Phd Thesis, University of Wollongong

Home , Amalosia, Cape rock gecko, Carpet chameleon, Carphodactylidae, Christinus, Cnemaspis phillipsi

Comparative phylogeography and diversity of

Australian Monsoonal Tropics lizards

Rebecca Jan Laver

ORCID ID 0000-0002-6319-7213

Doctor of Philosophy

January 2017

The School of BioSciences

Faculty of Science

The University of Melbourne

Submitted in total fulfilment of the requirements of the degree of Doctor of Philosophy

Produced on archival quality paper

Thesis Abstract

Tropical savannah biomes cover ~20% of the world’s landmass, however the biodiversity encompassed within these environments and the underlying processes that have shaped it remain poorly understood. Recent increased research to address this knowledge gap have begun to reveal surprisingly high amounts of deep, geographically- structured diversity, much of which is cryptic or hidden within morphologically similar species complexes. These patterns are especially emphasized in vertebrate taxa which are intrinsically linked to rock escarpments and ranges that dissect the savannah woodlands and grasslands of many of these biomes, hinting at a role of heterogeneous topography in structuring diversity.

The remote Australian Monsoonal Tropics (AMT) spanning the north of the

Australian continent is a particularly vast, and relatively undisturbed, tropical savannah region. Recent increased surveys are revealing numerous new species and endemism hotspots, indicating we are only just beginning to uncover the true biodiversity levels within this biome. Not only is there a relative paucity of knowledge regarding the present diversity within this region, but there is also limited understanding of how this diversity came to be. Phylogeographic studies can assist us in establishing current patterns of diversity and their evolutionary significance within regions and biomes.

Furthermore, by comparing and contrasting the patterns and timing of diversification within and between biomes for multiple ecologically diverse taxa, we can begin to elucidate the history of these biomes and the environmental processes that have shaped the diversity we observe today.

In this dissertation I aimed to better assess and establish true patterns of biodiversity and endemism within the Kimberley region of the AMT (Western Australia), and to place these patterns within a broader continental context using intra- and inter-biome comparisons in related taxa. Using geckos as a model system I took a comparative phylogeographic approach, integrating advanced next-generation genetics and morphology to establish patterns and timing of diversification across ecologically variable taxa. Within all Kimberley taxa I studied, I uncovered high levels of cryptic diversity. Much of this diversity involves especially short-range endemic lineages concentrated in key regions typically with one or more of the following factors: highly mesic conditions, island or insular environments, and unique or complex geological formations. In recognising these areas I have provided evidence of novel biodiversity hotspots and emphasised the significance of others as representing important “refugia” within the Kimberley that allow persistence and facilitate divergence of lineages through harsh periods of environmental change. These findings indicate diversification patterns are shaped by complex interactions of climatic variation, topography, and species’ ecology, allowing inference of biogeographic history and a greater ability to predict impacts of future environmental change.

iii

Declaration

This is to certify that:

This thesis is my own original work towards the PhD except where otherwise indicated in the Preface. Due acknowledgment has been made in the text to all other material used. The thesis is fewer than 100,000 words in length, exclusive of tables, maps, bibliographies and appendices.

______

Rebecca Jan Laver

The School of BioSciences

The University of Melbourne

Victoria, 3010

Australia

September, 2016

Preface

To facilitate publication, two of the chapters within this dissertation (Chapters 2 and

3) are presented as they were prepared for stand-alone scientific articles. For these manuscripts which are in preparation for publication I am the primary author as follows:

Chapter 2:

Laver, RJ, Nielsen, SV, Rosauer, DF, & Oliver, PM. Trans-biome diversity in

Australian grass-specialist lizards (Diplodactylidae: Strophurus). In review.

Chapter 3:

Laver, RJ, Doughty, P & Oliver, PM. Endemism, persistence and introgression across

an aridity gradient in two specialised lizard lineages from the Australian Monsoonal

Tropics (Oedura spp.). In review.

For these chapters and the remainder of this dissertation I i) contributed to conceptualising research projects; ii) conducted majority of the data collection – including collection of specimens, sequencing and assembly of molecular data, and collection of morphological data; iii) conducted majority of the data analyses; iv) conducted the literature review; v) wrote all manuscripts, and prepared figures and tables. My primary supervisor, Paul Oliver, served as co-author for Chapters 2 and 3, and is credited for his contribution to conceptualising and funding the projects covered in this dissertation, as well as assistance with interpretation of results and editorial comments.

Additional co-authors for the above mentioned manuscripts contributed the following:

Stuart Nielsen (University of Mississippi) – contributed to molecular data collection,

data analysis, and provided feedback on draft revisions (Chapter 2).

Dan Rosauer (Australian National University) – contributed to data analysis, and

provided feedback on the manuscript (Chapter 2).

Paul Doughty (Western Australia Museum) – contributed to field work, tissue

collection, and data collection of morphometric measures, and assisted with revisions

of the manuscript (Chapter 3).

Though presented in the form of a thesis chapter for this dissertation Chapter 4 was also research conducted as part of a broader collaboration. Contributions of collaborators for this work are detailed as follows:

Renae Pratt (Australian National University) – contributed to molecular laboratory work

and data collection (next-generation sequencing – exon capture; Chapter 4).

Craig Moritz (Australian National University) – contributed to funding, fieldwork,

tissue collection, and draft revisions (Chapters 2, 3 & 4).

All other assistances to data collection, fieldwork, and analytics are detailed in the

Acknowledgements.

The research detailed in this dissertation was conducted as part of a broader study on lizard diversity in the northern Australian Monsoonal Tropics, and was supported and funded by a linkage grant from the Australian Research Council (ARC) awarded to Paul

Oliver, Michael Lee and Paul Doughty, a McKenzie Postdoctoral fellowship to Paul

Oliver from the University of Melbourne, and a Discovery Early Career Researcher

Award (DECRA) to Paul Oliver. In addition, funding from an Australian Laureate

Fellowship from the ARC awarded to Craig Moritz assisted with fieldwork, tissue sampling, and next-generation sequencing. Sampling for this project was covered under a Western Australian license to take fauna, permit number SF009863, issued to Craig

Moritz by the Department of Parks and Wildlife, WA.

vii

Acknowledgements

Firstly, I would like to thank my supervisors: Tim Jessop, Jane Melville and Paul

Oliver, for providing the opportunity to work on this amazing project and supporting me throughout to ensure I managed to finish it. I greatly appreciated Tim always being available to provide support, advice, guidance, and an ear to listen whenever it was needed. Jane ensured a fantastic start to my PhD by ensuring I was welcomed into the science community at Museum Victoria, and provided access to an office and laboratory facilities. This experience helped me make many important connections and afforded several great educational opportunities which greatly enhanced my student experience and gave me a real appreciation for working in a museum environment. Paul I thank for funding and conceiving the project, providing the chance to conduct fieldwork in some spectacular landscapes and with awesome critters, and enabling me to attend an international conference and visit universities and colleagues overseas. Thank you to the members of my advisory committee, Devi Stuart-Fox and Michael Keough, for keeping track of my progress and being accommodating of delayed reviews and paperwork.

In addition I am so grateful to the many people who perused drafts and provided valuable feedback to enhance my written work; in particular Craig Moritz, Richard

Harrison, Dominique Potvin, Paul Doughty, and Renee Catullo. Special mention to the late Richard Harrison whose encouragement and guidance re-ignited the passion for my work during the final trying stretch of writing up.

Thank you to the many colleagues and collaborators who provided guidance or assisted with various analyses or data collection, greatly enhancing the quality of the work within this thesis. To my co-authors Stuart Nielsen, Paul Doughty, and Dan

viii

Rosauer, thank you so much for additional analyses and data collection that greatly enhanced the studies covered in Chapters 2 and 3. To my amazing and endlessly patient teachers, Katie Smith and Renae Pratt, thank you for ensuring my competence in the genetics laboratory from traditional Sanger sequencing through to next-generation methods, and for spending many, many hours contributing to collection of molecular data. Thanks to Sumitha Hunjan, Maggie Haines and Bee Gunn, for conducting additional sequencing to assist in completing genetic datasets and ensuring final troublesome loci were included. Thanks to Mozes Blom, Jason Bragg, and Luisa

Teasdale for assistance and guidance with the frighteningly complex world of bioinformatics, and for creating fantastic pipelines crucial for dealing with the otherwise overwhelming exon-capture datasets. Thanks to Paul Doughty and Ryan Ellis for assisting with collection of morphometric datasets; I greatly appreciate the hours spent measuring and trying to distinguish even the tiniest of differences in outwardly identical geckos. Thanks to Ana Catarina Silva, Marta Vidal Garcia, and Emma Sherratt for guidance and assistance with choosing the best methods for analysing various molecular and morphometric datasets, and then answering my many questions when I struggled to interpret results. Thank you to the amazing photographers whose fantastic snaps of some fabulous gecko models I simply had to include in figures or title pages: Henry

Cook, Jordan de Jong, Ryan Francis, Rich Glor, Stephen Richards, Brendan Schembri,

Jordan Vos, and Stephen Zozaya.

Thanks to everyone I was privileged enough to work with in the field – be it for my own work, or a chance to escape and enjoy a break with an alternate taxonomic adventure – this list is long yet includes various members of the Australian Wildlife

Conservancy, the Moritz Lab at ANU, the Museums of Western Australian (WAM) and

ix the Northern Territory (MAGNT), and the Bioscan team from Museum Victoria.

Important thanks to the Traditional Owners of Country for access to the land upon which much of my fieldwork in the Kimberley and Top-End was conducted and the invaluable specimens used in my studies were collected. Thanks to Paul Doughty, Craig

Moritz, and Paul Oliver for allowing me to conduct fieldwork under your various collection permits; this made my life in the field a lot easier and I in no way envy having to have sorted out all that paperwork. Special mention to Gaye Bourke, field manager extraordinaire, who organised absolutely everything for field surveys with the

Moritz Lab, made fieldwork extremely smooth sailing, and always maintained the stocks of chocolate, coke and beer. Also to Paul Doughty who was so patient and took time to teach me about the various taxa and the process of specimen preparation during my first trip to the Kimberley; as a developing scientist I appreciated you sharing your knowledge and expertise immensely. Thanks to the museums MAGNT, MV, QMJ,

SAMA, and WAM, and to their collections managers for access to voucher specimens and tissues without which this work simply would not have been possible. Thank you to all the geckos who dedicated their lives to the betterment of scientific knowledge; your contributions were invaluable and will live on in your honour.

Having moved around different institutions throughout the course of my PhD I cannot emphasise enough how much it meant to feel welcomed and that I had a place at each institution, which made adjusting and continuing to be productive a lot easier.

Thanks to Joanna Sumner for providing me with access to desk space and the genetics facilities at Museum Victoria, and for your patience and kind assistance whenever I had questions about the general running of the lab. Thanks to Scott Keogh for being so welcoming when I joined the Ecology, Evolution and Genetics department at ANU. I

x greatly appreciated you always offering advice, to look over drafts or job applications, or simply someone to talk to. Your time and support meant a great deal to me during the most difficult stages of my PhD; thank you for your kindness. Huge thanks to the many members of the Moritz Lab at ANU. You took me in as one of your own from the very beginning; I felt so welcome and lucky to be part of a pretty special community. Thank you all for collecting additional gecko samples for me over the many fieldtrips to the

Monsoonal Tropics and for the memories shared in the field, or at lab functions and

BBQs at Craig’s farm. Thank you for so much useful feedback, advice and support throughout my PhD, and for many stimulating discussions at lab meetings or over beers.

Special thanks to Renae Pratt who really built my confidence in the genetics lab during the years spent fighting next-gen methods, and also ensured I felt at home both in the lab and in Canberra. In particular, incredible thanks to Craig Moritz for taking me under your wing, including me as part of your extended science-family, and providing me with so many amazing opportunities throughout my PhD. I cannot express how grateful

I am that you were so gracious in extending your funding and resources to assist in furthering my research and allowing me to greatly enhance the quality of my project and future career prospects with the training in next-gen techniques. Thank you for all the advice, support, and mentorship, and for all the time you took assisting with thesis drafts.

I would like to acknowledge the funding provided by various sources to support my work throughout this PhD project. This included funding from various grants awarded to my supervisor, Paul Oliver; namely, a Linkage Grant from the Australian Research

Council, a McKenzie Postdoctoral Fellowship from the University of Melbourne, and a

Discovery Early Career Researcher Award. Additional funding was generously

xi provided by Craig Moritz, from his Australian Laureate Fellowship awarded by the

Australian Research Council. All sources enabled various field trips to the Kimberley and Top-End for specimen collection, and assembly of molecular datasets via both traditional Sanger sequencing and next-generation sequencing exon-capture methods. In addition, for financial support which gave me the chance to attend an international conference and visit various universities in North America, I am very grateful to the

University of Melbourne for several bursaries (Melbourne Abroad Travelling

Scholarship, Department of Zoology Research Higher Degree Student Travel Support, and the Drummond Travel Award), and also to the Australian Government’s Australian

Biological Resources Study (ABRS) National Taxonomy Research Grant Program

(NTRGP) for an ABRS Student Travel Grant. In addition, I cannot thank enough the administration officers at my various institutions who assisted with any issues that may have arisen, provided access to facilities and equipment of various kinds, were always available to answer queries when I had no idea what forms I required, made organising field work and travel such a breeze, deciphered and traversed the world of finances so I could survive this far, and provided me with opportunities to communicate my science to the general public.

Thanks to various office-mates and department colleagues I have had over the past five or more years at Melbourne University, Museum Victoria, and the Australian

National University. I have enjoyed the company, extensive chats, coffee breaks, Friday drinks, and various celebrations throughout the years. There were always great people around to bounce ideas around or sympathise with technology-/analysis-related frustrations. To all my academic friends – thank you for the great memories, the conversations, the ideas, for sharing your research journeys with me and alongside

xii mine. To the friendly coffee staff at my different institutions – thank you for providing stimulants, for the cheerful smiles when days were tough, for genuine friendships outside of the academic bubble, or for simply remembering my order; you all helped me make it through this most challenging chapter of university to date.

Finally, so many thank yous to my family for your support and understanding. For being there to listen or comfort me when I needed it most, and for financial aid (and gifts of socks sans holes) to see me through when scholarships and savings dried up.

Thank you all for being proud of my achievements and my work when I was so worn down I struggled to feel that myself. Special mention to my best friends and loved ones

Amelia Brennan, Ryan Tessier, and Dominique Potvin. Your constant support, encouragement and care ensured I made it through this. You were always there to listen, comfort, distract, and share laughs. You each lit the way for me to find the end of the

PhD tunnel. To anyone I may have forgot to mention, a sincere apology, but know how much I appreciated the help of any and all who contributed to me being able to make it through. This could not have been done without the help of so many.

xiii

Table of Contents

Title page ...... i

Abstract ...... ii

Declaration ...... iv

Preface ...... v

Acknowledgements ...... viii

Table of contents ...... xiv

List of tables ...... xix

List of figures ...... xxi

Chapter 1: General introduction ...... 1

1.1. Diversity as a foundation of biology ...... 2

1.2. Phylogeography ...... 4

1.3. Tropical savannah biomes ...... 5

1.4. The Kimberley ...... 7

1.5. Geckos as a model system ...... 9

1.6. Thesis overview ...... 10

1.6.1. Aims & approach ...... 10

1.6.2. Broad methods of data collection ...... 13

1.6.2.1. Fieldwork ...... 13

1.6.2.2. Museum collections ...... 14

1.6.2.3. Genetics ...... 15

1.6.3. Synthesis ...... 16

xiv

Figures ...... 17

References ...... 18

Chapter 2: The role of biome ecology in shaping diversification patterns in a clade of Australian spinifex-dwelling lizards (Strophurus) ...... 28

Abstract ...... 29

2.1. Introduction ...... 30

2.2. Materials & methods ...... 34

2.2.1. Sampling ...... 34

2.2.2. Nucleotide data and phylogenetic analyses ...... 35

2.2.3. Phylogenetic dating ...... 38

2.2.4. Distribution modelling ...... 39

2.3. Results ...... 40

2.3.1. Phylogenetic relationships and lineage diversity ...... 40

2.3.2. Divergence dates ...... 42

2.3.3. Distribution modelling ...... 42

2.4. Discussion ...... 43

2.4.1. Contrasting lineage diversity in neighbouring biomes ...... 44

2.4.2. Definition and evolutionary history of biomes ...... 46

2.4.3. Conclusions ...... 48

Tables & Figures ...... 49

References ...... 54

Appendix ...... 69

Chapter 3: The contribution of climate and topography to diversification patterns in two clades of Kimberley geckos (Oedura spp.) ...... 84

Abstract ...... 85

3.1. Introduction ...... 86

3.2. Materials & methods ...... 90

3.2.1. Sampling ...... 90

3.2.2. Nucleotide data and phylogenetic analyses ...... 91

3.2.3. Phylogenetic dating ...... 93

3.2.4. Establishing nominal evolutionarily distinct lineages ...... 94

3.2.5. Population structure ...... 95

3.2.6. Coalescent-based analyses ...... 95

3.2.7. Phenotypic analyses ...... 97

3.3. Results ...... 98

3.3.1. Phylogenetic relationships and divergence times ...... 98

3.3.2. Mitochondrial diversity ...... 98

3.3.3. Phylogenetics of nuclear loci ...... 100

3.3.4. Population structure – clustering and coalescent methods ...... 101

3.3.5. Phenotypic variation ...... 101

3.4. Discussion ...... 102

3.4.1. Emphasising the evolutionary distinctiveness and endemism of the central-

west Kimberley ...... 103

3.4.2. Persistence and divergence in isolated ranges of the semi-arid southern

Kimberley ...... 104

xvi

3.4.3. Introgression and delimitation breakdown at the arid-monsoon biome

interface ...... 106

3.4.4. Discovery of cryptic diversity within a tropical savannah biome ...... 107

Tables & Figures ...... 110

References ...... 120

Appendix ...... 136

Chapter 4: Continental islands support ancient genetic diversity within gekkonid lineages (Gehyra spp.) in the Australian Monsoonal Tropics ...... 153

Abstract ...... 154

4.1. Introduction ...... 155

4.2. Materials & methods ...... 162

4.2.1. Sampling and mitochondrial sequencing ...... 162

4.2.2. Exon capture ...... 163

4.2.2.1. Library design ...... 163

4.2.2.2. In solution capture ...... 163

4.2.2.3. Bioinformatics pipeline ...... 166

4.2.3. Preliminary phylogenetics ...... 167

4.2.4. Phylogenetic dating ...... 168

4.2.5. Nuclear assessment of divergent mitochondrial lineages ...... 169

4.2.6. Morphometric comparisons ...... 171

4.3. Results ...... 172

4.3.1. Mitochondrial diversity ...... 172

4.3.2. Timeframes of diversification ...... 173

xvii

4.3.3. Nuclear recovery and relationships of mitochondrial lineages ...... 174

4.3.4. Further nuclear support for distinctness of lineages ...... 175

4.3.5. Morphometrics ...... 176

4.4. Discussion ...... 177

4.4.1. Old endemic lineages in land-bridge islands ...... 178

4.4.2. Contrasting diversification responses to environmental change ...... 180

4.4.3. Continental islands as evolutionary refugia ...... 183

4.4.4. Conclusion ...... 185

Tables & Figures ...... 187

References ...... 195

Appendix ...... 211

Chapter 5: General discussion ...... 227

5.1. Key findings ...... 229

5.1.1. High diversity ...... 229

5.1.2. Scaling of lineage depth and range-size with aridity ...... 230

5.1.3. Refugia ...... 230

5.1.4. The role of ecology in environmentally-driven diversification ...... 231

5.2. Future considerations ...... 232

5.2.1. Introgression ...... 232

5.2.2. Additional trait data ...... 233

5.3. Theoretical directions ...... 233

5.4. Summary ...... 234

Figures ...... 235

References ...... 236

xviii

List of Tables

Chapter 2

Table 2.1. Environmental predictors and their contribution to species distributions

models for the major clades of phasmid Strophurus ...... 49

Table 2.2. Prior settings and crown-age estimates from phylogenetic dating analyses

of diversification within phasmid Strophurus ...... 50

Appendix

Table A2.1. Limbed gekkotans exhibiting thin, contrasting longitudinal striped

patterning ...... 69

Table A2.2. Details of specimens used in molecular and phylogenetic analyses ..... 70

Table A2.3. Characteristics of the loci sequenced for molecular and phylogenetic

analyses ...... 72

Table A2.4. Details of primers and protocols for molecular data assembly ...... 73

Table A2.5. Optimal partitioning and model schemes for each dataset and

phylogenetic analysis ...... 74

Table A2.6. Mitochondrial genetic diversity and population structure within phasmid

Strophurus ...... 75

Chapter 3

Table 3.1. Prior settings and crown-age estimates from phylogenetic dating analyses

of Kimberley Oedura and allied taxa ...... 110

Table 3.2. Genetic diversity data for major lineages of Oedura gracilis and allied

taxa ...... 111

xix

Table 3.3. Estimated geographic range sizes for mitochondrial lineages of Kimberley

Oedura and allied taxa ...... 112

Appendix

Table A3.1. Details of specimens used in molecular and phylogenetic analyses ... 136

Table A3.2. Characteristics of genetic data sequenced for Oedura gracilis ...... 139

Table A3.3. Details of primers and protocols for molecular data assembly ...... 140

Table A3.4. Optimal partitioning and model schemes for each dataset and

phylogenetic analysis ...... 141

Table A3.5. Morphometric variables measured for phenotypic analyses ...... 142

Table A3.6. Phenotypic measures for Oedura gracilis specimens ...... 143

Chapter 4

Table 4.1. Mitochondrial genetic diversity and population structure within the clades

of the Gehrya occidentalis complex and G. xenopus ...... 187

Table 4.2. Prior settings and crown-age estimates Gehyra phylogenetic dating .... 188

Appendix

Table A4.1. Details of specimens used in molecular and phylogenetic analyses ... 211

Table A4.2. Details of primers and protocols for molecular data assembly ...... 217

Table A4.3. Optimal partitioning and model schemes for datasets and phylogenetic

analyses ...... 218

Table A4.4. Morphometric variables measured for phenotypic analyses ...... 219

Table A4.5. Phenotypic measures for Gehyra specimens ...... 220

List of Figures

Chapter 1

Figure 1.1. Maps of Australian regions integral to this dissertation; including

pertinent topographic and climatic features of the Kimberley, and locations of key

Kimberley islands ...... 17

Chapter 2

Figure 2.1. Distribution map of phasmid Strophurus species records ...... 51

Figure 2.2. Phylogenetic networks for phasmid Strophurus species ...... 52

Figure 2.3. Chronogram of divergence dates for lineages of phasmid Strophurus and

congeners ...... 53

Appendix

Figure A2.1. Summary of the phylogenetic distribution of graminicolous (grass-

dwelling) ecology and associated specialised morphology ...... 76

Figure A2.2. Phylogenetic reconstruction of nuclear concatenated (three loci) dataset

for Strophurus and outgroups ...... 77

Figure A2.3. Phylogenetic reconstruction of mitochondrial and nuclear concatenated

(four loci) dataset for Strophurus and outgroups ...... 78

Figure A2.4. Phylogenetic reconstruction of single mitochondrial locus (ND2) for

Strophurus and outgroups ...... 79

Figure A2.5. Phylogenetic reconstruction of single nuclear locus (PDC) for

Strophurus and outgroups ...... 80

Figure A2.6. Phylogenetic reconstruction of single nuclear locus (PRLR) for

Strophurus and outgroups ...... 81

xxi

Figure A2.7. Phylogenetic reconstruction of single nuclear locus (RAG1) for

Strophurus and outgroups ...... 82

Figure A2.8. Species distribution model (SDM) maps for the three major phasmid

Strophurus lineages ...... 83

Chapter 3

Figure 3.1. Distribution maps for mitochondrial lineages of Kimberley Oedura and

allied taxa ...... 114

Figure 3.2. Mitochondrial and nuclear phylogenies for Oedura gracilis and allied

taxa ...... 116

Figure 3.3. Distribution map of diversity within Oedura gracilis and summary

phylogeny indicating delimitation results for various analyses ...... 117

Figure 3.4. PCAs and box-and-whisker plots indicating phenotypic variation in

Oedura gracilis ...... 119

Appendix

Figure A3.1. GMYC species delimitation across Oedura gracilis ...... 144

Figure A3.2. Phylogenetic reconstruction of single nuclear locus (K1AA1549) for

Oedura gracilis and outgroups ...... 145

Figure A3.3. Phylogenetic reconstruction of single nuclear locus (KIF24) for Oedura

gracilis and outgroups ...... 146

Figure A3.4. Phylogenetic reconstruction of single nuclear locus (PDC) for Oedura

gracilis and outgroups ...... 147

Figure A3.5. Phylogenetic reconstruction of single nuclear locus (PRLR) for Oedura

gracilis and outgroups ...... 148

xxii

Figure A3.6. Phylogenetic reconstruction of single nuclear locus (PTPN12) for

Oedura gracilis and outgroups ...... 149

Figure A3.7. Phylogenetic reconstruction of single nuclear locus (RAG1) for Oedura

gracilis and outgroups ...... 150

Figure A3.8. Phylogenetic reconstruction of mitochondrial and nuclear concatenated

(seven loci) dataset for Oedura gracilis and outgroups ...... 151

Figure A3.9. PCA result of multidimensional scaling analysis of nuclear data for

Oedura gracilis ...... 152

Chapter 4

Figure 4.1. Maps of the distribution of the Gehyra focal clades and rainfall isohyets

within the Kimberley ...... 189

Figure 4.2. Distribution maps of genetic sampling of Gehyra clades ...... 190

Figure 4.3. Mitochondrial phylogenies for the two clades of Gehyra ...... 191

Figure 4.4. Phylogenies of nuclear (exon-capture) data ...... 192

Figure 4.5. PCAs of morphometric analyses for two Gehyra clades ...... 193

Figure 4.6. Map of Kimberley continental islands ...... 194

Appendix

Figure A4.1. Morphological (body-size) and pre-cloacal pore count variability across

Gehyra lineages ...... 226

Chapter 5

Figure 5.1. Map of key biodiversity hotspots within the Kimberley ...... 235

xxiii

Chapter 1: General introduction and thesis overview

General introduction and thesis overview

Chapter 1: General introduction and thesis overview

1.1. Diversity as a foundation of biology

The recognition and classification of diversity is one of the foundations of biological research and understanding. The Greek philosopher Aristotle in the 4th century BCE established one of the first systems for basic division of organisms into distinct groups, including the use of the terms “genus” and “species” (Mayr et al. 1963). The first biological definition of “species” was provided by English naturalist John Ray in the

17th century, and in the 18th century Linnaeus devised the binomial nomenclature system for taxonomic classification used today (Wilkins 2009). Scientists since have aimed to accurately classify elements across the tree of life, however to do this effectively requires a definitive concept for identifying “species”, which is actually somewhat problematic. A plethora of different species concepts and definitions have been proposed to date, and much controversy exists as to which ultimate delimitation method should be used (Wilkins 2011). Traditional taxonomy was built primarily upon the ability to identify species using phenotypic traits such as variable morphology or colour pattern. However, phenotypic variation within species can be extreme, particularly if certain traits vary with changing environments across a species’ geographic range. Conversely, some species complexes exhibit highly conserved morphology between species, usually a result of limited exertion of selective pressure upon body-form when populations occupy similar habitat niches yet are otherwise isolated from one another. Both examples illustrate a need for a taxonomic approach which considers more than basic morphological identification alone (de Queiroz 1998,

2007; Packer et al. 2009).

Chapter 1: General introduction and thesis overview

Speciation is rarely clear-cut because it is a process rather than an event. At any time throughout that process different characteristics may arise that provide a means to distinguish lineages as distinct, however the timing or order in which the characters evolve can vary. Taxonomists have a wealth of information to draw upon for making decisions about species distinctions: most recently, technological advances and the ability to incorporate information from genetic material has been an important element of this process. It therefore becomes a necessity for taxonomists to adopt an integrative approach (de Queiroz 2007; Miralles & Vences 2013; Oliver et al. 2014; Padial et al.

2010), implementing multiple independent sources of evidence to infer the evolutionary independence of entities in question. Early work included the study of allozymes— variation in enzymes coded for by different alleles of the same locus—and was improved upon with sequencing of microsatellites and single loci from fast evolving organellar genomes (chloroplasts and mitochondria; Avise 2012). Such datasets not only provided greater resolution of distinctly evolving units, but additionally began to reveal more about diversification histories of populations. Further advancements in the ability to assemble small multi-locus nuclear genetic datasets, however, have uncovered complicated patterns within these histories. Different molecular markers may have varied modes of inheritance or rates of evolution, and throughout (and even post-) the speciation process individual genes may diverge along quite disparate trajectories compared to the “true” species tree. In addition, selection, recent or rapid diversification events, incomplete lineage sorting, and varying forms of hybridisation are just a few examples of situations where species delimitation can become highly challenging due to inconsistent or discordant signatures across differently evolving molecular datasets (e.g.

Avise 2000, 2012; Wendel & Doyle 1998). Recent advances in the ability to gather

Chapter 1: General introduction and thesis overview

molecular datasets of increasing size and power via next-generation sequencing techniques hold the promise of streamlining the taxonomic process somewhat by improving accuracy in detecting truly independent evolutionary lineages in more complicated and non-model systems (McCormack et al. 2013). Independently in this instance does not imply strictly reproductively isolated, but that the lineage or population is evolving along an independent trajectory away from those of congeners.

Whilst the major focus of this thesis is on recognising and describing patterns of diversity rather than specifically defining species, the general species concepts considered when discussing nominal species herein include the evolutionary species concept (Wiley 1978) and the general lineage concept (de Queiroz 1998). The ability to accurately recognise and describe patterns of diversity is fundamental to understanding the evolutionary processes underpinning diversification and allows us to address questions such as why certain environmental regions accumulate higher diversity than others.

1.2. Phylogeography

First coined in 1987 by John Avise, phylogeography is the consideration of contemporary patterns of species- or population-level genetic diversity across a landscape, and the historical processes which may have led to the formation of such patterns (Avise 2000). Understanding of diversification patterns within an environment is best achieved via comparative phylogeography, combining information of multiple taxa which occupy varied distributions (Bermingham & Moritz 1998). Elucidating the taxonomic or evolutionary relevance of diversity is a key element of accurate phylogeographic inference. Establishing comprehensive datasets including genotypic,

Chapter 1: General introduction and thesis overview

phenotypic, and physiological information for taxa with varying ecologies provides a good foundation for being able to address not only the relevance of the diversity observed, but in turn the processes which have generated it (Avise 2000). Much phylogeographic work to date has focussed on northern hemisphere environments subjected to extreme impacts of historical glacier formation, known to have significantly impacted distributions of taxa (Beheregaray 2008). Similarly, biomes recognised for high levels of biodiversity, such as rainforests, receive a great deal of interest in terms of why certain regions may be considered so productive (e.g. Martin et al. 2012). Comparatively less attention has been afforded to tropical savannah and grassy biomes, although increasing work in such regions are revealing uniquely associated floral and faunal assemblages, and significant underestimation of biodiversity and endemism (Carnaval et al. 2014; Martin et al. 2012; Parr et al. 2014;

Pennington et al. 2006).

1.3. Tropical savannah biomes

Tropical savannah woodlands and grasslands occur naturally over ~20% of the world’s land surface and support a unique assemblage of associated and often specialised flora and fauna (Bond & Parr 2010; Grace et al. 2006). Yet these grassy biomes have long been underappreciated due to misconception they were the result of anthropogenic alteration of other habitat types (Bond & Parr 2010; Parr et al. 2014).

Savannah environments are under significant threat from impacts of agriculture, development, changing fire regimes, and introduction of pest species; hence it is particularly important that the true biodiversity of these regions is recognised and understood (Miles et al. 2006; Waeber et al. 2015). The increasingly recognised

Chapter 1: General introduction and thesis overview

diversity within tropical savannahs tends to be highly geographically structured, and especially in rock-specialised (saxicoline) taxa, where diversity is linked to topographically heterogeneous regions of the landscape (e.g. Lorenzen et al. 2012;

Nogueira et al. 2011; Werneck et al. 2012). In addition, much of this diversity tends to be cryptic (morphologically conserved), and is therefore difficult for us to initially identify (e.g. Domingos et al. 2014; Moritz et al. 2015). This phenomenon emphasises the need for increased survey effort in these often remote regions, along with the assembly of larger datasets to accurately assess the relevance of the diversity being uncovered.

The Australian Monsoonal Tropics (AMT) is a vast tropical savannah spanning the top of the Australian continent (Fig. 1a). The region is strongly characterised by its summer monsoon climate and consists of woodland and grassland habitats, intersected with rock escarpments and plateaux (Bowman et al. 2010; Woinarski et al. 2007). The

AMT comprises three major regions (Fig. 1a)—the Kimberley (Western Australia

[WA]), Top End and Gulf region (Northern Territory [NT]), and Cape York

(Queensland)—each of which is remote, and relatively understudied compared to biomes such as the Australian Wet Tropics (AWT; Moritz et al. 2015). There is, however, a growing body of work suggesting the AMT supports high levels of biodiversity and particular hotspots of localised endemism (e.g. Catullo et al. 2014;

Criscione et al. 2012; Doughty 2011; Smith et al. 2011). In addition, there is a clear relationship between many taxa of the AMT and the Australian Arid Zone (AAZ) biome located to the south (Fig. 1a; Byrne et al. 2011; Byrne et al. 2008). Taxa of the AAZ have long been considered derived from AMT ancestors, are often younger and less diverse – characteristics attributed to the comparatively younger age of the present-day

Chapter 1: General introduction and thesis overview

deserts (Byrne et al. 2008; Fujita et al. 2010). However, recent studies indicate at least some elements of the AAZ biodiversity are significantly older, and there may have been multiple transitions of taxa between these biomes in both directions (e.g. Catullo &

Keogh 2014; Oliver et al. 2014a; Toon et al. 2012; Toon et al. 2015).

1.4. The Kimberley

The Kimberley region (Western Australia) is a particularly remote and topographically complex region of the AMT (Fig. 1b; Pepper & Keogh 2014; Woinarski et al. 2007). In contrast to its stable geological history, the region has endured a dynamic climatic history. The environment is presumed to have been significantly impacted by extreme drying during the process of aridification and formation of the central Australian deserts (Fujioka et al. 2009; Nix 1982). This process would have resulted in a transition from a warm-wet to cool-dry climate, in turn causing habitat changes within the Kimberley (Proske et al. 2014; Reeves et al. 2013). Evidence suggests during the last 40–20 kya savannah woodlands shifted north and further along the continental shelf during periods of low sea-levels to be replaced by drier grassland habitats; resulting in displacement of the AMT/AAZ interface over a north-south latitudinal gradient (Denniston et al. 2013; Frawley & O'Connor 2010; Reeves et al.

2013; Woinarski et al. 2007). Pleistocene glacial cycles would not only have caused periodic drying of the Kimberley region, but also dramatically altered the area of exposed mainland via changing sea-levels (Proske et al. 2014). The west Kimberley coastal region comprises over 2600 continental or land-bridge islands, typically consisting of large escarpments and exposed rock structures and located close to the mainland (generally <4km, Fig. 1c; Lyons et al. 2014; Palmer et al. 2013). Having been

Chapter 1: General introduction and thesis overview

cyclically dis- and re-connected to the mainland and each other throughout history, the most recent formation of these islands is estimated at only ~8 ka (De Deckker &

Yokoyama 2009; Nix & Kalma 1972). In addition to extensive historic climate variation, the Kimberley region presently experiences a strong rainfall and aridity gradient from the highly mesic central-west (~1500 mm/yr precipitation) towards the south-east (~500 mm/yr, Fig. 1d–e; Gibson & Köhler 2012; Palmer et al. 2013; Pepper

& Keogh 2014). The Kimberley region is already recognised broadly as a region of high diversity, but given the extreme climatic and environmental change, and the topographic complexity of the Kimberley landscape, it could be assumed that additional hotspots of biodiversity may exist within the region itself.

Due to the remoteness of the Kimberley the biodiversity of the region has remained poorly sampled until fairly recently, particularly in the central region (Moritz et al.

2013). Increasing surveys have revealed higher than expected levels of geographically- structured mtDNA diversity within the Kimberley, yet the taxonomic relevance of these diversity patterns is somewhat unclear (e.g. Oliver et al. 2010; Pepper et al. 2011b;

Potter et al. 2012). Furthermore, much of the detailed phylogeographic work within the

Kimberley so far has focussed on ecologically generalist or terrestrial taxa, which tend to have more extensive sampling (Catullo et al. 2014; Moritz et al. 2015; Potter et al.

2016). As a result of increased surveys there have been discoveries of new reptile and amphibian species, and evidence of localised endemism in more topographically complex regions such as the Mitchell Plateau in the mesic central-west (Doughty 2011;

Doughty et al. 2012; Köhler 2011). Many regions of the Kimberley, and particularly the west islands, are of increasing conservation concern due to threats from development, increased mining and tourism, altered fire regimes, and introduced pests such as the

Chapter 1: General introduction and thesis overview

cane toad (Carwardine et al. 2011; Collins 2008; Nias et al. 2010). It is likely historic climate and environmental variation significantly impacted taxa within the Kimberley, which could respond to these changes in a variety of ways; in particular, distributions of taxa may shift to track habitat, undergo extinction-re-colonisation processes, or fragment and persist in more suitable regions throughout the landscape. Improving our understanding of diversification responses of taxa to past environmental change can inform us of important regions for the preservation of biodiversity in the present and also enhance our ability to predict effects of future environmental change. A taxonomic system intrinsically connected to rock formations (Fig. 1f) provides a particularly valuable model to investigate phylogeographic patterns and diversification processes that shape them within heterogeneous landscapes such as the Kimberley.

1.5. Geckos as a model system

Gekkotans globally exhibit high levels of diversity and genetic structuring, and a propensity for evolution of cryptic complexes; particularly within tropical rainforest and savannah environments (e.g. Domingos et al. 2014; Gamble et al. 2012; Leaché &

Fujita 2010; Werneck et al. 2012). The highly speciose gekkotan community within

Australia is no exception with evidence of hidden diversity (e.g. Fujita et al. 2010;

Moritz et al. 2015; Oliver et al. 2010; Oliver et al. 2007; Pepper et al. 2011a). Within the savannah environments of the AMT geckos are especially prolific and relatively easy to sample in large quantities. In especially heterogeneous habitats many different gecko species can co-occur, hence geckos represent an ideal system for comparative phylogeographic studies in many different biomes.

Chapter 1: General introduction and thesis overview

Within Australia two separate major radiations of gekkotans constitute the predominant gecko fauna. The first, the pygopodoid geckos, has Gondwanan origins

(~60 Ma) and consists of three families: Carphodactylidae, Diplodactylidae and

Pygopodidae, with the majority of constituents being endemic to Australia (Oliver &

Sanders 2009). The many taxa spanning this radiation exhibit highly diverse morphological (e.g. size differences, digital and toe-pad variation, limb loss, etc.) and ecological variation (e.g. habitat specialisation – saxicolous, arboreal, fossorial, terrestrial, graminicolous), and inhabit a majority of the biomes within the Australian continent. The second radiation is a result of in situ diversification of taxa from a single genus, Gehrya, derived from a comparatively recent dispersal from Asia (<30 Ma) subsequent to the separation of Australia from Antarctica (Heinicke et al. 2011; Sistrom et al. 2009). Though taxonomically this radiation rivals the level of diversity of the

Australian pygopodoids, Gehyra are comparatively morphologically conserved, and have established complex community assemblages based on size- and ecological niche- partitioning within the environments they inhabit (Sistrom et al. 2013; Sistrom et al.

