ECOLOGY AND EXTINCTION OF SOUTHEAST ASIA’S MEGAFAUNA

JULIEN LOUYS

Thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy in the School of Biological, Earth and Environmental Sciences University of New South Wales Sydney, Australia

December 2007

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Louys

First name: Julien Other name/s: Claude Alexandre

Abbreviation for degree as given in the University calendar: PhD

School: Biological, Earth and Environmental Sciences Faculty: Science

Title: Ecology and Extinction of Southeast Asia’s Megafauna

Abstract 350 words maximum: (PLEASE TYPE)

The Quaternary megafauna of Southeast Asia are among the world’s poorest known. Throughout the Pleistocene, continental collisions, active volcanic systems and fluctuations in sea level have had dramatic effects on the region’s geography, from southern China to Indonesia. Many Southeast Asian megafauna experienced geographical range reduction or complete extinction during that interval. This thesis explores the relative influence of environmental change and human interaction in these extinctions. There is currently no direct evidence to suggest that humans had a negative impact on Southeast Asian megafauna until the Holocene. Rather, extinctions and geographical range reduction experienced by megafauna are likely to have resulted from of loss of suitable habitats, in particular the loss of more open habitats. Environmental change throughout the Pleistocene of Southeast Asia is reconstructed on the basis of discriminant functions analysis of megafauna from twenty-seven Southeast Asian Quaternary sites, as well as Gongwangling, an early Pleistocene hominin site previously interpreted as paleoarctic. The discriminant functions were defined on the basis of species lists drawn from modern Asian nature reserves and national parks, and were analysed using both taxonomic and phylogeny-free variables. Biases present in these species lists were mitigated against using a range of mulitvariate techniques. The reconstructions show that Pleistocene environments in Southeast Asia varied from open (e.g. savannah), mixed (woodland) and closed (e.g. rainforest) habitats. Changes in habitats through time are likely to have been driven, at least in part, by changes in sea-level, in turn related to oscillations between glacial and interglacial conditions. The environmental changes associated with these oscillations are likely to have adversely affected many of Southeast Asia’s megafauna. The Toba super-eruption (~74kya) is unlikely to have been responsible for any of the megafauna extinctions of the Late Pleistocene.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

…………………………………………………………… ……………………………………..……………… ……….……………………...…….… Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

ii

ORIGINALITY STATEMENT ‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………………………......

Date ……………………………………………......

iii

To my parents Pierre and Chantal Louys who provided boundless inspiration, moral and financial support

iv Acknowledgments

This thesis could not have been completed without the support and encouragement of many people. I would firstly like to thank my supervisors Mike Archer and Darren

Curnoe, and my co-supervisor Sue Hand, who gave me the complete freedom to pursue whatever topic interested me, and to construct this thesis on my own terms. Many thanks to them for helpful discussions, critical readings and support provided from initial conception to finished product. A special thanks to Darren, who initially took me on, supported my move back to BEES, and helped finance my data collection trip.

Many people have provided support, inspiration, helpful comments, reviews and discussions on the various ideas, papers and chapters produced throughout this thesis, including anonymous reviewers, Mina Bassarova, Robin Beck, Pip Brewer, Andy Coate,

Darren Curnoe, Henk Godthelp, Yamila Gurovich, Sue Hand, Andy Herries, Scott

Hocknull, Pierre Louys, Geoff Mulhearn, Gilbert Price, Jim Rose, Rachael Tomlins,

Haowen Tong, Kenny Travouillon, John de Vos, Vera Weisbecker, Jolee West and Steve

Wroe. Many thanks to Guillaume Louys and Kenny Travouillon for providing much needed assistance in generating the various figures; and to Diane Louys for her boundless enthusiasm in tackling all the tedious formatting chores I put to her.

I thank the staff at the Museum of Natural History, Leiden (Naturalis), in particular John de Vos and Reiner van Zelst, for kindly hosting me, and making my trip to Holland so enjoyable. I also thank the staff at the Institute of Vertebrate Palaeontology and

v Palaeoanthropology, Beijing for their kind hospitality, in particular Haowen Tong for his welcome and constant support during my time in China.

The time spent working on this project would have been much duller had it not been for the students and friends in both the Biological Anthropology (Andy Coate, Hayley Green,

Andy Herries, Diana Neuweger) and Vertebrate Palaeontology (Rick Arena, Mina

Bassarova, Hayley Bates, Robin Beck, Karen Black, Pip Brewer, Kirstin Crosby, Anna

Gillespie, Yamila Gurovich, Zac Kirkham, Karen Marino, Jackie Nguyen, Liz Price,

Karen Roberts, Kenny Travouillon, Vera Weisbecker, Steve Wroe) Labs, in particular

Robin and Kenny. Thanks guys.

And finally, I wish to thank my family; my brothers Guillaume and Greg, my sister

Diane, my sister-in-law Trinette and my parents Pierre and Chantal. If it wasn’t for their constant financial, moral, and emotional support throughout my many years of study, none of this would have been possible.

vi Table of Contents

1. Introduction 1 1.1 Background 1 1.2 Aims 5 1.3 Specific objectives 5 1.4 Chapter outline 6 1.5 A note on the nature of the chapters 8

Part 1. Extinction 10

2. Characteristics of Pleistocene megafauna Extinctions in Southeast Asia 11 2.1 Abstract 11 2.2 Introduction 12 2.3 Methodology 13 2.4 Results and Discussion 15 2.4.1 Indochinese province 16 2.4.2 Sundaic Province 29 2.4.3 Taxa 37 2.4.4 Influence of humans in the region 46 2.4.5 Climate and vegetation 52 2.5 Conclusions 54

3. Southeast Asian Quaternary Extinctions: Position in the Debate and Further Analysis 56 3.1 Introduction 56 3.2 The Quaternary “megafauna” extinctions 57 3.2.1 What are megafauna? 57 3.2.2 The debate 58 3.2.3 Human-induced extinctions 59 3.2.4 Climate-induced extinctions 62 3.2.5 Towards a reconciliation 64 3.3 Quaternary extinctions in Southeast Asia 64 3.3.1 Geography of Southeast Asia 66 3.3.2 Southeast Asia’s megafauna 74 3.3.3 Human overhunting in Southeast Asia? 81 3.3.4 Climate change and megafauna 83 3.3.5 The modern extinction crisis 87 3.4 Summary 89

4. The Role of Volcanism in Extinctions: the Case of the Toba Eruption 90 4.1 Abstract 90 4.2. Introduction 90 4.3 Methods 92 4.4 Results 98

vii 4.4.1 Borneo 98 4.4.2 Vietnam 99 4.4.3 Sumatra 99 4.4.4 Java 100 4.4.5 Southern China 101 4.5 Discussion 102 4.5.1. General discussion 102 4.5.2 Volcanic winter 104 4.5.3 No volcanic winter 107 4.6 Conclusions 109

Part 2. Ecology 110

5. Palaeoecology Multivariate Methods: Examining Modern Habitat Biases and Application to Gongwangling Hominin Site, China 111 5.1 Abstract 111 5.2 Introduction 112 5.2.1 Geological context 115 5.2.2 Previous palaeoenvironmental analyses 115 5.3 Materials and Methods 116 5.3.1 Materials 116 5.3.2 Taphonomy 123 5.3.3 Zoogeographic identity 124 5.3.4 Palaeoecology 125 5.4 Results and Discussion 137 5.4.1 Taphonomy 137 5.4.2 Zoogeography 146 5.4.3 Palaeoecology 152 5.5 Summary 165

6. Palaeoecology of Quaternary Southeast Asian Sites Revealed Through Phylogeny-Free Discriminant Functions Analysis 167 6.1 Abstract 167 6.2 Introduction 168 6.3 Methods 171 6.4 Results 176 6.5 Discussion 191 6.5.1 Early Pleistocene sites 193 6.5.2 Middle Pleistocene sites 194 6.5.3 Late Pleistocene and Holocene sites 196 6.6 Conclusion 198

7. Summary 200 7.1 Indochinese Province 200 7.1.1 Southern China 200

viii 7.1.2 Burma 202 7.1.3 Laos 203 7.1.4 Vietnam 204 7.1.5 Cambodia 207 7.1.6 Thailand 208 7.1.7 Indochinese summary 210 7.2 Sundaic Province 211 7.2.1 Malaysia 211 7.2.2 Borneo 211 7.2.3 Sumatra 212 7.2.4 Java 212 7.2.5 Sundaic summary 215 7.3 Conclusions 216

8. Conclusions 218 8.1 Major implications of this research 218 8.2 Concluding remarks 219

References 221

Appendix A 263 Appendix B 281 Appendix C 291 Appendix D 298 Appendix E 311

ix List of Figures

Figure 1.1 Southeast Asia 2 Figure 2.1 Generic level extinctions in the Early Pleistocene 37 Figure 2.2 Generic level extinctions in the Middle Pleistocene 38 Figure 2.3 Generic level extinctions in the Late Pleistocene 39 Figure 3.1 Southeast Asia at the LGM 67 Figure 3.2 Global sea-level for the last 130 ka 72 Figure 3.3 Global sea-level for the last 1.2 Ma 84 Figure 4.1 Map showing the approximate locations of fossil sites preserving an extinction signature for Toba 98 Figure 5.1 Location of Gongwangling with respect to the Qinling Mountains 113 Figure 5.2 Locations of the nature reserves and national parks used in this study 126 Figure 5.3 Pie graph showing minimum number of individuals (MNI) of large-bodied species from GWL 138 Figure 5.4 Proportion of non-dental minimum number of elements (MNE) 139 Figure 5.5 V5419 Cervus grayi antler showing extensive rodent gnaw marks 142 Figure 5.6 Box plot showing the relative abundance of modern faunas within the groups defined by Soligo and Andrews (2005) 144 Figure 5.7 Confidence funnels for taxonomic distinctness of the nature reserves and national parks listed in Table 5.2 154 Figure 5.8 Dendogram showing the clustering of the reserves and parks based on families present 156 Figure 5.9 Scatter plot of the principle coordinates analysis 158 Figure 5.10 Graphical representation of the canonical discriminant functions showing the clustering of the different reserves and parks 163 Figure 6.1 Location of Pleistocene Southeast Asian sites considered in this chapter 174 Figure 6.2 PCA of modern reserves and parks using ecological variables 177 Figure 6.3 Loadings for the first two PCA components 178 Figure 6.4 Box plots of modern and Pleistocene distribution of ecological variables 185 Figure 6.5 DFA showing classification of Southeast Asian Pleistocene sites 186

x List of Tables

Table 2.1 Mean body mass of the five orders examined 14 Table 2.2 Extinct species for Southern China 16 Table 2.3 Extinct species for Burma 21 Table 2.4 Extinct species for Laos 23 Table 2.5 Extinct species for Vietnam 26 Table 2.6 Extinct species for Cambodia 27 Table 2.7 Extinct species for Thailand 28 Table 2.8 Extinct species for Malaysia 30 Table 2.9 Extinct species for Borneo 31 Table 2.10 Extinct species for Sumatra 32 Table 2.11 Extinct species for Java 33 Table 2.12 List of species where extinctions range over more than one country 35 Table 2.13 Landmass area for islands within the range of this study 36 Table 3.1 Species becoming locally extinct in more than one Southeast Asian country 66 Table 4.1 List of sites discussed in text, their respective dates and corresponding references 93 Table 4.2 Species list of sites discussed in text 94 Table 5.1 List of species and minimum number of individuals (MNI) present at GWL 118 Table 5.2 List of nature reserves and national parks used in this study with a summary of the published vegetation type 127 Table 5.3 Percentage of each of the large-bodied mammalian families present in the nature reserves and national parks 133 Table 5.4 Percentage of the large-bodied families present at GWL 136 Table 5.5 Minimum Number of Elements (MNE)/Number of Identified Specimens (NISP) for large- and small-bodied mammals 139 Table 5.6 List of surface modifications present in sample studied 142 Table 5.7 Zoogeography of the species present at GWL 148 Table 5.8 Faunal list for Yangguo (derived from Ji (1975), with revisions by Qiu et al. (2004)) and Chenjiawo (derived from Chow (1964) and Chow and Li (1965)) 151 Table 5.9 Summary of DFA results for the three habitat types 162 Table 5.10 DFA classification results 163 Table 6.1 Ecological variables used in discriminant functions analysis, and examples of mammals from Southeast Asia falling within those categories 172 Table 6.2 Relative proportions of species in each of the ecological categories for the park and reserve used in Chapter 5 179 Table 6.3 Proportion of species in each ecological category for 30 Southeast Asian sites 181 Table 6.4 PCA loading for the first two components 183 Table 6.5: Dice similarity index between closed, mixed and open habitats 183

xi Table 6.6: Minimum numbers of species necessary to confidently differentiate between the major habitat types 183 Table 6.7: Palaeoecological classification of Pleistocene sites on the basis of the DFA 187

xii CHAPTER 1

INTRODUCTION

1.1 Background

Southeast Asia is one of the richest regions on Earth in terms of its biodiversity. It stretches from southern China to Indonesia, comprising both continental and island areas

(Figure 1.1). Ecologically it has been, until modern deforestation, a land of forests

(Corlett, 2005), in particular rainforests. It hosts some of the world’s most fascinating, but little studied megafauna. Among the better known examples are the orangutan, the

Asian elephant and the tiger. It has also hosted throughout the Pleistocene a wide variety of large-bodied mammals that are now extinct. Among the most fascinating are the giant ape, Gigantopithecus, whose closest extant relative is the orangutan, the giant tapir, the giant pangolin, various elephant-like stegodons, and the giant hyena. The presence of these large mammals in Pleistocene Southeast Asia is consistent with almost every other continent, where there existed during the Pleistocene a variety of large-bodied mammals, the so-called “megafauna”. The nature of the disappearance of these megafauna has been the subject of intense debate (see, for example, the various papers in Martin and Klein,

1984; Reed et al., 2005). Many researchers advocate the direct influence of humans in the extinctions (e.g. Martin, 1984; Miller et al., 2005), while others advocate a more influential role for climate (e.g. Guthrie, 1984; Graham, 2006). Debate on the causes of extinction has, however, almost entirely ignored the megafauna from Southeast Asia, largely because of the poor chronology of many of the taxa and sites in the region

1 (Martin, 1984). Nevertheless, Southeast Asia’s Pleistocene megafauna are of unique value to this debate. Unlike the megafauna of North America, Europe, Australia and

South America, they have had contact with hominins for one of the longest periods for any continent, save of course Africa. Like Africa, many Southeast Asian megafauna are

still extant. However, unlike for Africa, hominins are not endemic to Southeast Asia, and

as such megafauna there have had only limited time to evolve defences against humans.

China Burma

Vietnam

Laos Thailand Cambodia Malaysia

Borneo Sumatra

Java

Figure 1.1: Southeast Asia showing the countries and sites examined in this thesis. Not to scale

2 Hominins most likely colonised Southeast Asia sometime during the Early

Pleistocene. Hominin artifacts from Nihewan, northern China, have been dated at 1.66

Ma (Zhu et al., 2004), while dental remains of early Homo from northern China have been dated to possibly as old as 1.96 Ma (Huang et al., 1995). Early to Middle

Pleistocene hominins are most common in Indonesia (e.g. Trinil, Ngandong, Flores), however, the dating and nature of these deposits remains controversial (see Yokoyama and Falgueres, 2002, Swisher et al., 1996, van den Bergh et al., 2001, Westaway et al.,

2007). Archaic Homo sapiens are probably less than 400 ka old in China (Wu & Poirier,

1995; Chen et al., 1994; Rink et al., 2003). Modern H. sapiens have been dated at 42-39 ka in northern China (Shang et al., 2007), 45-39 ka in Borneo (Barker et al., 2007), and possibly 70 ka in Australia (Thorne et al., 1999) suggesting a widespread distribution of modern H. sapiens by the Late Pleistocene. Potential overlap between various species of hominin in Southeast Asia (e.g. H. erectus, H. floresiensis, H. sapiens) has yet to be conclusively shown, and remains much debated.

Southeast Asia is also of unique value for examining the effects of environmental change on mammalian evolution and extinction. The region is the result of a dynamic geological regime, with continental collisions, active volcanic systems and changes in sea-level having a dramatic effect on its geography throughout the Pleistocene.

Alternating glacial/interglacial cycles have continuously exposed and submerged the low- lying Sunda shelf. Periods of exposure have connected otherwise isolated islands, allowing active migration of many species. The last period of low sea levels also brought about dramatic changes in vegetation, including the development of savannahs in the centre of Southeast Asia (e.g. Heaney, 1991; Bird et al., 2005). Continental collisions

3 have caused the uplift of the Tibetan plateau, which set about the beginning of the Asian

monsoon system (Ferguson, 1993). An active volcanic regime has produced many

islands, among the largest including Java. It also produced the Quaternary’s largest

volcanic eruption, the Toba super-eruption. This super-volcano erupted approximately 74

kya, and has been suggested as responsible for a global human population bottleneck (e.g.

Ambrose, 1998, 2003).

Since Dubois went in search of the “missing link” at the close of the 19th century, the Pleistocene mammals of the region have been the subject of intermittent study.

Taxonomically, biostratigraphically and palaeoenvironmentally, the faunas of Java and

Southern China have received attention at a level starting to approach those of Africa and

Europe. These attentions can be fairly confidently linked to the discovery of hominins, in

particular Homo erectus, from these regions. Other geographical areas in Southeast Asia have received comparatively little attention. Despite some new areas of research in

Southeast Asia, other major scientific fields remain little examined. Examples of these

include dating, taphonomic analyses, and palaeoenvironmental reconstructions of

Pleistocene sites, as well as detailed study of the functional morphology and systematics

of extinct taxa. In many cases, research has been hampered by historical, political or

economic constraints. In addition, until recently much palaeontological research in the

region has been published in obscure or regional journals and books, further impeding

study of the region’s prehistory, as well limiting international attention.

This is problematic because Southeast Asia is of direct relevance to a number of

scientific disciplines of international importance. The significance of Southeast Asia in

understanding early hominin evolution and migration is one that has long been

4 recognised (e.g. Dubois, 1891). In addition, research on the biogeography of the region is

paramount in understanding the process of evolution itself – the long recognised biogeographical divide between east and west Indonesia (Müller, 1839; Wallace, 1869), precipitated many ideas on evolution, zoogeography and ecology.

Sadly, the continued existence of many species in Southeast Asia is currently under threat (Sodhi et al., 2004). While an understanding of the ecology and biology of extant mammals is paramount in any attempt to prevent further extinctions, so too is an understanding of the past factors that may have acted detrimentally to the survival of mammals in the region. In particular, it is important to be able to understand the ecology of past Southeast Asian communities, and the factors which may have contributed to their extinction. Today, East Asia is home to half of the world’s population. Thus the region is of critical importance to matters of conservation and to human long term sustainability during an era of global climate change and widespread poverty.

1.2 Aims

This thesis aims to examine the extinction and/or geographical range reduction of

Pleistocene megafauna in Southeast Asia. The extinction of megafauna in Southeast Asia will be examined primarily in the framework of the global megafaunal, or Quaternary, extinction event (see Koch and Barnosky, 2006; Martin and Klein, 1984). In particular, the roles of humans and especially climate change will be examined for their possible effects on large-bodied mammalian communities. Additional factors falling outside the range of the traditional megafauna debate, but of unique relevance to Southeast Asia will also be examined.

5 The palaeoenvironments represented by particular sites in Southeast Asia will be reconstructed. These sites span the Early Pleistocene to Holocene, and from China to

Java. They are reconstructed on the basis of modern Asian faunal communities, using

multivariate palaeoecology methods. In many cases, these analyses represent the first

palaeoenvironmental reconstructions of any type to be completed for those countries. In sites where reconstructions of palaeoenvironments have previously been made, the methods described in this thesis represent the first quantitative analyses to be completed.

On the basis of the results arrived at in this thesis, changes in environments over temporal

and geographical dimensions can be mapped for the Quaternary of Southeast Asia, and its

impacts on survival of megafauna assessed.

1.3 Specific objectives

This thesis has four primary objectives:

1. To document which megafauna from Southeast Asia experienced extinction or

geographical range reduction during the Pleistocene.

2. To evaluate the potential roles of humans and climate in these extinctions, thereby

addressing the Southeast Asian megafaunal extinctions within the context of the

global megafaunal extinction event.

3. To evaluate the potential effects of the Toba super-eruption on megafauna.

4. To examine the ecology of Southeast Asia’s megafauna, in particular of

individual extinct species, and examine the habitats megafaunal communities

occupied during the Pleistocene.

6 1.4 Chapter outline

Chapter 1 details the background, aims and specific objectives of this thesis, as well as

providing an overview of the chapters.

Chapter 2 provides an introduction to megafauna extinctions in Southeast Asia. It details

the extinct megafauna recovered from a number of sites in Southeast Asia, as well as

briefly reviewing the sites themselves. It examines previous extinction hypotheses

proposed to explain the disappearance and range reduction of certain taxa in the region.

The arrival of humans in the region, and their possible impacts are explored. The nature,

and temporal and geographic extent, of environmental change in the region is also

introduced. This chapter defines the genera and species of megafauna that disappeared over more than one region in Southeast Asia, and provides a minimum estimate of extinction. This chapter has been published (Louys et al., 2007).

Chapter 3 examines in more detail the nature of the global megafauna debate, and reviews Southeast Asia’s position within that debate; in particular, it examines the specific role of climate change in extinction. The environmental and geographical characteristics of Southeast Asia are detailed. What is known of the ecology of those extinct and extant megafauna defined in Chapter 1 is reviewed. General ecological traits are examined, and their relationship to climatic changes occurring through the Quaternary are explored. The modern extinction crisis is introduced. This chapter has been published

(Louys, 2008).

7 Chapter 4 examines the specific role of the super-eruption of Toba on the megafauna. In

particular, this chapter examines whether or not any extinctions can be tied with to the

hypothetical environmental changes associated with this eruption. The implications of

this analysis for the theory of a human population bottleneck, as associated with Toba,

are examined. In particular, two scenarios are examined, depending on whether or not the

effects of the eruption can be likened to a nuclear winter scenario. This chapter has been

published (Louys, 2007).

Chapter 5 introduces the concept of multivariate palaeoecological analyses for Asian

Quaternary faunas. A database of mammal species from various parks and reserves from

Asia is built. Limitations associated with using these sorts of data are detailed, and ways

of mitigating them are explored and implemented. The database is subsequently used as a

basis for the palaeoecological reconstruction of the early Pleistocene fossil site of

Gongwangling, in Lantian county, China. This site is examined in detail. A brief

examination of its geology, taphonomy and zoogeography is made, in order to assess

potential biases in the deposit. This is an important site because it represents the most

northerly extent of the Southeast Asian biogeographic zone during the Early Pleistocene.

Chapter 6 examines the palaeoenvironment of particular Southeast Asian sites.

Limitations of data from Quaternary Southeast Asian sites are assessed, in particular the

heavy bias towards large mammals, and the limited accessibility of materials from some sites. These limitations are addressed through the development of phylogeny-free

variables that can be calculated from species lists alone. These variables are used to

8 reconstruct the palaeohabitats of 27 fossil sites. The results of these analyses are

compared with previous palaeoenvironmental reconstructions for the region.

Chapter 7 provides a summary of the extinctions and environmental changes occurring in

Southeast Asia throughout the Pleistocene. It reviews whether palaeoenvironmental

changes can be tied to the extinctions of megafauna.

Chapter 8 lists the major implications arising from this thesis, and proposes areas of

further research.

1.5 A note on the nature of the chapters

Most of the chapters comprising this thesis were submitted and/or published

during a period of three and a half years. Information in some of the earlier written

chapters has been superseded or has been subsequently shown to be erroneous. The previously published chapters are, however, included as printed, the only changes being

of a formatting (and titling) nature. Any developments or corrections to earlier chapters

are detailed in subsequent chapters. Furthermore, because these chapters were written as

independent publications, common elements may be found within them, for example, in

the introductions.

Some of the publications herein represent multi-authored works. In these cases,

the contributions by the various authors are indicated at the beginning of each respective

chapter. I am the senior author of all chapters included in the body of this thesis.

9

PART 1

EXTINCTION

It is looking into the details of an extinction that you discover the really useful information that seems to apply to our own predicament as a species with suicidal tendencies.

- Robert Twigger 2001

CHAPTER 2

CHARACTERISTICS OF PLEISTOCENE MEGAFAUNA

EXTINCTIONS IN SOUTHEAST ASIA1

2.1 Abstract

The extinction of large-bodied taxa from the Pleistocene in Southeast Asia is examined.

Although the chronological resolution of these extinctions is poor, and number of excavations in the region relatively few, broad characteristics of these extinctions can be described. Many taxa which became extinct appear to have been endemic to regions within Southeast Asia, while some taxa which experienced extinction or severe range reduction occurred in several regions. Members of the latter group include proboscideans

(Stegodon and Palaeoloxodon), the pygmy hippopotamus (Hexaprotodon), the orangutan

(Pongo), hyenas (Crocuta and Hyaena), the giant panda (Ailuropoda), tapirs (Tapirus and

Megatapirus), rhinoceroses (Rhinoceros), and the giant Asian ape Gigantopithecus. The loss of these species cannot be assigned to a single cause. Rather their disappearance is likely tied to both climatic and human agents. Unlike other regions which experienced megafauna extinctions, eustatic changes in sea level in Southeast Asia seems to have been an important factor.

1 This chapter is published as: Louys, J., Curnoe, D., Tong., H. 2007. Characteristics of Pleistocene megafauna extinctions in Southeast Asia. Palaeogeogr. Palaeoclimat. Palaeoecol. 243, 152-173. DC and I contributed equally to the section entitled “Influence of humans in the region”. HT verified the species lists for Southern China, as well as providing general comments. The remainder of this chapter is my own work.

11 2.2 Introduction

The extinction of large-bodied species in the Late Pleistocene has been observed on every continent save Antarctica. The causes of extinctions have traditionally been divided into two camps – human agents or climatic agents, although increasingly researchers are opting for a multi-agent (including humans and climate) causative model (e.g. Barnosky et al., 2004; Wroe et al., 2004). Despite an ever growing body of work concerned with megafaunal extinctions in many parts of the world, the history of their disappearances remain virtually unstudied in several geographic regions. In particular, the extinction of megafauna in Southeast Asia and South America has received little scrutiny. In the case of Southeast Asia, this has largely been the result of a poor chronology for these extinctions (Martin, 1984). However, poor chronology does not prevent discussion of the extinction. While several studies have examined extinctions in Southeast Asia (e.g.

Medway, 1972, 1977; Sondaar, 1987; Tougard et al., 1996; Cranbrook, 2000), these have been restricted to particular taxonomic or geographic groups. This region provides an interesting case study whereby several possible mechanisms of extinction can be examined.

It has long been thought that extinctions in Africa were less severe than in other regions of the world due to the long term coevolution of humans and megafauna

(Barnosky et al., 2004, Martin 1966). In Southeast Asia however, this period of coevolution did not occur, although the region has experienced the greatest duration of hominin habitation outside of Africa. If anything, the extinction record of Southeast Asia should parallel Europe, where it is commonly assumed that hunting for over 400,000 years of megafauna was not detrimental to their survival, and that it was only the

12 emergence of behaviourally modern humans which led to acceleration of extinctions

(Barnosky et al., 2004). Unlike Europe, but like Africa, Southeast Asia continues to host a number of extremely large-bodied species, including the Asian elephant and the rhinoceros.

Southeast Asia also holds special significance for the extinctions in Greater

Australia (Sahul). The hypothesis that extinctions occurred soon after the first arrival of humans in Sahul is commonly accepted (e.g. Miller et al., 1999; Roberts et al., 2001;

Miller et al., 2005), and is dependent on the assumption that the first Australians either hunted the megafauna or created significant habitat alteration resulting in their extinction.

We argue that in order to understand the role of humans on the Australian ecosystem, it is necessary to also examine the effect and ecology of their ancestors – in this case,

Southeast Asians.

2.3 Methodology

We decided to restrict this analysis to five orders of (mostly) large bodied taxa (Table

2.1) owing to an often arbitrary and subjective application of the term “megafauna” (see

Marshall, 1984; Wroe et al., 2004), combined with a scarcity of papers dealing with body mass estimates for extinct species in Southeast Asia. As mammals within the same order often exhibit a similar range of size variation (Smith et al., 2004), it can be safely assumed that those extinct species examined will not exhibit radically different body proportions to those of their most closely related living forms. The orders studied are

Primates, Carnivora, Proboscidea, Artiodactyla and Perissodactyla. Only the latter three orders have a mean body mass greater than the average definition of megafauna (>44kg as per Martin (1984); Table 2.1). However, it was felt that an examination of Carnivora

13 would be prudent for two reasons: firstly, any loss of herbivores from the overall biomass would undoubtedly affect carnivores, regardless of body mass. Secondly, it is generally recognised that large body mass in mammals is a response to the need to digest low quality vegetation (Owen-Smith, 1992). It is axiomatic that these principles will not affect the Carnivora (with the notable exception of the giant panda), and that (terrestrial) carnivores are unlikely to achieve the body masses of their herbivorous prey. With respect to the primates, although they generally have a smaller body mass, this is not true of the apes: their mass falls within the common definition of “megafauna”.

Order Mean Body Mass (grams) Artiodactyla 47,863 Carnivora 3,548 Primates 1,778 Perissodactyla 398,107 Proboscidea 5,128,614

Table 2.1: Mean body mass of the five orders examined. Adapted from Smith et al.(2004) and Smith et al. (2003).

Extinction lists were compiled from published literature (Tables 2.2-2.11), by comparing Pleistocene faunal assemblages with extant taxa (compiled from Corbet and

Hill, 1992; Nowak, 1999). Taxonomic nomenclature follows Nowak (1999) for extant species, McKenna and Bell (1997) for extinct genera and Corbet and Hill (1992) for extant genera but extinct species, except for Proboscidean taxonomy, which follows

Shoshani and Tassy (2005). The subdivision of the Pleistocene into Early (2.5 Mya to

780 kya), Middle (780 kya to 128 kya) and Late (128 kya to 11 kya) follows Jablonski and Whitford (1999). Extinction lists are sorted according to country, period and site

(Tables 2.2-2.11). In the following analysis, any species designated “c.f.” is treated as the

14 conferred species; a species designated “sp.” is ignored if another species of the same genus is present in the site, as these taxa may be conspecific. Our analysis is confined to the species and genus levels, with subspecies ignored. Taxa unassigned above genus level have not been included. Thus, our analysis provides a minimum estimate of extinction.

Southeast Asia as discussed in this paper is defined as the area of land south of the

Yangtze river in China, and west of Huxley’s line (which runs between Bali and Lombok in the south, between Borneo and Sulawesi, and west of the Philippines; Figures 2.1-2.3).

Biogeographically, Southeast Asia can be split into two distinct provinces – the

Indochinese Province (consisting of southern China, Burma/Myanmar, Laos, Cambodia,

Vietnam and northern Thailand) and the Sundaic Province (consisting of southern

Thailand, Malaysia, Sumatra, Java and Borneo) on the basis of climatic, botanical and zoological factors (Tougard 2001, Lekugal and McNeely 1988, Gray et al. 1994). Each province was separated into their individual countries (Thailand was grouped within the

Indochinese province as all sites examined from this country are situated in northern

Thailand) and taxa were examined independently for each. Countries share a number of extinctions or range reductions of the same genera (Table 2.12). Among these taxa, some are discussed in detail in the text. These include Stegodon, Hexaprotodon,

Palaeoloxodon, Pongo, Crocuta, Hyaena, Ailuropoda, Rhinoceros, Tapirus, Megatapirus and Gigantopithecus.

2.4 Results and Discussion

The extinction list for each country is provided in Tables 2.2-2.11. Generic-level extinctions for the Early, Middle and Late Pleistocene are shown in Figures 2.1, 2.2 and

2.3 respectively.

15 2.4.1 Indochinese Province

Southern China

The extinction list for Southern China (Table 2.2) is the most extensive for the region, largely because of the larger number of excavations that have taken place there (Han and

Xu, 1985). The extinction list and the chronology of the sites are based on Han and Xu

(1985) and Kahlke (1961), and as a discussion of each site would unnecessarily encumber this manuscript, the interested reader is directed to these works for more information. The relatively large sample should allow for detection of diachronic trends not discernible within other (smaller) regional samples. It should be noted however that this list is by no means exhaustive.

Early Pleistocene Gigantopithecus Cave †Gigantopithecus blacki †Hesperotherium “praesinensis”a *Pongo sp. *Tapirus †peii Cuon †dubius Rhinoceros †”chiai”a Arctonyx †minor †Dicoryphochoerus ultimus Ailuropoda †microta Sus †xiaozhu *Hyaena †brevirostris Sus †peii Felis †teilhardi †Dorcabune liuchengensis *Acinonyx †pleistocaenicus Muntiacus †lacustris †Gomphotherium serridentoides †Cervavitus fenqii †Stegodon preorientalis Cervus †yunnanensis *Equus †yunnanensis †Megalovis guangxiensis

Yuanmou Canis †yuanmoensis Muntiacus †nanus Vulpes cf. †chikushenensis Axis †shansius *Hyaena †brevirostris Axis cf. †rugosa †Megantereon meganteron Cervus †yunnanensis †Stegodon elephantoides Cervus †stehlini †Stegodon zhaotongensis †Procapreolus stenos †Stegodon yuanmoensis *Gazella sp. *Equus †yunnanensis †Eostyloceros longchuanensis †Hesperotherium sp. †Metacervulus capreolinus Rhinoceros †sinensis †Paracervulus attenuatus Muntiacus †lacustris † Cervavitus ultimus

Jianshi

16 †Gigantopithecus blacki *Hyaena †brevirostris †Stegodon sp. *Equus †yunnanensis *Tapirus †indicus

Bijiashan *Hyaena †brevirostris †Gomphotherium serridentoides †Stegodon cf. preorientalis †Stegodon orientalis *Tapirus †indicus Rhinoceros †sinensis Sus †xiaozhu Sus †bijashanensis

Middle Pleistocene Daxin Hoshantung Hsingan Koloshan †Gigantopithecus blacki *Pongo pygmaeus Macaca †robustus *Pongo pygmaeus *Pongo sp. *Crocuta crocuta Cuon †simplicidens *Crocuta crocuta †Stegodon orientalis Felis *lynx Martes †sinensis †Palaeoloxodon sp. Rhinoceros †sinensis †Palaeoloxodon †Melodon simplicidens †Stegodon orientalis †Megatapirus augustus namadicus Rhinoceros †sinensis Rhinoceros †sinensis †Dicoryphochoerus †Megatapirus augustus †Megatapirus augustus *Tapirus †indicus ultimus Muntiacus †Megalovis †szechuanensis guangxiensis Bubalus †brevicornis Sus †bijashanensis *Tapirus indicus Wuming †Gigantopithecus blacki *Hyaena sp. †Stegodon sp. Rhinoceros †sinensis Naemorhedos cf. †sumatraensis Bama †Gigantopithecus blacki *Pongo sp. Rhinoceros †sinensis *Tapirus sp. †Stegodon sp.

Guanyindong †Stegodon cf. orientalis *Crocuta crocuta †Stegodon guizhouensis †Megatapirus augustus Rhinoceros †sinensis Naemorhedos †sumatraensis Vulpes cf. †vulgaris *Equus sp. Shilontou

17 *Hyaena †brevirostris †Stegodon orientalis Rhinoceros †sinensis

Late Pleistocene Maba *Hyaena sp. †Stegodon sp. †Palaeoloxodon namadicus *Tapirus sp.

Tongzi *Pongo sp. *Crocuta crocuta †Megatapirus augustus Rhinoceros †sinensis †Stegodon orientalis *Tapirus sp.

Changyang *Hyaena †brevirostris †Megatapirus augustus Rhinoceros †sinensis †Stegodon orientalis

Jiande *Crocuta crocutab †Megatapirus augustus Rhinoceros †sinensis †Stegodon orientalis †Palaeoloxodon cf. namadicus *Ovis sp.

Liujiang †Megatapirus augustusb Rhinoceros †sinensis †Stegodon orientalisb a Not formally described b Survived into the Holocene (Tong and Liu, 2004).

Table 2.2: Extinct species for Southern China; sites ordered into Early, Middle and Late Pleistocene, listed chronologically (oldest to youngest; for full discussion of chronology see text); * indicates local extinction; † indicates global extinction; bold indicates last appearance. For the full list of species from each site see Han and Xu (1985), except Hoshantung, Hsingan and Koloshan see Kahlke (1961).

The southern Chinese and Burmese extinction lists are the only ones to show progressively fewer extinctions through the Pleistocene (Figures 2.1-2.3), suggesting that

Early Pleistocene extinctions were more significant for the northern parts of Southeast

18 Asia. Many of the species that became extinct during the Early Pleistocene in southern

China appear to be the last surviving members of lineages that had been in decline since the Miocene (e.g. Hesperotherium). Of interest during this period is the extinction of

Acinonyx and Gazella, which are currently restricted to Africa (Nowak, 1999). In addition to these, most species of Stegodon became extinct during the Early Pleistocene.

The Middle Pleistocene has far fewer generic extinctions. This period saw the extinction of the last members of Equus and Gigantopithecus in the region. Northern

Chinese hominins specialised in predation on Equus until at least Middle Pleistocene times (Keates, 2003), and evidence suggests that early settlers roasted horse heads at

Zhoukoudian (Binford and Stone, 1986). Equus is also one genus thought to have been hunted to extinction by early Clovis hunters in North America (Martin, 1984), as it coincides with the colonisation of the Americas by humans (Barnosky et al., 2004). The colonisation of China by humans may have also affected the horse. Equally the extinction of the horse may mark a change in foraging strategy by humans in the region. This change may also have adversely affected the distribution of Gigantopithecus (but see below). The Late Pleistocene saw the disappearance of many species that seem to characterise the Southeast Asian Pleistocene extinctions at an interregional level (see below), however, some species that became extinct in the Late Pleistocene in other regions appear to have survived into the Holocene in China (Ma and Tang, 1992; Tong and Liu, 2004). These include Crocuta ultima, Stegodon orientalis and Megatapirus augustus (see Tong and Liu (2004) for a discussion on sites and dates).

19 Burma/Myanmar

Deposits in Burma are described in detail by Colbert (1943), and the extinction list provided in Table 2.3 is drawn from his species lists. Colbert (1943) describes three principal faunas, those being from Terrace deposits, Mogok Caves and the Irrawaddy beds. The Terrace deposits are considered by Colbert (1943) to be Upper Pleistocene in age and consist solely of extant fauna, they are therefore not discussed further here.

The Mogok caves are fissures found in the Mogok district, and are interpreted as

Middle Pleistocene by Colbert on the basis of the presence of Stegodon, although Hooijer

(1950) considers them ‘Lower’ Pleistocene. Colbert provides details of two caves, however of the extinct species listed in Table 2.3, Colbert mentions finding only the proboscideans in situ. The presence of Ailuropoda is based on collection by others

(notably villagers) not from either cave described, but rather from nearby caves. However given the ubiquitous nature of this species during the Middle Pleistocene in Southeast

Asia, its presence in Burma during this time is warranted.

20

Early Pleistocene Irrawaddy beds †Stegodon elephantoides †Stegalophodon latidens †Palaeoloxodon namadicus Rhinoceros †sivalensis *Equus †yunnanensis †Hipparion cf. antelopinum *Potamochoerus sp. *Hexaprotodon †iravaticus *Gazella sp. †Hemibos triquetricornis †Proleptobos birmanicus †Merycopotamus dissimilis

Middle Pleistocene Mogok Caves *Ailuropoda melanoleuca †Stegodon orientalis †Palaeoloxodon namadicus

Table 2.3: Extinct species for Burma; sites ordered into Early and Middle Pleistocene, listed chronologically (oldest to youngest; for full discussion of chronology see text); * indicates local extinction; † indicates global extinction. For the full list of species from each site see Colbert (1943).

The Irrawaddy beds represent both rewashed and in situ fossils, which likely span the ages of Late Pliocene to Early Pleistocene (Colbert 1943). They are likely derived from the Upper Irrawadies, and share many species in common with the Siwalik faunas of India (Colbert 1943). The Mogok faunas represent more eastern Asian characteristics

(Colbert 1943), although Hooijer (1950) argues that the fauna agrees with that of the

Tatrot zone of the Siwaliks. If Colbert’s hypothesis is borne out, then these two deposits record a faunal turnover between the Lower and Middle Pleistocene.

Laos

The extinction list for Laos (Table 2.4) comes from deposits found in the “Chaîne

Annamitique septentrionale” (northern Annamese Cordillera), described by Fromaget

21 (1936) and Arambourg and Fromaget (1938). Although these authors mention two

“sites”, Tam Hang (Fromaget 1936) and Tam Nang (Arambourg and Fromaget 1938), the specific locations of these sites are not provided.

The site of Tam Hang as recorded by Fromaget (1936) is presumably the same as that dated to the Late Pleistocene by Shackelford (2003). Fromaget (1936) does not record the specific fauna recovered from this site, remarking only on its similarity to that of other Pleistocene Southeast Asian faunas including those of Trinil and Zhoukoudian.

He also records the recovery of further Pleistocene mammals from a silt and gravel layer twelve kilometres to the north of Tam Hang, at 1000m altitude. It is from this horizon that the faunal list comes (although this is not entirely clear), and Fromaget notes that the two faunas are the same. To our knowledge, the fauna from Tam Hang has not been further investigated, and among the countries studied here, Laos remains one of the most starved of palaeontological research. Based on the presence of Pongo and Megatapirus within the faunal list, it is likely that this fauna belongs to the Middle Pleistocene.

Tougard (1998) in her biochronological scheme places Tam Hang in the middle of the

Middle Pleistocene, and tentatively dates this period as between 170 ka and 475 ka

(however this determination appears to be based on a combined fauna between Tam Hang and Tam Nang, discussed in more detail below). If this is the case, either Fromaget and

Shackelford’s “Tam Hang” represent separate deposits within the same site, or the two faunas (“Tam Hang” proper, and the site north of Tam Hang) described by Fromaget

(1936) aren’t contemporaneous (unlikely if his contention that the faunas are the same is borne out). The fauna is therefore interpreted here as Middle Pleistocene.

22

Middle Pleistocene Tam Nang, tuff layer, lower horizon Tam Hang Sus cf. †brachygnathus Macaca †robustus Naemorhedus cf. *goral *Pongo pygmaeus Cervus cf. †leptodus Ursus cf. †angustidens †Proboselaphus sp. †Megatapirus augustus Canis cf. †cynoides Rhinoceros cf. †plicidens *Crocuta crocuta Rhinoceros cf. †sivalensis Felis †micotis Sus cf. †lydekkeri *Pongo pygmaeus Cervus †grayi *Ailuropoda melanoleuca

Tam Nang, tuff layer, upper horizon

Ursus †premalayanus †Stegodon sinensis †Palaeoloxodon namadicus Rhinoceros †sivalensis †Megatapirus augustus Sus †lydekkeri Arctonyx †rostratus Panthera *leo Felis aff. †issidiorensis ?Bibos †geron ?Bubalus †teilhardi Cervus †orientalis Macaca cf. †andersonni Macaca †robustus

Tam Nang, red layer *Tapirus sp. †Spiroceros sp. Ursus †premalayanus Macaca †robustus

Table 2.4: Extinct species for Laos; sites listed chronologically (oldest to youngest; for full discussion of chronology see text); bold indicates last appearance; * indicates local extinction; † indicates global extinction. For the full list of species from: Tam Hang see Fromaget (1936); Tam Nang see Arambourg and Fromaget (1938).

The second site examined, that of Tam Nang (Arambourg and Fromaget 1938), is often overlooked or combined with that of Tam Hang when faunal lists of Laos are compiled. We have found no evidence to suggest that Tam Hang and Tam Nang are the same site, and proceed on the understanding that they represent separate sites. The fauna recovered from Tam Nang comes from two Quaternary layers, a red silt layer which sits

23 unconformably above a calcareous tuff divided into two “horizons” (an upper and a lower horizon). The extinction list is separated in accordance with these layers and horizons

(Table 2.4). Arambourg and Fromaget (1938) compare the fauna from these deposits with those of Pleistocene sites in Java and China. They assign an age of “Pleistocene inférieur” to the calcareous layers, and that of “Quaternaire moyen ou supérieur” for the red silt layer. We interpret the calcarious layers and the red silt layers as Middle Pleistocene, due to the presence of typical Middle Pleistocene fauna within the list (such as Ailuropoda,

Megatapirus, Stegodon, Pongo, Macaca robustus and Palaeoloxodon) although we acknowledge the tenuousness of such dating. We further interpret the red silt layers as being younger than the calcareous ones based on their relative stratigraphic positions as described by Arambourg and Fromaget (1938).

Of interest in this extinction list is the presence of Panthera leo, the Asiatic lion.

If the specific diagnosis made by Arambourg and Fromaget is borne out, this would represent the most eastern example of this species; its current range in Asia restricted to western India, although subfossil species have been found as far east as Bengal (Corbet and Hill 1992). In addition to taxa mentioned above, Laos during this time saw the disappearance of two species each of Rhinoceros, Macaca, Ursus and Sus.

Vietnam

The extinction list for Vietnam (Table 2.5) is compiled on the basis of three Middle

Pleistocene and four Late Pleistocene sites. Thum Khuyen and Tham Hai were roughly estimated at between 300 ka and 200 ka on the basis of biostratigraphic correlations

(Olsen and Ciochon 1990); however subsequent absolute dating assigned an age of 475 ka for Tham Khuyen on the basis of ESR dating of tooth enamel (Ciochon et al. 1996).

24 Again on the basis of faunal correlations dates of 140 – 80 ka and 30 – 20 ka were suggested for Hang Hum and Keo Leng respectively (Olsen and Ciochon 1990). Dating of Ma U’Oi is even more tenuous, being tentatively assigned to the Late Pleistocene by

Bacon et al. (2004) on the basis of it similarity with other Late Pleistocene faunas of

Southeast Asia. With regards to palaeoenvironments of the various sites, Bacon et al.

(2004) suggest that the absence of Pongo from Ma U’Oi indicates an open woodland environment, while the presence of Niviventer fulvescens and Leopoldamys sabinus, two

Murid species, suggests the presence of lowlands and foothills of evergreen forests. The presence of Gigantopithecus and Pongo in Tham Khuyen and Tham Hai, and the latter also in Hang Hum and Keo Leng suggests the presence of tropical to sub-tropical vegetation, while the presence of Ailuropoda suggests temperate bamboo species.

25

Middle Pleistocene Tham Khuyen Tham Hai †Gigantopithecus blacki † ?Gigantopithecus blacki *Pongo pygmaeus *Pongo pygmaeus † ?Palaeoloxodon namadicus †Stegodon orientalis †Stegodon orientalis Rhinoceros †sinensis Rhinoceros †sinsensis Sus cf. †lydekkeri *Ailuropoda melanoleuca Ursus *malayanus †Megatapirus augustus Tham Om †Palaeoloxodon cf. namadicus †Stegodon orientalis Rhinoceros †sinensis *Elaphodus sp. *Ailuropoda melanoleuca Bos (Bibos) *gaurus †Megatapirus augustus †Naemorhedus sumatraensis Late Pleistocene Hang Hum I Hang Hum II Ma U’Oi †Stegodon orientalis *Pongo pygmaeus Rhinoceros. cf.* Rhinoceros †sinensis †Palaeoloxodon cf. namadicus unicornis †Megatapirus augustus Sus cf. †lydekkeri Sus cf. ?officinalis Keo Leng *Pongo pygmaeusa †Stegodon orientalis *Ailuropoda melanoleuca Rhinoceros †sinensis †Megatapirus augustus Sus cf. †lydekkeri †Naemorhedus sumatraensis

Table 2.5: Extinct species for Vietnam; sites ordered into Early, Middle and Late Pleistocene, listed chronologically (oldest to youngest; for full discussion of chronology see text); bold indicates last appearance; * indicates local extinction; † indicates global extinction. For the full list of species from: Tham Khuyen, Tham Hai, Tham Om, Hang Hum I, Hang Hum II and Keo Leng see Olsen and Ciochon (1990); Ma U’Oi cave see Bacon et al. (2004).

Vietnam and Southern China are, at present, the only two countries where

Gigantopithecus fossils have been found (Tables 2.2 and 2.5). This ape became extinct during the Middle Pleistocene (Figure 2.2). As with China, this could mark the time of the first occurrence of Homo sapiens, or alternatively, a change in foraging strategies of

26 humans in the region. Apart from Pongo, Ailuropoda, Stegodon, Palaeoloxodon and

Megatapirus (discussed in more detail below), Vietnam saw little in the way of additional generic extinction.

Cambodia

Cambodia shows relatively few extinctions on the basis of the sites examined here. Due to the difficulty in locating references for Cambodia, the list provided by Tougard (2001) is the primary reference used, however this is supplemented by the species list provided by Carbonnel and Guth (1968). Although both sites are from Phnom Loang, it is unlikely

(on the basis of the faunas recorded) that these sites are the same. In Table 2.6, we have followed the naming conventions used by both authors: i.e. Phnom Loang for those species listed by Tougard (2001) and Préah Vihear for those listed by Carbonnel and

Guth (1968). Phnom Loang is described as Middle Pleistocene by Tougard (2001), while

Carbonnel and Guth (1968) attribute a lower to middle Pleistocene age on the basis of the presence of Dicerorhinus mercki and Crocuta crocuta. The site is interpreted as Middle

Pleistocene here. Among the extinction list, Cambodia sees the loss of Pongo, Crocuta,

Tapirus, Dicerorhinus mercki and Cervus grayi, disappearances characteristic of

Southeast Asia.

Middle Pleistocene Phnom Loang Préah Vihear *Pongo pygmaeus *Crocuta crocuta *Crocuta crocuta Cervus †grayi *Tapirus indicus *Dicerorhinus †mercki

Table 2.6: Extinct species for Cambodia; * indicates local extinction; † indicates global extinction. For the full list of species from: Phnom Loang see Tougard (2001); Préah Vihear see Carbonnel and Guth (1968).

27 Thailand

The extinction lists for Thailand (Table 2.7) comes from four sites: Kao Pah Nam (Pope et al. 1981), the Cave of the Monk of Ban Fa Suai (Zeiton et al. 2005), Thum Wiman

Nakin and Thum Phra Khai Phet (Tougard 1998), all being Middle Pleistocene in age.

The only site to be given an absolute age is that of Thum Wiman Nakin. Calcified clay sediments from above the fossiliferous layer have been dated to 169 ka ± 11 ka, thereby providing an absolute minimum age to the fossils (Esposito et al. 2002). Thum Phra Khai

Phet is thought to be of a comparable age based on taxonomic similarity (Tougard 1998).

Zeiton et al. (2005) consider the Cave of the Monk to be Middle to late Middle

Pleistocene based on the presence of Stegodon-Ailuropoda fauna at the site, although they acknowledge the tenuousness of such dating. The Kao Pah Nam site is estimated at 690 ka based on its geological position and fauna (Pope et al. 1981).

Middle Pleistocene Kao Pah Nam Cave of the Monk (Ban Fa Suai) * ?Pongo sp. *Pongo cf. pygmaeus *Crocuta sp. *Ailuropoda cf. melanoleuca Cervus †eldii Cervus cf. †eldii ? Nemorhaedus c.f. *goral Thum Wiman Nakin Thum Phra Khai Phet Nemorhaedus c.f. *caudatus *Pongo pygmaeus *Ailuropoda melanoleuca †Stegodon sp. *Ailuropoda melanoleuca *Crocuta crocuta *Crocuta crocuta Sus *barbatus Rhinoceros *unicornis Sus *barbatus Cervus *eldii

Table 2.7: Extinct species for Thailand; sites listed chronologically (oldest to youngest; for full discussion of chronology see text); bold indicates last appearance; * indicates local extinction; † indicates global extinction. For the full list of species from: Thum Wiman Nakin and Thum Phra Khai Phet see Tougard (1998); Kao Pah Nam see Pope et al. (1981); Cave of the Monk see Zeiton et al. (2005).

