 146 Jurassic 136–143 Jurassic– Cretaceous 153– 168 taphonomy 153–154 turtle fauna 2 –3, 99, 125 China 142– 143, 144, 145, 148 Japan 145, 146, 147 Laos 146, 148 Mongolia 142, 145, 148 Russia 145 unconformity Jurassic 52 Mid-Cretaceous 46 Upper Nam Phong Formation 44, 46, 47, 49, 53, 82 vertebrae sauropod 203–208, 209– 213 theropod 230– 231 vertebrate 1, 18, 53, 69–70 distribution 2, 16– 17 footprints 245– 253 vertebrate assemblages, biogeographical analysis data protocol 287 method 287–290 palaeogeographic interpretation 290, 294– 298 taxonomic analysis 291– 294 Vientiane Basin 48, 50, 53 volcanism, Khorat Group 63 Wan Din So, fish locality 129, 131 Wat Sak Kawan, sauropod locality 189, 191 West Burma Block 12, 51 West Sumatra Block 12 wood 47, 86, 87–92 xylological analysis 87 Yunnan Basin 60 zircon 63