2012). Whilst there are other gekkotan lineages present in Australia (e.g. Christinus,

Hemidactylus, Lepidodactylus, Nactus) majority have not diversified like the pygopodoids or Gehyra (though see Heteronotia; e.g. Moritz et al., 2015).

1.6. Thesis overview

1.6.1. Aims & approach

The broad objective of this work was to understand how diversification patterns vary both within the complex tropical savannah landscape of the Kimberley, and compare these patterns to those of related regions and biomes. To do this I took a comparative

Chapter 1: General introduction and thesis overview

phylogeographic approach, integrating independent molecular (mitochondrial and nuclear loci) and morphological datasets, and used multiple analytic methods to estimate phylogenetic relationships and timeframes of diversification to compare the structure of diversity across ecologically variable taxonomic groups. Considering taxa with different ecologies or degrees of specialisation provided an opportunity to make predictions regarding the expected degree of genetic diversity or structure which might be observed across a landscape, depending on the level of connectivity different habitat- types (e.g. rock versus trees) may provide populations. I aimed to compare patterns of diversity to infer how climate and topology interact within an environment to shape diversification patterns, and how different ecologies may influence responses to these drivers. This work addresses the questions of why certain regions may promote or harbour higher diversity and whether such regions simply facilitate persistence or also drive in situ diversification.

To investigate and compare diversity patterns and infer the processes shaping them I considered three taxonomic groups with different origins, distributions and ecologies.

In Chapter 2 I focus on a lineage of grass-specialist (graminicolous) pygopodoid geckos from the genus Strophurus and consider variation in structure of diversity between biomes (AMT vs. AAZ). I use a small multi-locus dataset to investigate genetic diversity and relationships in the different regions of the AMT (Kimberley vs. Top

End), as well as the northern region of the AAZ. I estimate timeframes for diversification for the Strophurus geckos using multiple datasets, and also compare these results with recently estimated timeframes for the diversification of the Triodia spinifex grasses (Toon et al. 2015) to which they are adapted. Finally I implement

Chapter 1: General introduction and thesis overview

species distribution modelling to gauge whether the ecological variables which best correlate with the distribution of Strophurus diversity vary in different regions.

To contrast the role of ecology in diversification patterns, in Chapter 3 I look at phylogeographic patterns in two congeneric rock-specialist (saxicolous) lineages of the genus Oedura (also pygopodoids) which have contrasting distributions within the

Kimberley. For this study I assemble a larger multi-locus dataset and compare molecular diversity patterns between the two lineages. To taxonomically assess diversity within the widespread lineage I implement multiple species delimitation and ordination methods, and compare phenotypic variation using a suite of morphological measurements. Previous work investigating diversity across a closely related complex occupying much of continental Australia also allowed further comparison of genetic structuring between the Kimberley and related regions (Top End, Gulf region, and

AAZ; Oliver et al. 2014b).

Finally, to compare genetic and morphological diversification patterns with geckos from a distinct radiation (Gehyra), I conduct my final study (Chapter 4) on two broadly sympatric congeners from the mesic west-coast and Kimberley islands. The distributions of these lineages provide an opportunity to compare and contrast diversity patterns within a region recognised for high localised endemism (Doughty et al. 2012), and to investigate the role of the Kimberley islands in either maintaining or shaping diversity. This work implements next-generation sequencing methods, dramatically increasing the amount of genetic data available to confirm the validity of diverse lineages identified in preliminary mtDNA analyses. I use a number of phylogenomic and population genomic analysis methods to accurately assess phylogenetic relationships. Ecologically relevant morphological measurements are gathered to assess

Chapter 1: General introduction and thesis overview

the correlation of genetic and morphometric variability between the two lineages, and between insular and mainland populations.

1.6.2. Broad methods of data collection

1.6.2.1. Fieldwork

Two major field trips (March–April 2012 and May–June 2013) were conducted to assist in acquisition of voucher specimens and genetic material for this project, with surveys occurring across the southern Kimberley (WA) and western Top End (NT).

Specimens were collected predominantly using active search methods—fossicking through leaf litter and spinifex grasses, turning rocks and debris, including sheet metal and other items in station tip sites, and by spot-lighting for eye-shine at night. In addition, some surveys conducted with personnel from the Australian Wildlife

Conservancy (AWC) and Western Australian Museum (WAM) involved collection of specimens using pit-fall and funnel-trapping methods.

Processing of specimens collected included identifying sex where possible, measuring a suite of morphometric characters using callipers, recording details of discrete features such as counts of lamellae on finger and toe pads, and taking photographs. At any site, two to four individuals of a species were kept as voucher specimens to be accessioned into collections at various research institutions or museums, and genetic samples (tail tips) were collected for up to seven individuals.

Voucher specimens were euthanized using MS 222, and both liver and tail tissues were taken and stored in RNALater, and refrigerated in the field. Vouchers were then fixed using formalin, and later washed and stored in ethanol for return to museum collections.

Chapter 1: General introduction and thesis overview

Additional field surveys conducted by various members of the Moritz Lab at the

Australian National University (ANU), or researchers from the National Museum of

Victoria (NMV) targeted many other sites throughout the Kimberley and Top End periodically from 2012–2016, and collected extra specimens and tissue samples also incorporated in this project.

1.6.2.2. Museum collections

In addition to specimens collected in the field for this project, I made use of the voucher and tissue collections of several Australian museums, either personally visiting the collections or receiving loans to the institutions where I was studying. I took detailed morphological measurements using calipers, took counts of pre-cloacal pores and lamellae on digits, and compared scales and colour patterns for voucher specimens of several species during work to delimit cryptic species. Colleagues and collaborators assisted in measuring specimens from particular species complexes studied in Chapters

3 and 4.

For molecular analyses, sub-samples of specimen tissues were acquired from several

Australian museums (see Acknowledgements), for extraction of genetic material.

Additional molecular data for outgroup taxa, both within and outside Australia, were also provided by collaborators and colleagues from various research institutions.

Museum records were invaluable, providing detailed information on all specimens available within the collections, as well as general sightings information to assist with establishing true species distributions, confirming accuracy of locality data, and planning targeted sampling for rare taxa. In particular, the Kimberley Islands Biological

Survey (2006–2013) of the west Kimberley islands conducted by the Department of

Chapter 1: General introduction and thesis overview

Parks and Wildlife (DPAW) provided crucial sampling which would otherwise be unavailable for this research due to financial cost and restricted access to many of these localities.

1.6.2.3. Genetics

Initial analyses for this project involved assembly of large-scale mitochondrial

(mtDNA) datasets to establish the broad patterns of genetic diversity and how this was structured across the landscape for multiple taxa. Once preliminary mitochondrial diversity patterns were established, I was able to conduct targeted sub-sampling within several taxonomic groups to develop a subset of both genetically- and geographically- representative individuals for which additional independent molecular and morphological datasets were to be gathered. Using traditional Sanger sequencing methods, I developed small multi-locus nuclear (nuDNA) datasets to attempt to confirm the genetic structure observed in mtDNA for species complexes studied in Chapters 2 and 3.

To significantly enhance ability to assess mtDNA structure, attempt to confirm evolutionary distinctness of lineages, and more accurately infer the diversification processes involved within this landscape, I chose to expand molecular datasets for my final study (Chapter 4) using next-generation sequencing methods. This was conducted in collaboration with the Moritz Lab at ANU, where collaborators had developed targeted exon-capture arrays for a number of reptile taxa (skinks and geckos). The exon capture method is explained in detail in Chapter 4, to summarize briefly it involved the development of a large probe-set designed to target thousands of protein-coding exon regions, using transcriptomes of closely related taxa. The process then involved

Chapter 1: General introduction and thesis overview

preparation, amplification, pooling and hybridisation of genetic material for ~56 individual specimens, each uniquely barcoded so sequences could be linked back to specific individuals after sequencing. This method produced sub-genomic datasets with thousands of nuDNA sequences for potentially hundreds of individuals.

1.6.3. Synthesis

To conclude the thesis I will discuss the major findings of my work and how each data chapter (Chapters 2–4) contributes to further knowledge of the diversification patterns and processes within the tropical savannahs of the Kimberley. I will go into some detail on four key observations regarding Kimberley phylogeography; namely, i) the Kimberley supports high biodiversity, ii) multiple “refugia” regions of short-range endemism are evident, iii) there is evidence that depth and range-size of diversity scales over an aridity gradient, and iv) ecology plays a role in diversification response yet taxa with different ecologies still display higher genetic structure in the Kimberley than related regions. Finally, I will briefly discuss some additional avenues of research which would further enhance our understanding of the diversification processes occurring within the Kimberley and how climate, topography and the ecologies of taxa interact to shape the diversity patterns we observe.

Chapter 1: General introduction and thesis overview

Figure 1.1. a) Location of Australian Monsoonal Tropics (AMT) and Arid Zone (AAZ) biomes, including the three major regions within the AMT – key to the work covered in this thesis are the Kimberley (northwest), and Top End (central north). b) Topographic relief map of the Kimberley from the Atlas of Living Australia’s (ALA) online spatial portal, warmer colours indicate higher elevation. c) Map indicating some of the larger west Kimberley islands, many of which were important sampling localities for this thesis; map adapted from Lyons et al. 2014. d) Rainfall gradient across the Kimberley region, solid lines indicate rainfall isohyets; map adapted from the work of Palmer et al., 2013. e) Map indicating aridity gradient across the Kimberley region via an annual mean aridity index as taken from the ALA online spatial portal, warmer colours indicate higher aridity. f) Example of a gecko species (Oedura gracilis) which relies on the rock habitats of the Kimberley and is a key taxon considered in the phylogeographic work of this thesis (Chapter 3); photograph courtesy of Stephen Zozaya.

Chapter 1: General introduction and thesis overview

References

Avise JC (2000) Phylogeography: the history and formation of species Harvard

university press.

Avise JC (2012) Molecular Markers, Natural History and Evolution Springer Sicence

& Business Media.

Beheregaray LB (2008) Twenty years of phylogeography: the state of hte field and the

challenges for the Southern Hemisphere. Molecular Ecology 17, 3754-3774.

Bermingham E, Moritz C (1998) Comparative phylogeography: concepts and

applications. Molecular Ecology 7, 367-369.

Bond WJ, Parr CL (2010) Beyond the forest edge: ecology, diversity and conservation

of the grassy biomes. Biological Conservation 143, 2395-2404.

Bowman DMJS, Brown GK, Braby MF, et al. (2010) Biogeography of the Australian

monsoon tropics. Journal of Biogeography 37, 201-216.

Byrne M, Steane DA, Joseph L, et al. (2011) Decline of a biome: evolution, contraction,

fragmentation, extinction and invasion of the Australian mesic zone biota.

Journal of Biogeography 38, 1635-1656.

Byrne M, Yeates DK, Joseph L, et al. (2008) Birth of a biome: insights into the

assembly and maintenance of the Australian arid zone biota. Molecular Ecology

17, 4398-4417.

Carnaval AC, Waltari E, Rodrigues MT, et al. (2014) Prediction of phylogeographic

endemism in an environmentally complex biome. Proceedings of the Royal

Society B: Biological Sciences 281, 20141461.

Carwardine J, O'Connor T, Legge S, et al. (2011) Priority threat management to protect

Kimberley wildlife CSIRO Ecosystem Sciences, Brisbane.

Chapter 1: General introduction and thesis overview

Catullo RA, Keogh JS (2014) Aridification drove repeated episodes of diversification

between Australian biomes: Evidence from a multi-locus phylogeny of

Australian toadlets (Uperoleia: Myobatrachidae). Molecular Phylogenetics and

Evolution 79, 106-117.

Catullo RA, Lanfear R, Doughty P, Keogh JS (2014) The biogeographical boundaries of

northern Australia: evidence from ecological niche models and a multi-locus

phylogeney of Uperoleia toadlets (Anura: Myobatrachidae). Journal of

Biogeography 41, 659-672.

Collins JH (2008) Marine tourism in the Kimberley region of Western Australia.

Geographical Research 46, 111-123.

Criscione F, Law ML, Köhler F (2012) Land snail diversity in the monsoon tropics of

Northern Australia: revision of the genus Exiligada Iredale, 1939 (Mollusca:

Pulmonata: Camaenidae), with description of 13 new species. Zoological

Journal of the Linnean Society 166, 689-722.

De Deckker P, Yokoyama Y (2009) Micro-palaeontological evidence for Late

Quaternary sea-level changes in Bonaparte Gulf, Australia. Global and

Planetary Change 66, 85-92. de Queiroz K (1998) The general lineage concept of species, species criteria, and the

process of speciation: a conceptual unification and terminological

recommendations. In: Endless Forms: Species and Speciation (eds. Howard DJ,

Berlocher SH), pp. 57-75. Oxford University Press, Oxford. de Queiroz K (2007) Species concepts and species delimitation. Systematic Biology 56,

879-886.

Chapter 1: General introduction and thesis overview

Denniston RF, Wyrwoll K-H, Asmerom Y, et al. (2013) North Atlantic forcing of

millenial-scale Indo-Australian monsoon dynamics during the Last Glacial

period. Quaternary Science Reviews 72, 159-168.

Domingos FMCB, Bosque RJ, Cassimiro J, et al. (2014) Out of the deep: cryptic

speciation in a Neotropical gecko (Squamata, Phyllodactylidae) revealed by

species delimitation methods. Molecular Phylogenetics and Evolution 80.

Doughty P (2011) An emerging frog diversity hotspot in the northwest Kimberley of

Western Australia: another new frog species from the high rainfall zone.

Records of the Western Australian Museum 26, 209-216.

Doughty P, Palmer R, Sistrom MJ, Bauer AM, Donnellan SC (2012) Two new species

of Gehyra (Squamata: Gekkonidae) geckos from the north-west Kimberley

region of Western Australia. Records of the Western Australian Museum 27,

117-134.

Frawley S, O'Connor S (2010) A 40,000 year wood charcoal record from Carpenter's

Gap 1: New insights into palaeovegetation change and indigenous foraging

strategies in the Kimberley, Western Australia. Altered Ecologies: Fire, Climate

and Human Influence on Terrestrial Landscapes (Terra Australis 32), 257-279.

Fujioka T, Chappell J, Fifield LK, Rhodes EJ (2009) Australian desert dune fields

initiated with Pliocene-Pleistocene global climatic shift. Geology 37, 51-54.

Fujita MK, McGuire JA, Donnellan SC, Moritz C (2010) Diversification and

persistence at the arid-monsoonal interface: Australia-wide biogeography of the

Bynoe's Gecko (Heteronotia binoei: Gekkonidae). Evolution 64, 2293-2314.

Chapter 1: General introduction and thesis overview

Gamble T, Colli GR, Rodrigues MT, Werneck FP, Simons AM (2012) Phylogeny and

cryptic diversity in geckos (Phyllopezus; Phyllodactylidae; Gekkota) from South

America's open biomes. Molecular Phylogenetics and Evolution 62, 943-953.

Gibson LA, Köhler F (2012) Determinants of species richness and similarity of species

composition of land snail communities on Kimberley islands Records of the

Western Australian Museum Supplement 81, 40-65.

Grace J, José SJ, Meir P, Miranda HS, Montes RA (2006) Productivity and carbon

fluxes of tropical savannas. Journal of Biogeography 33, 387-400.

Heinicke MP, Greenbaum E, Jackman TR, Bauer AM (2011) Phylogeny of a trans-

Wallacean radiation (Squamata, Gekkonidae, Gehyra) supports a single early

colonization of Australia. Zoologica Scripta 50.

Köhler F (2011) Descriptions of new species of the diverse and endemic land snail

Amplirhagada Iredale, 1933 from rainforest patches across the Kimberley,

Western Australia (Pulmonata: Camaenidae). Records of the Australian Museum

63, 167-202.

Leaché AD, Fujita MK (2010) Bayesian species delimitation in West African forest

geckos (Hemidactylus fasciatus). Proceedings of the Royal Society B: Biological

Sciences 277, 3071-3077.

Lorenzen E, Heller R, Siegismund HR (2012) Comparative phylogeography of African

savannah ungulates. Molecular Ecology 63, 534-542.

Lyons MN, Keighery GJ, Gibson LA, Handasyde T (2014) Flora and vegetation

communities of selected islands off the Kimberley coast of Western Australia.

Records of the Western Australian Museum Supplement 81, 205-244.

Chapter 1: General introduction and thesis overview

Martin LJ, Blossey B, Ellis E (2012) Mapping where ecologists work: biases in the

global distribution of terrestrial ecological observations. Frontiers in Ecology

and the Environment 10, 195-201.

Mayr E, Mayr E, Mayr E, Mayr E (1963) Animal species and evolution Belknap Press

of Harvard University Press, Cambridge, Massachusetts.

McCormack JE, Hird SM, Zellmer AJ, Carstens BC, Brumfield RT (2013) Applications

of next-generation sequencing to phylogeography and phylogenetics. Molecular

Phylogenetics and Evolution 66, 526-538.

Miles L, Newton AC, Defries RS, et al. (2006) A global overview of the conservation

status of tropical dry forests. Journal of Biogeography 33, 491-505.

Miralles A, Vences M (2013) New metrics for comparison of taxonomies reveal striking

discrepancies among species delimitation methods in Madascincus lizards. PLoS

ONE 8, e68242.

Moritz C, Ens EJ, Potter S, Catullo RA (2013) The Australian monsoonal tropics: an

opportunity to protect unique biodiversity and secure benefits for Aboriginal

communities. Pacific Conservation Biology 19, 343-355.

Moritz C, Fujita MK, Rosauer D, et al. (2015) Multilocus phylogeography reveals

nested endemism in a gecko across the monsoonal tropics of Australia. Mol

Ecol.

Nias RC, Burbidge AA, Ball D, Pressey RL (2010) Island arks: the need for an

Australian national island biosecurity initiative. Ecological Management and

Restoration 11, 166-167.

Nix H, Kalma J (1972) Climate as a dominant control in the biogeography of northern

Australia and New Guinea. In: Bridge and barrier: the natural and cultural

Chapter 1: General introduction and thesis overview

history of Torres Strait (ed. Walker D), pp. 61-91. Research School of Pacific

Publication E6/3. Australian National University Press, Canberra.

Nix HA (1982) Environmental determinants of biogeography and evolution in Terra

Australis. Evolution of the flora and fauna of arid Australia 47.

Nogueira C, Ribeiro Sr, Costa GC, Colli GR (2011) Vicariance and endemism in a

Neotropical savana hotspot: distribution patterns of Cerrado squamate reptiles.

Journal of Biogeography 38, 1907-1922.

Oliver PM, Adams M, Doughty P (2010) Molecular evidence for ten species and Oligo-

Miocene vicariance within a nominal Australian gecko species (Crenadactylus

ocellatus, Diplodactylidae). BMC Evolutionary Biology 10, 386.

Oliver PM, Couper PJ, Pepper M (2014a) Independent transitions between monsoonal

and arid biomes revealed by systematic revision of a complex of Australian

geckos (Diplodactylus; Diplodactylidae). PLoS ONE 9, e111895.

Oliver PM, Hugall A, Adams M, Cooper SJB, Hutchinson MN (2007) Genetic

elucidation of cryptic and ancient diversity in a group of Australian

diplodactyline geckos; the Diplodactylus vittatus complex. Molecular

Phylogenetics and Evolution 44, 77-88.

Oliver PM, Keogh JS, Moritz C (2015) New approaches to cataloguing and

understanding evolutionary diversity: a perspective from Australian herpetology.

Australian Journal of Zoology 62, 417-430.

Oliver PM, Sanders K (2009) Molecular evidence for Gondwanan origins of multiple

lineages within a diverse Australasian gecko radiation. Journal of Biogeography

36, 2044-2055.

Chapter 1: General introduction and thesis overview

Oliver PM, Smith KL, Laver RJ, Doughty P, Adams M (2014b) Contrasting patterns of

persistence and diversification in vicars of a widespread Australian lizard

lineage (the Oedura marmorata complex). Journal of Biogeography 41, 2068-

2079.

Packer L, Gibbs J, Sheffield C, Hanner R (2009) DNA barcoding and the mediocrity of

morphology. Molecular Ecology Resources 9, 42-50.

Padial JM, Miralles A, De la Riva I, Vences M (2010) The integrative future of

taxonomy. Frontiers in zoology 7, 1.

Palmer R, Pearson DJ, Cowan MA, Doughty P (2013) Islands and scales: a

biogeographic survey of reptiles on Kimberley islands, Western Australia.

Records of the Western Australian Museum Supplement 81, 183-204.

Parr CL, Lehmann CER, Bond WJ, Hoffmann WA, Andersen AN (2014) Tropical

grassy biomes: misunderstood, neglected, and under threat. TRENDS in Ecology

and Evolution 29, 205-213.

Pennington RT, Lewis GP, Ratter JA (2006) An overview of the plant diversity,

biogeography and conservation of Neotropical savannas and seasonally dry

forests. In: Neotropical savannas and seasonally dry forests: plant diversity,

biogeography and conservation (eds. Pennington RT, Lewis GP, Ratter JA), pp.

1-29. CRC Press, Boca Raton, FL.

Pepper M, Doughty P, Hutchinson MN, Keogh JS (2011a) Ancient drainages divide

cryptic species in Australia's arid zone: Morphological and multi-gene evidence

for four new species of Beaked Geckos (Rhynchoedura). Molecular

Phylogenetics and Evolution 61, 810-822.

Chapter 1: General introduction and thesis overview

Pepper M, Fujita MK, Moritz C, Keogh JS (2011b) Palaeoclimate change drove

diversification among isolated mountain refugia in the Australian arid zone.

Molecular Ecology 20, 1529-1545.

Pepper M, Keogh JS (2014) Biogeography of the Kimberley, Western Australia: a

review of landscape evolution and biotic response in an ancient refugium.

Journal of Biogeography 41, 1443-1455.

Potter S, Bragg JG, Peter BM, Bi K, Moritz C (2016) Phylogenomics at the tips:

inferring lineages and their demographic history in a tropical lizard, Carlia

amax. Molecular Ecology.

Potter S, Eldridge MDB, Taggart DA, Cooper SJB (2012) Multiple biogeographical

barriers identified across the monsoon tropics of northern Australia:

phylogeographic analysis of the brachyotis group of rock-wallabies. Molecular

Ecology 21, 2254-2269.

Proske U, Heslop D, Haberle S (2014) A Holocene record of coastal landscape

dynamics in the eastern Kimberley region, Australia. Journal of Quaternary

Science 29, 163-174.

Reeves JM, Bostock HC, Ayliffe LK, et al. (2013) Palaeoenvironmental change in

tropical Australasia over the last 30,000 years–a synthesis by the OZ-

INTIMATE group. Quaternary Science Reviews 74, 97-114.

Sistrom MJ, Donnellan SC, Hutchinson MN (2013) Delimiting species in recent

radiations with low levels of morphological divergence: a case study in

Australian Gehyra geckos. Molecular Phylogenetics and Evolution 68, 135-143.

Chapter 1: General introduction and thesis overview

Sistrom MJ, Edwards DL, Donnellan SC, Hutchinson MN (2012) Morphological

differentiation correlates with ecological but not with genetic divergence in a

Gehyra gecko. Journal of Evolutionary Biology 25, 647-660.

Sistrom MJ, Hutchinson MN, Hutchinson RG, Donnellan SC (2009) Molecular

phylogeny of Australian Gehyra (Squamata: Gekkonidae) and taxonomic

revision of Gehyra variegata in south-eastern Australia. Zootaxa 2277, 14-32.

Smith KL, Harmon LJ, Shoo LP, Melville J (2011) Evidence of constrained phenotypic

evolution in a cryptic species complex of agamid lizards. Evolution 65, 976-992.

Toon A, Austin JJ, Dolman G, Pedler L, Joseph L (2012) Evolution of arid zone birds in

Australia: Leapfrog distribution patterns and mesic-arid connections in quail-

thrush (Cinclosoma, Cinclosomatidae). Molecular Phylogenetics and Evolution

62, 286-295.

Toon A, Crisp MD, Gamage H, et al. (2015) Key innovation or adaptive change? A test

of leaf traits using Triodiinae in Australia. Scientific reports 5.

Waeber PO, Wilmé L, Ramamonjisoa B, et al. (2015) Dry forests in Madagascar:

neglected and under pressure. International Forestry Review 17, 127-148.

Wendel JF, Doyle JJ (1998) Phylogenetic incongruence: window into genome history

and molecular evolution. In: Molecular systematics of plants II, pp. 265-296.

Springer US.

Werneck FP, Gamble T, Colli GR, Rodrigues MT, Sites Jr. JW (2012) Deep

diversification and long-term persistence in the South American 'dry diagonal':

integrating continent-wide phylogeography and distribution modeling of geckos.

Evolution 66, 3014-3034.

Chapter 1: General introduction and thesis overview

Wiley EO (1978) The evolutionary species concept reconsidered. Systematic Biology

27, 17-26.

Wilkins JS (2009) Species: a history of the idea University of California Press.

Wilkins JS (2011) Philosophically speaking, how many species concepts are there.

Zootaxa 2765, 58-60.

Woinarski J, Mackey B, Nix H, Traill B (2007) The Nature of Northern Australia:

Natural Values, Ecological Processes and Future Prospects. The Australian

National University E Press, Canberra.

Chapter 2: Role of biome ecology in shaping diversification patterns

The role of biome ecology in shaping

diversification patterns in a clade of Australian

spinifex-dwelling lizards (Strophurus)

Photograph courtesy of Henry Cook

Chapter 2: Role of biome ecology in shaping diversification patterns

Abstract

Inter-biome comparisons of biodiversity patterns can inform us of the relative roles of history, landscape, and environmental change in shaping diversification patterns within biomes. The Australian continent is dominated by the Monsoonal Tropics

(AMT) and Arid Zone (AAZ) biomes which share an extensive border region and a high diversity of related taxa. Genetic diversity of lineages spanning both biomes tends to be higher and older within the AMT, and usually associated with heterogeneous rock escarpments. The AAZ is considered a relatively young biome in comparison, and exhibits younger, more widespread diversity often derived from mesic ancestry.

Phylogeographic comparisons between the AMT and AAZ have recently shown, however, that this is not always the case: in generalist and rock-specialist lizard taxa there is some evidence of long histories and highly structured lineages associated with isolated montane refugia.

Here, I compared the timing and patterns of diversification within and between biomes across a different ecologically-specialised clade of Australian lizards with

Gondwanan ancestry which are associated with spinifex grasses (Strophurus: phasmid geckos). My results indicated that despite this taxon having an extensive history within the AAZ there was limited genetic structure to reflect this age. In contrast, genetic diversity was higher, older and more finely geographically structured within phasmid gecko lineages of the AMT. This implies that diversity in this biome is not necessarily explained by the age of the biome or history of occupancy, but perhaps more significantly by the structure of the landscape and habitat, which may support greater persistence through climatic and environmental change.

Chapter 2: Role of biome ecology in shaping diversification patterns

2.1. Introduction

Comparing biodiversity patterns between biomes allows us to better establish the relative roles of history, landscape and environmental change in shaping diversification of taxa within biomes (Byrne et al. 2011; Byrne et al. 2008; Crisp et al.

2009; Oliver et al. 2014a; 2014c). The majority of the Australian continent is dominated by either Monsoonal Tropics (AMT) or Arid Zone (AAZ) desert biomes, both sharing an extensive border region across the north of Australia, as well as a high diversity of related taxa (Bowman et al. 2010; Byrne et al. 2008). In lineages which span both biomes there is a tendency for diversity within the AMT to be higher and older, compared to younger, widespread, shallow diversity in the AAZ (Byrne et al. 2011;

Byrne et al. 2008; Crisp et al. 2004; Fujita et al. 2010; Kuch et al. 2005; Oliver et al.

2014c; Sakaguchi et al. 2013). Arid Zone lineages are often found to be derived from mesic ancestry (Byrne et al. 2008; Chapple & Keogh 2004; Ladiges et al. 2011; Toon et al. 2012; Williams et al. 2010), hence signatures of younger diversity and range- expansion within this region fit with the idea that the AAZ is a relatively young biome

(Fujita et al. 2010; Jennings et al. 2003; Kuch et al. 2005; Marin et al. 2013; Pepper et al. 2011a; Pepper et al. 2011c). With this understanding, it could be assumed that the comparative patterns of diversity levels between the AMT and AAZ biomes are due to relative timescales over which lineages have occupied each region. However, despite contemporary deserts having formed fairly recently (<4 Ma; Fujioka et al. 2009), significant structure within rocky isolates indicate persistence of much older diversity

(Byrne 2008; Chapple & Keogh 2004; Maryan et al. 2007; Melville et al. 2011; Oliver

& McDonald 2016; Pepper et al. 2008; Pepper et al. 2011; Shoo et al. 2008).

Furthermore, evidence is beginning to accumulate indicating a more complex history of

Chapter 2: Role of biome ecology in shaping diversification patterns

transitions between the AMT and AAZ biomes in both directions (e.g. Catullo & Keogh

2014; Nielsen et al. 2016; Oliver et al. 2014a; Toon et al. 2015).

Phylogeographic work of related AMT and AAZ lineages has tended to focus on widespread terrestrial or generalist vertebrate taxa (Fujita et al. 2010; Oliver & Bauer

2011; Oliver et al. 2014a; Pepper et al. 2011c). Recent increased focus on rock- specialised (saxicoline) taxa particularly within the topographically complex AMT biome is recovering especially fine-scale geographic structuring of lineage diversity (see

Chapters 3 and 4; Oliver et al. 2009; Oliver et al. 2014b; Potter et al. 2012; Rosauer et al. 2016). When these rock specialised groups are compared across rock isolates of both biomes the contrast of diversity patterns becomes even starker (e.g. Oliver et al. 2014c;

Pepper et al. 2011b). Rock escarpments provide typically stable, and somewhat climatically buffered, habitat over broad timescales; and hence associated rock- specialist taxa are apt for phylogeographic biome comparisons. However, studies of additional types of ecological specialists are required to further improve our understanding of diversification histories within both biomes.

A globally unique feature of the Australian biota is the abundant, drought-tolerant, sclerophyllous grasses of the genus Triodia (spinifex). This genus is a dominant habitat component across ~30% of Australia, especially disjunct rocky plateaux of the AMT and larger expanses of the AAZ (Bowman et al. 2010; Crisp & Cook 2013).

Interestingly, despite the ecological dominance of spinifex within the AMT and AAZ biomes, recent studies have shown the sub-tribe Triodiinae only colonised Australia around the mid-Miocene (Toon et al. 2015). Furthermore, contra the general paradigm that AAZ lineages are derived from mesic ancestors now restricted to peripheral biomes

(e.g. Byrne et al. 2008), the initial radiation of spinifex may have arid origins, with

Chapter 2: Role of biome ecology in shaping diversification patterns

subsequent colonisation and radiation into more mesic monsoonal environments of northern Australia (<10 Ma; Toon et al. 2015). Juxtaposed against this history, extant diversity appears more finely structured in the AMT. The AMT has multiple relatively divergent lineages endemic to topographically complex regions (Toon et al. 2015), and emerging patterns within the AAZ show highest structuring associated with rocky refugia such as the Pilbara (Anderson et al. 2016); mirroring the patterning of vertebrate diversity distributed across these biomes (Byrne 2008; Byrne et al. 2011; Byrne et al.

2008; Fujita et al. 2010; Oliver et al. 2014a; Oliver et al. 2014c; Pepper et al. 2011c).

Spinifex provides food, shelter, and a thermally-buffered microhabitat for a diverse associated biotic community (Pianka 1981; Wilson 2012), including numerous dependent lineages (e.g. birds, (Christidis et al. 2010); mammals, (Haythornthwaite &

Dickman 2006); reptiles, (Gordon et al. 2010); and invertebrates (Dessen 2008)). Many

Australian lizard taxa are closely associated with spinifex, taking refuge and foraging within spinifex clumps (Rabosky et al. 2007b; Wilson & Swan 2013; Wilson 2012).

Whilst majority of these lizards are terrestrial, some are scansorial (climbing specialists), living on spinifex itself.

The most diverse group of Australian spinifex-dwelling (graminicolous) lizards are the colloquially named ‘phasmid’ geckos from the genus Strophurus (see Nielsen et al.

2016). These geckos are highly dependent and entirely known to occur on spinifex habitat (King & Horner 1993; Storr 1978; Wilson & Swan 2013; Wilson 2012). The taxa within this group all have a distinctive morphology: small to very small size compared to congenerics (44–56mm SVL; see Appendix: Fig. A2.1.); an elongate body- form; and, most distinctively, a ‘pin-striped’ dorsal, and sometimes ventral, colour pattern (Nielsen et al. 2016; Wilson & Swan 2013; Wilson 2012). Among gekkotans,

Chapter 2: Role of biome ecology in shaping diversification patterns

thin, longitudinal and highly contrasting stripes are particularly rare (≤ 25 examples in over 1,500 species; Appendix: Table A2.1.). More generally such patterning is also often associated with utilisation of longitudinally-oriented vegetation (e.g. bamboo, sedges, grasses, etc.) suggesting a specialised disruptive or camouflage function (Losos

2009).

The genus Strophurus is part of a Gondwanan radiation with a history pre-dating the isolation of Australia from Antarctica (Oliver et al. 2009; Oliver & Sanders 2009).

These Gondwanan origins therefore preclude Asian immigration into northern habitats as an explanation for patterns of biome diversification within this gecko group.

Furthermore, recent work by Nielsen et al. (2016) indicated Strophurus, including the phasmid group, have a long history (~20 My) in the AAZ; and although support for phylogenetic relationships between clades was ambiguous, ancestral reconstructions suggested a potential origin for this genus within the AAZ. The phasmid Strophurus include five recognised species distributed across the AMT and AAZ. Specifically: i)

Strophurus jeanae – distributed throughout the northern deserts of the AAZ, from the

Pilbara craton to the central ranges; ii) S. mcmillani – west Kimberley, iii) S. robinsoni

– east Kimberley (Ord Region), iv) S. horneri – Arnhem Land (Top End), and v) S. taeniatus – Northern Deserts Region from the southwest Kimberley to the Selwyn

Ranges (Fig. 2.1.). However, many of these taxa are difficult to distinguish; S. jeanae and S. taeniatus having been confounded for majority of the last century (Storr 1988), and their geographic ranges are still poorly known (Vanderduys et al. 2012).

Furthermore, given the recent discovery and description of S. horneri (Oliver & Parkin

2014) and the propensity for cryptic species complexes within the AMT (e.g. Moritz et

Chapter 2: Role of biome ecology in shaping diversification patterns

al. 2015; Oliver et al. 2016; Oliver et al. 2012), further hidden diversity may exist within the phasmid geckos.

Given the potential origins and long histories of both spinifex and spinifex-dwelling phasmid geckos in the AAZ, it could be predicted that phasmid geckos may show higher genetic structuring within this biome compared to younger immigrants from the mesic north. Alternatively, if landscape structure plays a stronger role than history within a biome in shaping diversity patterns, the reverse could be expected and phasmid geckos in the AMT will display higher genetic structure. In this chapter I set out to further place patterns of diversity in the Kimberley in a broader continental context. To do this I compared diversification patterns between and within the AMT and AAZ biomes in an ecologically specialised clade of geckos which may have originated within the AAZ. Using all available tissues for the clade of phasmid Strophurus I collated a multi-locus dataset and used phylogenetic and species delimitation analyses to assess and compare levels and patterns of genetic diversity. I implemented dating estimation methods to further compare the timescales over which diversification may have occurred in different regions. Finally, I also used simple species distribution modelling to see if the environmental variables that best correlated with the distributions of phasmid Strophurus lineages varied between regions and biomes.

2.2. Materials & methods

2.2.1. Sampling

Tissues used in genetic analyses are listed in Table A2.2. Sampling included all tissues registered in Australian museums (47 individuals) across all five currently recognised species of phasmid Strophurus geckos (Fig. 2.1., Table A2.2.a). Sequence

Chapter 2: Role of biome ecology in shaping diversification patterns

data for an additional 57 outgroup taxa (22 from Strophurus) were included to provide calibration nodes for dating analyses (Table A2.2.b). We sequenced a portion of the mitochondrial locus ND2 for all samples, and three nuclear loci (PDC, PRLR, RAG-1) for majority of individuals (including outgroups). Details and characteristics of genetic data are listed in Table A2.3. Primer sequences and amplification protocols are provided in Table A2.4.

Genomic DNA was extracted from liver or tail tip samples using a Qiagen DNeasy extraction kit or a Qiaxtractor (Qiagen, Valencia, CA). PCR products were purified using 1uL of a 20% dilution of ExoSAP-IT (US78201, Amersham Biosciences,

Piscataway, NJ), incubated at 37°C for 30 min, followed by 80°C for 15 min. Clean products were then sent to genetic services companies (Macrogen, Seoul, South Korea and DNASU, Arizona State University, USA) with amplicons sequenced in both directions. Gene sequences were assembled and edited using GENEIOUS v.6.1.7

(Drummond et al. 2008), and alignments visually examined and translated into amino acids to confirm correct reading frames and full translation. Previously published sequences were also used in analyses and all new sequences were deposited to GenBank

(Table A2.2.).

2.2.2. Nucleotide data and phylogenetic analyses

Congruence between mitochondrial and nuclear data was assessed by estimating phylogenies for i) single loci, ii) concatenated nuclear dataset, and iii) combined four locus dataset (mtDNA + nuDNA). Models of nucleotide substitution and partitioning strategies (Table A2.5.) were selected for each locus in PARTITIONFINDER v1.1.1

(Lanfear et al. 2012) using the Bayesian information criterion (BIC). Phylogenetic

Chapter 2: Role of biome ecology in shaping diversification patterns

relationships were estimated using Maximum-Likelihood (RAXML v8.0.24; Stamatakis

2006; Stamatakis et al. 2008) and Bayesian (MRBAYES v3.2.2; Ronquist &

Huelsenbeck 2003) analyses implemented through the CIPRES Science Gateway 3.1 for online phylogenetic analysis (Miller et al. 2010). Default RAXML settings were used in

CIPRES following selected partition strategies. Bayesian analyses in MRBAYES also used selected models and partitions, running four independent Markov Chain Monte

Carlo (MCMC) chains with 4 x 10 million generations sampling every 1000. ARE WE

THERE YET (AWTY; Wilgenbusch et al. 2004) and TRACER v1.6 (Rambaut et al. 2014) were used to ensure convergence, stability of log likelihoods, and stationarity and

Effective Sample Size (ESS) values >100 for all parameters. Maximum Clade

Credibility (MCC) trees were then constructed after 20% of samples were discarded as burn-in.

To assess and compare diversity levels across regions and biomes average, minimum, and maximum genetic distances (Tamura Nei [TN]; Tamura & Nei 1993) within and between major mtDNA (genetic divergence ≧ 8%) lineages within the phasmid geckos were calculated in MEGA v6.06 (Tamura et al. 2013). Phylogenetic tree methods can be problematic at the level of population genetics or lower diversity where reticulation and recombination can lower resolution and increase uncertainty of relationship estimation (Posada & Crandall 2002; Vriesendorp & Bakker 2005). For this reason, lineage relationships were also visualized by generating phylogenetic networks of mtDNA and concatenated nuDNA separately, using the Neighbor-Net algorithm

(Bryant & Moulton 2004) in SplitsTree v4.10 (Huson & Bryant 2006). As this method only compares the pairwise divergences between sequences rather than estimating evolutionary history, concatenating the nuclear loci is not an issue. I assessed support

Chapter 2: Role of biome ecology in shaping diversification patterns

for inferred splits with 1000 bootstrap pseudoreplicates. To determine if demographic histories varied in different biomes I also tested for signatures of range expansion in mtDNA using Tajima’s D (Tajima 1989) and Fu’s Fs (Fu 1997) neutrality tests calculated in DNASP v5.10.01 (Librado & Rozas 2009) for the major lineages identified

(see below).