28 Thailand saw the disappearance of Ailuropoda melanoleuca, Crocuta crocuta,

Pongo pygmaeus, Sus cf. barbatus, Rhinoceros cf. unicornis, Cervus eldii, Nemorhaedus c.f. goral, Nemorhaedus c.f. caudatus, Tapirus sp. and Stegodon sp. during the Middle

Pleistocene (Table 2.7). Tougard (1998) briefly addressed the extinction question for

Thailand however her examination was restricted to only four species – Ailuropoda melanoleuca, Crocuta crocuta, Pongo pygmaeus and Rhinoceros cf. unicornis. She advocated a multi-causal mode of extinction, with Ailropoda and Crocuta affected by changes in climate and vegetation, while Pongo and Rhinoceros were argued to have been adversely affected by the negative ecological influence of humans.

Palaeoenvironmental analysis of the region based on the mammals from Thum Wiman

Nakin suggest that the region was composed of slightly open forested landscape with relatively humid conditions, representing a more temperate and open environment than presently found (Tougard and Montuire 2006).

2.4.2 Sundaic Province

Malaysia

The Malaysian fossil record is poorly known for the Pleistocene and the extinction list provided (Table 2.8) is based on only one site, Tambun, being Middle Pleistocene in age

(Medway 1972). Besides Palaeoloxodon, Malaysia saw the local extinction of

Hexaprotodon and Duboisa during this time.

29

Middle Pleistocene Tambun †Palaeoloxodon namadicus *Hexaprotodon sp. *Duboisia santeng

Table 2.8: Extinct species for Malaysia; * indicates local extinction; † indicates global extinction. For the full list of species see Medway (1972).

Borneo/ Kalimantan

The late Pleistocene of Borneo saw the local extinction of four genera, Cuon, Panthera,

Tapirus and Rhinoceros, based on the faunal records of Niah Caves (Table 2.9), dated on the basis of 14C at 40 ka ± 1 ka (Medway 1964, Harrison 1996, Barker, 2002). Cranbrook addressed the question of megafauna extinctions in Borneo in a number of publications

(Medway, 1972, 1977; Cranbrook, 2000). He noted that the only animal to become extinct from the Niah caves faunas was the giant pangolin, Manis palaeojavanica. He also noted the disappearance of Hexaprotodon and proboscideans from Borneo

(Cranbrook et al., 2000), however their specific provenance could not be determined (and hence they are not listed in Table 2.9). Given that the majority of these were grazers or browsers dependent on short vegetation, he concluded their extinctions were largely the result of changes in the vegetative mosaic of the region. Medway (1972) postulated that the loss of most grazing animals by the close of the Late Pleistocene were tied to the loss of large tracts of grasslands, while Holocene changes in vegetation were detrimental to forest dwelling species such as Tapirus and Pongo. These floristic changes are thought to result from global climatic changes acting regionally (Medway, 1972, 1977; Cranbrook,

30 2000), in particular the rising mean ambient temperatures and an increasing, less seasonal rainfall (Medway, 1972). The change from a more complex and heterogenous vegetation structure present in the Pleistocene to a more uniform rainforest environment today may have resulted in many taxa becoming extinct. This is likely to have happened during the transition to the last glacial (see below). Interestingly, this causative model is also posited for a number of other megafauna extinctions in both North America and Australia (e.g.

Guthrie, 1984; Lundelius, 1989; Graham et al., 1996; Price and Sobbe, 2005).

Late Pleistocene Niah Caves *Cuon alpinusa *Panthera tigris *Tapirus indicusb *Rhinoceros sondaicusb aMay in fact be Canis familiaris (Cranbrook, 2000) bMay have survived until Holocene (Medway 1960, Cranbrook 2000)

Table 2.9: Extinct species for Borneo; * indicates local extinction. For the full list of species see Medway (1960).

Sumatra

The extinction list for Sumatra (Table 2.10) comes from one site, Lida Ajer, being Late

Pleistocene in age (de Vos 1983). All but one species (Panthera pardus) are currently extant on the island. Meijaard (2004) suggested that P. pardus could not have survived on

Sumatra due to the limited ungulate biomass and carnivore competition once Sumatra became isolated from other Southeast Asian islands from the last glacial maximum onwards.

31

Late Pleistocene Lida Ajer Panthera *pardus

Table 2.10: Extinct species for Sumatra; * indicates local extinction. For the full list of species see de Vos (1983).

Java

The palaeontological record for Java is, along with that of Southern China, the best known for the region. Extinction lists (Table 2.11) were complied from the following faunal units (listed in chronological order): Satir, Ci Saat, Trinil H.K., Kedung Brubus,

Ngangdong, Punung and Wajak (Leinders et al. 1985, Theunissen at al. 1990, van den

Bergh et al. 2001). The oldest faunas, those of Satir and Ci Saat, have been dated to 1.5 and 1.2 Ma respectively (de Vos and Long, 2001, de Vos et al. 1994). The Satir fauna is thought to represent a swampy environment based on pollen records, while that of Ci Saat represents a drier environment with more grasses (Semah 1984, de Vos and Long, 2001).

A faunal turnover is evident from these sites, with the more primitive Hexaprotodon simplex and Tetralophodon bumiajuensis replaced by the more derived Hexaproton sivalensis and Stegodon trigonocephalus respectively (de Vos et al. 1994). These replacements are likely tied to the environmental change recorded by the pollen records, and/or through competition with invading species (de Vos et al. 1994). Satir and Ci Saat are relatively poor in species, which indicates predominately isolated island conditions for Java during this time period (de Vos et al. 1994, de Vos and Long, 2001, van den

Bergh et al. 2001). This relative isolation is thought to prevail until 0.8 Ma, where a large immigration event is evident in the Kedung Brubus fauna (van den Bergh et al. 2001).

32 Early Pleistocene Satir †Sinomastodon bumiajuensis *Hexaprotodon †sivalensis

Ci Saat †Stegodon trigonocephalus *Hexaprotodon †sivalensis Axis †lydekkeri †Duboisia santeng Bubalus †palaeokerabau Bos (Bibos) †palaesondaicus

Trinil H.K. †Stegodon trigonocephalus Axis †lydekkeri †Duboisia santeng Bubalus †palaeokerabau Bos (Bibos) †palaesondaicus Sus †brachgnathus Cuon †(Mececyon) trinilensis

Middle Pleistocene Kedung Brubus *Hyaena †brevirostris †Stegodon trigonocephalus †Stegodon? Hypsilophus Elephas †hysudrindicus Rhinoceros *unicornis *Hexaprotodon †sivalensis Axis †lydekkeri †Duboisia santeng †Epileptobos groeneveldtii Bubalus †palaeokerabau Bos (Bibos) †palaesondaicus Sus †macrognathus

Late Pleistocene Ngangdong †Stegodon trigonocephalus Elephas †hysudrindicus *Hexaprotodon †sivalensis Bubalus †palaeokerabau Bos (Bibos) †palaesondaicus ?Sus †brachgnathus Sus †macrognathus

Punung Macaca *nemestrina Hylobates *syndactylus *Pongo pygmaeus

33 *Ursus malayanus *Helarctos malayanus Sus *barbatus *Naemorhedus (Capricornis) sumatraensis

Table 2.11: Extinct species for Java; sites ordered into Early, Middle and Late Pleistocene, listed chronologically (oldest to youngest; for full discussion of chronology see text); bold indicates last appearance; * indicates local extinction; † indicates global extinction. For the full list of species from: Satir, Ci Saat, Trinil H.K., Kedung Brubus and Ngangdong see van den Bergh et al (2001); Punung see de Vos (1983).

Trinil H.K., Kedung Brubus and Ngangdong are thought to represent an open woodland habitat (de Vos 1983, 1995, Sondaar 1984, van den Bergh et al. 2001), as indicated by the large number of bovids present, as well as the lack of typically tropical rainforest species such as Pongo, Hylobates and Helacrtos (van den Bergh et al. 2001,

Storm 2001). The later Punung fauna, however, presents more modern, humid rainforest characteristics. The faunal turnover recorded between Kedung Brubus and Punung indicates the disappearance of 17 taxa, and the appearance of ten new species (van den

Bergh et al. 2001). This turnover is thought to have occurred between 110 ka and 70 ka, when the sea levels were sufficiently low enough to allow passage of fauna from the mainland (van den Bergh et al. 2001). Among those disappearing during this time are

Stegodon trigonocephalus, Hexaprotodon sivalensis, Elephas hysudrindus and Sus macrognathus; disappearances which are likely tied to both environmental changes as well as competition with new taxa which had migrated southwards. The extinction of

Pongo and other species such as Hylobates syndactylus, Macaca nemestrina and Sus barbatus from Java most likely occurred during the Late Pleistocene, as these species are absent in the Wajak fauna (10,560 ka) (Storm et al. 2005). These local extinctions may be

34 a result of the drier and cooler climate in Java present after 81 ka (van der Kaars and

Dam, 1995), and the resultant loss of rainforest habitat (Storm et al., 2005).

Species of Number Countries China South Burma Laos Cambodia Vietnam Thailand Malaysia Java Borneo Duboisia santeng 2 X X Equus yunnanensis 2 X X Gazella sp. 2 X X Gigantopithecus blacki* 2 X X Macaca robustus 2 X X Hyaena brevirostris* 2 X X Rhinoceros sinensis* 2 X X Rhinoceros sivalensis* 2 X X Sus barbatus 2 X X Sus lydekkeri 2 X X Stegodon elephantoides* 2 X X Megatapirus augustus* 3 X X X Rhinoceros unicornis* 3 X X X Stegodon orientalis* 3 X X X Tapirus indicus* 3 X X X Naemorhedus sumatraensis 3 X X X Ailuropoda melanoleuca* 4 X X X X Crocuta crocuta* 4 X X X X Palaeoloxodon namadicus* 5 X X X X X Pongo pygmaeus* 6 X X X X X X

Table 2.12: List of species where extinctions range over more than one country; * indicates taxa discussed in detail in the text.

Of the countries studied here, only Sumatra, Borneo and Java are islands. Humans have been implicated in the extinctions of island endemics in Southeast Asia (Sondaar, 1987).

However, the factors involving extinctions on continents are unlikely to be as simple as extinctions occurring on islands (Wroe et al., 2004), and as such are unlikely to apply for much of the region. Furthermore, only one island within this study (Java) is smaller than

35 the largest island where overkill has been relatively universally accepted, the south island of New Zealand (Wroe et al., 2004) (Table 2.13). Java was connected to Sumatra and

Borneo during periods of low sea level (Voris, 2000), making it easier for human settlement. If overkill did occur on these islands, then it is more likely to have taken place when island extinction dynamics, as opposed to continental factors, were in play – the period when sea levels were high. If this was the situation, then humans could be a causative agent in extinctions there. However, extinctions would not ensue without being preceded by sea level changes, which also brought about changes in vegetation, as well as allowing the immigration of new species in the region.

Island Landmass (km2) Borneo 744 000 Sumatra 474 000 Java 132 000 South Island, New Zealand 150 000

Table 2.13. Landmass areas for islands within the range of this study. Only Java is smaller than the largest island where overkill is generally accepted, the South Island of New Zealand (Wroe et al., 2004). However, during the periods of low sea levels, these islands were joined (see text for discussion).

36 2.4.3 Taxa

*Acinonyx †Gomphotherium *Tapirus †Dicoryphochoerus †Dorcabune †Cervavitus †Megalovis †Stegodon †Megantereon †Stegalophodon †Hesperotherium *Equus †Procapreolus †Hipparion *Gazella *Potamochoerus. †Eostyloceros *Hexaprotodon †Metacervulus *Gazella †Paracervulus †Hemibos *Equus †Proleptobos †Merycopotamus

†Sinomastodon

Huxley’s Line

Figure 2.1 Generic level extinctions in the Early Pleistocene. † - complete extinction; * - local extinction. For complete extinction list and a break down of sites refer to Tables 2.2-2.11.

37 †Dicoryphochoerus †Megalovis †Melodon †Gigantopithecus

†Gigantopithecus *Elaphodus

*Ailuropoda †Palaeoloxodon †Proboselaphus †Stegodon *Crocuta *Pongo *Ailuropoda *Pongo †Stegodon *Ailuropoda †Palaeoloxodon *Crocuta †Megatapirus †Stegodon *Tapirus. * ?Tapirus *Pongo †Spiroceros *Tapirus *Crocuta *Dicerorhinus †Palaeoloxodon *Hyaena *Hexaprotodon †Stegodon *Duboisia †Duboisia Huxley’s Line †Epileptobos

Figure 2.2 Generic level extinctions in the Middle Pleistocene. † - complete extinction; * - local extinction. For complete extinction list and a break down of sites refer to Tables 2.2-2.11.

38 *Pongo *Hyaena †Palaeoloxodon *Tapirus *Ovis

*Pongo *Ailuropoda †Stegodon †Palaeoloxodon †Megatapirus †Naemorhedus

†Stegodon *Hexaprotodon *Pongo *Cuon *Panthera *Ursus *Helarctos *Tapirus *Naemorhedus *Rhinoceros

Huxley’s Line

Figure 2.3 Generic level extinctions in the Late Pleistocene. † - complete extinction; * - local extinction. For complete extinction list and a break down of sites refer to Tables 2.2-2.11.

Stegodon and Palaeoloxodon

Stegodon remains are found in Late Pleistocene deposits in Burma, Laos, Borneo and

Java, and survived into the Holocene in Southern China (Tong and Patou-Mathis, 2003;

Tong and Liu, 2004; Ma and Tang, 1992) (Figures 2.1 and 2.2). Although absent from other faunal lists, given the wide chronological and geographical distribution of

Stegodon, it is reasonable to assume that this is a result of limited field work as opposed to real absence. This abundance and distribution has led to the “Stegodon-Ailuropoda faunas”, which are often cited as characteristic of the Middle Pleistocene of Southeast

Asia. However, no study has thus far examined the extinction of this genus across its

39 entire range. Sondaar (1987) postulated that the arrival of early humans on islands in

Southeast Asia (specifically Flores) led to the extinction of the pygmy stegodon,

Stegodon sondaari. Stegodon florensis, however, appears to have had a long-term coexistence with Homo in Flores (Morwood et al., 2004). The more likely explanation is that Stegodon was adversely affected by the southward latitudinal shifts of the tropical and subtropical zones, decreases in the areal extent of these zones (Jablonski and

Whitford, 1999) and the corresponding reduction in heterogeneity of vegetation during the Pleistocene. This corresponds to the gradual reduction in the number of species of

Stegodon evident from the Early through to Late Pleistocene, especially when the southern Chinese data are examined. Although the impact of humans on these taxa cannot be discounted, this would probably have been more pronounced on islands rather than on the continent.

Palaeoloxodon is likely to have shared a similar fate to that of Stegodon. Much more closely related to Elephas than Stegodon, during the Pleistocene (in China at least)

Palaeoloxodon exhibited a similar distribution to that of Elephas, but became extinct before the Holocene (Tong and Patou-Mathis, 2003). It is found until the Middle

Pleistocene in Burma, Laos and Malaysia (Figure 2.2), and until the Late Pleistocene in

Southern China and Vietnam (Figure 2.3). Why these two proboscideans became extinct in the region, while Elephas continued to survive to today, is not satisfactorily explained by human- or climate-mediated models of extinction; it may perhaps be tied to subtle differences in ecology and body size. Stegodon abundances appear to be greater than for those of Elephas in Late Pleistocene Southern China, providing some support to a generalist vs. specialist hypothesis (Saegusa, 2001). Without an in-depth study of the

40 ecology and behaviour of these extinct taxa, however, it would be unwise to pronounce anything more than tentative suggestions.

Pongo and Gigantopithecus

Pongo is currently confined to Sumatra and Borneo, however, Pleistocene fossil specimens are known from Southern China, Cambodia, Thailand, Vietnam, Laos and

Java (Table 2.12, Figures 2.2 and 2.3). It is known only from Middle Pleistocene deposits of Laos and Cambodia (Figure 2.2), but this is likely to be a reflection of the small number of excavations rather than actual absence. It is also possible that Pongo survived into the Holocene in Vietnam (Kahlke, 1972). Pleistocene Pongo was larger than the modern orangutan (Nowak, 1999), a similar condition to that in a number of other

Pleistocene Southeast Asian taxa (Medway, 1972; Pope, 1995; Tougard, 2001; Brown et al., 2004). A reduction in body size is another characteristic of the megafauna extinctions on other continents (e.g. North America, Guthrie, 2003). It has been suggested that during the Pleistocene, the orangutan was group living and terrestrial, rather than solitary and arboreal, as seen today (MacKinnon, 1971). The hypothetical change from terrestriality to arboreality has been linked to competition with an invading terrestrial ape, humans, and thought to include hunting pressure for at least 35,000 years (MacKinnon, 1971).

However, it is more likely that the orangutan’s solitary and arboreal habitats were present from before the incursion of humans, given its arboreal limb adaptations and frugivorous diet (Galdikas, 1988). The negative influence of humans, through hunting pressure or habitat alteration, which has been observed in historical times (Rijksen, 1978), should not be discounted when looking at the Pleistocene range reduction of this species.

41 Gigantopithecus is the only extinct primate genus known from the Pleistocene in

Southeast Asia, known only from teeth and mandibular fragments. Based on comparisons of these with those of Gorilla, its hypothesised long bones would be 20-25% longer and more robust than the extant ape’s (Johnson, 1979). It shows a general increase in size throughout the Pleistocene (Ciochon et al., 1996). Although it is currently known only from deposits in China and Vietnam (Figure 2.2), the limited number of excavations for the region in general suggest that it may have had a much wider geographical distribution. Faunas associated with Gigantopithecus suggest it inhabited a tropical or subtropical forest environment (Kahlke, 1984; Ciochon et al., 1990), and it has been reconstructed as subsisting on a variety of grasses and fruits (Ciochon et al., 1990).

Both Pongo and Gigantopithecus exhibit range reductions, and extinction in the case of the latter, throughout the Pleistocene (Jablonski and Whitford, 1999; Jablonski et al., 2000). The southward latitudinal shifts of the tropical and subtropical vegetation zones coupled with a decrease in area of these zones are the likely causes of these reductions (Jablonski and Whitford, 1999; Jablonski et al., 2000). Other catarrhine species with different life history parameters were less adversely affected (Jablonski et al., 2000). Although the influence of hunting and possible competition with humans is not discounted, these factors are, we suggest, unlikely to have been the sole contributing factor to the range reductions or extinctions experienced by these apes.

Crocuta and Hyaena

While we understand that taxonomically, Hyaena from Southern China has recently been replaced by Pachycrocuta, and Crocuta crocuta with Crocuta ultima, for the sake of consistency with older works the older names are utilised here. Crocuta is known until

42 the Middle Pleistocene in Cambodia, Thailand and Laos (Figure 2.2), and survived into the Holocene in China (Tong and Liu, 2004). Hyaena is known until the Middle

Pleistocene in Java and Late Pleistocene in Southern China (Figures 2.2 and 2.3). Kurtén

(1956) suggested the extinction of Hyaena brevirostris in China was the result of competition with the better adapted Crocuta crocuta, a point based on their respective stratigraphic positions. This hypothesis is supported by Tougard (1998), who noted that

Crocuta was found in more recent deposits in Southern China, Laos and Cambodia, and tied its dominance over Hyaena to climatic cooling which occurred during the middle

Pleistocene (480-440 ka). What caused the eventual extinction of Crocuta? Perhaps the introduction of the domestic dog commensurate with modern humans led to competitive exclusion. Or, perhaps the increase in rainforest during the Late Pleistocene adversely affected Crocuta, as its preferred habitat of open areas (Nowak, 1999) gave way to more enclosed forest. The possibility also exists that a combination of these two factors ultimately led to the disappearance of Crocuta from Southeast Asia.

Ailuropoda

The widespread distribution of Ailuropoda during the Pleistocene is well documented, exemplified by its pairing to Stegodon to form the characteristic elements of Middle

Pleistocene fauna in Southeast Asia. Tougard et al. (1996) specifically examined the disappearance of the giant panda from Southeast Asia. They attribute reduction in range of Ailuropoda to changes in bamboo distribution during the Pleistocene, a phenomenon intrinsically tied to changes in climate. They postulated that during the Pleistocene, temperate bamboo species, the food of Ailuropoda, were widespread in Thailand, and presumably wherever else Ailuropoda occurred. However, climatic conditions became

43 unfavourable for temperate bamboo in the latest Middle Pleistocene and it subsequently retreated northwards, with Ailuropoda following its food until it reached its present range, restricted to the Southwest of China (Nowak, 1999). Ailuropoda is known until the

Middle Pleistocene in Burma, Thailand and Laos, and Late Pleistocene in Vietnam (Table

2.12; Figures 2.2 and 2.3).

Tapirus and Megatapirus

Tapirus is currently restricted to Burma, Thailand, Malaysia, Sumatra and possibly Laos.

It is adapted for locomotion in thick undergrowth, with its habitat in nearly any wooded or grassed area near water (Nowak, 1999). It is known until the Middle Pleistocene in

Cambodia (Figure 2.2), and until the Late Pleistocene in Southern China and Borneo

(Figure 2.3), where it could have survived until at least 8 Ka (Medway, 1960).

Megatapirus is much larger than the current tapir – its teeth and skull are one-quarter larger than Tapirus indicus (Tong, 2005). Megatapirus derived in the later part of the

Early Pleistocene, and is known from Middle and Late Pleistocene deposits in Laos and

Vietnam, respectively, as well as from two Holocene sites in China, Chongqing (4235 ybp) and Zheijiang (7815 ± 385 ybp) (Tong and Liu, 2004).

Hexaprotodon

The hippopotamus is known until the Early Pleistocene in Burma, Middle Pleistocene in

Malaysia and Late Pleistocene in Java (Figures 2.1, 2.2 and 2.3). It is also represented by a single fragment from Borneo (Cranbrook et al., 2000), however the provenance of this find (and hence its age) has not been determined. Due to its critical dependence on open water sources, Jablonski (2004) has tied the extinction of Hexaprotodon in Southeast

44 Asia to a drying up of river and lake systems, a result of increasing seasonality of rainfall and river flow. These changes are also associated with the intensification of summer and winter monsoons. It is highly likely that the isolation of many islands, and the ensuing disruption to river systems resulting from changes in sea level (Voris, 2000), contributed equally to the extinction of this genus.

Rhinoceros and Dicerorhinus

Rhinoceros is represented by at least five species during the Pleistocene: R. sondaicus, R. sivalensis, R. unicornis, R. plicideus, R. sinensis, as well as a sixth (referred to as “R. chaii”, though yet to be formally described). Dicerorhinus is represented by two species:

D. sumatrensis and D. merki. Of all these rhinocerotids, only R. sondaicus, R. unicornis and D. sumatrensis are not extinct, although all three are highly endangered. Rhinoceros unicornis still survives in Assam and Nepal, while R. sondaicus can be found in Java,

Borneo, Malaysia, Burma and Indochina (Groves and Guérin, 1980). Dicerorhinus survives today in Sumatra, Borneo, Malaysia, Burma and Indochina (Nowak, 1999).

Rhinoceros unicornis is able to graze, while R. sondaicus and D. sumatrensis are browsers, inhabiting swamps and dense forests (Prothero et al. 1989).

Rhinoceros sondaicus probably became extinct in Borneo in the Late Pleistocene

(however it may have survived into the Holocene; see Medway 1960, Cranbrook 2000);

R. sivalensis became extinct in the Early Pleistocene in Burma and Middle Pleistocene in

Laos; R. unicornis disappeared from Java and Thailand in the Middle Pleistocene, as well as the Late Pleistocene of Vietnam; R. plicideus became extinct in the Middle Pleistocene of Laos; and R. sinensis disappeared in the Late Pleistocene of Southern China and

Vietnam. Dicerorhinus mercki is last recorded from the Middle Pleistocene of Cambodia.

45 Along with Stegodon, Palaeoloxodon, Tapirus, Megatapirus and Hexaprotodon, rhinocerotids constitute a likely food source for hunting or scavenging hominids. In

China, rhinoceros remains are found in 78% of anthropogenic sites (Tong, 2000). Most sites in China contain only fragmentary remains of Rhinoceros, with complete postcranial and cranial remains rare (Tong 2000). However, this is not the case in at least one

Indonesian site, Trinil, where numerous complete and almost complete postcranial elements where observed (Louys, personal observation). At the Middle Pleistocene site of

Nanjing, just south of the Yangtze River and west of Shanghai, Dicerorhinus material is found alongside human remains. Age profile analysis indicates that the majority (74%) of the individuals represented are juvenile (Tong 2001). A predominantly juvenile profile is also observed at Guanyindong, a lower Palaeolithic site in the Guizhou province, China

(Li and Wen, 1986). Tong (2000, 2001) suggests that human hunting rather than carnivore activity accounts for these distributions; however as in other regions a lack of direct evidence of hunting renders this evidence circumstantial. That early humans in

Asia utilized rhinocerotids is however not debated. Examples include Zhoukoudian and

Yanjiagang in North China (Chow, 1978, Jiang, 1990). As with other taxa however, questions remain as to whether these animals were hunted, or their remains scavenged

(Boaz et al., 2000, 2004; Binford and Ho, 1985; Binford and Stone, 1986).

2.4.4 Influence of humans in the region

The human colonisation of Southeast Asia most probably occurred during the

Early Pleistocene. Evidence from southern China and Indonesia in the form of human fossils suggest colonisation before 1.5 Ma (Larick et al., 2001; Sémah et al., 2000).

Precisely which hominin species settled the region is a matter of debate. The earliest East

46 Asians are argued to be of the general grade of either early Homo (i.e. H. habilis: Huang et al., 1995, but see Wu (2000)) or early H. sapiens sensu lato [=H. erectus/H. ergaster:

Wanpo et al., 1995; Anton & Swisher, 2004; for a discussion of taxonomic issues see

Curnoe & Tobias (2006)]. However, evidence for colonisation by early Homo relies heavily on limited dental evidence, making it controversial. In this discussion, we assume that early H. sapiens sensu lato was the first hominin to have colonised East Asia, including Southeast Asia.

Fossils from the Middle Pleistocene continue to sample early H. sapiens sensu lato, the youngest dated to greater than 400 ka in southern China (Chen et al., 1998; Zhou et al., 1999; Zhao et al., 2001). In Indonesia, stone-tools associated with early hominins have been recovered from Flores and dated to about 800 ka (Morwood et al., 1998).

Uranium-series dating of fossil humans from Ngandong in Indonesia have provided contradictory ages – mid-Middle Pleistocene (Yokoyama et al., 1998) and late Upper

Pleistocene (Yokoyama & C. Falgueres, 2002; Falgueres, Personal Communication).

While electron spin resonance dating of associated fauna is also consistent with a late

Upper Pleistocene age (Swisher et al., 1996), these dates have been challenged on several grounds (Grün & Thorne, 1997). Most recently, a new species of hominin (H. floresiensis) has been described for Flores and dated to the late Upper Pleistocene (Brown et al., 2004; Morwood, et al., 2004; Morwood et al., 2005). This taxon is claimed to have descended from early H. sapiens sensu lato or even an australopithecine (Morwood et al.,

2005) in isolation on Flores. It should be noted, however, that this research is controversial (see Henneberg & Thorne, 2004; Weber et al., 2005). Homo is represented in Thailand by a single tooth from the late Middle Pleistocene (Tougard et al., 1998). In

47 Vietnam, H. sapiens sensu lato is associated with Gigantopithecus in deposits dated to around 475 ka (Ciochon et al., 1996).

Earliest so-called archaic H. sapiens is less than 400 ka in China (Wu & Poirier,

1995; Chen et al., 1994; Rink et al., 2003), being unknown from other sites in East Asia and Southeast Asia. Oldest so-called modern H. sapiens in the region could be up to 220 ka, but an estimate of less than 200 ka for this group is probably more realistic (Wu &

Poirier, 1995; Shen et al., 2001, 2002). Apart from southern China, direct evidence for earliest modern H. sapiens is absent from Southeast Asia until the Holocene. However, the earliest Aboriginal Australians must have colonised Sahul via Southeast Asia, and

Australian fossils demonstrate their presence in the region by at least 70 ka (Thorne et al.,

1999; Thorne & Curnoe, 2000; however some authors advocate human presence no earlier than c 50 ka [e.g. Bowler et al., 2003]).

Both early Homo and early H. sapiens sensu lato produced sophisticated stone tools and exploited animal matter, and there is a strong case for them having hunted to acquire meat and intramuscular fat (see Domínguez-Rodrigo & Pickering, 2003).

However, the small (australopith-size) body of early Homo suggests important ecological and behavioural differences to early H. sapiens sensu lato, with its modern human-size body. While the cause of this change remains unknown, it has been associated with a greater commitment to a terrestrial lifestyle, inhabitation of more open environments, increased foraging distances and/or dietary shifts (Foley, 1987; Wheeler, 1992; McHenry,

1994; Leonard & Robertson, 1997; Klein, 1999). Increased body size is also associated with higher mobility, larger territory size and increased meat eating (Foley, 1987), and linked by some workers with the emergence of a modern human-like hunting and

48 gathering lifestyle (Leonard & Robertson, 1997). Larger body size has also been suggested as a major factor in the successful colonisation of the non-African Old World by hominins (Foley, 1987).

Earliest direct evidence for animal exploitation by humans in Southeast Asia is indicated by fossils associated with early H. sapiens sensu lato in southern China. As discussed above, mortality profile analysis of rhinoceros remains have been used to suggest human exploitation more than 500 ka (Tong, 2001). At the Chinese mid-late

Pleistocene site of Panxian Dadong (Shen et al., 1997; Rink et al., 2003), Schepartz et al.

(2005) found that early humans at least scavenged Stegodon carcasses and are likely to have played an important role in the formation of the faunal assemblage at this site.

Despite this level of interaction the continuation of this species does not seem to have been aversely affected; Stegodon in China survived into the Holocene (Tong and Patou-

Mathis, 2003). The site of Tongzi in China dates to around 111-180 ka and also has stone tools and evidence for fire in association with the remains of Stegodon and Rhinoceros

(Wu et al., 1975; Wu & Lin, 1985; Wu & Wu, 1985; Yuan et al., 1986).

Other examples of interaction between humans and megafauna in Southeast Asia have not produced immediate extinctions. It is suggested that Homo selectively hunted juvenile Stegodon on Flores (Morwood et al., 2004). However, it appears that hunting was sustained over thousands of years without causing its extinction, which has instead been attributed to volcanic eruptions on the island (Morwood et al., 2004).

The megafaunal losses on Borneo were likewise unlikely to have been a result of human over-hunting (Medway 1977). Based on the excavations at Niah, Medway (1977) concluded that the effects of prehistoric humans on the mammalian fauna were minimal

49 at best. He based this on two main lines of evidence. Firstly, based on the level of faunal accumulation in the site, he postulated that the density of pre-historic humans in the region was too low to impact greatly on either the fauna or the environment. His second line of evidence is based on faunal distributions. Medway observed that the abundance of the mammal most hunted, the Bearded Pig (Sus barbatus), shows little change over the depositional history of the site. By contrast the abundance of two rat species (Rattus muelleri and R. sabanus) changes significantly, even though he suggests they were not significant food sources.

As discussed above, the hunting of the orangutan is known from the historical record (see also Rijksen, 1978). In Southeast Asia, the hunting of this species has been associated with both subsistence and trophy hunting (Rijksen, 1978). While there is no direct evidence that Pleistocene Southeast Asians hunted Pongo pygmaeus, indirect evidence exists. It has been suggested that fossil Pongo, which is often associated with human remains, is indicative of their being hunted (MacKinnon, 1971; Rijksen, 1978). If that is the case then it is possible they have been hunted for as long as 500,000 years

(Rijksen, 1978). However, this line of argument is based only on association, and more direct evidence is necessary if it is to be tested and substantiated.

What of the subsistence ecology of early humans in the region? It has been suggested that Pleistocene southern Chinese diets were orientated to small animals including invertebrates, while large animal consumption was restricted to the more open north (Keates, 2003). Observations by Watanabe (1985) suggest that rainforest hunter- gatherers are almost exclusively vegetarian, and when hunting are more likely to prey on small mammals, despite the presence large mammals in the area. In addition, stone tools

50 in Southeast Asian are notoriously scarce, and Southeast Asia does not have a clear handaxe-cleaver tradition, unlike Europe and Africa (Corvinus, 2004). The choice of material for tool manufacture appears to have been largely opportunistic rather than considered (Schepartz et al., 2000; Yamei et al., 2000; Keates, 2004). In China the only tools associated with H. sapiens sensu lato are a cobble industry associated with small flake tool assemblages (Corvinus, 2004). No direct association of stone tools with H. sapiens sensu lato exists for the rest of Southeast Asia; however, a small flake industry has been provisionally associated with hominins from Sangiran and Ngandong (Corvinus,

2004). More sophisticated stone tool technologies seem only to be associated with so- called modern humans of uppermost Pleistocene age (Corvinus, 2004). This does not mean that early H. sapiens sensu lato lacked big game hunting technologies; however, due to evidence of a scavenging behaviour (e.g. Boaz et al., 2004), a case could be made that this hominin in Asia practiced a predominantly scavenging rather than hunting existence.

Finally, environmental cores from the Southeast Asian region show increasing amounts of charcoal from 225 ka onwards that could be associated with burning in southern Indonesia (Kershaw et al., 2006). Thus, human modification of the environment, a possible catalyst of megafauna extinction (e.g. Miller et al. 2005), could have been part of the ecology of Southeast Asia since the Middle Pleistocene. However, processes other than human burning might also explain the presence of charcoal increases (Kershaw et al., 2006), making such inferences uncertain.

Can the timing of the extinctions in Southeast Asia be associated with human migrations, as it has in North America? The current state of the evidence is too

51 incomplete for such an assessment. The dearth of excavations in the region coupled with poor existing chronologies makes it premature for such an analysis to be completed.

Perhaps the disappearance of megafauna can be tied to a cultural revolution, where the unsophisticated tool technology of H. sapiens sensu lato is replaced by a more sophisticated, and hunting orientated technology. Evidence of this change may be discernable in the anatomical record of hominins, the archaeological record and also the faunal (extinction) record. However, more evidence of all three from Southeast Asia is necessary before such a conclusion can be reached. While these results are preliminary, we conclude by suggesting that human over-hunting in Southeast Asia, while possibly contributing to extinctions of certain taxa in the region, could not be solely responsible. It is at least clear from this analysis that a blitzkrieg model isn’t applicable for Southeast

Asia. Rather, gradual over-exploitation of faunal resources and ecological changes brought about by human settlement might be more appropriate factors contributing to extinctions for this region. The need for more research in Southeast Asia is again demonstrated.

2.4.5 Climate and vegetation

A detailed review of climatic changes in Southeast Asia is outside the scope of this paper, and as such only the major points are covered below. The region has experienced a long- term trend towards more variable and/or drier climatic conditions over last 200,000 years, accompanied by major changes in the intensity of the monsoon and the El Nino-Southern oscillation (Hope et al., 2004). Due to the growth of continental glaciers, sea level reductions of up to 160 m were experienced during the Middle Pleistocene, and 120 m

52 during the last glacial maximum (Heaney, 1991). Temperatures were also cooler than today, perhaps by around 2-6 oC (Heaney, 1991).

Changes in vegetation are linked to glacial cycles, and Southeast Asia is no exception.

The region exhibits vegetational changes in glacial cycles of a similar scale to those of

North America (Hope et al., 2004). Southeast Asia is characterised by a number of native vegetation types, the major elements of which are evergreen tropical rainforest, seasonal or monsoon forest, montane forest and natural (as opposed to human-induced) savannah

(Heaney, 1991). The extent of these zones is controlled by a number of factors: temperature and rainfall, and in particular for Southeast Asia (but unlike Africa or

America), sea level changes (Heaney, 1991; Gathorne-Hardy et al., 2002). This last factor is the result of a reduction in the extent of shallow seas during periods of high sea level, reducing evaporation and therefore moisture in the area (Heaney, 1991). Heaney (1991) hypothesised the existence of a corridor of low rainfall running through the centre of the

Sunda shelf during periods of lower sea level. He further postulated that this area would be home to a mosaic of vegetation types. This is supported by research by Hope et al.

(2004), who noted that the, “greatest variation in vegetation is shown both within and between interglacial periods, while glacial periods (apart from interstadials) have been characterized by greater vegetation homogeneity” (Hope et al. 2004, p120). It has also been pointed out that the assemblage of mammals present throughout Southeast Asia could only exist in a mosaic of closed forest with alternating areas of scrub, bush or parkland (Medway, 1972, 1977; Cranbrook, 2000). As discussed above, this is a characteristic of many late Pleistocene communities that suffered extinctions (Guthrie,

1984; Lundelius, 1989; Graham et al., 1996; Price and Sobbe, 2005). Continued

53 alternation between savannah and rainforest, as well as the fragmentation of the landmass through rises in sea level, may have been a major contributing factor to range reduction and/or extinction of fauna in Southeast Asia, in a self-organising, threshold model of the type envisaged for North America (Graham, 2005). Although the Pleistocene ended with an extremely cold phase (Ferguson, 1993), an expansion of humid vegetation is recorded at the Pleistocene-Holocene boundary (Hope et al., 2004).

2.5 Conclusions

The Southeast Asian megafauna extinctions share certain characteristics with megafauna extinctions recorded in other parts of the world, including the presence of disharmonious assemblages, surviving megafauna exhibiting larger body size in the Pleistocene compared with modern forms, and a dearth of kill sites. The Southeast Asian megafauna extinctions are further characterised by the extinction or reduction in range over at least two countries of a number of genera, among them Stegodon, Hexaprotodon, Pongo,

Crocuta, Hyaena, Palaeoloxodon, Ailuropoda, Tapirus, Rhinoceros and Megatapirus.

Other generic extinctions seem to be more restricted, although in general this observation is more likely to result from an incomplete/poorly sampled fossil record rather than actual absence or disappearance of taxa. The chronological resolution of the assemblages is still too coarse for us to make more than preliminary predictions about causes of extinction in the region. However, given the complexity of this extinction event, it is likely that the extinctions were not the result of a single cause but more likely a combination of human induced and climatic factors. This contrasts with the extinctions on small Southeast Asian islands where humans are likely to have had a more direct and disastrous impact

(Sondaar, 1987).

54 Another equally important factor in these extinctions is likely to have come from eustatic changes in sea level. These changes in sea level may have allowed easier routes for early colonising mammals, humans among them. It also brought about changes in the structure of the vegetation, disrupted river systems, and isolated islands like Java and

Sumatra. This aspect of extinctions is unique to Southeast Asia – it is unlikely that changes in sea level affected the Americas, Europe or Australia in quite the same way.

The lists provided herein are at times based on a single site which was excavated before modern palaeontological techniques were developed. In many cases there exists limited information on taphonomic and palaeoecological factors affecting deposition and preservation. If nothing else, this study highlights the extremely limited number of excavations of Quaternary sites that have been carried out in Southeast Asia. Although a smaller number of genera became extinct in Southeast Asia relative to other continents, examinations of these extinctions will undoubtedly shed new light and, it is hoped, stimulate new avenues of research in this debate.

55 CHAPTER 3

SOUTHEAST ASIAN QUATERNARY EXTINCTIONS:

POSITION IN THE DEBATE AND FURTHER ANALYSIS1

3.1 Introduction

The Quaternary extinctions have held the fascination of scientists since the concept of

extinction became widely acceptable. In fact, it was the remains of those large beasts,

such as the wooly mammoth, who wandered the plains of prehistoric Europe which became one of the integral arguments in the evolutionary debates of the 19th and early

20th centuries (Grayson 1984b). Interest in the megafauna extinction debate has ranged

from the purely academic to the highly political, and has been studied by scientists from a range of disciplines including archaeologists, biologists, climatologists, conservationists, geologists, paleontologists, paleoanthropologists, zoologists as well as many others. The extinctions occurred on all continents save Antarctica, and at various times throughout the Pleistocene. Traditionally, the focus of research and debate has been on the Eurasian and North American extinctions but increasingly there has been considerable interest in the Australian extinctions (see, for example, the many references in Reed et al. 2006).

The African extinctions have received comparatively less attention, due largely to the fact that they were less severe than any others. Lastly, the South American and, even more so, the Asian Quaternary extinctions have received the least amount of attention. The

1 This chapter has been published as Louys, J., 2008. Quaternary extinctions in Southeast Asia. In Elewa AMT (ed) Mass extinction - the danger around us. Springer-Verlag, Heidelberg, Germany pp 159-189. This chapter was not subject to peer-review.

56 question of megafauna extinctions has, however, much relevance for today. Increased

climatic variability and human-induced environmental degradation occurring throughout

the world has resulting in the rapid extinction of many species. An understanding of

extinctions, particularly one where we may have played a part, is integral to our ability to

mitigate against further loses.

3.2 The Quaternary “megafauna” extinctions

3.2.1 What are megafauna?

The term “megafauna” is one that is not easily defined. Marshall (1984) points out that four different definitions are provided by different authors in Martin and Klein’s

1984 Quaternary Extinctions, one of the definitive works in the field. He goes on to criticize the use of the term, concluding that it “is a hindrance to understanding the extinction process itself” (Marshall 1984: 796). The use of the term is also examined by

Wroe et al. (2004), who demonstrate that its imprecise nature can lead to very different results in extinction studies. In particular, the subjective use of this term can be made to reaffirm any given hypothesis as its injudicious use allows the removal of problematic species and the addition of affirming ones. What then is the solution? Is the term too entrenched in modern work to be discounted? Marshall (1984) suggests that if the term should continue to be employed (and he makes it clear that this is doubtful), then fauna should be divided into “dwarfed megafauna” and “extinct megafauna” to distinguish between those mammals that experienced significant body-size decreases at the end of the Pleistocene, and those that became extinct. However, even if this solution were adopted, then the term “megafauna” still needs to be defined. Furthermore, it is now evident that many species experienced dwarfing prior to extinction (e.g. Guthrie 2003),

57 so these two terms could be used synonymously. Perhaps, as Martin and Klein have done, it is more prudent to refer to this global faunal extinction event as the Quaternary, rather than megafauna extinctions. This is the approach I have taken, with the understanding that “Quaternary extinctions” refers to the extinctions of the generally large-bodied mammals (i.e. megafauna) which occurred throughout the last 1.8 Ma. In particular, my analyses are restricted to five orders of mainly large-bodied mammals found in Southeast

Asia: Carnivora (e.g. tigers), Proboscidea (e.g. elephants), Artiodactyla (e.g. deer),

Perrisodactyla (e.g. tapirs) and Primates (e.g. orangutans). For a discussion on the treatment of these orders as megafauna see Louys et al. (2007).

3.2.2 The debate

The debate on the demise of the megafauna is often characterized by two highly polarized points of view: (1) human-induced extinction; and (2) climate-induced extinction. However, the nature and usefulness of these diametrically opposed views has been questioned (e.g. Wroe and Field 2006, Grayson 1984a). It may be that they represent an example of Kuhn’s paradigm hypothesis where the core of each view is unassailable (Grayson 1984a). These two points of view are extreme, however, and most megafauna extinction hypotheses fall somewhere in between them. Indeed, there is an increasing awareness that the nature of the extinction will not be in the form of “humans did it” or “climate did it”, but rather an understanding that extinctions on each continent were fundamentally different, and that each extinction event will be influenced by different factors, some having a greater impact than others (e.g. Barnosky et al. 2004). A summary of the major points of both paradigms are covered below.

58 3.2.3 Human-induced extinctions

Epitomized by prehistoric hunters exterminating megafauna upon colonizing new

continents, this paradigm covers such theories as blitzkrieg (e.g. Mosimann and Martin

1975, Martin 1984), over-hunting (e.g. Whittington and Dyke 1984), hyper-disease (e.g.

MacPhee and Marx 1997), ecological alterations (e.g. Miller et al. 2005, Miller et al.

1999) as well as any other hypothesis that posits extinction were a direct result of human

influence. Whether humans were active (as in the former two theories) or passive (as in

the latter two theories) in the extinctions is, for the purposes of this review, a moot point.

The highlights of the human-induced paradigm are that it provides a ready explanation

for:

• why extinctions seem to quickly follow human colonization

• a correlation between Pleistocene extinctions and historical (generally island)

extinctions

• why the extinction event was not as severe in Africa relative to the rest of the

world; a product of the co-evolution of Homo and megafauna such that

megafauna were not “naïve” to humans as predators

• why large-sized mammals were more likely to become extinct rather than smaller-

sized animals

• why the extinctions affected terrestrial rather than aquatic biotas

However this paradigm fails to account for or fully explain:

• why dwarfing and extinctions occurred sometimes synchronously, and sometimes

asynchronously between taxa

• why some extinctions are not invariably coincident with human colonization

59 • why megafauna extinctions should even have occurred in Africa

• the presence of non-contemporaneous faunal assemblages during the Pleistocene

(i.e. faunal assemblages representing communities which have no modern

analogues), and their subsequent disappearance

Several lines of evidence are often cited as proof of human-induced extinction. These include chronology of extinction, the archeological and paleontological records, and computer simulations (Barnosky et al. 2004). Chronology of extinction refers to the apparent synchrony of human colonization and megafaunal extinction. The evidence of close association of these temporal events are strongest for New Zealand (Barnosky et al.

2004). Some authors would argue that a strong case can also be made between these two events for Australia at ~40-50 ka (e.g. Miller et al. 2005, Miller et al. 1999, Roberts et al.

2001). However both timing for megafauna extinction (e.g. Field and Fullagar 2001,

Trueman et al. 2006) and human colonization (e.g. Thorne et al. 1999) have been brought into question.

The archaeological evidence most often debated is the presence (or more accurately absence) of kill sites. These sites are the result of hunting, and often butchering, of megafauna by humans. The paucity of these sites has been used by proponents of climatic driven extinction as evidence that overkill is unlikely. Martin

(1984), however, has argued that a lack of kill sites would be expected if the extinctions occurred quickly, leaving little time for fossilization to occur. The lack of kill sites therefore does not damage the arguments of proponents of overkill, although as Wroe et al. (2004) point out, nor does it constitute evidence for it. Kill sites are too few and their

60 interpretation in the debate too ambiguous to provide any salient value beyond proving that prehistoric hunters hunted certain taxa.

Globally, the paleontological record is generally used by proponents of climate change, although this type of evidence is equivocal for Australia. Some Australian researches maintain that megafauna are absent in Australian fossil sites dated younger than ~46 ka, the suggested date of colonization by humans (e.g. Roberts et al. 2001).

Others (e.g. Wroe and Field 2006) maintain that the date from at least one megafauna bearing site (Cuddie Springs), dated at 30-36 ka, is reliable. If the latter’s contention is sustained, this would seriously undermine the potential of a rapid, human-induced extinction of megafauna in Australia. The presence of more reliably dated deposits as well as accepted kill sites in North America and Eurasia means that paleontological datasets are not generally used in arguments advocating human-induced extinctions on those continents.

The last line of evidence to be examined for human induced extinction in this review is that of computer simulations. Models for extinction have been based on simple predator-prey models, where different variables are used to model population numbers.

The most basic variables are those of reproduction, mortality and predation rates. Within this basic setup, more complex variables can be added, and different parameters introduced, including rates of migration, effect of dwindling biomass, etc. These are varied in order to determine what ecological parameters might potentially drive animals to extinction via human hunting. There are many criticisms of this line of evidence, the least of which is that most treat all prey as a single species (Barnosky et al. 2004).

61 3.2.4 Climate-induced extinctions

Proponents of this paradigm propose that the Pleistocene extinctions occurred as a direct result of climate change. Largely those points irreconcilable with the human- induced paradigm as described above are well explained by these theories. Equally, those points well explained by the human-induced paradigm provide difficulties to the climate- induced paradigm. Some lines of evidence are however readily explained by both paradigms, for example, the lack of kill sites, as discussed above.

The Pleistocene was dominated by a succession of glacial and interglacial cycles, with the former associated with more arid climatic conditions, and the latter with more humid ones (see also the more detailed treatment of this issue as it applies to Southeast

Asia, below). The most frequent argument used against climate change as a causal factor in the extinction is that similar, and even greater, climatic fluctuations occurred previous to those associated with the extinction event; however these failed to produce significant reductions in megafauna. Another integral argument against climate change as a force in extinction is that it provides no ready explanations of why extinctions seem to follow so closely human colonization of new continents.

A solution to the first criticism follows an ecological threshold model as proposed by a number of authors, in particular Graham (2006). This hypothesis essentially suggests that successive climate changes gradually reduced population numbers of megafauna, until the last major climatic change (usually assumed to be the Last Glacial Maximum

(LGM) at ~20 ka), when population levels dropped below a critical level (the threshold).

With unsustainable population levels, megafauna eventually became extinct.

62 The second criticism is less well handled by proponents of climate change, and

they cite poor existing chronology of invasions as their primary objection. However, a

possible solution to both criticisms was recently advocated by Price and Webb (2006).

They cite evidence which suggests that the LGM was in fact associated with the most

severe levels of aridity in Australia. This therefore provides a ready explanation for why

megafauna became extinct at this time (for Australia at least), and they suggest that the

apparent synchronicity of human colonization and extinction of megafauna may in fact be

nothing more than a coincidence.

The most important point that the proponents of the climate change paradigm

have highlighted is the need to explain disharmonious or non-contemporaneous assemblages of Pleistocene faunas (e.g. Guthrie 1984, Lundelius 1989, Price and Sobbe

2005, Graham et al. 1996). These assemblages lack modern analogues because they exhibit the co-occurrence of taxa which today are found in radically different environments (e.g. Medway 1972, 1977, Cranbrook 2000). The most convincing theory explaining these puzzling co-occurrences suggests that changes in climate during the

Pleistocene altered vegetation structures from heterogeneous mosaics to homogenous slates. Proponents of climate change-induced extinctions suggest that the transition from this heterogeneous vegetative structure to the more homogenous conditions we see today resulted in megafauna range reductions, and in many cases, extinctions (e.g. Guthrie

1984). The presence of these disharmonious assemblages is one that human over-hunting proponents find hard to explain, although theories of vegetation change through loss of megaherbivores (e.g. Owen-Smith 1987), or through large scale habitat alteration are notable exceptions.

63 3.2.5 Towards a reconciliation

Despite the polarity between these two paradigms, there is now an increasing awareness that a single cause is unlikely to be solely responsible for all extinctions, and that both human and climate would have an effect (e.g. Barnosky et al. 2004). The debate has now largely shifted towards how much each factor contributed to the extinction.

However, this complex interplay of factors will likely never be completely understood, and Wroe and Field (2006) recently questioned even the validity of this approach. They cite the example of a study of the Caribbean Island lizard (Schoener et al. 2001) where in the wake of a hurricane, one species of lizard was at an elevated risk of extinction from another. As both the hurricane and the presence of the predatory lizard were necessary for extinction, Wroe and Field (2006) argue that trying to determine which of these was the primary factor provides no worthy insight, but in fact only serves to stifle the science.

Trying to understand how the various factors are likely to have influenced the biota, rather than trying to attach the lion’s share of blame to a particular extinction agent represents a more constructive avenue of approach. Only in this way may we gain insights into extinction processes, which may in turn be of benefit in understanding the modern conservation crisis.

3.3 Quaternary extinctions in Southeast Asia

Although the extinction of large mammals in Southeast Asia has been reported by several authors (e.g. Medway 1972, 1977, Sondaar 1987, Tougard et al. 1996, Cranbrook

2000), it was not until very recently that these extinctions were placed into the global

“megafauna extinction” context. Southeast Asia is a complex biogeographical unit, a result of various biotic migrations and endemisms, periods of insularity and unity,

64 orogenies and changes in climate. Therefore, in order to understand the megafauna extinctions in Southeast Asia, one must first understand some of its complex biogeographical history. I will firstly describe those geographical and environmental changes that have affected Southeast Asia’s biota. Secondly, I will provide a review of the ecological requirements of those taxa that experienced extreme geographical range reductions and, where possible, those taxa which became extinct. Twenty species of

megafauna have an extinction record in more than one Southeast Asian country (Table 1).

Mechanisms of extinction will be examined for Southeast Asia within the context of the

two extinction paradigms described above. In particular, the possibility of over-hunting

of Southeast Asia’s megafauna by early hominids, as well as the impacts of climate,

vegetation and sea level changes, will be discussed. Finally, a brief examination of the

modern extinction crisis in Southeast Asia will be presented.

65

Common Name Species Name Extinct/Extant China South Burma Laos Cambodia Vietnam Thailand Malaysia Java Borneo Dubois’s antelope Duboisia santeng † X X Yunnan horse Equus yunnanensis † X X Asian gazelle Gazella sp. ∞ X X Giant ape Gigantopithecus blacki † X X Robust macaque Macaca robustus † X X Pachycrocuta Giant hyena † X X brevirostris Chinese rhino Rhinoceros sinensis † X X South Asian rhino Rhinoceros sivalensis † X X Bearded pig Sus barbatus † X X Lydekker’s pig Sus lydekkeri † X X Elephant stegodon Stegodon elephantoides † X X Giant tapir Megatapirus augustus † X X X Indian rhino Rhinoceros unicornis ∞ X X X Stegodon Stegodon orientalis † X X X Malayan tapir Tapirus indicus ∞ X X X Naemorhedus Serow ∞ X X X sumatraensis Giant panda Ailuropoda melanoleuca ∞ X X X X Asian spotted hyena Crocuta crocuta ∞ X X X X Palaeoloxodon Archaic elephant † X X X X X namadicus Orangutan Pongo pygmaeus ∞ X X X X X X

Table 3.1: Species becoming locally extinct in more than one Southeast Asian country. † indicates globally extinct; ∞ indicates globally extant; X indicates country of extinction. Adapted from Louys et al. (2007).

3.3.1 Geography of Southeast Asia

Southeast Asia, as discussed here, includes Southern China, Myanmar, Thailand,

Laos, Cambodia, Vietnam, Malaysia, Borneo and Java. It is generally separated into two distinct provinces, those of Sunda and Indochina (Figure 3.1), on the basis of a number of biogeographical differences (e.g. Tougard 2001, Louys et al. 2007). Southeast Asia can

66 also be separated climatically (Chuan 2005) into continental (consisting of Southern

China, Myanmar, Thailand, Laos, Cambodia and Vietnam) and insular (consisting of

Malaysia, Singapore, Indonesia and the Philippines).The geography of Southeast Asia has been explored in a number of publications (e.g. Hall and Holloway 1998, Gupta

2005), such that this chapter will provide only a brief description of the salient geographical points.

Figure 3.1: Southeast Asia at the LGM showing the major rivers and lakes present during that time. Heaney’s (1991) proposed savannah corridor is shown in yellow. The dotted line represents the division between Indochina (north) and Sunda (south). Adapted after Bird et al. (2005).

Geological history

The Southeast Asian continental block is largely composed of elements which had broken off from the southern super-continent Gondwanaland (Gatinsky and Hutchinson

1987, Metcalf 1990, 1996). Much of the modern geographical aspects of Southeast Asia occurred as a result of the fusion between the Sinoburmalaya and Cathaysia plates, in an event known as the Late Triassic Indosinian Orogeny (Hutchinson 2005). More plate

67 collisions followed, including the collision of the Burma plate with Shan highlands in the

Cretaceous and the collision between India and Eurasia in the Eocene (Hutchinson 2005).

The latter was considered a “gentle affair” (Ferguson 1993) until the Miocene, where it

resulted in the North-South trending mountain ranges of Western Yunnan, Burma and the

Malay peninsula to the east, and the uplift of the Himalayas and the Qinghai-Tibetan plateau to the north (Whitmore 1987). The uplift of the Himalayas and the Qinghai-

Tibetan plateau during the Miocene significantly altered the established climate, and resulted in the initiation of the Southeast Asian monsoon system (Ferguson 1993).

Furthermore, the continued uplift of the Qinghai Plateau during the Pleistocene resulted

in a loss of intensity of the summer monsoons during glacial periods, while at the same

time increasing the intensity of the winter monsoons (Ferguson 1993). Evidence for these

changes in monsoon intensity are also supported by data from ocean cores (e.g. Wang et

al. 1999, Gingele et al. 2002). Another, relatively more recent plate collision occurred

some 15 Mya between Southeast Asia and Australia, leading to the formation of the

Lesser Sunda islands, as well as parts of Sulawesi and the Philippine Archipelago

(Hutchinson 1989).

Climate

Three factors have had a major impact on the climate in Southeast Asia over the

Quaternary: the position of the Inter-Tropical Convergence Zone (ITCZ), changes in

temperature and changes in sea level (Verstappen 1980). The ITCZ and temperature

changes will be dealt with here; changes in sea level will be dealt with in a separate

section below.

68 The ITCZ is a low pressure belt situated at or near to the equator, although its position fluctuates throughout the year. In Southeast Asia, its position in December –

February passes south of Java and over the northern part of Australia, while in July –

September it passes near the Himalayas and the Bay of Bengal, then extends south to near

the equator east of New Guinea (Verstappen 1980). Since the maximum amount of

precipitation occurs at or near the ITCZ, it has a major impact on Southeast Asia’s

climate. Verstappen (1980) suggests that the ITCZ had a more southerly position during

glacial periods, thereby resulting in an increased dryness in the northern latitudes during

these times.

The temperature changes in the Quaternary of Southeast Asia are also related to

glacial and interglacial periods. In general the temperature in Southeast Asia has been

cooler than is present today. This is supported largely by deep sea cores, which indicate

temperatures approximately 3-5oC colder on land, and 2-4oC in the seas (e.g. Chappell et

al. 1996, Heaney 1991), however these temperature may have been raised 1-2oC during

interglacials (e.g. Verstappen 1997).

Current climate in Southeast Asia varies geographically. Continental Southeast

Asia is characterised by greater seasonality, temperature and rainfall extremes, as well as

more pronounced dry spells, whereas insular Southeast Asia has a more equable climate

(Chuan 2005). The monsoon weather system determines, more so than temperature, the

precipitation patterns in Southeast Asia. Two monsoon seasons are present: the summer

or southwest monsoon (March – October) and the winter or northeast monsoon

(November – April). The summer monsoon brings increased precipitation to continental

Southeast Asia, as well as Sumatra and Borneo, while the winter monsoon brings

69 increased precipitation to insular Southeast Asia and Australia. However, distribution of rainfall is also strongly dependent on topographical relief (Chuan 2005).

Vegetation

Vegetation in Southeast Asia is governed by two major factors: water availability

and vertical gradient (Corlett 2005). Southeast Asia has been described as the “region of

forests climates”, which refers to the fact that, until human- generated deforestation

within the last few thousand years, the climate of Southeast Asia was such that it

supported little else but forest (Corlett 2005). Three major vegetation types for Southeast

Asia have been described by Corlett (2005): lowland vegetation (including, among

others, tropical rainforests, tropical deciduous forests, savannahs and shrublands),

montane vegetation (including montane forests and alpine vegetation) and wetlands.

Tropical rainforests are arguably the most famous of Southeast Asia’s vegetation

types. As a global percentage, Southeast Asia’s rainforests are second only to the

neotropics (Heaney 1991, Primack and Corlett 2004). Southeast Asia’s tropical

rainforests are predominately found in a large block situated in the western part of the

Sunda shelf (including Sumatra, the Malay Peninsular and Borneo) (Corlett 2005).

Rainforests are also present (or were until recently) in scattered parts of both insular and continental Southeast Asia, including Java and Myanmar (Corlett 2005). The Southeast

Asian rainforest is distinguished from other types of rainforests around the world in its dominance of dipterocarps, such that the forest canopy reaches heights of around 30-40 m

(Primack and Corlett 2004, Corlett 2005). Tropical rainforests largely occur in regions devoid of regular dry seasons (Primack and Corlett 2004). In regions that experience more extensive dry seasons (i.e. one-four months), the vegetation type changes from

70 tropical rainforest to tropical seasonal forest. These forests are characterised by an

increase in the number of deciduous trees, which can at times compose up to 50% of the canopy tree species, although the majority of the forest remains evergreen (Corlett 2005).

Tropical deciduous forests occur where rainfall is too low to support a predominantly evergreen forest, generally when the length of the dry season ranges from three to seven months (Corlett 2005). The demarcation between evergreen and deciduous forests is not a gradual one, but usually sharply defined due to the respective forests’ tolerance to fire conditions (Corlett 2005).

Other types of vegetation in Southeast Asia include montane vegetation, characterised by shorter, evergreen trees, which are often converted to open woodlands and savannahs by fire during dry seasons; and wetlands, including vegetation especially adapted for soil saturation. Fire regimes seem to be of particular importance in the type of vegetation currently present in Southeast Asia, in particular in the formation of savannahs and shrublands (Corlett 2005). Although these fires can be natural, the last few thousand years have seen an increase in the use of fire by Southeast Asia peoples (Anshari et al.

2001). The effects of humans on the fire regime will be explored in more detail below.

Sea level changes

Changes in sea level have had such an unparalleled influence on Southeast Asia’s geography and biota that a detailed treatment of these changes was necessary. Sea level changes occur due to the movement of huge volumes of water from oceans to ice sheets, and back again (Lambeck et al. 2002). These movements are a result of oscillations between glacial and interglacial climatic conditions, and have seen sea levels fall as low

as 140 m below present (Lambeck et al. 2002). Regional changes in sea level can be

71 approximated using global sea level fluctuations, which are in turn commonly derived from oxygen isotope studies (e.g. Chappell and Shackleton 1986). Because these oxygen isotopes are derived from marine core drillings, their record can be limited by the depth of the cores. The resulting resolution is therefore much greater for more recent sea level changes but becomes coarser further back in time. Oxygen isotopes are divided into

Oxygen Isotope Stages (OIS) (also called Marine Isotope Stages, or MIS), which in turn relate to glacial and interglacial periods. By convention, odd numbered OIS are associated with interglacials, and even numbered OIS associated with glacials (Figure

3.2).

Figure 3.2: Global sea-level for the last 130 ka, with associated Oxygen Isotope Stages (OIS). The points at which the major Indonesian islands were connected are indicated by the colored lines: Sumatra (blue), Borneo (green) and Java (red). Relative sea-levels from Chappell & Shackleton (1986); OIS from Martinson et al. (1987).

During periods of lower sea level, many of the present islands in Southeast Asia were connected to the mainland. Based on modern bathymetric lines, Sumatra would be

72 connected to the Malaysian peninsula if the sea level dropped 30m below present, Borneo at 40m below present, and Java at 50m below present (Bird et al. 2005). These sea-levels are indicated on Figures 3.2 and 3.3. However, these values are somewhat simplified: uplift, subsidence and volcanic eruptions occurring throughout the Pleistocene make them an approximation at best (Voris 2000, Bird et al. 2005). Nevertheless, this model serves as a suitable estimate for land connections.

Fluctuations in sea level have resulted in dramatically altered hydrology in the region. Major rivers crossed the exposed Sunda shelf during lower sea levels

(Mollengraaff 1921, Voris 2000). The largest of these is the North Sunda River, which drained northeastern Borneo, the northern parts of the Java Sea and southern Sumatra

(Figure 3.1). Another large river, the East Sunda River, drained southern Borneo and northern Java, draining in a southeasterly direction. The Siam River is also shown on

Figure 3.1, providing drainage for the eastern Malaysian peninsular as well as the Gulf of

Thailand. Many other, smaller tributaries drained other parts of the exposed shelf

(Mollengraaff 1921, Bird et al. 2005, Voris 2000). In addition to these rivers, exposed basins on the shelf would have formed large freshwater lakes (Bird et al., 2005).

The existence of a low-lying continental shelf in Southeast Asia resulted in a dramatically different landscape than that found today. Heaney (1991) proposed a

‘savannah corridor’ running through the middle of Sundaland during the LGM, surrounded on both sides by tropical forests (Figure 3.1). Essentially, this savannah corridor would result from a decrease in the surface area of the Sunda Sea, with a resulting loss of evaporation, and hence precipitation in the region (Heaney 1991). This theory has received support from Bird et al. (2005), who examined data from a large

73 number of scientific fields, including geomorphology, biogeography, palynology and

vegetation modelling. Although the actual extent of the corridor cannot yet be determined

with any level of accuracy, Bird et al. (2005) suggest that a minimal-sized corridor would

be 50-150km wide, running between the islands of Bangka, Belitung and Karimata and along the Java Sea through to Java. They also note that little evidence for the existence of

such a corridor north of the equator exists.

Although Heaney (1991) proposed the existence of this corridor for the LGM,

presumably this corridor would not be restricted to this period, but in all likelihood would

also have occurred during other glacial periods. Indeed, evidence for drier conditions in

Southeast Asia during the Late Pliocene and Early Pleistocene exist, including the

deposition of thick boulder beds and braided river and alluvial sand sediments

(Verstappen 1975, 1997, Batchelor 1979, 1988), as well as palynology records (de Vos et

al. 1994, de Vos and Long 2001). The loss of this corridor, resulting from the transition

from glacial to interglacial conditions, and its associated effects on the megafauna, is

explored below.

3.3.2 Southeast Asia’s megafauna

Unlike North America and Australia, many species of megafauna can still be

found living in Southeast Asia. These include such charismatic species as the giant panda

(Ailuropoda melanoleuca), the orangutan (Pongo pygmaeus), the Javan and Sumatran rhinoceroses (Rhinoceros sondaicus and Dicerorhinus sumatrensis), the Asian elephant

(Elephas maximus) and the tiger (Panthera tigris). Southeast Asia’s extinct megafauna is less well known to general audiences, but include several species of stegodons, hyaenas and bovids (e.g. Stegodon orientalis, Pachycrocuta brevirostris and Naemorhedus

74 sumatraensis respectively), as well as the giant tapir (Megatapirus augustus) and the

giant ape (Gigantopithecus blacki). The ecological requirements of those twenty species

of megafauna listed in Table 3.1 are explored below.

Dubois’s Antelope

Dubois’s antelope (Duboisia santeng) is an extinct antelope known from fossil deposits in Java and Malaysia (Louys et al. 2007). At various times this species has been referred to the genera Bos, Anoa and Capra (see Brongersma (1936) for a discussion on the history of this species), and as such is likely to have shared certain similarities in ecology with these genera. In particular, as a close relative of the extant antelopes, it is likely to have subsisted as a grazer, although this inference should be subject to more detailed study.

Yunnan Horse

The Yunnan horse, (Equus yunnanensis), is known only from the northern parts of

Southeast Asia during the early Pleistocene (Louys et al. 2007). Like other members of the genus Equus, it is likely to have been a grazer, and probably occupied open tracts of grasslands.

Asian Gazelle

Although not identified to the species level, the Asian gazelle (genus Gazella sp. indet.) is recorded from Southern China and Burma during the Early Pleistocene (Louys et al. 2007). Although Corbet and Hill (1992) do not record the occurrence of the Asian gazelle in China, the goitered gazelle (Gazella subgutturosa) is recorded in the faunal list of Bogdhad Mountain Biosphere Reserve (Information Centre for the Environment

75 2007), a reserve situated in the far northwest of China (44°00'N/83°00'E). In addition to this species, which is also found in the far west of the Indomalayan region, the Indian gazelle (Gazella bennettii) is found in the drier parts of peninsular India (Corbet and Hill

1992). Both extant species are found in open, arid to semi-arid plains (Corbet and Hill

1992).

Giant Ape

An understanding of the ecology of the Asian giant ape (Gigantopithecus) has

advanced considerably since it was interpreted as “a hunter of large ungulates who

apparently dragged his prey back to the cave” (Livingstone 1964: 1284). However, direct

interpretations of this fascinating ape’s environment have been limited, due largely to the

paucity of the fossil material available for study. Ciochon et al.’s (1990) study remains

one of a few which deal with this subject matter. Opal phytoliths (opalized remains of

plants) found on the molars of Gigantopithecus revealed a diet of a variety of fruits and

grasses. Furthermore, faunas associated with Gigantopithecus suggest a tropical to sub-

tropical environment (Kahlke, 1984; Ciochon et al., 1990). It is likely that this ape lived

in a similar way to other extant large bodied apes, namely the orangutan and the gorilla,

although its large size would in all likelihood have restricted its arboreal capabilities.

Robust Macaque

The robust macaque (Macaca robustus) has been found largely in Chinese deposits, but is also known from deposits in Laos (Fromaget 1936). While the palaeoecology of the robust macaque has not been studied in any great detail, an examination of fossil macaques in China indicates that they were all restricted to

76 deciduous broad-leafed forests of sub-tropical to warm temperate zones during the

Pleistocene (Pan and Jablonski 1987, Jablonski and Pan 1988). The robust macaque is most closely related to the Japanese macaque (Macaca fuscata) and the Formosan rock macaque (M. cyclopsis) (Jablonski and Pan 1988), and as such may share certain ecological similarities with these species.

Giant Hyena

The giant hyena (Pachycrocuta brevirostris, formally Hyaena brevirostris) is the largest of the true hyenas (Turner and Anton 1996). The postcranial skeleton of the giant hyena indicates that it was not built for running; and although it was only slightly taller at the shoulder than the spotted hyena (Crocuta crocuta), it was longer, and its skull was larger and more powerful (Turner and Anton 1996). It is hypothesised as having been an occasional hunter and aggressive scavenger, preying on medium-sized carcasses (Turner and Anton 1996). Its extinction during the Pleistocene has been tied to the world-wide faunal turnover in the felid guild, itself a result of the change in structure of ungulate fauna (Turner and Anton 1996).

Rhinoceroses

Both extant species of Southeast Asian rhinos, the Javan rhinoceros (Rhinoceros sondaicus) and the Sumatran rhinoceros (Dicerorhinus sumatraensis) are critically endangered. Several species of rhino become extinct in Southeast Asia during the

Pleistocene (Louys et al., 2007), including, among others, the extant Indian rhino

(Rhinoceros unicornis), the South Asian rhino (R. sivalensis) and an intermediate form between the Asian and Indian members of Rhinoceros, R. sinensis (the Chinese rhino).

77 The extant Southeast Asian rhinoceroses are browsers, while the Indian rhino is predominately a grazer (Parr 2003, Groves and Kurt, 1972, Laurie et al. 1983). All extant species of rhino wallow, and their preferred habitats include a ready source of water (e.g. river, wetland, streams, etc) (Parr 2003, Groves and Kurt 1972, Laurie et al. 1983). The

Southeast Asian rhinos can be found in tropical and evergreen forests, while the Indian rhino prefers alluvial plains (Parr 2003, Groves and Kurt 1972, Laurie et al. 1983).

Pigs

Two species of pig are listed by Louys et al. (2007) as becoming extinct in more than one country, the bearded pig (Sus barbatus) and Lydekker’s pig (Sus lydekkeri). Pigs subsist in a wide variety of habitats, but prefer those where there is some vegetative cover is available (Nowak 1999). Lekagul and McNeely (1988) noted that they are more common in wet forests, and during the dry season are usually found in riparian environments. They are omnivorous, eating anything from carrion to green vegetation, and can be found in droves of over 100 (Nowak 1999).

Stegodons

The stegodons were a characteristic element of many of the Southeast Asian faunas, although a detailed analysis of its ecology has yet to be completed. In a study of the diet of modern and fossil elephants, Cerling et al. (1999) reported on tooth enamel from a stegodon, Stegodon sp., a species from Dhok Pathan, Pakistan, dated at 7.4 Ma

13 old. The δ C values obtained for this specimen indicates that C4 grasses would likely

constitute a large part of its diet. Extrapolating the likely diet of Pleistocene stegodons

from this one data point is speculative at best. Alternatively, van den Bergh (1999)

78 assumes a more browsing diet for continental (as opposed to insular) stegodons, due to

the presence of their low-crowned teeth. This is an area which clearly requires more

research. The most common species of stegodon, Stegodon orientalis, survived until the

early Holocene in Southern China (Tong and Liu 2004).

Malayan Tapir

The Malayan tapir (Tapirus indicus) is the only extant species of tapir found

outside of South America. It favours well-watered environments with dense forest

vegetation (Novarino et al. 2005, Lekagul and McNeely 1988). Although Lekagul and

McNeely (1988) suggest they inhabit only primary forests, recent research by Novarino

et al. (2005) suggest they actually prefer dense secondary forests. They are generally

solitary and most of their activity is tied to rivers and wetlands; they are excellent

swimmers and often take refuge in water (Novarino et al. 2005, Lekagul and McNeely

1988). Tapirs are browsers and feed on a variety of aquatic and low-lying vegetation

(Nowak 1999, Lekagul and McNeely 1988). In Thailand they are found in tropical evergreen, mixed dipterocarp and mixed deciduous forests (Parr 2003).

Giant Tapir

Despite its description over 80 years ago, little work has been done on the ecology of the giant tapir. Originally described as a subgenus of Tapirus, it would appear likely that its habitat and diet were similar to that of the Malayan tapir. Based on its dentition, it was approximately 25% larger than the extant tapir (Tong 2005).

79 Serow

Members of the genus Naemorhedus include the goral and the serow. These goat-

like animals are found in a range of forest types, but appear to prefer steep limestone

terrain with thick forest (Lekagul and McNeely 1988). They are established climbers, and

often shelter in deep forest cover and caves (Parr 2003, Lekagul and McNeely 1988).

Like goats they eat a variety of vegetation, but appear to prefer leaves and shoots

(Lekagul and McNeely 1988).

Giant Panda

The giant panda, so characteristic of Southern China, was once widespread throughout most of Southeast Asia (Louys et al. 2007). Although subsisting almost exclusively on a diet of bamboo, it has however been observed taking small vertebrate carcasses (Sheng 1999). Due to the heavy dependence of the panda on its preferred food, its prehistoric distribution has been tied to the distribution of temperate bamboo (Tougard et al. 1996). It is currently found in high altitude montane forests, consisting largely of mixed coniferous and broad leafed vegetation (Sheng 1999).

Asian Spotted Hyena

Members of the (extant) spotted hyena, Crocuta crocuta, currently restricted to

Africa, have also been found in Pleistocene Southeast Asia. In Africa, the spotted hyaena reaches its greatest densities in flat, open country, although it can be found from as diverse habitats as hot and arid areas to dense mist forests (Kruuk 1972).

80 Archaic elephant

Often put in the in same genus as the modern Asian elephant (Elephas),

Palaeoloxodon is actually an extinct, closely related genus (Shoshani and Tassy 2005).

However, along with the stegodons and the giant tapir, little research has focused on the specific ecology of this species. Given its high similarity with the modern Asian elephant, however, it is likely to share many habitat and environmental traits. The modern Asian elephant is largely a grazer (Eltringham 1982), and lives in a diverse range of habitats from the forests of Malaysia to the grasslands of Sri Lanka (Eltringham 1982).

Orangutan

The orang-utan is the only extant species of ape (apart from humans) endemic to

Southeast Asia. It has a solitary and arboreal lifestyle, although it has been known to descend from the trees. It subsists largely on a diet of fruits, although fungus, leaves, bark honey and insects are sometimes included (Galdikas 1988). Currently restricted to the rainforests of Sumatra and Borneo, like the panda and stegodons, it was much more widely distributed during the Pleistocene (Louys et al. 2007).

3.3.3 Human overhunting in Southeast Asia?

Synchrony of colonisation and extinctions in Southeast Asia is particularly

difficult to establish, and is exacerbated by questions of human evolution and poor

chronology of sites (Louys et al. 2007). Unlike North America and Australia, the earliest colonisers in Southeast Asia were not modern humans but Homo erectus. Evidence as to whether Homo erectus in Asia subsisted predominantly on a scavenging or hunting diet is still equivocal, although current evidence is suggestive that the former scenario is more

81 likely (e.g. Boaz et al. 2004, Schepartz et al. 2005). If this evidence is borne out, it is unlikely that these hominids would have adversely impacted on fauna through over- exploitation.

The earliest records for Homo sapiens in Southeast Asia come from Niah Great

Caves, Sarawak, dated to approximately 35 ka (Barker et al. 2007). While evidence of hunting is generally accepted for Niah caves, there is no evidence that this was done at an unsustainable rate (Corlett 2007). The predominant prey species appears to have been pigs, followed by primates (Barker at al. 2007, Medway 1977). Bearded pigs are present in all levels of excavation of Niah Caves, suggesting that they were not adversely affected by traditional hunting practices (Corlett 2007). This also suggests, however, that they were not adversely affected by any climate change. Orangutans comprise more than

30% of the non-human primate specimens at Niah, and this high frequency has been explained as a result of human occupants specializing in hunting these apes (Harrison et al. 2006). However given that Borneo is one of only two places in the entirety of

Southeast Asia where orangutans still survive, it seems quite unlikely that prehistoric humans could overhunt this species to extinction, as has been suggested for other countries in Southeast Asia (e.g. Harrison et al. 2006). Attributing the extreme geographical decline suffered by the orangutan during the Pleistocene to human overhunting seems premature at this time, given the lack of clear evidence to that fact, coupled with the adverse effects climate change has had on this species (see below).

Furthermore, there is no evidence to suggest that the Pleistocene toolkit of

Southeast Asians was adapted to hunting of big game for almost the entirety of the

Pleistocene. Stone tool technology in Southeast Asia consisted largely of choppers, and

82 did not share the sophistication of other regions, in particular Europe, until the advent of the Haobinhian during the Late Pleistocene (Reynolds 1990, Corvinus 2004). One explanation for this apparent lack of sophistication suggests the ubiquitous use of bamboo as a substitute for stone tools (e.g. Pope, 1989). However, bamboo appears particularly unsuitable for large game hunting and butchering (West and Louys 2007). In fact, it appears hunting did not become unsustainable in Southeast Asia until the last 2-3000 years, largely as a result of increasing human populations and the advent of modern hunting arms (Sodhi et al. 2004, Corlett 2007).

3.3.4 Climate change and megafauna

Climate change, and associated fluctuations in sea levels, has dramatically altered

the region’s vegetation, as discussed above. In particular, the existence of a savannah corridor running through the middle of Sundaland, a concept which seems foreign when we think of Southeast Asia, appears to have been the norm during the Pleistocene. Like

other continents, Southeast Asia hosts a number of disharmonious assemblages, and the

presence of this corridor provides a ready explanation for these assemblages (Louys et al.

2007, Medway 1972). It is also notable that for the majority of the Pleistocene, Sumatra,

Borneo and Java are likely to have been connected to the mainland (Figure 3.3).

83

Figure 3.3. Global sea-level (x-axis) for the past 1.2 Ma (y-axis). The points at which the major Indonesian islands were connected are indicated by the vertical lines: Sumatra (dashed), Borneo (dotted) and Java (solid). Snapshots of Southeast Asia at selected relative sea-levels are shown. Relative sea-level calculated from Chappell & Shackleton (1986), relative ages calculated from Shackelton and Opdyke (1976), maps of Southeast Asia from the Field Museum of Natural History (2006) (Voris 2000; Sathiamurthy and Voris 2006).

84 Interglacials, and associated high sea levels (i.e. comparable to present levels), occurred for only 10% of the Pleistocene (Lambeck et al. 2002). In fact, for over half of the past 250 ka, sea levels have been at least 40m below present (Voris 2000). Because glacial conditions and lower sea levels dominated during the Pleistocene, the principal type of vegetation was likely to have been characteristic glacial vegetation, that is, a heterogeneous vegetative structure including a central savannah corridor, more temperate forests, rainforest refugia, and more extensive river and lake systems, with concomitant drier and cooler climes. Animals inhabiting Sundaland would need to be versatile enough to cope with these dramatic fluctuations in vegetation, or alternatively retreat to refugia during periods of ecological stress, whereupon they could return once conditions become more optimal.

Of the animals examined above, some prefer open environments over closed.

Both the horse and gazelle were largely restricted to the more open north of China throughout the Pleistocene, a result of their preference for more open environments

(Jablonski and Whitford 1999). Stable isotope studies show that more open and arid environments were present during the Late Pliocene of central China (Kaakinen et al.

2006), and, given that there existed no physical barrier between the north and south of

China during this time, it is likely the southern migration of the horses and gazelles occurred then. The Early Pleistocene of China showed a retreat of steppe-like and temperate zones northwards, with concomitant advances of the tropical and sub-tropical zones (Jablonski and Whitford 1999). The tropical and subtropical zones then subsequently retreated in the Middle to Late Pleistocene, however, steppe-like environments and their associated faunas remained restricted to the north of China due to

85 the physical barrier presented by the rising Qinling Mountains (Ferguson 1993), a barrier which continues to act today (Xie at al. 2004). The horse and gazelle were therefore most

likely relicts from a Pliocene southern incursion, becoming eventually extinct as a result

of loss of their preferred habitat.

Further south, the Indian rhino became extinct in Thailand, Vietnam and Java in

the Middle to Late Pleistocene. Being predominately a grazer, it would conceivably be

adversely affected by the loss of the central savannah corridor. Indeed, its current decline

in India has been tied largely to loss of grazing land (Laurie et al. 1983). Likewise,

grazers and other open-adapted faunas such as Dubois’s antelope, perhaps the stegodons,

and the archaic elephant Palaeoloxodon would also be adversely affected by the loss of

grasslands. The loss of these grazers would detrimentally affect the major scavengers of

the region, namely the hyenas. Modern spotted hyenas, although inhabiting diverse

habitats, are more common in open grasslands, where carcasses are more easily accessed, and their abundance is directly proportional to the abundance of ungulates (Kruuk 1972).

Many extinct Southeast Asian mammals appear to be intrinsically tied to freshwater sources. In particular, the rhinoceroses, the Asian hippopotamus (not discussed here, but for an account of this species see Jablonski, 2004 and Louys et al.,

2007), the tapir, the giant tapir, and to a lesser extent the pigs, all required ready access to either standing bodies of water or riparian environments. The extensive loss of these sources of water through rises in sea-level, and concomitant changes in the hydrological regime of the region could only have had a negative ecological impact on these taxa.

These changes are likely to have contributed, at least in part, to their range reductions, and in some cases, extinctions.

86 Gigantopithecus and the orang-utan were also likely adversely affected by

changes in climate. The ultimate cause of their extinction has been reviewed by a number

of researchers (e.g. Jablonski et al. 2000, Jablonski and Whitford 1999). According to

their research, it appears likely that these species were driven to extinction or range

reduction as a result of the deterioration of their preferred environments during the

Middle to Late Pleistocene, in particular the reduction in the tropical and subtropical

zones experienced in the north of Southeast Asia.

3.3.5 The modern extinction crisis

Although the Pleistocene shows relatively little effects of humans on Southeast

Asia’s biota, the last 10 000 years or so has seen unprecedented levels of anthropogenic pressure; in almost all cases negative. There is general consensus that the massive deforestation which has occurred in earnest since the 1800s as a result of an expanding agricultural economy has been the driving force behind most extinctions (e.g. Sodhi et al.

2004). However there has also been an increasing awareness of the detrimental effects of unsustainable hunting and wildlife trade in the region (e.g. Sodhi et al. 2004, Corlett

2007). Five species of mammals are currently listed as extinct in Southeast Asia (IUCN

2007), with many more critically endangered. It has been estimated that if human- induced deforestation continues unabated, three-quarters of Southeast Asia’s original forest cover will be gone by the end of the century (Sodhi et al. 2004), wreaking unknown devastation on its biota. Furthermore, wildlife is being removed from Southeast

Asia’s tropical forests through hunting and wildlife trade at six times the sustainable rate

(Sodhi et al. 2004); such that the density of large mammals even in protected areas are lower than they should be expected (Steinmetz et al. 2006, Corlett 2007).

87 The timeline of anthropological escalation of environmental destruction can be

roughly traced out, and is very much restricted to the Holocene. Traditional Chinese

medicine dates back approximately 5000 years, and is associated with the continued

reduction of a number of large vertebrates including tigers, bears, rhinos, monkeys and

pangolins (Sodhi et al. 2004). Although evidence exists of hunting in the region from the

Late Pleistocene (Niah), evidence for unsustainable hunting does not appear until the last

2-3000 years (Corlett 2007). Extensive habitat destruction is also tied to the recent past -

the intensity of man-made fires becomes important only within the last 1400 years

(Anshari et al. 2001), while deforestation for agriculture, and more recently logging, has

been largely restricted to the last 200 years.

While the number of extinct mammals in Southeast Asia is currently relatively

small (compared to the rest of the world), this is likely to change dramatically in the near future should current deforestation and wildlife trade practices continue. Even if they do

change, however, it is quite likely that the momentum of past activities will have a

detrimental impact on the survival of Southeast Asian biota in the future (Brook et al.

2006). But the severity of such potential impacts may be partly mitigated if conservation

practices are implemented now. Although this will be difficult due to Southeast Asia’s

complicated socioeconomic status, it needs to be driven in part by a larger global

awareness of this looming disaster. As Dudgeon (2000) notes, tropical Asia does not

invoke the same conservation concern as Africa and the neotropics. This is astounding

considering that Indonesia, for example, hosts more birds and flowering plants than the

whole of Africa (Dudgeon, 2000), and that up to 42% of Southeast Asia’s biodiversity

could be extinct by 2100 (Sodhi et al. 2004). I certainly agree with Corlett (2007), that

88 unless something is done soon, we will witness the mechanisms of megafauna extinctions

in action.

3.4 Summary

Limited evidence suggests that humans contributed to the extinction of Southeast

Asia’s megafauna until the Holocene. Rather, major changes in vegetation, hydrology

and geography, resulting in large part from fluctuations in sea level, have had severe

impacts on many large-bodied mammals, in many cases leading to their eventual

extinction. Those taxa that did not become extinct often underwent severe restrictions in geographic distribution. Southeast Asia’s extant megafauna have not been left unscathed by these Pleistocene environmental throes, and the beginning of the Holocene finds many of them, in a figurative sense, severely battered and just holding on. Enter modern humans, and this precarious situation has been severely tested. In some cases, this test has already failed. However, as Sodhi et al. (2004) state, this is just the tip of the iceberg. It

may in fact be already too late to save those species which are critically endangered. And

if nothing is changed, many more than these will follow.

89 CHAPTER 4

THE ROLE OF VOLCANISM IN EXTINCTIONS: THE

CASE OF THE TOBA ERUPTION1

4.1 Abstract

The effect of the Toba super-eruption at ~74ka on the mammals of Southeast Asia is examined. Although few Late Pleistocene sites from Southeast Asia have been described, an analysis of those which pre-and post-date Toba reveals relatively few species became extinct following the eruption. It is suggested that species survived in refugia immediately following the eruption, and that they repopulated vast areas following a probable short period (a few centuries) of environmental devastation. This study suggests that mammals are more robust at coping with catastrophic events than previously acknowledged, and questions the perceived human monopoly in overcoming ecological adversity.

4.2 Introduction

Lake Toba, Sumatra, a 100 x 30 km caldera complex, is the site of the Quaternary’s largest explosive eruption (Figure 4.1). This super-eruption expelled an estimated 7 x

1015 kg of rhyolitic magma (Rose and Chesner, 1987, 1990; Chesner and Rose, 1991) and injected ~1015g of fine ash into the stratosphere (Zeilinski et al., 1996; Bühring and

Sarnthein 2000) over an estimated period of 9 to 14 days (Ledbetter and Sparks, 1979).

1 This chapter has been published as Louys, J., 2007. Limited effect of the Quaternary’s largest super- eruption (Toba) on land mammals from Southeast Asia. Quaternary Science Reviews 26, 3108-3117 doi:10.1016/j.quascirev.2007.09.008.

90 Rhyolitic ash deposits have been found as far afield as the seabed between Borneo and

Indochina (Bühring et al., 2000), and suspected Toba tephra have been found on the

Indian sub-continent, more than 3000 km away from Sumatra (Oppenheimer, 2002).

Dating of the various deposits associated with Toba has consistently returned ages of approximately 74 ± 2 ka (Oppenheimer, 2002).

Climatic impacts of the Toba eruption are related to the emission of stratospheric sulphur (Ambrose, 2003). Estimates of these emissions range from 2.3-4.7 x 1012 kg

(Zeilinski et al., 1996) to 3.5 x 1010 kg (Scaillet et al., 1998), a difference of two orders of magnitude. Atmospheric modeling using the upper range of sulphuric emission suggests a

5-6 year period of high stratigraphic sulphate aerosol abundance (Bekki et al., 1996).

Predictions of temperature changes resulting from the Toba eruption range from no more than those observed for previous ice age oscillations (Oppenheimer, 2002; Gathorne-

Hardy and Harcourte-Smith, 2003) to those associated with a nuclear winter (Rampino and Ambrose, 2000; Rampino and Self, 1992, 1993). In the latter scenario, regional temperatures are expected to drop ~10-15oC below normal, while global decreases of 3-

5oC could be expected (Rampino and Ambrose, 2000). Simulations of super-eruptions suggest that climatic conditions resulting from volcanic aerosol are strongly dependent on the season of eruption (Timmreck and Graf, 2006).

Climatic changes associated with Toba could have resulted in a volcanic winter of such magnitude as to devastate vegetation and cause a human population bottleneck

(Ambrose, 1998; Rampino and Self, 1992, 1993; Rampino and Ambrose, 2000; Rampino

2003). In modeling the effects of the eruption on vegetation, Rampino and Self (1992,

1993) utilized studies of the effects of a nuclear winter on the environment (e.g. Harwell

91 and Hutchinson, 1985; Harwell, 1984). Such studies suggest near complete decimation of above-ground tropical plants, and little better survival of deciduous and temperate forests

(Harwell et al., 1985). These studies, however, have little to say on the survivability of mammals caught in nuclear winter (e.g. Hutchinson et al., 1985; Harwell et al., 1985) beyond lamenting their probable demise. Studies of the effects of volcanic super- eruptions suggest that they alone would be insufficient in producing mass extinction events (Erwin and Vogel, 1992; Prothero, 2004). However, even proponents of the volcanic winter hypothesis have not claimed mass extinctions would follow on from

Toba (Ambrose, 2003). Given the magnitude and devastation associated with Toba, however, one would expect that regional extinctions would follow, especially if the ecological devastation posited for the volcanic winter were accurate. Species especially vulnerable to temperature changes and those dependent on tropical vegetation should, in particular, be adversely affected by the Toba super-eruption.

4.3 Methods

This analysis is restricted to Late Pleistocene (128-11 ka) sites from Southeast

Asia. Definitions of Southeast Asia, taxonomy and chronology follow those of Louys et al. (2007). Faunal lists were derived from the published literature (Table 4.1). A number of listings provide difficulties when examining whether particular taxa survived through the time of eruption, especially when analyses (such as this one) are focused at the specific level. The first of these are taxa unassigned above the level of genus. These specimens could in fact represent any member of its designated group, and thus provide limited information in terms of ecology and species extinctions. Those taxa unassigned

92 above the level of genus are therefore omitted from the list provided here. The second involves

Faunal list Reference Age (ka) Reference (faunal list) (chronology) Niah Medway (1972) 39-45 Barker et al. (2006) Hang Hum Olsen and 80-140 Olsen and (combined faunal Ciochon (1990) Ciochon (1990) list from Hang Hum I and Hanh Hum II) Keo Leng Olsen and 20-30 Olsen and Ciochon (1990) Ciochon (1990) Lida Adjer de Vos (1983) 118-128 Based on correlation with Punung, as per de Vos (1983) Sibrambang de Vos (1983) 118-128 Based on correlation with Punung, as per de Vos (1983) Kedung Brubus van den Bergh 700-800 van den Bergh et al. (2001) et al. (2001) Ngangdong van den Bergh Late See text et al. (2001) Pleistocene? Punung van den Bergh 118-128 Westaway et al. et al. (2001) (2007) Jiande Han and Xu 90-117 Etler (1991) (1985) Liujiang Han and Xu 68-153 Shen et al. (1985) (2002)

Table 4.1: List of sites discussed in text, their respective dates and corresponding references.

those species assigned to genus level, but unassigned to any particular species. This can occur for a number of reasons, including (but not limited to): 1, the material could be too incomplete to determine species, while still being informative at the generic level; 2, the author describing the material is unfamiliar with the particular group examined; or 3, it could possibly represent a new species, although not enough material is preserved to be

93 sure. (Note these are not mutually exclusive.) Difficulties arise when trying to determine if these taxa survived the extinction: the unassigned individual may belong to an established species (either extinct or extant), or could be a new species (in which case it is likely that any such species would be extinct). Those individuals unassigned at the species level may represent the only example of that genus at that time or even in that country. Given these different scenarios, should these taxa be scored as present or absent in the Holocene? The approach taken in this study is to list taxa assigned only to the generic level at the sites from which they were recovered (Table 4.2). Holocene presence for those genera finding a modern representative is marked with “*”. The final problematic listing is that of a conferred species. In this analysis a taxon designated “c.f.” is treated as the conferred species, and any Holocene presence is therefore based on that species.

Vietnam Sumatra Java Southern China Borneo

Jiande Jiande Punung Liujiang Liujiang Holocene Holocene Holocene Holocene Holocene Holocene Lida Ajer Keo Leng Leng Keo Hang Hum Hum Hang Niah Caves Niah Caves Ngangdong Ngangdong Sibrambang Sibrambang Kedung Brubus

PRIMATES Macaca cf. assamensis x x Macaca fascicularis x x x x x x Macaca cf. mulatta x x Macaca nemestrina x x x x x Macaca sp. x x * x x * x * Presbytis cristata x Presbytis melalophus x Presbytis sp. x x * Trachypithecus x x x x x cristatus Hylobates syndactylus x x x x Hylobates moloch x Hylobates sp. x x * x x *

94 Pongo pygmaeus x x x x x x x

CARNIVORA Nyctereutes sp. x * Cuon sp. x * Cuon alpinus x x Ursus thibetanus x x x x x Ursus malayanus x x x x x Ursus sp. x * Ailuropoda x x x x melanoleuca Arctonyx collaris x x x x x Lutrogale perspicillata x Paradoxurus cf. x x x hermaphroditus Paradoxurus sp. x * Paguma larvata x x x x Crocuta crocuta x x Panthera tigris x ? x x x x x x x x Melogale orientalis x Lutra sumatrana x Arctictis binturong x Felis bengalensis x Mustela nudipes x

PROBOSCIDEA Stegodon orientalis x x x x x Stegodon x x trigonocephalus Elephas maximus x x x x x x x Elephas hysudrindicus x x Palaeoloxodon cf. x namadicus Palaeoloxodon x namadicus

PERISSODACTYLA Tapirus indicus x x x x x x x x x x Megatapirus augustus x x x x x Rhinoceros sondaicus x x x x x x x Rhinoceros sinensis x x x x Rhinoceros unicornis x Dicerorhinus x x x x x x sumatrensis

ARTIODACTYLA Sus barbatus x x x x x Sus scrofa x x x x x x x x x Sus sp. x * x * x x *

95 Sus cf. barbatus Sus cf. lydekkeri x x Sus macrognathus x x Sus cf. officinalis x x Hexaprotodon x x sivalensis Cervus unicolor x x x x x Cervus sp. x x * x x * x x * x * Axis porcinus x x Muntiacus muntjak x x x x x x x x x x Muntiacus sp. x * x * Bos frontalis x x Bos javanicus x x x x ? x Bos sauveli x Bos palaeosondaicus x x Bos gaurus x x x Bubalus bubalis x x x x x x x x Bubalus sp. x * Bubalus palaeokerabau x x Tragulus napu x Duboisia santeng x Naemorhedus x x x x x x x sumatraensis Epileptobos x groeneveldtii Ovis sp. x

PHOLIDOTA Manis javanica x Manis palaeojavanica x x

Table 4.2: Species list of sites discussed in text; x represents presence, * represents generic presence during Holocene. List drawn from references in Table 4.1 for Pleistocene sites and from references cited in Methods for Holocene.

The chronology of the sites was obtained from the published record (Table 4.1).

All sites discussed have been included. Any disparity in accepted chronology is explored

in the discussion. Holocene distribution of mammals was compiled from Corbet and Hill

(1992), Nowak (1999), van den Bergh et al. (2001), and Tong and Liu (2004).

This analysis is restricted to large bodied taxa, in this case represented by six

orders: Artiodactyla, Proboscidea, Perissodactyla, Carnivora, Primates and Pholidota.

96 This was for two reasons. Firstly, most excavations described from the region, completed earlier in the last century, did not generally attach significance to microfauna, resulting in a bias towards large, charismatic species. Furthermore, large bodied taxa are easiest to recognize in the paleontological record, leading to faunal lists where only large-bodied taxa are significantly represented. Secondly, larger bodied taxa are more susceptible to extinction (Cardillo et al., 2005): they generally have lower reproduction rates, longer gestation periods, attain sexual maturity more slowly and have lower population densities

(Cardillo et al., 2005). Recovery from a catastrophic event should favor those animals who can re-establish their population quickly.

97 4.4 Results

Figure 4.1: Map showing the approximate locations of fossil sites preserving an extinction signature for Toba. Pie graphs represent proportion of species not represented after the eruption for each respective site. Extinct species for each country are illustrated in the boxes next to the pie graphs. Circles show recovered ash-fall deposits in the region (after Oppenheimer, 2002). Not to scale.

4.4.1 Borneo

The site of Niah Great Caves, Borneo, postdates the Toba super-eruption (Table

4.1). Therefore, any mammals found at Niah would necessarily have survived the

98 climatic after affects of the eruption. Of note among those taxa found at Niah are two species each of langur and macaque (Presbytis and Macaca), and one each of gibbon and orangutan (Hylobates and Pongo) (Table 4.2).

4.4.2 Vietnam

The site of Hang Hum (Figure 4.1), based on biochronological dating (Table 4.1), would pre-date the Toba eruption. Seven species recorded by Hang Hum (representing

26% of species from the site) are now extinct in Vietnam. These are the orangutan

(Pongo pygmaeus), two species of proboscidean (Stegodon orientalis and Palaeoloxodon namadicus) the giant tapir (Megatapirus augustus), the extinct Chinese rhino (Rhinoceros sinensis) and two species of pig (Sus lydekkeri and Sus officinalis). However, of these, only Sus officinalis, Palaeoloxodon namadicus are not represented at Keo Leng (Table

4.2), another Vietnamese site dated well after Toba (Table 4.1). Therefore, the only species occurring at Hang Hum whose subsequent loss could be attributed to the Toba eruption is a suid and an elephant, neither of which could be considered as rainforest obligates. In addition to the species mentioned above, many species found in Hang Hum are still extant in Vietnam. These include several species of macaque and gibbon

(Macaca and Hylobates), all the Carnivora listed, the tapir (Tapirus indicus), pig (Sus scrofa), a deer (Cervus unicolour) and two bovids, Bos gaurus and Bubalus bubalis.

4.4.3 Sumatra

Based on faunal similarity, Lida Ajer (Figure 4.1) is considered contemporaneous with Punung from Java (de Vos, 1983), which has a suggested age of 118-128 ka

(Westaway et al., 2007). Therefore it pre-dates the Toba eruption. Louys et al. (2007)

99 incorrectly stated that the leopard Panthera pardus becomes extinct from Sumatra during the Late Pleistocene; however, as Meijaard (2004) points out, there are no reliable fossils of this felid from Sumatra. The only species described from Lida Ajer which could have become extinct is the bovid Bos javanicus, although the domesticated form of this animal is currently found on the island (Corbet and Hill, 1992). Lida Ajer therefore records at least 13 species which survived the extinction (Table 4.2).

Sibrambang (Figure 4.1), shares the majority of its species with Lida Ajer, and as such is most likely contemporaneous (de Vos, 1983). In addition to species shared with

Lida Ajer, the tiger (Panthera tigris), the Javan rhino (Rhinoceros sondaicus) and a buffalo (Bubalus bubalis) are also found at Sibrambang, all of which are currently extant in Sumatra.

4.4.4 Java

Java hosts two sites which likely pre-dated the Toba eruption, and therefore provide an indication of the impact of the eruption on the mammalian fauna. The first of these is

Punung (Figure 4.1), which has recently been dated to 118-128 ka by luminescence and uranium series dating (Westaway et al. 2007). The second site, that of Ngangdong, is more insecurely dated. Swisher et al. (1996), on the basis of ESR and U-series dating of bovid teeth have provided an age of ~27-53 ka, however this date has been questioned by van den Bergh (1999), who considers an age of ~ 135 ka more reasonable. Certainly, the

Ngangdong fauna is extremely similar to the older (0.8 – 0.7 Ma) Kedung Brubus fauna, sharing with it ten (out of eleven) of its species (Table 4.2). In addition to this, species recorded from Kedung Brubus and Ngangdong evince very different environments to those of Punung. As van den Bergh (1999) points out, it is far less parsimonious for the

100 faunal succession and paleoenvironment of Java to shift from an archaic fauna representing an open forest environment (Kedung Brubus) to a more derived rainforest fauna (Punung), reverting back to an archaic fauna (Ngangdong), then developing the more modern fauna as represented by sites such as Wajak (see Storm (2001), for a description of Wajak). Particularly if, as van den Bergh (1999) suggests, the layer dated by Swisher et al. (1996) consists of reworked material intruding into the older, and principal fossil bearing layers. Given these doubts on the dating recorded by Swisher et al. (1996), and given the similarity in fauna between Kedung Brubus and Ngangdong, any further discussion of the extinctions herein will be based on the basis of chronology of sites proposed by van den Bergh et al. (1999), i.e. Ngangdong predates Punung.

Therefore Ngangdong is not further considered here.

Those species listed from Punung which are currently no longer found in Java include a macaque (Macaca nemestrina), a gibbon (Hylobates syndactylus), the orangutan

(Pongo pygmaeus), a bear (Ursus malayanus), the serow (Naemorhedus sumatraensis) and a pig (Sus barbatus) (Figure 4.1). Those which are found on Java after the eruption include the tiger (Panthera tigris), the Asian elephant (Elephas maximus), the tapir

(Tapirus indicus), the Javan rhino (Rhinoceros sondaicus), a pig (Sus scrofa), a muntjac

(Muntiacus muntjack) and a buffalo (Bubalus bubalis).

4.4.5 Southern China

China is host to possibly the largest number of Pleistocene sites in the region. However, the only sites examined for the purpose of this analysis are those which most likely and most closely pre-date the Toba eruption. Therefore, of the sites considered characteristic of Late Pleistocene Southern China by Han and Xu (1985), only two are considered here

101 – Jiande and Luijiang (Figure 4.1). Based on their accepted chronology (Table 4.1),

Jiande completely pre-dates the eruption, and while the range of dates for Liujiang encompasses the date of the Toba eruption, it was also most likely deposited prior to the eruption. Jiande records the extinction of two species – the archaic elephant

Palaeoloxodon namadicus and the Chinese rhino Rhinoceros sinensis – the latter representing the only extinction recorded by Luijiang (Figure 4.1). Species recorded from

Jiande which survived until the Holocene include a bear (Ursus thibetanus), a hyena

(Crocuta crocuta), badger (Arctonyx collaris), a stegodon (Stegodon orientalis), the giant tapir (Megatapirus augustus) and the giant panda (Ailuropoda melanoleuca), with the latter three also represented by Liujiang.

4.5 Discussion

4.5.1 General discussion

A surprisingly limited number of extinctions are recorded during the time of the Toba eruption. Vietnam saw the extinction of a pig and elephant, Southern China the extinction of an elephant and a rhinoceros, while Sumatra seems to have recorded no extinctions at all. Java provides by far the largest list of extinct species: two species of monkey, one ape, one bear, one serow and a species of pig (Figure 4.1).

Of significance is the fact that the vast majority of these extinctions do not involve characteristic rainforest taxa. The vast majority of animals which do become extinct appear to be characteristic of open forests and environments, in particular the elephants and bovids.