For accurate comparison of diversity patterns between regions it is important that the distinctness of the genetic structure observed is confirmed as genuine rather than a signal artefact of historical structure persisting in certain genetic markers. I therefore tested the independence of divergent mtDNA lineages detected using coalescent species delimitation with the program Bayesian Phylogenetics and Phylogeography (BP&P) v2.1 (Yang 2015). I conducted multiple Bayesian species delimitation analyses using a nuDNA dataset (three loci) I had phased with the PHASE v2.1 program (Stephens &

Donnelly 2003; Stephens et al. 2001) and SeqPHASE online (Flot 2010). The starting tree topology used matched the mtDNA tree and I explicitly tested for the distinctness of nine major lineages (see Results; Fig. 2.2., 2.3., and Table A2.6.). I ran multiple

BP&P analyses varying population size parameters (θs) and divergence time the species tree root (τ0). Gamma priors I assigned as i) both θs and τ0 equal to G(1, 10), mean =

0.1; ii) both θs and τ0 equal to G(2,2000), mean = 0.001; iii) θs – G(1,10), τ0 –

G(2,2000); and iv) θs – G(2,2000), τ0 – G(1,10). For other divergence time parameters I assigned the Dirichlet prior (Yang & Rannala 2010]: equation 2). I used the species delimitation rjMCMC algorithm 0 with finetune ε = 2, and the species model prior set to

0 so as not to favour symmetric trees. I ran analyses for 100,000 generations, sampling every two, with a burn-in of 8,000; and conducted two independent runs differing starting seeds to ensure consistency of results.

Chapter 2: Role of biome ecology in shaping diversification patterns

2.2.3. Phylogenetic dating

To understand the evolutionary history of the phasmid geckos within different biomes I estimated timescales of diversification using BEAST v1.8.0 (Drummond et al.

2012). The same models and partitioning strategies were used as in the Bayesian analyses above (Table A2.5.). I compared different clock and speciation models using path-sampling and stepping-stone methods (Baele et al. 2012; Baele et al. 2013), and strict clock with Yule process speciation priors resulted in highest Maximum-

Likelihood values. Final analyses I ran for 10 million generations, sampling every 1000.

I assessed convergence, stability and adequate ESS values (>100) in TRACER v1.6

(Rambaut et al. 2014) and AWTY (Wilgenbusch et al. 2004), then discarded 20% of samples as burn-in, and summarised remaining MCC trees with TREEANNOTATOR v1.8.0 (Drummond et al. 2012).

I wanted to build upon and improve the robustness of the phylogenetic dating analysis conducted in previous work by Nielsen (2016) by comparing estimates calculated from different datasets, and including an additional nuclear locus. I estimated diversification times within Strophurus using three strategies: i) a nuDNA (PDC, PRLR, and RAG1) only dataset with secondary constraints derived from previous fossil calibrated squamate phylogenies (i.e. the mean and standard deviations obtained in previous phylogenetic dating analyses) for key nodes (Lee et al. 2009; Oliver et al.

2014c) – two normally distributed age priors for Diplodactylidae (‘core

Diplodactylidae’ [mean 35 Ma; standard deviation 6] and the New

Caledonian/Pseudothecadactylus clade [43 Ma; SD 9]), with normal root prior [mean

75 Ma; SD 13] (Oliver et al. 2010); ii) a combined nuDNA and mtDNA (minus third codons in case of saturation) dataset with the same calibrations as above; and iii) a

Chapter 2: Role of biome ecology in shaping diversification patterns

mtDNA dataset with mean molecular rate of pairwise sequence divergence constrained to 3% pairwise per one million years (normal distribution, [mean 0.015; SD 0.003]).

Though a 2% pairwise rate is commonly used (Zamudio & Greene 1997) recent studies have implied 3% is potentially more accurate and realistic for squamates (Eo &

DeWoody 2013; Oliver et al. 2010).

2.2.4. Distribution modelling

Whilst comparing genetic diversity patterns between biomes I wanted to see if the distributions of phasmid Strophurus lineages in different regions were closely correlated with the same or different ecological variables. The distribution of spinifex varies in the

AMT – where it tends to be patchy and associated with oligotrophic soils in areas such as escarpments and surrounds (Bowman et al. 2010), compared to the AAZ – where it is more continuous and widespread (Crisp et al. 2004; Crisp & Cook 2013). With this in mind, it could be predicted that environmental variables associated with topographic complexity (rock escarpments) might correlate more strongly with distributions of phasmid geckos in the AMT than the AAZ. I used MAXENT v3.3.3 (Phillips & Dudík

2008) to estimate species distribution models (SDMs) for the three major clades of phasmid Strophurus (see results) and to compare the correlation of climate, geology and vegetation features with distributions of different taxa between biomes. Analyses were conducted at major clade level as taxonomic level, particularly within the Kimberley region, would have resulted in too reduced sample sizes to accurately model distributions. I chose eleven environmental variables (see Table 2.1.a) which could be considered important determinants of distributions for these geckos to compare in the

SDMs. Modelling was conducted using all available museum records to ensure

Chapter 2: Role of biome ecology in shaping diversification patterns

adequate sample sizes and most accurate estimation of distributions. For each lineage the model was repeated 20 times using the MAXENT bootstrapping option, with the median result retained.

2.3. Results

2.3.1. Phylogenetic relationships and lineage diversity

Monophyly of the clade comprising the five recognised phasmid gecko species relative to other Strophurus was strongly supported in all analyses with concatenated datasets (i.e. nuDNA only vs. all loci [mtDNA + nuDNA]; Fig. A2.2–3.). The monophyly of this clade was similarly supported in individual locus analyses for two loci (PRLR, RAG1). Though the remaining two loci (ND2, PDC) did not recover strong support (i.e. <75 Maximum-Likelihood bootstraps, <0.90 Bayesian posterior probabilities) for this node there was also no evidence of strongly supported incongruence with this relationship. Within the phasmid Strophurus clade, three geographically cohesive lineages were also strongly supported in all analyses: two clades in the AMT – i) mcmillani/robinsoni (Kimberley/Ord Region), and ii) horneri/taeniatus (Arnhem Land/Northern Deserts); and a single lineage from northern

AAZ iii) jeanae (Fig. A2.2–7.). The pattern of relationships between these three lineages was not resolved.

I recovered deep mtDNA structure within all nominal species except S. jeanae. The mcmillani/robinsoni clade comprised five lineages with mean Tamura-Nei (TN) mtDNA divergences ranging from 10.5–14.4% (Figs. 2.2.a and 2.3., lineages 1–5; Table

A2.6.). The distribution and relationships of these five lineages were inconsistent with current taxonomy of material in museums, and samples identified as S. robinsoni render

Chapter 2: Role of biome ecology in shaping diversification patterns

S. mcmillani paraphyletic. Within the Arnhem Land/Northern Deserts clade the two currently recognised species (S. horneri and S. taeniatus) formed two discrete mtDNA lineages, also showing further ‘intraspecific’ divergences (5.6–9.2%; Figs. 2.2.a and

2.3.). Mitochondrial divergences within the relatively restricted S. horneri from the

Arnhem Plateau were comparable to or deeper than those within the much more widely distributed Northern Deserts taxon S. taeniatus (S. taeniatus ‘west’ and S. taeniatus

‘east’, Fig. 2.2.a, Table A2.6.). Analyses based on nuDNA (PDC, PRLR and RAG1) provided no evidence of further subdivisions or recognition of distinct taxa within the

Kimberley/Ord Region and Arnhem Land/Northern Deserts clades (Fig. 2.2.b).

However, these loci typically used in phylogenetic studies across highly divergent genera often lack sufficient resolution for phylogeographic studies and were only used here due to ease of amplification and to ensure comparability to similar datasets. In contrast to high mtDNA diversity in the AMT, diversity within S. jeanae occupying the

AAZ is comparatively shallow (0–4.7%) despite the fact this taxon has by far the widest distribution.

Tajima’s D tests for range expansion within all three clades of phasmid geckos were not significant: Kimberley/Ord Region (0.010); Arnhem Land/Northern Deserts (0.196); and AAZ (-0.700). Although also not significant, the Fu’s Fs statistic for the AAZ jeanae lineage (-6.448) when compared to those of the AMT clades

(mcmillani/robinsoni – 1.338, horneri/taeniatus – -0.225) does imply a stronger signal of demographic expansion.

Species delimitation tests in Bayesian Phylogenetics and Phylogeography (BP&P) consistently supported the distinctness of the nine divergent mtDNA lineages in the

Kimberley and Top End (≧0.98 Posterior Probability), despite limited recovery in

Chapter 2: Role of biome ecology in shaping diversification patterns

nuDNA-based phylogenetic analyses. Further work is required to clarify the evolutionary significance of these lineages, but these results suggest the presence of additional cryptic candidate taxa within both the Kimberley and Top End regions.

2.3.2. Divergence dates

Despite tests suggesting mitochondrial saturation was not significant, preliminary analyses suggested dramatic inflation of age estimates for deep phylogenetic divergences within Strophurus when mitochondrial data was included. Hence I chose to conservatively remove third codons from the mtDNA in the combined nuDNA and mtDNA analyses, and in light of the possible mtDNA saturation issue (Brandley et al.

2011) I subsequently largely focus on dates derived from the secondarily calibrated dating analyses of the combined and nuDNA datasets. Secondarily calibrated age estimates indicated the crown radiation of the phasmid geckos dates to the early- to mid-Miocene (15–16 Ma, 95% Highest Posterior Density [HPD] 11–20 Ma). Estimated crown ages for the AMT clades tended to date to the very late-Miocene or Pliocene (4–5

Ma, HPD 2–7 Ma), while diversity within the AAZ lineage is comparatively younger

(1–3 Ma; Table 2.2). Relative age estimates were broadly concordant across methods

(Table 2.2).

2.3.3. Distribution modelling

The environmental variables which contributed most to the species distribution models (SDMs; Fig. A2.8.) differed across the three clades of phasmid Strophurus (see

Table 2.1.b), although total annual precipitation contributed the highest percentage to all models. Broad-scale climatic factors accounted for 88% of the distribution of the jeanae

Chapter 2: Role of biome ecology in shaping diversification patterns

lineage in the AAZ. Lineages in the AMT showed a stronger signal from local-scale variables; most notably slope (related to local topographic relief) contributed ~19% of the SDMs for both AMT lineages, as opposed to <3% for the AAZ lineage. Within the

AMT lineages further differentiation was evident. Broad-scale climatic factors and local-scale topographic and vegetation variables each contributed ~50% to prediction of the horneri/taeniatus distribution in the Top End/Northern Deserts; while broad-scale climatic variables, especially those associated with precipitation (including rainfall seasonality), explained almost 80% of the distribution of mcmillani/robinsoni within the

Kimberley.

2.4. Discussion

I present here a detailed phylogeographic and phylogenetic study of phasmid

Strophurus, a lineage of spinifex-specialised geckos from the Australian Monsoonal

Tropics (AMT) and Arid Zone (AAZ) biomes. My analyses strongly support that this group is monophyletic (concordant with previous work; Nielsen et al. 2016) and composed of three major clades: i) jeanae – a single species widespread in the AAZ; ii) horneri/taeniatus – two species occurring in the semi-arid Northern Deserts

(widespread) and Monsoonal Tropics (AMT) (restricted); and iii) mcmillani/robinsoni complex – entirely restricted to the AMT. Though not overt for slowly evolving nuDNA loci sequenced here, the latter includes at least five deeply divergent mtDNA lineages consistently supported as evolutionarily distinct by BP&P. Relationships between the three major clades remain ambiguous.

Chapter 2: Role of biome ecology in shaping diversification patterns

2.4.1. Contrasting lineage diversity in neighbouring biomes

An abundance of deeply divergent and highly geographically-structured unrecognised lineages is a striking, and until recently overlooked, pattern emerging from multiple phylogeographic analyses of northern Australian vertebrates (Moritz et al. 2015; Oliver et al. 2014c; Potter et al. 2016; Rosauer et al. 2016). This geographic structuring appears particularly pronounced in topographically complex escarpments of both the Kimberley (Doughty 2011; Oliver et al. 2010; Oliver et al. 2014b; Potter et al.

2012) and Arnhem Land (Catullo et al. 2014b; Cracraft 1991; Ladiges et al. 2003;

Oliver & Parkin 2014). The phasmid geckos broadly conform to these patterns despite having an ecology not linked (at least not directly) to rocks. I identified five divergent mitochondrial lineages of phasmid Strophurus geckos (mcmillani/robinsoni complex) in the Kimberley/Ord Region (divergences deeper or close to species-level divergences in other Diplodactylid geckos [e.g. 8–16%; Oliver & Sanders 2009; Fig. 2.3., lineages 1–5;

Table A2.6.; Oliver et al. 2014c), and at least two (within S. horneri) on the Arnhem

Plateau. Though most lineages were not recovered by slowly-evolving nuclear loci used in this study, analyses with rapidly-evolving nuclear exons and greater sampling could potentially better define this diversity (see Potter et al. 2016). In contrast, just one lineage with low genetic diversity (S. jeanae), also showing evidence of recent range expansion, is widespread across much of the AAZ – again a pattern replicated in many taxa (Chapple & Keogh 2004; Fujita et al. 2010; Jennings et al. 2003; Oliver et al.

2014c). Although some groups, including spinifex grasses, show evidence of genetic structure within rocky isolates in the AAZ, especially the Pilbara (Anderson et al. 2016;

Maryan et al. 2007; Melville et al. 2011; Pepper et al. 2008; Pepper et al. 2011b), phasmid geckos show no evidence of this.

Chapter 2: Role of biome ecology in shaping diversification patterns

Contrasting patterns of regional diversity may be attributed to multiple factors including differing biome age (i.e. time for speciation; Fujita et al. 2010; Pepper et al.

2011c) and/or differing net diversification (speciation - extinction) rates in different regions (Hutter et al. 2013; Rabosky et al. 2007a). Given the small number of lineages involved and uncertainty about ancestral biomes, my data provide no strong evidence phasmid geckos have a longer history in the AMT. However, in conjunction with spatial modelling, they do suggest topographic complexity is an important correlate with lineages in the AMT, but not in the AAZ. Empirically, the most geographically restricted lineages are associated with regions of geologically stable, ancient, and exposed escarpment in the Kimberley and Arnhem Land. Unlike many other highly geographically-structured taxa of this region, phasmid geckos are not rock specialised

(saxicoline). However, in heavily burnt and wetter areas of the northern AMT, large areas of (often) fire-sensitive spinifex are closely associated with stable rocky outcrops and often isolated by grassy savannah woodlands or black-soil plains (Bowman et al.

2010). In the AAZ, spinifex is more widespread, occurring across many different substrates (Crisp & Cook 2013) and there is also strong evidence desert habitats have been highly mobile and changed significantly throughout the Plio-Pleistocene (Fujioka et al. 2009). My findings would suggest the variable effect of underlying topography and climatic variation on the distribution of spinifex in different biomes may have played a role in shaping contrasting patterns of phylogeographic diversity in phasmid geckos.

Chapter 2: Role of biome ecology in shaping diversification patterns

2.4.2. Definition and evolutionary history of biomes

The distributions of phasmid gecko clades broadly correspond with and support the boundaries of two major Australian biomes and bioregions. One is restricted to the arid biome (jeanae), one to the monsoonal tropics (mcmillani/robinsoni), whilst the third

(horneri/taeniatus) includes sub-lineages associated with the Arnhem Plateau of the

AMT (S. horneri) and another widespread over the seasonally wet semi-arid zone

(Northern Deserts) along the northern edge of the AAZ (S. taeniatus). This last region has long been recognised as a broad interzone between biomes (Cracraft 1991; Nix

1982) and under different definitions could be viewed as either part of the AAZ (e.g. aridity index < 0.5) or AMT (> 85% of rainfall concentrated in summer; Bowman et al.

2010). My findings implying a Northern Deserts endemic complement a growing number of phylogenetic studies indicating the Northern Deserts region has fostered development of a unique localised biota (Catullo et al. 2014a; Catullo et al. 2014b;

Fujita et al. 2010; Ladiges et al. 2011; Melville et al. 2011; Smith et al. 2011).

The distribution of lineages across biomes prompts the question of the history of transitions between, and consequently time available for diversification within, the

Australian monsoonal and arid biomes. There is a general and longstanding trend to place the trajectory of evolution in the Australian continent into a framework of a broad directional transition from ancestral mesic environments into derived and younger arid environments (Byrne et al. 2011; Byrne et al. 2008; Crisp et al. 2004; Rix et al. 2015).

This framework is generally supported when comparing the central arid and mesothermal biomes along the east coast (Byrne et al. 2011; Byrne et al. 2008; Chapple

& Keogh 2004; Hugall et al. 2008; Oliver & Bauer 2011). However, there is potential for much nuance to this overall theme, and the history and patterns of biotic interchange

Chapter 2: Role of biome ecology in shaping diversification patterns

between seasonally mesic or arid areas (such as the AMT and AAZ) remain rather poorly understood (Crisp & Cook 2013; Jabaily et al. 2014; Toon et al. 2015).

Previous work on the genus Strophurus suggests the arid biome may be ancestral

(Nielsen et al. 2016); a pattern matching results for Triodia (Toon et al. 2015). The depth of divergences (mean estimates dating back to the mid- to late-Miocene) between the three major clades of phasmid geckos further emphasises that the seasonally arid or arid biomes to which this lineage is now restricted date back well into the Miocene.

This concurs with a growing number of paleo-ecological and phylogenetic datasets suggesting at least seasonally arid biomes – and/or fire-affected biomes – may have a long history on the Australian continent (Crisp & Cook 2013; Oliver et al. 2010; Oliver

& Bauer 2011; Toon et al. 2015). This result also adds to the mounting list of studies indicating an extensive history of biotic interchange between the AMT and AAZ

(Catullo & Keogh 2014; Fujita et al. 2010; Oliver et al. 2014a; Toon et al. 2015).

Interestingly however, the lack of genetic structure within the phasmid gecko lineage (S. jeanae) in the AAZ is surprisingly at odds with an inferred long history within this biome (Nielsen et al. 2016); particularly compared to patterns of high structure observed across the Pilbara in the spinifex grasses they inhabit (Anderson et al. 2016) and other gecko taxa distributed in the AAZ (Pepper et al. 2008; Pepper et al. 2013).

The especially long branch from the divergence of S. jeanae from other phasmid geckos to within-lineage diversity could result from extinction of other lineages, or may instead reflect long-term connectivity of this widespread population, perhaps facilitated by the nature and more continuous distribution of its spinifex habitat within this biome. Either way, the comparative patterns of genetic diversity in phasmid geckos between the AMT and AAZ biomes cannot necessarily be explained by differences in age of the two

Chapter 2: Role of biome ecology in shaping diversification patterns

biomes, considering evidence for a long history of Strophurus in arid biomes (Nielsen et al. 2016). Furthermore, mounting evidence for relatively young (mid- to late-Miocene) crown-ages of vertebrate lineages within the Kimberley and Top End regions suggests diversity within the Monsoonal Tropics biome is actually relatively young (see also

Chapters 3 and 4). My findings suggest that the heterogeneity of the landscape, and potentially the associated differences in distribution of spinifex within the AMT has played an important role in facilitating diversification and persistence of short-range endemic lineages within the more topographically complex escarpment regions.

2.4.3. Conclusions

Spinifex-specialised phasmid Strophurus geckos exhibit higher and older genetic structuring associated with rock regions in the AMT and wider distributions with shallower divergences in the AAZ which do not reflect the inferred long history within this arid biome. Despite shallow present structure in the AAZ, the old crown age of the phasmid geckos, along with old diversity within the AMT support the growing body of data indicating that at least seasonally arid environments may have a long history within

Australia dating back well into the Miocene. Finally, this study emphasizes biome age or history of occupancy alone cannot explain the especially high genetic diversity being uncovered in the AMT compared to the AAZ, and implies an important role of the topographically variable landscape in shaping and preserving diversity within the AMT.

Chapter 2: Role of biome ecology in shaping diversification patterns

Table 2.1. a) Environmental predictors used in species distribution models (SDMs), and b) percentage contribution of predictors to SDMs for the three major clades of phasmid Strophurus geckos.

Table 2.1. a)

Variable Theme Attribute Broad-scale PTA Moisture Annual precipitation PTI Moisture Precipitation of driest month PTX Moisture Precipitation of wettest month PTSL1 Moisture Rainfall seasonality Dec-Feb v Jun-Aug (Williams et al., 2010) TNI Temperature Daily minimum temperature of coldest month TXX Temperature Daily maximum temperature of hottest month Local-scale CTIDEPTH Soil Soil depth rescaled using TWI (Teng et al., 2005) WII Soil Wilford weathering index (Wilford, 2012) FGreen Vegetation fPAR (Fraction of Photosynthetically Active Reflectance; Mackey et al., 2012) RSM Radiation Annual mean shortwave radiation (Reside et al., 2013) SLOPE Terrain Slope

Table 2.1. b)

AAZ AMT AAZ Top End/Northern Deserts Kimberley Variable % Contribution Variable % Contribution Variable % Contribution PTA 39.70 PTA 25.30 PTA 60.70 PTI 22.80 PTI 15.10 PTI 8.30 PTX 1.70 PTX 1.40 PTX 5.10 PTSL1 1.20 PTSL1 1.30 PTSL1 3.10 RSM 14.90 RSM 4.40 RSM 1.30 TNI 2.00 TNI 0.60 TNI 0.20 TXX 5.70 TXX 0.10 TXX 0.10 CTIDEPTH 3.20 CTIDEPTH 3.90 CTIDEPTH 0.70 WII 1.20 WII 18.00 WII 0.10 FGreen 5.00 FGreen 10.90 FGreen 1.40 SLOPE 2.60 SLOPE 18.90 SLOPE 19.10

Chapter 2: Role of biome ecology in shaping diversification patterns

Table 2.2. Prior and posterior distributions for crown-age estimates of major lineages within the Australian phasmid Strophurus complex and related taxa based on three different combinations of calibration strategy and data: i) nuclear (PDC, PRLR and RAG1) dataset, with secondary calibrations; ii) combined nuclear (PDC, PRLR and RAG1) and mitochondrial (ND2 minus 3rd codon) dataset, with secondary calibrations; and iii) mitochondrial (ND2) data, calibrated at rate of 3% pairwise per million years.

Nuclear Combined ND2 (3%) Priors Root (Uniform) 75 (54–96) 75 (54–96) n.a. New Cal/Aus (Pseudothecadactylus) 43 (28–58) 43 (28–58) n.a. Core Diplodactylidae' 35 (25–45) 35 (25–45) n.a. ND2 pairwise divergence rate n.a. n.a. 3% Posteriors (crown ages) Core Diplodactylidae' 29 (21–37) 29 (21–36) 40 (26–58) Strophurus 19 (14–25) 20 (14–25) 27 (17–39) Spiny-tail Strophurus clade 13 (9–17) 15 (10–19) 15 (9–21) Phasmid Strophurus complex 16 (11–20) 15 (11–20) 25 (16–36) 1) Arid Zone – Strophurus jeanae n.a. n.a. n.a. 2) Kimberley/Ord – Strophurus mcmillani/robinsoni 4 (2–7) 4 (2–6) 11 (7–16) 3) Arnhem/N Deserts – Strophurus horneri/taeniatus 5 (3–7) 5 (2–7) 10 (6–14) n.a., no rate or calibration prior applied, or no estimate available

Chapter 2: Role of biome ecology in shaping diversification patterns

Figure 2.1. Distribution of all phasmid gecko specimen records in Australian museums, acquired from the Atlas of Living Australia online (ALA; http://ozcam.ala.org.au/). Strophurus jeanae (pink) – Arid Zone; S. mcmillani (blue) – Kimberley, S. robinsoni (yellow) – Ord Region; and S. horneri (purple) – Arnhem Land, S. taeniatus (green) – Northern Deserts. Approximate boundary of the Australian Arid Zone (AAZ) is indicated by the dotted line as modified from Byrne et al., (2008), whilst the Australian Monsoonal Tropics (AMT) spans the region to the north. Brown dashed regions indicate the boundaries of the Kimberley Plateau and Arnhem escarpment, whilst the blue dashed line surrounds the approximate Ord Region as modified from Catullo et al., (2014b). Photographs courtesy of Henry Cook, Ryan Francis, Stephen Richards, Brendan Schembri, and Stephen Zozaya.

Chapter 2: Role of biome ecology in shaping diversification patterns

Figure 2.2. SplitsTree networks of phasmid Strophurus geckos for a) mitochondrial (ND2), and b) concatenated nuclear (PDC, PRLR, RAG1) datasets. Number symbols indicate lineages within the mcmillani/robinsoni clade, circles are taxa identified as S. mcmillani, whilst squares are identified as S. robinsoni. Triangle symbols indicate lineages within the horneri/taeniatus clade (S. horneri – purple, S. taeniatus – green). Red ellipse in b) highlights samples of S. horneri nested within S. taeniatus. Maps display geographic distributions of the three major phasmid gecko lineages.

Chapter 2: Role of biome ecology in shaping diversification patterns

Figure 2.3. Chronogram of divergence dates between lineages of the Australian phasmid Strophurus complex and congeners estimated with BEAST v1.8.0 using the combined mitochondrial (ND2, 3rd codon removed) and nuclear (PDC, PRLR, RAG1) dataset and secondary calibrations. Major genetically divergent clades (≧8% Tamura- Nei divergence) within the phasmid geckos are highlighted, black circles on key nodes indicate Maximum-likelihood (RAXML) and posterior probability (MRBAYES) support values ≧95/1.0. Red shaded rectangles over key nodes indicate minimum to maximum of 95% highest posterior densities (HPD) from the various dating analyses for key nodes. Green column indicates range from minimum to maximum of 95% HPD estimates for the crown age of Triodia within Australia (Toon et al. 2015). Taxa labels in brown indicate lineages within the phasmid Strophurus clade.

Chapter 2: Role of biome ecology in shaping diversification patterns

References

Anderson BM, Barrett MD, Krauss SL, Thiele K (2016) Untangling a species complex

of arid zone grasses (Triodia) reveals patterns congruent with co-occurring

animals. Molecular Phylogenetics and Evolution 101, 142-162.

Baele G, Lemey P, Bedford T, et al. (2012) Improving the accuracy of demographic and

molecular clock model comparison while accommodating phylogenetic

uncertainty. Molecular Biology and Evolution 29, 2157-2167.

Baele G, Li WLS, Drummond AJ, Suchard MA, Lemey P (2013) Accurate model

selection of relaxed molecular clocks in Bayesian phylogenetics. Molecular

Biology and Evolution 30, 239-243.

Bauer AM, de Silva A, Greenbaum E, Jackman T (2007) A new species of day gecko

from high elevation in Sri Lanka, with a preliminary phylogeny of Sri Lanka

Cnemaspis (Reptilia, Squamata, Gekkonidae). Mitteilungen aus dem Museum

für Naturkunde in Berlin, Zoologische Reihe 83, 22-32.

Bowman DMJS, Brown GK, Braby MF, et al. (2010) Biogeography of the Australian

monsoon tropics. Journal of Biogeography 37, 201-216.

Branch WR (1998) Field Guide to Snakes and Other Reptiles of Southern Africa Struik

Publishers, Cape Town.

Brandley MC, Wang Y, Guo X, et al. (2011) Accommodating heterogenous rates of

evolution in molecular divergence dating methods: an example using

intercontinental dispersal of Plestiodon (Eumeces) lizards. Systematic Biology

60, 3-15.

Chapter 2: Role of biome ecology in shaping diversification patterns

Bryant D, Moulton V (2004) Neighbor-net: an agglomerative method for the

construction of phylogenetic networks. Molecular Biology and Evolution 21,

255-265.

Byrne M (2008) Evidence for multiple refugia at different time scales during

Pleistocene climatic oscillations in southern Australia inferred from

phylogeography. Quaternary Science Reviews 27, 2576-2585.

Byrne M, Steane DA, Joseph L, et al. (2011) Decline of a biome: evolution, contraction,

fragmentation, extinction and invasion of the Australian mesic zone biota.

Journal of Biogeography 38, 1635-1656.

Byrne M, Yeates DK, Joseph L, et al. (2008) Birth of a biome: insights into the

assembly and maintenance of the Australian arid zone biota. Molecular Ecology

17, 4398-4417.

Callaway RM, Hierro JL, Thorpe AS (2005) Evolutionary trajectories in plant and soil

microbial communities: Centaurea invasions and the geographic mosaic of

coevlution. In: Species invasions: insights into ecology, evolution and

biogeography (eds. Sax DF, Stachowicz JJ, Gaines SD), pp. 341-363. Sinauer,

Sunderland, MA.

Catullo RA, Doughty P, Keogh JS (2014a) A new frog species (Myobatrachidae:

Uperoleia) from the Northern Deserts region of Australia, with a redescription

of U. trachyderma. Zootaxa 3753, 251-262.

Catullo RA, Keogh JS (2014) Aridification drove repeated episodes of diversification

between Australian biomes: Evidence from a multi-locus phylogeny of

Australian toadlets (Uperoleia: Myobatrachidae). Molecular Phylogenetics and

Evolution 79, 106-117.

Chapter 2: Role of biome ecology in shaping diversification patterns

Catullo RA, Lanfear R, Doughty P, Keogh JS (2014b) The biogeographical boundaries

of northern Australia: evidence from ecological niche models and a multi-locus

phylogeney of Uperoleia toadlets (Anura: Myobatrachidae). Journal of

Biogeography 41, 659-672.

Chapple DG, Keogh JS (2004) Parallel adaptive radiations in arid and temperate

Australia: molecular phylogeography and systematics of the Egernia whitii

(Lacertilia: Scincidae) species group. Biological Journal of the Linnean Society

83, 157-173.

Christidis L, Rheindt FE, Boles WE, Norman JA (2010) Plumage patterns are good

indicators of taxonomic diversity, but not of phylogenetic affinities, in

Australian grasswrens Amytornis (Aves: Maluridae). Molecular Phylogenetics

and Evolution 57, 868-877.

Cracraft J (1991) Patterns of diversification within continental biotas: Hierarchical

congruence among the areas of endemism of Australian vertebrates. Australian

Systematic Botany 4, 211-227.

Crisp M, Cook L, Steane D (2004) Radiation of the Australian flora: what can

comparisons of molecular phylogenies across multiple taxa tell us about the

evolution of diversity in present-day communities? Philosophical Transactions

of the Royal Society B: Biological Sciences 359, 1551-1571.

Crisp MD, Arroyo MTK, Cook LG, et al. (2009) Phylogenetic biome conservatism on a

global scale. Nature 458, 754-756.

Crisp MD, Cook LG (2013) How was the Australian flora assembled over the last 65

million years? A molecular phylogenetic perspective. Annual Review of

Ecology, Evolution, and Systematics 44, 303-324.

Chapter 2: Role of biome ecology in shaping diversification patterns

de Vienne DM, Refrégier G, López-Villavicencio M, et al. (2013) Cospeciation vs host-

shift speciation: methods for tesing, evidence from natural assocations and

relation to coevolution. New Phytologist 198, 347-385.

Dessen B (2008) Fantastic Phasmids. The Australian Invertebrate Forum Newsletter 1,

2-4.

Dogra KS, Sood SK, Dobhal PK, Sharma S (2010) Alien plant invasion and their

impact on indigenous species diversity at global scale: a review. Journal of

Ecolog and the Natural Environment 2, 175-186.

Doughty P (2011) An emerging frog diversity hotspot in the northwest Kimberley of

Western Australia: another new frog species from the high rainfall zone.

Records of the Western Australian Museum 26, 209-216.

Doughty P, Ellis RJ, Oliver PM (2016) Many things come in small packages: Revision

of the clawless geckos (Crenadactylus: Diplodactylidae) of Australia. Zootaxa

4168, 239-278.

Drummond AJ, Ashton B, Cheung M, et al. (2008) Geneious v6.0, Available from

http://www.geneious.com/.

Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with

BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29, 1969-1973.

Duchene D, Klanten SO, Munday PL, Herler J, van Herwerden L (2013) Phylogenetic

evidence for recent diversification of obligate coral-dwelling gobies compared

with their host corals. Molecular Phylogenetics and Evolution 69, 123-132.

Eo SH, DeWoody JA (2013) Evolutionary rates of mitochondrial genomes correspond

to diversification rates and to contemporary species richness in birds and

reptiles. Proceedings of the Royal Society B: Biological Sciences 227.

Chapter 2: Role of biome ecology in shaping diversification patterns

Flot J-F (2010) SeqPHASE: a web tool for interconverting PHASE input/output files

and FASTA sequence alignments. Molecular Ecology Resources 10, 162-166.

Fu YX (1997) Statistical tests of neutrality of mutations against population growth,

hitchhiking and background selection. Genetics 147, 915-925.

Fujioka T, Chappell J, Fifield LK, Rhodes EJ (2009) Australian desert dune fields

initiated with Pliocene-Pleistocene global climatic shift. Geology 37, 51-54.

Fujita MK, McGuire JA, Donnellan SC, Moritz C (2010) Diversification and

persistence at the arid-monsoonal interface: Australia-wide biogeography of the

Bynoe's Gecko (Heteronotia binoei: Gekkonidae). Evolution 64, 2293-2314.

Glaw F, Vences M (2007) A Field Guide to the Amphibians and Reptiles of

Madagascar, 3rd edn. Vences & Glaw Verlag, Cologne.

Gordon CE, Dickman CR, Thompson MB (2010) What factors allow opportunistic

nocturnal activity in a primarily diurnal desert lizard (Ctenotus pantherinus)?

Comparative Biochemistry and Physiology, Part A 156, 255-261.

Groth JG, Barrowclough GF (1999) Basal divergences in birds and the phylogenetic

utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution 12,

115-123.

Haythornthwaite AS, Dickman CR (2006) Long-distance movements by a small

carnivorous marsupial: how Sminthopsis youngsoni (Marsupialia: Dasyuridae)

uses habitat in an Australian sandridge desert. Journal of Zoology 270, 543-549.

Heinicke MP, Greenbaum E, Jackman TR, Bauer AM (2012) Evolution of gliding in

Southeast Asian geckos and other vertebrates is temporally congruent with

dipterocarp forest development. Biology Letters, rsbl20120648.

Chapter 2: Role of biome ecology in shaping diversification patterns

Hugall AF, Foster R, Hutchinson M, Lee MSY (2008) Phylogeny of Australasian

agamid lizards based on nuclear and mitochondrial genes: implications for

morphological evolution and biogeography. Biological Journal of the Linnean

Society 93, 343-358.

Huson DH, Bryant D (2006) Application of phylogenetic networks in evolutionary

studies. Molecular Biology and Evolution 23, 245-267.

Hutter CR, Guayasamin JM, Wiens JJ (2013) Explaining Andean megadiversity: the

evolutionary and ecological causes of glassfrog elevational richness patterns.

Ecology Letters 16, 1135-1144.

Jabaily RS, Shepherd KA, Gardner AG, et al. (2014) Historical biogeogrpahy of the

predominantly Australian plant family Goodeniaceae. Journal of Biogeography

41.

Jennings WB, Pianka ER, Donnellan SC (2003) Systematics of the lizard family

Pygopodidae with implications for the diversification of Australian temperate

biotas. Systematic Biology 52, 757-780.

Keast JA (1981) Ecological biogeography of Australia W. Junk, Hague Boston, Boston.

King M, Horner P (1993) Family Gekkonidae. In: Fauna of Australia (eds. Glasby CG,

Ross GJB, Beesley PL), pp. 1-33. AGPS, Canberra, Australia.

Kuch U, Keogh JS, Weigel J, Smith LA, Mebs D (2005) Phylogeography of Australia's

king brown snake (Pseudechis australis) reveals Pliocene divergence and

Pleistocene dispersal of a top predator. Naturwissenschaften 92, 121-127.

Ladiges P, Parra-O C, Gibbs A, et al. (2011) Historical biogeographical patterns in

continental Australia: congruence among areas of endemism of two major clades

of eucalypts. Cladistics 27, 29-41.

Chapter 2: Role of biome ecology in shaping diversification patterns

Ladiges PY, Udovicic F, Nelson G (2003) Australian biogeographical connections and

the phylogeny of large genera in the plant family Myrtaceae. Journal of

Biogeography 30, 989-998.

Lanfear R, Calcott B, Ho SYW, Guindon S (2012) PartitionFinder: combined selection

of partitioning schemes and substitution models for phylogenetic analyses.

Molecular Biology and Evolution 29, 1695-1701.

Lee MSY, Oliver PM, Hutchinson MN (2009) Phylogenetic uncertainty and molecular

clock calibrations: A case study of legless lizards (Pygopodidae: Gekkota).

Molecular Phylogenetics and Evolution 50, 661-666.

Librado P, Rozas J (2009) DnaSP v5: A software for comprehensive analysis of DNA

polymorphism data. Bioinformatics 25, 1451-1452.

Losos JB (2009) Lizards in an Evolutionary Tree: Ecology and Adaptive Radiation of

Anoles University of California Press.

Macey JR, Larson A, Ananjeva NB, Fang Z, Papenfuss TJ (1997) Two novel gene

orders and the role of light-strand replication in the rearrangement of the

vertebrate mitochondrial genome. Molecular Biology and Evolution 14, 91-104.

Mackey B, Berry S, Hugh S, et al. (2012) Ecosystem greenspots: identifying potential

drought, fire, and climate-change micro-refuges. Ecological Applications 22,

1852-1864.

Marin J, Donnellan SC, Hedges SB, et al. (2013) Tracing the history and biogeography

of the Australian blindsnake radiation. Journal of Biogeography 40, 928-937.

Maryan B, Aplin KP, Adams M (2007) Two new species of the Delma tincta group

(Squamata: Pygopodidae) from northwestern Australia. Records of the Western

Australian Museum 23, 273-305.

Chapter 2: Role of biome ecology in shaping diversification patterns

Melville J, Ritchie EG, Chapple SNJ, Glor RE, Schulte II JA (2011) Evolutionary

origins and diversification of dragon lizards in Australia's tropical savannas.

Molecular Phylogenetics and Evolution 58, 257-270.

Miller MA, Pfeiffer W, Schwartz T (2010) Creating the CIPRES Science Gateway for

inference of large phylogenetic trees. In: Proceedings of the Gateway

Computing Environments Workshop (GCE), pp. 1-8, New Orleans, LA.

Mooney HA, Cleland EE (2001) The evolutionary impact of invasive species.

Proceedings of the National Academy of Science USA 98, 5446-5451.

Moritz C, Fujita MK, Rosauer D, et al. (2015) Multilocus phylogeography reveals

nested endemism in a gecko across the monsoonal tropics of Australia. Mol

Ecol.

Nielsen SV, Oliver PM, Laver RJ, Bauer AM, Noonan BP (2016) Stripes, jewels and

spines: further investigations into the evolution of defensive strategies in a

chemically defended gecko radiation (Strophurus, Diplodactylidae). Zoologica

Scripta.

Nix HA (1982) Environmental determinants of biogeography and evolution in Terra

Australis. Evolution of the flora and fauna of arid Australia 47.

Oliver PM, Adams M, Doughty P (2010) Molecular evidence for ten species and Oligo-

Miocene vicariance within a nominal Australian gecko species (Crenadactylus

ocellatus, Diplodactylidae). BMC Evolutionary Biology 10, 386.

Oliver PM, Adams M, Lee MSY, Hutchinson MN, Doughty P (2009) Cryptic diversity

in vertebrates: molecular data double estimates of species diversity in a radiation

of Australian lizards (Diplodactylus, Gekkota). Proceedings of the Royal Society

B: Biological Sciences 276, 2001-2007.