102 Rainforest taxa seem to be relatively unaffected in distribution following the Toba eruption. The orangutan survived in Borneo, Vietnam and Sumatra, while the crab-eating macaque (Macaca fascicularis) survived in Borneo, Java and Sumatra. At least one other species of macaque, the pigtail macaque, also survives on Borneo and Sumatra. Gibbons, represented by the genus Hylobates, survive in Borneo, Vietnam and Sumatra. These apes are arboreal obligates, and occur predominately in evergreen and deciduous monsoon rainforests (Nowak, 1999). The Asian tapir, Tapirus indicus, as well as its giant relative,

Megatapirus augustus, do not seem to have been adversely affected by the eruption, surviving in Borneo, Vietnam, Java and Sumatra in the case of the former, and Vietnam and Southern China in the case of the latter. While almost nothing is known of the ecology of the giant tapir, the little that is known of the Asian tapir suggests it has a preference for flat and damp environments and is predominantly restricted to closed forests (Novarino et al., 2005). It generally prefers secondary forest to primary and/or disturbed forests (Novarino et al., 2005), and the most serious threat to its survival remains deforestation (Holden et al., 2003).

The list of surviving fauna is not only restricted to rainforest taxa. A species of elephant (Stegodon orientalis) survives in Vietnam, the former also surviving in Southern

China. Several species of suids, cervids and bovids also survive, all of which are usually associated with less densely wooded environments. The fact that most carnivores survive is also not surprising given the abundance of carcasses following any catastrophe

(Harwell et al., 1985), coupled with the suggested relatively short period of vegetative recovery (Harwell, 1984). The ensuing grace period for carnivores in this case would, however, be necessarily short, and any extension in the recovery time of flora (and

103 corresponding fauna) would likely adversely affect carnivores (Harwell et al., 1985).

Gathorne-Hardy and Harcourte-Smith (2003) also noted the survival of nine endemic species of rainforest obligates on the Mentawai islands, situated 350km south of Toba.

The nature of these survivals suggest one of two possibilities, depending on whether we accept that the aftermath of the volcanic eruption can be likened to a nuclear winter and the environmental response suggested for this event.

4.5.2. Volcanic winter

If we accept that the impact of the volcanic winter produced by a super-eruption can be likened to a nuclear winter, and that the hypothesized aftermath (see Rampino and

Self, 2000, for a detailed description) will be, although highly speculative, at least within reason, then this suggests that fauna in the immediate area of the volcanic eruption are surprisingly resilient. The devastation of vegetation in the region would surely affect the vast majority of each respective community’s population, and it is likely that the surviving fauna would do so only in geographically isolated and environmentally protected refugia.

Perhaps because of the severe structural damage and ensuing human misery associated with catastrophic events, studies into the effects of catastrophes on mammals have been few (Willig and McGinley, 1999). Modern analogues to the Toba eruption are necessarily magnitudes smaller than Toba, and include the eruption of Krakatau and

Mount St Helens, and also possibly the Chernobyl disaster. The 1883 Krakatau eruption is thought to have extinguished all life on its associated islands directly following its eruption (Simkin, 1983; Gathorne-Hardy et al., 2000). In its immediate vicinity this is thought to result from the falling tephra, lava and hot gases. Heavy tephra falls are also

104 thought to be responsible for the felling of trees within a wider area, at least 19km

(Gathorne-Hardy and Harcourte-Smith, 2003). The Mount St Helens eruption resulted in the deaths of 5000 deer (Odocoileus hemionus), 1500 elk (Cervus canadensis), 200 black bears (Ursus americanus) and 15 mountain goats (Oreamnos americanus) (Del Moral and Grishin, 1999). In the case of the Chernobyl disaster, the first plume released by the explosion resulted in the loss of 400 hectares of pine forest west of the reactor and the extermination of almost all avian and mammalian fauna within five kilometres (Chesser et al., 2005).

The ability of animals to survive disturbances is dependent on the timing, magnitude and intensity of the disturbance as well as the phenotype and life-history of the animal (Willig and McGinley, 1999). Studies into catastrophic disturbances have demonstrated that even the largest of these do not produce “blank slates”, but leave mosaic patches of different species composition (see Platt and Connell, (2003) and references therein). Although these studies primarily deal with flora, this should equally be applicable to fauna. In fact, animal populations can recover rapidly after volcanic events, and many species thrive in post-eruption environments providing forage recovers quickly (Del Moral and Grishin, 1999, Edwards, 2005).

Many species can avoid death by hiding out in refugia. In the path of destruction produced by the Chernobyl explosion, for example, many such refugia were created from which local animal communities were able to repopulate the region (Chesser et al., 2005).

Flora and fauna continue to demonstrate a surprising ability to recuperate from catastrophe; note the recolonization of Krakatau in less than 100 years (Gathorne-Hardy et al., 2000); and as Chesser states with respect to the Chernobyl disaster “ecosystems

105 [inside the exclusion zone] have shown remarkable recovery and resilience to this insult”

(Chesser, 2006, p. 7). At Mt St Helens, fauna re-entered the zone of devastation within days of the eruption, and recolonisation of the blast zone occurred within five years of the eruption (Edwards, 2005). Extrapolating these studies to the Toba eruption, and given the low extinction signature discussed above, it would appear that fauna have an underappreciated capacity for recovery from catastrophe. If mammals did become extinct as a result of Toba, it is likely these animals were already very much in decline, and that the eruption only pushed them over the edge. This has implications for other studies of megafauna extinctions: the hypothesis that humans were responsible for the extinction of large-bodied mammals in the Late Pleistocene in a quick blitzkrieg fashion (e.g. Martin,

1984) is rendered less likely given the ability for most animals to survive sudden and catastrophic events.

The volcanic winter hypothesis has been used as a causative model explaining the hypothesized Late Pleistocene human bottleneck (e.g. Ambrose, 1998, 2003; Rampino and Ambrose, 2000). In the scenario proposed by these authors, climatic changes resulting from the Toba eruption would have cut the global human population down to between 3000 and 10,000 individuals (Rampino and Ambrose, 2000). Recovery of humans is thought to be the result of an unprecedented adoption of cooperative skills and communication strategies (Ambrose, 2003). Evidence of population expansion of the chimpanzee at ~ 67ka is also cited as supportive evidence (Ambrose, 2003; Rampino and

Ambrose, 2000). Furthermore, orangutan diversity in Borneo seems to indicate a period of recovery following the Toba eruption (Steiper, 2006, although Steiper acknowledges this could just as likely result from environmental fluctuations). Based on the current

106 paleontological record, however, no mammals appear to have become extinct on Sumatra immediately following the eruption, either as a result of the initial blast, or the proceeding climatic changes. While the deposits of Sibrambang and Lida Ajer could post-date the eruption, it is perhaps more likely they pre-dated it. While Java, Sumatra and Borneo could have been repopulated following their connection to the mainland during periods of lower sea-level following the eruption, this is not the case for the Mentawai Islands, which preserve nine rainforest endemics (Gathorne-Hardy and Harcourte-Smith, 2003).

The perceived human monopoly in overcoming ecological disasters should be reconsidered. If the emerging picture from the paleontological record is correct, then there is no need to evoke unprecedented levels of corporation within prehistoric societies when calculating the response of humans to ecological adversity. It appears many mammals in the immediate area, without the need to revert to unusual or unprecedented behavior, were able to cope with resultant climatic changes without visible effect on their distributions.

4.5.3 No volcanic winter

The other alternative is if the climatic and ecological damage resultant from Toba is not considered comparable to that of a nuclear disaster. While the initial impact of the blast would certainly have been catastrophic, in this case the resultant changes to climate are not seen as permanent or as global as that predicted above. The survival of many taxa, some quite sensitive to deforestation, indicates that although vegetation may have been affected by the eruption, enough trees would have survived to allow their dependent fauna to also survive. In fact, forests are generally resilient to tephra events, since the growth form diversity of trees improves the possibility that some individuals will survive

107 (Del Moral and Grishin, 1999). This finding casts doubts on the theory that human populations would have been reduced to a few thousand individuals as a result of the volcanic eruption (Rampino and Ambrose, 2000). If the level of deforestation and climatic impacts were not sufficient to cause major extinctions or significant range reduction in the taxa studied here, it is unlikely that it would be sufficient in and of itself to reduce the global human population as has been suggested, these being furthermore restricted to the African continent (Rampino and Ambrose, 2000). While it is possible that such a population crash could be expected for Southeast Asia, it is unlikely to have affected the populations of Eurasia or Africa to quite the same extent.

The immediate aftermath of the volcanic eruption would have been quite dire for vegetation and fauna alike, especially regionally. While the paleontological record for

Southeast Asia is still relatively scarce, it is not uninformative. The immediate period of devastation would likely be short-lived, as suggested by the high survivability of mammalian megafauna. However, there is no need to evoke Toba as the sole extinction agent: analysis of sediment cores of Indonesia suggests wetter conditions developed from

74 – 47 ka (van der Kaars and Dam, 1995), possibly contributing to the extinction of the elephants and bovids. Thus, while the Toba eruption certainly would have produced less favorable conditions, these conditions would not have been outside the range of tolerance for most species. In fact, simulations of the climatic effects of super-eruptions suggest that although dramatic changes of climate could be expected following an eruption event, these would be short-lived (in the order of decades), and not enough in themselves to bring about a glacial cycle (Jones et al., 2005; Jones et al., 2007). The volcanic winter scenario is certainly possible, and is not discounted here; however, based on the

108 survivability of the mammals studied, it is more likely that the global effects of this eruption have been overestimated.

4.6 Conclusions

The ecological aftermaths of super-eruptions are understandably difficult to examine directly. Even when massive eruptions have occurred within recorded history, their effects on mammals have been little studied. Analyses such as this one are often the only means of exploring the ecological effects of super-eruptions on fauna. While the data presented here are not as constrained as could be hoped, the results are nevertheless suggestive of a far greater resilience of mammals coping with ecological adversity. Based on the current paleontological record, the total number of species recorded for the period of the super-eruption of Toba is small. While a mass extinction event for the region as a result of the eruption has been discounted by a number of authors (Ambrose, 2003; Erwin and Vogel, 1992), the small number of extinctions recorded is still surprising, particularly given the ferocity of the eruption event, and its proposed ecological effects. Two different scenarios are explored in this paper based on whether these proposed ecological effects are accepted. If the volcanic winter did proceed after the eruption, then this study suggests the previously underappreciated ability of mammals to recover from catastrophic events, and has implications for other mass extinction events, especially where large-bodied taxa feature prominently. If the volcanic winter were not as dire as predicted, then it is unlikely that the human bottleneck observed was caused by the super- eruption of Toba and the resultant climate change, and that as noted by other authors (e.g.

Prothero, 2004), the ecological effects of super-eruptions are not critical to mammalian extinctions.

109

PART 2

PALAEOECOLOGY

“Not only does the marvellous structure of each organised being involve the whole past history of the earth, but such apparently unimportant facts as the presence of certain types of plants or animals in one island rather than in another, are now shown to be dependent on the long series of past geological changes--on those marvellous astronomical revolutions which cause a periodic variation of terrestrial climates--on the apparently fortuitous action of storms and currents in the conveyance of germs--and on the endlessly varied actions and reactions of organised beings on each other. And although these various causes are far too complex in their combined action to enable us to follow them out in the case of any one species, yet their broad results are clearly recognisable; and we are thus encouraged to study more completely every detail and every anomaly in the distribution of living things, in the firm conviction that by so doing we shall obtain a fuller and clearer insight into the course of nature”

- Alfred Russel Wallace 1880

CHAPTER 5

PALAEOECOLOGY MULTIVARIATE METHODS:

EXAMINING MODERN HABITAT BIASES AND

APPLICATION TO GONGWANGLING HOMININ SITE,

CHINA1

5.1 Abstract

The palaeoecology and zoogeography of the large-bodied mammals from Gongwangling,

Lantian County, China (1.15 Mya) are examined. The palaeoenvironment is inferred through the use of multivariate methods. Results indicate that Gongwangling represents a closed forest environment. This study is the first time in which multivariate methods have been used in an Asian fossil site. Methods for mitigating limitations associated with the use of species lists from national parks and nature reserves in multivariate palaeoecological analyses are implemented. Habitat descriptions for the species lists provided are not accepted a priori.

Rather, cluster and principle components analyses enable the lack of consistency in habitat descriptions to be taken into account. In addition, taxonomic distinctness measures are used as a basis of testing for possible taxonomic bias in the published reserve and park species lists.

1 This chapter is submitted as: Louys, J., Tong., H., Bassarova, M., Travouillon, K. J. Palaeoecology and zoogeographic identity of the Gongwangling hominin site, China, and addressing limitations of national parks and nature reserves as palaeoecological analogues. Palaeogeography, Palaeoclimatology, Palaeoecology. HT facilitated access to fossil material, verified systematics and provided assistance in translating key texts. MB provided the idea of how to treat biases in descriptions of vegetation of modern habitats. KJT and I contributed equally to adapting the taxonomic distinctness measure for use in palaeoecology. The remainder of this chapter is my own work.

111 An analysis of the zoogeography of mammals from Gongwangling shows differences in identity along a body-size gradient, with large-bodied faunas being largely sub-tropical and small bodied mammals largely palearctic. Coupled with their different relative stratigraphic positions, this implies different depositional histories and environments for these two faunas. This was confirmed by preliminary taphonomic observations that suggest different preservational biases for small versus large mammal taxa. The results suggest that the large-bodied mammals were deposited during a warm, humid phase of the Early

Pleistocene.

The site of Gongwangling, both preceded and followed by typically palearctic faunas, is unique for the region. It represents one of the most northern occurrences of typically

Southeast Asian megafauna in China. The absence of these megafauna in more recent sites in the region is consistent with models of environmental change during the Pleistocene of

China. Hominins, however, continue to occur in later sites, suggesting their continued presence in the region throughout the change from sub-tropical to palearctic conditions.

5.2 Introduction

The Pleistocene in China was a period of great climatic and environmental changes, often precipitated by an active tectonic regime present in the region, in turn (partly) a result of the continuing collision between the Indian and Eurasian plates (Ferguson, 1993).

Climatic, environmental and geological changes throughout the Pleistocene have affected the flora and fauna of the region dramatically. In particular, the continuing uplift of the Tibetan

Plateau has resulted in an increase in seasonality and the southward migration of more temperate and sub-tropical zones (Jablonski and Whitford, 1999; Ferguson, 1993).

Furthermore, the long-recognized (e.g. Pei, 1957, 1958) north/south biogeographical divide

112 in Eastern China was suggested to be advanced by the continued uplift of the Qinling

Mountain chain, as it established a physical barrier for the dispersal of plants and animals

(Zhu and Zhou, 1994), and which now demarcates, along with the Huai River, the temperate and sub-tropical zones (Figure 5.1).

Within this complex ecological backdrop, hominins first enter the region. Their interaction with, and dependence on, various Chinese faunas is still not well understood, and it is still equivocal as to whether these individuals participated in a predominately scavenging or hunting subsistence (e.g. Boaz et al., 2004; Louys et al., 2007). That the palaeoecology of

Pleistocene faunas associated with early Chinese humans has received fairly limited treatment (e.g. Zhu and Zhou, 1994), especially when compared with the much better studied

African faunas, does not help to alleviate this ambiguity.

Figure 5.1: Location of Gongwangling with respect to the Qinling Mountains. The current division between northern (palearctic; temperate) and southern (oriental; subtropical) China is also indicated.

Gongwangling (GWL), Lantian county, Shaanxi Province, is an Early Pleistocene site from Northern China bearing hominin remains and artifacts, as well as other mammalian faunas. It is situated on the northern slope of the Qinling Mountains (Figure 5.1). Hominins are represented by an almost complete skull cap and an isolated left upper second molar

113 found in association with the large-bodied mammal fauna. They were assigned to Homo erectus, and most likely represent one individual (Wu, 1964; Wu, 1966). Twenty stone artifacts have also been recovered from GWL (Dai, 1966, Dai and Xu, 1973), although these were recovered from a stratigraphically higher layer than the hominin material. The other mammalian fauna found in association with the human skull provides an unusual and surprising mix of both northern, steppe-adapted faunas (e.g. Equus), and more typically southern, sub-tropical animals (e.g. Tapirus).

This study focuses on two aspects of the GWL fauna. Firstly, the zoogeographic identity of the site (i.e. whether it should be classified, based on faunal composition, with northern or southern Chinese sites) will be examined. In particular, the question of whether

GWL represents an unusual incursion into the north by warm adapted species or whether they represent a relict population not found in younger deposits will be addressed. Secondly, this study will seek to examine the palaeoecology of the GWL megafauna using discriminant functions analysis. Although this method has been utilized for both European (e.g. Mendoza et al., 2005) and African (e.g. Reed, 1997, 1998) faunas, it has not until now been employed for any Asian fossil site. In addition to providing an indication of the structure of vegetation present during the deposition of GWL, a methodological framework for palaeoecological studies will be developed for future use by palaeontological researchers of Pleistocene China.

In order to properly conduct these analyses, the taphonomic bias present at GWL will be assessed. Taphonomic observations are also noted. The results of this study have biogeographical implications for the distribution and dispersal patterns of Pleistocene

Chinese mammals.

114 5.2.1 Geological context

The site of GWL is situated on the northern slopes of the Qinling Mountains, Lantian

County, China (Figure 5.1). The profile of GWL is represented by 12 clay units which contain five paleosol layers, underlain by a basal gravel unit dating to the Early Pleistocene

(Wu et al., 1966, Wu and Poirer, 1995). The unit containing the hominin and other mammalian remains is a light yellow-grey laminated of silty loess 2.6m thick. The hominin and large-bodied mammal material was found associated with medium sized to large carbonate concretions in the middle and upper parts of this layer. Wu and Poirer (1995) indicated that these fossils had accumulated in piles, but we found no further information to confirm this. Taken at face value, it suggests a high energy deposit. The small mammal fauna was found in the lower part of this section, and appear unassociated with the carbonate concretions. The fossil bearing clay layer has been correlated with the middle sandy loess of the Lower Lishi formation (L15). This loess has been dated to 1.15 Ma on the basis of palaeomagnetism (An and Ho, 1989; An et al., 1990).

5.2.2 Previous palaeoenvironmental analyses

Previous palaeoenvironmental reconstructions of GWL have provided a mixed picture at best. The first reconstructions, based on the large fauna present at the site, indicated a warm sub-tropical climate (Chow et al., 1965; Hu and Qi, 1978). These reconstructions were heavily influenced by the presence of the many southern species present in the site, but were only qualitative in nature. In particular, the presence of many forest animals at the site

(such as tapirs (Tapirus and Megatapirus), monkey (Rhinopithecus) and tiger (Panthera sp. cf. P. tigris)) indicated to these authors the presence of a sub-tropical forested environment.

115 Pollen records, however, recorded the presence of a warm temperate forest habitat

(Hsu, 1966). This reconstruction was based, in particular, on pollen recovered from pine

(Pinus), hackberry (Celtis), elm (Ulmus), oak (Quercus), hornbeam (Carpinus) and birch

(Betula) (Hsu, 1966).

Further apparent discrepancies in palaeoenvironmental interpretation occurred when the loess record was examined. GWL, being correlated the middle sandy loess (L15) of the

Luochuan section, a section most likely formed in cold, dry conditions during a glacial period

(Liu, 1985; An and Ho, 1989). In an attempt to synthesize the conflicting data, Zhu and Zhou

(1994) proposed that because the hominin and other mammalian fossils accumulated in a carbonate-cemented horizon within the sandy loess, the deposit must have formed during a warm and humid period that followed after the deposition of L15. Therefore, GWL would represent a transitional environment between a cold, steppe-like environment to a warmer, sub-tropical environment. Further work, involving studies on the stable isotopes present in the GWL palaeosol, indicated the presence of a warm, temperate forest, with a probable cold, dry winter and mild, semi-humid summer (Wang et al., 1997).

5.3 Materials and Methods

5.3.1 Materials

The fossil mammals from GWL were originally described by Hu and Qi (1978)

(Table 5.1). The detail in the descriptions varied between mammal groups, as did the cataloguing of species, with in some cases a single catalogue number was used for many elements of the same species (Table 5.1). Resulting inferences based only on the published descriptions are therefore problematical, and a direct examination of all fossil material would be best. Unfortunately, not all the fossils described by Hu and Qi (1978) were available for

116 examination by the authors. The specimens directly examined are indicated in Table 5.1, and represent approximately 10% of the originally described material. A combination of first hand observations and published results are therefore used throughout this study. Since the original publication of the species descriptions, several of the groups have undergone taxonomic revision. Along with the original faunal list provided by Hu and Qi (1978), given in Table 5.1, taxonomic revisions and accompanying references are also listed. Throughout this study, the updated taxonomy is used. All material studied is housed in the Institute of

Vertebrate Palaeontology and Palaeoanthropology (IVPP), Beijing, China.

117

Order Family Species Original Taxonomic Material Described by Hu Material Catalogue MNI designation (Hu Reference and Qi (1978) observed numbers and Qi, 1978) Insectivora Talpidae Scaptochirus 2 incomplete right mandibles 5 isolated V5416.1; 2 moschatus teeth, 2 V5416.2 mandibular fragments Primates Cercopithecidae Rhinopithecus Megamacaca Jablonski incomplete mandible with left V2934.1; 1 lantianensis lantianensis and Gu P3-M3 and right C, M1-M3; V2934.3; (1991) right maxilla with P3-M3; left V2934.3; maxilla with C-M2; isolated V2934.4 RM1; isolated LM1 Rodentia Cricetidae Proedromys sp. Arvicola terra- Zheng and Left M2 and M3 V5395 1 cf.P. pliocaenicus rubrae Li (1990) Rodentia Cricetidae Allophaiomys sp. Microtus Zheng and 5 broken mandibles with M1- V5396 5 cf. A. epiratticeps Li (1990) M2; 16 isolated lower molars pliocaenicus Rodentia Cricetidae Yangia tingi Myospalax tingi Zheng Broken mandible M1-M3 1 partial V5397 1 (1993) mandible Rodentia Cricetidae Eospalax Myospalax Zheng Almost complete skull; 4 1 almost V5398.1; 3 fontanieri fontanieri (1997) broken maxillas complete V5398.2 skull, 2 isolated teeth Rodentia Cricetidae Eospalax sp. Myospalax sp. Zheng 1 right humerus 1 humerus V5399 1 (1997) Rodentia Cricetidae Cricetulus sp. 2 RM1; 1 LM1; RM2; LM2; 2 2 partial V5400 2 cf.C. griseus mandibles with LM1-3 mandibles Rodentia Cricetidae Allocricetus Cricetulus Zheng 1 right mandible with I, M1-3 1 partial V5401 1 teilhardi varians (1984) mandible Rodentia Cricetidae Bahomys Broken right maxilla with M1- V5403.1; 5 hyposodonta 3; broken right maxilla with V5403.2 M2-3; 5 broken left mandibles with M1-3; broken right mandible with M1-3

118 Rodentia Cricetidae Cricetulus sp. Broken illium; distal left V5402 1 femur; distal left tibia; right metatarsus; 2nd right phalanx Rodentia Cricetidae Meriones sp. Gerbillus sp. Broken left mandible with M2-3 1 partial V5406 1 mandible Rodentia Sciuridae Petaurista sp. RM2 1 isolated V5394 1 tooth Rodentia Muridae Apodemus sp. Broken right mandible with I, 1 partial V5404 1 M1-3 mandible Rodentia Hystricidae Hystrix sp. cf. H. Hystrix sp. cf. C. Corbet and Broken skull with RP4, M1-2; V2935 1 brachyura subcristata Hill, 1992 right mandible with P4, M1-2; RM1; some incisors and fragments Lagomorphia Ochotonidae Ochotonoides Ochotona Almost complete skull with V5407 1 complicidens complicidens mandible; 2 other pieces of broken skull; 10 mandibular pieces; left femur Lagomorphia Ochotonidae Ochotona sp. 2 broken mandibles with P3- 1 partial V5408 1 cf.O. thibetana M2 mandible Carnivora Canidae Canis lupus Canis variabilis Corbet and Broken skull with RP1-4, LP2- 7 V2936 4 Hill, 1992 M1; 2 right broken mandibles; mandibular 1 left broken mandible; several fragments, 7 isolated teeth isolated teeth, 3 maxillary fragments Carnivora Ursidae Selenarctos Ursus sp. cf. U. Broken maxilla with RP4-M2, V2938 1 thibetanus etruscus LP4-M1; isolated RM2 Carnivora Ursidae Ailuropoda LM1 1 isolated 1 melanoleuca tooth Carnivora Mustelidae Meles sp. cf. M. Meles sp. cf. M. ? Complete left mandible with 2 V5411; 2 leucurus meles some incisors missing; isolated mandibular V5410 LP3; isolated LP4 fragments Carnivora Hyaenidae Pachycrocuta Hyaena sinensis Colbert and Almost complete skull; broken 1 almost V2939.1; 3 sinensis Hooijer, maxilla with LP4 and RP4; complete V2939.2; 1953 broken maxilla with LDP3, skull, 1 V2939.3; LP4-M1 and RDP2, RP4-M1; mandibular V2939.4; broken left mandible with P2- fragment V2939.5

119 M1; broken left mandible with with P2-M1 DP2-DP4 Carnivora Felidae Panthera sp. cf. RP4 1 maxillary V5421 1 P. tigris fragment Carnivora Felidae Panthera pardus Broken left mandible with C- V2980 1 M1 Carnivora Felidae Sivapanthera Broken right mandible with P3- V5411.1; 1 pleistocaenicus M1; LP2 V5411.2 Carnivora Felidae Meganteron Almost complete right upper 2 isolated V2979.1; 1 lantianensis canine; broken left upper teeth V2979.2 canine 2 Proboscidea Stegodontidae Stegodon LM ; RM2; 2 broken lower 1 isolated V2978.1; 1 orientalis molars; left astragulus tooth, 1 V2978.2; autopodia V2978.3 Perrisodactyla Equidae Equus Almost complete skull; 28 V2940.1; 8 sanmeniensis seriously depressed skull; autopodia, 3 V2940.2; broken maxilla; 5 broken humeri V2940.3; mandibles with complete fragments, V2940.4; dentition; > 50 isolated teeth; 11 V2940.5 some broken limb and vertebra mandibular fragments, 3 maxillary fragments, 1 radius, 55 isolated teeth, 5 vertebral fragments Perrisodactyla Rhinocerotidae Dicerorhinus Dicerorhinus sp. Broken maxilla with P2-M2; 1 isolated V2942.1; 3 kirchbergensis cf. D. mercki left mandible with P3-4; broken tooth V2942.2; maxilla with DP1-DP4; 3 V2942.3; broken mandibles with DP1- V2942.4; DP4; some isolated teeth; distal V2942.5 ulna; almost complete radius; 2 left distal humeri; 2 right proximal humeri; 2 distal femurs; 2 proximal femurs; left distal tibia; right

120 metatarsal; left metatarsal; right carpal Perrisodactyla Rhinocerotidae Dicerorhinus Almost complete skull with 1 humerus V5413 1 lantianensis most teeth in position Perrisodactyla Tapiridae Megatapirus RP3 (or possibly P4) V5414 1 augustus Perrisodactyla Tapiridae Tapirus sinensis LP4-M2 1 maxillary V2941 1 fragment Perrisodactyla Chalicotheriidae Hesperotherium Nestoritherium Qiu, 2002 Broken maxilla with P4-M2 1 maxillary V5417 1 sinense sinense fragment Artiodactyla Suidae Sus lydekkeri 2 mandibles 2 cranial V2943.1; 2 fragment, 3 V2943.2 mandibular fragments, 1 radius Artiodactyla Cervidae Elaphodus 2 mandibles V5418 1 cephalophus Artiodactyla Cervidae Cervus nippon Pseudaxis Corbet and Most dominant with > 20 3 antler V5419 25 grayi Hill, 1992 mandibles and maxillas; 2-3 fragments, pieces of antler 33 maxillary fragment, 1 autopodia, 12 cranial fragments, 53 mandibular fragments, 1 rib, 30 isolated teeth Artiodactyla Cervidae Sinomegaceros Almost complete right antler; 1 V2944; 2 konwanlinensis left antler; broken mandible mandibular V5414; without teeth; 2 right fragment, 1 V2944.1; mandibles; left maxilla with maxillary V2944..2; P4-M3 fragment, 3 V2944.3 autopodia, 4 antler fragments

121 Artiodactyla Bovidae Leptobos 9 broken skulls; 2 maxilla 3 maxillary V2946.1; 9 brevicornis fragments; I complete fragments, 3 V2946.2 mandible; > 10 pieces broken mandibular mandible; some isolated teeth; fragments, 2 a few postcranials partial skulls, 1 isolated tooth Artiodactyla Bovidae Leptobos sp. Right radius V5420 1 Artiodactyla Bovidae Naemorhedus Capricornis Corbet and Almost complete skull 1 almost V2947 1 sumatraensis sumatraensis Hill, 1992 complete skull

Table 5.1: List of species and minimum number of individuals (MNI) present at GWL, including any taxonomic revisions of species since the original descriptions by Hu and Qi (1978). The material examined by these authors and catalogue numbers are indicated. Species listed in bold indicate where material was available for direct examination. MNI was calculated from both examined and published descriptions (see text).

122 5.3.2 Taphonomy

All material described by Hu and Qi (1978) or examined directly in this study had already been assigned to a skeletal element. However, like many faunal assemblages collected before the 1980’s, the fate of all the fragments that could not be assigned to a particular skeletal element is unknown, and are likely to have been discarded (Marean and

Kim, 1998). Each specimen directly examined in this study was classified according to skeletal element, side and in the case of dentitions, which teeth/alveoli were present. The minimum number of elements (MNE), and hence the MNI were calculated for each species as per Marean et al. (2000). Where material for a particular species was unavailable for study, the MNI was calculated based on the descriptions provided by Hu and Qi (1978). This was straightforward when the species was represented by a single specimen. Where more than one specimen was present, the degree of detail in the description was taken into account.

If no detail was provided in the description, a minimalist approach was taken; for example,

Ochotonoides complicidens is represented by an almost complete skull, 2 other pieces of broken skull, 10 mandibular pieces and a left femur. Because it is not stated whether these mandibular fragments represent left or right, anterior, posterior, etc, it is possible (although unlikely) that they all belong to a single individual. In this case, the absolute minimum number of individuals this could represent is one. Where fossil material was available for direct study, MNI was calculated on the basis of the specimens examined. In almost all cases involving MNI values greater than 1, the MNI was calculated on representation of mandibular or maxillary fragments.

Although the MNI calculated from directly examined material is more accurate, not all material was available for study. Where the calculated MNI is greater on the basis of the

123 published record than that from direct examination, it is likely to result from the incomplete nature of the examined material. Where the MNI is greater for examined material, it is likely due to the limited original descriptions provided. Where published and observed values for

MNI differed, the greater value was used.

Minimum number of elements and number of identified specimens (NISP) was tabulated for large and small mammals, using only material available for direct examination, as described for MNI. All fossil material available was also examined for surface modifications including markings attributable to carnivores and hominins. Taphonomic bias in the GWL deposit was examined using the methodology described by Soligo and Andrews

(2005). This comparative method is particularly useful for this site, because it can be calculated from the published record alone, and is therefore unaffected by the sometimes limited availability of material.

5.3.3 Zoogeographic identity

The general biogeographical ranges, i.e. restriction to the north or south of China, of all fauna present at GWL were assessed. In the case of extant species, their current biogeographical zones were used. For extinct species, ranges were determined based on fossil occurrences and/or comparisons with closely related extant groups. Biogeographical ranges were based on Zhang et al. (1997), Sheng et al. (1999), Corbet and Hill (1992) and Tong

(2006). In addition to species preserved at GWL, the zoogeography of taxa found at two additional Lantian sites, those of Yangguo and Chenjiawo, were examined. These two sites bracket GWL in ages with Yangguo considered Early Pleistocene in age (1.4 Ma: Qiu, 2006), and Chenjiawo Middle Pleistocene (0.65 Ma: An et al., 1990).

124 5.3.4 Palaeoecology

Multivariate analyses have the advantage over more traditional, descriptive methods in palaeoecology in being able to deal with the complex, multidimensional nature of palaeoecological data sets (Shi, 1993). Because of this, multivariate analyses are becoming more common in the reconstruction of palaeoenvironments (e.g. Reed 1997, 1998,

Alemseged, 2003, Kovarovic et al., 2003, Bassarova, 2005). This study uses discriminant functions analysis (DFA) for the palaeoenvironmental classification of GWL assemblage.

This method relies on pre-defined groups defined on the basis of a number of variables. The

DFA creates linear combinations of these variables, such that the likelihood of assigning observations to their correct group is maximised. It can subsequently be used to classify new observations. In the case of palaeoecological analyses, the pre-determined groups are major habitat types. These are defined on the basis of faunal communities from modern ecosystems.

The closest modern examples of unaltered ecosystems are national parks and nature reserves.

Sampling from zoogeographically similar ecosystems should provide greater degrees of accuracy in the reconstructions. In this study, classification of modern faunal communities was derived from species lists from national parks and nature reserves from Asia (Figure 5.2,

Table 5.2). Only reserves and parks with more than five mammals listed were selected. In addition, sampling of habitat types was as wide as possible, from rainforest to desert.

However, the nature of species lists from these reserves and parks is imperfect, and is in fact subject to a number of limitations. These limitations, and potential ways of mitigating against them, are outlined below.

125

Figure 5.2: Locations of the nature reserves and national parks used in this study. 1. Gobi 20. Daweishan 2. Bogdhad 21. Maolan 3. Wulate 22. Dinghushan 4. Maojingba 23. Poyang 5. Changbai 24. Tianmushan 6. Altun Mountain 25. Wuyishan 7. Yunwushan 26. Longqi Mountain 8. Yellow River 27. Ailaoshan 9. Changtang 28. Xishaungbanna 10. Jiuzhaigou 29. Bandhavgarh 11. Funiushan 30. Erawan 12. Yancheng 31. Khao Yai 13. Medog 32. Sai Yok 14. Nuijiang 33. Kaeng Krachan 15. Baima Xueshan 34. Gunung Mulu 16. Wolong 35. Gunung Leuser 17. Fanjingshan 36. Kerinci Seblat 18. Gaolingong 37. Bukit Barisan 19. Changshanerhai

126 Vegetation Reserve/Park Species reference description Vegetation reference Ailaoshan ICE, 2007 Subtropical forest CSIS, 2007 National Environment National Environment Protection Agency Protection Agency and Environmental and Environmental Protection Bureau of Protection Bureau of Xinjiang Uygur Xinjiang Uygur Autonomous Region, Autonomous Region, Altun Mountain 1988 Desert/Grassland 1988 Baima Xueshan ICE, 2007 Temperate forest CSIS, 2007 Bandhavgarh ICE, 2007 Deciduous forest Sawhney, 2003 Bogdhad ICE, 2007 Desert UNESCO, 2007 Directorate General of Directorate General of Forest Protection and Forest Protection and Nature Conservation, Nature Conservation, Bukit Barisan 2003 Rainforest 2003 Changbai ICE, 2007 Temperate forest UNESCO, 2007 Changshanerhai ICE, 2007 News Release Office News Release Office of the Tibetan Local of the Tibetan Local Changtang Government, 2003 Grassland Government, 2003 Daweishan ICE, 2007 Tropical mountain CSIS, 2007 Dinghushan ICE, 2007 Subtropical forest UNESCO, 2007 Erawan ICE, 2007 Deciduous forest Gray et al., 1991 Fanjingshan ICE, 2007 Evergreen forest Zhou, 1990 Subtropical/temperate Funiushan Song, 1994 broadleaf Song, 1994 Gaolingong ICE, 2007 Evergreen forest UNESCO, 2007 Gobi ICE, 2007 Desert UNESCO, 2007 Directorate General of Directorate General of Forest Protection and Forest Protection and Nature Conservation, Nature Conservation, Gunung Leuser 2003 Rainforest 2003 Hazebroek and Hazebroek and Gunung Mulu Morshidi, 2002 Rainforest Morshidi, 2002 Jiuzhaigou ICE, 2007 Temperate forest UNESCO, 2007 Tropical evergreen Kaeng Krachan ICE, 2007 forest Gray et al., 1991 Directorate General of Directorate General of Forest Protection and Forest Protection and Nature Conservation, Nature Conservation, Kerinci Seblat 2003 Rainforest 2003 Tropical evergreen Khao Yai ICE, 2007 forest Gray et al., 1991 Longqi Huang, 1993 Subtropical forest Huang, 1993 Maojingba Wu et al., 2006 Deciduous forest Wu et al., 2006 Maolan ICE, 2007 Subtropical forest UNESCO, 2007 Medog ICE, 2007 Tropical forest UNEP WCMC, 2007 Nujiang ICE, 2007 Conifer forest CSIS, 2007 Poyang Wu and Ji, 2002 Wetland Wu and Ji, 2002 Tropical evergreen Sai Yok ICE, 2007 forest Gray et al., 1991 Tianmushan ICE, 2007 Temperate forest UNESCO, 2007

127 Wolong ICE, 2007 Evergreen forest UNESCO, 2007 Wulate Song and Jia, 2000 Desert Song and Jia, 2000 Wuyishan ICE, 2007 Rainforest UNESCO, 2007 Xishuangbanna ICE, 2007 Tropical forest UNESCO, 2007 Yancheng ICE, 2007 Wetland UNESCO, 2007 YellowRiver Zhao and Song, 1995 Temperate forest Zhao and Song, 1995 Department of Ningxia Department of Ningxia Yunwushan Grassland Yunwushan Grassland Yunwushan Nature Reserve (2000) Semi-arid steppe Nature Reserve (2000)

Table 5.2: List of the nature reserves and national parks used in this study, with a summary of the published vegetation type. A full list of species from these sites is provided in Appendix A.

Limitations in the use of Nature Reserves and National Parks as analogues for palaeoecological studies

The most obvious limitation when using faunal lists from reserves and parks as a basis for defining the mammalian community structure in specific habitats is that they are not comprehensive (Voss et al., 2001; Voss and Emmons, 1996). Furthermore, in some cases these may not in fact be representative of a particular faunal complex, and thus of the habitat type under study. This is largely a result of differences in sampling effort, methodologies, and differences incurred by the use of different observers (inter-observational errors). Such factors may lead to species lists from certain parks favouring certain taxa, ecological niches and/or body sizes. The latter is perhaps the most predominant (and axiomatic) cause of non- representative species lists, because large, iconic mammals will, in general, be more likely observed than smaller, elusive mammals. Some of these limitations can be avoided or mitigated against in ecological studies if a detailed observational methodology is provided along with the species lists. However, these are rarely provided. Furthermore, some species lists for parks and reserves may be compiled from several, disparate sources.

Another limitation in using the mammalian community structure of modern nature reserves as a basis for palaeoecological analysis is again due in part to inter-observational

128 error, but this time is related to the habitats they represent. One observer’s definition of what constitutes a rainforest, for example, can be significantly different from another’s. Table 5.2 lists the descriptions of habitats and/or vegetation provided on the respective reserves and parks used in this study. In many cases, the description listed in Table 5.2 is a summary of a detailed habitat report. In other cases, only a single word description was provided. In order for habitat data to be useful in palaeoecological studies, it is necessary to have some level of consistency between the habitats described. Further, rarely, if ever, will a reserve or park represent a single habitat type; in all likelihood it will represent a mosaic of different habitats, one of which may be dominant. Despite the increasing use of species lists and habitat descriptions from parks and reserves in palaeoecological analyses, especially analyses involving multivariate methods, it is apparent that their limitations have not until now been addressed.

Methods of minimising limitations

In order to deal with the potential sampling effort problem, we restricted our analyses to large-bodied mammals. As discussed above, large-bodied mammals are more likely to be better sampled than small mammals. There were two additional reasons for including only large-bodied mammals, especially as they relate to palaeoecology. Firstly, large-bodied mammals have been shown to be a good predictive tool in the reconstruction of palaeohabitats (e.g. Reed, 1998, Plummer and Bishop, 1994, Kovarovic and Andrews, 2007), and the exclusion of small mammals has been shown to not adversely affect multivariate palaeoecological studies (Kovarovic et al., 2003). Secondly, many sites excavated before modern archaeological and palaeontological techniques were developed tend to under-

129 represent small mammals. This is particularly true for deposits in China, where sieving of material from sites was not widely practiced until recently. Large-bodied, as used in this study, refers to those families whose members’ average weight falls above 0.5kg. Therefore our analyses are restricted the following families: Ailuridae, Bovidae, Canidae,

Cercopithecidae, Cervidae, Cynocephalidae, Elephantidae, Equidae, Felidae, Herpestidae,

Hyaenidae, Hylobatidae, Homininae (excluding members of Homo), Hystricidae, Leporidae,

Manidae, Moschidae, Mustelidae, Rhinocerotidae, Suidae, Tapiridae, Tragulidae, Ursidae,

Vivveridae, these representing the families present in species lists of the Asian reserves and parks.

In order to test whether the species lists from the parks and reserves were taxonomically biased, we subjected our species lists to a taxonomic distinctness analysis.

This type of analysis tests the average degree to which species in an assemblage are related to each other (average taxonomic distinctness, Δ+) and the evenness of the spread of taxa across the taxonomic spectrum (variation in taxonomic distinctness, Λ+) (Clarke and Warwick,

2001). All mammals occurring in the modern parks and reserves were assigned to superorder, order, suborder, family, subfamily, genus and species. Average taxonomic distinctness is calculated by summing the path lengths through a taxonomic tree and dividing by the number of path lengths (Clarke and Warwick, 2001). Therefore, two species from the same genus will, for example, have a lower taxonomic distinctness than two species from different genera. In effect, this value gives an indication of the phylogenetic diversity of a particular geographic region (sensu Clarke and Warwick, 2001). The list of modern faunas from the reserves and parks were used to create ‘confidence funnels’ of taxonomic distinctness. These were calculated using Taxonomic Distinctness Analysis software (Escarguel and Legendre,

130 2006). Sites which fall outside these funnels will therefore have higher or lower than the average taxonomic diversity for the region, and could therefore be considered taxonomically biased.

Variation of taxonomic distinctness calculates whether certain taxa are over- or under-represented within a geographical region, and therefore provides another test for taxonomic bias of the faunal lists. The use of these methods is particularly useful for species lists from reserves or parks because it is independent of sample size or sampling effort

(Clarke and Warwick, 2001).

In order to obtain consistent habitat types from the species lists of the various parks and reserves, no habitat type was assumed for any park a priori. Instead, the respective percentage of each large-bodied family in each reserve or park (Table 5.3) was calculated, and a cluster analysis using correlation similarity index was run using PAST

(Palaeontological Statistics) version 1.67 (Hammer et al., 2001). This analysis tests which reserves and parks are most similar on the basis of families present. The groups identified by this analysis are assumed to result from similarities in habitats, in turn a reflection of the taxonomic composition of the parks and reserves. In other words, it is assumed that certain combinations of mammalian families will characterise certain habitats, and that the groupings thus derived from the cluster analysis will be a result of these shared characteristic families.

This assumption is a logical one, and underpins all taxonomic uniformitarian or taxonomic analogy palaeoecological methods. However, because a cluster analysis will cluster any group of objects, regardless of whether these groups actually exist, the same data matrix was subjected to a principle coordinates analysis (PCO), again using a correlation similarity index. The results of the PCO should confirm whether the groupings identified by the cluster

131 analysis are indeed a result of similarities between the parks and reserves. Once each cluster was defined, each reserve and park within these clusters was examined for shared similarity of habitat types, with the most parsimonious habitat used to describe that cluster.

132 Ailuridae Bovidae Canidae Cercopithecidae Cervidae Cynocephalidae Elephantidae Equidae Felidae Herpestidae Hyaenidae Hylobatidae Homininae Hystricidae Leporidae Manidae Moschidae Mustelidae Rhinocerotidae Suidae Tapiridae Tragulidae Ursidae Viverridae Ailaoshan 0 7.4 3.7 15 15 0 0 0 11 0 0 3.7 0 3.7 0 3.7 3.7 15 0 3.7 0 0 3.7 11

Altun Mountain 0 38 13 0 0 0 0 6.3 19 0 0 0 0 0 6.3 0 0 13 0 0 0 0 6.3 0

Baima Xueshan 14 43 0 14 0 0 0 0 0 0 0 0 0 0 0 14 14 0 0 0 0 0 0 0

Bandhavgarh 0 15 15 7.4 11 0 0 0 11 7.4 3.7 0 0 3.7 3.7 3.7 0 3.7 0 3.7 0 0 3.7 7.4

Bogdhad 0 44 0 0 11 0 0 11 0 0 0 0 0 0 0 0 0 22 0 0 0 0 11 0

Bukit Barisan 0 2.1 2.1 8.5 4.3 2.1 2.1 0 15 2.1 0 4.3 0 4.3 2.1 2.1 0 17 2.1 4.3 2.1 4.3 2.1 17

Changbai 0 5 10 0 15 0 0 0 15 0 0 0 0 0 0 0 5 35 0 5 0 0 10 0

Changshanerhai 5.6 17 0 11 5.6 0 0 0 22 0 0 0 0 5.6 5.6 5.6 0 11 0 0 0 0 5.6 5.6

Changtang 2.1 27 2.1 10 6.3 0 0 2.1 19 0 0 0 0 0 0 2.1 8.3 8.3 0 0 0 0 6.3 6.3

Daweishan 0 2.4 7.1 7.1 9.5 0 0 0 12 2.4 0 2.4 0 4.8 2.4 2.4 2.4 21 0 2.4 0 0 2.4 19

Dinghushan 0 5.6 17 0 5.6 0 0 0 17 5.6 0 0 0 5.6 0 5.6 0 22 0 5.6 0 0 0 11

Erawan 0 7 2.3 14 9.3 0 2.3 0 16 4.7 0 2.3 0 2.3 2.3 2.3 0 4.7 0 2.3 2.3 2.3 4.7 19

Fanjingshan 0 6.3 13 9.4 6.3 0 0 0 13 3.1 0 0 0 3.1 3.1 3.1 3.1 22 0 3.1 0 0 3.1 9.4

Funiushan 0 7.4 15 0 7.4 0 0 0 15 0 0 0 0 3.7 3.7 0 3.7 30 0 3.7 0 0 0 11

133 Gaolingong 3.3 6.7 6.7 17 6.7 0 0 0 10 0 0 3.3 0 6.7 3.3 3.3 3.3 10 0 3.3 0 0 6.7 10

Gobi 0 33 17 0 0 0 0 8.3 25 0 0 0 0 0 8.3 0 0 0 0 8.3 0 0 0 0

Gunung Leuser 0 2.5 2.5 10 5 2.5 2.5 0 10 2.5 0 5 2.5 7.5 0 0 0 15 5 2.5 0 5 2.5 18

Gunung Mulu 0 0 0 26 11 5.3 0 0 0 0 0 5.3 0 5.3 0 0 0 5.3 0 5.3 0 5.3 5.3 26

Jiuzhaigou 11 33 0 0 11 0 0 0 11 0 0 0 0 0 0 0 11 11 0 0 0 0 11 0

Kaeng Krachan 0 9.4 6.3 16 9.4 0 3.1 0 9.4 0 0 3.1 0 3.1 0 3.1 0 9.4 0 3.1 3.1 3.1 6.3 13

Kerinci Seblat 0 2.4 2.4 9.5 4.8 2.4 2.4 0 17 0 0 4.8 0 4.8 2.4 2.4 0 17 2.4 4.8 2.4 2.4 2.4 14

Khao Yai 0 6.9 3.4 3.4 6.9 0 3.4 0 17 6.9 0 6.9 0 6.9 3.4 0 0 6.9 0 3.4 0 3.4 6.9 14

Longqi 0 4.8 9.5 4.8 9.5 0 0 0 14 4.8 0 0 0 4.8 4.8 4.8 0 19 0 4.8 0 0 4.8 9.5

Maojingba 0 0 21 0 7.1 0 0 0 21 0 0 0 0 0 7.1 0 0 29 0 7.1 0 0 7.1 0

Maolan 0 5.6 11 5.6 17 0 0 0 11 0 0 0 0 5.6 5.6 0 5.6 11 0 5.6 0 0 0 17

Medong 5.3 16 0 16 11 0 0 0 21 0 0 0 0 0 5.3 0 16 0 0 0 0 0 0 11

Nujiang 3 6.1 12 9.1 6.1 0 0 0 12 0 0 0 0 0 3 0 6.1 27 0 0 0 0 3 12

Poyang 0 0 13 0 6.7 0 0 0 13 0 0 0 0 6.7 6.7 6.7 0 33 0 6.7 0 0 0 6.7

Sai Yok 0 8.6 2.9 17 8.6 2.9 2.9 0 17 0 0 2.9 0 2.9 0 0 0 5.7 0 2.9 2.9 2.9 5.7 14

Tianmushan 0 6.7 13 3.3 10 0 0 0 17 3.3 0 0 0 3.3 3.3 3.3 0 23 0 3.3 0 0 0 10

Wolong 2.7 8.1 11 5.4 11 0 0 0 16 0 0 0 0 2.7 2.7 0 5.4 22 0 2.7 0 0 5.4 5.4

Wulate 0 25 25 0 0 0 0 8.3 25 0 0 0 0 0 8.3 0 0 8.3 0 0 0 0 0 0

Wuyishan 0 3.4 6.9 6.9 10 0 0 0 17 3.4 0 0 0 3.4 3.4 3.4 0 24 0 3.4 0 0 3.4 10

134 Xishuangbanna 5 5 0 10 5 0 5 0 20 0 0 10 0 5 0 5 0 5 0 0 0 10 0 15

Yancheng 0 0 0 0 33 0 0 0 0 0 0 0 0 0 17 0 0 33 0 17 0 0 0 0

YellowRiver 0 0 17 0 0 0 0 0 17 0 0 0 0 0 17 0 0 50 0 0 0 0 0 0

Table 5.3: Percentage of each of the large-bodied mammalian families present in the nature reserves and national parks. For a complete species list, see Appendix A.

135 Multivariate palaeoecological methods

The GWL mammals examined were restricted to the large-bodied mammalian families discussed above. The habitat types identified by the cluster analysis were used as the basis for the discriminate functions analysis (DFA). The DFA, using percentage-abundance matrix of the different parks and reserves was classified by SPSS for Windows version 11.0.1

(www.spss.com). The Mahalonobis distance measure was used, and results were cross- validated by leaving one reserve out, re-running the analysis using the remainder of the matrix, and classifying the omitted reserve. The functions calculated in this way were then used as a basis for classifying GWL, using the relative percentages of each family present at

GWL (Table 5.4).

Ailuridae 0 Homininae 0

Bovidae 13 Hystricidae 4.2

Canidae 4.2 Leporidae 0

Cercopithecidae 4.2 Manidae 0

Cervidae 13 Moschidae 0

Cynocephalidae 0 Mustelidae 4.2

Elephantidae 4.2 Rhinocerotidae 8.3

Equidae 4.2 Suidae 4.2

Felidae 17 Tapiridae 8.3

Herpestidae 0 Tragulidae 0

Hyaenidae 4.2 Ursidae 8.3

Hylobatidae 0 Viverridae 0

Table 5.4: Percentages of the large-bodied families present at GWL. For a complete list of species see Table 5.1.

136 5.4 Results and Discussion

5.4.1 Taphonomy

The MNI for each species in the GWL assemblage is given in Table 5.1. The MNI of the large mammals is also provided in Figure 5.3. Of the large mammals, the greatest contributions come from the Cervidae, with Cervus comprising 34% of the assemblage.

Other high numbers of individuals are represented by Leptobos brevicornis and Equus sanmanensis at 12% and 11% respectively.

137

Figure 5.3: Pie graph showing minimum number of individuals (MNI) of large-bodied species from GWL. Relative percentages are indicated in brackets.

138

Large Mammals Small Mammals Antler 7/7 0/0 Autopodia 35/35 2/2 Crania 23/28 3/4 Femur 2/4 2/2 Humerus 4/7 1/1 Illium 0/0 1/1 Mandible 52/102 19/34 Maxilla 32/48 4/6 Radius 4/4 0/0 Rib 1/1 0/0 Tibia 1/1 1/1 Ulna 1/1 0/0 Vertebra 5/5 0/0 Isolated teeth -/109 -/29

Table 5.5: Minimum Number of Elements (MNE)/Number of Identified Specimens (NISP) for large- and small- bodied mammals.

Figure 5.4: Proportion of non-dental minimum number of elements (MNE) shown for small- (grey bars) and large-bodied (white bars) mammals. Proportions calculated from the total non-dental MNE for each body-size group.

139 The NISP and MNE for large and small mammals are given in Table 5.5, and shown in Figure 5.4. The number of large mammal elements outnumbers the small mammal elements. However, this is likely to be a result of the collecting bias. For both large and small mammals, isolated teeth make up the majority of the elements, followed by mandibles and maxilla. For large mammals, crania and autopodia also make up a significant percentage of the elements present.

Surface Modifications

A total of 11 instances of surface modifications were observed in the bone sample studied (Table 5.6). Of these, nine were likely rodent gnaw marks, and of these all but one are likely to have been caused by Hystrix (Figure 5.5), a rodent present in the deposit, and furthermore a bone surface alteration commonly recorded in Pleistocene sites around the world (e.g. de Vos 1983, Lyman, 1994, Maguire, 1976, Brain, 1981). This represents 2.8% of the specimens directly examined, and probably rules out porcupines as a collecting agent, because these often gnaw between 60 and 100% of bones they accumulate (Maguire, 1976,

Brain, 1981). No gnaw marks were observed on any of the isolated teeth. A total of four likely carnivore tooth pits were observed on the bones studied, representing 1.2% of the specimens examined. Two of these were also found on bones bearing rodent gnaw marks. It is possible that more bone surface alterations were present; however the poor preservation of the surface of some of the elements precludes definitive identification. None of the specimens observed preserved any kind of human alteration, in particular in the form of cut marks, percussion marks and/or burn marks. Mid-shaft fragments, however, which can record the best indication of hominin versus carnivore activity (Blumenschine, 1988;

140 Blumenschine and Marean, 1993; Capaldo, 1997, 1998; Marean and Kim, 1988; Selvaggio,

1998; Marean et al. 2000) are unfortunately almost completely absent from the collection examined. It is not clear whether their absence is a result of the initial collecting bias, or from subsequent loss, but it makes it almost impossible to test whether hominins were involved in the deposition. Further, the presences of cut marks on bones from Early Pleistocene hominin localities are rare. Some of the bones examined showed considerable surface weathering, some showed a degree of cracking, and others a fairly unweathered profile. However, exact percentages of stages of weathering are of limited value because most specimens examined directly were represented by isolated teeth or toothrows that weather differently to bones

(Andersson and Kaakinen, 2004). We do note, however, that some of the weathering observed at GWL appears identical to the heavily weathered bones from Locality 6, a Late

Miocene site in Lantian, China (see Figure 6A, Andersson and Kaakinen (2004)). Given the geographical proximity of Andersson and Kaakinen’s (2004) study sites to GWL, it is likely that a similar type of aerial weathering acted on the bones from both areas. It is also noteworthy that small mammals were not found in the same taphonomic or depositional setting as large mammals at this Miocene locality (Andersson, pers. comm. 2007).

141

Specimen Gnaw Tooth Genus Element number mark pits Equus V2940.5 right partial innominate 1 0 Equus V2940.5 left partial innominate 1 0 Equus V2940.5 left proximal humerus 0 1 Equus V2940.5 right distal metatarsal 1 0 Leptobos V2946 partial mandible 1 1 left cranial fragment with partial Cervus V5419 1 0 antler Cervus V5419 right partial mandible 1 0 Cervus V5419 left partial mandible 0 1 Cervus V5419 cranial fragment 1 1 Cervus V5419 rib fragment 1 0 Cervus V2978.5 antler 1 0

Table 5.6: List of surface modifications present in sample studied. 1 = presence, 0 = absence, gnaw marks refer to likely rodent activity, and tooth pits to likely carnivore activity.

Figure 5.5: V5419 Cervus grayi antler showing extensive rodent gnaw marks. Scale bar = 3cm.

142 Bias

The analysis revealed several biases in the GWL deposit (Figure 5.6). The most significant of these, which fell outside the maximum and minimum ranges observed by

Soligo and Andrews (2006), suggests an under-representation of the smaller carnivores (in both the <1 kg and 1-10 kg weight classes) and an over-representation of very large and massive (i.e. >45 kg) carnivores and massive (i.e. >180 kg) herbivores. The over- representation of large-bodied mammals and under-representation of small mammals could be due to original collection biases, particularly if the site was not sieved during excavation.

However, small mammals were not completely overlooked, such that the abundance of certain categories of microfauna does fall within the modern range (Figure 5.6).

The site preserves an over-representation of large carnivores. Although an over- abundance (i.e. at least 20% of species) of carnivores in a deposit can be an indication that they were involved in its deposition (Cruz-Uribe, 1991), the percentage of carnivores as a proportion of the large-bodied mammals at GWL (as calculated from MNI) does not exceed

20% (although this could also be due to the under-representation of smaller carnivores). The

GWL fauna also falls outside the 25th and 75th percentile bands observed by Soligo and

Andrews (2006) in a number of categories: the medium herbivores (under-represented), large carnivores (under-represented) and very large herbivores (over-representated).

143

Figure 5.6: Box plot showing the relative abundance of modern fauna within the groups defined by Soligo and Andrews (2005): A: <1kg, B: 1-10kg, C: 10-45kg, D: 45-180kg, E: >180kg; P: primary consumer, S: secondary consumer. Location of GWL faunas indicated by closed circles.

Taphonomic implications

Because only a portion of the original GWL material was available for study, we consider any inferences on the taphonomy of this site as preliminary. Our taphonomic observations were conducted for two primary reasons. Firstly, and most importantly, it was carried out in order to establish a firm basis for any subsequent palaeoecologcial analysis. In this endeavor, our results can be considered accurate, because they are based only on species’ presence – something already well established by the original publications about the site. The biases found were not unexpected. The site showed an over-representation of large mammals and under-representation of small mammals, in all likelihood an artifact of original collection biases. However, some small mammal abundance categories fall within the modern range.

Although generally showing over-representation, most large mammal faunas are within the range found in modern environments. Because the subsequent palaeoenvironmental

144 reconstruction (detailed below) used only large-bodied mammals, the under-representation of smaller mammals should not be adversely affect the palaeohabitat reconstructions.

Secondly, our taphonomic analysis sought to examine whether hominins contributed to the accumulated remains, in particular through hunting and/or butchery practices (Wang,

1996, Wang et al., 1997). Unfortunately, because of the limited nature of the material available for study, it is currently impossible to assess hominin involvement. Certainly, no signs of hominin modifications to the bones were observed. In particular, no cut marks were visible, and many of the limb elements examined were complete. In addition, many teeth were examined for evidence of burning, a sign of head roasting for consumption, which has been suggested for other Chinese sites, in particular Zhoukoudian (Binford and Ho, 1985;

Binford and Stone, 1986). No such alterations were found. Zhang (1985) does refer to the presence of charcoal granules at the site, although no further information regarding these has been uncovered. The paucity of mid-shaft fragments in the material examined makes it very difficult to accurately assess the level of involvement of hominins versus carnivores in the formation of the site (Marean 1998, Marean and Kim, 1998). This is compounded further by the fact that only about 10% of the material described by Hu and Qi (1978) could be examined directly in our study.

Carnivore and rodent damage was observed on several bones. It is unlikely, based on the sample examined, that rodents contributed significantly to the formation of the assemblage. The contribution by carnivores in the formation of the site is difficult to assess.

Although there is a high percentage of carnivores represented in the deposit, it is not a definitive trait of carnivore accumulations. The apparent dominance of foot and head

145 elements may be indicative of a carnivore accumulation (Stiner, 1991), but as discussed above, this more probably reflects original collection biases.

GWL shows an apparent depositional preference for large mammals, a common occurrence for surface sites (Lyman, 1994). Again, this may be an artifact of collection bias.

However, this could also be a result of different taphonomic histories between small and large mammals. Given that the small mammals were found in the lower parts of the stratigraphic profile, it is likely that the site represents two different periods of deposition, with the small mammals deposited during a glacial period in the sandy loess, and the larger mammals deposited in a palaeosol during a later, interglacial period. Different taphonomic histories could be suggested on the basis of the skeletal profiles of the small and large mammals. However, based on the small sample size of fossils studied here, this is necessarily speculative and confirmation of this model would require a more detailed taphonomic study that unfortunately is not possible on the basis of the data available.

Based on the preliminary observations reported here, together with its documented derivation from the Chinese loess, we suggest that this faunal assemblage may represent a surface deposit. If it is a surface site, the large numbers of isolated teeth, mandibular and cranial fragments suggests a channel, levee or floodplain deposit (Alemseged, 2003).

However, testing this prediction on the basis of the skeletal part profile (e.g. Behrensmeyer,

1975, Lyman, 1994, Norton et al., 2007) will require examination of, at least, all the material originally described by Hu and Qi (1978). The presence of aerial weathering suggests some bones spent a significant amount of time exposed at the surface before being buried. The suggestion by Wang (1996) that the unique appearance of sub-tropical species found at GWL was not due to climatic effects, but rather was a direct result of hominin hunting activity

146 cannot be discounted on the basis of the observations reported herein. However, based on the limited sample studied, we found no direct evidence of human involvement in this faunal accumulation.

5.4.2 Zoogeography

An examination of the zoogeography of the constituent species of GWL shows a clear distinction between body sizes (Table 5.7). Most of the small-bodied mammal fauna is characteristically palearctic in identity, while the majority of the large mammals are oriental.

On the basis of published accounts of the geology of the region, it appears that the small and large mammals may have been deposited at different times. Comparison of the skeletal elements directly examined also suggests different depositional profiles and by inference different times of deposition. (Figure 5.4). If the presence of the concretions, associated with the large mammals but not with the microfauna is indicative of more humid conditions (as suggested by Zhu and Zhou (1994)), then the lower part of the clay layer, associated with the microfauna, is likely to have formed during less humid climes, while the upper part, associated with the megafauna, formed in more humid conditions. This would be consistent with a depositional model wherein the small mammals were deposited during a glacial period, followed by the deposition of large mammals during an interglacial. This suggests a difference of age between the two faunas. Palaeomagnetic dating of the site correlated the

Lishi section with the GWL hominin material (An and Ho, 1989), such that the small-bodied mammals are probably older than 1.15 Ma.

147

Biogeographic Weight region class Small Mammals Scaptochirus moschatus Palearctic Small Proedromys sp. cf. P. pliocaenicus ? Small Allophaiomys sp. cf. A. Palearctic? Small pliocaenicus Yangia tingi Palearctic Small Eospalax fontanieri Palearctic Small Eospalax sp. Palearctic Small Cricetulus sp.cf. C. griseus Palearctic Small Allocricetus teilhardi Palearctic Small Bahomys hyposodonta ? Small Cricetulus sp. Palearctic Small Meriones sp. Palearctic? Small Petaurista sp. Oriental? Medium Apodemus sp. Widespread Small Ochotonoides complicidens Palearctic? Small Ochotona sp. cf. O. thibetana Widespread Small Large Mammals Rhinopithecus lantianensis Oriental Large Hystrix sp. cf. H. subcristata Oriental Medium Canis variabilis Widespread Very large Selenarctos thibetanus ?Widespread Very large Ailuropoda melanoleuca Oriental Very large Meles sp. cf. M. leucurus Widespread Very large Pachycrocuta sinensis Widespread Very large

148 Panthera sp. cf. P. tigris Widespread Very large Panthera pardus Widespread Very large Sivapanthera pleistocaenicus ? Very large Meganteron lantianensis Palearctic? Massive Stegodon orientalis Oriental Massive Equus sanmenensis Palearctic Massive Dicerorhinus kirchbergensis Oriental Massive Dicerorhinus lantianensis Oriental Massive Megatapirus augustus Oriental Massive Tapirus sinensis Oriental Massive Hesperotherium inense Oriental Massive Sus lydekkeri Widespread Very large Elaphodus cephalophus Oriental Very large Cervus grayi Oriental? Very large Sinomegaceros konwanlinensis Palearctic Massive Leptobos brevicornis Palearctic Massive Leptobos sp. Palearctic Massive Capricornis sumatraensis Oriental Very large

Table 5.7: Zoogeography of the species present at GWL (based on Zhang et al. (1997), Sheng et al. (1999), Corbet and Hill (1992) and Tong (2006)) with associated weight class (as per taphonomy methods: A-small; B- medium; C-large; D-very large; E-massive).

Zoogeographic implications

The biogeography of China has been the subject of many recent studies (e.g.

Hoffmann, 2001; Zhang, 2002; Xiang et al., 2004; Xie et al., 2004; Zhao et al., 2006; Zhao and Fang, 2006). The general distinction between the Eurasian (palearctic) and the

Indochinese (oriental) regions has long been recognized (e.g. Pei, 1957, 1958). Most

149 researchers further subdivide China into a number of distinct biogeographical zones/units based on either readily distinguished faunas or floras (Hoffmann, 2001, Zhao and Fang,

2006), quantitative analyses (Xiang et al., 2004, Zhao et al., 2006) or GIS-analytical analyses

(Xie et al., 2004). The palearctic and oriental faunas have generally been divided by the

Qinling Mountains to the west, and the Huai River to the east (Zhang, 2002). Mammals in

China are most strongly correlated with plant richness (Zhao et al., 2006), and it is suggested that this strong division is likely related to differences in humidity or climatic conditions between the north and south (Zhang, 2002; Tong, 2006). Oriental mammals, in particular, are tied to the distribution of evergreen broadleaf forests (Xie et al., 2004).

Pleistocene Chinese faunas have also traditionally been split into two distinct zoogeographic zones, on the basis of fauna, flora and climate (Pei, 1957, 1958). The northern

Chinese faunas represent the palaeoartcic zone, and are characterized by steppe- and cold- adapted mammals. Southern Chinese faunas represent the oriental zone, and are characterized by tropical to sub-tropical adapted mammals, the so-called Stegodon-Ailuropoda faunas, typical of Southeast Asia. The line of demarcation between these Pleistocene faunas is generally considered to be the Yangtze River (Pei, 1957, 1958).

150

Species Yangguo Chenjiawo Lepus wongi 1 Ochotonoides complicidens 1 Bahomys hypsodonta 1 Allocricetus teilhardi 1 Cricetulus barabensis 1 Myospalax chaoyatseni 1 Yangia tingi 1 1 Mysospalax sp. cf. M. fontanieri 1 Apodemus sp. cf. A.. sylvaticus 1 ? Hystrix sp. 1 Hystrix subcristata 1 Cuon alpinus 1 Felis sp. cf. F. paleosinensi 1 Felis shansius 1 Pachycrocuta sp. cf. P. licent 1 Meles sp. cf. M. meles 1 Panthera tigris 1 Ursus kokeni 1 Proboscihipparion sinense 1 Hipparion sp. 1 Equus sanmeniensis 1 Elephas sp. 1 Sus sp. cf. S. lydekkeri 1 Megaloceros sp. 1 Cervus grayi 1 Gazella sinensis 1 Elaphurus chinanensis 1 Nipponicervus elegans 1 Leptobos brevicornis 1

Table 5.8: Faunal list for Yangguo (derived from Ji (1975), with revisions by Qiu et al. (2004)) and Chenjiawo (derived from Chow (1964) and Chow and Li (1965)).

Located north of the Qinling Mountains, GWL, based on the delineation of the faunas described above, is usually classified as a northern Chinese fauna (e.g. Qiu, 2006, Han and

Xu, 1985). However the transitional nature of GWL has long been recognized (Hu and Qi,

1978) and it is also referred to the southern Chinese Stegodon-Ailuropoda fauna (e.g. Zhu

151 and Zhou, 1994). The biogeographic breakdown of the mammals from GWL shows a distinct difference along body sizes. What this means for the biogeographical history of the fauna is difficult to quantify. However, an examination of the faunal list of Yangguo (Table 5.8), an

Early Pleistocene site from Lantian pre-dating GWL, shows most species from this site having northern affinities (Qiu, 2006). The small mammals were preserved in loess sediments probably during a glacial period, while the large mammals were preserved in a palaeosol layer, most likely representing an interglacial (a likely candidate could be MIS 39).

This suggests that the larger, and southern-derived species present at GWL were the result of an incursion northwards, most likely due to the warmer and more humid environmental conditions present during Early Pleistocene interglacial periods (Jablonski and Whitford,

1999), with concomitant northern expansion of evergreen broadleaf forests. This is also supported by Wang et al.’s (1997) suggestion of pure C3 plants. Warmer conditions during the interglacial enabled the large mammals to occupy a broader geographical range, at least until the end of the interglacial. This finding also supports the contention by Tong (2006) that the division between the northern and southern Chinese faunas during the Pleistocene was more likely to be climatically driven, rather than the result of physical barriers.

5.4.3 Palaeoecology

Limitations of data from Nature Reserves and Parks

The taxonomic distinctness measure showed that some Asian parks have a lower average taxonomic distinctness (Δ+) than expected (Figure 5.7a), but all the parks have a variation in taxonomic distinctness (Λ+) within the expected boundaries (Figure 5.7b).

Daweishan, Gunung Leuser, Gunung Mulu, Xishuangbanna, Wolong, Gobi, Bandhavgarh and Wulate had Δ+ values lower than the 99% probability funnel, whilst Bukit Barisan,

152 Changbai and Changtang were inside the 99% probability funnel but outside the 95% probability funnel (Figure 5.7a). This measure was used in this study to identify potential taxonomic biases in the species lists of the reserves and parks. Originally, Warwick & Clarke

(1995) developed this measure to identify the effect of pollution on the taxonomic composition of a fauna, specifically a marine bivalve community. Clarke and Warwick

(2001) suggested further possible uses for this methodology, but the application presented in this study represents, to our knowledge, the first time this method has been used to identify a potential taxonomic bias in a modern faunal list. Warwick & Clarke (1995) suggested that a lower than average taxonomic distinctness of a community was the result of the action of pollution, but they also suggested it may also reflect different habitats or geographical

153

Figure 5.7: Confidence funnels for taxonomic distinctness of the nature reserves and national parks listed in Table 5.2; calculated from the species lists located in Appendix A. A: Average taxonomic distinctness; B: Variation in taxonomic distinctness.

154 regions. Given that, of the reserves and parks falling outside the confidence funnel (Figure

5.7a) there is no common geographical position or habitat type linking them, it is likely their lower than average taxonomic distinctness is a result of either poor/inadequate sampling or habitat destruction.

Polluted communities in Clarke and Warwick’s study (2001) were, however, shown to have fairly low variation in taxonomic distinctness (Λ+), meaning that the species were still evenly spread across the taxonomic tree. High values of Λ+ (uneven repartition of taxa in the taxonomical tree) were only identified on islands where habitat diversity is reduced

(Clarke and Warwick, 2001). It is not evident from the work of Clarke and Warwick (2001) which of average taxonomic distinctness or variation in taxonomic distinctness might be useful for identifying biases in a modern mammalian community, however, all parks and reserves in our study had Λ+ values within the probability funnel, meaning that all of them have taxa evenly distributed across the overall taxonomic tree. This suggests that none of the parks or reserves studied had reduced habitat diversity. Four of the parks and reserves are located on islands (three on Sumatra, and one on Borneo). According to Clarke and Warwick

(2001), island communities may have high variation in taxonomic distinctness due to reduced diversity, but our results do not show such a pattern. It may be that Sumatra and Borneo are large enough to not suffer any significant reduction in taxonomic diversity.

Low average taxonomic distinctness (Δ+) may simply be the result of poor sampling, but with variation in taxonomic distinctness (Λ+) falling within the probability funnel for all of the parks and reserves, this suggests that sampling was fairly evenly achieved. Therefore heavy sampling biases are suggested as unlikely, and it is perhaps more probable that these parks and reserves represent areas of environmental degradation. Λ+ may be a better measure

155 of sampling bias than Δ+, but since none of the parks or reserves showed any unexpected values of Λ+, we are unable to test this hypothesis. Given this uncertainty, and given that the parks and reserves identified as falling outside the Δ+ confidence funnel may represent degraded environments, we felt it most prudent to exclude these from any further analyses.

Figure 5.8: Dendogram showing the clustering of the reserves and parks based on families present. The most parsimonious habitats of each group are indicated.

The cluster analysis revealed three major groupings of habitat types (Figure 5.8), with only one reserve, Yancheng, not clearly clustering with any of the major habitat types. These results were confirmed by the PCO, which revealed clear separation between the habitat

156 types, as well as the separation of Yancheng (Figure 5.9). These groupings were compared with the published descriptions of habitat types of the reserves and parks. The three clusters were therefore assigned as closed forest, open/mixed forest and savannah/desert, based on the most parsimonious reconciliation between published habitat descriptions. These classifications are broadly consistent with previous palaeohabitat classification schemes (e.g.

Kappelman, 1988, 1991, Plummer and Bishop, 1994). Yancheng, due to its uncertain status, was excluded from further analyses. A more detailed description of these habitats is provided below.

157

Figure 5.9: Scatter plot of the principle coordinates analysis; A: coordinate 1 vs. coordinate 2; B: coordinate 2 vs. coordinate 3. Minimum span trees for each habitat type are also marked.

Habitat type 1 - Closed forest

The closed forest habitat is represented by the following reserves and parks: Kerinci

Seblat, Maolan, Erawan, Kaeng Krachan, Khao Yai, Sai Yok, Gaolinggong, Medog,

Changshanerhai and Ailaoshan.

158 Ailoshan is a subtropical evergreen humid broadleaf forest (CSIS, 2007). Khai Yai and Kaeng Krachan National Parks are predominately covered by dense, tropical broadleaf forests up to 1000m, and submontane broadleaf forest between 1000 and 1500m (Gray et al.,

1991). Erawan National Park is lower in altitude, and is characterised by mixed species deciduous forests as well as sizable tracts of bamboo growth (Gray et al., 1991). Sai Yok

National Park is cut by many rivers, and is also covered in large part by tropical, broadleaved evergreen forest (Gray et al., 1991). Medog Nature Reserve is the northern most tropical forest. River valleys support tropical forest, with shrubs and small trees comprising the forest floor. Montane subtropical broadleaf forest, with shrubs and dense bamboo groves occur at

2400m. Montane temperate coniferous forest occurs between 2400-3800m, while above 3800 alpine scrub and meadow predominate (UNEP, 2007).

Habitat type 2 - Open/Mixed forest

The open forest habitat is represented by the following reserves and parks:

Maojingba, Dinghushan, Funiushan, Longqi Mountains, Wuyishan, Tianmushan, Yellow

River, Nuijiang, Fanjingshan, Poyang and Yunwushan.

Yunwushan is predominately covered by herbages, and woodlands only occupy 2% of the reserve, situated largely in gullies and revines (Department of Ningxia Yunwushan

Grassland Nature Reserve, 2000). Lonqi Mountains, on the other hand, has forest coverage of

97%, and is located in the subtropical zone (Huang, 1993). Yellow River Delta Nature

Reserve and Jiangxi Poyang Lake Nature Reserve are wetlands. The Yellow River Nature

Reserve belongs to the warm, temperate broadleaf forest zone (Zhao and Song, 1995).

159 Funiushan could perhaps be considered most similar to GWL based on its geographic position. It straddles the line between the subtropical and warm temperate zone, with its southern slopes hosting vegetation characteristic of the former, and its northern slopes of the latter (Song, 1994). It is predominately covered in broadleaf forest and needle-leaved forest, but also hosts bamboo forests as well as bush and bush grasses. Maojingba is also largely covered in forest, and the following vegetation types are found there: coniferous forest, deciduous broad-leafed forest, deciduous broad-leafed shrubland and meadows (Wu and

Zhao, 2006). Nujiang is described as a sub-mountain conifer ecosystem (CSIS, 2007).

Wuyishan is considered a sub-tropical to temperate rainforest, but also contains coniferous forest and sub-tropical mountain meadow (UNESCO, 2007). Fanjingshan includes nine vegetation types: evergreen broadleaf forest, low mountain and warm coniferous forest, low mountain broadleaf forest, mid-mountain and temperate coniferous forests, evergreen and deciduous broadleaf forest, mountaintop mossy forest, subalpine scrub, fir and hemlock broadleaf forest and bamboo forest (Zhu and Yang, 1990). Tianmushan is predominately characterised by temperate broadleaf forests and woodlands (UNESCO, 2007).

Habitat type 3 - Savannah/Desert

This habitat is represented by the following reserves and parks: Jiuzhaigou, Altun

Mountains, Bogdhad and Baima Xueshan. The vegetation at Bogdhad (also known as

Bogeda) is characterised according to topography: alpine cushion vegetation above 3100m; alpine and sub-alpine meadows 2000-2900m; mid-mountain forest-meadow 1600-2700m; desert landscape 600-1000m (UNESCO, 2007). Baima Xueshan is described as a temperate and frigid-temperate forest ecosystem (CSIS, 2007) but, its clustering with other

160 savannah/desert habitats suggests that the number of forest-adapted mammalian families is likely to be minimal. The Altun Mountain National Nature Reserve is largely covered with semishrub desert and alpine-arctic steppe. The flora is dominated by cushion plants and alpine-arctic herbs (National Environment Protection Agency and Environmental Protection

Bureau of Xinjiang Uygur Autonomous Region, 1998). Wulate has vegetation typical of deserts (Song and Jia, 2000).

Discussion of habitat types

The clearest signatures of vegetative structure appear in the closed forest and savannah/desert habitats, although these are not all homogenous. Closed forests are generally characterized by evergreen broafleaf forests. The vegetation structure varies with altitude: high canopies present at low altitudes and decreasing canopy heights associated with higher altitudes. Vegetation can be quite dense, and this trait tends to increase with increasing altitude. The open/mixed forests display the most heterogeneous structures, and encompass wetlands, open woodlands and shrubland. Forest cover ranges from 2% to 97%. Broadleaf and coniferous forests appear to be fairly ubiquitous in these reserves, and zones not represented by tree cover tend to be covered with shrubs. Savannah/desert habitats are characterized by areas of desert vegetation, shrublands and grasslands. Each of the three broad habitat types shows altitudinal changes in vegetation, characterized largely by reduction in vegetation height and concomitant changes in floral species.

161 Palaeoecology of Gongwangling

On the basis of these habitat types, the DFA was run, revealing a total of four most informative families selected by the analysis (Table 5.9). This analysis yielded two discriminant functions that accounted for 100% of the variance (Table 5.9). Ninety-six percent of the reserves were correctly reclassified based on the variables used (Table 5.10).

GWL was assigned to the closed forest habitat type on the basis of these functions (Figure

5.10).

Eigenvalues Function Eigenvalue % of variance Canonical correlation 1 15.593 78 0.969 2 4.388 22 0.8902 Functions at Group Centroids Habitat Function 1 Function 2 Closed 0.146 -2.405 Open/mixed -2.979 -1.555 Desert/savannah -7.829 1.739 Canonical Discriminant Function Coefficients Family Function 1 Function 2 Bovidae 28.379 9.077 Hylobatidae 48.405 -51.570 Herspestidae -21.874 35.210 Mustelidae -1.337 14.551 Constant -3.237 -3.517

Table 5.9: Summary of DFA results for the three habitat types.

162

Predicted Group Membership Habitat Closed Forest Open/Mixed Savannah/Desert Original Count Closed Forest 10 0 0 Open/Mixed 0 11 0 Savannah/desert 0 0 4 Gongwangling 1 0 0 Percentage Closed Forest 100 0 0 Open/Mixed 0 100.0 .0 Savannah/desert 0 0 100.0 Gongwangling 100.0 0 0 Cross-validated Count Closed Forest 9 1 0 Open/Mixed 0 11 0 Savannah/desert 0 0 4 Percentage Closed Forest 90 10 0 Open/Mixed 0 100.0 0 Savannah/desert 0 0 100.0

Table 5.10: DFA classification results. In the case of cross validation, each case is classified by the functions derived from all cases other than that case.

Figure 5.10: Graphical representation of the canonical discriminant functions showing the clustering of the different reserves and parks. GWL is classified according to the discriminant functions calculated (see text), and its position relative to the other parks and reserves is indicated.

163

It should be firstly noted that the methodology used here does not provide clues to the aridity or humidity of the site, but rather the nature of the vegetation present at the site. The former may perhaps be revealed by other palaeoecological analyses. Cenograms, for example, have been used to infer the relative vegetation and humidity of a site from Thailand

(Tougard and Montuire, 2006), although the use of such a method would require better representation of the microfauna.

However, given that the distribution of oriental species may be driven by humidity (as discussed above), the presence of so many oriental species at GWL suggests an increase in humidity during its deposition. The results of the DFA are not surprising considering the results of both the taphonomic study and zoogeographic analysis. Given that the Qinling

Mountains were at this time less than 1000m in altitude (Ji, 1980), and given its northern geographical position, GWL probably most closely resembles Medog Nature Reserve

(described above). The presence of the giant panda in the deposit suggests the presence of bamboo in this forest, although the current habitat preferred by this mammal is a mixed forest, composed of conifers and broad-leafed forests, with an undergrowth of temperate bamboo (Farrelly, 1984, Reid et al., 1989). Oriental species, however, are often associated with broadleaf evergreen forests (Xie et al., 2004), suggesting these could also have made up a significant proportion of the forest.

The Late Pliocene in Lantian witnessed a change from drier and more open environments to more humid, closed habitats (Kaakinen et al., 2006). The introduction of these conditions, as well as the then lower height of the Qinling Mountains (Ji, 1980) may have enabled the migration of southern species into the north. The warm, humid, closed forest environment suggested by this analysis, however, did not last. By the later part of the

164 Early Pleistocene, permafrost had formed over the Tibetan Plateau (Zhou et al., 1991), and by the Middle Pleistocene, the sub-tropical and tropical zones had moved southwards

(Jablonski and Whitford, 1999).

The hominin locality of Chenjiawo, is located in the geographic vicinity of GWL, but dated at 650 ka. Its fossil faunas are typically palearctic in nature (Qiu, 2006). By the time this deposit formed, most of the southern elements present at GWL had therefore become locally extinct. The notable exception is the hominins, with Homo erectus remains also found at Chenjiawo. Environmental conditions were different between the deposition time of GWL and Chenjiawo, such that the continued survival of oriental mammals in the region was untenable.

5.5 Summary

The hominin locality of GWL represents a closed forest habitat that was most likely deposited during warm, moist conditions. Preliminary taphonomic observations provide no evidence of hominin involvment in the accumulation. The geological descriptions and the taphonomic observations suggest two phases of deposition, the first during a colder period involving the small-bodied mammals, the second during a more humid phase involving the large mammals and hominins. The zoogeographic roots of the large-bodied mammals of

GWL indicate they are predominately of an oriental nature, and were therefore most likely result from northern migration during a warmer, more subtropical interglacial. The palaeoenvironment of GWL is likely to represent a forest with a medium canopy height, abundant ground herbs and some bamboo. Hominins may have migrated north into GWL along with the oriental mammal species, taking advantage of the then lower Qinling

Mountains and concomitant warmer and more humid conditions. Whether hominins

165 remained in the area after the deterioration of climatic conditions is not testable with current data. Extinction of megafauna at this locality is likely tied to changes in environmental conditions. At present, there is no evidence to suggest that hominins were either responsible for the representation of oriental species in GWL, or that they were involved in the latter’s local extinction.

166 CHAPTER 6

PALAEOECOLOGY OF QUATERNARY SOUTHEAST

ASIAN SITES REVEALED THROUGH PHYLOGENY-

FREE DISCRIMINANT FUNCTIONS ANALYSIS

6.1 Abstract

Multivariate, phylogeny-free palaeoecological analyses are applied to thirty Southeast

Asian fossil sites spanning the Early Pleistocene to Holocene. These analyses focus on

megafauna because these are often the only species adequately represented in Southeast

Asian Pleistocene sites. These methods rely on minimum number of species to differentiate between closed (e.g. rainforest), mixed (e.g. open woodland) and open (e.g.

savannah) habitats. In twenty-seven sites, sufficient species were present to distinguish

open habitats from other habitat types. However, almost all sites preserve insufficient species to confidently differentiate between mixed and closed palaeohabitats.

Nevertheless, these results represent the first palaeoenvironmental reconstructions for

many Pleistocene sites in Southeast Asia. Where previous reconstructions exist, comparisons between those qualitative studies and the results of analyses conducted here reveal a high level of congruence. Support for methods outlined in this study indicates that palaeoenvironments can be successfully reconstructed on the basis of species lists

alone and that it is possible to reconstruct the palaeohabitats of sites for which published

material has been lost or is currently unavailable for study.

167 6.2 Introduction

Of all the regions on Earth, Southeast Asia remains, due to environmental, political and historical reasons, a place with very little known about its palaeontological history. Recent research has, however, begun to shed light on this fascinating region, its biogeography, palaeoecology and palaeontology (e.g. Zeiton et al., 2005; Bacon et al.,

2004; Tougard, 2001; van den Bergh et al., 2001; Louys, 2007). Despite this increased awareness, however, much general information for prehistoric Southeast Asia is very restricted geographically, and has a history of relying on outdated techniques and data.

Nonetheless, due largely to the presence of hominins in the region from the Early

Pleistocene, many attempts have been made to understand the palaeohabitats and palaeoenvironments occupied by these hominins. One of the best sources of information on palaeoenvironments, particularly as it applies to early hominins, is that of other mammalian faunas, and Southeast Asia is no exception. In fact, the use of fossil fauna from Southeast Asia to infer palaeoenvironments enjoys a long history (e.g. Dubois,

1908). However, with one exception, these studies have been qualitative, being based on the presence or absence of one or a small number of species (e.g. van den Bergh et al.,

2001; de Vos et al., 1994; Medway, 1977). While these studies are useful, they do not make full use of the entire faunal community preserved in an integrated way.

Most excavations undertaken in Southeast Asia, until recently, have been heavily biased towards larger, more iconic animals, largely due to the fact that these were undertaken at the beginning of last century, and therefore before modern excavation techniques were developed. A recent study by Tougard and Montuire (2006) attempted to reconstruct Pleistocene environments of Southeast Asia using the cenogram method. This

168 method relies on the spectrum of body mass distributions to reconstruct palaeoenvironments (see Legendre, 1986, 1989 for a discussion of this methodology).

However, as Tougard and Montuire (2006) point out in their study, the heavy bias towards large mammals in the palaeontological collections makes the use of the cenogram method in Southeast Asia problematic. As such, they were only able to assign a single palaeohabitat with any confidence to a single fossil fauna, that of Thum Wimam

Nakin, Thailand, which preserved adequate numbers of both small and large mammals.

Given the heavy bias towards large mammals in Southeast Asian collections, any analytical methods that seek to analyse the palaeohabitats of such communities must be equipped to deal with them. Such a method, specifically using only data from large- bodied mammals was developed to study the palaeoenvironments of Africa (Reed, 1997,

1998), and this methodological framework was recently adapted for use in China, on the basis of taxonomic (at the familial level) variables (Chapter 5).

These analyses, using discriminant functions analysis (DFA), rely on modern faunal communities (in the form of species lists from nature reserves and national parks) to classify extinct communities. A database of species lists from parks and reserves has been compiled for Asia (Chapter 5). These parks and reserves span a range of habitats, from deserts to rainforests. However, of these parks and reserves, those from Southeast

Asia include only rainforest (i.e. closed) habitats. Therefore, in order to sample the full range of habitats, the database contains species lists from parks zoogeographically distinct from Southeast Asia (e.g. Eurasian faunal communities). The exclusive use of taxonomic variables from this database could therefore produce skewed results.

169 One potential way to overcome this problem involves the use of phylogeny-free variables (sensu Reed, 1997). This method assigns ecological properties to each species, which are then used to define habitats. These types of variables have been used to classify both African (Reed, 1997, 1998) and European (Mendoza et al., 2005) fossil communities. In fact, zoogeographically distinct regions will preserve habitat structures when examined under phylogeny-free variables (Mendoza et al, 2005), such that it is theoretically possible to use the phylogeny-free discriminant functions defined by either

Reed (1997) or Mendoza et al. (2005) to determine palaeohabitats of Southeast Asian sites. However, because zoogeographical habitats defined on the basis of phylogeny-free variables will nonetheless also preserve a geographical signature (Louys and Bassarova,

2006), more accurate classifications will be obtained by using faunal communities closer biogeographically to the palaeocommunity under examination. Hence, this study will use the faunal lists described in Chapter 5.

The phylogeny-free variables defined by Reed (1997) and Mendoza et al. (2005) require access to original fossils, or at least detailed descriptions of individual specimens, in order to determine palaeodiets. Due to the nature of the palaeontological record of

Southeast Asia, access to fossils cannot always be guaranteed, and in some cases descriptions of fossil material by early excavators have been either perfunctory or even non-existent (for example, the fossil faunas of Laos are, to the best of my knowledge, only described as species lists). Hence there is a need to develop not only phylogeny-free, but also “material-free” palaeoecological variables. The variables defined in this study have been chosen such that they can be determined, in almost all cases, directly from

170 species lists themselves. The use of this method should allow the palaeoenvironments of

most Southeast Asian sites to be determined.

6.3 Methods

Definitions of habitat types, reserves and parks characteristic of these habitats,

and species lists for each park follows Chapter 5. As discussed above, in order to account

for the fact that most parks and reserves listed are zoogeographically distinct from

Southeast Asia, ecological variables were determined for each species. Thus, each species

was categorised using three basic criteria. Firstly, they were categorised according to

average body mass. These were calculated from published accounts (Nowak, 1999), or

estimated by comparison with most closely related taxa. These body size categories are

arbitrary but the variables have been used in many other taphonomic and

palaeoecological studies (e.g. Andrews et al., 1979; Gagnon, 1997; Andrews and

Humphries, 1999; Soligo and Andrews, 2005) and as such allow for direct comparison

with these studies. The categories are as follows: tiny (A) < 1kg; small (B) 1-10kg;

medium (C) 10-45kg; large (D) 45-180kg; and very large (E) >180kg. Taxa belonging to

body-weight category A were excluded from further analysis, because they are not

considered megafauna and therefore likely to be significantly under-represented (if at all)

in Pleistocene sites in Southeast Asia.

The second category is that of trophic guild, specifically, primary (P) and

secondary (S) consumers. Primary consumers are defined as those whose intake of non-

plant matter is very small or non-existent. Secondary consumers include hyper-

carnivores, carnivores, piscivores, omnivores, insectivores, as well as any other trophic

guild where non-plant matter makes up a significant part of the diet. Again, these

171 categories have been used by other researchers (e.g. Soligo and Andrews, 2005). The

final grouping categorises species on the basis of their arboreality. Those species

spending some to all of their time in trees are classified as arboreal (A). This includes

species that are considered scansorial. Terrestrial species (T) are defined as those who

spend no time in trees. Again, where information on arboreality was not available, it was

estimated by comparison with closest related taxa. While some species may have been

incorrectly classified due to the limited amount of ecological information available, the

use of multivariate analyses ensures that, unless the majority of species are incorrectly

classified, significant results will be obtained. Therefore, each species is allocated to one

of twenty variables (examples of species belonging to each of these categories are

provided in Table 6.1). Species from each nature reserve and national park used in

Chapter 5 were therefore allocated to these categories, and percentage representation

calculated. Fossil taxa from thirty sites in Southeast Asia were likewise defined by these

twenty variables based either on their current forms (in the case of extant taxa), or based

on comparisons with their nearest living relatives (in the case of extinct taxa). The geographical location of the fossil sites examined in this analysis are shown in Figure 6.1.

Category Description Examples BPA Small, arboreal primary Hylobates, Presbytis consumer

BPT Small, terrestrial primary Tragulus, Lepus consumer BSA Small, arboreal Ailurus, Melogale secondary consumer

BST Small, terrestrial Vulpes, Lutra secondary consumer CPA Medium, arboreal Macaca primary consumer

172 CPT Medium, terrestrial Naemorhedus, Hystrix primary consumer

CSA Medium, arboreal Neofelis, Arctictis secondary consumer

CST Medium, terrestrial Cuon, Lynx secondary consumer

DPA Large, arboreal primary Pongo consumer

DPT Large, terrestrial primary Cervus, Axis consumer

DSA Large, arboreal Manis palaeojavanica secondary consumer

DST Large, terrestrial Ursus, Panthera tigris secondary consumer

EPA Very large, arboreal None primary consumer EPT Very large, terrestrial Bos, Elephas primary consumer

ESA Very large, arboreal None secondary consumer EST Very large, terrestrial Sus secondary consumer

Table 6.1: Ecological variables used in discriminant functions analysis, and examples of mammals from Southeast Asia falling within those categories

173

Figure 6.1: Location of Pleistocene Southeast Asian sites considered in this chapter

174 The ecological variables were subjected to a principle coordinates analysis (PCA) in order to verify that the habitat groupings were reflected by the ecological categories defined above. This analysis was run on PAST software (Hammer et al., 2001). In addition to confirming groupings, an examination of the loadings for each of the components allows a determination of which ecological variables characterise each habitat type. Changes from one habitat type to another should be reflected in changes to

these loadings.

The species lists for Quaternary sites in Southeast Asia are represented by, in

some cases, quite a small number of megafauna species. Multivariate methods require a

minimum number of data-points in order to accurately assign a variable to a particular

category. In the case of palaeoecological methods, a minimum number of species from a

site is required before a determination of a habitat type for that site can be made. In order

to determine this number, the megafauna species in common for each major modern

habitat type (i.e. closed, mixed and open) was determined. The similarity of each habitat

type to the other was calculated using the Dice similarity index, using PAST software

(Hammer et al., 2001). These similarity values were subsequently used to determine the

minimum number of species necessary to differentiate between closed, mixed and open

habitat types, using equations described by Travouillon et al. (2007). These values were

compared with the number of megafauna species present in each site. Those sites falling

below the minimum number of species were excluded from further analysis.

A box-plot of the range of the ecological variables across all modern species lists

was calculated. These were compared to the Pleistocene species lists in order to identify

any further biases. Biases were subsequently mitigated by standardisation of both the

175 modern and fossil lists. The standardised values from modern habitats were used to define the discriminant functions, calculated using SPSS for Windows software. These

functions were then used to classify all Pleistocene sites.

6.4 Results

Table 6.2 lists the percentages of each ecological variable for each of the modern

reserves and parks, while Table 6.3 lists the percentages for these variables present in

Pleistocene sites. The results of the PCA (Figure 6.2) confirm the characterisation of the

three habitat types as established in Chapter 5 on the basis of the ecological variables

defined in this paper. In particular, the first two components (explaining 40% and 20% of

the variation, respectively) best separate the different habitat types. Negative values of

PC 1 are generally indicative of closed and open habitats, with the more negative values

associated with open habitats. Positive values for this component are generally indicative

of mixed habitats. PC 2 most clearly differentiates closed habitats from mixed and open

habitat types, with the latter values negative or close to zero. An analysis of the loadings

for these two components (Table 6.4, Figure 6.3) suggests the variables CPT and EPT

help separate open habitats, BPT, BSA, BST, CST help separate mixed habitats, and

BPA, CPA, CSA, EST help separate closed habitats, while DPT and DST separate

mixed/open from closed habitats.

176

Figure 6.2: PCA of modern reserves and parks using ecological variables. A: PC1 vs. PC2; B: PC1 vs. PC3; C: PC2 vs. PC3. Blue represents open habitats, red closed habitats and green represents mixed habitats.

177

Figure 6.3: Loadings for the first two PCA components. Colour scheme follows Figure 6.2, with the addition of brown representing mixed/open habitats.

178 Reserve/Park Habitat BPA BPT BSA BST CPA CPT CSA CST DPA DPT DSA DST EPT EST

Ailaoshan Closed 0.14 0.07 0.14 0.1 0.1 0.21 0.14 0 0 0 0 0.03 0.03 0.03

Changshanerhai Closed 0.09 0.18 0.14 0.09 0.09 0.05 0.14 0 0 0.05 0.05 0.05 0.09 0

Gaolingong Closed 0.07 0.1 0.14 0.1 0.14 0.1 0.1 0.03 0 0 0.03 0.07 0.07 0.03

Maolan Closed 0.11 0.16 0.11 0.16 0.05 0.26 0 0 0 0 0.05 0.05 0 0.05

Medong Closed 0 0.05 0.05 0.16 0.16 0.26 0.05 0 00.05 0.05 0.05 0.11 0

Sai Yok Closed 0.14 0.06 0.11 0.08 0.08 0.08 0.11 0.03 0 0 0.03 0.11 0.14 0.03

Khao Yai Closed 0.09 0.09 0.16 0.13 0.03 0.06 0.13 0.06 0 0 0.03 0.09 0.09 0.03

Kaeng Krachan Closed 0.09 0.09 0.15 0.03 0.09 0.09 0.09 0.09 0 0 0.03 0.09 0.15 0.03

Erawan Closed 0.09 0.09 0.13 0.13 0.09 0.07 0.11 0.02 0 0.02 0.02 0.09 0.11 0.02

Kerinci Seblat Closed 0.16 0.09 0.18 0.09 0.04 0.04 0.13 0.09 0 0 0 0.04 0.09 0.04

Dinghushan Mixed 0 0.06 0.13 0.31 0 0.13 0.13 0.06 0 0 0.06 0.06 0 0.06

Fanjingshan Mixed 0.09 0.09 0.09 0.18 0.09 0.15 0.09 0.09 0 0 0.03 0.09 0 0.03

Wuyishan Mixed 0.03 0.13 0.1 0.2 0.07 0.13 0.13 0.07 0 0 0.03 0.07 0 0.03

Funiushan Mixed 0 0.08 0.17 0.21 0 0.21 0.04 0.13 0 0 0.04 0.08 0 0.04

Nujiang Mixed 0.09 0.09 0.15 0.18 0.06 0.12 0.03 0.12 0 0.03 0.03 0.09 0.03 0

Tianmushan Mixed 0 0.07 0.11 0.21 0.04 0.18 0.14 0.11 0 0 0.04 0.07 0 0.04

Yellow River Mixed 0 0.25 0.25 0.25 0 0 0 0.25 0 0 0 0 0 0

179 Maojingba Mixed 0 0.08 0.08 0.25 0 0.08 0 0.17 0 0 0.08 0.17 0 0.08

Yunwushan Mixed 0 0.13 0.13 0.38 0 0 0 0.25 0 0 0 0.13 0 0

Longqi Mixed 0.05 0.14 0.14 0.14 0.05 0.1 0.1 0.05 00.05 0.05 0.1 0 0.05

Poyang Mixed 0 0.14 0.29 0.14 0 0.07 0.14 0.14 0 0 0 0 0 0.07

Baima Xueshan Open 0.13 0 0.13 0 0.13 0.25 0.13 0 0 0.25 0 0 0 0

Bogdhad Open 0 0 0.11 0.11 0 0.11 0 0 00.33 0 0.11 0.22 0

Jiuzhaigou Open 0.1 0 0.1 0.1 0 0.2 0 0 0 0.3 0.1 0 0.1 0

Altun Mountain Open 0 0.13 0.06 0.13 0 0.13 0 0.06 0 0.19 0 0.19 0.13 0

Table 6.2: Relative proportions of species in each of the ecological categories for the park and reserve used in Chapter 5.

180

Pleistocene Site BPA BPT BSA BST CPA CPT CSA CST DPA DPT DSA DST EPT EST # megagauna # megagauna species Tham Khuyen 27 0.04 0.11 0.04 0.07 0.07 0.04 0 0.04 0.04 0.15 0 0.07 0.22 0.11 Tham Hai 20 0 0.05 0.05 0.1 0.05 0 0 0.05 0.05 0.25 0.05 0.1 0.2 0.05 Tham Om 31 0 0.06 0.1 0.1 0.03 0.16 0.03 0.03 0 0.13 0.03 0.06 0.23 0.03 Hang Hum 1 25 0 0.08 0.08 0.08 0.04 0 0.04 0.12 0 0.12 0.04 0.08 0.2 0.12 Hang Hum 2 19 0 0.11 0.11 0 0.05 0 0.05 0.05 0.05 0.16 0.05 0.05 0.26 0.05 Lang Trang 21 0.05 0.05 0 0.1 0.05 0.1 0.05 0.05 0.05 0.1 0.05 0.14 0.19 0.05 Keo Leng 24 0.04 0.08 0.04 0.08 0.13 0.13 0 0 0.04 0.13 0 0.04 0.17 0.13 Tjidulang 3 0 0 0 0 0 0 0 0 0 0 0 0 1 0 Kali Glagah 11 0 0 0 0.09 0 0 0 0 0 0.27 0 0 0.45 0.18 Satir 6 0 0 0 0 0 0 0 0 0 0.33 0 0 0.5 0.17 Bukuran 6 0 0 0 0 0 0 0 0 0 0.33 0 0 0.5 0.17 Ci Saat 9 0 0 0 0 0 0.22 0 0 0 0.33 0 0 0.33 0.11 Trinil 21 0.05 0.05 0 0 0.05 0.1 0.05 0.05 0 0.24 0 0.05 0.29 0.1 Kedung Brubus 23 0.04 0.04 0 0.04 0.04 0.04 0.04 0 0 0.13 0.04 0.09 0.39 0.09 Ngangdong 15 0.07 0.07 0 0 0.07 0.07 0.07 0 0 0.13 0 0.07 0.4 0.07 Punung 18 0.11 0.06 0 0 0.11 0.11 0.06 0 0.06 0.06 0 0.11 0.22 0.11 Holocene Caves Java 26 0.04 0.12 0.12 0.08 0.04 0.04 0.08 0.04 0 0.12 0 0.08 0.19 0.08 Kinta Valley 8 0 0 0 0.13 0 0 0 0 0 0.13 0 0 0.63 0.13

181 Niah 22 0.14 0.09 0.05 0.09 0.09 0.05 0.14 0 0.05 0.09 0.05 0.05 0.09 0.05 Lida Ajer 18 0.22 0.06 0 0 0.11 0.11 0 0 0.06 0.11 0 0.06 0.17 0.11 Sibrambang 20 0.2 0.05 0 0 0.05 0.1 0 0 0.05 0.1 0 0.1 0.25 0.1 Thum Wimam Nakin 29 0.07 0.03 0.07 0.1 0.03 0.07 0 0.03 0.03 0.21 0 0.03 0.24 0.07 Thum Phra Khai Phet 8 0 0 0 0 0 0.25 0 0.13 0 0.25 0 0 0.25 0.13 Ban Fa Suai 34 0.03 0.03 0 0.06 0.03 0.18 0 0.03 0.03 0.18 0 0.09 0.26 0.09 Kao Pah Nam 7 0 0.14 0 0 0 0 0 0.14 0 0.29 0 0.14 0.29 0 Phnom Loang 11 0 0 0 0.09 0 0.09 0 0.09 0.09 0.18 0 0.09 0.36 0 Mogok Caves 8 0 0.13 0 0 0 0 0 0 0 0.25 0 0 0.5 0.13 Irrawaddy beds 19 0 0 0 0 0 0 0 0 0 0.16 0 0 0.79 0.05 Tam Nang 36 0 0.06 0.11 0.03 0.08 0.14 0 0.08 0.03 0.08 0 0.08 0.22 0.08 Tam Hang 13 0 0.08 0 0 0.08 0 0 0 0.08 0.08 0.08 0.15 0.38 0.08

Table 6.3 Proportion of species in each ecological category for thirty Southeast Asian sites. For the geographical position of these sites, see Figure 6.1. Full species list for Pleistocene sites can be found in Appendix B.