Chapter 2: Role of biome ecology in shaping diversification patterns

Oliver PM, Bauer AM (2011) Systematics and evolution of the Australian knob-tail

geckos (Nephrurus, Carphodactylidae, Gekkota): Plesiomorphic grades and

biome shifts through the Miocene. Molecular Phylogenetics and Evolution 59,

664-674.

Oliver PM, Couper PJ, Pepper M (2014a) Independent transitions between monsoonal

and arid biomes revealed by systematic revision of a complex of Australian

geckos (Diplodactylus; Diplodactylidae). PLoS ONE 9, e111895.

Oliver PM, Hugall A, Adams M, Cooper SJB, Hutchinson MN (2007) Genetic

elucidation of cryptic and ancient diversity in a group of Australian

diplodactyline geckos; the Diplodactylus vittatus complex. Molecular

Phylogenetics and Evolution 44, 77-88.

Oliver PM, Laver RJ, Smith KL, Bauer AM (2014b) Long-term persistence and

vicariance within the Australian Monsoonal Tropics: The case of the giant cave

and tree geckos (Pseudothecadactylus). Australian Journal of Zoology 61, 462-

468.

Oliver PM, McDonald P (2016) Young relicts and old relicts: a novel paleoendemic

vertebrate from the Australian Central Uplands. Royal Society Open Science 3,

160018.

Oliver PM, Parkin T (2014) A new phasmid gecko (Squamata: Diplodactylidae:

Strophurus) from the Arnhem Plateau: more new diversity in rare vertebrates

from northern Australia. Zootaxa 3878, 037-048.

Oliver PM, Sanders K (2009) Molecular evidence for Gondwanan origins of multiple

lineages within a diverse Australasian gecko radiation. Journal of Biogeography

36, 2044-2055.

Chapter 2: Role of biome ecology in shaping diversification patterns

Oliver PM, Smith KL, Laver RJ, Doughty P, Adams M (2014c) Contrasting patterns of

persistence and diversification in vicars of a widespread Australian lizard

lineage (the Oedura marmorata complex). Journal of Biogeography 41, 2068-

2079.

Paterson AM, Banks J (2001) Analytical approaches to measuring cospeciation of host

and parasites: through a glass, darkly. International Journal for Parasitology 31,

1012-1022.

Pepper M, Doughty P, Arculus R, Keogh JS (2008) Landforms predict phylogenetic

structure on one of the world's most ancient surfaces. BMC Evolutionary

Biology 8, 152.

Pepper M, Doughty P, Fujita MK, Moritz C, Keogh JS (2013) Speciation on the rocks:

Integrated systematics of the Heteronotia spelea species complex (Gekkota;

Reptilia) from western and central Australia. PLoS ONE 8, e78110.

Pepper M, Doughty P, Hutchinson MN, Keogh JS (2011a) Ancient drainages divide

cryptic species in Australia's arid zone: Morphological and multi-gene evidence

for four new species of Beaked Geckos (Rhynchoedura). Molecular

Phylogenetics and Evolution 61, 810-822.

Pepper M, Fujita MK, Moritz C, Keogh JS (2011b) Palaeoclimate change drove

diversification among isolated mountain refugia in the Australian arid zone.

Molecular Ecology 20, 1529-1545.

Pepper M, Ho SYW, Fujita MK, Keogh JS (2011c) The genetic legacy of aridification:

Climate cycling fostered lizard diversification in Australian montane refugia and

left low-lying deserts genetically depauperate. Molecular Phylogenetics and

Evolution 61, 750-759.

Chapter 2: Role of biome ecology in shaping diversification patterns

Phillips SJ, Dudík M (2008) Modeling of species distributions with Maxent: new

extensions and a comprehensive evaluation. Ecography 31, 161-175.

Pianka ER (1981) Diversity and adaptive radiations of Australian desert lizards. In:

Ecological Biogeography of Australia (ed. Keast A).

Posada D, Crandall KA (2002) The effect of recombination on the accuracy of

phylogeny estimation. Journal of Molecular Evolution 54, 396-402.

Potter S, Bragg JG, Peter BM, Bi K, Moritz C (2016) Phylogenomics at the tips:

inferring lineages and their demographic history in a tropical lizard, Carlia

amax. Molecular Ecology.

Potter S, Eldridge MDB, Taggart DA, Cooper SJB (2012) Multiple biogeographical

barriers identified across the monsoon tropics of northern Australia:

phylogeographic analysis of the brachyotis group of rock-wallabies. Molecular

Ecology 21, 2254-2269.

Rabosky DL, Donnellan SC, Talaba AL, Lovette IJ (2007a) Exceptional among-lineage

variation in diversification rates during the radiation of Australia's most diverse

vertebrate clade. Proceedings of the Royal Society B: Biological Sciences 274,

2915-2923.

Rabosky DL, Reid J, Cowan MA, Foulkes J (2007b) Overdispersion of body size in

Australian desert lizard communities at local scales only: no evidence for the

Narcissus effect. Oecologia 154, 561-570.

Rambaut A, Suchard MA, Xie D, Drummond AJ (2014) Tracer v1.6, Available from

http://beast.bio.ed.ac.uk/Tracer.

Chapter 2: Role of biome ecology in shaping diversification patterns

Read K, Keogh JS, Scott IAW, Roberts JD, Doughty P (2001) Molecular phylogeny of

the Australian frog genera Crinia and Geocrinia and allied taxa (Anura:

Myobatrachidae). Molecular Phylogenetics and Evolution 21, 294-308.

Reside AE, VanDerWal J, Phillips BL, et al. (2013) Climate change refugia for

terrestrial biodiversity.

Rix MG, Edwards DL, Byrne M, et al. (2015) Biogeography and speciation of

terrestrial fauna in the south-western Australian biodiversity hotspot. Biological

Reviews 90, 762–793.

Robb J (1980) New Zealand Amphibians and Reptiles in Colour Collins, Auckland.

Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesian phylogenetic inference

under mixed models. Bioinformatics 19, 1572-1574.

Rosauer DF, Blom M, Bourke G, et al. (2016) Phylogeographic hotspots and

conservation priorities: an example from the Top End of Australia. In press.

Sakaguchi S, Bowman DMJS, Prior LD, et al. (2013) Climate, not Aboriginal landscape

burning, controlled the historical demography and distribution of fire-sensitive

conifer populations across Australia. Proceedings of the Royal Society B:

Biological Sciences 280, 20132182.

Schwartz A, Henderson RW (1991) Amphibians and Reptiles of the West Indies

University of Florida Press, Gainesville.

Shoo LP, Rose R, Doughty P, Austin JJ, Melville J (2008) Diversification patterns of

pebble-mimic dragons are consistent with historical disruption of important

habitat corridors in arid Australia. Molecular Phylogenetics and Evolution 48,

528-542.

Chapter 2: Role of biome ecology in shaping diversification patterns

Smith KL, Harmon LJ, Shoo LP, Melville J (2011) Evidence of constrained phenotypic

evolution in a cryptic species complex of agamid lizards. Evolution 65, 976-992.

Spawls S, Howell K, Drewes RC, Ashe J (2004) A Field Guide to the Reptiles of East

Africa: Kenya, Tanzania, Uganda, Rwanda and Burundi Bloomsbury Publishing

PLC, London, UK.

Stamatakis A (2006) RAxML-VI-HPC: Maximum likelihood-based phylogenetic

analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688-

2690.

Stamatakis A, Hoover P, Rougemont J (2008) A rapid bootstrap algorithm for the

RAxML web servers. Systematic Biology 57, 758-771.

Stephens M, Donnelly P (2003) A comparison of Bayesian methods for haplotype

reconstruction from population genotype data. American Journal of Human

Genetics 73, 1162-1169.

Stephens M, Smith N, Donnelly P (2001) A new statistical method for haplotype

reconstruction from population data. American Journal of Human Genetics 68,

979-989.

Storr GM (1978) Seven new gekkonid lizards from Western Australia. Records of the

Western Australian Museum 6, 337-352.

Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA

polymorphism. Genetics 123, 585-595.

Tamura K, Nei M (1993) Estimation of the number of nucleotide substitutions in the

control region of mitochondrial DNA in humans and chimpanzees. Molecular

Biology and Evolution 10, 512-526.

Chapter 2: Role of biome ecology in shaping diversification patterns

Tamura K, Stecher G, Peterson D, Filipski A, Kumar S (2013) MEGA6: Molecular

evolutionary genetics analysis version 6.0. Molecular Biology and Evolution 30,

2725-2729.

Teng J, Vaze NK, Tuteja NK, Gallant J (2005) CLASS spatial analyst: A GIS based

tool for distributed hydrological modelling. In: MODSIM 2005; International

Congress on Modelling and Simulation (eds. Zerger A, Argent RM), pp. 1485–

1491. Modelling and Simulation Society of Australia and New Zealand.

Toon A, Austin JJ, Dolman G, Pedler L, Joseph L (2012) Evolution of arid zone birds in

Australia: Leapfrog distribution patterns and mesic-arid connections in quail-

thrush (Cinclosoma, Cinclosomatidae). Molecular Phylogenetics and Evolution

62, 286-295.

Toon A, Crisp MD, Gamage H, et al. (2015) Key innovation or adaptive change? A test

of leaf traits using Triodiinae in Australia. Scientific reports 5.

Townsend TM, Alegre ER, Kelley ST, Wiens JJ, Reeder TW (2008) Rapid

development of multiple nuclear loci for phylogenetic analysis using genomic

resources: An example from squamate reptiles. Molecular Phylogenetics and

Evolution 47, 129-142.

Vriesendorp B, Bakker FT (2005) Reconstructing patterns of reticulate evolution in

angiosperms: what can we do? Taxon 54, 593-604.

Wilford J (2012) A weathering intensity index for the Australian continent using

airborne gamma-ray spectrometry and digital terrain analysis. Geoderma 183-

184, 124-142.

Wilgenbusch JC, Warren DL, Swofford DL (2004) AWTY: A system for graphical

exploration of MCMC convergence in Bayesian phylogenetic inference.

Chapter 2: Role of biome ecology in shaping diversification patterns

Williams KJ, Ferrier S, Rosauer D, et al. (2010) Harnessing continent-wide biodiversity

datasets for prioritising national conservation investment. In: A report prepared

for the Department of Sustainability, Environment, Water, Population and

Communities, Australian Grovernment, Canberra, by CSIRO Ecosystem

Sciences: Canberra – synthesis report.

Wilson S, Swan G (2013) A Complete Guide to Reptiles of Australia, 4th edn. New

Holland Publishers, Sydney.

Wilson SK (2012) Australian Lizards a Natural History CSIRO Publishing,

Collingwood, VIC, Australia.

Yang Z (2015) The BPP for species tree estimation and species delimitation. Current

Zoology 61.

Yang Z, Rannala B (2010) Bayesian species delimitaiton using multilocus sequence

data. Proceedings of the National Academy of Sciences 107, 9264-9269.

Zamudio KR, Greene HW (1997) Phylogeography of the bushmaster (Lachesis muta:

Viperidae): Implications for neotropical biogeography, systematics, and

conservation. Biological Journal of the Linnean Society 62, 421-442.

Chapter 2: Role of biome ecology in shaping diversification patterns

Appendix

Table A2.1. Summary of limbed gekkotan taxa exhibiting thin, contrasting longitudinal striped patterning.

Species Habitat Biogeographical region Reference Crenadactylus horni Spinifex Arid Zone, Australia Doughty et al. (2016) Crenadactylus naso Spinifex, vine thickets Monsoonal Tropics, Australia Doughty et al. (2016) Crenadactylus occidentalis Spinifex, shrubs Arid Zone, Australia Doughty et al. (2016) Crenadactylus pilbarensis Spinifex Arid Zone, Australia Doughty et al. (2016) Crenadactylus rostralis Spinifex Monsoonal Tropics, Australia Doughty et al. (2016) Crenadactylus tuberculatus Spinifex, shrubs Arid Zone, Australia Doughty et al. (2016) Cryptactites peringueyi Sedge, grass, Restio clumps Eastern Cape Province, South Africa Branch (1998) Goggia lineata Terrestrial, rubble and vegetation Western Cape Province, South Africa Branch (1998) Lygodactylus capensis Savannah and subtropical thicket Central and southern Africa Spawls et al. (2004) Lygodactylus grotei Arboreal, shrubs and bushes Southeast Africa Spawls et al. (2004) Lygodactylus kimhowelli Arboreal, coastal forest Tanzania, Africa Spawls et al. (2004) Lygodactylus mirabilis Grassland, ericoid thickets Madagascar, Africa Glaw & Vences (2007) Lygodactylus scheffleri uluguruensis Arboreal, shrubs and bushes Tanzania, Africa Spawls et al. (2004) Sphaerodactylus cochranae Leaf litter, bromeliads Dominican Republic, Hispaniola Schwartz & Henderson (1991) Sphaerodactylus roosevelti Arboreal, shrubs and bushes Southwest Puerto Rico Schwartz & Henderson (1991) Sphaerodactylus semasiops Bromeliads, rock and ground Jamaica Schwartz & Henderson (1991) Strophurus horneri Spinifex Monsoonal Tropics, Australia Wilson & Swan (2013) Strophurus jeanae Spinifex Arid Zone, Australia Wilson & Swan (2013) Strophurus mcmillani Spinifex Monsoonal Tropics, Australia Wilson & Swan (2013) Strophurus michaelseni Spinifex and sedge Arid Zone, Australia Wilson & Swan (2013) Strophurus robinsoni Spinifex Monsoonal Tropics, Australia Wilson & Swan (2013) Strophurus taeniatus Spinifex Northern Deserts, Australia Wilson & Swan (2013) Toropuku stephensi Palms and other vegetation New Zealand Robb (1980) Uroplatus lineatus Bamboo Eastern Madagascar Glaw & Vences (2007) Woodworthia chrysosireticus Flax and other vegetation New Zealand Robb (1980)

Chapter 2: Role of biome ecology in shaping diversification patterns

Table A2.2. Specimen numbers, localities, & GenBank accession details for a) ingroup specimens and b) outgroups included in molecular and phylogenetic analyses.

Table A2.2. a)

Specimen GenBank accession numbers Species Locality Lat Long number ND2 PDC PRLR RAG1 Strophurus horneri CCM2502 NT: Kakadu -13.64 132.60 √ √ √ – Strophurus horneri CCM2572 NT: Kakadu -12.86 132.98 √ √ √ – Strophurus horneri NMVD72591 NT: Yirrakak, Arnhemland Plateau -12.20 133.80 KU680191 KU680074 KU680134 – Strophurus horneri TS0053 NT: Narmargon Gorge, Arnhemland Plateau -12.91 132.93 KU680190 KU680073 KU680135 – Strophurus jeanae NMVD74413 NT: 35k N Sandfire Roadhouse -19.59 121.34 √ – – – Strophurus jeanae NMVD74414 NT: 35k N Sandfire Roadhouse -19.59 121.34 √ – – – Strophurus jeanae NMVD74415 NT: 35k N Sandfire Roadhouse -19.59 121.34 √ – – – Strophurus jeanae NTM R14930 NT: 12k SW Sangsters Bore -20.87 130.27 AY369005 – – – Strophurus jeanae SAMAR53984 NT: 11k S Wycliffe Well -20.90 134.22 KJ685532 – – – Strophurus jeanae WAMR100772 WA: Woodstock Station -21.61 118.96 √ √ √ √ Strophurus jeanae WAMR102051 WA: 32k SSE Mundabullangana Homestead -20.75 118.25 √ √ √ √ Strophurus jeanae WAMR102301 WA: Hermite Is (South), Montebello Islands -20.47 115.52 √ √ √ – Strophurus jeanae WAMR106161 WA: 32k SSW Turee Ck Homestead -23.87 118.57 √ √ √ √ Strophurus jeanae WAMR110614 WA: Tanami Desert -19.90 128.87 KU680195 KU680079 KU680140 KU679994 Strophurus jeanae WAMR131954 WA: Nifty Mine -21.67 121.58 √ √ √ – Strophurus jeanae WAMR135281 WA: Cape Lambert -20.59 117.18 √ √ √ – Strophurus jeanae WAMR146684 WA: 248k SSW Port Hedland -22.50 119.09 KU680197 KU680081 KU680141 KU679996 Strophurus jeanae WAMR154145 WA: Barrow Is -20.79 115.45 KU680196 KU680080 KU680142 KU679995 Strophurus jeanae WAMR154791 WA: 36k ESE Wittenoom -22.35 118.65 √ √ √ – Strophurus jeanae WAMR157049 WA: Fortescue Marsh -22.46 119.04 √ √ √ – Strophurus jeanae WAMR157072 WA: Fortescue Marsh -22.46 119.04 √ √ √ – Strophurus jeanae WAMR157338 WA: Tanami Desert -19.90 128.87 √ √ √ – Strophurus jeanae WAMR158113 WA: 20.5k WNW Mt Marsh -22.47 119.02 √ √ √ – Strophurus jeanae WAMR160004 WA: 46k NNE Whim Ck Hotel -20.48 118.00 √ √ √ – Strophurus jeanae WAMR160043 WA: 85k E Meentheena Outcamp -21.30 121.26 √ √ √ √ Strophurus mcmillani AMSR126185 WA: Mitchell Plateau, Upstream of Mitchell Falls -14.82 125.68 AY369008 KU680084 KU680145 KU680000 Strophurus mcmillani CCM0964 WA: 30k N King Edward Rvr Homestead, Theda -14.52 126.45 KU680200 KU680085 KU680146 – Strophurus mcmillani CCM0965 WA: 30k N King Edward Rvr Homestead, Theda -14.52 126.45 √ √ √ – Strophurus mcmillani WAMR164906 WA: Katers Is -14.46 125.52 √ √ √ √ Strophurus mcmillani WAMR164907 WA: Katers Is -14.46 125.52 √ √ √ – Strophurus mcmillani WAMR168891 WA: Bigge Is -14.60 125.12 √ √ √ – Strophurus mcmillani WAMR168892 WA: Bigge Is -14.60 125.12 KU680199 KU680083 KU680144 KU679999 Strophurus mcmillani WAMR171689 WA: Lachlan Is -16.62 123.47 √ √ √ – Strophurus mcmillani WAMR171690 WA: Unnamed Is -15.93 124.47 √ √ √ – Strophurus mcmillani WAMR171691 WA: Unnamed Is -15.91 124.46 KU680202 KU680086 KU680147 KU680001 Strophurus mcmillani WAMR172333 WA: Theda Station -14.81 126.51 KU680201 √ √ – Strophurus mcmillani WAMR172795 WA: Manning Gorge, Mt Barnett Station -16.66 125.93 KU680203 KU680087 KU680148 KU680002 Strophurus mcmillani WAMR172855 WA: Manning Gorge, Mt Barnett Station -16.66 125.93 √ √ √ – Strophurus robinsoni Data Pending NT: S Keep Rvr National Park -16.19 129.11 KU680206 KU680089 KU680152 KU680006 Strophurus robinsoni WAMR156743 WA: Mt Parker -17.17 128.31 KU680207 KU680090 KU680153 KU680007 Strophurus taeniatus BP00761 WA: Sunday Is -16.43 123.18 √ √ √ – Strophurus taeniatus CCM3074 WA: Mt Nyulasy -17.76 129.10 √ – – – Strophurus taeniatus NTMR36343 NT: Mt Sanford, Victoria Rvr Region -17.48 129.88 KU680216 KU680094 KU680163 KU680019 Strophurus taeniatus NTMR36750 NT: Wongalara Sanctuary -14.27 134.62 KU680213 KU680091 KU680164 KU680016 Strophurus taeniatus NTMR36751 NT: Wongalara Sanctuary -14.27 134.62 √ √ √ – Strophurus taeniatus SAMAR55298 QLD: Phosphate Hill -21.80 139.91 KU680214 KU680092 KU680165 KU680017 Strophurus taeniatus WAMR162452 WA: 20k W Kununurra -15.77 128.63 KU680215 KU680093 KU680166 KU680018

Chapter 2: Role of biome ecology in shaping diversification patterns

Table A2.2.b)

GenBank accession numbers Outgroup species Specimen number Locality Lat Long ND2 PDC PRLR RAG1 Amalosia lesueurii AMSR159546 NSW: Moonbi Lookout, Moonbi Ranges -30.99 151.08 JQ173642 JQ173688 – JQ173736 Amalosia obscura AMSR136124 WA: 4k NE Surveyors Pool, Mitchell Plateau -14.65 125.77 JQ173655 JQ173701 – JQ173748 Amalosia rhombifera SAMAR55604 QLD: Kroombit Tops -24.33 150.94 JQ173661 JQ173709 – JQ173755 Aprasia parapulchella AMSR127439 NSW: 20k N Tarcutta -35.17 147.88 GU459941 GU459741 – GU459539 Crenadactylus ocellatus SAMAR22245 NT: 10k S Barrow Ck -21.63 133.88 AY369016 JQ945367 – AY662627 Diplodactylus conspicillatus AMSR158426 NSW: Sturt National Park -29.38 142.04 JQ173627 JQ173673 – JQ173721 Diplodactylus granariensis AMSR150637 WA: Dedari -31.08 120.68 JQ173628 JQ173674 – JQ173722 Diplodactylus ornatus AMSR140546 WA: Denham -25.92 113.53 JQ173629 JQ173675 – JQ173723 Diplodactylus tessellatus AMSR143855 QLD: 7.9k SW Landborough Hwy on Boulia Rd -24.37 143.32 JQ173631 JQ173677 – JQ173725 Hesperoedura reticulata WAMR114757 WA: Kalgoorlie -30.87 121.25 KU680225 KU680099 KU680105 KU680028 Hoplodactylus duvaucellii FT(VUW)174 NZ: Mercury Is -36.63 175.90 GU459843 GU459642 KU680106 GU459440 Lucasium maini AMSR150647 WA: Dedari -31.08 120.68 JX024362 KU680100 – JX024503 Lucasium stenodactylum AMSR139897 WA: El Questro Station -16.02 128.00 JQ173630 JQ173676 – JQ173724 Mokopirirakau granulatus RAH363 NZ: Trass -41.41 172.92 GU459812 GU459611 – GU459409 Naultinus gemmeus RAH188 NZ: Otago Peninsula -45.89 170.62 GU459761 GU459560 – GU459358 Nebulifera robusta ABTC3938 QLD: near Rathdowney -28.21 152.86 JQ173662 JQ173710 KU680107 JQ173756 Oedodera marmorata AMSR161254 NC: Sommet Noir, Paagoumène, 11k NW Koumac -20.49 164.20 JQ173632 JQ173678 – JQ173726 Oedura castelnaui AMSR143917 QLD: 4.9k E Georgetown -18.28 143.59 JQ173633 JQ173679 – JQ173727 Oedura coggeri AMSR143918 QLD: 20k W Gulf Development Rd & Kennedy Hwy -18.13 144.63 JQ173635 JQ173681 – JQ173729 Oedura filicipoda AMSR126183 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.72 JQ173636 JQ173682 KU680108 JQ173730 Oedura gemmata NTMR35680 NT: Kakadu National Park -13.86 132.98 KJ803589 KJ803710 KU680109 KJ803710 Oedura gracilis WAMR172904 WA: Doongan Station -15.20 125.90 KU680226 KU680101 KU680110 KU680029 Oedura cincta SAMAR55902 QLD: 9k N NSW/QLD border on Mitchell Hwy -28.96 145.73 KJ803578 KJ803707 √ KJ803666 Oedura monilis SAMAR54560 QLD: Dawson Development Rd, 18k E Alpha T/O -24.53 146.55 JQ173653 JQ173699 – JQ173747 Oedura murrumanu NMVD76948 WA: Oscar Range, quarries E Leopold Downs Rd -17.92 125.30 KM016840 √ √ √ Oedura tryoni AMSR152045 NSW: Moonbi Lookout, Moonbi Ranges -30.99 151.08 JQ173663 JQ173711 – JQ173757 Paniegekko madjo MNHN1998.0467 NC: Mt Ignambi -20.47 164.60 GU459950 GU459751 – GU459549 Pseudothecadactylus australis QMJR57120 QLD: Heathlands Ranger Station -11.75 142.58 HQ288425 KU680102 – FJ855449 Pseudothecadactylus cavaticus WAMR171550 WA: Prince Regent Nature Reserve -15.76 125.26 KJ685523 KU680103 KU680111 KJ685534 Pseudothecadactylus lindneri MVZ99544 NT: Kakadu National Park -13.09 123.40 GU459946 GU459747 – GU459545 Rhacodactylus chahoua AMSR161238 NC: Dome de Tiebaghi, 14k NW Koumac -20.46 164.19 JQ173665 JQ173713 – JQ173759 Rhacodactylus leachianus AMB7189 NC: Ilot Moro -22.65 167.38 GU459949 GU459750 – GU459548 Rhynchoedura eyrensis AMSR155371 NSW: Sturt National Park -29.04 141.31 GU459954 GU459755 – GU459553 Strophurus ciliaris WAMR158369 WA: Giralia Station -22.89 114.56 KU680176 KU680062 KU680126 KU679972 Strophurus assimilis AMSR149832 WA: 17.6k W Bonny Vale Railway Station -30.80 120.98 JQ173666 JQ173714 KU680112 JQ173760 Strophurus assimilis WAMR126398 WA: Bungalbin Sandplain -30.28 119.75 KU680174 – KU680116 KU679969 Strophurus ciliaris AMSR147216 NT: Barkley Hwy -19.92 137.24 JQ173668 JQ173716 KU680118 JQ173762 Strophurus ciliaris NTMR16041 NT: 20k SW Katherine on Victoria Hwy -14.65 132.12 KU680185 KU680068 KU680120 – Strophurus ciliaris WAMR164701 WA: 15k SE Fitzroy Crossing -18.31 125.64 KU680180 KU680064 KU680117 KU679976 Strophurus elderi AMSR118604 NSW: 20k N Coombah Roadhouse -32.80 141.60 KU680186 KU680070 KU680128 KU679984 Strophurus elderi AMSR130987 NSW: 17.9k N Coombah Roadhouse -32.85 141.62 JQ173669 JQ173717 – JQ173763 Strophurus elderi SAMAR62147 WA: 18.4k NE Blackstone, Morgan Range -25.89 128.43 KU680188 KU680071 KU680131 KU679986 Strophurus intermedius SAMAR28963 SA: Gawler Ranges -32.60 136.20 KU680194 KU680077 KU680138 KU679992 Strophurus intermedius WAMR157858 WA: Balladonia Roadhouse -32.28 123.48 KU680193 KU680076 KU680139 KU679991 Strophurus krisalys SAMAR54523 QLD: Hughenden Dump -20.35 144.46 KU680198 KU680082 KU680143 KU679998 Strophurus rankini AMSR140490 WA: Coral Bay -23.13 113.77 JQ173670 JQ173718 – JQ173764 Strophurus rankini SAMAR22889 WA: Warroora Homestead -23.48 113.80 AY369002 – KU680151 KU680005 Strophurus spinigerus AMSR149815 WA: Buckland Hill -32.02 115.77 JQ173671 JQ173719 KU680155 JQ173765 Strophurus spinigerus WAMR154065 WA: Bindoon Military Training Area -31.23 116.30 KU680208 – KU680157 KU680008 Strophurus strophurus AMSR140536 WA: Denham -25.92 113.53 JQ173672 JQ173720 – JQ173766 Strophurus taenicauda ABTC27723 QLD: 5k E Miles -26.66 150.18 KU680217 – KU680167 KU680020 Strophurus taenicauda QMJR76397 QLD: Coominglah State Forest, near Monto -24.80 150.98 HQ171996 – KU680216 KU680021 Strophurus wellingtonae WAMR145495 WA: Lorna Glen Station -26.08 121.45 KU680219 KU680095 KU680169 KU680022 Strophurus williamsi QMJR48398 QLD: N Brigalow Belt, Townsville -19.27 146.82 KU680221 KU680097 KU680170 KU680024 Strophurus williamsi SAMAR25518 SA: Danggalli Conservation Park -33.60 140.88 AY369007 KU680096 KU680171 KU680023 Woodworthia maculatus RAH 292 NZ: Titahi Bay -41.06 174.50 GU459852 GU459651 – GU459449 Tissue sources are as follows: ABTC, Australian Biological Tissue Collection; AMB - Aaron M. Bauer - University of Villanova; AMS, Australia Museum; BP - Russel Palmer field number - lodged with Western Australia Museum (WAM); CCM; Craig C. Moritz collection - Australian National University; FT(VUW), frozen tissue collection of Victoria University of Wellington; MNHN, Museum National d'Histoire Naturelle; MVZ, Museum of Vertebrate Zoology - University of California at Berkeley; NMV, National Museum Victoria; NTM, Northern Territory Museum; QMJ, Queensland Museum; RAH, Rodney A. Hitchmough; SAMA, South Australia Museum; TS, Stuart Young field number - lodged with MAGNT (Museum and Art Gallery of the Northern Territory); WAM, Western Australia Museum. Locality acronyms: NC, New Caledonia; NSW, New South Wales; NT, Northern Territory; NZ, New Zealand; QLD, Queensland; SA, South Australia; WA, Western Australia.

Chapter 2: Role of biome ecology in shaping diversification patterns

Table A2.3. Characteristics of loci sequenced for molecular and phylogenetic analyses of the Australian phasmid Strophurus complex.

Locus bp n S V PI ND2 861 47 43 407 364 PDC 393 41 14 29 15 PRLR 561 41 20 79 59 RAG1 1065 13 14 41 27 bp, length (in base pairs); n, number of sequences; S, singleton sites; V, variable sites; PI, parsimony-informative sites

Chapter 2: Role of biome ecology in shaping diversification patterns

Table A2.4. Details of primers and protocols used for molecular dataset assembly.

Fragment Primer name Primer sequence (5' to 3') Source PCR conditions length ND2 ND2.F GCC CAT ACC CCG AAA ATS TTG Oliver et al. (2007) c. 900 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; 72 °C, ND2.R TTA GGG TRG TTA TTT GHG AYA TKC Oliver et al. (2007) 1 min); 72 °C, 5 min; hold at 15 °C L4437 AAG CTT TCG GGG CCC ATA CC Macey et al. (1997) c. 1200 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; 72 °C, tRNA-Asn CTA AAA TRT TRC GGG ATC GAG GCC Read et al. (2001) 1 min); 72 °C, 5 min; hold at 15 °C PDC PHO.F2 AGA TGA GCA TGC AGG AGT ATG A Bauer et al. (2007) c. 450 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; 72 °C, PHO.R1 TCC ACA TCC ACA GCA AAA AAC TCC T Bauer et al. (2007) 1 min); 72 °C, 5 min; hold at 15 °C PRLR PRLR.F1 GAC ARY GAR GAC CAG CAA CTR ATG CC Townsend et al. (2008) c. 600 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; 72 °C, PRLR.R3 GAC YTT GTG RAC TTC YAC RTA ATC CAT Townsend et al. (2008) 1 min); 72 °C, 5 min; hold at 15 °C RAG1 RAG1f.Stroph379 GTG AGA GGA GAC ATT GAY ACA Nielsen et al. (2016) c. 800 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; 72 °C, RAG1R18.R GAT GCT GCC TCG GTC GGC CAC CTT T Groth & Barrowclough (1999) 1 min); 72 °C, 5 min; hold at 15 °C RAG1R13.F TCT GAA TGG AAA TTC AAG CTG TT Groth & Barrowclough (1999) c. 800 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; 72 °C, RAG1r.Stroph840 AAG TGC TTG CAT GTT GTT TC Nielsen et al. (2016) 1 min); 72 °C, 5 min; hold at 15 °C

Chapter 2: Role of biome ecology in shaping diversification patterns

Table A2.5. Optimal partition and model schemes for each dataset and analysis type selected by Bayesian Information Criterion (BIC) scores.

Optimal partition and model scheme Dataset RAxML MrBayes/BEAST

ND2 ND2_1 (GTR+I+G); ND2_2 (GTR+I+G); ND2_3 (GTR+G) ND2_1 (GTR+I+G); ND2_2 (HKY+I+G); ND2_3 (GTR+G) PDC PDC_1 PDC_2 (GTR+G); PDC_3 (GTR+G) PDC_1 PDC_2 (HKY+G); PDC_3 (K80+G) PRLR PRLR_1 PRLR_2 PRLR_3 (GTR+G) PRLR_1 PRLR_2 PRLR_3 (K80+G) RAG1 RAG1_1 RAG1_2 (GTR+G); RAG1_3 (GTR+G) RAG1_1 RAG1_2 (HKY+G); RAG1_3 (K80+G) 3 gene concatenation PDC_1 PDC_2 RAG1_1 RAG1_2 (GTR+I+G); PDC_3 PDC_1 PDC_2 RAG1_1 RAG1_2 (GTR+I+G); PDC_3 (nuDNA) PRLR_1 PRLR_2 PRLR_3 RAG1_3 (GTR+G) (K80+G); PRLR_1 PRLR_2 PRLR_3 RAG1_3 (K80+G) ND2_1 (GTR+I+G); ND2_2 (GTR+I+G); ND2_3 (GTR+G); ND2_1 (GTR+I+G); ND2_2 (HKY+I+G); ND2_3 (GTR+G); 4 gene concatenation PDC_1 PDC_2 RAG1_1 RAG1_2 (GTR+G); PDC_3 PDC_1 PDC_2 RAG1_1 RAG1_2 (HKY+G); PDC_3 (mtDNA+nuDNA) (GTR+G); PRLR_1 PRLR_2 PRLR_3 RAG1_3 (GTR+G) (HKY+G); PRLR_1 PRLR_2 PRLR_3 RAG1_3 (K80+G)

Chapter 2: Role of biome ecology in shaping diversification patterns

Table A2.6. Mitochondrial genetic diversity and population structure within the Australian phasmid Strophurus complex.

Clade n Avg Min Max 8% d Arid Zone (Strophurus jeanae) 21 0.027 0.000 0.047 0 0.191 Arnhem Land/Northern Deserts 11 0.093 0.003 0.141 2 0.191 Kimberley/Ord Region 15 0.117 0.000 0.164 5 0.219 Arnhem/Northern Deserts lineages Strophurus horneri 4 0.056 0.012 0.092 0 0.125 Strophurus taeniatus 7 0.057 0.003 0.081 0 0.125 Kimberley/Ord Region lineages Strophurus mcmillani 1 5 0.021 0.000 0.025 0 0.105 Strophurus robinsoni 2 1 n.a. n.a. n.a. 0 0.105 Strophurus robinsoni 3 1 n.a. n.a. n.a. 0 0.131 Strophurus mcmillani 4 5 0.032 0.001 0.052 0 0.134 Strophurus mcmillani 5 3 0.068 0.008 0.091 0 0.144 n, number of samples; avg, average Tamura-Nei (TN) pairwise divergence between individuals; min, minimum TN divergence between individuals; max, maximum TN divergence between individuals; 8%, number of lineages showing mean divergence over 8% within grouping; d, average TN divergence to nearest relative; n.a., value not calculable from single sample.

Chapter 2: Role of biome ecology in shaping diversification patterns

Figure A2.1. Summary phylogenetic tree of Strophurus and related taxa adapted from Nielsen et al. (2016) summarising the phylogenetic distribution of graminicolous ecology, small body size and longitudinal striping. Evolutionary relationships and character evolution estimated using BEAST from concatenated mitochondrial (ND2, 3rd codon removed) and nuclear loci (PDC, PRLR, RAG1). Branch colours indicate ecology of taxa as labelled, orange taxon labels indicate longitudinal striped body patterning, and branch width is scaled proportionally to log maximum body-size of each species. Photographs courtesy of Jordan de Jong and Stephen Zozaya.

Chapter 2: Role of biome ecology in shaping diversification patterns

75/0.83

100/1.0 88/0.96

97/1.0 91/0.9

85/0.99 83/0.93 94/1.0

96/1.0 83/0.96 100/1.0 100/1.0 83/1.0 Oedura_murrumanu530

79/0.95 100/1.0

100/1.0 91/0.75 85/0.99

97/1.0 89/0.86

99/0.91

87/1.0

80/0.89

99/1.0 88/1.0

82/0.47 100/1.0 80/0.91 91/1.0

100/1.0

99/1.0 Phasmid Strophurus horneri345_ArnhemLand horneriCM280_Kakadu 100/1.0 horneri254_Namarvgon horneriCM275_Kakadu 93/1.0 93/1.0 99/1.0

Figure A2.2. Phylogenetic reconstruction for 100/1.0 Strophurus and outgroups based on a three loci concatenated nuclear dataset (PDC, PRLR, RAG1) generated in MrBayes. Topology is concordant with RAxML analyses. Node values with high support in either analysis are shown as maximum-likelihood bootstraps/posterior probabilities (≧75 BS/≧0.95 PP).

Chapter 2: Role of biome ecology in shaping diversification patterns

94/0.99 100/1.0 58/0.97 100/1.0 90/0.9 87/0.99 100/1.0 99/1.0

88/0.98 94/1.0 99/1.0 94/1.0 87/1.0

100/1.0 100/1.0 84/1.0 Oedura_murrumanu530 94/1.0 100/1.0 100/1.0 100/1.0 86/1.0 95/0.99

100/1.0 97/1.0 58/1.0 57/0.97 100/1.0 96/1.0 99/1.0 100/1.0 95/1.0 100/1.0

100/1.0 94/1.0 80/1.0 99/1.0 81/0.96 76/1.0 99/1.0

100/1.0 67/0.98 100/1.0 100/1.0 100/1.0 86/1.0 99/1.0 100/1.0

100/1.0

Phasmid Strophurus

98/1.0 100/1.0 horneri345_ArnhemLand horneriCM275_Kakadu horneri245_Namarvgon 100/1.0 horneriCM280_Kakadu 100/1.0

96/1.0 84/0.96 Figure A2.3. Phylogenetic reconstruction for Strophurus and outgroups based on a four loci 100/1.0 concatenated dataset (mitochondrial – ND2 100/1.0 [minus 3rd codons], nuclear – PDC, PRLR, RAG1) generated in MrBayes. Topology is concordant with 70/1.0 RAxML analyses. Node values with high support in 99/1.0 100/1.0 either analysis are shown as maximum-likelihood bootstraps/posterior probabilities (≧75 BS/≧0.95 PP).

100/1.0 Chapter 2: Role of biome ecology in shaping diversification patterns

100/1.0

100/1.0 63/0.93 93/0.99 100/1.0 99/1.0 63/0.93 88/0.99 75/0.81 81/0.98 87/0.89 98/1.0 74/0.95 97/1.0 100/1.0 48/0.99 Oedura_murrumanu530 85/0.98 99/1.0 87/1.0 94/1.0 100/1.0 94/1.0 100/1.0

98/1.0 97/1.0 100/1.0 98/1.0 98/1.0 88/1.0 96/1.0 89/1.0 49/0.99 88/1.0 100/1.0 100/1.0 98/1.0 80/1.0 100/1.0 97/1.0 100/1.0 100/1.0 100/1.0 horneri345_ArnhemLand 100/1.0 horneriCM275_Kakadu 100/1.0 horneri254_Namarvgon horneriCM280_Kakadu 100/1.0 90/1.0

100/1.0

Phasmid Strophurus

100/1.0 100/1.0

94/1.0 99/1.0 100/1.0

Figure A2.4. Phylogenetic reconstruction for Strophurus and outgroups based on a single mitochondrial locus (ND2) generated in MrBayes. Topology is concordant with RAxML analyses. Node values with high support in either analysis are shown as maximum-likelihood bootstraps/posterior probabilities (≧75 BS/≧0.95 PP).

Chapter 2: Role of biome ecology in shaping diversification patterns

89/0.99

100/1.0 92/0.99

94/1.0 86/1.0

91/0.97 Oedura_murrumanu530

63/0.99 82/0.99 82/1.0

89/0.99

99/1.0

46/0.95

93/1.0 79/0.90

60/0.98 96/0.99 horneri254_Namarvgon horneriCM280_Kakadu horneriCM275_Kakadu horneri345_ArnhemLand

88/0.95 Phasmid Strophurus

Figure A2.5. Phylogenetic reconstruction for Strophurus and outgroups based on one nuclear locus (PDC) generated in MrBayes. Topology is concordant with RAxML analyses. Node values with high support are shown as maximum-likelihood bootstraps/posterior probabilities (≧75 BS/≧0.95 PP).

Chapter 2: Role of biome ecology in shaping diversification patterns

95/1.0 93/1.0

96/0.98 Oedura_murrumanu530 100/1.0 98/1.0

78/0.98

100/1.0

98/1.0 78/1.0 100/1.0

100/1.0 100/1.0

92/1.0 Phasmid Strophurus

horneri345_ArnhemLand 100/1.0 horneriCM280_Kakadu horneri254_Namarvgon horneriCM275_Kakadu

99/1.0

92/1.0

Figure A2.6. Phylogenetic reconstruction for Strophurus and outgroups based on one nuclear locus (PRLR) generated in MrBayes. Topology is concordant with RAxML analyses. Node values with high support are shown as maximum-likelihood bootstraps/posterior probabilities (≧75 BS/≧0.95 PP).

Chapter 2: Role of biome ecology in shaping diversification patterns

100/1.0

100/1.0 99/1.0 88/0.84 94/1.0

99/1.0 100/1.0 91/0.85 84/0.99 61/1.0 52/0.96

71/1.0 59/0.99

81/0.9 72/0.98 95/1.0 Oedura_murrumanu530 80/0.95 100/1.0 79/0.99

93/1.0 98/1.0 100/1.0 95/0.85

99/1.0 66/0.96 95/0.98 Phasmid Strophurus

99/1.0

Figure A2.7. Phylogenetic reconstruction for Strophurus and outgroups based on one nuclear locus (RAG1) generated in MrBayes. Topology is concordant with RAxML analyses. Node values with high support in either analysis are shown as maximum-likelihood bootstraps/posterior probabilities (≧75 BS/≧0.95 PP).

Chapter 2: Role of biome ecology in shaping diversification patterns

Figure A2.8. Species distribution models (SDMs) estimated in MaxEnt using eleven environmental variables relating to climate, geology and vegetation features. Maps displayed are the final SDMs for the three major clades of phasmid Strophurus: a) S. mcmillani/robinsoni—Kimberley/Ord Region, b) S.horneri/taeniatus—Top End/Northern Deserts, and c) S. jeanae—Australian Arid Zone (AAZ).

Chapter 3: Contribution of climate and topography to diversification

The contribution of climate and topography to

diversification patterns in two clades of

Kimberley geckos (Oedura spp.)

Photograph courtesy of Stephen Zozaya

Chapter 3: Contribution of climate and topography to diversification

Abstract

Tropical savannahs cover ~20% of the world’s land surface, yet patterns of biodiversity within these biomes – and the underlying processes that have shaped them

– remain poorly understood. Recent work indicates many taxa in savannah landscapes are characterised by high levels of phylogeographic structuring, especially taxa associated with rocky regions. In this study I use mitochondrial and nuclear loci combined with morphological data to elucidate and contrast patterns of spatial and temporal divergence in two ecologically distinct and broadly sympatric clades of congeneric saxicoline lizards (genus Oedura) from the topographically complex

Kimberley region in the Australian Monsoonal Tropics (AMT). My results suggest similar crown ages for both lineages but otherwise contrasting (fragmented relictual vs. widespread persistence) patterns of diversification. One lineage consists of two allopatric, morphologically distinct populations that are reciprocally monophyletic at both mt- and nu-DNA. The other comprises a widespread species complex of parapatric lineages with a small number of deeply divergent populations and evidence of at least two introgression events at contact zones. Evidence of microendemism, persistence and diversification emerges in high-rainfall or geologically unique regions of the central- and south-west, and the lowland east Kimberley. Lineages become more widespread, with evidence of introgression at their boundaries, over a gradient of increasing aridity in the central and south-east Kimberley. The interaction between topography and

Miocene-to-Pleistocene climate change appears to have contributed to high genetic diversity in tropical savannahs, however ecological or physiological differences among lineages may have also driven idiosyncratic responses to these extrinsic factors.

Chapter 3: Contribution of climate and topography to diversification

3.1. Introduction

Tropical savannahs cover ~20% of the world’s land surface (Grace et al. 2006) and are particularly prominent in all major southern continents (Cerling et al. 1997; da Silva

& Bates 2002; Waeber et al. 2015). These biomes are thought to be relatively young, with evidence suggesting a rapid spread of C4 grasses away from equatorial regions world-wide during the mid- to late-Miocene roughly ~8 Ma (Cerling et al. 1997; Tipple

& Pagani 2007). This timing correlates with a global increase in aridity and fire frequency, suggesting a possible causal link between climatic and vegetational change

(Bond et al. 2008; Keeley & Rundel 2005). How the biota of savannah biomes has evolved and diversified through Quaternary climatic cycles remains poorly understood

(da Silva & Bates 2002; Mayle et al. 2000). Detailed phylogeographic analyses of various tropical savannah taxa are revealing cryptic and geographically localised diversity (Criscione et al. 2012; Lorenzen et al. 2012; Moritz et al. 2015; Werneck et al.

2012) and this seems especially pronounced in saxicoline (rock-specialist) mammals and reptiles (e.g. Domingos et al. 2014; Guarnizo et al. 2016; Nogueira et al. 2011;

Pepper et al. 2011b; Potter et al. 2012). Research into the diversity and evolutionary history of these biomes is particularly important as savannahs are becoming increasingly threatened by agricultural expansion, introduced species, and changing fire regimes (Bond & Parr 2010; Miles et al. 2006; Parr et al. 2014; Waeber et al. 2015).

The Australian Monsoonal Tropics (AMT; Fig. 3.1.) is one of the largest intact savannah regions in the world (Woinarski et al. 2007). Despite a lack of high mountains, the landscape is highly heterogeneous – comprising a mosaic of ancient exposed sandstone plateaus and interspersed lowland woodlands and grasslands

(Bowman et al. 2010; Pepper & Keogh 2014). Climate in the AMT is hot and seasonal,

Chapter 3: Contribution of climate and topography to diversification

with a predictable high-rainfall wet season and a much longer dry season (Bowman et al. 2010). Paleontological and phylogenetic evidence suggest a dynamic climatic history, akin to other savannah regions, with Pleistocene glacial cycles causing periodic transitioning between cool, dry and warm, wet conditions (Proske et al. 2014; Reeves et al. 2013). There is a suggestion that changing climatic conditions may have led to shifts in distribution of both savannah woodland and arid grassland habitat along a north- south cline, likely reflecting similar movement of the boundary delineating the AMT and the Australian Arid Zone (AAZ) to the south (Denniston et al. 2013; Frawley &

O'Connor 2010; Proske et al. 2014; Reeves et al. 2013; Woinarski et al. 2007). How climatic and habitat variation affected distributions of associated savannah taxa within the AMT is poorly understood, though evidence for strong fluctuations of spatial distributions is emerging (Potter et al. 2016).

The north-western component of the AMT, the Kimberley, is both geologically and topographically complex, and characterised by a strong rainfall and aridity gradient from very distinctly monsoonal central-west to semi-arid regions in the south and east

(Pepper & Keogh 2014; Wende 1997). “Hotspots” of localised endemism have been identified in the most heterogeneous areas of the Kimberley including the high-rainfall zone of the central-west coast and associated islands (e.g. see Chapters 2 and 4;

Doughty 2011; Köhler 2011a; Oliver et al. 2014c; Palmer et al. 2013; Slatyer et al.

2007), southern Devonian Reef System (DRS) limestone ranges (Cameron 1992; Oliver et al. in press; Oliver et al. 2014b), and the Ord Region separating the Kimberley

Plateau and Top End (e.g. see Chapter 2; Catullo et al. 2014; Criscione et al. 2012;

Eldridge et al. 2011; Moritz et al. 2015; Potter et al. 2012). In addition to these endemism hotspots, recent research has revealed many examples of mammals, reptiles,

Chapter 3: Contribution of climate and topography to diversification

amphibians and invertebrates (e.g. snails) in the AMT (and especially the Kimberley) which exhibit extremely fine-scale geographic structuring of mitochondrial (mtDNA) haplotypes (e.g. Catullo et al. 2014; Moritz et al. 2015; Potter et al. 2016; Smith et al.

2011), and this is particularly pronounced in low-dispersal or specialist taxa (Köhler

2011b; Oliver et al. 2010; 2014d; Potter et al. 2012). It is posited that Pleistocene glacial cycles caused significant cooling and drying of the AMT, and that observed patterns of fine-scale fragmented diversity in rock-associated lineages is indicative of persistence through such environmental change within topographically localised microrefugia; in contrast with more widespread extinction-colonisation patterns that tend to be noted in more generalist taxa (Byrne 2008). Multi-locus studies in the AMT to date have investigated the phylogeographic patterns of widespread generalists

(Catullo et al. 2014; Moritz et al. 2015; Potter et al. 2016); however, further work across ecologically specialised (e.g. saxicolous) taxa are required to better understand how historical climatic variation drove diversification within this biome.

Velvet geckos (Oedura) are a lineage of scansorial (climbing-specialist; including arboreal and saxicoline taxa) lizards that are widespread across tropical and arid

Australia. They are part of an endemic Australasian family of geckos (Diplodactylidae) with Gondwanan affinities. These lizards exhibit exceptionally high levels of mtDNA structuring across the AMT and mountain ranges of the AAZ (Oliver et al. 2014b;

2014d). Across most of the range of Oedura only a single lineage is present in any particular area; however, in the Kimberley there are two endemic divergent (not sister taxa; Oliver et al. 2014d) clades: i) the widespread O. gracilis that spans the entire

Kimberley region (Fig. 3.1.a), and ii) a comparatively restricted clade comprising two allopatric species—O. filicipoda (central-west coast) and O. murrumanu (south

Chapter 3: Contribution of climate and topography to diversification

Kimberley, DRS limestone ranges; Fig. 3.1.b). Though both Kimberley clades are saxicolous they display clear morphological differences, possibly indicating varying ecologies or different forms of rock-specialisation. Oedura filicipoda/O. murrumanu exhibit highly-specialised morphologies for climbing large, open rock-faces (i.e. elongated limbs, expanded toe pads; P. Skipwith pers. comm.). In comparison, O. gracilis are crevice-specialists displaying significant dorso-ventral flattening of the head and body (pers. obs.; P. Skipwith pers. comm.). Published mitochondrial data suggests that O. gracilis is a species complex; however, this previous work involved limited sampling and the significance of this lineage diversity was not tested using independent nuclear or morphological datasets (Oliver et al. 2014d). A third non-Kimberley clade, the O. marmorata complex (comprising five arboreal and saxicoline species; Oliver &

Doughty 2016), is allied to the two Kimberley endemic clades, and spans a wide distribution across the Top End of the AMT as well as rocky ranges and woodlands of the AAZ (Fig. 3.1.c). The three Oedura clades considered herein provide an opportunity to compare and contrast diversification patterns—the depth and geographic scale of genetic diversity—across related taxa, and further develop a framework to understand how climatic change and underlying topographic variation may have shaped regional patterns of diversity in ecologically specialised taxa.

In this chapter I aimed to assess how historical shifting of the climatic gradient

(mesic–arid) at the AMT/AAZ boundary may have shaped accumulation of diversity in two congeneric but ecologically different saxicoline gecko clades (Oedura) within the

Kimberley. I predicted that, in response to Pleistocene climate change, rock specialist taxa would demonstrate signatures of persistence in micro-refugia offered by the topographic complexity of the region, in contrast to genetic evidence of expansion-

Chapter 3: Contribution of climate and topography to diversification

contraction dynamics typically observed in more generalist, terrestrial species. I investigated this using three approaches: i) I first compiled an expanded (mitochondrial and nuclear loci, and morphology) dataset to assess the taxonomic relevance of highly structured mtDNA diversity within the widespread Kimberley endemic, O. gracilis; ii) I then used increased molecular and taxonomic sampling to compare timeframes of diversification for the two endemic Kimberley Oedura clades (O. filicipoda/murrumanu and O. gracilis), and iii) I finally placed the endemic diversity observed within the

Kimberley and savannah regions of north-western Australia into a broader continental- scale context by comparing levels of diversity and range sizes of lineages in the

Kimberley to those of the closely related O. marmorata complex from the Top End

(AMT) and AAZ (Fig. 3.1.c). I hypothesised that the mesic (high-rainfall) and/or topographically complex Top End and west Kimberley would exhibit the highest genetic diversity indicating localised persistence of rock-specialists in micro-refugia, whilst more arid regions such as the south/south-east Kimberley would show lower and more widespread genetic diversity, reflecting the expansion and contraction of the arid biome.

3.2. Materials & methods

3.2.1. Sampling

Details of tissues used in genetic analyses are covered in the Appendix: Table A3.1.

My sampling included the vast majority of tissues registered in Australian museums, as well as my own recent collections, for the three recognised endemic Kimberley Oedura

(O. filicipoda—6 tissues, O. gracilis—107, and O. murrumanu—11) spanning 64 unique localities (many of them newly surveyed; Fig. 3.1.a–b). I used mitochondrial (mtDNA)

Chapter 3: Contribution of climate and topography to diversification

sequences from 33 representatives across the O. marmorata complex for comparisons of lineage diversity between Kimberley endemic Oedura and related non-endemic taxa.

Additionally I included sequence data (mtDNA and selected nuclear loci [nuDNA:

PDC, PRLR, and RAG-1]) for 53 outgroup taxa (including 12 non-Kimberley Oedura) to calibrate nodes for dating analyses (Table A3.1.b).

I extracted genomic DNA from liver or tail tip samples using a Qiagen DNeasy extraction kit or a Qiaextractor (Qiagen, Valencia, CA), and sequenced a portion of the mtDNA locus ND2 for all samples. I then selected a subsample of major lineages within

Kimberley Oedura (83 individuals; see Table A3.1.) spanning geographic and mtDNA diversity, plus 11 outgroup samples, which I also sequenced for six segments of nuDNA loci (KIAA1549, KIF24, PDC, PRLR, PTPN12, RAG-1). Genetic data characteristics, and primer sequences and amplification protocols, are described in the Appendix

(Tables A3.2. and A3.3., respectively). I purified PCR products using 1uL of a 20% dilution of ExoSAP-IT (US78201, Amersham Biosciences, Piscataway, NJ), incubated at 37°C for 30 min, then 80°C for 15 min. A genetic services company (Macrogen,

Seoul, South Korea) sequenced each amplicon in both directions from the cleaned products. Using GENEIOUS v.6.1.7 (Drummond et al. 2008) I assembled and edited sequences, visually examining alignments and amino acids to ensure correct reading frame and translation. All new sequences are available on GenBank (Table A3.1.).

3.2.2. Nucleotide data and phylogenetic analyses

I initially assessed genetic diversity within Kimberley Oedura and congruence of genetic (mtDNA and nuclear [nuDNA]) datasets by estimating phylogenies for i) individual loci, ii) concatenated nuclear data (six loci), and iii) the combined seven loci

Chapter 3: Contribution of climate and topography to diversification

dataset (mtDNA + nuDNA). Nucleotide substitution models and partitioning strategies

(Table A3.4.) I selected in PARTITIONFINDER v1.1.1 (Lanfear et al. 2012) using

Bayesian information criterion (BIC).

I used Maximum-Likelihood (RAxML v8.0.24; Stamatakis 2006; Stamatakis et al.

2008) and Bayesian (MrBayes v3.2.2; Ronquist & Huelsenbeck 2003) analyses to estimate phylogenetic relationships, both of which I conducted on CIPRES Science

Gateway 3.1 for online phylogenetic analysis (Miller et al. 2010). I used the default

CIPRES settings for RAXML with the PARTITIONFINDER partitioning strategies and using the GTR-CAT model (Stamatakis et al. 2008), ceasing bootstrapping when MRE- bootstrapping criteria were reached. Bayesian analyses I ran with four independent

Markov chain Monte Carlo (MCMC) chains and 4 x 10 million generations, sampling every 1000. Then I used TRACER v1.6 (Rambaut et al. 2014) and ARE WE THERE YET

(AWTY; Wilgenbusch et al. 2004) to confirm convergence and log likelihood stability, and that effective sample sizes (ESS) were >100 for all parameters, after which I constructed Maximum Clade Credibility (MCC) trees following removal of 20% burn- in.

To investigate and compare diversity within Kimberley Oedura and related non-

Kimberley taxa (O. marmorata complex) I used MEGA v6.06 (Tamura et al. 2013) to calculate average, minimum, and maximum Tamura-Nei genetic divergences (Tamura

& Nei 1993) within and between major mtDNA (genetic divergence ≧ 8%) lineages.

Additionally I estimated approximate range sizes of major mtDNA lineages by mapping distributions in ArcGIS Online (Esri) then using the measuring tool to draw minimum polygons around data points. Single points I assigned 10 km2 ranges, and lineages from individual islands I assigned the island size for range (Gibson & McKenzie 2012).

Chapter 3: Contribution of climate and topography to diversification

3.2.3. Phylogenetic dating

Although previous work estimated crown ages of various Oedura clades from the

AMT and AAZ (Oliver et al. 2014d), my study builds on earlier datasets with inclusion of an extra nuclear locus and increased within-clade sampling, including the addition of a newly described species (O. murrumanu) from the Kimberley. I estimated diversification timeframes for endemic Kimberley clades and major lineages (O. filicipoda/O. murrumanu and O. gracilis), as well as related taxa occurring outside the

Kimberley (O. marmorata complex; see Fig. 3.1.c) using BEAST v1.8.0 (Drummond et al. 2012). Models and partitioning strategies were the same as in Bayesian analyses above (Table A3.4.). I used path-sampling and stepping-stone methods (Baele et al.

2012) to compare various clock (strict versus relaxed: uncorrelated lognormal) and speciation (Yule versus birth-death) models to determine those that best fit the data.

Final analyses used a strict clock with Yule process speciation prior and ran for 10 million generations, sampling every 1000. TRACER v1.6 (Rambaut et al. 2014) and

AWTY (Wilgenbusch et al. 2004) were again used to assess stability, convergence,

ESS >100; burn-in of 20% was then discarded, with remaining MCC trees summarised in TREEANNOTATOR v2.3.1 (Rambaut & Drummond 2002–2015).

I used three strategies to estimate divergence timescales within Oedura: i) nuDNA

(PDC, PRLR, and RAG-1) combined with secondary constraints (estimated mean and standard deviations) derived from previous fossil calibrated squamate phylogenies for key nodes (Gamble et al. 2015; Skipwith et al. 2016) – a normal root prior [mean 75

Ma; SD 13] (Oliver et al. 2010), and two normal priors for Diplodactylidae (‘core

Diplodactylidae’ [mean 35 Ma; standard deviation 6] and the New

Caledonian/Pseudothecadactylus clade [43 Ma; SD 9]); ii) a combined dataset (mtDNA

Chapter 3: Contribution of climate and topography to diversification

[minus third codons in case of saturation] + nuDNA) and above secondary calibrations; and iii) mtDNA constrained to 3% pairwise per one million years (normal distribution,

[mean 0.015; SD 0.003]) mean molecular rate of pairwise sequence divergence – indicated as potentially more realistic for squamates, and in particular diplodactylid geckos (Eo & DeWoody 2013; Oliver et al. 2010).

3.2.4. Establishing nominal evolutionarily distinct lineages

Previous work based on limited sampling revealed high mtDNA diversity within O. gracilis (Oliver et al. 2014d). Given my greatly expanded spatial sampling I aimed to better characterise this diversity by assessing whether it represents molecularly defined

(based on genetic divergence) evolutionarily significant units (ESUs; Moritz 1994).

Since all delimitation methods involve assumptions that can be violated in any given system, I used multiple approaches to delimit evolutionarily distinct units (Carstens et al. 2013). I then compared the results to assess their validity. Initially I took two approaches identify candidate lineages using a single mtDNA locus (ND2). The first strategy defined candidate evolutionarily significant units (CESUs) on the basis of mtDNA divergence – specifically all lineages with ≧ 8% Tamura-Nei (TN) genetic divergence (this being the minimum divergence observed between currently recognised species in this genus; Oliver et al. 2009; Oliver & Doughty 2016). This led to somewhat extreme splitting across the range of O. gracilis and could be viewed as an initial, heuristic way to define CESUs. My second approach used a common method for predicting nominal species; the general mixed Yule-coalescent model (GMYC;

Fujisawa & Barraclough 2013; Pons et al. 2006). The GMYC model detects the depth within the tree at which the branching pattern shifts from a Yule process (inter-specific)

Chapter 3: Contribution of climate and topography to diversification

to a coalescent process (intra-specific) to provide a hypothesis for species delimitation.

This analysis I ran in R (R Development Core Team, 2011) using the package ‘splits’

(Ezard et al. 2009) and an ultrametric O. gracilis summary tree of all unique mtDNA sequences (n=71; including identical sequences is known to cause issues; Monaghan et al. 2009) produced in BEAST v1.8.0 (Drummond et al. 2012).

3.2.5. Population structure

Phylogenetic analyses may not accurately reflect population-level processes in the presence of reticulation (Eckert & Carstens 2008; Edwards 2009; Edwards et al. 2016).

To account for this possibility I performed a clustering analysis on the nuDNA dataset

(six loci) for O. gracilis using multi-dimensional scaling (MDS) followed by principal components analysis (PCA). This method I chose over STRUCTURE (Pritchard et al.

2000) as it is more appropriate for this type of data (low diversity nuclear sequences) and because sample sizes for lineages were in certain cases too limited (n = 1). The

MDS analysis I conducted by creating a distance matrix in POFAD v1.07 (Joly &

Bruneau 2006) using the Genpofad measure, weighting data from each locus equally. I included all individuals with data for at least five of the six loci (n = 61). Finally I conducted a principal components analysis (PCA) of the distance matrix in MVSP v3.1

(Kovach 2007).

3.2.6. Coalescent-based analyses

Coalescent methods are potentially more appropriate for analysing multi-locus datasets, as they explicitly model coalescent variation among loci across recently diverged taxa (Fujita et al. 2012). I therefore further assessed support for candidate

Chapter 3: Contribution of climate and topography to diversification

lineages within O. gracilis using the species delimitation method implemented in

Bayesian Phylogenetics and Phylogeography (BP&P v3.2; Yang 2015). I conducted multiple analyses for joint Bayesian species delimitation and species tree estimation in

BP&P. For these analyses I tested two different starting trees (8% candidate ESUs – 16 groups, or GMYC clades – 6 groups; see Results). I estimated each starting tree topology using all samples with at least five nuDNA loci for the concatenated dataset

(mtDNA + six nuDNA loci) in *BEAST (Heled & Drummond 2010), run in BEAST v1.8.0 through the CIPRES portal (Miller et al. 2010). For this analysis I phased the nuDNA dataset (six loci) using PHASE v2.1 (Stephens & Donnelly 2003; Stephens et al. 2001) in conjunction with SEQPHASE online (Flot 2010). *BEAST analyses I ran for 20 million generations, sampling every 20,000. Then I used TRACER v1.6 (Rambaut et al. 2014) and AWTY (Wilgenbusch et al. 2004) to test convergence and stability, and

I conducted multiple repeat analyses to ensure consistency of the resulting topology.

Within BP&P population size parameters (θs) and divergence time at the root of the species tree (τ0) I assigned the gamma priors G(1, 10), with mean = 0.1, while the other divergence time parameters I assigned the Dirichlet prior (Yang & Rannala 2010: equation 2). I used the species delimitation rjMCMC algorithm 0 with finetune ε = 2, and the species model prior set to 0 so that symmetric trees were not favoured. Prior knowledge that divergences within the O. gracilis complex date to the late-

Miocene/early Pleistocene (see Results) and preliminary tests with BP&P indicated these parameter settings were appropriate and any changes, including differing starting trees had negligible effect on the results. Final analyses I ran for 100,000 generations, sampling every two, with burn-in of 8,000; and I ran two independent analyses with different starting seeds to confirm consistency.

Chapter 3: Contribution of climate and topography to diversification

3.2.7. Phenotypic analyses

To ascertain whether any genetically divergent populations in the taxonomically challenging O. gracilis group were phenotypically distinctive I measured morphometric

(e.g. size measures) and colouration (i.e. banding pattern) traits following standard methods (e.g. Oliver et al. 2014b; see Table A3.5. for list of measures). I had a total of

49 genotyped adult individuals (18 male, 31 female) measured representing majority of the 8% mtDNA CESUs and sampling localities. Voucher specimens were unavailable for two of the 8% mtDNA CESUs (Barnett River, and Lake Argyle; see Fig. 3.1.a), and a further two 8% mtDNA CESUs had limited sampling consisting solely of juveniles

(Augustus Is. and Boongaree Is.; see Fig. 3.1.a; four and one specimens, respectively).

These groups were therefore not included in analyses. I tested for sexual dimorphism using a multivariate analysis of covariance (MANCOVA) in SPSS v23.0 (IBM_Corp.

Released 2015) with sex as a fixed effect, assessing for variation in shape of males and females of equal size. I then analysed raw measurement data, and also normalised linear measures to account for size-related variation using log-shape ratio transformation

(Mosimann 1970). For each individual I calculated the geometric mean of all measurements, then I divided each linear measure by this mean to produce the shape ratios. My final dataset consisted of the log values for each shape ratio, and I analysed this data using a PCA undertaken in R (R Development Core Team, 2011). Quantitative measures such as number of dorsal bands I excluded from PCAs and instead compared using box and whisker plots, also in R.

Chapter 3: Contribution of climate and topography to diversification

3.3. Results

3.3.1. Phylogenetic relationships and divergence times

The two endemic Kimberley Oedura clades (O. filicipoda/O. murrumanu, the O. gracilis complex) and the allied non-Kimberley O. marmorata complex were each strongly supported as monophyletic in all analyses. The relationships among these three clades were not resolved in mtDNA phylogenies, although concatenated nuDNA phylogenies supported the three Kimberley taxa (O. filicipoda, O. gracilis, and O. murrumanu) as monophyletic (Fig. 3.2.). Oedura filicipoda and the recently described

O. murrumanu were strongly supported as sister taxa, but well differentiated from one another in mtDNA, nuDNA, and morphological (Oliver et al. 2014b) datasets.

Age estimates were broadly concordant across methods, and suggested crown radiations for the three focal Oedura clades (O. filicipoda/O. murrumanu, O. gracilis, and O. marmorata) occurred over a broadly contemporaneous timeframe (Table 3.1.).

Recognising the potential for mtDNA to inflate divergence times for deep branches I focus on the conservative nuDNA derived dating estimates. Late-Miocene diversification is estimated both within the Kimberley, for the O. filicipoda/O. murrumanu (mean estimate 7 Ma, 3–10 Highest Posterior Density [HPD]) and O. gracilis (6 Ma, 4–8 HPD) clades, and across much of continental Australia—the O. marmorata complex (7 Ma, 4–10 HPD; Table 3.1.).

3.3.2. Mitochondrial diversity

Phylogenetic analyses of mtDNA recovered particularly high genetic structure across the Kimberley within the widespread O. gracilis, while the deepest divergences were observed between O. filicipoda and O. murrumanu (Fig. 3.2.a). Diversity within the

Chapter 3: Contribution of climate and topography to diversification

geographically restricted O. murrumanu was shallow. Oedura filicipoda however comprised two lineages that were at least 8% divergent (Tamura-Nei [TN]): one restricted to the Mitchell Plateau (Filicipoda MP; Fig 3.1.b) and the second ranging from the southern edge of the Mitchell Plateau to the south of the Prince Regent River area (Filicipoda PR; Fig. 3.1.b). My two methods for establishing preliminary hypotheses of lineage limits within the more diverse O. gracilis complex implied a hierarchy of diversity, within which some deeper clades showed species-level divergences. I distinguished 16 candidate evolutionary significant units (CESUs) using an 8% mtDNA genetic divergence limit (8–15% mean TN; Figs. 3.1.a, 3.2.a, 3.3.a, see

Table 3.2. for comparative genetic diversity). All 8% CESUs were strongly supported in

Maximum-Likelihood and Bayesian analyses (100 bootstraps/1.00 posterior probabilities). The more conservative general mixed Yule-coalescent model (GMYC) analysis recovered six putative clades generally corresponding to broader geographic clusters of the 8% CESUs: East, Mitchell Plateau, South (King Leopold Range + West- coast), Augustus Is., North, and Central (see Fig. 3.3.; Fig. A3.1.).

Average range sizes for 8% CESU mtDNA lineages in the Kimberley were relatively small—O. filicipoda/O. murrumanu average = 693 km2 (12–1,804 minimum–maximum area), O. gracilis = 2,185 km2 (10–24,202)—compared to distributions of corresponding

(≥8% divergent) lineages in the allied non-Kimberley clade—O. marmorata complex, average = 163,418 km2 (10–863,692; Table 3.3.). The South (King Leopold Rng) 8%

CESU of O. gracilis had the most widespread distribution of all lineages within the

Kimberley region (~6x the distribution size of the second-most widespread lineage;

Table 3.3.a). Partitioning lineages differently to compare average range sizes of mtDNA lineages in the AMT (O. filicipoda, O. gracilis, O. murrumanu, and Top End and Gulf

Chapter 3: Contribution of climate and topography to diversification

O. marmorata lineages) to those in the AAZ (AAZ O. marmorata lineages only) indicated that average range sizes across the AMT were also notably smaller compared to ranges across the AAZ (Table 3.3.b).

3.3.3. Phylogenetics of nuclear loci

Phylogenetic analyses of the concatenated nuDNA dataset for O. gracilis recovered five major clades: East, West (Mitchell Plateau + West-coast), South (King Leopold

Rng), Augustus Is., and North/Central (four with high support; >80 BS/100 PP; see

Figs. 3.2.b, 3.3.). These clades are broadly similar to the GMYC mtDNA clades.

However, two localised discordances were apparent (i.e. strongly supported conflict in mtDNA versus nuDNA phylogenies): i) one 8% CESU in the central Kimberley

(Barnett River CESU, see Figs. 3.1.a, 3.2.) shares mtDNA affinities with lineages of the

West/South while nuDNA-based analyses place it with the North/Central lineages; and ii) an 8% CESU in the south Kimberley (Nyulasy CESU, see Figs. 3.1.a, 3.2.) shares mtDNA affinities with East lineages while nuDNA places it within the South (King

Leopold Rng) 8% CESU. Analyses for individual loci were limited in recovery of lineage structure within O. gracilis except for the Augustus Is. CESU and the East clade

(minus the Nyulasy CESU; Figs. A3.2–7.), whereas the concatenated mtDNA+nuDNA analysis recovered all mtDNA lineages and general topology with mostly strong support for lineages (>75 BS/100 PP; Fig. A3.8.)

The most divergent and well-supported divisions within the O. gracilis complex are the Augustus Is. 8% mtDNA CESU, which is estimated to have split from the rest of the complex in the late-Miocene (6 Ma, 4–8 HPD), and the East clade (minus the Nyulasy

8% CESU) which dates to the Plio-Pleistocene (2 Ma, 1–3 HPD; Table 3.1.).

100

Chapter 3: Contribution of climate and topography to diversification

3.3.4. Population structure – clustering and coalescent methods

PCA of the POFAD multi-dimensional scaling (MDS) method based on the six nuclear loci recovered six clusters (Fig. 3.3.; Fig. A3.9.). Four were consistent with the phylogenetic nuDNA clustering–East, West (Mitchell Plateau + West-coast), Augustus

Is., and North/Central–whilst the South was further split in two (King Leopold Rng, and

Nyulasy; see Fig. 3.3.).

When conducting species delimitation tests using Bayesian Phylogenetics and

Phylogeography (BP&P) I took a conservative approach and only considered support of posterior probabilities = 1.00 (PP). At the finest-scale splitting (16 [8% CESUs]) the evolutionary distinctiveness of 13 lineages was supported; the remaining three 8%

CESUs collapsed into a single Central Kimberley clade (see Fig. 3.3.). More conservative tests of only six GMYC mtDNA clades supported the evolutionary distinctiveness of all six. Based on these results, and following the molecular definition for recognition of evolutionarily significant units (Moritz 1994), I propose that the

Oedura gracilis complex comprises 14 ESUs as supported by BP&P (see Fig. 3.3.).

Localised cases of introgression aside, five broad clades were consistently recovered across analyses: East, West (Mitchell Plateau + West-coast), South (King Leopold

Rng), Augustus Is., and North/Central (indicated on the map in Fig. 3.3.), which probably represent candidate species.

3.3.5. Phenotypic variation

Tests indicated no sexual dimorphism within Oedura gracilis (F = 1.107, p = 0.358), hence phenotypic variation was analysed across all adult specimens. Majority of the variation within the complex was attributed to overall body-size (snout-vent-length)

101

Chapter 3: Contribution of climate and topography to diversification

when raw measures were considered (Fig. 3.4.). There was a trend towards the Mitchell

Plateau, North, and Nyulasy ESUs having larger body-size, although this is less obvious when broader geographic clades are compared, with signal most maintained for the

Mitchell Plateau ESU (Fig. 3.4.). PCA of the log-shape transformed data indicated considerable overlap of lineages and/or clades. However, the East clade and Mitchell

Plateau ESU show evidence of some distinction from O. gracilis of the central

Kimberley (Fig. 3.4.). PC1 explained 23.2% of the variation and was mostly weighted by finger- and toe- widths, whilst PC2 explained 15.7% of the variation, mostly loaded by head-depth and also slightly by finger-width. The East clade of O. gracilis appears to have wider digits, whilst the Mitchell Plateau ESU have narrower digits. PC3 (15.3% variation) was most driven by lengths of fingers and toes, which again were larger in the

East clade. PCs 4–7 each explained >5% of variation, however no further dispersion of clades were evident when these were considered, hence they are not shown here.

In colouration, the Mitchell Plateau ESU had a greater number of dorsal bands (7–

11), the East clade had fewer (5–6), and remaining O. gracilis spanning the geographic range in between (central Kimberley) had an intermediate number (5–7; Fig. 3.4.d).

Though the highly genetically divergent Augustus Is. ESU could not be included in formal morphological analysis (only juvenile specimens available), the banding pattern of these individuals was distinctive compared to other ESUs (banding pattern interspersed with broad light bands in Augustus Is., versus banding pattern interspersed with thin faded light bands in mainland lineages).

3.4. Discussion

Crown radiations for the two endemic saxicoline (rock specialised) clades of Oedura

102

Chapter 3: Contribution of climate and topography to diversification

in the Kimberley region (O. filicipoda/O. murrumanu, and O. gracilis) are both estimated to date to the late Miocene, whilst most within clade diversity (O. gracilis) dates to the Pliocene to early-Pleistocene. Oedura filicipoda and O. murrumanu occupy restricted allopatric distributions, and are strongly supported as distinct taxa by mtDNA, nuclear loci (nuDNA), and morphology (Oliver et al. 2014b). In contrast, the widespread O. gracilis comprises multiple parapatric and morphologically conservative lineages, many of which show conflicting partitioning across different molecular and morphological datasets and analyses. Despite these differences, consistent signals of short-range endemism and persistence emerge in mesic regions of the central-west

(including islands), and refugial ranges of the south-west Kimberley. This is in opposition to lower diversity, signatures of lineage break-down (introgression or collapse) and comparatively widespread distributions in more arid regions of the central and south-east Kimberley.

3.4.1. Emphasising the evolutionary distinctiveness and endemism of the central-west

Kimberley

The high-rainfall zone of the central-west Kimberley and nearby islands is considered a biodiversity hotspot with several endemic species of reptiles, amphibians, plants and snails recently described (Doughty 2011; Köhler 2011a; Maslin et al. 2013;

Palmer et al. 2013; Slatyer et al. 2007). Oedura filicipoda and O. gracilis are locally sympatric in the central-west Kimberley (a rare occurrence for this genus); and O. filicipoda appears to be endemic to the region. Here I have identified evidence of further endemism in the central-west; specifically, the Mitchell Plateau and Augustus Is. evolutionarily significant units (ESUs) were each consistently recovered across multiple

103

Chapter 3: Contribution of climate and topography to diversification

independent datasets (mtDNA, nuDNA, morphometric).

Patterns of snail endemism in the Kimberley islands are already recognised as exceptional (Gibson & Köhler 2012; Köhler 2011a) – however, the Augustus Is. O. gracilis adds to the presently small number of highly divergent vertebrate lineages recognised on Kimberley islands (see Chapter 4; Ellis 2016; Oliver et al. 2012; Palmer et al. 2013; Silva et al. in review); emphasising the biodiversity value of this region.

Recent work elsewhere in the AMT has also identified previously undetected divergent lineages on islands off the coast of the Northern Territory (Marin et al. 2013; Moritz et al. 2015; Potter et al. 2016) such that these islands represent an endemism hotspot

(Rosauer et al. 2016). Most of the Kimberley islands are close to the mainland (0.1–18.9 km; Gibson 2014), and only recently isolated by rising sea levels (~8 kya; De Deckker

& Yokoyama 2009; Palmer et al. 2013). However, genetic divergences between some island and mainland populations suggest their initial divergence substantially pre-dates glacial cycles (~late Miocene/early Plio-Pleistocene; e.g. this study, Chapter 4; Oliver et al. 2012). Understanding the divergence histories of these endemic lineages requires more work – but together, these studies point to future research into how topographical and climatic variability can contribute to diversification.

3.4.2. Persistence and divergence in isolated ranges of the semi-arid southern

Kimberley

Other “insular”-type environments exist across the Kimberley mainland in the form of isolated topographic refugia. Increased survey efforts have identified the presence of possible refugial regions in Devonian Reef System (DRS) limestone ranges in the south-west Kimberley with discoveries of multiple highly-restricted, divergent endemic

104

Chapter 3: Contribution of climate and topography to diversification

lineages (see Chapter 4; Cameron 1992; Oliver et al. in press; 2014b); some of which have sister taxa to the north. Oedura murrumanu is endemic to these ranges and is consistently supported across datasets as distinct from its sister species O. filicipoda, distributed to the north-west. The restricted, fragmented and relictual distribution patterns within this lineage contrasts with the comparatively widespread distribution of the genetically uniform South ESU of O. gracilis (see below) across the King Leopold

Range which intersects the distributions of O. filicipoda (north) and O. murrumanu

(south; see Fig. 3.1.a–b). The unique structure and microhabitats of the DRS limestone ranges inhabited by O. murrumanu have likely provided refugia within the southern

Kimberley where multiple saxicoline taxa have been able to persist during harsh periods of climatic cycling (Cameron 1992; Oliver et al. in press).

Contention exists over how the lowlands of the Ord River Basin in the south-east of the Kimberley function in regards to biogeographic processes. This zone is sometimes recognised as a break in distributions between the Kimberley and the Top End faunas, but is also known to support divergent endemic lineages between the two regions

(Catullo et al. 2014; Criscione et al. 2012; Eldridge et al. 2011; Moritz et al. 2015;

Potter et al. 2012; Woinarski 1992). The most divergent clade of the O. gracilis complex (East) spans the broader Ord Region and is consistently recovered across independent datasets as distinguishable from O. gracilis of the west Kimberley. My findings lend increasing support to the hypothesis that topographic and climatic factors have combined to isolate lineages occurring in the east- and west-Kimberley and that this has resulted in further endemism within the Ord Region (see Chapter 2; but also see discussion of introgression below).

105

Chapter 3: Contribution of climate and topography to diversification

3.4.3. Introgression and delimitation breakdown at the arid-monsoon biome interface

Late Pleistocene latitudinal and elevational habitat shifts have played a central role in shaping recent range shifts and population changes in rainforests (Carnaval et al. 2009;

Hoskin et al. 2011; Moussalli et al. 2009), temperate zones (Hewitt 1993; Wang et al.