182

Ecological Category PC 1 PC 2 BPA -0.172 0.3368 BPT 0.2708 0.09701 BSA 0.1526 0.1214 BST 0.4617 -0.3508 CPA -0.1407 0.3226 CPT -0.2937 0.03998 CSA -0.01904 0.4438 CST 0.4389 -0.2123 DPA 3.25 x 10-24 6.86 x 10-22 DPT -0.5316 -0.5895 DSA -0.01516 -0.02521 DST 0.04906 -0.1732 EPT -0.2681 -0.08747 EST 0.06724 0.07696

Table 6.4: PCA loading for the first two components. Bold values indicate the highest correlation value for the each ecological variable.

Habitat Closed Mixed Open Closed 1 0.50314 0.18045 Mixed 0.50314 1 0.33333 Open 0.18045 0.33333 1

Table 6.5: Dice similarity index between closed, mixed and open habitats.

Minimum Minimum number of Max Dice Max number of number of species Correlation species species (rounded value) Closed/Mixed 50.314 98 32.89956 32 Closed/Open 18.045 98 7.747749 7 Mixed/Open 33.333 61 11.1154 11

Table 6.6: Minimum numbers of species necessary to confidently differentiate between the major habitat types.

Table 6.5 lists the Dice similarity indices of the three major habitat types, while

Table 6.6 lists the relative species richness and the minimum number of species necessary to discriminate between these. A much higher number of species (32) is necessary to

183 differentiate between mixed and closed habitats than between mixed and open (11) and closed and open (7). Very few fossil sites from Southeast Asia record 32 species of megafauna (see Table 6.2). Therefore, the lowest minimum species richness value (7) will be used in this analysis, which allows differentiation between closed and open environments. Even allowing for this lower value, some species lists are still too small to enable multivariate analysis (Table 6.2). These sites are not further considered here.

The box-plots of raw data are shown in Figure 6.4A. As predicted, the Pleistocene data showed a large bias towards large and very large mammals, most likely resulting from collecting bias (see Chapter 5). In order to minimise this positive skew, the ecological variables were standardised by taking the square-root of all proportions. The subsequent box-plot of standardised data for each ecological variable (for both modern and Pleistocene lists) is shown in Figure 6.4B. The range and mean of the Pleistocene variables are clearly more congruous with those obtained for modern sites.

184

Figure 6.4: Box plots of modern and Pleistocene distribution of ecological variables. A, raw data; B, standardised data.

The results of the discriminate functions analysis (DFA) of modern faunas are shown in Figure 6.5. Five variables were selected for use in the discrimant function, these

185 being BST (small, terrestrial secondary consumers), CSA (medium, arboreal secondary consumers), DPT (large, terrestrial primary consumers), CST (medium, terrestrial secondary consumers) and EPT (very large, terrestrial primary consumers). The first function calculated accounted for 62.1% of the variance observed (0.953 canonical correlation), while the second accounted for 37.9% of the variance (0.926 canonical correlation). All reserves were correctly classified in a discriminant function calculated using all lists except that one. Each fossil site was subsequently allocated to a habitat on the basis of these functions (Table 6.7).

Figure 6.5: DFA showing the classification of Southeast Asian Pleistocene sites. Colour scheme follows that of Figure 6.2. Squares, diamonds and crosses represent modern parks and reserves.

186

Other habitat Number of Habitat possibilities Age Glacial/ Period Country Site megafauna (primary DF (more likely/ (ka) Interglacial (OIS) species classification) less likely Early Irrawaddy Burma 19 closed mixed Pleistocene beds Early ?Glacial Java Ci Saat 9 closed mixed/open ~1200 Pleistocene (OIS36) Early Java Kali Glagah 11 closed Pleistocene Early Java Trinil 21 open ~900 ?Intergacial (OIS23) Pleistocene

Middle Burma Mogok Caves 8 closed mixed/open Pleistocene Middle Cambodia Phnom Loang 11 mixed closed Pleistocene Middle Kedung ~700- Largely Glacial Java 23 closed mixed Pleistocene Brubus 800 (OIS17-20) Middle Java Ngangdong 15 closed mixed Pleistocene Middle Laos Tam Nang 36 mixed Pleistocene Middle Laos Tam Hang 13 closed mixed Pleistocene Middle Malaysia Kinta Valley 8 closed mixed/open Pleistocene Middle Thailand Kao Pah Nam 7 open mixed ~690 ?Interglacial (OIS17)

187 Pleistocene Thum Middle Thailand Wimam 29 mixed closed 169 Glacial (OIS6) Pleistocene Nakin Middle Thum Phra Thailand 8 open mixed 169 Glacial (OIS6) Pleistocene Khai Phet Middle Thailand Ban Fa Suai 34 mixed Pleistocene Middle 140- ?Glacial (OIS6,8)/ Vietnam Tham Om 31 mixed closed Pleistocene 250 ?Interglacial (OIS7) Boundary between Middle Vietnam Tham Khuyen 27 mixed closed 475 glacial (OIS12) and Pleistocene interglacial (OIS13) Boundary between Middle Vietnam Tham Hai 20 mixed closed 475 glacial (OIS12) and Pleistocene interglacial (OIS13)

Late Borneo Niah 22 mixed closed 39-45 Interglacial (OIS3) Pleistocene Boundary between Late Java Punung 18 closed mixed 128 interglacial (OIS5) Pleistocene and glacial (OIS6) Boundary between Late Sumatra Lida Ajer 18 closed mixed 128 interglacial (OIS5) Pleistocene and glacial (OIS6) Boundary between Late Sumatra Sibrambang 20 closed mixed 128 interglacial (OIS5) Pleistocene and glacial (OIS6) Late ?Interglacial (OIS5)/ Vietnam Hang Hum I 25 mixed closed 80-140 Pleistocene ?Glacial(OIS4,6)

188 Late ?Interglacial (OIS5)/ Vietnam Hang Hum II 19 open 80-140 Pleistocene ?Glacial(OIS4,6) Boundary between Late Vietnam Lang Trang 21 mixed closed 128 interglacial (OIS5) Pleistocene and glacial (OIS6) Late ?Glacial (OIS2)/ Vietnam Keo Leng 24 closed mixed 20-30 Pleistocene ?Interglacial (OIS3)

Holocene Holocene Java 26 mixed closed 10-6 Interglacial (OIS1) Caves Java

Table 6.7: Palaeoecological classification of Pleistocene sites on the basis of the DFA. Where number of megafauna species precludes the unambiguous classification of palaeohabitat, other possibilities are indicated. Chronology of sites follows Louys et al. (2007) and Louys (2007).

189 The sites of Thum Phra Khai Phet and Kao Pah Nam (Thailand) are classified as open habitats. Although these sites have sufficient numbers of species to differentiate between open and closed habitats, they do not have enough to differentiate between open and mixed environments. On the basis of their discriminant scores, however, these are more likely to represent open, rather than mixed, habitats. Trinil and Hang Hum II are also classified by the DFA as open, although an examination of their relative position to the habitat group centroids show that they fall much closer to mixed habitats than the aforementioned Thai sites. They preserve, however, sufficient numbers of species to differentiate between open and mixed habitats, and as such should be classified as open.

Due to the generally low number of megafauna species currently recorded for

Pleistocene deposits in Southeast Asia, only two sites could confidently be allocated to a mixed habitat: Tam Nang, Laos and Ban Fa Suai, Thailand. Their positions are, however, close to those of Trinil and Hang Hum II, both classified as open sites. These four sites therefore in all likelihood preserve a gradation from more open habitats (Trinil and Hang

Hum II) to slightly more wooded habitats (Tam Nang and Ban Fa Suai), on the basis of the mammals present.

All other sites, classified as either mixed or closed habitats, do not preserve sufficient numbers of species to confidently attribute them to either habitat. What can be determined, however, is that these sites, with the exception of Ci Saat, Mogok Caves and

Kinta Valley, are unlikely to represent open habitats. These last three sites, although preserving sufficient numbers of species to differentiate between closed and open habitats, do not preserve sufficient numbers to differentiate between mixed and open and

190 mixed and closed habitats, therefore making unambiguous classifications difficult to establish.

6.5 Discussion

An examination of the PCA results shows that open habitats, such as those represented by savannahs and deserts, are characterized by the smallest number of ecological variables. The variables separating open habitats from all other habitats are the medium terrestrial primary consumers and very large terrestrial primary consumers.

These results are congruous to that expected: that open habitats would have a greater relative proportion of larger, terrestrial, primary consumers, especially grazers. Of these variables, very large terrestrial primary consumers were also a selected ecological variable with regards to the DFA, again highlighting this characteristic of open habitats.

Mixed habitats were separated in the PCA by large proportions of four ecological variables: small terrestrial primary consumers, small secondary consumers, both arboreal and terrestrial, and medium terrestrial secondary consumers. Of these only the proportion of small and medium terrestrial secondary consumers was selected by the DFA in differentiating between habitats.

Closed habitats are differentiated from open and mixed habitats by large representation in four variables, and a smaller representation in two variables. Of the larger proportions, these are mostly arboreal categories (small and medium primary consumers, and medium secondary consumers), as well as very large, terrestrial secondary consumers. Smaller proportions of large, terrestrial, primary and secondary consumers differentiates between closed and mixed/open habitats along principle component 2. Of the ecological variables helping to separate closed habitats, only

191 medium arboreal, secondary consumers and large, terrestrial primary consumers were selected by the DFA in the classification of habitats.

On the basis of these results, some expectations of changes in proportion of these ecological variables can be determined across suspected ecological changes. In particular, the presence of a central savannah corridor running through the middle of Sundaland, as proposed by a number or researchers (e.g. Bird et al., 2005; Heaney, 1991) should be traceable through an ecological analysis of fossil habitats. In particular, a change from a closed (i.e. rainforest) habitat to either a mixed or open (i.e. open forest or savannah) habitat should be reflected by a significant decrease in the proportion of medium, arboreal secondary consumers. Mammals from Pleistocene sites described by this variable include the small cats (e.g. Catopuma and Prionailurus species), the pangolin

(Manis), the clouded leopard (Neofelis) and the binturong (Arctictis). Concomitant with these are relative increases in the proportion of large and very large terrestrial, primary consumers and small and medium terrestrial secondary consumers. Mammals from

Pleistocene sites described by these variables include the dhole (Cuon), Asian spotted hyena (Crocuta), as well as the cervids, bovids, rhinos and elephants.

Differentiating between open and mixed habitats will be more difficult than differentiating between closed and open habitats. This is reflected firstly in the higher species richness required (Table 6.6). Further, although component 2 in the PCA separates closed habitats, it does not separate mixed from open habitats. Nevertheless, significant relative changes between proportions of large, terrestrial primary consumers and small and medium terrestrial secondary consumers should indicate transition from mixed habitat to open.

192 6.5.1 Early Pleistocene sites

The classifications of Pleistocene sites on the basis of the DFA are largely congruous with those proposed by qualitative analyses. However, the classification of some sites is unexpected. In particular, most of the Early Pleistocene sites are classified as closed habitats. This is surprising because Ci Saat has previously been interpreted to represent a more open habitat on the basis of pollen evidence (de Vos et al., 1994).

However, an examination of Ci Saat’s position relative to the group centroids for the habitats (Figure 6.5) shows that its position approaches that of open habitats. Perhaps it represents a mostly closed forest with, nevertheless, significant grassland components.

The older Kali Glagah fauna represents a more likely closed forest, as evidenced firstly by its position relative to habitat group centroids (Figure 6.5), as well as preserving sufficient numbers of species to confidently differentiate between habitats. The transition from closed to more open environments through the Early Pleistocene of Java agrees broadly with previous palaeoenvironmental reconstructions that assigned mangrove/swampy conditions to Satir fauna (2.0-1.5 Ma) and drier, more grassland conditions for the Ci Saat fauna (1.2-1 Ma) (de Vos et al., 1994; van den Bergh et al.,

2001). Trinil’s classification agrees largely with previous palaeoenvironmental reconstructions, which, on the basis of the large numbers of bovids present at the site, was classified as open woodland (de Vos et al., 1994; van den Bergh et al., 2001; Storm,

2001). The presence of grasslands during the deposition of Trinil is also suggested by the avian fauna present at the site (Meijaard, 2004). Its almost intermediate position between open and mixed habitats in the DFA suggests predominately grassland conditions, but does not exclude the presence of some woodland. The classification of the Irrawaddy

193 Beds is congruous with research suggesting tropical forests were present throughout the

Indochinese province, and as far north as the Qinling Mountains (Jablonski and Whitford,

1999; Chapter 5).

6.5.2 Middle Pleistocene sites

Of Southeast Asian Middle Pleistocene sites, only Kao Pah Nam and Thum Phra

Khai Phet (Thailand) are classified as open habitats. Kao Pah Nam is confidently assigned to an open habitat on the basis of both species numbers as well as its canonical position (Figure 6.5). This largely supports Pope et al.’s (1981) palaeoenvironmental interpretation of “relatively open, dry dipterocarp woodland” (Pope et al., 1981, p. 161).

Thum Phra Khai Phet has been considered, on the basis of taxonomic similarity, to be contemporaneous with Thum Wimam Nakin (Tougard, 1998, 2001). In this analysis

Thum Wimam Nakin is classified as representing a mixed habitat, which agrees with previous palaeoenvironmental reconstructions suggesting it represents “a slightly open forested environment” (Tougard and Montuire, 2006). Thus, the classification of Thum

Phra Khai Phet as open is likely to be a result of the inadequate number of large-bodied taxa preserved at the site (Table 6.7), which do not allow differentiation between open and mixed habitats.

The faunal turnover between Trinil and Kedung Brubus at approximately 0.8 Ma

(van den Bergh et al., 2001) is reflected in the results, with both Kedung Brubus and the later Ngandong faunas being classified as likely closed habitats. However, these results are at odds with previous reconstructions that suggested that Kedung Brubus and

Ngandong faunas represented open woodland environments (de Vos et al., 1994; van den

Bergh et al., 2001). These previous reconstructions were based on the number of bovids

194 present in the fauna, as well as the absence of primates (de Vos et al., 1994). An analysis of the DFA shows that that closed and mixed habitats are separated along discriminant function (DF) 1, with closed habitats having generally negative DF1 values. A breakdown of DF1 reveals that a higher proportion of large, terrestrial primary consumers will contribute to lower DF1 values. Equally, lower (to zero) values of the other four ecological categories (BST, DPT, CSA and CST) will also drive DF1 towards the negative. In fact, it appears that the lower than average numbers of small and medium terrestrial secondary consumers (e.g. viverrids, mustelids, canids), more than any other variable, has driven DF1 towards the negative, and thereby produced the closed habitat classification. This is confirmed by an examination of the PCA loadings for each habitat, which indicates that small and medium terrestrial secondary consumers are principle components in mixed habitats.

The palaeoenvironmental reconstructions for Laos represent the first for this country. Tam Nang is one of two sites confidently assigned, on the basis of both DFA and species numbers, as a mixed habitat. However, its position relative to the DFA group centroids also reveals it approaches open habitats. Therefore, it is herein interpreted as a mixed habitat with significant open areas. Interestingly, Tam Hang is classified as a closed habitat; however, the small number of species preserved at this site makes this classification uncertain, and it may also in fact represent a mixed habitat.

Likewise, Vietnam and Cambodia have received little previous palaeoecological attention. The three Vietnamese sites examined in this analysis (Tham Khuyen, Tham

Hai, Tham Om) are each classified as mixed, but they do not preserve sufficient species to confidently distinguish between mixed and closed forests. Of these three sites, Tham

195 Om is the most southern geographically. With respect to the DFA, it is closest to that of the mixed habitat group centroid. In terms of age, it most closely approaches that of

Tham Wimam Nakin, Thailand, which is also considered a mixed habitat. The site of

Phnom Loang, Cambodia, is also classified as mixed habitat, but again too few species are preserved to rule out that it may represent a closed habitat. The discriminant functions calculated for this site are almost identical to that of Tham Hai, Vietnam, suggesting that these two sites represent the same habitat.

Kinta Valley, Peninsula Malaysia, and Mogok Caves, Burma, are both classified as closed, but the number of species preserved at both sites means they could in fact represent mixed habitats, or, less likely, open habitats. In the DFA, Kinta Valley is positioned close to the closed habitat group centroid than Mogok Caves, which has a much higher DF2 score. This suggests a stronger likelihood for Kinta Valley representing a closed habitat, while Mogok Caves’ proximity to the open habitat group centroid suggests its classification as closed may reflect its limited species list. The interpretation of Mogok Caves representing an open habitat would indicate an opening up of vegetative structure in Burma from the Early to Middle Pleistocene, and is consistent with the suggestion that the tropical/subtropical zones shifted southwards during this period

(Jablonski and Whitford, 1999).

6.5.3 Late Pleistocene and Holocene sites

Niah Cave, Borneo, is classified as a mixed habitat, although not enough species are preserved at this site to confidently distinguish it from a closed habitat. However this classification agrees with previous qualitative reconstructions on the basis of both fauna

(e.g. Medway, 1973, 1977; Harrison, 1996) and flora (e.g. Barker et al., 2007).

196 Punung, Java, is classified as closed (again, however, it could represent a mixed habitat on the basis of numbers of species present). This agrees with previous palaeoenvironmental reconstructions that propose it represents a tropical rainforest (de

Vos, 1983; Storm, 2001; Storm et al., 2005; van den Bergh et al., 2001). It has been suggested that a faunal turnover occurred on Java between the Middle Pleistocene site of

Ngandong and the Late Pleistocene site of Punung, representing a shift from open woodlands to rainforest (Storm, 2001; Storm et al., 2005; van den Bergh et al., 2001).

However, on the basis of the DFA results reported, this faunal turnover (which is not herein disputed) does not appear to be accompanied by any large-scale habitat changes.

As such, another agent may possibly be responsible. This agent may be simply a reconnection of Java with the mainland. It could also represent the results of the activities of modern Homo sapiens, the earliest record of which may be in Punung (Storm et al.,

2005).

The sites of Lider Ajer and Simbrambang, Sumatra, share the same classification as Punung. This is not surprising because they share a high level of taxonomic similarity with Punung (de Vos, 1983). Furthermore, the fauna from this site is essentially modern – neither Lider Ajer nor Simbrambang record any extinctions (Louys, 2007) – and it is expected that these faunas would correspond to Sumatra’s present-day habitat type.

The sites from Vietnam represent a variety of habitat types. Interestingly, the only site to preserve sufficient numbers of species to confidently assign a single habitat type is

Hang Hum II, which is classified as open. An examination of the discriminant functions shows that Hang Hum II is the furthest “open” site from the open habitat group centroid.

On the basis of its discriminant functions, it is quite close to Tam Nang, a site confidently

197 assigned as mixed. The faunas preserved in these two sites therefore most likely represent an ecotone between mixed and open habitats. Other sites from Vietnam are classified as either mixed (Hang Hum I, Lang Trang) or closed (Keo Leng); however, numbers of species preserved at these sites can only confidently exclude them from being open. The two older sites (Hang Hum I and Lang Trang) are likely to represent mixed habitats, while the younger site is likely to represent a closed habitat, suggesting an increase in habitat density during the time period spanned by these sites.

The Holocene caves fauna from Java are classified as most likely representing a mixed habitat. This agrees with previous interpretations of this fauna as open woodland

(van den Bergh et al., 2001; Storm et al., 2005), as well as with modern habitats from

Java (van den Bergh et al., 2001; Storm et al., 2005).

6.6 Conclusion

The results of the DFA have for the first time enabled habitat reconstructions of most Pleistocene sites from Southeast Asia. In most cases insufficient numbers of species were present to allow confident differentiation between mixed and closed habitats, but nevertheless even this information is extremely valuable in palaeoenvironmental studies.

A comparison between previous palaeoecological reconstructions and the reconstructions presented in this study show a high level of congruence. Most importantly, classification of Late Pleistocene and Holocene sites that preserve a modern faunal signature match the habitats currently found there.

The following general trends for the Indo-Chinese and Sundaic provinces can be suggested on the basis of these results. The Indo-Chinese province is characterized by closed habitats in the Early Pleistocene. The Middle Pleistocene sees vegetation density

198 decreasing, with mixed habitats dominating, but some evidence for open habitats is also present. The beginning of the Late Pleistocene sees the continual dominance of mixed habitats. By the end of the Late Pleistocene these give way to the modern, closed habitat structure seen today. The Sundaic province is characterized by closed habitats at the beginning of the Early Pleistocene, but conditions become drier, and more open habitats dominate by the end of the Early Pleistocene. Closed habitats return at the beginning of the Middle Pleistocene, and these continue for most of this period and into the Late

Pleistocene. The end of the Late Pleistocene and Holocene sees the dominance of mixed habitats, with the return to closed habitats in the Sundaic province probably only occurring after the Last Glacial Maximum.

199 CHAPTER 7

SUMMARY

7.1 Indochinese Province

7.1.1 Southern China

Palaeoenvironmental changes occurring during the Pleistocene of China have

been well documented by a number of authors (e.g. Jablonski and Whitford, 1999;

Ferguson, 1993). The large scale changes associated with the Pleistocene of China include the southern retreat of the tropical and subtropical zones, until their restriction to the most southern parts of China at the beginning of the Holocene (Jablonski and

Whitford, 1999).

The northern limit of the subtropical zone during the Early Pleistocene, however, is a topic which has been subject to some ambiguity. In particular, the site of

Gongwangling, in Lantian county, China, previously classified as representing either a northern (Eurasian) or a southern (Indochinese) fauna (see Chapter 5), suggests the exact extent of this zone was unknown. The analysis of this site provided in Chapter 5 strongly suggests that the megafauna represents a southern Chinese fauna, and as such

Gongwangling represents one of the most northern zoogeographic points of Southeast

Asia during the early Pleistocene. The extinction of the megafauna from Gongwangling can be ascribed to a loss of suitable habitat (Chapter 5), which can in turn be related to the climatic shifts documented for this period.

200 The ecology of early hominins in Asia is still unresolved. There is no evidence

from the study of Gongwangling that the hominins recovered from that site interacted

with megafauna (Chapter 5). Although some Asian sites do show clear signs of hominin

utilization of animals (e.g. Panxian Dadong; Schepartz et al., 2005), these traces are still

rare, and it is equivocal as to whether these hominins practiced a predominantly hunting or scavenging ecology. Even assuming that hominins in Asia employed a successful hunting-based ecology, there is no evidence suggesting any taxa were hunted to extinction, at least for the majority of the Pleistocene. Those taxa that were utilized by

hominins are either still extant or survived until the Holocene (e.g. Stegodon, Rhinoceros,

Equus). Based on current evidence, it seems most likely that the majority of the

extinctions of megafauna in Pleistocene Southern China were driven by loss of suitable

habitat, as the example of Gongwangling intimates.

Southern China shares a number of Quaternary megafauna extinctions with other

countries in Southeast Asia. A species each of horse, gazelle and stegodon (Equus

yunnanensis, Gazella sp. and Stegodon elephantoides, respectively) became extinct in both Southern China and Burma; the giant ape Gigantopithecus and the Chinese rhino

(Rhinoceros sinensis) also became extinct in Vietnam; the robust macaque (Macaca robustus) also in Laos; and the giant hyena (Pachycrocuta brevirostris) also in Java. The giant tapir (Megatapirus augustus) also disappeared from Laos and Vietnam; while the stegodon (Stegodon orientalis), in addition to becoming extinct from Southern China in the Holocene, disappeared from Burma and Vietnam during the Pleistocene. The extinction of the Malayan tapir (Tapirus malayanus) is one also shared with Cambodia and Borneo; the serow (Naemorhedus sumatraensis) from Java and Vietnam; the Asian

201 spotted hyena (Crocuta crocuta) from Laos, Cambodia and Thailand; the archaic

elephant (Palaeoloxodon namadicus) from Burma, Laos, Vietnam and Malaysia. Finally

the orangutan (Pongo pygmaeus) became extinct in five countries other than China: Laos,

Cambodia, Vietnam, Thailand and Java.

7.1.2 Burma

At least twelve species of megafauna recovered from the Early Pleistocene

Irrawaddy beds of Burma are no longer currently found in the region. Nine of these

twelve species represent global extinction of the species. Five of these species also

became extinct in other Southeast Asian countries: the Yunnan horse (Equus

yunnanensis, also becoming extinct in Southern China), a species of gazelle (Gazella sp.,

this genus also becoming extinct in Southern China), the archaic elephant

(Palaeoloxodon namadicus, also becoming extinct in Southern China, Laos, Vietnam and

Malaysia), the South Asian rhino (Rhinoceros sivalensis, also becoming extinct in Laos)

and the elephant stegodon (Stegodon elephantoides, also becoming extinct in Southern

China). Based on the palaeoenvironmental reconstruction described in Chapter 6, the

fauna present in the Irrawaddy beds most likely inhabited a closed habitat, although the possibility remains that this fauna represents a mixed (e.g. woodland) habitat.

The Middle Pleistocene Burmese Mogok Caves fauna preserves three species of megafauna no longer found in the region. All three extinctions are shared by other regions in Southeast Asia: the giant panda (Ailuropoda melanoleuca, also becoming extinct in Laos, Vietnam and Thailand), the archaic elephant (Palaeoloxodon namadicus, also becoming extinct in Southern China, Laos, Vietnam and Malaysia) and the stegodon

(Stegodon orientalis, also becoming extinct in Southern China and Vietnam). Two of

202 these extinctions represent global extinctions. The multivariate palaeoecology analysis

(Chapter 6) classifies the Mogok Caves fauna as a closed habitat, although insufficient numbers of species are present to rule out that this represents either a mixed or open habitat. An open classification would in fact be more parsimonious for the region (see

Chapter 6).

The extinction of the majority of the megafauna from Burma may be the result of a shift in environmental conditions from closed to more open habitats during the late

Early Pleistocene to the Middle Pleistocene. Conversely, the presence of the horse and gazelle in the Irrawaddy beds may suggest the presence of more open environments during this time. Data is currently insufficient for this region to link environmental changes directly with extinctions. Likewise, determining a role for hominins in extinctions is not possible on the basis of the available evidence.

7.1.3 Laos

The Middle Pleistocene fossil site of Tam Nang includes 25 extinct species of megafauna. Nineteen of the 25 represent global extinctions. Eight of the 25 represent species extinct in other regions: the robust macaque (Macaca robustus, also becoming extinct in Southern China), the South Asian rhino (Rhinoceros sivalensis, also becoming extinct in Burma), Lydekker’s pig (Sus lydekkeri, also becoming extinct in Vietnam), the giant tapir (Megatapirus augustus, also becoming extinct in Southern China and

Vietnam), the giant panda (Ailuropoda melanoleuca, also becoming extinct in Burma,

Vietnam and Thailand), the spotted hyena (Crocuta crocuta, also becoming extinct in

Southern China, Cambodia and Thailand), the archaic elephant (Palaeoloxodon namadicus, also becoming extinct in Southern China, Burma, Vietnam and Malaysia) and

203 the orangutan (Pongo pygmaeus, also becoming extinct in Southern China, Cambodia,

Vietnam, Thailand and Java). Tam Nang preserves sufficient numbers of megafauna to confidently assign it to a mixed habitat, albeit with significant open areas (Chapter 6).

The site of Tam Hang records the extinction of eight species of megafauna. Of these, seven are global extinctions. Five of the eight species recorded as extinct also become extinct in other Southeast Asian regions. These five species are equally represented at Tam Nang (i.e. the robust macaque, the archaic elephant, Lydekker’s pig, the giant tapir and the South Asian rhino). The palaeoenvironmental classification of Tam

Hang is that of a closed habitat; however it could also represent a mixed habitat.

Unfortunately, dating of Laotian sites is currently non-existent and therefore it is impossible to establish detailed changes in environment over time, and as such link these with specific extinctions in that country. Changes from the mixed habitats represented by

Tam Hang and Tam Nang to the more closed habitats currently found in Laos may have driven some extinctions. Evidence for a human role in extinctions is currently lacking, although this could be due to the very limited palaeoanthropological research that has been conducted in this country.

7.1.4 Vietnam

The Middle Pleistocene site of Tham Khuyen records the extinction of nine species of megafauna. Six of the nine species represent global extinctions. Eight of the nine species are extinctions recorded in other regions: the giant ape (Gigantopithecus blacki, also becoming extinct in Southern China), the orangutan (Pongo pygmaeus, also becoming extinct in Southern China, Cambodia, Laos, Thailand and Java), the archaic elephant (Palaeoloxodon namadicus, also becoming extinct in Southern China, Burma,

204 Laos and Malaysia), the stegodon (Stegodon orientalis, also becoming extinct in

Southern China and Burma), the Chinese rhino (Rhinoceros sinsensis, also becoming extinct in Southern China), Lydekker’s pig (Sus lydekkeri, also becoming extinct in

Laos), the giant panda (Ailuropoda melanoleuca, also becoming extinct in Burma, Laos and Thailand) and the giant tapir (Megatapirus augustus, also becoming extinct in

Southern China and Laos). The results of the palaeoenvironmental reconstruction suggest that this site may have been a mixed habitat, but it could equally have represent a closed habitat.

The site of Tham Hai, considered contemporaneous with Tham Khuyen (Ciochon et al., 1996), records the extinction of four species of megafauna, three of which represent

global extinctions. All four species are extinctions recorded by sites in other regions.

These include the giant ape, the orangutan, the stegodon, and the Chinese rhino. Tham

Hai is equally classified as mixed, but like Tham Khuyen, it could equally represent a

closed habitat.

The site of Tham Om is considered younger than either Tham Khuyen or Tham

Hai (Ciochon et al., 1996). It preserves eight species of megafauna that are no longer found in Vietnam. Five of the eight extinctions represent global extinctions. Six of the

eight extinctions are extinctions recorded in other countries. In addition to sharing with

Tham Khuyen and Tham Hai the extinction of the archaic elephant, the stegodon, the

Chinese rhino, the giant panda, and the giant tapir, Tham Om also records the extinction

of the serow (Naemorhedus sumatraensis, also becoming extinct in Southern China and

Java). Although younger than Tham Khuyen and Tham Hai, Tham Om’s fauna is

classified as belonging to a mixed (but also possibly a closed) habitat. Based on these

205 results, it does not appear that any large scale habitat changes occurred between the

depositions of these three fossil sites. The only extinctions recorded between their

depositions are that of the giant ape and the sun bear (Ursus malayanus).

The Late Pleistocene site of Hang Hum I records the extinction of five species of

extinct megafauna. All of these represent global extinctions, and four of them represent

extinctions found in other Southeast Asian countries (stegodon, Chinese rhino, giant tapir

and Lydekker’s pig). The palaeoecological reconstruction of Hang Hum I suggests it represents a mixed habitat, but low numbers of species present means that it may also represent a closed habitat.

Hang Hum II, considered roughly contemporaneous with Hang Hum I (Olsen and

Ciochon, 1990) records the extinction of two species of extinct megafauna, the orangutan and the archaic elephant. Interestingly, the discriminant functions analysis (DFA) classifies this site as an open habitat, although it is close in the spread of discriminant

functions scores to Tam Nang, which represents a mixed habitat. Considering that the orangutan is found in Hang Hum II, some tree cover must have been present during the deposition of this site. Assuming Hang Hum I is slightly older than Hang Hum II, then the time between their respective depositions would record a measure of opening up of the vegetative structure. Extinct species present in Hang Hum I are not recorded in Hang

Hum II (although most are found in the younger site of Keo Leng), and as such this may

indicative of local extinctions, in turn associated with the recorded changes in vegetation

structure. The extinctions recorded at Hang Hum I and II are unlikely to be a result of the

Toba eruption (Chapter 4).

206 Keo Leng records the extinction of seven species of megafauna. Five of these represent global extinctions, and all these extinctions are shared by other countries in the region. The palaeoecology analysis classifies Keo Leng as a closed habitat, although insufficient numbers of megafauna are preserved to rule out mixed habitats. It therefore records an increase in the denseness of the vegetation structure. Extinctions occurring between the depositions of the older sites (representing mixed to open habitats) and Keo

Leng (representing a closed habitat) could be associated with environmental changes.

Most Vietnamese extinct megafauna in are recorded in Keo Leng, suggesting their extinctions occurred in the Late Pleistocene. If the palaeoecological reconstruction of

Keo Leng is confirmed, then the subsequent megafaunal extinctions cannot be directly associated with habitat change. Although no direct evidence exists, and the argument is highly speculative, it is possible that Late Pleistocene hominins were a factor here. More direct evidence of hominin interaction with extinct megafauna is necessary before such an assertion can be made, lest the problems associated with the megafauna debate in other continents (see Chapter 3) plague future investigations into the extinction of

Southeast Asian megafauna.

7.1.5 Cambodia

Although in Chapter 2 (published as Louys et al., 2007) I considered Phnom

Loang and Préah Vihear separate sites, upon further research, I discovered that they actually represent the same site (Beden and Guérin (1973); amended in Chapter 6). Five species became extinct after Phnom Loang, two of which represent global extinctions.

Four of the five species are also found in other countries in the region: the Malayan tapir

(Tapirus indicus, also becoming extinct in Southern China and Borneo), the orangutan

207 (also found in Southern China, Laos, Vietnam, Thailand and Java), the spotted hyena

(also found in Southern China, Laos and Malaysia) and Sika deer (Cervus nippon, also found in Laos1). The DFA performed in Chapter 6 classifies Phnom Loang as a mixed

habitat, which is in keeping with other Middle Pleistocene sites of this region. Lack of additional fossil sites, and inadequate dating, prohibits direct assessment of the causes of the extinction of the taxa recorded at this site. Extinctions could be related to the change from mixed habitats to the closed habitats present today. However, these sorts of changes are unlikely to have adversely affected the orangutan or the tapir. No evidence currently exists for a human role in extinctions in Pleistocene Cambodia, although this is again compounded by limited palaeoanthropological research in this country.

7.1.6 Thailand

Kao Pah Nam records the extinction of three megafauna, none of these representing a global extinction. Representatives of the orangutan and the spotted hyena genera (Pongo and Crocuta, respectively), listed for this site, also become extinct in other regions (Pongo: Southern China, Burma, Laos, Cambodia, Vietnam and Java; Crocuta:

Southern China, Laos, Cambodia). Based on the palaeoecological analysis, Kao Pah Nam is reconstructed as an open habitat. However, the possible presence of the orangutan in these deposits suggests some measure of tree cover (it is unlikely that Pleistocene orangutans had a different ecology than those of today, for reasons outlined in Chapter 2).

This site is considered significantly older than other Thai sites examined (Chapter 2).

1 Cervus nippon was inadvertently omitted from Tables 2.12 and 3.1. Because these tables have been published, and in order to preserve continuity with published records, this omission is not corrected in the chapters, but is instead corrected here.

208 Thum Wimam Nakin records the subsequent extinction of six species of megafauna, all of which are regional extinctions only. Five of these six extinctions occur in other countries in the region: the orangutan (see above), the giant panda (Ailuropoda melanoleuca, also becoming extinct in Burma, Laos and Vietnam), the spotted hyena (see above), the Indian rhino (Vietnam and Java) and the bearded pig (Sus barbatus, also becoming extinct in Java). Based on palaeoenvironmental reconstructions, Thum Wimam

Nakin represents a mixed habitat. Although the time interval between the deposition of

Kao Pah Nam and Thum Wimam Nakin in all likelihood involved closing of the vegetative structure, all of the extinct megafauna recorded by Kao Pah Nam are also recorded in later sites.

Thum Phra Khai Phet records the extinction of three species of megafauna (the giant panda, the spotted hyena and the bearded pig). All these are equally represented at

Thum Wimam Nakin. Although the DFA classifies Thum Phra Khai Phet as representing an open habitat, this is probably a result of the small number of megafauna present at the site (Chapter 6).

Ban Fa Suai records six species of extinct megafauna. In addition to the orangutan and the giant panda, this site records the extinction of Stegodon sp., this genus being common throughout Pleistocene Southeast Asia. The DFA classifies this site as representing a mixed fauna.

The Middle Pleistocene of Thailand is characterized by mixed habitats. These differ from the habitats currently found there, which are predominately closed (Chapters

3 and 4). The extinction of the megafauna from the Middle Pleistocene Thai sites could be a result of changes from mixed to closed habitats sometime during the Late

209 Pleistocene/Holocene. Testing this hypothesis will involve description of Late

Pleistocene faunas from Thailand. There is currently no evidence to suggest that humans were involved in these extinctions.

7.1.7 Indochinese summary

The Early Pleistocene in Indochina was most likely dominated by closed habitats, and extended further north than it does today. Gongwangling, in Lantian county, China represents one of the most northerly sites where megafauna are of a Southeast Asian nature. By the beginning of the Middle Pleistocene, environmental conditions had begun to change. This period saw a general southern movement of the tropical and sub-tropical zones (Jablonski and Whitford, 1999). These changes are reflected in changes in habitats recorded for this period (Chapter 6). Most Indochinese sites for the Middle Pleistocene are reconstructed as mixed habitats, representing an opening, structurally, of the vegetation complex. This may have impacted negatively on some megafauna, either through loss of habitat, or alternatively by allowing the successful migration of new species perhaps out-competing the old. Mixed habitats continue to dominate throughout most of the Indochinese Late Pleistocene. Closed habitats become established only towards the end of the Late Pleistocene. Habitat changes, brought about by changes in climate, are, based on current evidence, the most likely cause of most megafauna extinctions in the Indochinese province during most of the Pleistocene. The super- eruption of Toba seems to have had little to no effect on extinction of megafauna. The presence of extinct megafauna in the Late Pleistocene site of Keo Leng, whose environment was probably similar to today’s, is the first possible indication of an active human role in Late Pleistocene extinctions in this region. Historical evidence suggests,

210 however, that over the last few thousand years, humans have had an increasingly negative

effect on megafauna (Chapter 3).

7.2 Sundaic Province

7.2.1 Malaysia

The Kinta Valley fauna preserves three species of megafauna currently extinct in

the region. One of these (the archaic elephant) is now globally extinct. Two of these three

species are also extinct in other countries in the region: the archaic elephant

(Palaeoloxodon namadicus, also becoming extinct in Southern China, Burma, Laos and

Vietnam), and Dubois’s antelope (Duboisia santeng, also becoming extinct in Java). The

palaeoecological analysis classifies the Kinta Valley fauna as representing a closed

habitat. Due to the uncertain age of the Kinta Valley fauna, and given the lack of any

other sites in Peninsula Malaysia, any hypotheses on causes of extinction in this region

are premature. Although this site would lie in the proposed savannah corridor during

periods of lower sea-level, due to its uncertain age, it cannot yet be associated with either

glacial or interglacial conditions.

7.2.2 Borneo

The Niah Caves fauna preserves four species of extinct megafauna, all of which

became regionally extinct. In addition to these, Niah Caves also preserves the giant

pangolin (Manis palaeojavanica), a globally extinct congener of the modern pangolin

(not examined in Chapters 2 and 3). Borneo shares at least one extinction with another country in the region, that of the Malayan tapir (Tapirus indicus, also becoming extinct in

Southern China and Cambodia). The DFA classifies the Niah Caves fauna as a mixed

211 habitat. As such, these extinction could be a result of changes in habitat, from a mixed habitat to the more closed habitat found today (this hypothesis has been previously proposed by a number of authors, e.g. Medway, 1972; Cranbrook, 2000; see also

Chapters 2 and 3). Niah Caves preserves one of the few instances of direct human interaction with megafauna, with cut marks found on several bones from the site (Barker et al., 2007). However, this interaction appears to have had no adverse affects on those megafauna utilized by humans (see discussion in Chapter 3).

7.2.3 Sumatra

Although I state in Chapter 2 that Lider Ajer records the extinction of the leopard, this is an error corrected in Chapter 3. The current evidence suggests that no megafauna became extinct on Sumatra, although this is more likely due to the limited number of sites described from Sumatra. The DFA classifies Lider Ajer and Simbrambang as representing closed habitats. This suggests that the Toba super-volcano, erupting at ~74 kya, had no long-term effects on the survival of megafauna species, or for that matter on the structure of the vegetation, either as a result of the initial blast, or through subsequent climatic changes at a geological timescale. The association between the Toba eruption and the proposed human population bottleneck has been challenged in Chapter 4.

7.2.4 Java

Ci Saat preserves six species of extinct megafauna, all of which later became globally extinct. Only one of these, Dubois’s antelope (Duboisia santeng), is also found in another country (Malaysia). Ci Saat is classified by the DFA as a closed habitat; however its position relative to other sites in the analysis, as well as pollen studies

212 (Semah, 1984) suggests significant grassland components (Chapter 6). All of the species of megafauna found in Ci Saat do not become extinct until the Middle to Late

Pleistocene, and these extinctions are likely to be driven by changes in habitat (see below).

Trinil is classified as an open habitat by the DFA, and preserves many of the extinct megafauna preserved at Ci Saat. Like Ci Saat. Almost all the extinctions from this site occur in the Middle to Late Pleistocene. The only exception is that of the Trinil dog

(Cuon trinilensis).

The site of Kedung Brubus records twelve species of extinct megafauna. Only one of these does not represent a global extinction (the Indian rhino, Rhinoceros unicornis).

Three of the twelve extinctions are also recorded in other countries: the Indian rhino (also becoming extinct in Vietnam and Thailand), the giant hyena (Pachycrocuta (Hyaena) brevirostris, also becoming extinct in Southern China) and Dubois’s antelope (also becoming extinct in Malaysia). Six of the twelve occurrences of extinct megafauna at

Kedung Brubus represent last appearances in Java. The DFA classifies Kedung Brubus as a closed habitat. The faunal turnover recorded between Trinil and Kedung Brubus is reflected in the change in habitats between these two sites. However, all extinct species of megafauna from Trinil are also found in later sites, such that the impact of faunal turnover recorded between the deposition of Trinil and Kedung Brubus appears to have had minimal detrimental effect on species survival. The faunal turnover occurring between Kedung Brubus and the younger site of Ngandong, does appear, however, to have included with significant extinctions. Surprisingly, these extinctions do not appear to be accompanied by large-scale changes in habitat structure – both Kedung Brubus and

213 Ngandong are likely to represent closed habitats. Furthermore, there is no evidence of

human mediated extinction between these two faunas, nor is it suggested here. Three possibilities remain. In the first scenario, the faunas of Kedung Brubus and Ngandong were deposited during periods of high sea-levels (i.e. interglacial), explaining their closed habitats. An intermediate glacial, unrecorded in any faunas, would have enabled the migration of species present in Ngandong through a possible savannah corridor, as well as precipitating the extinctions recorded in Kedung Brubus. The second possibility is similar, but would not require an intermediate glacial, suggesting the faunal migration occurring not through a savanna corridor, but rather through closed habitats. The third possibility would posit that all species recorded in Kedung Brubus became extinct, but the new fauna recolonising the region (represented by Ngandong) shared some species in common with the older site.

Ngandong preserves seven species of extinct megafauna. All of these represent global extinctions, and none is currently shared further than Java. This site preserves some of the last extinct megafauna present since Trinil. All extinctions from this site represent the last record of these megafauna for Java. The same three possibilities suggested for the faunal turnover between Kedung Brubus and Ngandong could account for the faunal turnover recorded between Ngandong and Punung, if the interpretation that they both represent closed habitats is accepted.

Punung records six species of now extinct megafauna. All of these extinctions represent local extinctions only. Three extinctions are shared by more than one country in the region: the orangutan (also becoming extinct in Southern China, Laos, Cambodia,

Vietnam and Thailand), the bearded pig (also becoming extinct in Thailand) and the

214 serow (also becoming extinct in Southern China and Vietnam). Due to the rainforest

ecology of many of the extinct taxa, Storm et al. (2005) suggested that these extinctions were the result of changes to more open habitats during the Holocene. This is certainly supported by the palaeoenvironmental reconstructions, which suggest a closed habitat for

Punung, but a mixed habitat for the Holocene caves. Evidence of human interaction with

megafauna is again limited, but, a case could certainly be made for human mediated

habitat alterations in the Holocene (Chapter 3). Despite Java’s excellent fossil record

relative to other Southeast Asian countries, it is still difficult to confidently associate

ecological changes with extinctions.

7.2.5 Sundaic summary

At the beginning of the Early Pleistocene the Sundaic province is likely to have been largely made up of closed habitats. Evidence suggests that these would give way to more open habitats by the time Trinil was deposited. The faunal community present at

Trinil would likely have migrated during periods of lower sea-level, and hence is indicative of glacial conditions. Following the deposition of Trinil, however, it seems that in Java and Sumatra closed habitat conditions prevailed until the Holocene. If this were the case, then extinctions occurring on these islands could be related to the influx of new species during periods of connection with the mainland. The Late Pleistocene fossil site of Niah Cave, Borneo, is likely to represent a mixed habitat. Although Niah preserves some of the only direct evidence of human interaction with megafauna for Southeast

Asia, these interactions do not appear to have impacted on the survival of species.

Changes from a mixed habitat to closed or vice versa can feasibly be associated with at least some extinctions. A possible mechanism for these changes in habitat, and

215 hence extinctions is proposed as a result of this study. Fluctuations in sea-level, driven by

oscillations between glacial and interglacial conditions, are expected to have been at least

partly responsible for the appearance and disappearance of savannahs and woodlands

throughout the Pleistocene. However, a direct association between glacial/interglacial

cycles and extinctions cannot currently be reconstructed on the basis of the available

evidence, due in large part to the limited temporal framework present in Southeast Asia.

There is no evidence that the super-eruption of Toba had any long-term effects on

megafauna in the Sundaic province.

7.3 Conclusions

Pleistocene sites in Southeast Asia are reconstructed as representing a range of habitats currently found largely outside Southeast Asia, from open (e.g. savannah), mixed

(e.g. woodland) and closed (e.g. rainforest). The changes in palaeoenvironment reconstructed by this study are likely to have adversely affected the megafauna through

loss of suitable habitats. The limited evidence of human interaction with megafauna in

Southeast Asia suggests that humans were not a significant threat to extinction until the

Holocene.

This study proposes that megafauna extinctions in Southeast Asia were driven by

changes in habitat, in turn precipitated by fluctuations in sea level, which in themselves

were the result of alternating glacial/interglacial cycles. However, the existence of

savannah corridors during glacial periods prior to the LGM, although suggested by this

study, could not be demonstrated. This is largely due to a dearth of exposed sites in the

geographical vicinity of any proposed savannah corridors, as well as a generally poor

chronology for sites in the region. Direct dating of more Southeast Asian fossil sites will

216 enable more accurate tracking of habitat changes through time, and may resolve whether specific habitats can be tied to glacial or interglacial periods. Recovery, identification and documentation of more species from most fossil sites in Southeast Asia would also enable greater confidence in palaeohabitat reconstructions, as many of the sites studied record insufficient numbers of megafauna to confidently differentiate between closed and mixed habitats. Nevertheless, on the basis of the evidence examined in this thesis, environmental change remains the most likely cause of megafauna extinctions in

Southeast Asia until the Holocene.

217 CHAPTER 8

CONCLUSIONS

8.1 Major implications of this research

1. The Quaternary in Southeast Asia saw the extinction of a number of species of

megafauna. In some cases, these extinctions do not represent global extinctions,

but rather geographical range reductions. Species experiencing extinctions or

range reductions across more than one country in Southeast Asia include, among

others, proboscideans (Stegodon and Palaeloxodon), hippopotamuses

(Hexaprotodon), the orangutan (Pongo), hyenas (Pachyrocuta and Crocuta), the

giant panda (Ailuropoda), tapirs (Tapirus and Megatapirus), rhinoceroses

(Rhinoceros), and the giant Asian ape Gigantopithecus.

2. There is no evidence to suggest that humans contributed in any significant way to

extinctions of megafauna in Southeast Asia until the Holocene. Extinctions

associated with climate change are more likely for most of the Pleistocene, and

were probably driven at least in part by changes in sea-level. That climate change

can reasonably be associated with megafauna extinction is of direct relevance to

questions of megafauna extinctions on other continents.

3. The Toba super-eruption cannot be convincingly associated with any megafauna

extinction. The hypothesis that humans were globally reduced to a few thousand

individuals as a result of the climatic aftermath of the eruption is unlikely given

Toba’s limited affect on megafauna

218 4. Extinct megafauna share certain ecological traits. Many were grazers, and some

were intimately tied to freshwater sources. Palaeoecological reconstructions of

Pleistocene sites show that megafauna occupied a variety of habitats during the

Quaternary, including open (e.g. savannah), mixed (e.g. woodland) and closed

(e.g. rainforest) habitats. These reconstructions represent the most comprehensive

analyses of environmental change across the Pleistocene of Southeast Asia. As

such, they are of direct value to studies examining the biogeography of the region,

as well as hominin evolution and migrations.

8.2 Concluding remarks

This thesis proposes that fluctuations between glacial and interglacial conditions produced changes in vegetative structure, which would in turn have adversely affected megafauna. However, it is difficult to establish whether Sundaland, in the glacials preceding the Last Glacial Maximum (LGM), witnessed a more open habitat structure as a result of changes in sea-level. The results of the palaeoecological analyses (Chapter 6) certainly suggest more open habitats were present during the Middle and Late Pleistocene in the Indochinese province. Nevertheless, tying the changes associated with fluctuating glacial/interglacial cycles and vegetation structure is tenuous at present. The fact is that any potential sites that would lie in the proposed LGM savannah corridor during glacials would currently be submerged. The only definitive exception, that of Kinta Valley on the

Malaysian Peninsula, is not directly dated. It is therefore impossible to determine whether this site was deposited during glacial or interglacial conditions, and hence test whether previous glacials were associated with more open environments.

219 The taxa examined in this thesis do, however, appear to share certain ecological

traits. In particular, the loss of more open environments (suggested by the ecological

reconstructions, independent of any arguments of savannah corridors), suggests that the

loss of suitable habitat detrimentally affected certain Southeast Asian taxa. Correlating

changes in hydrology, associated with fluctuations in sea-level, and extinctions of certain megafauna must be more circumspect, particularly for periods before the LGM.

However, a potential role for humans, in particular for Early and Middle Pleistocene extinctions, is not currently supported by any available evidence. Current records suggest little to no impact on megafauna by the activities of humans until the end of the

Pleistocene and/or the beginning of the Holocene. While the limited data from Southeast

Asia makes it premature to rule out humans as a primary cause, the results of this thesis suggests environmental change a more likely cause.

This thesis has for the first time enabled reconstructions of many Pleistocene

Southeast Asian fossil sites. Nevertheless, the biogeographical history of the region

remains elusive. The discovery of new sites, and the re-examination of previously

excavated material may help resolve some of the issues highlighted in this thesis. One of

the most important areas of future research in Southeast Asia must be in establishing a

detailed chronology of Pleistocene sites. Without a detailed chronology for the region,

any new palaeontological findings can only be put in limited context. Fortunately, it

appears this area is the subject of renewed research (e.g. Westaway et al., 2007). The

detailed study of certain extinct taxa (e.g. Stegodon, Megatapirus), which at the moment

remains very limited, will help resolve the issues of taxonomy, chronology and ecology

also highlighted through this research.

220 REFERENCES

Alemseged, Z., 2003. An integrated approach to taphonomy and faunal change in the

Shungura Formation (Ethiopia) and its implication for hominid evolution. Journal

of Human Evolution 44, 451-478.

Ambrose, S. H., 1998. Late Pleistocene human population bottlenecks, volcanic winter,

and differentiation of modern humans. Journal of Human Evolution 34, 623-651.

Ambrose, S. H., 2003. Did the super-eruption if Toba cause a human population

bottleneck? Reply to Gathorne-Hardy and Harcourt-Smith. Journal of Human

Evolution 45, 231-237.

An, Z. S., Ho., C. K., 1989. New magnetostratigraphic dates of Lantian Homo erectus.

Quaternary Research 32, 213-221.

An, Z. S., Gao, W. Y., Zhu, Y. Z., 1990. Magnetostratigraphic dates of Lantian Homo

erectus. Acta Anthropologica Sinica 1, 1-7 (in Chinese with English summary).

Andersson, K., Kaakinen, A., 2004. Floodplain processes in the shaping of fossil bone

assemblages: an example from the Late Miocene, Bahe Formation, Lantian,

China. GFF 126, 279-287.

Andrews, P., Lord, X., Evans, E.M.N., 1979. Patterns of ecological diversity in fossil and

modern mammalian faunas. Biological Journal of the Linnean Society 11, 177-

205.

Andrews, P., Humphrey, L., 1999. African Miocene environments and the transition to

early hominines. In: Bromage, T.G., Schrenk, F. (Eds.), African Biogeography,

Climate Change, and Human Evolution. Oxford University Press, Oxford, pp.

282-300.

221 Anshari, G., Kershaw, A. P., van der Kaars, S., 2001. A Late Pleistocene and Holocene

pollen and charcoal record from peat swamp forest, Lake Sentarum Wildlife

Reserve, West Kalimantan, Indonesia. Palaeogeography, Palaeoclimatology,

Palaeoecology 171, 213-228.

Antón, S. C. & Swisher, C. C. III., 2004. Early dispersals of Homo form Africa. Annual

Review Anthropology 33, 271-296.

Arambourg, C., Fromaget, J., 1938. Le gisement quaternaire de Tam Nang (Chaîne

Annamitique septentrionale). Sa stratigraphie et ses faunes. Comptes Rendus de

l'Academie des Sciences 207, 793-795.

Bacon, A.-M., Demeter, F., Schuster, M., Long, V. T., Thuy, N. K., Antoine, P.-O., Sen,

S., Nga, H. H., Huong, N. M., 2004. The Pleistocene Ma U’Oi cave,

northernVietnam: palaeontology, sedimentology and palaeoenvironments. Geobios

37, 305-314.

Barker, G., 2002. Prehistoric foragers and farmers in South-East Asia: renewed

investigations at Niah Cave, Sarawak. Proceedings of the Prehistoric Society 68,

147-164.

Barker, G., Barton, H., Bird, M., Daly, P., Datan, I., Dykes, A., Farr, L., Gilbertson, D.,

Harrison, B., Hunt, C., Higham, T., Kealhofer, L., Krigbaum, J., Lewis, H.,

McLaren, S., Paz, V., Pike, A., Piper, P., Pyatt, B., Rabett, R., Reynolds, T., Rose,

J., Rushworth, G., Stephens, M., Stringer, C., Thompson, J., Turney, C., 2007.

The ‘human revolution’ in lowland tropical Southeast Asia: the antiquity and

behaviour of anatomically modern humans at Niah Cave (Sarawak, Borneo).

Journal of Human Evolution 52, 243-261.

222 Barnosky, A., Koch, P., Feranec, R., Wing, S., Shabel, A., 2004. Assessing the Causes of

Late Pleistocene Extinctions on the Continents. Science 306, 70-75.

Bassarova, M., 2005. Taphonomic and palaeoecological investigations of Riversleigh

Oligo-Miocene fossil sites. Ph.D. Dissertation, University of New South Wales.

Batchelor, B. C., 1979. Discontinuous rising Late Cainozoic eustatic sea-levels, with

special reference to Sundaland, southeast Asia. Geologie en Mijnbouw 58, 1-20.

Batchelor, D. A. F., 1988. Dating of Malaysian fluvial tin placers. Journal of Southeast

Asian Earth Sciences 2, 3-14.

Beden, M., Guérin, C., 1973. Le gisement de Vertébrés du Phnom Loang (province du

Kampot, Cambodge) – Faune pléistocène moyen terminal. Travaux et Documents

de l’O.R.S.T.O.M, pp 1-97.

Behrensmeyer, A. K., 1975. Taphonomy and paleoecology in the hominid fossil record.

Yearbook of Physical Anthropology 19, 36-50.

Bekki, S., Pyle, J. A., Zhong, W., Toumi, R., Haigh, J. D., Pyle, D. M., 1996. The role of

microphysical and chemical processes in prolonging the climate forcing of the

Toba eruption. Geophysical Research Letters 23, 2669-2672.

Binford, L. R., Ho, C. K., 1985. Taphonomy at a distance: Zhoukoudian, "the Cave

Home of Beijing Man?". Current Anthropology 26, 413-442.

Binford, L. R., Stone, N. M., 1986. Zhoukoudian: A closer look. Current Anthropology

27, 453-475.

Bird, M. I., Taylor, D., Hunt, C., 2005. Palaeoenvironments of insular Southeast Asia

during the Last Glacial Period: a savanna corridor in Sundaland? Quaternary

Science Reviews 24, 2228-2242.

223 Blumenschine, R. J., 1988. An experimental model of the timing of hominid and

carnivore influence on archaeological bone assemblages. Journal of

Archaeological Science 15, 483-502.

Blumenschine, R. J., Marean, C. W., 1993. A canivore’s view of archaeological bone

assemblages. In Hudson, J. (Ed.) From Bones to Behaviour. Carbondale: Center

for Archaeological Investigations, pp. 273-301.

Boaz, N.T., Ciochon, R.L., Xu, Q., Liu, J., 2000. Large mammalian carnivores as a

taphonomic factor in the bone accumulation at Zhoukoudian. Supplement to Vol.

19, 224-234.

Boaz, N.T., Ciochon, R.L., Xu, Q., Liu, J., 2004. Mapping and taphonomic analysis of

the Homo erectus loci at Locality 1 Zhoukoudian, China. Journal of Human

Evolution 46, 519-549.

Bowler, J.M., Johnston, H., Olley, J.M., Prescott, J.R., Roberts, R.G., Shawcross, W., &

Spooner,N.A. 2003. New ages for human occupation and climatic change at Lake

Mungo, Australia. Nature 421, 837-840.

Brain, C. K., 1981. The Hunters or the Hunted? University of Chicago Press, Chicago.

Brongersma, L. D., 1936. Some comments upon H. C. Raven’s paper: “Wallace’s line

and the distribution of Indo-Australian mammals.” Arch Néerl zool 2, 240-256.

Brook, B. W., Bradshaw, C. J. A., Koh, L. P., Sodhi, N. S., 2006. Momentum drives the

crash: mass extinction in the tropics. Biotropica 38, 302-305.

Brown, P., Sutikna, T., Morwood, M.J., Soejono, R.P., Saptomo, E.W., Due, R.A., 2004.

A new small-bodied hominin from the Late Pleistocene of Flores, Indonesia.

Nature 431, 1055-1061.

224 Bühring, C., Sarnthein, M., Leg 184 Shipboard Scientific Party, 2000. Toba ash layers in

the South China Sea: evidence of contrasting wind directions during eruption ca.

74kyr. Geology 28, 275-278.

Capaldo, S. D., 1997. Experimental determinations of carcass processing by Plio-

Pleistocene hominids and carnivores at FLK 22 (Zinjanthropus), Olduvai Gorge,

Tanzania. Journal of Human Evolution 33, 555-597.

Capaldo, S. D., 1998. Simulating the formation of dual-patterned archaeofaunal

assemblages with experimental control samples. Journal of Archaeological

Science 25, 311-330.

Carbonnel, J-P., Guth, C., 1968. Le gisement pléistocène infériuer du Phnom Loang

(Cambodge), stratigraphie, et faune. Comptes Rendus de l'Academie des Sciences

Série D 267, 2077-2080.

Cardillo, M., Mace, G.M., Jones, K.E., Bielby, J., Bininda-Emonds, O.R.P.,

Sechrest, W., Orme, C.D.L., Purvis, A., 2005. Multiple causes of high

extinction risk in large mammal species. Science 309, 1239–1241.

Cerling, T. E., Harris, J. M., Leakey, M. G., 1999. Browsing and grazing in elephants: the

isotope record of modern and fossil proboscideans. Oecologia 120, 364-374.

Chappell, J., Shackleton, N. J., 1986. Oxygen isotopes and sea level. Nature 324, 137-

140.

Chappell, J., Omura, A., Esat, T., McCulloch, M., Pandolfi, J., Ota, Y., Pillans, B., 1996.

Reconcilliation of late Quaternary sea levels derived from coral terraces at Huon

Peninsula with deep sea oxygen isotope records. Earth Planet Science Letters 141,

227-236.

225 Chen, T., Yang, Q. & Wu, E. 1994. Antiquity of Homo sapiens in China. Nature 368, 55-

56.

Chen, Q.Y., Wang, Y. & Liu, Z. 1998. U-series dating of stalagmite from Hulu Cave in

Nanjing, Jiangsu Province, China. Acta Anthropologica Sinica 17, 171-176.

Chesner, C. A., Rose, W. I., 1991. Stratigraphy of the Toba Tuffs and the evolution of the

Toba Caldera Complex, Sumatra, Indonesia. Bulletin of Volcanology 53, 343-

356.

Chesser, R. K., 2006. The legacy of Chornobyl: 1986 to 2006 and beyond. Briefing

presented at U.S. Capitol Building, April 27, 2006.

Chesser, R., Rodgers, B., Baker, R., 2005. Temporal trends in radiation doses, survival,

and recovery in wildlife populations at Chernobyl. Ecological Society of America

2005 Annual Meeting, Montreal, Canada (Abstract).

China Species Information Service (CSIS), 2007. http://www.chinadiversity.com.

Chow, B., 1978. The distribution of the woolly rhinoceros and woolly mammoth.

Vertebrata PalAsiatica 16, 47-59 (in Chinese with English summary).

Chow, M. C., 1964. Mammals of “Lantian Man” locality at Lantian, Shensi. Vertebrata.

PalAsiatica 2, 301-307 (in Chinese with English summary).

Chow, M. C., Li, C. K., 1965. Mammalian fossils in association with the mandible of

Lantian Man at Chenchiaou, in Lantian, Shensi. Vertebrata. PalAsiatica 9, 377-

393 (in Chinese with English summary).

Chuan, G. K., 2005. The climate of Southeast Asia. In: Gupta, A. (Ed) The physical

geography of Southeast Asia. Oxford University Press, Oxford pp. 80-93.

226 Ciochon, R.L., Piperno, D.R., Thompson, R.G., 1990. Opal phytoliths found on the teeth

of the extinct ape Gigantopithecus blacki: Implications for paleodietary studies.

Proceedings of the National Academy of Sciences 87, 8120-8124.

Ciochon, R.L., Long, V.T., Larick, R., González, L., Grün, R., de Vos, J., Yonge, C.,

Taylor, L., Yoshida, H., Reagan, M., 1996. Dated co-occurrence of Homo erectus

and Gigantopithecus from Tham Khuyen Cave, Vietnam. Proceedings of the

National Academy of Sciences 93, 3016-3020.

Clark, K. R., Warwick, R. M., 2001. A further biodiversity index applicable to species

lists: variation in taxonomic distinctness. Marine Ecology Progress Series 216,

265-278.

Colbert, E., 1943. Pleistocene vertebrates collected in Burma by the American Southeast

Asiatic expedition. Transactions of the American Philosophical Society 32(3),

395-429.

Colbert, E. H., and Hooijer, D. A., 1953. Pleistocene mammals from the limestone

fissures of Szechwan, China. Bulletin of the American Museum of Natural

History 101, 1-135.

Corbet, G.B., Hill, J.E., 1992. The Mammals of the Indomalayan Region. Oxford

University Press, New York.

Corlett, R. T., 2005. Vegetation. In: Gupta, A. (Ed) The physical geography of Southeast

Asia. Oxford University Press, Oxford, pp. 105-119.

Corlett, R. T., 2007. The impact of hunting on the mammalian fauna of tropical Asian

forests. Biotropica 39: 292-303.

227 Corvinus, G., 2004. Homo erectus in East and Southeast Asia, and the questions of the

age of the species and its association with stone artifacts, with special attention to

hand-axe-like tools. Quaternary International 117, 141-151.

Cranbrook, E.o., 2000. Northern Borneo environments of the past 40,000 years:

archaeozoological evidence. The Sarawak Museum Journal 55, 61-109.

Cranbrook, E.o., Currant A.P., Davison G.W., 2000. Quaternary mammal fossils from

Borneo: Stegodon and Hippopotamus. Sarawak Museum Journal 55, 215-233.

Cruz-Uribe, K., 1991. Distinguishing hyena from hominid bone accumulations. Journal

of Field Archaeology 18, 467-486.

Curnoe, D. & Tobias, P.V. 2006. Description, new reconstruction, comparative anatomy,

and classification of the Sterkfontein Stw 53 cranium, with discussions about the

taxonomy of other southern African early Homo remains. Journal of Human

Evolution 50, 36-77.

Dai, E., 1966. The paleoliths found at Lantian Man locality of Gongwangling and its

vicinity. Vertebrata PalAsiatica10, 30-34.

Dai, E., Xu, C., 1973. New finds of Palaeoliths from Lantian. Acta Archaeologica Sinica

2, 1-12.

Del Moral, R., Grishin, S. Y., 1999. Volcanic disturbances and ecosystem recovery. In:

Walker, L. R. (Ed.) Ecosystems of Disturbed Ground. Elsevier Science B.V.,

Amsterdam, pp. 137-160.

Department of Ningxia Yunwushan Grassland Nature Reserve, 2000. A Collection Work

on Scientific Investigation and Management in Ningxia Yunwushan Nature

Reserve. Ningxia People Publishing House, Yinchuan, China.

228 de Vos, J., 1983. The Pongo faunas from Java and Sumatra and their significance for

biostratigraphical and paleo-ecological interpretations. Proc. Koninklijke

Nederlandse Akadademie van Wetenschappen 86, 417-425.

de Vos, J., Sondaar, P. Y., van den Bergh, G. D., Aziz, F., 1994. The Homo bearing

deposits of Java and its ecological context. Courier Forschung-Institut

Senckenberg 171, 129-140. de Vos, J., 1995. The migration of Homo erectus and Homo sapiens in south-east Asia

and the Indonesian Archipelago. In: Bower, JRF, Sartono, S. (Eds.), Human

evolution in its ecological context. Evolution and Ecology of Homo erectus, vol.

I. DSWO Press, Leiden University, Leiden, pp. 239-260.

de Vos, J., Long, V. T., 2001. First settlements: relations between continental and insular

Southeast Asia. In: Sémah F, Falguères C, Grimaud-Hervé D, Sémah A, (Eds)

Origine des peuplements et chronologie des cultures Paléolithiques dans le Sud-

est Asiatique. Semenanjung, Paris, pp. 225-249.

Directorate General of Forest Protection and Nature Conservation, 2003. Submission for

Nomination of Tropical Rainforest Heritage of Sumatra by the Government of the

Republic of Indonesia to be included in the World Heritage List. Ministry of

Forestry, Indonesia.

Domínguez-Rodrigo, M. & Pickering, T. R. 2003. Early hominid hunting and

scavenging: a zooarcheological review. Evolutionary Anthropology 12, 275-282.

Dubois, E., 1891. Palaeontologische onderzoekingen op Java.- Versl. mijnwez. 41, 14-18.

Dubois, E., 1908. Das geologische Alter der Kendeng- oder Trinil fauna. – T. Kon.

Nederl. Aardr. Gen. 25, 1235-1270.

229 Dudgeon, D., 2000. The ecology of tropical Asian rivers and streams in relation to

biodiversity conservation. Annual Review of Ecology and Systematics 30, 239-

263.

Escarguel, G., Legendre, S., 2006. New methods for analysing deep-time meta-

community dynamics and their application to the Paleogene mammals from the

Quercy and Limagne area (Massif Central, France). Strata 13, 245-273.

Edwards, J. S., 2005. Animals and volcanoes: survival and revival. In: Martí, J., Ernst, G.

G. J. (Eds.) Volcanoes and the Environment. Cambridge University Press,

Cambridge, pp. 250-272.

Eltringham, S. K., 1982. Elephants. Blandford Press, Dorset, U.K.

Erwin, D. H., Vogel, T. A., 1992. Testing for causal relationships between large

pyroclastic volcanic eruptions and mass extinctions. Geophysical Research

Letters 19, 893-896.

Esposito, M., Reyss, J-L., Chaimanee, Y., Jaeger, J-J., 2002. U-series dating of fossil

teeth and carbonates from Snake Cave, Thailand. Journal of Archaeological

Science 29, 341-349.

Etler, D. A., 1996. The fossil evidence for human evolution in Asia. Annual Review of

Anthropology 25, 275-301.

Farrelly, D., 1984. The book of bamboo. Sierra Club books.

Ferguson, D.K., 1993. The impact of Late Cenozoic environmental changes in East Asia

on the distribution of terrestrial plants and animals. In: Jablonski, N.G. (Ed.),

Evolving landscapes and evolving biotas of East Asia since the Mid-Tertiary.

Centre of Asian Studies, Hong Kong, pp. 145-196.

230 Field, J., Fullager, R., 2001. Archaeology and Australian megafauna. Science 294: 7a

Field Museum of Natural History, 2006. http://www.fmnh.org/research_collections/

zoology/zoo_sites/seamaps/ Accessed 25th May 2007.

Foley, R. 1987. Another unique species. U.K.: Longman Scientific and Technical.

Fromaget, J., 1936. Sur la stratigraphie des formations récentes de la Chaine annamitique

septentrionale et sur l'existence de l'Homme dans le Quaternaire inférieur de cette

partie de l'Indochine. Comptes Rendus de l'Academie des Sciences 203, 738-741.

Gagnon, M., 1997. Ecological diversity and community ecology in the Fayum sequence

(Egypt). Journal of Human Evolution 32, 133-160.

Galdikas, B.M.F., 1988. Orangutan diet, range, and activity at Tanjung Puting, Central

Borneo. International Journal of Primatology 9(1), 1-35.

Gathorne-Hardy, F. J., Jones, D. T., Mawdsley, N. A., 2000. The recolonization of the

Krakatau islands by termites (Isoptera), and their biogeographical origins.

Biological Journal of the Linnean Society 71, 379-393.

Gathorne-Hardy, F.J., Syaukani, Davies, R.G., Eggleton, P., Jones, D.T., 2002.

Quaternary rainforest refugia in south-east Asia: using termites (Isoptera) as

indicators. Biological Journal of the Linnean Society 75, 453-466.

Gathorne-Hardy, F. J., Harcourt-Smith, W. E. H., 2003. The super-eruption of Toba, did

it cause a human bottleneck? Journal of Human Evolution 45, 227-230.

Gatinsky, Y. G., Hutchinson, C. S., 1987. Cathaysia, Gondwanaland and the Palaeotethys

in the evolution of continental Southeast Asia. Geological Society of Malaysia

Bulletin 20, 179-199.

231 Gingele, F. X., de Deckker, P., Girault, A., Guichard, F., 2002. History of the South Java

Current over the past 80ka. Palaeogeography, Palaeoclimatology, Palaeoecology

183, 247-260.

Graham, R.W., Lundelius, E.L.J., Graham, M.A., Schroeder, E.K., Toomey, R.S.,

Anderson, E., Barnosky, A.D., Burns, J.A., Churcher, C.S., Grayson, D.K.,

Guthrie, R.D., Harington, C.R., Jefferson, G.T., Martin, L.D., McDonald, H.G.,

Morlan, R.E., Semken, H.A.J., Webb, S.D., Werdelin, L., Wilson, M.C., 1996.

Spatial response of mammals to late Quaternary environmental fluctuations.

Science 272, 1601-1606.

Graham, R. W., 2005. A self-organising, threshold model for the environmental cause of

terminal Pleistocene extinctions in North America. 10th Conference on

Australasian Vertebrate Evolution, Palaeontology & Systematics and Quaternary

Extinctions Symposium. Naracoorte, South Australia.

Graham, R. W., 2006. A self-organizing, threshold model for the environmental cause of

terminal Pleistocene extinctions in North America. Alcheringa Special Issue 1,

443-444.

Gray, D., Piprell, C., Graham, M., 1991. National Parks of Thailand. Industrial Finance

Corporation of Thailand, Bangkok, Thailand.

Grayson, D. K., 1984a. Explaining Pleistocene extinctions: thoughts on the structure of a

debate. In: Martin, P. S., Klein, R. G. (Eds.) Quaternary Extinctions: A prehistoric

revolution. University of Arizona Press, Tucson, pp 807-823.

232 Grayson, D. K., 1984b Nineteenth century explanations of Pleistocene extinctions: a

review and analysis. In: Martin, P. S., Klein, R. G. (Eds.) Quaternary Extinctions:

A prehistoric revolution. University of Arizona Press, Tucson, pp 5-39.

Groves, C. P., Kurt, F., 1972. Dicerorhinus sumatrensis. Mammalian Species 21: 1-6.

Groves, C., Guérin, C., 1980. Le Rhinoceros sondaicus annamiticus (Mammalia,

Perissodactyla) d’Indochine : distinction taxonomique et anatomique, relations

phylétique. Geobios 13, 199-208.

Grün, R., Thorne, A., 1997. Dating the Ngandong Humans. Science 277, 753.

Gupta, A., 2005. The physical geography of Southeast Asia. Oxford University Press,

Oxford.

Guthrie, R.D., 1984. Mosaics, Allelochemics, and Nutrients: An ecological theory of late

Pleistocene megafaunal extinctions. In: Martin, P. S., Klein, R. G. (Eds.)

Quaternary Extinctions: A prehistoric revolution. University of Arizona Press,

Tucson, pp.259-298.

Guthrie, R.D., 2003. Rapid body size decline in Alaskan Pleistocene horses before

extinction. Nature 426, 169-171.

Hall, R,, Holloway, J. D., 1998. Biogeography and geological evolution of SE Asia.

Backhuys Publishers, Leiden.

Hammer, Ø., Harper, D. A. T., Ryan, P. D., 2001. PAST: Palaeontological Statistics

software package for education and data analysis. Palaeontologia Electronica 4,

9pp.

233 Han, D., Xu, C., 1985. Pleistocene mammalian faunas of China. In: Wu, R., Olsen J,

(Eds). Palaeoanthropology and Palaeolithic Archaeology in the People's Republic

of China. Academic Press, Orlando, pp 267-289.

Harrison, T., 1996. The palaeoecological context at Niah Cave, Sarawak: evidence from

the primate fauna. Indo-Pacific Prehistory Association Bulletin 14, 90-100.

Harrison, T., Krigbaum, J., Manser, J., 2006. Primate biogeography and ecology on the

Sunda shelf islands: a paleontological and zooarchaeological perspective. In

Lehman, S. M., Fleagle, J. G. (Eds.) Primate biogeography: progress and

prospects. Springer Science, New York, pp. 331-372.

Harwell, M. A., 1984. The human and environmental consequences of nuclear war. New

York, Springer-Verlag, 179p.

Harwell, M. A., Hutchinson, T. C., 1985. Environmental consequences of nuclear war,

Vol. II: Ecological and agricultural effects. New York, Wiley.

Harwell, M. A., Hutchinson, T. C., Cropper, W. P. Jr., Harwell, C. C., 1985.

Vulnerability of ecological systems to climatic effects of nuclear war, in: Harwell,

M. A., Hutchinson, T. C. (Eds.). Environmental consequences of nuclear war,

Vol. II: Ecological and agricultural effects. Wiley, New York, pp. 81-171.

Hazebroek, H. P., Morshidi, A. K. A., 2002. A guide to Gunung Mulu National Park.

Natural History Publications (Borneo) Sdn. Bhd., Malaysia.

Heaney, L.R., 1991. A synopsis of climatic and vegetational change in Southeast Asia.

Climatic Change 19, 53-61.

Henneberg, M., Thorne, A., 2004. Flores human may be pathological Homo sapiens.

Before Farming 2004/4 article 1.

234 Hoffmann, R. S., 2001. The southern boundary of the Palaearctic realm in China and

adjacent countires. Acta Zoologica Sinica 47, 121-131.

Holden, J., Yanuar, A., Martyr, J. R., 2003. The Asian Tapir in Kerinci Seblat National

Park, Sumatra: evidence collected through photo-trapping. Oryx 37, 34-40.

Hooijer, D. A., 1950. The fossil Hippopotamidae of Asia, with notes on the recent

species. Zoologische Verhandelingen 8, 1-123.

Hope, G., Kershaw, A.P., van der Kaars, S., Xiangjun, S., Liew, P., Heusser, L.E.,

Takahara, H., McGlone, M., Miyoshi, N., Moss, P.T., 2004. History of vegetation

and habitat change in the Austral-Asian region. Quaternary International 118-119,

103-126.

Hope, G., 2005. The Quaternary in Southeast Asia. In: Gupta, A., (Ed) The physical

geography of Southeast Asia. Oxford University Press, Oxford, pp. 24-37.

Hsu, J., 1996. The climatic condition in north China during the time of Sinanthropus.

Science Sinica 15, 410-414 (in Chinese).

Hu, C. & Qi, T., 1978. Gongwangling Pleistocene mammalian fauna of Lantian, Shaanxi.

Pal. Sin. 155, 1-64 (in Chinese with English summary).

Huang, C., 1993. Animals of Longqi Mountain. China Forestry Publishing House,

Beijing, China.

Huang, W., Ciochon, R., Yumin, G., Larick, R., Qiren, F., Schwarcz, H., Yonge, C., de

Vos, J. & Rink, W. 1995. Early Homo and associated artefacts from Asia. Nature

378, 275-278.

Hutchinson, C. S., 1989. Geological evolution of South-East Asia. Oxford Monographs

on Geology and Geophysics, Clarendon Press, Oxford.

235 Hutchinson, C. S., 2005. The geological framework. In: Gupta, A., (Ed) The physical

geography of Southeast Asia. Oxford University Press, Oxford, pp. 3-23.

Hutchinson, T. C., Cropper, W. P. Jr., Grover, H. D., 1985. Ecological principles relevant

to nuclear war, in: Harwell, M. A., Hutchinson, T. C. (Eds.), 1985. Environmental

consequences of nuclear war, Vol. II: Ecological and agricultural effects. New

York, Wiley, p. 3-80.

Information Centre for the Environment (ICE), 2007. http://www.ice.ucdavis.edu/

bioinventory/bioinventory.html. Accessed 10th May 2007.

IUCN (2007) http://www.redlist.org. Accessed 10th May 2007.

Jablonski, N.G., Gu, Y., 1991. A reassessment of Megamacaca lantianensis, a large

monkey from the Pleistocene of north-central China. Journal of Human Evolution

20, 51-66.

Jablonski, N. G., Pan, Y., 1988. The evolution and palaeobiogeography of monkeys in

China. In Whyte, P., Aigner, J. S., Jablonski, N. G., Taylor, G., Walker, D.,

Wang, P. X. (Eds) The palaeoenvironment of East Asia from the mid-Tertiary,

Centre of Asian Studies, University of Hong Kong, pp 849-867.

Jablonski, N., Whitford, M., 1999. Environmental change during the Quaternary in East

Asia and its consequences for mammals. Records of the Western Australian

Museum. Supplement No. 57, 307-315.

Jablonski, N.G., Whitford M.J., Roberts-Smith N., Qinqi X., 2000. The influence of life

history and diet on the distribution of catarrhine primates during the Pleistocene in

eastern Asia. Journal of Human Evolution 39, 131-157.

236 Jablonski, N.G., 2004. The hippo's tale: how the anatomy and physiology of Late

Neogene Hexaprotodon shed light on Late Neogene environmental change.

Quaternary International 117, 119-123.

Ji, H. X., 1975. The lower Pleistocene mammalian fossils of Lantian District, Shensi.

Vert. PalAsiat. 13, 169-177 [in Chinese with English summary].

Ji, H. X., 1980. The classification of Quaternary mammalian fauna in Lantian area.

Paleovertebrata and Paleoanthropologica Sinicia 18, 220-228.

Jiang, P., 1990. The Late Pleistocene Mammuthus-Coelodonta fauna in Song-Nun Plain.

In: Environment evolution in the north-east China Grand Plain during Quaternary.

Harbin Map Publishing House, pp. 24-29 (in Chinese).

Johnson, A.E., 1979. Skeletal estimates of Gigantopithecus based on a Gorilla analogy.

Journal of Human Evolution 8(6), 585-587.

Jones, G. S., Gregory, J. M., Stott, P. A., Tett, S. F. B., Thorpe, R. B., 2005. An AOGCM

simulation of the climate response to a volcanic super-eruption. Climate

Dynamics 25, 725-738.

Jones, M. T., Sparks, R. S. J., Valdes, P. J., 2007. The climatic impact of supervolcanic

ash blankets. Climate Dynamics, doi: 10.1007/s00382-007-0248-7.

Kaakinen, A., Sonninen, E., Lunkka, J.P., 2006. Stable isotope record in paleosol

carbonates from the Chinese Loess Plateau: implications for late Neogene

paleoclimate and paleovegetation. Palaeogeography, Palaeoclimatology,

Palaeoecology 237, 359-369.

237 Kahlke, H. D., 1961. On the complex of the Stegodon-Ailuropoda fauna of Southern

China and the chronological position of Gigantopithecus blacki V.Koenigswald.

Vertebrata PalAsiatica 6, 83-108.

Kahlke, H.D., 1972. A review of the Pleistocene history of the Orang-Utan (Pongo

Lacépède 1799). Asian Perspectives 15, 5-14.

Kahlke, H.D., 1984. Paleo-environment of Pleistocene Gigantopithecus blacki of

continental Southeast Asia. International Journal of Primatology 5(4), 395.

Kappelman, J., 1988. Morphology and locomotor adaptations of the bovid femur in

relation to habitat. Journal of Morphology 198, 119-130.

Kappelman, J., 1991. The paleoenvironment of Kenyapithecus at Fort Ternan. Journal of

Human Evolution 20, 95-129.

Keates, S.G., 2003. Biostratigraphy, taphonomy, palaeoenvironment and hominid diet in

the Middle and Late Pleistocene of China. In: Shen, C., Keates, S., (Eds) Current

Research in Chinese Pleistocene Archaeology. Archaeopress, Oxford, pp 37-56.

Keates, S., 2004. Home range size in Middle Pleistocene China and human dispersal

patterns in Eastern and Central Asia. Asian Perspectives 43, 227-247.

Kershaw, P., van der Kaars, S., Moss, P., Opdyke, B., Guillard, F., Rule, S. & Turney,

C., 2006. Environmental change and the arrival of people in the Australian region.

Before Farming 2006/1 article 2.

Klein, R.G. 1999. The human career. University of Chicago Press, Chicago.

Kovarovic, K., Andrews, P., Aiello, L., 2003. The palaeoecology of the Upper Ndolanya

Beds at Laetoli, Tanzania. Journal of Human Evolution 43, 395-418.

238 Kovarovic, K., Andrews, P., 2007. Bovid postcranial ecomorphological survey of the

Laetoli paleoenvironment. Journal of Human Evolution 52, 663-680.

Kruuk, H., 1972. The Spotted Hyena: a study of predation and social behaviour. The

University of Chicago Press, Chicago.

Kurtén, B., 1956. The status and affinities of Hyaena sinensis Owen and Hyaena ultima

Matsumoto. American Museum Novitates 1764, 1-48.

Lambeck, K., Esat, T. M., Potter, E., 2002. Links between climate and sea levels for the

past three million years. Nature 419, 199-206.

Larick,R., Ciochon, R.L., Zaim, Y., Sudijono, Suminto, Rizal, Y., Aziz, F., Reagan, M. &

Heizler, M. 2001. Early Pleistocene 40Ar/39Ar ages for Bapang Formation

hominins, Central Jawa, Indonesia. Proceedings National Academy Sciences 98,

4866–4871.

Laurie, W. A., Lang, E. M, Groves, C. P., 1983. Rhinoceros unicornis. Mammalian

Species 211, 1-6.

Ledbetter, M., Sparks, R. S. J., 1979. Duration of large-magnitude explosive eruptions

deduced from graded bedding in deep-sea ash layers. Geology 7, 240-244.

Legendre, S., 1986. Analysis of mammalian communtities from the late Eocene and

Oligocene of southern France. Palaeovertebrata 16, 191-212.

Legendre, S., 1989. Les communautés de mammifères du Paléogène (Eocène et

Oligocène) d’Europe occidentale : stuctures, milieu et évolution. Münchner

Geowissenschaftliche Abhandlungen 16, 1-110.

Leinders, J. J., Aziz, M. F., Sondaar, P. Y., de Vos, J., 1985. The age of the hominid-

bearing deposits of Java: state of the art. Geologie en Mijnbouw 64, 167-173.

239 Lekagul, B., McNeely, J.A., 1988. Mammals of Thailand. Association for the

Conservation of Wildlife, Bangkok, Thailand.

Leonard, W.R. & Robertson, M.L. 1997. Comparative primate energetics and hominid

evolution. American Journal of Physical Anthropology 102, 265-281.

Li, Y., Wen B., 1986. Guanyindong – A lower Paleolithic site at Qianxi County, Guizhou

Province. Cultural Relics Publishing House, Beijing, pp. 1-181 (in Chinese with

English summary).

Liu, T. S. (Ed.), 1985. Loess and the environment. China Ocean Press, Beijing.

Livingstone, F. B., 1965. Controversy regarding Gigantopithecus. American

Anthropologist 67, 1283-1284.

Louys, J. & Bassarova, M., 2006. Community structure of rainforest mammals:

implications for fossil deposits. Riversleigh Symposium 20th -22nd November

2006.

Louys, J., Curnoe, D., Tong, H., 2007. Characteristics of Pleistocene megafauna

extinctions in Southeast Asia. Palaeogeography, Palaeoclimatology,

Palaeoecology 243, 152-173.

Louys, J., 2007. Limited effect of the Quaternary’s largest super-eruption (Toba) on land

mammals from Southeast Asia. Quaternary Science Reviews 26, 3108-3117.

Louys, J., 2008. Quaternary extinctions in Southeast Asia’s megafauna. In: Elewa, A. M.

T., (Ed.) Mass extinction - the danger around us. Springer-Verlag, Heidelberg,

Germany, pp. 159-189.

Lundelius, E.L.J., 1989. The implications of disharmonious assemblages for Pleistocene

extinctions. Journal of Archaeological Science 16, 407-407.

240 Lyman, R. L., 1994. Vertebrate Taphonomy. Cambridge University Press, Cambridge.

Ma, A., Tang, H., 1992. On discovery and significance of a Holocene Ailuropoda-

Stegodon fauna from Jinhua, Zhejinag. Vertebrata PalAsiatica 30, 295-312 (in

Chinese with English Abstract).

MacKinnon, J., 1971. The Orang-utan in Sabah Today. Oryx 11, 141-191.

MacPhee, R. D. E., Marx, P.A., 1997. The 40, 000-year plague: humans, hyperdisease,

and first-contact extinctions. In: Goodman, S. M., Patterson, B. D. (Eds.) Natural

Change and Human Impact in Madagascar. Smithsonian Institute Press, pp 169-

217.

Maguire, J., 1976. A taxonomic and ecological study of the living and fossil Hystricidae

with particular reference to southern Africa. Ph.D. Dissertation, Univeristy of

Witwatersrand, South Africa.

Marean, C. W., 1998. A critique of the evidence for scavenging by Neandertals and Early

Modern Humans: new data from Kobeh Cave (Zagros Mountains, Iran) and Die

Kelders Cave 1 Layer 10 (South Africa). Journal of Human Evolution 35, 111-

136.

Marean, C. W., Kim, S. Y., 1998. Mousterian faunal remains from Kobeh cave (Zagros

Mountains, Iran): behavioural implications for Neanderthals and early modern

humans. Current Anthropology 38, S79-S113.

Marean, C. W., Abe, Y., Frey, C. J., Randall, R. C., 2000. Zooarchaeological and

taphonomic analysis of the Die Kelders Cave 1 Layers 10 and 11 Middle Stone

Age larger mammal fauna. Journal of Human Evolution 38, 197-233.

241 Marshall, L.G., 1984. Who killed Cock Robin? An investigation of the extinction

controversy. In: Martin, P. S., Klein, R. G., (Eds). Quaternary extinctions. A

prehistoric revolution. The University of Arizona Press, Tucson, pp 785-806.

Martin, P. S., 1966. Africa and Pleistocene overkill. Nature 212, 339-342.

Martin, P.S., 1984. Prehistoric overkill: The global model. In: Martin, P. S., Klein, R. G.,

(Eds). Quaternary extinctions. A prehistoric revolution. The University of Arizona

Press, Tucson, pp 354-403.

Martinson, D. G., Pisias, N. G., Hays, J. D., Imbrie, J., Moore, T. C., Shackelton, N. J.

1987. Age dating and the orbital theory of the Ice Ages: development of a high-

resolution 0 to 300,000-year chronostratigraphy. Quaternary Research 27: 1-29.

McHenry, H.M., 1994. Behavioral ecological implications of early hominid body size.

Journal of Human Evolution 27, 77-88.

McKenna, M., Bell, S., 1997. Classification of Mammals above the species level.

Columbia University Press, New York.

Medway, L., 1960. The Malay Tapir in late Quaternary Borneo. Sarawak Museum

Journal 9, 356-360.

Medway, L., 1964. Post-Pleistocene changes in the mammalian fauna of Borneo.

Archaeological evidences from the Niah Caves. Studies in Speleology 1, 33-37.

Medway, L.,1972. The Quaternary era in Malesia. In: Ashton PS, Ashton M, editors.

Miscellaneous series; Aberdeen, Scotland. University of Hull; University of

Aberdeen. p 63-83.

Medway, L., 1977. The Niah excavations and an assessment of the impacts of early man

on mammals in Borneo. Asian Perspectives 20(1), 51-69.

242 Meijaard, E., 2004. Biogeographic history of the Javan leopard Panthera pardus based on

a craniometric analysis. Journal of Mammalogy 85(2), 302-310.

Meijaard, E., 2004. Solving mammalian riddles. A reconstruction of the Tertiary and

Quaternary distribution of mammals and their palaeoenvironments in island

South-East Asia. PhD dissertation, School of Archaeology and Anthropology, The

Australian National University.

Mendoza, M., Janis, C. M., Palmqvist, P., 2005. Ecological patterns in the trophic-size

structure of large mammal communities: a ‘taxon-free’ characterization.

Evolutionary Ecology Research 7, 505–530.

Metcalf, I., 1990. Allochthonous terrane processes in Southeast Asia. London: Royal

Society Philosophical Transaction A331, 625-640.

Metcalf, I., 1996. Pre-Cretaceous evolution of SE Asian terranes. In Hall R, Blundell D

(eds.) Tectonic evolution of Southeast Asia. Geological Society Special

Publication 106, London, pp 97-122.

Miller, G.H., Magee, J.W., Johnson, B.J., Fogel, M.L., Spooner, N.A., McCulloch, M.T.,

Ayliffe, L.K., 1999. Pleistocene extinction of Genyornis newtoni: Human impact

on Australian Megafauna. Science 283, 205-208.

Miller, G.H., Fogel, M.L., Magee, J.W., Gagan, M.K., Clarke, S.J., Johnson, B.J., 2005.

Ecosystem collapse in Pleistocene Australia and a human role in megafauna

extinction. Science 309, 287-290.

Mollengraaff, G. A. F., 1921. Modern deep-sea research in the East Indian archipelago.

Geographical Journal 57, 95-121.

243 Morwood, M.J., O'Sullivan, P.B., Aziz, F. & Raza, A. 1998. Fission-track ages of stone

tools and fossils on the east Indonesian island of Flores. Nature 392, 173-176.

Morwood, M.J., Soejono, R.P., Roberts, R.G., Sutikna, T., Turney, C.S.M., Westaway,

K.E., Rink, W.J., Zhao, J.-x., van den Bergh, G.D., Due, R.A., Hobbs, D.R.,

Moore, M.W., Bird, M.I., Fifield, L.K., 2004. Archaeology and age of a new

hominin from Flores in eastern Indonesia. Nature 431, 1087-1091.

Morwood, M.J., Brown, P. Sutikna, J.T., Wahyu Saptomo, E., Westaway, K.E., Awe

Due, R., Roberts, R.G., Maeda, T., Wasisto, S. & Djubiantono, T. 2005. Further

evidence for small-bodied hominins from the Late Pleistocene of Flores,

Indonesia. Nature 437, 1012-1017.

Mosimann, J. E., Martin, P. S., 1975. Simulating overkill by Paleoindians. American

Scientist 63, 303-313.

Müller, S., 1839-1844. Over de zoogdieren van den Indischen Archipel. Verhandelingen

van de Natuurlijke Geschiedenis der Nederlandsche Overzeesche Bezittingen.

National Environment Protection Agency and Environmental Protection Bureau of

Xinjiang Uygur Autonomous Region, 1988. The Altun Mountain National

Reserve. China Environmental Science Press, Beijing, China.

News Release Office of the Tibetan Local Government, 2003. Nature Reserves in Tibet,

China. Continental Media Press, China.

Norton, C. J., Hasegawa, Y., Kohno, N., Tomida, Y., 2007. Distinguishing archaeological

and paleontological faunal collections from Pleistocene Japan: taphonomic

perspectives from Hanaizumi. Anthopological Science 115, 91-106.

244 Novarino, W., Kamilah, S. N., Nugroho, A., Janra, M. N., Silmi, M., Syafri, M., 2005.

Habitat use and density of the Malayan Tapir (Tapirus indicus) in the Taratak

Forest Reserve, Sumatra, Indonesia. Tapir Conservation 14/2, 28-30.

Nowak, R.M., 1999. Walker's mammals of the world. The John Hopkins University

Press, London.

Olsen, J. W., Ciochon, R. L., 1990. A review of evidence for postulated Middle

Pleistocene occupations in Viet Nam. Journal of Human Evolution 19, 761-788.

Oppenheimer, C., 2002. Limited global change due to the largest known Quaternary

eruption, Toba ≈ 74kyr BP? Quaternary Science Reviews 21, 1593-1609.

Owen-Smith, N., 1987. Pleistocene extinctions: the pivotal role of megaherbivores.

Paleobiology 13, 351-362.

Owen-Smith, R. N., 1992. Megaherbivores: The influence of very large body size on

ecology. Cambridge University Press, Cambridge.

Pan, Y., Jablonski, N. G., 1987. The age and geographical distribution of fossil

cercopithecids in China. Human Evolution 2, 59-69.

Parr, J. W. K., 2003. Large mammals of Thailand. Sarakadee Press, Bangkok.

Pei, W. C.,1957. The zoogeographical divisions of Quaternary mammalian faunas in

China. Vertebrate PalAsiatica 1, 9-24.

Pei, W. C., 1958. The zoogeographical and biostratigraphical divisions of Quaternary

mammalian faunas in China. Quaternary Sinica 1, 23-29 (in Chinese).

Platt, W. J., Connell, J. H., 2003. Natural disturbances and directional replacement of

species. Ecological Monographs 73, 507-522.

245 Plummer, T. W., Bishop, L. C., 1994. Hominid palaeoecology at Olduvai Gorge,

Tanzania as indicated by antelope remains. Journal of Human Evolution 27, 47-

75.

Pope, G. G., Frayer, D. W., Liangchareon, M., Kulasing, P., Nakabanlang, S., 1981.

Palaeoanthropological investigations of the Thai-American expeditions in

Northern Thailand (1978-1980): an interim report. Asian Perspectives 21, 147-

163.

Pope, G. G., 1989. Bamboo and human evolution. Natural History 10, 49-56.

Pope, G. G., 1995. The influence of climate and geography on the biocultural evolution

of the Far Eastern Hominids. In: Vrba, E. S., Denton, G. H., Partridge, T. C.,

Burckle, L. H., (Eds). Paleoclimate and evolution, with emphasis on human

origins. Yale University Press, New Haven, pp 493-506.

Price, G., Sobbe, I.H., 2005. Pleistocene palaeoecology and environmental change on the

Darling Downs, Southeastern Queensland, Australia. Memoirs of the Queensland

Museum 51(1), 171-201.

Price, G. J, Webb, G. E., 2006. Late Pleistocene sedimentology, taphonomy and

megafauna extinction on the Darling Downs, southeastern Queensland. Australian

Journal of Earth Sciences 53, 947-970.

Primack, R., Corlett, R., 2005. Tropical Rain Forests: an ecological and biogeographical

comparison. Blackwell Publishing, Malden, USA.

Prothero, D. R., Guérin, C., and Manning, E., 1989. The history of the Rhinocerotoidea.

In: Prothero D.R. and Schoch R.M. (Eds). The evolution of the Perissodactyls.

Oxford University Press, New York.

246 Prothero, D. R., 2004. Did impacts, volcanic eruptions, or climate change affect

mammalian evolution. Palaeogeography, Palaeoclimatology, Palaeoecology 214,

283-294.

Qiu, Z. X., 2002. Hesperotherium – A new genus of the last Chalicotheres. Vert.

PalAsiat. 40, 317-325 (in Chinese with English summary).

Qiu, Z. X., Deng, T., Wang, B. Y., 2004. Early Pleistocene mammalian fauna from

Longdan, Dongxiang, Gansu, China. Palaeont. Sin. New Ser. C. 27, 1-198 (in

Chinese with English summary).

Qiu, Z. X., 2006. Quaternary environmental changes and evolution of large mammals in

North China. Vertebrata. PalAsiatica 44, 109-132.

Rampino, M. R., Self, S., 1992. Volcanic winter and accelerated glaciation following the

Toba super-eruption. Nature, 359, 50-52.

Rampino, M. R., Self, S., 1993. Bottleneck in human evolution and the Toba eruption.

Science 262, 1955.

Rampino, M. R., Ambrose, S. H., 2000. Volcanic winter in the Garden of Eden: the Toba

super-eruption and the Late Pleistocene human population crash. In: McCoy, F.

W., Heiken, G. (Eds.), Volcanic Hazards and Disasters in Human Antiquity.

Geological Society of America Special Paper 345, 71-82.

Reed, K. E., 1997. Early hominid evolution and ecological change through the African

Plio-Pleistocene. Journal of Human Evolution 32, 289-322.

Reed, K. E., 1998. Using large mammal communities to examine ecological and

taxonomic structure and predict vegetation in extant and extinct assemblages.

Paleobiology 24, 384-408.

247 Reed, L., Bourne, S., Megirian, D., Prideaux, G., Young, G., Wright, A., 2006.

Proceedings of CAVEPS 2005. Alcheringa Special Issue 1: 475 pp

Reid, D. G., Hu, J., Dong, S., Wang, W., Huang, Y., 1989. Giant panda, Ailuropoda

melanoleuca, behaviour and carrying capacity following a bamboo die-off.

Biological Conservation 49, 85-104.

Reynolds, T. E. G., 1990. The Hoabinhian: a review. In: Barnes, G. L., (Ed.)

Bibliographical reviews of Far Eastern archaeology. Oxbow Books, Oxford, pp.

1-30.

Rijksen, H. D., 1978. A Field Study on Sumatran Orang Utans (Pongo pygmaeus abelli

Lesson 1827): Ecology, behaviour and conservation. H. Veenman & Zonen B. V.,

Wageningen.

Rink, W. J., Schepartz, L. A., Miller-Antonio, S., Huang, W., Hou, Y., Bakken, D.,

Richter, D., Jones, H. L. 2003. Electron spin resonance (ESR) dating of

mammalian tooth enamel at Panxian Dadong Cave, Guizhou, China. In: Chen S.

& Keates S. G. (Eds.), Current Research in Chinese Pleistocene Archaeology.

Oxford: BAR Monograph Series, pp. 111-118.

Roberts, R. G., Flannery, T., Ayliffe, L. K., Yoshida, H., Olley, J. M., Prideaux, G. J.,

Laslett, G. M., Baynes, A., Smith, M. A., Jones, R., Smith, B. L., 2001. New ages

for the last Australian megafauna: continent-wide extinction about 46,000 years

ago. Science 292, 1888-1892.

Rose, W. I., Chesner, C. A., 1987. Dispersal of ash in the great Toba eruption, 75kyr.

Geology 15, 913-917.

248 Rose, W. I., Chesner, C. A., 1990. Worldwide dispersal of ash and gases from earth’s

largest known eruption: Toba, Sumatra, 75kyr. Palaeogeography,

Palaeoclimatology, Palaeoecology 89, 269-275.

Saegusa, H., 2001. Comparisons of stegodon and elephantid abundances in the Late

Pleistocene of southern China. In: Cavaretta, G., Gioia, P., Mussi, M., Palombo,

M. R. (Eds.), The World of Elephants, Consiglio Nazionale delle Ricerche, Rome,

pp. 345 - 349.

Sathiamurthy, E., Voris, H. K., 2006. Maps of Holocene Sea Level Transgression and

Submerged Lakes on the Sunda Shelf. The Natural History Journal of

Chulalongkorn University, Supplement 2, 1-43.

Sawney, P., 2003. People-Park interaction: A case of Bandhavgarh National Park.

Ecology and Development Series No. 5, 1-144.

Scaillet, B., Clemente, B., Evans, B. W., Pichavant, M., 1998. Redox control of sulfur

degassing in silicic magmas. Journal of Geophysical Research 103, 23937-23949.

Schepartz, L.A., Miller-Antonio, S., Bakken, D.A., 2000. Upland resources and the early

Palaeolithic occupation of Southern China, Vietnam, Laos, Thailand and Burma.

World Archaeology 32(1), 1-13.

Schepartz, L.A., Stoutamire, S., Bekken, D.A., 2005. Stegadon orientalis from Panxian

Dadong, a Middle Pleistocene archaeological site in Guizhou, South China:

taphonomy, population structure and evidence for human interaction. Quaternary

International 126-28, 271-282.

Schoener, T. W., Spiller, D. A., Losos, J. B., 2001. Predators increase the risk of

catastrophic extinction of prey populations. Nature 412, 183-186.

249 Selvaggio, M. M., 1998. Evidence for a three-stage sequence of hominid and carnivore

involvement with long bones at FLK Zinjanthropus, Olduvai Gorge, Tanzania.

Journal of Archaeological Science 25, 191-202.

Sémah, A-M., 1984. Palynology and Javanese Pithecanthropus palaeoenvironment.

Courier Forschungsinstitut Senckenberg 69, 237-243.

Sémah, F., Saleki, H. & Falgueres, C. 2000. Did early man reach Java during the late

Pliocene? Journal Archaeological Science 27, 763-769.

Shackelford, L.L., 2003. Late Pleistocene postcranial skeletal remains from Tam Hang

(Laos). American Journal of Physical Anthropology 36 (S), 189.

Shackelton, N. J., Opdyke, N. D. 1976. Oxygen-isotope and paleomagnetic stratigraphy

of Pacific core V28-239 late Pliocene to latest Pleistocene. Geological Society of

America Memoir 145, 449-464.

Shang, H., Tong, H., Zhang, S., Chen, F., Trinkhaus, E., 2007. An early modern human

from Tianyuan Cave, Zhoukoudian, China. Proceedings of the National Academy

of Sciences 104, 6573-6578.

Shen, G., Liu, J. & Jin, L. 1997. Preliminary results on U-series dating of Panxian

Dadong in Guizhou Province, S-W China. Acta Anthropologica Sinica 16, 221-

230.

Shen, G., Wang, W., Wang, Q., Zhao, J., Collerson, K., Zhou, C., Tobias, P. V., 2002. U-

series dating of Liujiang hominid site in Guangxi, Southern China. Journal of

Human Evolution 43, 817-829.

Sheng, H., Ohtaishi, N., Lu, H., 1999. The mammalian of China. China Forestry

Publishing House, Beijing.

250 Shi, G. R., 1993. Multivariate data analysis in palaeoecology and palaeobiogeography – a

review. Palaeogeography, Palaeoclimatology, Palaeoecology 105, 199-234.

Shoshani, J., Tassy, P., 2005. Advances in proboscidean taxonomy & classification,

anatomy & physiology, and ecology & behavior. Quaternary International 126-

128, 5-20.

Simkin, T., 1983. Krakatau, 1883 – the volcanic eruption and its effects. Smithsonian

Institution, USA.

Smith, F. A., Lyons, S. K., Ernest, S. K. M., Jones, K. E., Kauffman, D. M., Dayan, T.,

Marquet, P. A., Brown, J. H., Haskell,. J. P., 2003. Body mass of late Quaternary

mammals. Ecology 84, 3403.