2015), and arid regions (Byrne 2008; Kearns et al. 2014); however, the role such processes may have had in shaping diversity within the monsoonal biome is still poorly understood (although see Catullo et al. 2014; Moritz et al. 2015; Potter et al. 2016;

Sakaguchi et al. 2013). The southern Kimberley sits at the boundary of the Australian

Arid Zone (AAZ) and AMT. An already harsh environment with lower rainfall and higher aridity, this region would potentially be most impacted by historical climatic and environmental changes such as increased aridification and shifting savannah habitats associated with glacial dynamics (Pepper & Keogh 2014).

Historically limited sampling has resulted in a paucity of knowledge about diversity patterns within the relatively homogeneous landscape of the Kimberly’s central region

(Moritz et al. 2013; Pepper & Keogh 2014). Throughout the central Kimberley I observed low support for the distinctness of deep mtDNA lineages of O. gracilis using additional independent datasets. This suggests the high lineage structure detected in the central Kimberley by the mtDNA and BP&P delimitation method correspond to population structure within, rather than between, species. These patterns may be indicative of multiple micro-refugia within the heterogeneous rocky escarpment of the central Kimberley in which genetic diversity has persisted over long timescales (Byrne

2008). Where the distribution of the O. gracilis complex extends into its most arid habitats (e.g. towards the central-south and south-east Kimberley), lineage boundaries appear to degrade; with strongly supported discordances in the mtDNA versus nuDNA

106

Chapter 3: Contribution of climate and topography to diversification

partitioning of lineages (see Figs. 3.1.a, 3.2.; also Chapter 4). Introgression at lineage boundaries could arise through persistent hybridization of parapatric populations, or as a result of secondary contact due to repeated expansion-contraction events of allopatric populations. This is more likely to occur when formerly isolated populations reconnect in areas with histories of fluctuating climatic suitability (Moritz et al. 2015; Silva et al. in review), such as the southern Kimberley region bordering the AAZ. Greater sampling and larger molecular datasets are required to evaluate this hypothesis (e.g. Potter et al.

2016).

The most striking pattern observed here is that range sizes of genetically divergent lineages and clades increase along an aridity gradient, with the most widespread distributions occurring in the south and east Kimberley (see Fig. 3.2.a; also Chapters 2 and 4). In this aspect the Kimberley recapitulates at a fine-scale the pattern displayed by the widespread O. marmorata complex over a continental scale (Fig. 3.1.). The widespread, genetically uniform distribution of the South ESU of the O. gracilis complex may indicate persistence and connectivity across the King Leopold and Durack

Ranges orogens; i.e. these mountain ranges have functioned as additional southern refugia.

3.4.4. Discovery of cryptic diversity within a tropical savannah biome

Detection of highly structured phylogeographic patterns across tropical savannah biomes is increasing, and these patterns are particularly pronounced in saxicolous (rock specialist) gekkotan taxa (e.g. Domingos et al. 2014; Doughty et al. 2012; Guarnizo et al. 2016; Nogueira et al. 2011; Pepper et al. 2011a). My data contributes to this growing body of evidence. The fine scale geographic structuring of deeply divergent genetic

107

Chapter 3: Contribution of climate and topography to diversification

lineages of saxicoline taxa in the AMT compared to the AAZ implies that diversification processes in tropical savannah biomes are due to more than topographic structure alone (Moritz et al. 2015; Oliver et al. 2010; Wilmé et al. 2006). This is because an increasing number of taxa which occupy topographically complex regions within the AAZ biome exhibit lower structuring within these rock isolates than congeneric lineages in escarpment regions of the AMT (see Chapter 2; Kuch et al.

2005; Oliver et al. 2014a; Oliver et al. 2014d). Given this emerging pattern of contrasting diversity levels between montane regions of the AMT and AAZ I hypothesise that historical climatic change may be underpinning this variation.

Properly documenting and describing this taxonomic diversity poses major analytical and philosophical challenges. Whilst a multitude of species delimitation methods exist, the use of simplified models by definition require assumptions that are often violated in natural datasets, contributing to their limited ability to produce concordant conclusions

(Carstens et al. 2013). Detection of cryptic complexes is inherently difficult, due to functional morphological conservatism associated with occupying similar environmental niches (Florio et al. 2012). This hinders our ability to bring together multiple lines of evidence to support evolutionary distinctness, just as introgression tends to oppose evolutionary distinctness (Miralles & Vences 2013). My study encountered similar challenges, however, despite some discordance across delimitation methods I was able to identify hierarchical structuring within the Oedura gracilis complex. The unique size and sampling range of this dataset allowed me to consistently identify deeply divergent, independently evolving lineages, likely representative of currently undescribed species. Despite the difficulties surrounding delimitation of lineage boundaries, systems like O. gracilis provide an excellent opportunity to

108

Chapter 3: Contribution of climate and topography to diversification

investigate the processes shaping diversification within heterogeneous tropical savannah landscapes.

The AMT biome – comprising a unique combination of heterogeneous landscape and dynamic climatic history – appears to foster diversification and persistence of numerous independently evolving lineages in microrefugia; maintaining relict taxa, and allowing development of complex community assemblages. My findings add to a growing body of work highlighting the under-estimation of biodiversity in tropical savannahs, which may be comparable to levels recognised in well-studied tropical rainforest biomes

(Carnaval et al. 2014; Gamble et al. 2012; Gehring et al. 2012; Leaché & Fujita 2010;

Miraldo et al. 2016; Pennington et al. 2006; Werneck et al. 2012). In regions of the world where sampling is already difficult, cryptic diversity therefore poses a significant challenge; particularly if ongoing population reticulation processes make it difficult to assess the taxonomic relevance of deep mtDNA diversity. This underestimation of savannah biodiversity is an important factor to consider when it comes to conservation in the face of threats to these biomes (e.g. agricultural industrialisation, introduced pest species, altered fire regimes, and climate change; Moritz et al. 2015; Rosauer et al.

2016; Waeber et al. 2015) – emphasising the need for increased surveying combined with more comprehensive genetic analyses.

109

Chapter 3: Contribution of climate and topography to diversification

Table 3.1. Prior and posterior probability distributions for the crown ages of the two endemic Kimberley clades of Oedura (O. filicipoda/O. murrumanu and O. gracilis) and the nearest related non-Kimberley clade (O. marmorata) based on three different combinations of calibration strategy and data: i) nuclear (PDC, PRLR and RAG1) dataset, with secondary calibrations; ii) combined nuclear (PDC, PRLR and RAG1) and mitochondrial (ND2 minus 3rd codon) dataset, with secondary calibrations; and iii) mitochondrial (ND2), calibrated at rate of 3% pairwise per one million years.

Nuclear Combined ND2 (3%) Priors

Root (Uniform) 75 (54–96) 75 (54–96) n.a. New Cal/Aus (Pseudothecadactylus) 43 (28–58) 43 (28–58) n.a. 'Core Diplodactylidae' 35 (25–45) 35 (25–45) n.a. ND2 pairwise divergence rate n.a. n.a. 3% Posteriors (crown ages)

'Core Diplodactylidae' 31 (24–39) 33 (25–40) 40 (26–58) Oedura 17 (12–22) 24 (18–29) 28 (18–41) Oedura marmorata clade 7 (4–10) 10 (7–13) 13 (8–20) Oedura filicipoda clade 7 (3–10) 10 (7–14) 13 (8–20) Oedura gracilis clade 6 (4–8) 11 (8–14) 13 (8–19) O. gracilis "Augustus Is. ESU" 6 (4–8) 9 (7–11) 11 (7–17) O. gracilis "East clade"* 2 (1–3) 4 (2–5) 6 (4–9) n.a., no rate or calibration prior applied, or no estimate available; *Nyulasy CESU not included.

110

Chapter 3: Contribution of climate and topography to diversification

Table 3.2. Genetic diversity data for major lineages within the Oedura gracilis complex and allied clades.

n mtDNA nuDNA Clade n (species) Avg Min Max 8% d n Avg Min Max d Oedura gracilis complex 107 ? 0.141 0.000 0.210 16 0.242 61 0.007 0.000 0.017 Oedura filicipoda /O. murrumanu 17 2 0.102 0.000 0.204 3 0.230 Oedura marmorata complex 33 5 0.140 0.000 0.209 10 0.230 Mitochondrial lineages 1) Lake Argyle 1 n.a. n.a. n.a. 0 0.076 1 n.a. n.a. n.a. 0.004 2) Bungle Bungles 3 0.010 0.002 0.015 0 0.076 3 0.000 0.000 0.000 0.004 3) Bullo River 6 0.008 0.000 0.023 0 0.095 3 0.001 0.001 0.001 0.004 4) Nyulasy 3 0.016 0.001 0.024 0 0.104 3 0.001 0.001 0.002 0.004 5) South 23 0.045 0.000 0.073 0 0.121 12 0.003 0.000 0.006 0.004 6) Mitchell Plateau 11 0.029 0.000 0.041 0 0.122 6 0.001 0.000 0.004 0.003 7) Boongaree Island 1 n.a. n.a. n.a. 0 0.092 1 n.a. n.a. n.a. 0.003 8) South-west 5 0.014 0.000 0.027 0 0.084 5 0.004 0.001 0.008 0.006 9) Barnett River 6 0.004 0.000 0.007 0 0.084 3 0.001 0.000 0.004 0.001 10) Police Valley 6 0.001 0.000 0.002 0 0.106 2 0.002 0.002 0.002 0.002 11) Mt Elizabeth 2 0.030 0.030 0.030 0 0.078 2 0.002 0.002 0.002 0.001 12) Doongan 7 0.003 0.000 0.006 0 0.078 3 0.004 0.001 0.007 0.004 13) Ellenbrae 11 0.012 0.000 0.039 0 0.093 4 0.003 0.001 0.005 0.004 14) Theda 6 0.032 0.001 0.047 0 0.114 4 0.003 0.000 0.006 0.004 15) North 12 0.048 0.000 0.056 0 0.122 5 0.002 0.000 0.005 0.004 16) Augustus Island 4 0.001 0.000 0.001 0 0.148 4 0.000 0.000 0.000 0.006 Nuclear populations* 1) East 10 0.063 0.000 0.130 3 0.167 7 0.004 0.000 0.006 0.009 2) South 26 0.069 0.000 0.170 2 0.131 15 0.003 0.000 0.007 0.007 3) West-coast 17 0.083 0.000 0.155 3 0.131 12 0.004 0.000 0.013 0.007 4) North/Central 50 0.114 0.000 0.175 7 0.155 23 0.005 0.000 0.010 0.008 5) Augustus Is 4 0.001 0.000 0.001 0 0.159 4 0.000 0.000 0.000 0.008 n, number of samples; avg, average Tamura-Nei (TN) pairwise divergence between individuals; min, minimum TN divergence between individuals; max, maximum TN divergence between individuals; 8%, number of lineages showing mean divergence over 8% within grouping; d, average TN divergence to nearest relative; n.a., value not calculable from single sample. *Nyulasy and Barnett River are included with nuclear sister lineages (South and North/Central respectively).

111

Chapter 3: Contribution of climate and topography to diversification

Table 3.3. a) Geographic range sizes of ≥8% Tamura-Nei divergent mtDNA lineages within the Kimberley endemic Oedura clades (O. filicipoda/O. murrumanu and O. gracilis) and the allied non-Kimberley clade (O. marmorata complex). b) Comparisons of average lineage range sizes within the AMT and between the AMT and AAZ.

Table 3.3. a)

Clade Lineage Area (km 2) Oedura filicipoda/O.murrumanu Fili_Nth (Mitchell Plateau) 12 Fili_Sth (Prince Regent) 1,804 Murrumanu (Oscar Rng) 262 minimum 12 mean 693 maximum 1,804 Oedura gracilis Lake Argyle 10 Augustus Is 190 Barnett River 70 Boongaree Is 42 Bullo River 85 Bungle Bungles 641 Doongan 160 Ellenbrae 4,302 Mitchell Plateau 2,112 Mt Elizabeth 99 North 1,865 Nyulasy 111 Police Valley 10 South-west 570 Theda 498 South 24,202 minimum 10 mean 2,185 maximum 24,202 Oedura marmorata Marmorata_1 129,766 Marmorata_2 10 Marmorata_3 290 Marmorata_4 17 Gemmata 15,195 Bella1 25,800 Bella2 13,900 Cincta_East 863,692 Cincta_Central 124,378 Fimbria 461,136 minimum 10 mean 163,418 maximum 863,692

112

Chapter 3: Contribution of climate and topography to diversification

Table 3.3. b)

Non-Kimberley Kimberley (marmorata AMT AAZ complex) minimum 10 10 10 124,378 mean 1,950 163,418 8,539 483,069 maximum 24,202 863,692 129,766 863,692

113

Chapter 3: Contribution of climate and topography to diversification

Figure 3.1. Distribution of comparative genetic diversity (mitochondrial–ND2) of endemic Kimberley Oedura—a) O. gracilis, and b) O. filicipoda and O. murrumanu— and allied non-endemic clade—c) O. marmorata complex. Symbol colours denote

114

Chapter 3: Contribution of climate and topography to diversification

lineages ≥8% Tamura-Nei divergent, shapes indicate different described taxa (see legends). AMT and AAZ regions are labelled in c) with AAZ boundary indicated (grey dotted line); red box in c) indicates inset of Kimberley encompassed in a) and b). Kimberley Plateau and Arnhem escarpments (brown dashed lines), and the Ord Region (blue dashed lines). Lineages within O. gracilis exhibiting mt-/nu-DNA discordance are labelled in a), and the southern Kimberley DRS limestone ranges discussed in the text are also indicated with an arrow in b).

115

Chapter 3: Contribution of climate and topography to diversification

Figure 3.2. Phylogenetic trees based on a) mitochondrial (ND2), and b) concatenated six nuclear loci (K1AA, KIF24, PDC, PRLR, PTPN12, RAG1) datasets. Strongly supported nodes are labelled with Bayesian posterior probabilities (PP) ≧0.95, and Maximum-Likelihood bootstraps (BS) ≧0.75; (PP/BS). Colour scheme over branches indicates ≧8% (mtDNA, Tamura-Nei) candidate evolutionarily significant units (CESUs) within the Oedura gracilis complex and corresponds to colour scheme in Fig. 3.1., whilst broader clades recovered in majority of analyses are indicated by the dotted bars beside tips (all labelled in legend). Stars highlight lineages with strongly supported discordant relationships (closest population affinities) in mitochondrial versus nuclear datasets.

116

Chapter 3: Contribution of climate and topography to diversification

Figure 3.3. Map of the Kimberley region of the Australian Monsoonal Tropics (AMT) showing distribution of genetic diversity of Oedura gracilis. Symbol colours indicate divergent mitochondrial (ND2) lineages (≥8% Tamura-Nei [TN]; see legend). Coloured

117

Chapter 3: Contribution of climate and topography to diversification

ellipses indicate broad clades of O. gracilis (labels in black) supported across majority of analyses and datasets. Lineages exhibiting mt-/nu-DNA discordance are labelled in blue and indicated with arrows. Simplified tree summarises diversity within O. gracilis indicating various delimitations of evolutionarily distinct units based on different analysis methods and datasets. Tree topology and node support (Bayesian Posterior Probabilities) derived from a *BEAST analysis of mitochondrial (ND2) data, and tip structure follows the most extreme splitting where candidate evolutionarily significant units (CESUs) were designated based on a minimum 8% TN divergence. The 16 CESUs are named after localities where lineages were originally detected (see map); colours of populations are matched to sampling points. From left to right clustering follows: two methods using mtDNA data only (≥8% divergence, and GMYC method); three methods incorporating nuDNA data only (phylogenetic concatenation, POFAD, BP&P); and a single analysis using morphological data. Number of clusters identified by each analysis is listed beneath each column.

118

Chapter 3: Contribution of climate and topography to diversification

Figure 3.4. Variation in morphometric (size and shape) measures and dorsal banding patterns for the Oedura gracilis complex. Principal Components Analyses are displayed for log-shape transformed measurements: a) PC1 vs PC2, and b) PC3 vs PC2; whilst PCA for raw linear measures including body size (snout-vent-length) is displayed in c) PC1 vs PC2. d) Indicates variation in number of dorsal bands. Convex hulls indicate the PCA space covered by geographic clades. Arrows are labelled at each axis to indicate the variables loading the principal components.

119

Chapter 3: Contribution of climate and topography to diversification

References

(R Development Core Team, 2011) R: A language and environment for statistical

computing. R Foundation for Statistical Computing. Available at: http://www.R-

project.org, Vienna, Austria.

Baele G, Lemey P, Bedford T, et al. (2012) Improving the accuracy of demographic and

molecular clock model comparison while accommodating phylogenetic

uncertainty. Molecular Biology and Evolution 29, 2157-2167.

Bauer AM, de Silva A, Greenbaum E, Jackman T (2007) A new species of day gecko

from high elevation in Sri Lanka, with a preliminary phylogeny of Sri Lanka

Cnemaspis (Reptilia, Squamata, Gekkonidae). Mitteilungen aus dem Museum

für Naturkunde in Berlin, Zoologische Reihe 83, 22-32.

Bond WJ, Parr CL (2010) Beyond the forest edge: ecology, diversity and conservation

of the grassy biomes. Biological Conservation 143, 2395-2404.

Bond WJ, Silander Jr. JA, Ranaivonasy J, Ratsirarson J (2008) The antiquity of

Madagascar's grasslands and the rise of C4 grassy biomes. Journal of

Biogeography 35, 1743-1758.

Bowman DMJS, Brown GK, Braby MF, et al. (2010) Biogeography of the Australian

monsoon tropics. Journal of Biogeography 37, 201-216.

Byrne M (2008) Evidence for multiple refugia at different time scales during

Pleistocene climatic oscillations in southern Australia inferred from

phylogeography. Quaternary Science Reviews 27, 2576-2585.

Cameron RAD (1992) Land snail faunas of the Napier and Oscar Ranges, Western

Australia; diversity, distribution and speciation. Biological Journal of the

Linnean Society 45, 271-286.

120

Chapter 3: Contribution of climate and topography to diversification

Carnaval AC, Hickerson MJ, Haddad CFB, Rodrigues MT, Moritz C (2009) Stability

predicts genetic diversity in the Brazilian Atlantic forest hotspot. Science 323,

785-789.

Carnaval AC, Waltari E, Rodrigues MT, et al. (2014) Prediction of phylogeographic

endemism in an environmentally complex biome. Proceedings of the Royal

Society B: Biological Sciences 281, 20141461.

Carstens BC, Pelletier TA, Reid NM, Satler JD (2013) How to fail at species

delimitation. Molecular Ecology 22, 4369-4383.

Catullo RA, Lanfear R, Doughty P, Keogh JS (2014) The biogeographical boundaries of

northern Australia: evidence from ecological niche models and a multi-locus

phylogeney of Uperoleia toadlets (Anura: Myobatrachidae). Journal of

Biogeography 41, 659-672.

Cerling TE, Harris JM, MacFadden BJ, et al. (1997) Global vegetation change through

the Miocene/Pliocene boundary. Nature 389, 153-158.

Criscione F, Law ML, Köhler F (2012) Land snail diversity in the monsoon tropics of

Northern Australia: revision of the genus Exiligada Iredale, 1939 (Mollusca:

Pulmonata: Camaenidae), with description of 13 new species. Zoological

Journal of the Linnean Society 166, 689-722. da Silva JMC, Bates JM (2002) Biogeographic patterns and conservation in the South

American Cerrado: a tropical savanna hotspot. BioScience 52, 225-234.

De Deckker P, Yokoyama Y (2009) Micro-palaeontological evidence for Late

Quaternary sea-level changes in Bonaparte Gulf, Australia. Global and

Planetary Change 66, 85-92.

121

Chapter 3: Contribution of climate and topography to diversification

Denniston RF, Wyrwoll K-H, Asmerom Y, et al. (2013) North Atlantic forcing of

millenial-scale Indo-Australian monsoon dynamics during the Last Glacial

period. Quaternary Science Reviews 72, 159-168.

Domingos FMCB, Bosque RJ, Cassimiro J, et al. (2014) Out of the deep: cryptic

speciation in a Neotropical gecko (Squamata, Phyllodactylidae) revealed by

species delimitation methods. Molecular Phylogenetics and Evolution 80.

Doughty P (2011) An emerging frog diversity hotspot in the northwest Kimberley of

Western Australia: another new frog species from the high rainfall zone.

Records of the Western Australian Museum 26, 209-216.

Doughty P, Palmer R, Sistrom MJ, Bauer AM, Donnellan SC (2012) Two new species

of Gehyra (Squamata: Gekkonidae) geckos from the north-west Kimberley

region of Western Australia. Records of the Western Australian Museum 27,

117-134.

Drummond AJ, Ashton B, Cheung M, et al. (2008) Geneious v6.0, Available from

http://www.geneious.com/.

Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with

BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29, 1969-1973.

Eckert AJ, Carstens BC (2008) Does gene flow destroy phylogenetic signal? The

performance of three methods for estimating species phylogenies in the presence

of gene flow. Molecular Phylogenetics and Evolution 49, 832-842.

Edwards SV (2009) Is a new and general theory of molecular systematics emerging?

Evolution 63, 1-19.

122

Chapter 3: Contribution of climate and topography to diversification

Edwards SV, Xi Z, Janke A, et al. (2016) Implementing and testing the multispecies

coalescent model: A valuable paradigm for phylogenomics. Molecular

Phylogenetics and Evolution 94, 447-462.

Eldridge MDB, Potter S, Cooper SJB (2011) Biogeographic barriers in north-western

Australia: an overview and standardisation of nomenclature. Australian Journal

of Zoology 59, 270-272.

Ellis RJ (2016) A new species of blindsnake (Scolecophidia: Typhlopidae: Anilios)

from the Kimberley region of Western Australia. Herpetologica 72, 271-278.

Eo SH, DeWoody JA (2013) Evolutionary rates of mitochondrial genomes correspond

to diversification rates and to contemporary species richness in birds and

reptiles. Proceedings of the Royal Society B: Biological Sciences 227.

Ezard T, Fujisawa T, Barraclough TG (2009) Splits: species' limits by threshold

statistics. R package version 1.

Florio AM, Ingram CM, Rakotondravony HA, Louis Jr. EE, Raxworthy CJ (2012)

Detecting cryptic speciation in the widespread and morphologically conservative

carpet chameleon (Furcifer lateralis) of Madagascar. Journal of Evolutionary

Biology 25, 1399-1414.

Flot J-F (2010) SeqPHASE: a web tool for interconverting PHASE input/output files

and FASTA sequence alignments. Molecular Ecology Resources 10, 162-166.

Frawley S, O'Connor S (2010) A 40,000 year wood charcoal record from Carpenter's

Gap 1: New insights into palaeovegetation change and indigenous foraging

strategies in the Kimberley, Western Australia. Altered Ecologies: Fire, Climate

and Human Influence on Terrestrial Landscapes (Terra Australis 32), 257-279.

123

Chapter 3: Contribution of climate and topography to diversification

Fujisawa T, Barraclough TG (2013) Delimiting species using single-locus data and the

Generalized Mixed Yule Coalescent approach: a revised method and evaluation

on simulated data sets. Systematic Biology 62, 707-724.

Fujita MK, Leaché AD, Burbrink FT, McGuire JA, Moritz C (2012) Coalescent-based

species delimitation in an integrative taxonomy. TRENDS in Ecology and

Evolution 27, 480-488.

Gamble T, Colli GR, Rodrigues MT, Werneck FP, Simons AM (2012) Phylogeny and

cryptic diversity in geckos (Phyllopezus; Phyllodactylidae; Gekkota) from South

America's open biomes. Molecular Phylogenetics and Evolution 62, 943-953.

Gamble T, Greenbaum E, Jackman TR, Bauer AM (2015) Into the light: diurnality has

evolved multiple times in geckos. Biological Journal of the Linnean Society 115,

896-910.

Gehring P-S, Tolley KA, Eckhardt FS, et al. (2012) Hiding deep in the trees: discovery

of divergent mitochondrial lineages in Malagasy chameleons of the Calumma

nasutum group. Ecology and Evolution 2, 1468-1479.

Gibson LA (2014) Biogeographic patterns on Kimberley islands, Western Australia.

Records of the Western Australian Museum 81, 245-280.

Gibson LA, Köhler F (2012) Determinants of species richness and similarity of species

composition of land snail communities on Kimberley islands Records of the

Western Australian Museum Supplement 81, 40-65.

Gibson LA, McKenzie NL (2012) Identification of biodiversity assets on selected

Kimberley islands: background and implementation. Records of the Western

Australian Museum Supplement 81, 1-14.

124

Chapter 3: Contribution of climate and topography to diversification

Grace J, José SJ, Meir P, Miranda HS, Montes RA (2006) Productivity and carbon

fluxes of tropical savannas. Journal of Biogeography 33, 387-400.

Groth JG, Barrowclough GF (1999) Basal divergences in birds and the phylogenetic

utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution 12,

115-123.

Guarnizo CE, Werneck FP, Giugliano LG, et al. (2016) Cryptic lineages and

diversification of an endemic anole lizard (Squamata, Dactyloidae) of the

Cerrado hotspot. Molecular Phylogenetics and Evolution 94, 279-289.

Heled J, Drummond AJ (2010) Bayesian inference of species trees from multilocus

data. Molecular Biology and Evolution 27.

Hewitt GM (1993) Postglacial distribution and species substructure: lessons from

pollen, insects and hybrid zones. Evolutionary patterns and processes 14, 97-

123.

Hoskin CJ, Tonione M, Higgie M, et al. (2011) Persistence in peripheral refugia

promotes phenotypic divergence and speciation in a rainforest frog. The

American Naturalist 178, 561-578.

IBM_Corp. (Released 2015) IBM SPSS Statistics for Windows. IBM Corp., Armonk,

NY.

Joly S, Bruneau A (2006) Incorporating allelic variation for reconstructing the

evolutionary history of organisms from multiple genes: an example from Rosa in

North America. Systematic Biology 55, 623-636.

Kearns AM, Joseph L, Toon A, Cook LG (2014) Australia's arid-adapted butcherbirds

experienced range expansions during Pleistocene glacial maxima. Nature

Communications 5.

125

Chapter 3: Contribution of climate and topography to diversification

Keeley JE, Rundel PW (2005) Fire and the Micoene expansion of C4 grasslands.

Ecology Letters 8, 683-690.

Köhler F (2011a) The Camaenid species of the Kimberley islands, Western Australia

(Stylommatophora: Helicoidea). Malacologia 54, 203-406.

Köhler F (2011b) Descriptions of new species of the diverse and endemic land snail

Amplirhagada Iredale, 1933 from rainforest patches across the Kimberley,

Western Australia (Pulmonata: Camaenidae). Records of the Australian Museum

63, 167-202.

Kovach WL (2007) MVSP - A MultiVariate Statistical Package for Windows, ver. 3.1.

Kovach Computing Services, Pentraeth, Wales, U.K.

Kuch U, Keogh JS, Weigel J, Smith LA, Mebs D (2005) Phylogeography of Australia's

king brown snake (Pseudechis australis) reveals Pliocene divergence and

Pleistocene dispersal of a top predator. Naturwissenschaften 92, 121-127.

Lanfear R, Calcott B, Ho SYW, Guindon S (2012) PartitionFinder: combined selection

of partitioning schemes and substitution models for phylogenetic analyses.

Molecular Biology and Evolution 29, 1695-1701.

Leaché AD, Fujita MK (2010) Bayesian species delimitation in West African forest

geckos (Hemidactylus fasciatus). Proceedings of the Royal Society B: Biological

Sciences 277, 3071-3077.

Lorenzen E, Heller R, Siegismund HR (2012) Comparative phylogeography of African

savannah ungulates. Molecular Ecology 63, 534-542.

Macey JR, Larson A, Ananjeva NB, Fang Z, Papenfuss TJ (1997) Two novel gene

orders and the role of light-strand replication in the rearrangement of the

vertebrate mitochondrial genome. Molecular Biology and Evolution 14, 91-104.

126

Chapter 3: Contribution of climate and topography to diversification

Marin J, Donnellan SC, Hedges SB, et al. (2013) Hidden species diversity of Australian

burrowing snakes (Ramphotyphlops). Biological Journal of the Linnean Society

110, 427-441.

Maslin BR, Barrett MD, Barrett RL (2013) A baker's dozen of new wattles highlights

significant Acacia (Fabaceae: Mimosoideae) diversity and endemism in the

north-west Kimberley region of Western Australia. Nuytsia 23, 543-587.

Mayle FE, Burbridge R, Killeen TJ (2000) Millennial-scale dynamics of southern

Amazonian rainf forests. Science 290, 2291-2294.

Miles L, Newton AC, Defries RS, et al. (2006) A global overview of the conservation

status of tropical dry forests. Journal of Biogeography 33, 491-505.

Miller MA, Pfeiffer W, Schwartz T (2010) Creating the CIPRES Science Gateway for

inference of large phylogenetic trees. In: Proceedings of the Gateway

Computing Environments Workshop (GCE), pp. 1-8, New Orleans, LA.

Miraldo A, Li S, Borregaard MK, et al. (2016) An Anthropocene map of genetic

diversity. Science 353, 1532-1535.

Miralles A, Vences M (2013) New metrics for comparison of taxonomies reveal striking

discrepancies among species delimitation methods in Madascincus lizards. PLoS

ONE 8, e68242.

Monaghan MT, Wild R, Elliot M, et al. (2009) Accelerated species inventory on

Madagascar using coalescent-based models of species delineation. Systematic

Biology 58, 298-311.

Moritz C (1994) Defining 'evolutionarily significant units' for conservation. TRENDS in

Ecology and Evolution 9, 373-374.

127

Chapter 3: Contribution of climate and topography to diversification

Moritz C, Ens EJ, Potter S, Catullo RA (2013) The Australian monsoonal tropics: an

opportunity to protect unique biodiversity and secure benefits for Aboriginal

communities. Pacific Conservation Biology 19, 343-355.

Moritz C, Fujita MK, Rosauer D, et al. (2015) Multilocus phylogeography reveals

nested endemism in a gecko across the monsoonal tropics of Australia. Mol

Ecol.

Mosimann JE (1970) Size allometry: size and shape variables with characterizations of

the lognormal and generalized gamma distributions. Journal of the American

Statistical Association 65, 930-948.

Moussalli A, Moritz C, Williams SE, Carnaval AC (2009) Variable responses of skinks

to a common history of rainforest fluctuation: concordance between

phylogeography and palaeo-distribution models. Molecular Ecology 18, 483-

499.

Nielsen SV, Oliver PM, Laver RJ, Bauer AM, Noonan BP (2016) Stripes, jewels and

spines: further investigations into the evolution of defensive strategies in a

chemically defended gecko radiation (Strophurus, Diplodactylidae). Zoologica

Scripta.

Nogueira C, Ribeiro Sr, Costa GC, Colli GR (2011) Vicariance and endemism in a

Neotropical savana hotspot: distribution patterns of Cerrado squamate reptiles.

Journal of Biogeography 38, 1907-1922.

Oliver PM, Adams M, Doughty P (2010) Molecular evidence for ten species and Oligo-

Miocene vicariance within a nominal Australian gecko species (Crenadactylus

ocellatus, Diplodactylidae). BMC Evolutionary Biology 10, 386.

128

Chapter 3: Contribution of climate and topography to diversification

Oliver PM, Adams M, Lee MSY, Hutchinson MN, Doughty P (2009) Cryptic diversity

in vertebrates: molecular data double estimates of species diversity in a radiation

of Australian lizards (Diplodactylus, Gekkota). Proceedings of the Royal Society

B: Biological Sciences 276, 2001-2007.

Oliver PM, Couper PJ, Pepper M (2014a) Independent transitions between monsoonal

and arid biomes revealed by systematic revision of a complex of Australian

geckos (Diplodactylus; Diplodactylidae). PLoS ONE 9, e111895.

Oliver PM, Doughty P (2016) Systematic revision of the marbled velvet geckos

(Oedura marmorata species complex, Diplodactylidae) from the Australian arid

and semi-arid zones. Zootaxa 4088, 151-176.

Oliver PM, Doughty P, Palmer R (2012) Hidden biodiversity in rare northern Australian

vertebrates: the case of the clawless geckos (Crenadactylus, Diplodactylidae) of

the Kimberley. Wildlife Research 39, 429-435.

Oliver PM, Hugall A, Adams M, Cooper SJB, Hutchinson MN (2007) Genetic

elucidation of cryptic and ancient diversity in a group of Australian

diplodactyline geckos; the Diplodactylus vittatus complex. Molecular

Phylogenetics and Evolution 44, 77-88.

Oliver PM, Laver RJ, De Mello Martins F, et al. (in press) A novel hotspot of vertebrate

endemism and evolutionary refugium in tropical Australia. Diversity and

Distributions.

Oliver PM, Laver RJ, Melville J, Doughty P (2014b) A new species of Velvet Gecko

(Oedura: Diplodactylidae) from the limestone ranges of the southern Kimberley,

Western Australia. Zootaxa 3873, 49-61.

129

Chapter 3: Contribution of climate and topography to diversification

Oliver PM, Laver RJ, Smith KL, Bauer AM (2014c) Long-term persistence and

vicariance within the Australian Monsoonal Tropics: The case of the giant cave

and tree geckos (Pseudothecadactylus). Australian Journal of Zoology 61, 462-

468.

Oliver PM, Smith KL, Laver RJ, Doughty P, Adams M (2014d) Contrasting patterns of

persistence and diversification in vicars of a widespread Australian lizard

lineage (the Oedura marmorata complex). Journal of Biogeography 41, 2068-

2079.

Palmer R, Pearson DJ, Cowan MA, Doughty P (2013) Islands and scales: a

biogeographic survey of reptiles on Kimberley islands, Western Australia.

Records of the Western Australian Museum Supplement 81, 183-204.

Parr CL, Lehmann CER, Bond WJ, Hoffmann WA, Andersen AN (2014) Tropical

grassy biomes: misunderstood, neglected, and under threat. TRENDS in Ecology

and Evolution 29, 205-213.

Pennington RT, Lewis GP, Ratter JA (2006) An overview of the plant diversity,

biogeography and conservation of Neotropical savannas and seasonally dry

forests. In: Neotropical savannas and seasonally dry forests: plant diversity,

biogeography and conservation (eds. Pennington RT, Lewis GP, Ratter JA), pp.

1-29. CRC Press, Boca Raton, FL.

Pepper M, Fujita MK, Moritz C, Keogh JS (2011a) Palaeoclimate change drove

diversification among isolated mountain refugia in the Australian arid zone.

Molecular Ecology 20, 1529-1545.

Pepper M, Ho SYW, Fujita MK, Keogh JS (2011b) The genetic legacy of aridification:

Climate cycling fostered lizard diversification in Australian montane refugia and

130

Chapter 3: Contribution of climate and topography to diversification

left low-lying deserts genetically depauperate. Molecular Phylogenetics and

Evolution 61, 750-759.

Pepper M, Keogh JS (2014) Biogeography of the Kimberley, Western Australia: a

review of landscape evolution and biotic response in an ancient refugium.

Journal of Biogeography 41, 1443-1455.

Pons J, Barraclough TG, Gómez-Zurita J, et al. (2006) Sequence-based species

delimitation for the DNA taxonomy of undescribed insects. Systematic Biology

55, 595-609.

Portik DM, Bauer AM, Jackman TR (2010) The phylogenetic affinities of Trachylepis

sulcata nigra and the intraspecific evolution of coastal melanism in the Western

Rock Skink. African Zoology 45, 147-159.

Potter S, Bragg JG, Peter BM, Bi K, Moritz C (2016) Phylogenomics at the tips:

inferring lineages and their demographic history in a tropical lizard, Carlia

amax. Molecular Ecology.

Potter S, Eldridge MDB, Taggart DA, Cooper SJB (2012) Multiple biogeographical

barriers identified across the monsoon tropics of northern Australia:

phylogeographic analysis of the brachyotis group of rock-wallabies. Molecular

Ecology 21, 2254-2269.

Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using

multilocus genotype data. Genetics 155, 945-959.

Proske U, Heslop D, Haberle S (2014) A Holocene record of coastal landscape

dynamics in the eastern Kimberley region, Australia. Journal of Quaternary

Science 29, 163-174.

131

Chapter 3: Contribution of climate and topography to diversification

Rambaut A, Drummond AJ (2002–2015) TreeAnnotator v2.3.1,

http://beast.bio.ed.ac.uk/treeannotator.

Rambaut A, Suchard MA, Xie D, Drummond AJ (2014) Tracer v1.6, Available from

http://beast.bio.ed.ac.uk/Tracer.

Read K, Keogh JS, Scott IAW, Roberts JD, Doughty P (2001) Molecular phylogeny of

the Australian frog genera Crinia and Geocrinia and allied taxa (Anura:

Myobatrachidae). Molecular Phylogenetics and Evolution 21, 294-308.

Reeves JM, Bostock HC, Ayliffe LK, et al. (2013) Palaeoenvironmental change in

tropical Australasia over the last 30,000 years–a synthesis by the OZ-

INTIMATE group. Quaternary Science Reviews 74, 97-114.

Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesian phylogenetic inference

under mixed models. Bioinformatics 19, 1572-1574.

Rosauer DF, Blom M, Bourke G, et al. (2016) Phylogeographic hotspots and

conservation priorities: an example from the Top End of Australia. In press.

Sakaguchi S, Bowman DMJS, Prior LD, et al. (2013) Climate, not Aboriginal landscape

burning, controlled the historical demography and distribution of fire-sensitive

conifer populations across Australia. Proceedings of the Royal Society B:

Biological Sciences 280, 20132182.

Silva ACA, Bragg JG, Potter S, et al. (in review) Tropical specialist versus climate

generalist: diversification and demographic history of sister species of Carlia

skinks from northwestern Australia. Molecular Ecology.

Skipwith PL, Bauer AM, Jackman TR, Sadlier RA (2016) Old but not ancient:

coalescent species tree of New Caledonian geckos reveals recent post-inundation

diversification. Journal of Biogeography.

132

Chapter 3: Contribution of climate and topography to diversification

Slatyer C, Rosauer D, Lemckert F (2007) An assessment of endemism and species

richness patterns in the Australian Anura. Journal of Biogeography 34, 583-596.

Smith KL, Harmon LJ, Shoo LP, Melville J (2011) Evidence of constrained phenotypic

evolution in a cryptic species complex of agamid lizards. Evolution 65, 976-992.

Stamatakis A (2006) RAxML-VI-HPC: Maximum likelihood-based phylogenetic

analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688-

2690.

Stamatakis A, Hoover P, Rougemont J (2008) A rapid bootstrap algorithm for the

RAxML web servers. Systematic Biology 57, 758-771.

Stephens M, Donnelly P (2003) A comparison of Bayesian methods for haplotype

reconstruction from population genotype data. American Journal of Human

Genetics 73, 1162-1169.

Stephens M, Smith N, Donnelly P (2001) A new statistical method for haplotype

reconstruction from population data. American Journal of Human Genetics 68,

979-989.

Tamura K, Nei M (1993) Estimation of the number of nucleotide substitutions in the

control region of mitochondrial DNA in humans and chimpanzees. Molecular

Biology and Evolution 10, 512-526.

Tamura K, Stecher G, Peterson D, Filipski A, Kumar S (2013) MEGA6: Molecular

evolutionary genetics analysis version 6.0. Molecular Biology and Evolution 30,

2725-2729.

Tipple BJ, Pagani M (2007) The early origins of terrestrial C4 photosynthesis. Annual

Review of Earth and Planetary Sciences 35, 435-461.