Smith, F. A., Brown, J. H., Haskell, J. P., Lyons, S. K., Alroy, J., Charnov, E. L., Dayan,

T., Enquist, B. J., Ernest, S. K. M., Hadly, E. A., Jones, K. E., Kaufman, D. M.,

Marquet, P. A., Maurer, B. A., Niklas, K. J., Porter, W. P., Tiffney, B., Willig, M.

R., 2004. Similarity of mammalian body size across the taxonomic hierarchy and

across space and time. American Naturalist 163(5), 672-691.

Sodhi, N. S., Koh, L. P., Brook, B. W., Ng, P. K. L., 2004. Southeast Asian biodiversity:

an impending disaster. Trends in Ecology and Evolution 19: 654-660.

Soligo, C., Andrews, P., 2005. Taphonomic bias, taxonomic bias and historical non-

equivalence of faunal structure in early hominid localities. Journal of Human

Evolution 49, 206-229.

Sondaar, P. Y., 1984. Faunal evolution and the mammalian biostratigraphy of Java. In:

Andrews, P., Franzen, J.L. (Eds.), The Early Evolution of Man with Special

251 Emphasis on Southeast Asia and Africa. Courier Forschungs-Institut Senckenberg

69, 219-235.

Sondaar, P.Y., 1987. Pleistocene man and extinctions of island endemics. Mémoires de la

Société Géologique de France NS 150, 159-165.

Song, C. 1994. Scientific survey of the Funiushan Mountain National Reserve. China

Forestry Publishing House, Beijing, China.

Song, C., Jia, K., 2000. Scientific Survey of the Wulate Haloxylon ammodendron Nature

Reserve. China Forestry Publishing House, Beijing, China.

Steinmetz, R., Chutipong, W., Seuaturien, N., 2006. Collaborating to conserve large

mammals in Southeast Asia. Conservation Biology 20, 1391-1401.

Steiper, M. E., 2006. Population history, biogeography, and taxonomy of oranutans

(Genus: Pongo) based on a population genetic meta-analysis of multiple loci.

Journal of Human Evolution 50, 509-522.

Stiner, M. C., 1991. Food procurement and transport by human and non-human predators.

Journal of Archaeological Science 18, 455-482.

Storm, P., 2001. The evolution of humans in Australasia from an environmental

perspective. Palaeogeography, Palaeoclimatology, Palaeoecology 171, 363-383.

Storm, P., Aziz, F., de Vos, J., Kosasih, D., Baskoro, S., Ngaliman, van den Hoek

Ostende, L. W., 2005. Late Pleistocene Homo sapiens in a tropical rainforest

fauna in East Java. Journal of Human Evolution 49, 536-545.

Swisher, C.C. III, Rink, W.J., Antón, S.C., Schwarcz, H.P., Curtis, G.H., Suprijo, A.,

Widiasmoro. 1996. Latest Homo erectus of Java: potential contemporaneity with

Homo sapiens in Southeast Asia. Science 274, 1870-1874.

252 Theunissen, B., de Vos, J., Sondaar, P. Y., Aziz, F., 1990. The establishment of a

chronological framework for the hominid-bearing deposits of Java; a historical

survey. Geological Survey of America, Special Paper 242, 39-54.

Thorne, A. Grün, R. Mortimer, G. Spooner, N.A. McCulloch, M. Taylor, L. & Curnoe,

D., 1999. Australia’s Oldest Human Remains: Age of the Lake Mungo 3

Skeleton. Journal of Human Evolution 36, 591-612.

Thorne, A., Curnoe, D., 2000. The Sex and Significance of Lake Mungo 3: Reply to

Brown “Australian Pleistocene Variation and the Sex of Lake Mungo 3”. Journal

of Human Evolution 39, 587-600.

Timmreck, C., Graf, H.-F., 2006. The initial dispersal and radiative forcing of a Northern

Hemisphere mid-latitude super volcano: a model study. Atmospheric Chemistry

and Physics 6, 35-49.

Tong, H., 2000. Les Rhinocéros des sites à fossiles humains de Chine. L'Anthropology

104, 523-529.

Tong, H., 2001. Age profiles of rhino fauna from the Middle Pleistocene Nanjing Man

site, South China – explained by the rhino specimens of living species.

International Journal of Osteoarchaeology 11, 231-237.

Tong, H., Patou-Mathis M., 2003. Mammoth and other proboscideans in China during the

Late Pleistocene. Deinsea 9, 421-428.

Tong, H., Liu, J., 2004. The Pleistocene-Holocene extinctions of mammals in China. In:

Dong, W. (Ed.). Proceedings of the Ninth Annual Symposium of the Chinese

Society of Vertebrate Paleontology. China Ocean Press, Beijing, pp. 111-119 (in

Chinese with English abstract).

253 Tong, H., 2005. Dental characters of the Quaternary tapirs in China, their significance in

classification and phylogenetic assessment. Geobios. 38, 139-150.

Tong, H.W., 2006. Composition des faunes de mammifères quaternaires en Chine selon

un gradient Nord-Sud. L’anthropologie 110, 870-887.

Tougard, C., Chaimanee, Y., Suteethorn, V., Triamwichanon, S., Jaeger, J.-J., 1996.

Extension of the geographic distribution of the giant panda (Ailuropoda) and

search for the reasons for its progressive disappearance in Southeast Asia during

the Latest Middle Pleistocene. Comptes Rendus de l'Academie des Sciences 323,

973-979.

Tougard, C., Jaeger, J., Chaimanee, Y., Suteethorn, V., Triamwichanon, S., 1998.

Discovery of a Homo sp. tooth associated with mammalian cave fauna of Late

Middle Pleistocene age, Northern Thailand. Journal of Human Evolution. 35, 47-

54.

Tougard, C., 1998. Les faunes de grands mammifères du Pleistocene moyen terminal de

Thaïlande dans leur cadre phylogénétique, paléoécologique et biochronologique.

Ph.D. Thesis, Universite Montpellier II. Montpellier.

Tougard, C., 2001. Biogeography and migration routes of large mammal faunas in South-

East Asia during the Late Middle Pleistocene: focus on the fossil and extant

faunas from Thailand. Palaeogeography, Palaeoclimatology, Palaeoecology 168,

337-358.

Tougard, C., Montuire, S., 2006. Pleistocene paleoenvironmental reconstructions and

mammalian evolution in South-East Asia: focus on fossil faunas from Thailand.

Quaternary Science Reviews 25, 126-141.

254 Travouillon, K. J., Archer, M., Legendre, S., Hand, S. J., 2007. Finding the Minimum

Sample Richness (MSR) for multivariate analyses : implications for

palaeoecology. Historical Biology 19, 315-320.

Trueman, C. N. G., Field, J., Dortch, J., Charles, B., Wroe, S., 2005. Prolonged

coexistence of humans and megafauna in Pleistocene Australia. Proceedings of

the National Academy of Science 102, 381-8385.

Turner, A., Antón, M., 1996. The giant hyaena Pachycrocuta brevirostris (Mammalia,

Carnivora, Hyaenidae). Geobios 29, 455-468.

UNEP WCMC, 2007. http://unep-wcmc.org.

UNSECO 2007. Man and the Biosphere. http://www2.unesco.org/

mab/br/brdir/directory/biores.asp. van den Bergh, G., 1999. The late Neogene elaphantoid-bearing faunas of Indonesia and

their palaeozoogeographic implications. A study of the terrestrial faunal

succession of Sulawesi, Flores and Java, including evidence for early hominid

dispersal east of Wallace’s line. Scripta Geologica 117, 1-419. van den Bergh, G., de Vos, J., Sondaar, P., 2001. The Late Quaternary palaeogeography

of mammal evolution in the Indonesian Archipelago. Palaeogeography,

Palaeoclimatology, Palaeoecology 171, 385-408. van der Kaars, W. A., Dam, M. A. C., 1995. A 135,000-year record of vegetational and

climatic change from the Bandung area, West-Java, Indonesia. Palaeogeography,

Palaeoclimatology, Palaeoecology 117, 55-72.

Verstappen, H. Th., 1975. On palaeo climates and landform development in Malesia.

Modern Quaternary Research in SE Asia 1, 3-35.

255 Verstappen, H. Th., 1980. Quaternary climatic changes and natural environments in SE

Asia. Geojournal 4: 45-54.

Verstappen, H. Th., 1997. The effect of climatic change on Southeast Asian

geomorphology. Journal of Quaternary Science 12, 413-418.

Voris, H. K., 2000. Maps of Pleistocene sea levels in Southeast Asia: shorelines, river

systems and time durations. Journal of Biogeography 27, 1153-1167.

Voss, R. S., Emmons, L. H., Mammalian diversity in Neotropical lowland rainforests: a

preliminary assessment. Bulletin of the American Museum of Natural History

230, 1-115.

Voss, R. S., Darrin, P. L., Simmons, N. B., 2001. The mammals of Paracou, French

Guiana: a neotropical lowland rainforest fauna. Part 2. Nonvolant species.

Bulletin of the American Museum of Natural History 263, 1-236.

Wallace, A. R., 1869. The Malay Archipelago. Oxford University Press, Oxford.

Wang, H., 1996. Climate and habitat reconstruction of hominid sites in Northern China

with paleosol stable isotopes. Ph.D. Dissertation, University of Illinois at Urbana-

Champaign.

Wang, H., Ambrose, S. H., Liu, C. L., Follmer, L. R., 1997. Paleosol stable isotope

evidence for early Hominid occupation of East Asian temperate environments.

Quaternary Research 48, 228-238.

Wang, L., Sarnthein, M., Erlenkeuser, H., Grimalt, J., Grootes, P., Heilig, S., Ivanova, E.,

Keinast, M., Pelejero, C., Pflaumann, U., 1999. East Asian monsoon climate

during the Late Pleistocene: high resolution sediment records from the South

China Sea. Marine Geology 156, 245-284.

256 Warwick, R. M., Clarke, K. R., 1995. New ‘biodiversity’ measures reveal a decrease in

taxonomic distinctness with increasing stress. Marine Ecology Progress Series

129, 301-305.

Watanabe, H., 1985. The Chopper-Chopping tool complex of Eastern Asia: an

ethnoarchaeological-ecological reexamination. Journal of Anthropological

Archaeology 4, 1-18.

Weber, J., Czarnetzki, A., Pusch, C. M., 2005. Comment on “The Brain of LB1, homo

floresiensis.” Science 310, 236b.

Wheeler, P.E. 1992. The thermoregulatory advantages of large body size for hominids

foraging in savannah environments. Journal of Human Evolution 23, 351-362.

West, J.A., Louys, J., 2007. Differentiating bamboo from stone tool cut marks in the

zooarchaeological record, with a discussion on the use of bamboo knives. Journal

of Archaeological Science 34, 512-518.

Westaway, K. E., Morwood, M. J., Roberts, R. G., Rokus, A. D., Zhao, J.-x., Storm, P.,

Aziz, F., van den Bergh, G., Hadi, P., Jatmiko, de Vos, J., 2007. Age and

biostratigraphic significance of the Punung Rainforest Fauna, East Java,

Indonesia, and implications for Pongo and Homo. Journal of Human Evolution,

doi : 10.1016/j.hevol.2007.06.002.

Whitmore, T. C., 1987. Biogeographical evolution of the Malay Archipelago. Clarendon

Press, Oxford.

Whittington, S. L., Dyke, B., 1984. Simulating overkill: Experiments with the Mosimann

and Martin model. In Martin, P. S., Klein, R. G. (Eds.) Quaternary Extinction: A

Prehistoric Revolution. University of Arizona Press, Tucson, pp 451-465.

257 Willig, M. R., McGinley, M. A., 1999. The response of animals to disturbance and their

roles in patch generation. In: Walker, L. R. (Ed.) Ecosystems of Disturbed

Ground. Elsevier Science B.V., Amsterdam, p 633-657.

Wroe, S., Field, J., Fullagar, R., Jermiin, L.S., 2004. Megafaunal extinction in the late

Quaternary and the global overkill hypothesis. Alcheringa 28(1), 291-331.

Wroe, S., Field, J., 2006. A review of the evidence for a human role in the extinction of

Australian megafauna and an alternative interpretation. Quaternary Science

Reviews 25, 2692-2703.

Wu, M.L., Wang, L.H., Zhang, Y.Y., & Zhang, S.S. 1975. Fossil human teeth and

associated fauna from northern Guizhou. Vertebrata PalAsiatica 13, 14-23.

Wu, R. K., 1964. Mandible of the Sinanthropus-type discovered at Lantian, Shaanxi.

Vertebrata PalAsiatica8, 1-17.

Wu, R. K., 1966. The hominid skull of Lantian, Shaanxi. Vertebrata PalAsiatica10, 1-12.

Wu, R.K. & Lin, S.L. 1985. Chinese palaeoanthropology: retrospect and prospect. In:

Wu, R. K. and Olsen, J. W., (Eds.) Palaeoanthropology and Palaeolithic

archaeology in the People’s Republic of China. Academic Press, New York, pp.

1-17.

Wu, X., Yuan, Z., Han, D., Qi, T., 1966. Report of the excavation at Lantian Man locality

of Gongwangling in 1965. Vertebrata PalAsiatica.10, 23-29.

Wu, X.Z. & Wu, M.L. 1985. Early Homo sapiens in China. In: Wu, R. K. and Olsen, J.

W., (Eds.) Palaeoanthropology and Palaeolithic archaeology in the People’s

Republic of China. Academic Press, New York, pp. 91-106.

258 Wu, X., Poirier, F.E. 1995. Human evolution in China. Oxford University Press, New

York.

Wu, Y. H., Ji, W.T., 2002. Researches on Poyang Lake National Nature Reserve, Jiangxi

Province. China Forestry Publishing House, Beijing, China.

Wu, Y., Zhao, J., Cheng, J., 2006. Scientific survey and study on biodiversity in Hebei

Maojingba Nature Reserve, China. Science Press, Beijing, China.

Xiang, Z., Liang, X., Huo, S., Ma, S., 2004. Quantitative analysis of land mammal

zoogeographic regions in China and adjacent regions. Zoological Studies 43, 142-

160.

Xie, Y., MacKinnon, J., Li, D., 2004. Study on biogeographical divisions of China.

Biodiversity and Conservation 13, 1391-1417.

Yamei, H., Potts, R., Baoyin, Y., Zhengtang, G., Deino, A., Wei W., Clark J., Guangmao

X., Weiwen H., 2000. Mid-Pleistocene Acheulean-like stone technology of the

Bose Basin, South China. Science 287, 1622-1626.

Yokoyama, Y., Jacob, T., Falgueres, C. & Semah, F. 1998. Direct dating of Homo erectus

skulls of Solo man in Java by non-destructive gamma-ray. Abstracts of Dual

Congress of the Intl Assoc Study Hum Paleo and Intl Assoc Hum Biol South

Africa, p. 57 (Abstract).

Yokoyama, Y. & Falgueres, C. 2002. Direct dating of hominid remains in Eurasia.

EGS XXVII General Assembly, Nice, 21-26 April (Abstract).

Yuan, S., Chen, T.M. & Gao, S. 1986. Uranium series chronological sequence of some

Palaeolithic sites in South China. Acta Anthropologica Sinica 5, 179-190.

259 Zeiton, V., Seveau, A., Forestier, H., Thomas, H., Lenoble, A., Laudet, F., Antoine, P-O.,

Debruyne, R., Ginsberg, L., Mein, P., Winayalai, C., Chumdee, N., Doyasa, T.,

Kijngam, A., Nakbunlung, S., 2005. Découverte d'un assemblage faunique à

Stegodon-Ailuropoda dans une grotte du Nord de la Thaïlande (Ban Fa Suai,

Chiang Dao). C. R. Palevol. 4, 255-264.

Zhang, S. 1985. The early Palaeolithic of China. . In: Wu, R., Olsen J, (Eds.)

Palaeoanthropology and Palaeolithic Archaeology in the People's Republic of

China. Academic Press, Orlando, pp.147-186.

Zhang, Y., et al. 1997. Distribution of Mammalian Species in China. China Forestry

Publishing House, Beijing.

Zhang, Z., 2002. Geological events and mammalian distribution in China. Acta Zoologica

Sinica 48, 141-153.

Zhao, J.X., Hu, K., Collerson, K.D. & Xu, H.k. 2001. Thermal ionization mass

spectrometry U-series dating of a hominind site near Nanjing, China. Geology 29,

27-30.

Zhao, S., Fang, J., 2006. Patterns of species richness for vascular plants in China’s nature

reserves. Diveristy Distrib. 12, 364-372.

Zhao, Y.M., Song, C.S., 1995. Scientific Survey of the Yellow River Delta Nature

Reserve. Chinese Forestry Publishing House, Beijing, China.

Zhao, Z., Fang, J., Peng, C., Tang, Z., 2006. Relationships between species richness of

vascular plants and terrestrial vertebrates in China: analyses based on data of

nature reserves. Diversity and Distribution, 12, 189-194.

260 Zheng, S. H., 1984. A revision of the fossil Cricetines (Rodentia, Mammalia) from

Zhoukoudian, the Peking Man site. Vertebrata PalAsiatica 22, 179-197.

Zheng, S. H., 1993. Quaternary rodents of Sichuan-Guizhou area, China. Science Press,

Beijing, pp. 109-125 (in Chinese with English summary).

Zheng, S. H., 1997. Evolution of the Mesosiphneinae (Siphneidae, Rodentia) and

environmental change. In Tong, Y.S (Ed.) Evidence for evolution – essays in

honour of Prof. Chungchien Young on the hundredth anniversary of his birth.

China Ocean Press, Beijing, pp. 137-150 (in Chinese with English summary).

Zheng, S. H., Li, C. K., 1990. Comments on fossil arvicolids of China. International

Symposium on Evolution, Phylogeny and Biostratigraphy of Arvicolids

(Rodentia, Mammalia), pp. 431-442. Praha.

Zhou, C., Wang, Y.J., Cheng, H. & Liu, Z. 1999. Discussion on Nanjing Man’s age. Acta

Anthropologica Sinica 18, 255-262.

Zhou, Z., 1990. Research on the Fanjing Mountain. Forestry Bureau of Guizhou

Province, Guizhou People’s Publishing House, Guizhou, China.

Zhu, R. X., Potts, R., Xie, F., Hoffman, K. A., Deng, C. L., Shi, C. D., Pan, Y. X., Wang,

H. Q., Shi, R. P., Wang, Y. C., Shi, G. H., Wu, N. Q., 2004. New evidence on the

earliest human presence at high latitudes in northeast Asia. Nature 431, 559-562.

Zhu, S., Yang, Y., 1990. Forest types. In Zhou, Z (Ed.) Research on the Fanjing

Mountain, Forestry Bureau of Guizhou Province, Guizhou People’s Plublishing

House, Guizhou, China, pp. 17-21.

Zhu, Y., Zhou, M., 1994. The Asian monsoon variation and the habituation environment

of fossil man in Northern China. Catena 22, 121-131.

261 Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S., Twickler, M.S., and Taylor,

K., 1996. Potential atmospheric impact of the Toba mega-eruption ~71,000 years

ago Geophysical Research Letters 23, 837-840.

262 APPENDIX A

Appendix A lists all the mammals present in the national parks and nature reserves discussed in the thesis. For the source of these species lists, see Table 5.2.

263

r

r Kerinci Seblat Kerinci Seblat Bukit Barisan Bandhavgarh Maojingba Changtang Wulate Yunwushan Longqi Gunung Mulu Poyang Altun Mountain Altun Mountain Sai Yok Khao Yai Krachan Kaeng Erawan Gunung Leuse Species Ailaoshan Baima Xueshan Bogdhad Changbai Changshanerhai Daweishan Dinghushan Fanjingshan Wuyishan Funiushan Gaolingong Gobi Jiuzhaigou Maolan Medong Nujiang Tianmushan Wolong Xishuangbanna Yancheng Yellow Rive Family Ailuridae Ailurus fulgens 1 1 11111 1 1 Alticola Arvicolidae argentatus 1 Clethrionomys Arvicolidae rufocamus 1 1 Eothenomys Arvicolidae custos 1 Arvicolidae Eothenomys eva 1 Arvicolidae Eothenomys inez 1 Eothenomys Arvicolidae melanogaster 1 11 1 111 1 Eothenomys Arvicolidae olitor 1 Arvicolidae Microtus fortis 1 1 Arvicolidae Microtus irene 1 Microtus Arvicolidae oeconomus 1 Bovidae Bos frontalis 1 1111 1 Bovidae Bos gaurus 1 Bovidae Bos grunniens 1 Bovidae Bos javanicus 1 11 Bovidae Bos mutus 1 Boselaphus Bovidae tragocamelus 1 Budorcas Bovidae taxicolor 1 11111 1 Bovidae Capra ibex 1 11 Bovidae Capra sibirica 1 1 Capricornis Bovidae sumatraensis 1 1 1 1

264 Bovidae Gazella gazella 1 Gazella Bovidae subgutturosa 1 1 1 1 Hemitragus Bovidae jenlahious 1 Naemorhedus Bovidae baileyi 1 Naemorhedus Bovidae caudatus 1 11 Naemorhedus Bovidae cranbrooki 1 Naemorhedus Bovidae goral 1 1 1 11 1 Naemorhedus Bovidae sumatraensis 1 11111 111 1 1111 1 Bovidae Ovis ammon 1 1 1 1 1 Pantholops Bovidae hodgsoni 1 1 Procapra Bovidae picticaudata 1 1 Procapra Bovidae przewalskii 1 Procarpra Bovidae gutturosa 1 Bovidae Pseudois nayaur 1 1 1 1 1 1 1 Pseudois Bovidae schaeferi 1 Bovidae Saiga tatarica 1 Tetracerus Bovidae quadricornis 1 Canidae Canis aureus 11 1 Canidae Canis lupus 1 1 1 111 1 11 111 Canidae Cuon alpinus 1 11 11 111 111 1111 1 Nyctereutes Canidae procyonoides 1 11111 1 11 1 1 Vulpes Canidae bengalensis 1 Canidae Vulpes corsac 1 11 Canidae Vulpes ferrilata 1 Canidae Vulpes vulpes 1 1111111 1 111 11 1 111 1 Cercopithecid Macaca ae arctoides 1 1 1 1 1 1 1

265 Cercopithecid Macaca ae assamensis 11 1 1 11 Cercopithecid Macaca ae fascicularis 1 11111 1 Cercopithecid ae Macaca mulatta 1 1 11 11 1 11111 1 1 1 1 1 Cercopithecid Macaca ae nemestrina 1 1 1 1111111 1 1 Cercopithecid Macaca ae thibetana 1 1 1 Cercopithecid Presbytis ae cristata 11 1 Cercopithecid Presbytis ae melalophos 1 1 1 Cercopithecid Presbytis ae obscura 1 Cercopithecid Presbytis ae thomasi 1 Cercopithecid Prespytis ae entellus 1 Cercopithecid ae Prespytis geei 1 Cercopithecid ae Prespytis hosei 1 Cercopithecid Prespytis ae melalophus 1 Cercopithecid Prespytis ae rubicunda 1 Cercopithecid Semnopithecus ae entellus 1 1 Cercopithecid Trachypithecus ae cristatus 1 1 1 Cercopithecid Trachypithecus ae obscurus Cercopithecid Trachypithecus ae phayrei 1 1 1 1 1 Cercopithecid Rhinopithecus ae bieti 1 1 1 Cercopithecid Rhinopithecus ae roxellana 111 Cervidae Axis axis 1 Cervidae Axis porcinus 1 Capreolus Cervidae capreolus 1 1 1 1

266 Capreolus Cervidae pygargus 1 Cervus Cervidae albirostris 1 1 1 Cervidae Cervus elaphus 1 1 1 1 Cervidae Cervus nippon 1 1 Cervidae Cervus unicolor 1 11 1 1 1 11111111 1 1 Elaphodus Cervidae cephalophus 1 1 1 1 1 1 1 1 1 Hydropotes Cervidae inermis 1 1 Muntiacus Cervidae crinifrons 1 1 Cervidae Muntiacus feae 1 11 Muntiacus Cervidae muntjak 1 11 1 1 11111111 1 Muntiacus Cervidae reevesi 1 1 1 1 1 1 1 1 1 Cricetulus Cricetidae barabensis 1 1 1 11 Cricetulus Cricetidae longicaudua 1 Cricetulus Cricetidae migritorius 1 11 Cricetidae Cricetulus triton 1 1 1 1 Crictulus Cricetidae eversmanni 1 Cricetidae Lagurus luteus 1 Meriones Cricetidae meridianus 1 1 11 Meriones Cricetidae unguiculatus 1 1 Microtus Cricetidae mandarinus 1 Microtus Cricetidae maximowiczii 1 Myospalax Cricetidae cansus 1 Myospalax Cricetidae fontanieri 1 1 Myospalax Cricetidae psilurus 1 1 1 Phodopus Cricetidae roborovakii 11

267 Phodopus Cricetidae sungorus 1 Rhombomys Cricetidae opimus 1 1 Cynocephalid Cynocephalus ae variegates 1 111 1 Allactaga Dipodidae bullata 1 1 Cardiocraninae Dipodidae paradoxus 1 Dipodidae Dipus andrewsi 1 Dipodidae Dipus sagitta 1 Salpingotus Dipodidae kozlovi 1 Dipodidae Zapus trinotatus 1 Elephas Elephantidae maximus 1 1111111 Equidae Equus hemionus 1 1 1 Equidae Equus kiang 1 1 Echinosorex Erinaceidae gymnurus 1 1 1 Erinaceus Erinaceidae amurensis 1 Erinaceus Erinaceidae dauricus 1 Erinaceus Erinaceidae europaeus 1 1 1 1 1 1 1 Hemiechinus Erinaceidae auritus 1 Hemiechinus Erinaceidae dauricus 1 1 Erinaceidae Hylomys auillus 1 Erinaceidae Hylomys parvus 1 Hylomys Erinaceidae sinensis 1 1 Erinaceidae Hylomys suillus 1 1 1 1 1 Felis Felidae bengalensis 1 1 1 1 11 1 Felidae Felis bieti 11 Felidae Felis chaus 11 1 1 1 1 Felidae Felis lynx 1 111 Felidae Felis manul 1 1111

268 Felidae Felis manul 1 1 Felidae Felis temmincki 1 1 11 Felidae Felis temmincki 1 11 111 111 1 1 1 Felidae Felis viverrina 1 Felidae Lynx lynx 1 1 1 1 Neofelis Felidae nebulosa 1 11 11 1 11 1 1111111 1 1 1 Felidae Panthera pardus 11111111 11111 1 1111 111 1 Felidae Panthera tigris 1 11111 1 11 1 11111111 1 Felidae Panthera uncia 1 1 1 Felidae Panthera unica 1 1 Pardofelis Felidae marmorata 1 1 1 1 Prionailurus Felidae bengalensis 111 1 11 11 1 11 Prionailurus Felidae planiceps 11 Prionailurus Felidae viverrinus 1 1 1 Herpestes Herpestidae brachyurus 1 Herpestes Herpestidae edwardsii 1 Herpestes Herpestidae javanicus 1 1 1 Herpestes Herpestidae semitorquatus 1 Herpestes Herpestidae smithii 1 Herpestidae Herpestes urva 111 1 11 1 Pongo Hominidae pygmaeus 1 Hyaenidae Hyaena hyaena 1 Hylobatidae Hylobates agilis 11 Hylobates Hylobatidae concolor 1 1 1 Hylobates Hylobatidae hoolock 1 Hylobatidae Hylobates lar 11111 Hylobates Hylobatidae leucogenys 1

269 Hylobates Hylobatidae muelleri 1 Hylobates Hylobatidae pileatus 1 Hylobates Hylobatidae syndactylus 111 Atherurus Hystricidae macrourus 1 1 1 Hystrix Hystricidae brachyura 1 1111111 1 11 1 1111111 Hystrix Hystricidae hodgsoni 1 1 Hystricidae Hystrix indica 1 Hystrix Hystricidae sumatrae 1 Trichys Hystricidae fasciculate 111 Hystricidae Trichys lipura 1 Leporidae Lepus capensis 1 1 1 1 1 11 Leporidae Lepus comus 11 1 1 Leporidae Lepus nigricollis 1 Leporidae Lepus oiostolus 1 1 Leporidae Lepus peguensis 1 1 Leporidae Lepus sinensis 1 1 1 1 1 Leporidae Lepus timidus 1 Leporidae Lepus tolai 1 Nesolagus Leporidae netscheri 11 Nycticebus Lorisidae coucang 1 1 1 1 1 1 1 1 Nycticebus Lorisidae pygmaeus 1 Manis Manidae crassicaudata 1 Manidae Manis javanica 11 11 Manis Manidae pentadactyla 1 1 11111 1 1 1 1 1 1 Moschus Moschidae berezovskii 1 1 1 1 11 1111 1 1 Moschus Moschidae chrysogaster 1 1 Moschidae Moschus fuscus 11 1

270 Moschus Moschidae moschiferus 1 1 Moschus Moschidae sifanicus 1 Apodemus Muridae agrarius 111 1 1 1 1 1 Apodemus Muridae chevrieri 1 1 1 1 Muridae Apodemus draco 1 1 1 1 1 1 1 1 Apodemus Muridae latronum 1 1 Apodemus Muridae peninsulae 1 1 Apodemus Muridae speciosus 1 Bandicota Muridae bengalensis 1 Bandicota Muridae indica 1 1 1 1 Bandicota Muridae savilei 1 Berylmys Muridae bowersi 1 11 1 1 Cannomys Muridae badius 11 Chiropodomys Muridae gliroides 1 1 1 1 Chiropodomys Muridae sp. 1 Eolagurus Muridae przewalskii 1 Hapalomys Muridae longicaudatus 1 Muridae Lenothrix canus 1 Leopoldamys Muridae edwardsi 1 1 1 1 1 1 Leopoldamys Muridae sabanus 1 11 Maxomys Muridae hylomoides 11 Maxomys Muridae inflatus 1 Muridae Maxomys rajah 1 1 Muridae Maxomys surifer 1 111 1

271 Maxomys Muridae whiteheadi 111 Melomys Muridae rattoides 1 Micromys Muridae minutus 1 1 1 1 1 1 1 Muridae Mus booduga 1 Muridae Mus caroli 1 1 1 Muridae Mus cookii 1 Mus Muridae crociduroides 1 Muridae Mus musculus 1 1111 1 1 11 1 1 1 11 1 Muridae Mus pahari 1 1 1 1 1 Muridae Niniventer bukit 1 Muridae Niniventer rapit 1 1 Niviventer Muridae brahma 1 Niviventer Muridae confucianus 1 11 111 1 1 Niviventer Muridae coxingi 1 Niviventer Muridae cremoriventer 1 Muridae Niviventer eha 1 1 Niviventer Muridae fulvescens 1 11111 1 11 1 1 1 Niviventer Muridae niviventer 111 1 Muridae Niviventer rapit 1 Rattus Muridae argentiventer 1 1 Muridae Rattus bowersi 1 Rattus Muridae confucianus 1 1 Rattus Muridae cremoriventer 1 Muridae Rattus edwardsi 1 1 Muridae Rattus exulans 111 Rattus Muridae flavipectus 1 1 1 Rattus Muridae fulvescens 1 1

272 Rattus Muridae hoogerwerfi 1 Rattus Muridae infraluteus 1 Muridae Rattus losea 1 1 1 Muridae Rattus muelleri 1 1 Muridae Rattus nitidus 1 11111 111 1 1 Rattus Muridae niviventer 11 1 Rattus Muridae norvegicus 1 1111 11 1 1 11 1 Muridae Rattus rapit 1 Muridae Rattus rattus 11 1 1 1111 1 1 Muridae Rattus sabanus 1 Rattus Muridae sikkimensis 1 Muridae Rattus surifer 1 Muridae Rattus tanezumi 1 11111 1 11 1 Rattus Muridae tiomanicus 111 Rattus Muridae whiteheadi 1 Rhizomys Muridae pruinosus 1111 11 Rhizomys Muridae sinensis 1 1 1 1 1 1 Rhizomys Muridae sumatrensis 111 Sundamys Muridae infraluteus 11 Sundamys Muridae muelleri 1 1 Muridae Vernaya fulva 1 1 1 Volemys Muridae millicens 1 Abditomys Muridae latidens 1 Mustelidae Aonyx cinerea 1 Mustelidae Aonyx cinerea 1 111 1 Mustelidae Aonyx. sp. 1 Arctonyx Mustelidae collaris 1111 111 1 111111 1

273 Mustelidae Lutra lutra 1 1 1 11111 1 111 1 1 1 Lutra Mustelidae perspicillata 1 Mustelidae Lutra sp 1 Lutrogale Mustelidae perspicillata 1 1 1 Mustelidae Lutrogale. sp 1 Mustelidae Martes flavigula 1 111 111 111 1 1111111 1 1 1 Mustelidae Martes foina 1 1 1 1 Mustelidae Martes zibellina 1 Mustelidae Meles meles 1 1 111 111 1 1 11 1 Mellivora Mustelidae capensis 1 Melogale Mustelidae moschata 1 111111 1 11 1 1 Melogale Mustelidae personata 11 Mustelidae Mustela altaica 1 1 1 Mustela Mustelidae eversmanni 1 1 1 1 Mustelidae Mustela kathiah 1 111111 11 1 Mustela Mustelidae lutreolina 1 Mustelidae Mustela nivalis 1 Mustelidae Mustela nudipes 111 Mustela Mustelidae putorius 1 Mustelidae Mustela sibirica 11111111 1 111 1 1 1 1 Mustela Mustelidae strigidorsa 1 1 Mydaus Mustelidae javanensis 11 Vormela Mustelidae peregusna 1 Ochotonidae Ochotona alpina 1 Ochotona Ochotonidae curzoniae 1 Ochotona Ochotonidae dauurica 1 1 Ochotona Ochotonidae gaoligongensis 1

274 Ochotona Ochotonidae koslovi 1 Ochotona Ochotonidae pallasi 1 1 Ochotonidae Ochotona rutila 1 Ochotona Ochotonidae thibetana 1 1 1 1 Belomys Petauristidae pearsonii 1 1 Hylopetes Petauristidae alboniger 1 1 1 Hylopetes Petauristidae lepidus 1 Trogopterus Petauristidae xanthipes 1 1 1 Platacanthom Typhlomys yidae cinereus 1 1 1 1 Aeromys Pteromydae tephromelas 11 1 Petinomys Pteromydae genibarbis 1 Petinomys Pteromydae setosus 111 Rhinocerotida Dicerorhinus e sumatrensis 111 Rhinocerotida Rhinoceros e sondaicus 1 Callosciurus Sciuridae pygerythrus 1 1 Callosciurus Sciuridae quinquestriatus 1 Callosciurus Sciuridae albescens 1 Callosciurus Sciuridae baluensis 1 Callosciurus Sciuridae caniceps 111 Callosciurus Sciuridae erythraeus 1 11 111 1 11 1 1 Callosciurus Sciuridae finlaysonii 111 Callosciurus Sciuridae notatus 1 1 Callosciurus Sciuridae prevosti 111 1

275 Calloscriurus Sciuridae albiculus 1 Calloscriurus Sciuridae nigrovittatus 1 1 Calloscriurus Sciuridae notatus 11 Citellus Sciuridae dauricus 1 Dremomys Sciuridae everetti 1 Dremomys Sciuridae lokriah 1 Dremomys Sciuridae pernyi 11 11 1 111 1 Dremomys Sciuridae rufigenis 1 1 1 1 1 Eutamias Sciuridae sibiricus 1 Exilisciriurus Sciuridae exilis 1 Exilisciriurus Sciuridae whiteheadi 1 Funambulus Sciuridae pennantii 1 Lariscus Sciuridae insignis 111 1 Sciuridae Lariscus niobe 1 Marmota Sciuridae himalayana 1 1 1 Marmota Sciuridae sibirica 1 Menetes Sciuridae berdmorei 111 Nannosciurus Sciuridae melanotis 11 Petaurista Sciuridae alborufus 11 1 1 Petaurista Sciuridae elegans 1 11 1 1 1 111 Petaurista Sciuridae leucogenys 1 Petaurista Sciuridae petaurista 1 1 1 1 1 1 1 1 1 Sciuridae Pteromys volans 1 1

276 Sciuridae Ratufa affinis 11 1 Sciuridae Ratufa bicolor 11 1 1 111111 Rheithrosciurus Sciuridae macrotis 1 Rhinosciurus Sciuridae laticaudatus 1 1 Sciurotamias Sciuridae davidianus 1 1 1 Sciurotamias Sciuridae forresti 1 1 1 1 Sciuridae Sciurus vulgaris 1 Spermophilus Sciuridae alashanicus 1 Spermophilus Sciuridae dauricus 1 1 Spermophilus Sciuridae erythrogenys 1 1 Sciuridae Suncus etruscus 1 1 Sciuridae Suncus murinus 1 1 1 1 1 Sundasciurus Sciuridae altitudinis 1 Sundasciurus Sciuridae hippurus 111 1 Sundasciurus Sciuridae jentiki 1 Sundasciurus Sciuridae lowii 111 Sundasciurus Sciuridae tenuis 111 Sciuridae Tamias sibiricus 1 Sciuridae Tamias sibiricus 1 1 Tamiops Sciuridae macclellandi 1 1 1 111 Tamiops Sciuridae maritimus 1 Tamiops Sciuridae swinhoei 1 111111111 1 Anourosorex Soricidae squamipes 1 11 11 Blarinella Soricidae quadraticauda 1 1 1 1 Crocidura Soricidae attenuata 11 11 1 11 11 1

277 Crocidura Soricidae beccarii 1 Crocidura Soricidae fuliginosa 1 1 1 1 1 Crocidura Soricidae horsfieldii 1 1 1 Crocidura Soricidae monticola 1 Crocidura Soricidae russula 1 1 Crocidura Soricidae suaveolens 1 1 1 1 Nectogale Soricidae elegans 1 1 1 Soricidae Sorex araneus 1 Soricidae Sorex asper 1 Sorex Soricidae bedfordiae 1 1 Sorex Soricidae cylindricauda 1 1 1 1 Soricidae Sorex excelsus 1 Soricidae Sorex sinalis 1 Soriculus Soricidae caudatus 1 1 Soriculus Soricidae hypsibius 1 Soriculus Soricidae lamula 1 Soriculus Soricidae leucops 1 1 1 Soriculus Soricidae macrurus 1 1 1 Soriculus Soricidae salenskii 1 Suidae Sus barbatus 11 1 Suidae Sus scrofa 1 1 1111111 1 11 1 111111111 1 1 Euroscaptor Talpidae longirostris 1 Mogera Talpidae insularis 1 1 Talpidae Mogera robusta 1 Talpidae Mogera wagura 1 Talpidae Parascaptor 1 1 1

278 leucura Scaptochirus Talpidae moschatus 1 11 Scaptonyx Talpidae fusicaudus 1 1 1 1 Talpidae Talpa altaica 1 Talpa Talpidae moschatus 1 Uropsilus Talpidae gracilis 1 1 Uropsilus Talpidae investigator 1 Uropsilus Talpidae soricipes 1 Tapiridae Tapirus indicus 1 11 11 Tarsius Tarsiidae bancanus 1 1 Hyemoschus Tragulidae aquaticus 1 Tragulus Tragulidae javanicus 1 11111 1 1 Tragulidae Tragulus napu 111 Tupaiidae Anathana ellioti 1 Dendrogale Tupaiidae melanura 1 Tupaia Tupaiidae belangeri 1 1 1 1 1 Tupaiidae Tupaia glis 1111111 Tupaiidae Tupaia gracilis 1 Tupaiidae Tupaia javanica 11 Tupaiidae Tupaia minor 1 Tupaiidae Tupaia montana 1 1 Tupaiidae Tupaia picta 1 1 Tupaiidae Tupaia tana 111 Ailuropoda Ursidae melanoleuca 1 1 Helarctos Ursidae malayanus 1 1111111 1 1 Ursidae Ursus arctos 1 1 1 1 Ursus Ursidae thibetanus 1 111 11 1 1 1 1111 11 1 Ursidae Ursus ursinus 1

279 Arctictis Viverridae binturong 1 1 1111111 1 Arctogalidia Viverridae trivirgata 1 1 11 1 Chrotogale Viverridae owstoni 1 Cynogale Viverridae bennettii 1 1 Hemigalus Viverridae derbyanus 1 11 1 Viverridae Paguma larvata 1 111111 1 11 1 111111 1 1 Paradoxurus Viverridae hermaphroditus 1 1 11111111 1 Prionodon Viverridae linsang 1111 Prionodon Viverridae pardicolor 1 1 1 1 1 1 Viverra Viverridae megaspila 1 Viverra Viverridae tangalunga 1 1 1 Viverridae Viverra zibetha 1 1 1111 1111 1111 1 Viverricula Viverridae indica 111111 11111 1 11 1 1 1 Zapodidae Sicista concolor 1

280 APPENDIX B

Appendix B lists the complete species list for the Pleistocene sites considered in Chapter

6. These lists were complied from Fromaget (1936), Arambourg and Fromaget

(1938),Colbert (1943), Medway (1960, 1972), Beden and Guérin (1973), Pope et al.

(1981), de Vos (1983), Tougard (1998), Olsen and Ciochon (1990), van den Bergh et al.

(2001), Meijaard (2004), and Zeiton et al. (2005).

281 Country Vietnam Vietnam Vietnam Vietnam Vietnam Vietnam Vietnam Java Java Java Java Java Java Java Java Java Java Malaysia Borneo Sumatra Sumatra Thailand Thailand Thailand Thailand Cambodia Burma Burma Laos Laos Site Tham Khuyen Tham Hai Tham Om Hang Hum 1 Hang Hum 2 Lang Trang Keo Leng Tjidulang Kali Glagah Satir Bukuran Ci Saat Trinil Kedung Brubus Ngangdong Punung Holocene Caves Java Kinta Valley Niah Lider Ajer Sibrambang Thum Wimam Nakin Thum Phra Khai Phet Ban Fa Suai Kao Pah Nam Phnom Loang Mogok Caves Irrawaddy beds Tam Nang Tam Hang PRIMATES Gigantopithecus blacki 1 ? Homo erectus 1 1 ? 1 1 1 Homo sapiens 1 1 1 1 1 1 1 1 Homo sp. 1 Hylobates moloch 1 Hylobates sp. 1 1 1 Macaca anderssoni 1 Macaca cf. assamensis 1 1 Macaca cf. mulatta 1 Macaca cf. nemestrina 1 Macaca fascicularis 1 1 1 1 1 1 1 Macaca nemestrina 1 1 1 Macaca robustus 1 1 Macaca sp. 1 1 1 1 1 1 1 1 1 Macaca rhesus 1 Nomascus concolor 1 Nycticebus coucang 1 Pongo pygmaeus 1 1 1 1 1 ? 1 ? 1 1 1 1 1 1 1 1

282 Pongo sp. ? Presbytis cristata 1 Presbytis comata ? Prespytis (Semnopithecus) sp. ? ? 1 1 Prespytis melalophos 1 Prespytis sp. 1 1 Hylobates (Symphalangus) syndactylus ? 1 1 1 Trachypithecus auratus 1 1 1 1 1 Trachypithecus cristatus 1 1 Trachypithecus sp. 1

CARNIVORA Ailruropoda cf. melanoleuca 1 Ailuropoda melanoleuca 1 1 1 1 1 1 1 Amblonyx cinereus 1 Arcticitus binturong 1 Arctogalidia sp. 1 Arctonyx collaris 1 1 1 1 1 1 1 1 Arctonyx rostratis 1 Canis cf. cynoides 1 Catopuma temminckii 1 1 Crocuta crocuta 1 1 1 1 1 Crocuta sp. 1 Cuon alpinus 1 1 1 Cuon javanicus = C. alpinus 1 Cuon sp. 1 1 1 1 1 1 Cynogale sp. 1 Felis cf. micotis 1 Felis sp. 1 1

283 Felis issidiorensis 1 Homotherium 1 Hyaena brevirostris 1 Lutra lutra 1 Lutra robusta 1 Lutra sp. ? Lutra sumatrana 1 Lutrogale palaeoleptonyx = perspicillata 1 Lutrogale perspicillata 1 Martes flavigula 1 1 Martes sp. 1 Mececyon trinilensis 1 Melogale orientalis 1 Mustela nudipes 1 Neofelis nebulosa 1 1 Nycterertes sp. 1 Paguma cf. larvata 1 Paguma larvata 1 1 Panthera cf. tigris 1 Panthera pardus 1 1 1 1 1 ? ? ? 1 Panthera sp. ? Panthera tigris 1 1 1 1 ? 1 1 1 1 1 1 1 1 1 Panthera leo 1 Paradoxurus cf. hermaphroditus 1 1 1 1 1 1 Paradoxurus hermaphroditus 1 1 Paradoxurus sp. 1 1 Prionailurus bengalensis 1 1 1 1 1 1 Ursus angustidens 1 1 Ursus cf. thibetanus 1 Ursus kokeni 1

284 Ursus malayanus 1 1 1 1 1 1 1 1 Ursus thibetanus 1 1 1 1 1 1 1 1 Viverra cf. zibetha 1

PROBOSCIDEA Stegodon sp. 1 Elephas cf. maximus 1 Elephas hysudricus 1 Elephas hysudrindicus 1 1 Elephas maximus 1 1 1 1 Elephas planifrons 1 1 Elephas sp. 1 1 1 Mastodon bumiajuensis 1 Palaeoloxodon (Elephas) cf. namadicus 1 1 Palaeoloxodon (Elephas) namadicus ? 1 1 1 1 Stegodon elephantoides 1 1 Stegodon insignis 1 Stegodon orientalis 1 1 1 1 1 1 1 1 Stegodon sinensis 1 Stegodon trigonocephalus 1 1 1 1 1 1 1 Stegolophodon latidens 1 Tetralophodon bumiajuensis 1 1

PERISSODACT YLA Dicerorhinus 1 1 1 1

285 sumatrensis Equus yunnanensis 1 Hipparion cf. antelopinum 1 Megatapirus augustus 1 1 1 1 1 1 1 1 1 Rhinoceros cf. plicideus 1 Rhinoceros cf. sivalensis 1 Rhinoceros cf. unicornis 1 Rhinoceros sinensis 1 1 1 1 1 1 Rhinoceros sivalensis 1 1 Rhinoceros sondaicus 1 1 1 1 1 1 1 1 1 1 Rhinoceros sp. 1 1 Rhinoceros sp. 1 1 Rhinoceros sp. 2 1 Rhinoceros sp. 3 1 Rhinoceros unicornis 1 Tapirus indicus 1 1 1 1 1 1 1 1 1 1 1 Tapirus sp. 1 1 1 ?

ARTIODACTYL A Antilope saatensis 1 1 Axis kuhlii 1 Axis lydekkeri 1 1 1 1 1 1 Axis porcinus 1 1 1 Axis sp. 1 Bos (Bibos) cf. sondaicus = javanicus 1 Bos (Bibos) gaurus 1 1 1 ? 1 1 Bos (Bibos) palaesondaicus 1 1 1 Bos frontalis 1 Bos geron 1 1

286 Bos javanicus 1 1 1 1 1 1 Bos sauveli 1 1 Bos sp. 1 1 Boselaphine 1 Bovine 1 1 Bubalus arnee 1 1 Bubalus bubalus 1 1 1 1 1 1 1 1 1 1 Bubalus cf. bubalis 1 1 Bubalus palaeokerabau 1 1 1 Bubalus teilhardi 1 Bubalus sp. 1 Caprinae gen et sp indet 1 1 Cervinae indet 1 1 Cervus cf. unicolor 1 Cervus eldii 1 1 1 Cervus nippon 1 Cervus cf. leptodus 1 1 Cervus palaeojavanicus 1 1 Cervus problematicus 1 1 Cervus orientalis 1 Cervus sp. 1 1 1 1 1 1 1 1 1 1 1 1 Cervus sp. (Rusa sp.) 1 1 1 Cervus stehlini 1 Cervus hippelaphus (timorensis) 1 1 1 1 1 Cervus unicolor 1 1 1 1 1 1 1 1 1 1 1 Duboisia santeng 1 1 1 1 Elaphodus. sp. 1 Epileptobos groeneveldtii 1 Gazella sp. 1 Hemibos triquetricornis 1 Hexaprotodon cf. sivalensis 1

287 Hexaprotodon iravaticus 1 Hexaprotodon sivalensis 1 1 1 1 1 1 1 1 Hexaprotodon sp. 1 1 Hippotragine 1 Merycopotamus dissimilis 1 Merycopotamus nanus 1 Muntiacus muntjack 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Muntiacus sp. 1 1 1 1 Naemorhedus sumatraensis 1 1 1 1 1 1 1 1 1 Naemorhedus caudatus 1 Naemorhedus goral 1 1 1 Naemorhedus cinereus 1 Naemorhedus edwardsi 1 Naemorhedus sp. 1 Potamochoerus sp. 1 Proboselaphus sp. 1 Proleptobos birmanicus 1 Spirocerus sp. ? 1 Sus barbatus 1 1 1 1 Sus brachgnathus 1 1 1 1 1 1 ? 1 Sus cf. barbatus 1 Sus cf. lydekkeri 1 1 1 1 1 Sus cf. officinalis 1 Sus cf. scofa 1 Sus macrognathus 1 1 Sus scofa 1 1 1 1 1 1 1 1 1 1 1 Sus sp. 1 1 1 1 1 1 1 Sus sp. 1 1

288 Sus sp. 2 1 Sus stremmi 1 Sus verrucosus 1 Tragulus javanicus 1 Tragulus napu 1

PHOLIDOTA Manis palaeojavanica 1 1 Manis javanica 1 1

INSECTIVORA Crocidura fuliginosa 1 Hylomys suillus 1 Tupaia minor 1

RODENTIA Atherurus macrourus 1 1 Atherurus sp. 1 1 1 ? Bandicota savilei 1 Callosciurus finalysoni 1 Cannomys badicus 1 Chirmyscus chiropus 1 Chiropodomys gliroides 1 1 Hapalomys delacouri 1 Hylopetes spadiceus 1 Hystrix brachyura 1 1 1 1 1 1 1 1 1 Hystrix javanica 1 Hystrix sp. 1 1 1 1 Hystrix subcristata 1 1 1 1 1 1 1 1 1 Leopoldamys sabanus 1 Maxomys niviventer 1

289 Maxomys sp. 1 Maxomys surifer 1 Menetes berdmorei 1 Mus caroli 1 Mus pahari 1 Mus shortridgei 1 Mus sp. 1 Niviventer confuciacus 1 Petaurista petaurista 1 Rattus aff. Tiomanicus 1 Rattus argentiviventer 1 Rattus koratensis 1 Rattus rattus 1 Rattus sp. 1 1 1 1 Rattus sp. A 1 Rattus sp. B 1 Rattus sp. C 1 Rattus trinilensis 1 Ratufa affinis 1 Rheithrosciurus macrotis 1 Rhizomys sp. 1 Rhizomys troglodytes 1 1 1 1 1 Sundasciurus lowii 1 Vandeleuria oleracea 1

290 APPENDIX C

West, J. A., Louys, J., 2007. Differentiating bamboo from stone tool cut marks in the

zooarchaeological record, with a discussion on the use of bamboo knives. Journal

of Archaeological Science 34: 512-518. APPENDIX D

Louys, J., Black, K., Archer, M., Hand, S. & Godthelp, H., 2007. Descriptions of koala

material from the Miocene of Riversleigh, northwestern Queensland and its

implications for Litokoala (Marsupialia, Phascolarctidae). Alcheringa 31: 99-110. APPENDIX E

Archer, M., Arena, D. A., Bassarova, M., Beck, R., Black, K., Boles, W.E., Brewer, P.,

Cooke, B.N., Crosby, K., Gillespie, A., Godthelp, H., Hand, S.J., Kear, B., Louys,

J, Morrell, A., Muirhead, J., Roberts, K. K., Scanlon, J. D., Travouillon, K.T, and

Wroe, S., 2006. Current status of species-level representation in faunas from

selected fossil localities in the Riversleigh World Heritage Area, northwestern

Queensland. Alcheringa Special Issue 1: 1-17.