133

Chapter 3: Contribution of climate and topography to diversification

Townsend TM, Alegre ER, Kelley ST, Wiens JJ, Reeder TW (2008) Rapid

development of multiple nuclear loci for phylogenetic analysis using genomic

resources: An example from squamate reptiles. Molecular Phylogenetics and

Evolution 47, 129-142.

Waeber PO, Wilmé L, Ramamonjisoa B, et al. (2015) Dry forests in Madagascar:

neglected and under pressure. International Forestry Review 17, 127-148.

Wang Y-H, Jiang W-M, Comes HP, et al. (2015) Molecular phylogeography and

ecological niche modelling of a widespread herbaceous climber, Tetrastigma

hemsleyanum (Vitaceae): insights into Plio-Pleistocene range dynamics of

evergreen forest in subtropical China. New Phytologist 206, 852-867.

Wende R (1997) Aspects of the fluvial geomorphology of the Eastern Kimberley

Plateau, Western Australia PhD Thesis, University of Wollongong.

Werneck FP, Gamble T, Colli GR, Rodrigues MT, Sites Jr. JW (2012) Deep

diversification and long-term persistence in the South American 'dry diagonal':

integrating continent-wide phylogeography and distribution modeling of geckos.

Evolution 66, 3014-3034.

Wilgenbusch JC, Warren DL, Swofford DL (2004) AWTY: A system for graphical

exploration of MCMC convergence in Bayesian phylogenetic inference.

Wilmé L, Goodman SM, Ganzhorn JU (2006) Biogeographic evolution of Madagascar's

microendemic biota. Science 312, 1063-1065.

Woinarski J, Mackey B, Nix H, Traill B (2007) The Nature of Northern Australia:

Natural Values, Ecological Processes and Future Prospects. The Australian

National University E Press, Canberra.

134

Chapter 3: Contribution of climate and topography to diversification

Woinarski JCZ (1992) Biogeography and conservation of reptiles, mammals and birds

across north-western Australia: an inventory and base for planning an ecological

reserve system. Wildlife Research 19, 665-705.

Yang Z (2015) The BPP for species tree estimation and species delimitation. Current

Zoology 61.

Yang Z, Rannala B (2010) Bayesian species delimitaiton using multilocus sequence

data. Proceedings of the National Academy of Sciences 107, 9264-9269.

135

Chapter 3: Contribution of climate and topography to diversification

Appendix

Table A3.1. Specimen details for a) ingroup and b) outgroup samples.

Table A3.1. a)

Specimen GenBank accession numbers Species Locality Lat Long number ND2 K1AA KIF24 PDC PRLR PTPN12 RAG1 Oedura filicipoda AMSR126183 WA: Mitchell Plateau, Upstream of Little Mertens Falls -14.82 125.72 JQ173636 √ √ JQ173682 KU680108 √ JQ173730 Oedura filicipoda NMVD75792 WA: W Artesian Range -16.55 125.00 √ √ √ √ √ √ √ Oedura filicipoda NMVD75770 WA: W Artesian Range -16.55 125.00 √ √ √ √ √ √ √ Oedura filicipoda WAMR138874 WA: Donkins Hill -14.99 125.51 KJ803581 √ √ KJ803729 √ – KJ803729 Oedura filicipoda WAMR167805 WA: Surveyors Pool -14.67 125.73 KJ803579 √ √ √ √ √ √ Oedura filicipoda WAMR171552 WA: Prince Regent Nature Reserve -15.76 125.26 KJ803580 √ √ √ √ √ √ Oedura gracilis AMSR136067 WA: Bells Gorge, Bells Crk, Isdell Rvr -16.98 125.18 JQ173638 – – JQ173684 – – JQ173732 Oedura gracilis AMSR140309 WA: Manning Gorge, Mt Barnett Station -16.66 125.93 JQ173639 – – JQ173685 – – JQ173733 Oedura gracilis CCM0628 NT: Bullo Rvr Station -15.59 129.64 √ – – – – – – Oedura gracilis CCM0639 NT: Bullo Rvr Station -15.66 129.66 √ – – – – – – Oedura gracilis CCM0640 NT: Bullo Rvr Station -15.66 129.66 √ – – – – – – Oedura gracilis NTMR37356 NT: Bullo Rvr Station -15.66 129.66 √ – – – – – – Oedura gracilis CCM0729 WA: Spring Crk -15.20 126.09 √ – – – – – – Oedura gracilis ANWCR10207 WA: Spring Crk -15.20 126.09 √ – – – – – – Oedura gracilis CCM0784 WA: Spring Crk -15.20 126.09 √ – – – – – – Oedura gracilis CCM0785 WA: Spring Crk -15.20 126.09 √ √ √ √ √ √ √ Oedura gracilis CCM0821 WA: Old Mitchell, near Spring Crk -15.14 126.16 √ √ √ √ √ √ √ Oedura gracilis CCM1043 WA: 21.4k W Ellenbrae -15.95 126.87 √ – – – – – – Oedura gracilis ANWCR10208 WA: Gibb Rvr Rd, W Durack Rvr -15.85 127.41 √ √ √ √ √ √ √ Oedura gracilis CCM1074 WA: 21.4k W Ellenbrae -15.95 126.87 √ – – – – – – Oedura gracilis ANWCR10209 WA: Monorromboora Hill, Theda -14.77 126.58 √ √ √ √ √ √ √ Oedura gracilis CCM1187 WA: Mitchell Plateau Repeater Track -14.83 125.72 √ – – – – – – Oedura gracilis CCM1309 WA: Mt Hart Rd -16.97 125.03 √ – – – – – – Oedura gracilis ANWCR10210 WA: Mt Hart Rd -16.97 125.03 √ √ √ √ √ √ √ Oedura gracilis ANWCR10211 WA: Grevillea Gorge -16.50 125.34 √ – – – – – – Oedura gracilis CCM1429 WA: Lilly Pool -16.50 125.34 √ √ √ √ √ √ √ Oedura gracilis CCM1460 WA: Teronis Gorge -17.29 127.25 √ – – – – – – Oedura gracilis CCM1468 WA: Teronis Gorge -17.30 127.26 √ – – – – – – Oedura gracilis ANWCR10212 WA: Teronis Gorge -17.30 127.26 √ √ √ √ √ √ √ Oedura gracilis ANWCR10213 WA: Gibb Rvr Rd -17.17 125.35 √ – – – – – – Oedura gracilis CCM1541 WA: Police Valley -16.82 126.22 √ – – – – – – Oedura gracilis CCM1542 WA: Police Valley -16.82 126.22 √ – – – – – – Oedura gracilis CCM1543 WA: Police Valley -16.82 126.22 √ √ √ √ √ √ √ Oedura gracilis CCM1544 WA: Police Valley -16.82 126.22 √ – – – – – – Oedura gracilis CCM1556 WA: Police Valley -16.82 126.22 √ √ √ √ √ √ √ Oedura gracilis ANWCR10214 WA: Police Valley -16.82 126.22 √ – – – – – – Oedura gracilis CCM1570 WA: Barnett Rvr Gorge -16.54 126.13 √ √ √ √ √ √ √ Oedura gracilis CCM1585 WA: Barnett Rvr Gorge -16.54 126.13 √ √ √ √ √ √ √ Oedura gracilis CCM1586 WA: Barnett Rvr Gorge -16.54 126.13 √ – – – – – – Oedura gracilis CCM1609 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ √ √ – √ √ √ Oedura gracilis ANWCR10215 WA: Munja Track -16.21 126.01 √ √ √ √ √ √ √ Oedura gracilis CCM1740 WA: Mitchell Plateau Repeater Track -14.83 125.72 √ √ √ √ √ √ √ Oedura gracilis ANWCR10216 WA: Mitchell Plateau Repeater Track -14.83 125.72 √ – – – – – – Oedura gracilis CCM1755 WA: King Edward Rvr -14.88 126.17 √ √ √ √ √ √ √ Oedura gracilis CCM2832 WA: Mornington -17.51 126.11 √ – – – – – – Oedura gracilis CCM2833 WA: Mornington -17.51 126.11 √ – – – – – – Oedura gracilis CCM2834 WA: Mornington -17.51 126.11 √ – – – – – – Oedura gracilis CCM2844 WA: Maggie Springs -17.44 126.19 √ √ √ √ √ – √ Oedura gracilis CCM3034 WA: Saw Tooth Gorge -18.43 127.82 √ √ √ √ √ – √ Oedura gracilis CCM3196 WA: Mt Nyulasy -16.75 128.28 √ √ √ √ √ – √ Oedura gracilis CCM3422 WA: Doongan Station -15.16 126.26 √ – – – – – – Oedura gracilis N906 WA: Anjo Peninsula -14.08 126.35 √ √ √ √ √ √ √ Oedura gracilis N907 WA: Anjo Peninsula -14.08 126.35 √ – – √ √ – √ Oedura gracilis NMVD77001 WA: Gibb Rvr Rd, King Leopold Ranges -17.12 125.13 √ √ √ √ √ √ √ Oedura gracilis NMVD77024 WA: McSherry Gap -17.56 125.10 √ √ √ √ √ √ √ Oedura gracilis NTMR13438 NT: Bullo Rvr Station -15.47 129.75 KJ803620 √ √ KJ803741 √ √ KJ803741 Oedura gracilis NTMR13445 NT: Bullo Rvr Station -15.47 129.75 √ √ √ √ √ √ √ Oedura gracilis PMO152 WA: Barnett Rvr Gorge -16.53 126.13 √ – – – – – – Oedura gracilis PMO160 WA: Wire Springs -17.69 125.14 √ – – – – – – Oedura gracilis QMJR91439 WA: Lake Argyle Rd -16.04 128.77 √ √ √ √ √ √ √ Oedura gracilis TR821 WA: Chamberlin River -16.67 127.63 √ √ √ √ √ – √ Oedura gracilis WAMR106213 WA: Manning Gorge -16.67 125.95 √ √ √ √ √ √ √ Oedura gracilis WAMR138878 WA: Donkins Hill -15.00 125.50 √ √ √ √ √ – √ Oedura gracilis WAMR138891 WA: Donkins Hill -14.99 125.51 KJ803617 √ √ KJ803738 √ – KJ803738 Oedura gracilis WAMR151005 WA: Mt Nyulasy -16.75 128.29 KJ803619 √ √ KJ803740 √ √ KJ803740 Oedura gracilis WAMR151963 WA: SW Osborn Is -14.35 125.95 KJ803621 – – KJ803742 √ – KJ803742 Oedura gracilis WAMR151980 WA: SW Osborn Is -14.35 125.95 √ – – √ √ – √ Oedura gracilis WAMR156724 WA: Piccaninny Massif, Bungle Bungles -17.40 128.41 KJ803609 √ √ KJ803730 √ √ KJ803730 Oedura gracilis WAMR156725 WA: Piccaninny Massif, Bungle Bungles -17.40 128.41 √ √ √ √ √ √ √ Oedura gracilis WAMR156728 WA: Tunnel Crk -17.64 125.17 KJ803610 √ √ KJ803731 √ √ KJ803731 Oedura gracilis WAMR164863 WA: SW Osborn Is -14.37 125.94 √ √ √ √ √ √ √ Oedura gracilis WAMR164864 WA: SW Osborn Is -14.37 125.94 √ – – √ √ – √ Oedura gracilis WAMR164908 WA: Katers Is -14.46 125.52 √ – – √ √ – √ Oedura gracilis WAMR164909 WA: Katers Is -14.45 125.52 √ – – √ √ – √ Oedura gracilis WAMR164910 WA: Katers Is -14.45 125.52 KJ803611 √ √ KJ803732 √ √ KJ803732

136

Chapter 3: Contribution of climate and topography to diversification

Table A3.1. a) continued

Specimen GenBank accession numbers Species Locality Lat Long number ND2 K1AA KIF24 PDC PRLR PTPN12 RAG1 Oedura gracilis WAMR168057 WA: Quail Falls -15.75 125.37 √ √ √ √ √ √ √ Oedura gracilis WAMR168454 WA: Sir Graham Moore Is -13.88 126.57 √ – – √ √ – √ Oedura gracilis WAMR168463 WA: Sir Graham Moore Is -13.88 126.57 √ √ √ √ √ √ √ Oedura gracilis WAMR168564 WA: Boongaree Is -15.10 125.20 KJ803612 √ √ KJ803733 √ √ KJ803733 Oedura gracilis WAMR168565 WA: Augustus Is -15.35 124.53 KJ803613 √ √ KJ803734 √ √ KJ803734 Oedura gracilis WAMR168566 WA: Augustus Is -15.35 124.53 √ – – √ √ – √ Oedura gracilis WAMR168739 WA: Katers Is -14.47 125.53 √ – – √ √ – √ Oedura gracilis WAMR168903 WA: Bigge Is -14.60 125.12 √ – – √ √ – √ Oedura gracilis WAMR168904 WA: Bigge Is -14.60 125.12 √ √ √ √ √ √ √ Oedura gracilis WAMR171204 WA: Augustus Is -15.39 124.59 √ √ √ √ √ √ √ Oedura gracilis WAMR171205 WA: Augustus Is -15.35 124.53 √ – – √ √ – √ Oedura gracilis WAMR171668 WA: Storr Is -15.95 124.56 KJ803614 √ √ KJ803735 √ √ KJ803735 Oedura gracilis WAMR171669 WA: Storr Is -15.95 124.56 √ √ √ √ √ √ √ Oedura gracilis WAMR171670 WA: Lachlan Is -16.62 123.47 KJ803615 √ √ KJ803736 √ √ KJ803736 Oedura gracilis WAMR171671 WA: Long Is -16.56 123.36 √ – – √ √ – √ Oedura gracilis WAMR171672 WA: Long Is -16.56 123.36 √ √ √ √ √ √ √ Oedura gracilis WAMR171673 WA: Long Is -16.56 123.36 √ √ √ √ √ √ √ Oedura gracilis WAMR171674 WA: Sunday Is -16.43 123.18 √ – – √ √ – √ Oedura gracilis WAMR171675 WA: Storr Is -15.95 124.56 √ – – √ √ – √ Oedura gracilis WAMR171676 WA: Unnamed Is -15.91 124.46 √ √ √ √ √ √ √ Oedura gracilis WAMR171677 WA: NW Molema Is -16.26 123.82 √ √ √ √ √ √ √ Oedura gracilis WAMR172341 WA: Theda Station -14.81 126.51 KJ803616 √ √ KJ803737 √ √ KJ803737 Oedura gracilis WAMR172865 WA: Ellenbrae -15.98 127.05 KJ803618 – – KJ803739 √ – KJ803739 Oedura gracilis WAMR172866 WA: Ellenbrae -15.98 127.05 √ – – √ √ – √ Oedura gracilis WAMR172867 WA: Ellenbrae -15.98 127.05 √ – – √ √ – √ Oedura gracilis WAMR172868 WA: Ellenbrae -15.97 127.06 √ √ √ √ √ √ √ Oedura gracilis WAMR172869 WA: Ellenbrae -15.97 127.07 √ – – √ √ – √ Oedura gracilis WAMR172870 WA: Ellenbrae -15.97 127.07 √ – – √ √ – √ Oedura gracilis WAMR172871 WA: Ellenbrae -15.97 127.07 √ √ √ √ √ √ √ Oedura gracilis WAMR172903 WA: Doongan Station -15.20 125.90 √ – – – – – – Oedura gracilis WAMR172904 WA: Doongan Station -15.20 125.90 KU680226 √ √ KU680101 KU680110 √ KU680029 Oedura gracilis WAMR173941 WA: Coucal Gorge -15.01 126.84 √ – – – – – – Oedura gracilis WAMR173942 WA: Coucal Gorge -15.01 126.84 √ – – – – – – Oedura gracilis WAMR174023 WA: W King Edward Rvr -14.75 126.21 √ – – – – – – Oedura gracilis WAMR174025 WA: W King Edward Rvr -14.75 126.21 √ – – – – – – Oedura gracilis WAMR174030 WA: W King Edward Rvr -14.75 126.21 √ – – – – – – Oedura gracilis WAMR174049 WA: Craticus Falls -14.78 127.10 √ – – – – – – Oedura gracilis WAMR174106 WA: Worriga Falls -15.03 126.67 √ – – – – – – Oedura gracilis WAMR174114 WA: Euro Gorge -15.06 126.72 √ – – – – – – Oedura murrumanu CCM3313 WA: Geikie Gorge -18.10 125.69 KU522238 – – – – – – Oedura murrumanu CCM3314 WA: Geikie Gorge -18.10 125.69 √ – – – – – – Oedura murrumanu NMVD77002 WA: Oscar Range, W Leopold Downs Rd -17.91 125.28 KM016836 √ √ √ √ √ √ Oedura murrumanu WAMR173369 WA: Oscar Range, W Leopold Downs Rd -17.91 125.28 KM016837 √ √ √ √ √ √ Oedura murrumanu NMVD76947 WA: Oscar Range, W Leopold Downs Rd -17.91 125.28 KM016838 √ √ √ √ √ √ Oedura murrumanu WAMR173368 WA: Oscar Range, E Leopold Downs Rd -17.92 125.30 KM016839 √ √ √ √ √ √ Oedura murrumanu NMVD76948 WA: Oscar Range, E Leopold Downs Rd -17.92 125.30 KM016840 √ √ √ √ √ √ Oedura murrumanu WAMR173370 WA: Oscar Range, E Leopold Downs Rd -17.92 125.30 KM016841 √ √ √ √ √ √ Oedura murrumanu PMO177 WA: Oscar Range -17.68 125.07 KU522241 – – – – – – Oedura murrumanu PMO185 WA: Oscar Range -17.67 125.07 KU522239 – – – – – – Oedura murrumanu PMO190 WA: Oscar Range -17.68 125.07 KU522240 – – – – – –

137

Chapter 3: Contribution of climate and topography to diversification

Table A3.1. b)

Specimen GenBank accession numbers Species Locality Lat Long number ND2 K1AA KIF24 PDC PRLR PTPN12 RAG1 Amalosia lesueurii AMSR159546 NSW: Moonbi Lookout, Moonbi Ranges -30.99 151.08 JQ173642 – – JQ173688 – – JQ173736 Amalosia obscura AMSR136124 WA: 4k NE Surveyors Pool, Mitchell Plateau -14.65 125.77 JQ173655 – – JQ173701 – – JQ173748 Amalosia rhombifera SAMAR55604 QLD: Kroombit Tops -24.33 150.94 JQ173661 – – JQ173709 – – JQ173755 Aprasia parapulchella AMSR127439 NSW: 20k N Tarcutta -35.17 147.88 GU459941 – – GU459741 – – GU459539 Crenadactylus ocellatus SAMAR22245 NT: 10k S Barrow Ck -21.63 133.88 AY369016 – – JQ945367 – – AY662627 Diplodactylus conspicillatus AMSR158426 NSW: Sturt National Park -29.38 142.04 JQ173627 – – JQ173673 – – JQ173721 Diplodactylus granariensis AMSR150637 WA: Dedari -31.08 120.68 JQ173628 – – JQ173674 – – JQ173722 Diplodactylus ornatus AMSR140546 WA: Denham -25.92 113.53 JQ173629 – – JQ173675 – – JQ173723 Diplodactylus tessellatus AMSR143855 QLD: 7.9k SW Landborough Hwy on Boulia Rd -24.37 143.32 JQ173631 – – JQ173677 – – JQ173725 Hesperoedura reticulata WAMR114757 WA: Kalgoorlie -30.87 121.25 KU680225 √ √ KU680099 KU680105 – KU680028 Hoplodactylus duvaucellii FT(VUW)174 NZ: Mercury Is -36.63 175.90 GU459843 – – GU459642 KU680106 – GU459440 Lucasium maini AMSR150647 WA: Dedari -31.08 120.68 JX024362 – – KU680100 – – JX024503 Lucasium stenodactylum AMSR139897 WA: El Questro Station -16.02 128.00 JQ173630 – – JQ173676 – – JQ173724 Mokopirirakau granulatus RAH363 NZ: Trass -41.41 172.92 GU459812 – – GU459611 – – GU459409 Naultinus gemmeus RAH188 NZ: Otago Peninsula -45.89 170.62 GU459761 – – GU459560 – – GU459358 Nebulifera robusta ABTC3938 QLD: near Rathdowney -28.21 152.86 JQ173662 – – JQ173710 KU680107 – JQ173756 Oedodera marmorata AMSR161254 NC: Sommet Noir, Paagoumène, 11k NW Koumac -20.49 164.20 JQ173632 – – JQ173678 – – JQ173726 Oedura bella NTMR21288 QLD: Musselbrook Reservoir -18.61 138.08 KJ803591 – – – – – – Oedura bella SAMAR34188 NT: McArthur Rvr Station -16.67 135.85 KJ803592 √ √ KJ803712 √ √ KJ803671 Oedura bella SAMAR35425 QLD: Lawn Hill NP -18.58 138.50 KJ803594 – – – – – – Oedura castelnaui AMSR143917 QLD: 4.9k E Georgetown -18.28 143.59 JQ173633 – – JQ173679 – – JQ173727 Oedura cintca AMSR136049 NT: 49.7k NW Yuenduma T/O Tanami Rd, 8 Mile Bore -22.12 131.37 JQ173644 – – – – – – Oedura cintca AMSR138446 NSW: 1.2k W Warrego Rvr Brdg, Bourke-Wanaaring Rd -30.00 145.35 JQ173645 – – – – – – Oedura cintca NTMR18278 NT: Paddy's Rockhole -22.40 137.67 KJ803565 – – – – – – Oedura cincta SAMAR38842 NT: Honeymoon Gap, Alice Springs -23.75 133.75 JQ173648 √ √ JQ173694 √ √ JQ173742 Oedura cintca SAMAR41324 SA: Oakbank Stn -33.13 140.65 KJ803573 – – – – – – Oedura cintca SAMAR42883 QLD: 30k E Noonbah Stn -24.12 143.42 KJ803574 – – – – – – Oedura cincta SAMAR52203 SA: 2.7k W Lance Bore, Narrina Stn -30.96 138.77 JQ173649 – – – – – – Oedura cintca SAMAR54300 SA: W Bellbird Campsite, Gluepot Reserve -33.70 140.16 KJ803577 – – – – – – Oedura cincta SAMAR55902 QLD: Mitchell HWY, 9k N NSW/QLD Border -28.96 145.73 KJ803578 √ √ KJ803707 √ √ KJ803666 Oedura coggeri AMSR143918 QLD: 20k W Gulf Development Rd & Kennedy Hwy -18.13 144.63 JQ173635 – – JQ173681 – – JQ173729 Oedura fimbria WAMR105965 WA: 7k N Mt Magnet -28.00 119.12 JQ173650 – – – – – – Oedura fimbria WAMR132296 WA: Ulongunna Rock -27.12 117.23 KJ803640 – – – – – – Oedura fimbria WAMR154783 WA: Brockman Rdg -23.31 119.92 KJ803650 – – – – – – Oedura fimbria WAMR154797 WA: Walga Rock -27.40 117.47 KJ803653 √ √ KJ803726 √ √ KJ803726 Oedura fimbria WAMR160074 WA: 32.5k ESE Meentheena Outcamp -21.33 120.75 KJ803658 – – – – – – Oedura fimbria WAMR165242 WA: Burrup Peninsula, Hearson Cove -20.64 116.80 KJ803661 – – – – – – Oedura gemmata SAMAR34170 NT: UDP Falls -13.43 132.42 JQ173637 – – – – – – Oedura gemmata NTMR34896 NT: Mountain Valley Stn -13.98 133.18 KJ803587 – – – – – – Oedura gemmata NTMR34897 NT: Mountain Valley Stn -13.98 133.18 KJ803587 – – – – – – Oedura gemmata NTMR35680 NT: Kakadu National Park -13.86 132.98 KJ803589 √ √ KJ803710 KU680109 √ KJ803710 Oedura marmorata NTMR19029 NT: N Marchinbar Is -11.20 136.70 KJ803598 √ √ KJ803716 √ √ KJ803675 Oedura marmorata NTMR22444 NT: Limmen Gate National Park -15.78 135.33 KJ803608 – – KJ803723 √ – KJ803682 Oedura marmorata NTMR13222 NT: Urapunga Stn -14.51 134.57 KJ803596 – – – – – – Oedura marmorata NTMR13295 NT: Victoria Rvr Gregory NP -15.94 130.50 KJ803597 – – – – – – Oedura marmorata NTMR13619 NT: 3k S Katherine -14.47 132.27 JQ173647 – – – – – – Oedura marmorata NTMR19030 NT: N Marchinbar Is -11.28 136.63 KJ803599 – – – – – – Oedura marmorata NTMR22104 NT: Cox Peninsula Rd -12.52 130.82 KJ803606 – – – – – – Oedura marmorata NTMR22774 NT: English Company Isles, Astell Is -11.88 136.42 KJ803600 – – – – – – Oedura marmorata NTMR34156 NT: 1k N Katherine -14.46 132.26 KJ803604 – – – – – – Oedura marmorata NTMR34158 NT: 1k N Katherine -14.46 132.26 KJ803605 – – – – – – Oedura marmorata NTMR36746 NT: Wongalara Sanctuary -14.13 134.34 KJ803603 – – – – – – Oedura monilis SAMAR54560 QLD: Dawson Development Rd, 18k E Alpha T/off -24.53 146.55 JQ173653 – – JQ173699 – – JQ173747 Oedura tryoni AMSR152045 NSW: Moonbi Lookout, Moonbi Ranges -30.99 151.08 JQ173663 – – JQ173711 – – JQ173757 Paniegekko madjo MNHN1998.0467NC: Mt Ignambi -20.47 164.60 GU459950 – – GU459751 – – GU459549 Pseudothecadactylus australis QMJR57120 QLD: Heathlands Ranger Station -11.75 142.58 HQ288425 – – KU680102 – – FJ855449 Pseudothecadactylus cavaticus WAMR171550 WA: Prince Regent Nature Reserve -15.76 125.26 KJ685523 √ √ KU680103 KU680111 √ KJ685534 Pseudothecadactylus lindneri MVZ99544 NT: Kakadu National Park -13.09 123.40 GU459946 – – GU459747 – – GU459545 Rhacodactylus chahoua AMSR161238 NC: Dome de Tiebaghi, 14k NW Koumac -20.46 164.19 JQ173665 – – JQ173713 – – JQ173759 Rhacodactylus leachianus AMB7189 NC: Ilot Moro -22.65 167.38 GU459949 – – GU459750 – – GU459548 Rhynchoedura eyrensis AMSR155371 NSW: Sturt National Park -29.04 141.31 GU459954 – – GU459755 – – GU459553 Strophurus assimilis AMSR149832 WA: 17.6k W Bonny Vale Railway Station -30.80 120.98 JQ173666 – – JQ173714 KU680112 – JQ173760 Strophurus ciliaris WAMR164701 WA: 15k SE Fitzroy Crossing -18.31 125.64 KU680180 √ √ KU680064 KU680117 √ KU679976 Strophurus elderi AMSR130987 NSW: 17.9k N Coombah Roadhouse -32.85 141.62 JQ173669 – – JQ173717 – – JQ173763 Strophurus horneri NMVD72591 NT: Yirrakak, Arnhemland Plateau -12.20 133.80 KU680191 – – KU680074 KU680134 – – Strophurus intermedius SAMAR28963 SA: Gawler Ranges -32.60 136.20 KU680194 – – KU680077 KU680138 – KU679992 Strophurus jeanae WAMR110614 WA: Tanami Desert -19.90 128.87 KU680195 – – KU680079 KU680140 – KU679994 Strophurus krisalys SAMAR54523 QLD: Hughenden Dump -20.35 144.46 KU680198 – – KU680082 KU680143 – KU679998 Strophurus mcmillani WAMR171691 WA: Unnamed Is -15.91 124.46 KU680202 – – KU680086 KU680147 – KU680001 Strophurus rankini AMSR140490 WA: Coral Bay -23.13 113.77 JQ173670 – – JQ173718 – – JQ173764 Srophurus robinsoni Data pending NT: S Keep Rvr National Park -16.19 129.11 KU680206 √ √ KU680089 KU680152 √ KU680006 Strophurus spinigerus AMSR149815 WA: Buckland Hill -32.02 115.77 JQ173671 – – JQ173719 KU680155 – JQ173765 Strophurus strophurus AMSR140536 WA: Denham -25.92 113.53 JQ173672 – – JQ173720 – – JQ173766 Strophurus taeniatus NTMR36343 NT: Mt Sanford, Victoria Rvr Region -17.48 129.88 KU680216 – – KU680094 KU680163 – KU680019 Srophurus taeniatus NTMR36750 NT: Wongalara Sanctuary -14.27 134.62 KU680213 √ √ KU680091 KU680164 √ KU680016 Strophurus taenicauda QMJR76397 QLD: Coominglah State Forest, near Monto -24.80 150.98 HQ171996 – – – KU680216 – KU680021 Strophurus wellingtonae WAMR145495 WA: Lorna Glen Station -26.08 121.45 KU680219 – – KU680095 KU680169 – KU680022 Strophurus williamsi SAMAR25518 SA: Danggalli Conservation Park -33.60 140.88 AY369007 – – KU680096 KU680171 – KU680023 Woodworthia maculatus RAH292 NZ: Titahi Bay -41.06 174.50 GU459852 – – GU459651 – – GU459449 Tissue sources: ABTC, Australian Biological Tissue Collection; AMB - Aaron M. Bauer - University of Villanova; AMS, Australia Museum; BP - Russel Palmer - lodged at Western Australia Museum (WAM); CCM; Craig C. Moritz - Australian National University; FT(VUW), frozen tissue of Victoria University of Wellington; MNHN, Museum National d'Histoire Naturelle; MVZ, Museum of Vertebrate Zoology - University of California, Berkeley; NMV, Museum Victoria; NTM, Northern Territory Museum; QMJ, Queensland Museum; RAH, Rodney A. Hitchmough; SAMA, South Australia Museum; TS, Stuart Young - lodged at NTM.

138

Chapter 3: Contribution of climate and topography to diversification

Table A3.2. Characteristics of the genetic data sequenced for molecular and phylogenetic analyses of the Oedura gracilis complex.

Locus bp n S V PI ND2 861 107 24 379 355 KIAA1549 573 60 13 27 14 KIF24 519 60 24 80 56 PDC 393 75 5 14 9 PRLR 561 72 9 24 15 PTPN12 897 48 5 16 11 RAG1 1065 76 18 50 32 bp, length (in base pairs); n, number of sequences; S, singleton sites; V, variable sites; PI, parsimony-informative sites.

139

Chapter 3: Contribution of climate and topography to diversification

Table A3.3. Details of primers and protocols used in this study.

Fragment Primer name Primer sequence (5' to 3') Source PCR conditions length ND2 L4437 AAG CTT TCG GGG CCC ATA CC Macey et al. (1997) c. 1200 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; tRNA-Asn CTA AAA TRT TRC GGG ATC GAG GCC Read et al. (2001) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C ND2.F GCC CAT ACC CCG AAA ATS TTG Oliver et al. (2007) c. 900 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; ND2.R TTA GGG TRG TTA TTT GHG AYA TKC Oliver et al. (2007) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C ND2_grac.F TCC TTC CTA TAC TAA TAA ACC CAG T This study c. 1000 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; ND2_grac.R AGC TTG GTA ATG TGG ATA GGG CT This study 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C KIAA1549 KIAA1549f332 CAG GTT CCT ATC CAG TGA AGC CA Skipwith et al. (2016) c. 1000 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; KIAA1549r709 TTT CTC ATT AAA MAG ACT ATC AGC CAA AAT Skipwith et al. (2016) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C KIF24 KIF24f1 SAA ACG TRT CTC CMA AAC GCA TCC Portik et al. (2010) c. 550 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; KIF24r1zebra GCT GCT GRA RCT GCT GGT GAT AAA GRC G Skipwith et al. (2016) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C PDC PHO.F2 AGA TGA GCA TGC AGG AGT ATG A Bauer et al. (2007) c. 450 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; PHO.R1 TCC ACA TCC ACA GCA AAA AAC TCC T Bauer et al. (2007) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C PRLR PRLR.F1 GAC ARY GAR GAC CAG CAA CTR ATG CC Townsend et al. (2008) c. 600 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; PRLR.R3 GAC YTT GTG RAC TTC YAC RTA ATC CAT Townsend et al. (2008) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C PTPN12 PTPN12.F1 AGT TGC CTT GTW GAA GGR GAT GC Townsend et al. (2008) c. 1000 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 50 °C, 30 s; PTPN12.R6 CTR GCA ATK GAC ATY GGY AAT AC Townsend et al. (2008) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C RAG1 RAG1R13.F TCT GAA TGG AAA TTC AAG CTG TT Groth & Barrowclough (1999) c. 1000 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; RAG1R852.R GAG TCT GCA GAA TAA GTG CTT GCA Oliver et al. (2014) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C RAG1R13.F TCT GAA TGG AAA TTC AAG CTG TT Groth & Barrowclough (1999) c. 1000 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 58 °C, 30 s; RAG1R18.R GAT GCT GCC TCG GTC GGC CAC CTT T Groth & Barrowclough (1999) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C RAG1R13.F TCT GAA TGG AAA TTC AAG CTG TT Groth & Barrowclough (1999) c. 800 bp 95 °C, 5 min; 40 × (95 °C, 30 s; 55 °C, 30 s; RAG1r.Stroph840 AAG TGC TTG CAT GTT GTT TC Nielsen et al. (2016) 72 °C, 1 min); 72 °C, 5 min; hold at 15 °C

140

Chapter 3: Contribution of climate and topography to diversification

Table A3.4. Optimal partition and model schemes for each dataset and analysis as ranked by Bayesian Information Criterion (BIC) scores.

Optimal partition and model scheme Dataset RAxML MrBayes/BEAST ND2 ND2_1 (GTR+G); ND2_2 (GTR+I+G); ND2_3 (GTR+G) ND2_1 (GTR+G); ND2_2 (HKY+I+G); ND2_3 (GTR+G) KIAA1549 KIAA_1 KIAA_2 KIAA_3 (GTR+G) KIAA_1 KIAA_2 KIAA_3 (K80+G) KIF24 KIF24_1 KIF24_2 KIF24_3 (GTR+G) KIF24_1 KIF24_2 KIF24_3 (GTR+G) PDC PDC_1 PDC_2 (GTR+G); PDC_3 (GTR+G) PDC_1 PDC_2 (K80+I); PDC_3 (HKY+G) PRLR PRLR_1 PRLR_2 PRLR_3 (GTR+G) PRLR_1 PRLR_2 PRLR_3 (K80+G) PTPN12 PTPN12_1 PTPN12_2 (GTR+G); PTPN12 (GTR+G) PTPN12_1 (K80); PTPN12_2 (HKY); PTPN12_3 (HKY+G) RAG1 RAG1_1 RAG1_2 RAG1_3 (GTR+I+G) RAG1_1 RAG1_2 (HKY+I); RAG1_3 (K80+G) KIAA_1 KIAA_3 PRLR_1 PRLR_2 PRLR_3 PTPN12_3 KIAA_1 KIAA_2 KIF24_1 KIF24_2 PRLR_1 PRLR_2 6 gene RAG1_3 (GTR+I+G); KIAA_2 PDC_1 PDC_2 RAG1_3 (HKY+I+G); KIAA_3 PRLR_3 PTPN12_3 concatenation PTPN12_1 PTPN12_2 RAG1_1 RAG1_2 (GTR+G); (HKY+G); KIF24_3 PDC_3 (HKY+G); PDC_1 PDC_2 (nuDNA) KIF24_1 KIF24_2 KIF24_3 PDC_3 (GTR+I+G) PTPN12_1 PTPN12_2 RAG1_1 RAG1_2 (HKY+I) ND2_1 (GTR+G); ND2_2 (GTR+I+G); ND2_3 (GTR+G); ND2_1 (GTR+G); ND2_2 (HKY+I+G); ND2_3 (GTR+G); 7 gene KIAA_1 KIAA_3 PRLR_1 PRLR_2 PRLR_3 PTPN12_3 KIAA_1 PRLR_1 PRLR_2 RAG1_1 RAG1_2 (HKY+G); concatenation (GTR+G); KIAA_2 KIF24_1 KIF24_2 PDC_3 RAG1_3 KIAA_2 RAG1_3 (K80+I+G); KIAA_3 PRLR_3 PTPN12_3 (mtDNA+nuDNA) (GTR+I+G); KIF24_3 (GTR+G); PDC_1 PDC_2 (HKY+G); KIF24_1 KIF24_2 KIF24_3 PDC_3 (GTR+I+G); PTPN12_1 PTPN12_2 RAG1_1 RAG1_2 (GTR+G) PDC_1 PDC_2 PTPN12_1 (K80+I); PTPN12_2 (HKY)

141

Chapter 3: Contribution of climate and topography to diversification

Table A3.5. Morphological variables measured for voucher specimens of the Oedura gracilis complex.

Measure Definition Snout-vent-length (SVL) Length from tip of snout to anterior edge of vent Head width (HW) Maximum width of head Head depth (HD) Maximum depth of head just posterior to orbitals Head length (HL) Length from anterior edge of ear to tip of snout Head length-ret (HL-RET) Length from retroarticular process to tip of snout Eye-to-naris distance (EN) Length from anterior corner of eye to posterior edge of naris Internarial distance (IN) Distance between inner edges of nares Interorbital distance (IO) Distance between anterior-dorsal edges of eyes Transverse length of eye (EYE) Transverse length of eye Ear-to-eye distance (EE) Length from anterior edge of ear to posterior corner of eye Axilla to groin (trunk) distance (Trk) Length from posterior edge of forelimb insertion to anterior edge of hindlimb insertion Length of lower arm (Crus) Length from posterior edge of bent elbow to wrist Length of lower leg (Tibia) Length from anterior edge of bent knee to heel 4th finger length (4FL) Maximum length (including lamellae) of the 4th finger 4th finger width (4FW) Maximum width (including lamellae) of the 4th finger 4th toe length (4TL) Maximum length (including lamellae) of the 4th toe 4th toe width (4TW) Maximum width (including lamellae) of the 4th toe Number of supralabial scale (SuL) Number of supralabial scales more than twice size of surrounding granular scales Number of infralabial scales (InL) Number of infralabial scales more than twice size of surrounding granular scales Length of rostral crease (RC) Length of rostral crease as a percentage of rostral height Number of pre-cloacal pores (PP) Total number of expressed pre-cloacal pores on males Number of bands (Bands) Number of distinct light dorsal bands from nape to haunches 4th finger lamellae (4FLam) Number of enlarged lamellae (more than twice the width of surrounding scales) under the 4th finger 4th toe lamellae (4TLam) Number of enlarged lamellae (more than twice the width of surrounding scales) under the 4th toe

142

Chapter 3: Contribution of climate and topography to diversification

Table A3.6. Phenotypic measures for specimens of Oedura gracilis used in phenotypic analyses.

Specimen # Lineage Clade Sex SVL HW HD HL HL-RET EN IN IO EYE EE Trk Crus Tibia 4FL 4FW 4TL 4TW SuL InL RC PP Bands 4FLam 4TLam CCM0647 Bullo River East M 85.13 15.47 7.33 20.68 23.23 9.02 2.71 6.41 5.35 7.76 34.78 11.32 13.14 4.91 1.87 5.45 1.86 14 12 0.30 11 5 9 10 NTMR13438 Bullo River East F 73.89 14.41 6.66 20.16 21.59 8.89 2.54 6.26 3.66 6.52 32.27 10.93 12.60 4.61 1.27 5.21 1.75 14 13 0.50 5 10 9 NTMR13445 Bullo River East F 85.01 16.21 6.77 21.85 25.27 7.18 2.98 6.91 4.25 7.32 36.63 10.10 13.62 4.77 1.80 5.10 2.04 15 13 0.45 5 10 9 CCM3034 Bungle Bungles East F 86.47 15.81 6.61 20.30 23.78 8.97 2.78 7.41 4.16 7.33 40.77 11.87 13.56 4.67 1.87 5.58 2.18 11 12 0.50 5 9 10 WAMR156724 Bungle Bungles East F 81.07 14.83 7.10 19.64 21.61 8.60 2.76 6.83 4.28 7.21 34.31 10.51 13.07 3.99 2.05 5.12 2.35 13 12 0.10 5 10 11 WAMR156725 Bungle Bungles East M 79.13 14.54 6.82 19.37 22.11 8.61 2.83 6.64 3.67 6.89 36.24 11.21 12.71 4.14 1.73 5.36 2.21 14 11 0.40 9 5 9 10 CCM0729 Doongan N/C F 82.59 15.38 7.24 19.81 22.42 8.85 2.81 7.24 4.46 6.51 33.68 10.97 13.39 4.07 1.63 4.79 1.81 14 13 0.33 6 9 10 CCM1074 Ellenbrae N/C F 76.59 15.07 5.94 18.35 21.25 8.10 2.34 7.19 3.53 6.90 35.10 9.99 12.11 3.79 1.55 4.76 1.71 13 11 0.50 6 9 11 WAMR172865 Ellenbrae N/C F 81.65 15.82 6.34 19.60 23.35 8.50 2.67 6.83 3.74 7.39 35.18 9.87 12.62 3.89 1.65 4.48 1.97 12 11 0.15 6 9 11 WAMR172866 Ellenbrae N/C M 80.56 16.01 6.66 19.84 23.06 8.58 2.87 6.91 3.70 6.95 34.54 10.06 12.68 4.21 1.62 4.49 1.79 12 11 0.25 9 6 9 11 WAMR172867 Ellenbrae N/C F 82.96 15.71 6.34 20.20 24.37 8.75 2.71 7.25 3.77 7.46 34.62 10.31 12.26 4.08 1.60 4.65 1.99 10 11 0.33 6 8 10 WAMR172868 Ellenbrae N/C M 78.88 14.59 6.66 19.40 22.45 8.42 2.73 7.08 3.74 6.80 37.75 9.91 11.72 3.38 1.54 4.26 1.76 12 11 0.33 6 8 10 WAMR174049 Ellenbrae N/C F 79.22 16.44 5.57 19.15 22.33 8.67 2.66 7.42 3.77 7.04 35.06 9.74 13.11 3.94 1.60 4.34 1.92 13 11 0.50 6 9 9 CCM1740 Mitchell Plateau MP M 82.03 15.86 7.01 21.91 24.34 10.45 2.95 7.37 4.14 7.23 33.96 11.61 13.23 4.16 1.56 4.51 1.55 12 12 0.33 11 10 9 11 CCM1741 Mitchell Plateau MP F 86.99 15.42 7.76 22.16 25.42 9.65 3.18 7.23 4.34 7.58 37.81 11.93 14.21 4.09 1.82 5.41 1.67 15 13 0.40 8 9 12 WAMR138878 Mitchell Plateau MP F 89.74 15.10 7.83 21.61 24.73 9.77 2.93 7.24 4.41 7.55 39.10 10.73 14.54 4.15 1.63 1.92 14 13 0.40 7 10 10 WAMR164908 Mitchell Plateau MP F 89.30 16.37 7.94 22.39 25.37 10.02 3.17 7.65 4.21 10.27 38.98 11.78 14.94 4.16 1.67 5.29 2.02 14 12 0.40 10 9 10 WAMR164909 Mitchell Plateau MP F 92.80 16.13 7.70 22.92 26.14 10.20 2.92 6.57 4.83 8.39 42.76 11.86 14.61 3.88 1.66 4.41 2.09 13 12 0.33 11 9 10 WAMR164910 Mitchell Plateau MP F 94.30 16.78 8.79 23.33 25.85 10.55 2.99 7.78 4.43 9.23 43.29 11.43 14.62 4.19 1.94 5.20 1.98 13 13 0.25 8 9 11 WAMR168739 Mitchell Plateau MP F 89.17 14.78 7.15 21.77 25.29 9.69 3.01 7.43 4.69 7.43 39.43 11.69 14.55 4.03 1.47 4.84 1.71 14 13 0.25 11 9 11 WAMR168903 Mitchell Plateau MP F 83.84 15.20 6.98 21.81 25.19 9.63 3.00 6.67 4.52 7.57 36.62 11.23 13.86 3.70 1.49 3.94 1.72 13 12 0.33 11 9 11 WAMR168904 Mitchell Plateau MP M 86.04 16.08 7.37 22.82 25.54 10.00 3.09 7.54 4.45 7.80 35.55 11.51 14.29 3.82 1.80 5.24 2.02 12 12 0.60 6 10 9 11 CCM1639 Mt Elizabeth N/C F 86.65 15.72 7.53 20.38 23.90 8.90 2.79 7.19 4.10 6.93 40.34 10.88 13.85 4.37 1.79 4.84 1.79 14 12 0.50 7 8 10 CCM1755 North N/C M 87.03 16.96 5.65 20.85 23.88 9.01 2.88 7.69 3.95 7.00 40.60 10.80 13.02 3.82 1.54 4.82 1.94 12 11 0.33 6 9 10 WAMR151963 North N/C M 78.06 13.85 6.66 19.50 21.91 8.38 2.64 6.41 4.04 6.85 32.69 10.30 13.06 4.17 1.48 5.06 1.73 12 12 0.60 11 6 9 10 WAMR164863 North N/C M 93.65 17.51 6.94 22.22 25.45 10.05 3.10 7.52 4.61 7.92 41.22 10.34 13.91 4.40 2.09 5.42 2.25 13 12 0.50 14 5 8 9 WAMR164864 North N/C F 82.48 15.27 6.74 20.74 22.81 9.15 2.89 7.10 4.41 7.33 38.85 10.74 12.84 4.42 1.64 4.94 1.87 14 12 0.25 5 8 11 WAMR168463 North N/C F 91.69 16.82 7.74 21.73 25.58 9.55 3.19 6.55 4.49 8.32 42.24 11.18 14.17 4.38 1.54 5.86 2.01 12 12 0.25 6 9 11 WAMR172341 North N/C M 79.88 14.63 6.91 19.65 22.87 6.76 2.73 6.65 4.28 6.63 32.69 10.70 12.64 3.98 1.66 5.24 1.93 14 12 0.40 15 7 9 11 WAMR174023 North N/C F 92.42 15.84 6.61 21.60 24.54 9.44 3.06 7.65 4.13 7.13 43.06 11.30 13.01 4.18 1.58 13 12 0.50 6 9 9 WAMR174025 North N/C F 92.93 16.70 7.53 22.44 25.25 9.92 3.22 8.02 4.39 7.95 41.78 11.31 14.25 4.50 1.86 4.55 2.02 13 12 0.50 5 9 10 WAMR174030 North N/C M 91.87 17.00 8.37 22.20 24.51 9.89 3.13 7.50 4.43 7.91 40.63 11.71 14.38 4.52 1.89 4.87 2.12 15 13 0.40 14 6 8 9 CCM3196 Nyulasy East M 85.10 14.95 7.51 20.77 23.54 9.31 2.71 7.10 4.08 7.07 38.14 10.90 13.83 4.59 1.86 5.26 2.16 14 13 0.33 11 6 9 10 WAMR151005 Nyulasy East F 89.42 16.55 6.80 20.45 23.59 8.61 2.89 7.32 4.78 7.25 39.26 11.34 13.91 4.33 2.03 4.96 2.28 13 12 0.10 6 9 11 CCM1556 Police Valley N/C F 82.97 17.04 5.91 20.22 23.90 9.03 2.79 7.50 3.32 6.99 34.02 10.69 12.75 4.68 1.53 5.48 1.86 12 11 0.25 6 9 11 CCM1557 Police Valley N/C F 74.58 14.33 5.60 18.23 21.13 7.94 2.40 6.60 3.30 6.21 32.55 9.17 11.64 4.35 1.76 4.81 1.79 13 12 0.33 6 9 9 CCM1402 South South M 71.99 15.81 4.66 18.22 20.86 7.76 2.64 5.93 3.58 7.46 28.05 9.65 11.57 3.90 1.47 4.63 1.62 13 11 0.25 9 6 9 10 CCM1490 South South F 82.24 15.51 6.40 20.46 23.21 8.92 2.76 7.31 4.21 7.63 33.10 10.34 12.49 4.29 1.74 4.52 2.10 11 12 0.33 7 8 10 CCM2844 South South F 83.70 15.14 5.82 19.61 24.17 8.76 2.82 6.42 4.56 7.13 39.87 10.53 12.46 4.25 1.58 5.14 1.74 13 13 0.40 7 9 9 WAMR156728 South South F 86.18 16.59 7.66 21.51 25.14 9.35 2.87 6.77 4.54 7.75 36.13 11.69 14.12 4.78 1.73 5.18 1.86 14 13 0.50 6 9 10 WAMR171671 South South M 84.62 15.07 7.46 20.38 23.68 8.84 2.65 6.91 4.49 7.15 36.12 10.90 13.84 4.12 1.64 4.87 1.84 14 12 0.33 11 7 9 10 WAMR171672 South South M 81.86 15.71 6.70 20.80 24.14 8.95 2.83 6.70 4.21 7.07 35.44 10.72 13.53 4.42 1.86 4.65 1.85 12 13 0.33 12 6 9 10 WAMR171673 South South F 72.23 12.98 6.15 18.28 20.46 7.73 2.10 6.20 3.92 6.39 32.76 10.33 11.71 3.84 1.33 4.26 1.59 11 12 0.33 6 9 10 WAMR171674 South South M 76.26 13.45 6.05 18.87 20.99 8.77 2.60 5.33 3.51 6.56 33.12 9.43 12.71 4.17 1.48 4.27 1.79 13 12 0.25 10 6 9 10 WAMR171677 South South F 82.05 16.25 7.29 21.01 22.80 9.14 3.01 7.57 3.73 7.19 35.38 10.17 12.46 4.16 1.78 4.61 2.03 12 11 0.50 6 8 9 WAMR171675 South-west WC M 76.10 14.59 5.98 18.69 22.34 6.16 2.89 6.33 3.68 6.99 33.51 9.34 12.03 3.59 1.45 4.24 1.71 13 11 0.30 8 7 7 9 WAMR171676 South-west WC M 83.38 14.33 7.68 20.27 22.63 8.96 3.02 6.54 3.80 7.30 35.15 10.33 13.06 3.78 1.80 4.33 2.19 13 12 0.25 5 7 9 10 WAMR173942 Theda N/C F 81.97 15.30 6.37 19.95 22.74 8.82 2.83 7.23 4.17 7.45 36.49 10.28 12.69 4.00 1.73 4.67 1.89 11 12 0.33 6 9 11 WAMR174106 Theda N/C F 84.30 15.93 5.97 20.41 23.67 8.68 3.00 7.22 4.28 6.94 40.27 10.06 13.12 4.15 1.70

Clades are labelled as follows: East – East, MP – Mitchell Plateau, N/C – North/Central, South – South, WC – West-coast.

143

Chapter 3: Contribution of climate and topography to diversification

Figure A3.1. GMYC species delimitation of mitochondrial (ND2) data across the Oedura gracilis complex. Comparison of groups delineated using this method versus ≥8% Tamura-Nei divergences is indicated by the coloured bars; colour scheme matches that of populations shown in Fig. 3.2.

144

Chapter 3: Contribution of climate and topography to diversification

Figure A3.2. Phylogenetic reconstruction of one nuclear locus (K1AA1549) for the Oedura gracilis complex and allied taxa. Node support values are indicated where support was high (>75 Maximum-Likelihood bootstraps), Bayesian posterior probabilities not shown as no nodes were recovered with strong support. ≥8% Tamura- Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

145

Chapter 3: Contribution of climate and topography to diversification

Figure A3.3. Phylogenetic reconstruction of one nuclear locus (KIF24) for the Oedura gracilis complex and allied taxa. Node support values are indicated where support was high (>75 Maximum-Likelihood bootstraps, >95 Bayesian posterior probabilities). ≥8% Tamura-Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

146

Chapter 3: Contribution of climate and topography to diversification

Figure A3.4. Phylogenetic reconstruction of one nuclear locus (PDC) for the Oedura gracilis complex and allied taxa. Node support values are indicated where support was high (>75 Maximum-Likelihood bootstraps), Bayesian posterior probabilities not shown as no nodes were recovered with strong support. ≥8% Tamura-Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

147

Chapter 3: Contribution of climate and topography to diversification

Figure A3.5. Phylogenetic reconstruction of one nuclear locus (PRLR) for the Oedura gracilis complex and allied taxa. Node support values are indicated where support was high (>75 Maximum-Likelihood bootstraps, >95 Bayesian posterior probabilities). ≥8% Tamura-Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

148

Chapter 3: Contribution of climate and topography to diversification

Figure A3.6. Phylogenetic reconstruction of one nuclear locus (PTPN12) for the Oedura gracilis complex and allied taxa. Node support values are indicated where support was high (>75 Maximum-Likelihood bootstraps, >95 Bayesian posterior probabilities). ≥8% Tamura-Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

149

Chapter 3: Contribution of climate and topography to diversification

Figure A3.7. Phylogenetic reconstruction of one nuclear locus (RAG1) for the Oedura gracilis complex and allied taxa. Node support values are indicated where support was high (>75 Maximum-Likelihood bootstraps, >95 Bayesian posterior probabilities). ≥8% Tamura-Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

150

Chapter 3: Contribution of climate and topography to diversification

Figure A3.8. Phylogenetic reconstruction of concatenated mtDNA (ND2) + nuDNA (K1AA1549, KIF24, PDC, PRLR, PTPN12, RAG1) dataset for Oedura gracilis and allied taxa. Node support values with high support (>75 BS, >95 PP) are shown. ≥8% Tamura-Nei divergent mtDNA lineages within O. gracilis are indicated by colour scheme matching that of populations shown in Fig. 3.2.

151

Chapter 3: Contribution of climate and topography to diversification

Nyulasy

South

Figure A3.9. Principal Components Analysis (PCA) of POFAD multidimensional scaling results conducted on the six nuclear loci for the Oedura gracilis complex. Symbol colours correspond to the ≥8% divergent mitochondrial lineages and match population colour scheme in Fig. 3.2., whilst coloured ellipses indicate the broader clades recovered with majority of analyses and datasets, as labelled.

152

Chapter 4: Continental islands support ancient genetic diversity

Continental islands support ancient genetic diversity

within gekkonid lineages (Gehyra spp.) in the

Australian Monsoonal Tropics

Photograph courtesy of Jordan Vos

153

Chapter 4: Continental islands support ancient genetic diversity

Abstract

While speciation dynamics and evolutionary processes have been well-studied on oceanic islands, the potential for extreme isolated diversity within continental islands has been overlooked due to their relatively recent isolation and proximity to mainland populations. The Australian Monsoonal Tropics (AMT) is an ideal location to investigate potential levels of cryptic diversity and/or phenotypic divergence on such islands. Rock isolates on the mainland have acted as refugia supporting cryptic diversity in this region throughout a variable climatic history; whether land-bridge islands have played a similar role remains unclear. Here, I tested genetic and morphological divergence between island and mainland populations in two gecko clades (Gehyra spp.) distributed across the west Kimberley in the AMT. Using next-generation sequencing I assembled large molecular datasets and implemented population genomic and phylogenetic dating methods to compare the depth and age of genetic diversity within clades. I also assessed phenotypic divergence potentially associated with ecological selection in island environments. I found limited evidence of morphological divergence on islands. However, I did find support for deep genetic structure, including multiple divergences of single-island endemic lineages pre-dating the most recent island formation by millions of years. This implies that populations on these islands have remained isolated even at low sea-levels, though only for particular ecological specialists. These results show that continental islands are capable of supporting old, genetically divergent taxa as part of highly heterogeneous tropical savannah systems.

154

Chapter 4: Continental islands support ancient genetic diversity

4.1. Introduction

Despite the occurrence of large tropical savannah biomes within most major continents, the true levels of biodiversity within these biomes are still poorly known

(Carnaval et al. 2014; da Silva & Bates 2002; Grace et al. 2006; Martin et al. 2012;

Mayle et al. 2000; Waeber et al. 2015). However, there are emerging patterns of highly geographically-structured, and often cryptic diversity within these environments

(Criscione et al. 2012; Guarnizo et al. 2016; Lorenzen et al. 2012; Miraldo et al. 2016;

Moritz et al. 2015; Vieites et al. 2009; Werneck et al. 2012). A great deal of this diversity tends to be linked to rock escarpments and ranges that dissect the woodlands and grasslands of the savannah biomes (Domingos et al. 2014; Guarnizo et al. 2016;

Nogueira et al. 2011; Pepper et al. 2011b; Potter et al. 2012). As rock habitats are typically highly stable in the face of environmental variation, it is assumed that the genetic structuring associated with this topography is a result of rock isolates functioning as refugia which allow persistence and in situ diversification of lineages over long timeframes (e.g. Byrne 2008; Pepper et al. 2008; Pepper et al. 2011b). The complex structure of rock habitats can provide climatically-buffered microhabitats which protect against intense climatic variation which would otherwise lead to broad- scale extirpation of populations (Byrne 2008; Cameron 1992; Couper & Hoskin 2008;

Glaw et al. 2006). Rock habitats also often retain more water, and offer a greater degree of protection against the impacts of fire (Couper & Hoskin 2008; Glaw et al. 2006;

Oliver et al. in press). Furthermore, these conditions also support persistence of protective vegetation such as hummock grasses, which in turn support food sources (e.g. invertebrates and small reptiles) for many reptile and mammal species (Anderson et al.

2016; Cameron 1992; Pianka 1981; Rabosky et al. 2007; Wallis 2001).

155

Chapter 4: Continental islands support ancient genetic diversity

To an extent, rock isolates act as islands within savannah biomes, allowing highly specialised or climatically vulnerable species to endure harsh periods of environmental change within insular environments (Byrne 2008; Oliver et al. 2012; Oliver et al. in press; Pepper et al. 2013; Pepper et al. 2011a). We might ask, then, to what extent continental or land-bridge islands have played a similar role in the climatic and evolutionary history of such biomes, and how they in turn contribute to processes of diversification within tropical savannah environments. Oceanic islands have been well studied for decades in terms of the evolution of their unique endemic biotic assemblages, island speciation dynamics and community turn-over (e.g. Emerson 2002;

Losos & Ricklefs 2009; Patiño et al. 2015; Paulay 1994; Vitousek et al. 2013;

Whittaker & Fernández-Palacios 2007). However, the potential extreme isolated diversity on continental islands are comparatively overlooked, with research being focused instead on community assemblages as a subset of the related mainland biota

(e.g. Bell et al. 2012; Bittencourt-Silva et al. 2014; Si et al. 2015; Vitousek et al. 2013;

Wilcox 1978).

The assumption is that land-bridge islands are not expected to exhibit cases of long- term divergence, due to their relatively recent isolation and their proximity to mainland populations (Bittkau & Comes 2005; Velo-Antón et al. 2012). However, the insular nature of these habitats can select for ecological divergence by providing opportunities for taxa to expand into novel niches, which can result in rapid phenotypic divergence without corresponding levels of genetic divergence (e.g. Grazziotin et al. 2006; Rojas-

Soto et al. 2010). Studies of tropical savannah islands indicate varied patterns; for instance in Scinax frogs in Brazil, phenotypic divergence has occurred for multiple island species relative to mainland congeners, though not all phenotypic divergence has

156

Chapter 4: Continental islands support ancient genetic diversity

been coupled with genetic divergence (see Bell et al. 2012; also ; Grazziotin et al.

2006). Furthermore, whilst the small size, similarity of ecological niches to the mainland and the proximity of continental islands close to the coastline may be factors contributing to lower overall diversification, they may also contribute to our understanding of initial processes promoting early evolutionary pathways (Velo-Antón et al. 2012).

In contrast, new evidence is emerging, particularly within tropical regions, indicating continental islands can also harbour highly genetically divergent but morphologically cryptic diversity, such as has been observed in the Brookesia chameleons of

Madagascar (Glaw et al. 2012). This pattern is also emerging on multiple land-bridge islands associated with the tropical savannahs of the Australian Monsoonal Tropics

(AMT). In particular, multiple islands of the Top End (Northern Territory) region have recently begun to be recognised as hotspots for deeply divergent, cryptic reptile lineages

(Moritz et al. 2015; Potter et al. 2016; Rosauer et al. 2016). There is extensive evidence that many islands of the Kimberley region (Western Australia) also support cryptic and divergent endemic lineages, some restricted to single islands. However, this diversity has thus far been found predominantly in invertebrates (snails; Criscione & Köhler

2015; Gibson & Köhler 2012; Johnson et al. 2010; Köhler 2010, 2011a, b; although see

Lyons et al. 2014). In order to understand which of these processes – cryptic genetic diversification or ecologically-driven phenotypic diversification – is occurring in vertebrates, we need to obtain high-quality genetic and phenotypic data from a well- studied continental island system. These two processes may not even be mutually exclusive.

157

Chapter 4: Continental islands support ancient genetic diversity

The vast assemblage of remote Kimberley islands (~2633 islands) provides such a system: these islands are part of an ancient, submerged coastline (Gibson & McKenzie

2012). The majority of these land-bridge islands have particularly short separation distances (generally <4km) from the mainland and have experienced an extensive history of periodic isolation from the mainland and other islands; the most recent of which occurred relatively recently (~8 kya; Burbidge & McKenzie 1978; Nix & Kalma

1972) due to rising sea-levels after the last glacial maximum (De Deckker & Yokoyama

2009). The topographic and vegetative composition of the Kimberley islands closely resembles that of the nearest west-coast mainland region, albeit slightly simplified

(Burbidge & McKenzie 1978; Gibson & McKenzie 2012; Lyons et al. 2014). Likewise, they are generally subject to the same climatic gradient as the mainland of the west- coast; transitioning from high rainfall (>1000 mm annually) in the central- and north- west, to dryer conditions further south and east (Gibson 2014). In addition, the

Kimberley experienced a dynamic history of climatic variation with transitions between especially arid and mesic periods associated with Pleistocene glacial cycles (Proske et al. 2014; Reeves et al. 2013), which presumably resulted in the high genetic diversity within rock refugia discussed above.

These remote Kimberley islands are home to members of a genus of gekkonid prolific on the mainland. The genus Gehyra comprises a significant proportion of the

Australian gecko fauna, with ~20 recognised species and many divergent lineages yet to be described (e.g. Doughty et al. 2012b; Hutchinson et al. 2014; Oliver et al. 2016;

Sistrom et al. 2009). The genus is known to have colonised Australia from Asia, most likely between 20–30 Ma, and consequently is highly speciose within the northern AMT region of the continent (Heinicke et al. 2011). Despite relatively high morphological

158

Chapter 4: Continental islands support ancient genetic diversity

conservatism, within the northern Australian savannah environments community assemblages can consist of multiple species of Gehyra; often with partitioning of the environmental niche space by size (e.g. often small, intermediate and large species occur sympatrically) or ecology (e.g. arboreal, or rock-specialists – including occupying different kinds and sizes of rock habitat; Sistrom et al. 2013; Sistrom et al. 2012).

For this particular study I selected two clades—the Gehyra occidentalis complex and

G. xenopus—due to their broadly sympatric distributions across the high-rainfall zone of the west Kimberley including populations on several Kimberley islands. The most recent study of these groups hinted at greater genetic diversity within both clades, as well as possible morphological (size) variation within G. occidentalis (Doughty et al.

2012b). Given these findings and the presence of multiple related mainland and island lineages, these clades provide an excellent opportunity to test whether land-bridge island populations within the Kimberley show evidence of i) genetic, and/or ii) phenotypic differentiation from mainland congeners.

The Gehyra occidentalis complex encompasses two currently recognised species: G. multiporosa distributed along the mesic west-coast and associated islands, and G. occidentalis spanning islands in the south-west off Yampi peninsula, across the southern King Leopold and Durack Ranges, and also present in the southern Devonian

Reef System limestone ranges (see Fig. 4.1.a). Gehyra xenopus occupies a comparatively restricted distribution predominantly along the west-coast and nearby islands (Fig. 4.1.a). G. xenopus is a large-bodied species of Gehyra that is highly specialised for climbing (scansorial) with long limbs and large toe-pads, and is only known to occupy rock habitats with particularly large open rock faces (Doughty et al.

2012b). Within the Kimberley, habitat of this kind is mostly distributed along the west-

159

Chapter 4: Continental islands support ancient genetic diversity

coast and geographically proximate islands; a similar distribution to those occupied by equally specialised large scansorial gecko in the genera Oedura and

Pseudothecadactylus (see Chapter 3; Oliver et al. 2014b). This distribution is also tightly correlated with the highest-rainfall zone within the Kimberley (Fig. 4.1.b). In addition, G. xenopus has a very reduced scale-size, whereas scale-size tends to be larger in arid-adapted squamates as a means to reduce desiccation (see Wegener et al. 2014, and references therein). Considering these factors it could be predicted that G. xenopus is a more mesic-reliant lineage, and hence particularly vulnerable to the impacts of aridification.

The G. occidentalis complex has a much wider distribution than G. xenopus (Fig.

4.1.a) which extends into the more arid southern regions of the Kimberley (Fig. 4.1.).

The taxa comprising this clade are also smaller- to intermediate- body-size, and could be considered more ecologically-generalist, tending to be found in any form of available rock-outcropping and even on nearby trees. In addition, G. occidentalis in particular, which is distributed furthest south, tends to have larger scales; suggesting these lizards are better predisposed to more arid conditions. Following this assumption, I predicted that this complex would be less likely to suffer notable impacts of aridity within the most mesic area of the Kimberley (the distribution of G. xenopus for example).

Pursuing this theory further, it might be expected that significant climatic drying associated with broad-scale aridification would likely result in extirpation of arid- sensitive lineages throughout much of their range, causing sundering of a previously widespread population into several disconnected lineages persisting only in the most buffered micro-refugia regions (e.g. Byrne 2008; Oliver et al. 2014b). In contrast, more arid-adapted taxa would only begin to suffer similar impacts of drying in the extent of

160

Chapter 4: Continental islands support ancient genetic diversity

their range exposed to the harshest environmental variation (e.g. in this case the southern limits of the Kimberley).

In this study, I sampled Gehyra specimens from both island and mainland Kimberley populations and used next-generation sequencing methods (targeted exon capture) to assemble extensive molecular datasets (thousands of loci) for both focal clades. To test whether there was genetic divergence between island and mainland populations, I used phylogenetic and population genomic methods to assess genetic diversity across different molecular datasets (mtDNA and nuDNA). In addition I used phylogenetic dating to estimate the timeframes of diversification within clades, to determine the isolation ages for island lineages. I also assessed phenotypic divergence (e.g. changes in body size) within clades that may be associated with island diversification processes. I predicted that the more specialised and geographically restricted G. xenopus would exhibit greater genetic diversity and consequently higher likelihood of extirpation and isolation of lineages over the sampling area. In contrast, I expected lower genetic diversity in the more widespread and tolerant generalist taxon (G. occidentalis complex), due to its capacity to traverse a variety of habitat types and thereby increase population connectivity. However, an increased ability to colonize novel ecological niches may also lead to increased morphological differentiation in this taxon, particularly in insular environments. In short, I ask whether the biodiversity within the

Kimberley and its land-bridge islands is likely to reflect a closely related subset of the mainland biota, indicating a history of close connection and environmental similarity; or whether these islands are likely to have functioned as evolutionary refugia for cryptic diversification in the same way other rock isolates within the savannahs of the mainland have structured genetic diversity?

161

Chapter 4: Continental islands support ancient genetic diversity

4.2. Materials & methods

4.2.1. Sampling and mitochondrial sequencing

All specimens I used in genetic analyses are listed in the Appendix: Table A4.1. I assembled genetic data for 231 Gehyra multiporosa and G. occidentalis (hereon referred to as G. occidentalis complex), and 104 G. xenopus samples, the majority of which were collected in the field as part of this study or otherwise registered in Australian museum tissue collections (Fig. 4.2., Table A4.1a). I then selected subsets of 23 (G. occidentalis complex) and 32 (G. xenopus) representing both geographic and genetic diversity, that were included in the exon capture experiment (Table A4.1a). In addition,

I also sequenced four nuclear loci (BDNF, Cmos, PDC, RAG1) for 45 outgroup Gehyra taxa from within and outside Australia for inclusion in phylogenetic dating analyses in which both focal clades were considered (Table A4.1b).

I obtained genomic DNA extractions from liver or tail tissue using a Qiagen DNeasy extraction kit (Qiagen, Valencia, CA) or a Qiaextractor (Qiagen, Doncaster, Victoria,

Australia). I purified the PCR products using 1uL of a 20% dilution of ExoSAP-IT

(US78201, Amersham Biosciences, Piscataway, NJ), incubated at 37°C for 30 min, followed by 80°C for 15 min. A genetic services company (Macrogen, Seoul, South

Korea) was used to sequence cleaned products, with each amplicon sequenced in both directions. A portion of the mitochondrial locus ND2 (1029 base pairs [G. occidentalis complex], and 1038 bp [G. xenopus]) was sequenced for all individuals. I assembled and edited gene sequences in GENEIOUS v.6.1.7 (Drummond et al. 2008), and I translated alignments into amino acids and inspected them by eye to confirm translation and reading frame. Primer sequences and amplification protocols are provided in Table

162

Chapter 4: Continental islands support ancient genetic diversity

A4.2. GenBank numbers for all new mitochondrial sequences and details of nuclear datasets submitted to dryad are provided in Table A4.1.

4.2.2. Exon capture

4.2.2.1. Library design

The targeted exon capture design implemented for this work is covered in detail in

Bragg et al. (2015). Briefly, oligonucleotide probes were designed that were complementary to a selection of protein coding exons identified across two Gehyra transcriptomes (G. nana and G. oceanica) that were assembled and annotated against the Anolis genome used in Bragg et al. (2015) following the approach of Bi et al.

(2012). In total, 1717 exons (>200 bp) were targeted in this capture, and a probe set

(SeqCap EZ Developer Library) to target these exons was consequently designed and synthesized by Roche NimbleGen.

4.2.2.2. In solution capture

In preparation for this study I conducted two exon capture experiments over which my sampling was spread; the first capture had 56 and the second had 75 individuals pooled to assess performance of increased sample capacity per capture on the resulting data acquisition. Additional samples included in these captures were predominantly for use in a separate study; however I did include sequences from outgroup samples for use in the dating analyses (see Table A4.1b).

To prepare samples for creation of genomic libraries I estimated double-stranded

DNA concentrations using a DropSense96 spectrophotometer (Trinean), and assessed molecular weight of the DNA by running extractions out on 1% agarose gel and

163

Chapter 4: Continental islands support ancient genetic diversity

comparing against a 100 bp ladder (Axygen). Based on concentration estimations I then diluted extractions to equal concentrations (1400 ng) in 100 uL volumes. Dilutions with high molecular weights (>500 bp) I then sheared by sonication using a BioRuptor Plus

(Diagenode) for 4–12 cycles depending on the molecular weight. I re-ran samples on gels to confirm adequate size range for hybridisation (<500 bp), and sheared for a further three cycles if fragments were still too long.

Next I prepared genomic libraries for the 55 samples I had selected for further nuDNA analysis, following the protocol of Meyer & Kircher (2010) with modifications from Bi et al. (2012). This process consisted of cleaning, blunt-end repair, adapter ligation and fill-in, and indexing PCRs where I double-indexed samples with unique combinations of index primers (48 forward and seven reverse) so sequences could be identified to individual samples after multiplexing. The modified protocol involved keeping samples on Sera-Mag beads (GE Healthcare) throughout the process instead of eluting off for thermal cycling. I initially ran indexing PCRs for 12 cycles and then ran the PCR products out on gels to confirm successful amplification. Any samples that were too faint I ran for more cycles (up to 21) until I had two successful independent index-PCRs available for each sample, reducing chance of PCR bias. The two index-

PCRs I then pooled for each sample and again cleaned the products on Sera-Mag beads.

I quantified DNA concentrations of samples again with the DropSense96 spectrophotometer to calculate volumes required so I could pool all samples into a single library in equimolar concentrations (total concentration 1600 ng). The concentration of this library I then estimated using a Qubit 2.0 Fluorometer (Invitrogen) to calculate the volume of the pooled library I required for hybridisation (1200 ng desired).

164

Chapter 4: Continental islands support ancient genetic diversity

The procedure for hybridisation of the pooled DNA library to the target probes followed the SeqCap EZ Developer Library user’s guide (Roche NimbleGen), but was modified by replacing the standard oligos with barcode-specific blocking oligos designed to block the custom barcode adapters. To the pooled library I added 1.8 uL of each index-specific 10 uM blocking oligo (x56), 1 uL of 1000 uM universal blocking oligo, and 25.2 uL of 197.9 ng/uL skink Cot-1 DNA (prepared with Lampropholis coggeri DNA following Trifonov et al. (2009)). I next dried the library down using a

DNA vacuum concentrator, and then re-eluted with 7.5 uL of hybridisation buffer and 3 uL of hybridisation component A. I then denatured the library for 10 min at 95°C before

I combined it with 4.5 uL of the EZ Library probes and incubated at 47°C for 72 h.

After hybridisation I washed the library and conducted two independent enrichment

PCRs to amplify the post-capture library (17 cycles with Phusion High-Fidelity DNA

Polymerase [Thermo Scientific]; Roche NimbleGen).

I used quantitative PCRs (qPRCs) to assess the capture success, which I ran using 1 uL aliquots of pre- and post-hybridisation sample libraries. The qPRCs I conducted using the Roche LightCycler 480 SYBR Green I Kit (Thermo Fisher Scientific Inc.) and

I used two positive controls (target primers – CAP09082 and CAP2260) and one negative control (non-target – CAP13990), following methods of Bi et al. (2012).

Results ensured enrichment of target regions (positive controls) and de-enrichment of non-targets (negative control) in the post-capture compared to pre-capture library. I also measured concentrations of pre- and post-capture DNA libraries once more using the

Qubit 2.0 fluorometer.

165

Chapter 4: Continental islands support ancient genetic diversity

Once libraries were ready each capture was sequenced on a single Illumina

HiSeq2000 lane at the Biomolecular Resource Facility, John Curtin School of Medical

Research, ANU, Canberra.

4.2.2.3. Bioinformatics pipeline

I used scripts developed for the SCPP (Sequence capture using PCR-generated probes) pipeline (pre-cleanup and scrubReads; Peñalba et al. 2014) to clean and trim raw sequence reads by removing duplicate, low complexity and contaminant sequences, and trimming adaptors and low quality base reads. I then assembled clean sequence reads using a workflow detailed in Bragg et al. (2015). In brief, for each sample de novo assemblies of target exons were created using VELVET v1.2.08 (Zerbino & Birney 2008) and resulting contigs were combined using CAP3 v08/06/13 (Huang & Madan 1999).

EXONERATE v2.2.0 (Slater & Birney 2005) was then used to identify contigs that aligned to target exons with a minimum overlap margin of 65%, and trimmed anything beyond the target region. Contigs were successfully aligned for 1603 of the target exons for the target clades of this study.

I created two separate datasets—concatenated exon alignments, and alignments of biallelic single nucleotide polymorphisms (SNPs)—for each target group, both of which involved alignment and filtering of exons via the EAPHY v1.2 pipeline developed by

Blom (2015). EAPHY created sequence alignments with MUSCLE v3.8.31 (Edgar

2004), checked coding of amino acids, and trimmed missing data from alignment ends.

I set initial filtering conditions to remove any contigs or loci that did not meet the following criteria: minimum contig length 150 bp, maximum ratio of Ns:basecalls 0.40, minimum number of individuals (contigs) per locus 65%. I had alignments flagged to be

166

Chapter 4: Continental islands support ancient genetic diversity

visually inspected if stop codons were detected or if any sequences were >75% divergent from the alignment consensus. For each target clade ~500 exons were flagged which I visually inspected, and ~100 I ultimately removed as potential paralogues or because the initial contigs contained stop codons. I ran EAPHY a number of times for each clade, with and without outgroups, and with diplo- and haplo-type sequences, so as to provide datasets suitable for multiple analyses.

4.2.3. Preliminary phylogenetics

I initially assessed genetic structure and diversity within each target clade by estimating mitochondrial (mtDNA) phylogenies using Maximum-Likelihood (RAxML v8.0.24; Stamatakis 2006; Stamatakis et al. 2008) and Bayesian (MrBayes v3.2.2;

Ronquist & Huelsenbeck 2003) methods. These analyses I ran on the CIPRES Science

Gateway 3.1 for online phylogenetic analysis (Miller et al. 2010), using default CIPRES

RAxML settings and including the GTR-CAT model (Stamatakis et al. 2008), also ceasing bootstrapping once MRE-bootstrapping criteria were reached. I determined the partitioning strategies and nucleotide substitution models (Bayesian analyses) for these analyses based on the Bayesian information criterion (BIC) in PARTITIONFINDER v2.0.0

(Lanfear et al. Forthcoming; see Table A4.3. for details). I ran Bayesian analyses with four independent Markov Chain Monte Carlo (MCMC) chains for 4 x 10 million generations, sampling every 1000. Convergence and log-likelihood stability I assessed using TRACER v1.6 (Rambaut et al. 2014) and ARE WE THERE YET (AWTY;

Wilgenbusch et al. 2004), also ensuring effective sample sizes were adequate (>100) for all parameters. The first 20% of samples I then discarded as burn-in, and constructed

167

Chapter 4: Continental islands support ancient genetic diversity

Maximum Clade Credibility (MCC) trees using TREEANNOTATOR v2.3.1 (Rambaut &

Drummond 2002–2015).

Once I had established preliminary mtDNA structure within both target clades, I selected subsamples of geographically and genetically representative individuals (31

Gehyra occidentalis complex, 32 G. xenopus) for sub-genomic nuclear (nuDNA) sequencing using targeted exon-capture arrays. Some samples failed to sequence, were contaminated, or were not the species identified by mtDNA analysis which reduced the final numbers for nuDNA analyses across the two clades. In addition, not all exons were recovered across each clade, and some exons were recovered for one group though not the other. I experimented with stringency settings in the EAPHY pipeline to assess the sizes of datasets when the minimum number of individuals required to accept a locus was varied by 5% increments between 75–100%. As datasets were considered sufficient size when only loci sequenced for 100% of samples were included, and as large amounts of missing data can cause issues in certain analyses, I chose to only use complete datasets for final downstream analyses. RAxML analyses I ran for concatenated exon datasets with the same settings as above, including using a partitioning strategy as determined in PARTITIONFINDER (see Table A4.3.).

4.2.4. Phylogenetic dating

The time available for lineage diversification to occur is an important factor when comparing diversity levels between clades and also helps to put regional divergence patterns into a broader context. For these reasons, I chose to estimate diversification timeframes for my target groups. To estimate ages I analysed a four locus nuDNA dataset (BDNF, Cmos, PDC, and RAG1) combined with secondary constraints on key

168

Chapter 4: Continental islands support ancient genetic diversity

nodes (previously estimated mean and standard deviations) derived from other fossil calibrated squamate phylogenies (Gamble et al. 2015; Oliver et al. in press) – a normal root prior (mean 30 Ma; SD 4), and a normal prior for the divergence between the

Australian and Pacific Gehyra clades (mean 22 Ma, SD 3). I carried out dating estimates using BEAST v1.8.0 (Drummond et al. 2012) with molecular model GTR+G, and used a relaxed uncorrelated log-normal clock rate and Birth-death speciation prior, with the analysis run for 50 million generations, sampling every 50000. I ensured stability and convergence using TRACER (Rambaut et al. 2014) and AWTY (Wilgenbusch et al.

2004), and once 20% burn-in was removed I summarised final MCC trees with

TREEANNOTATOR v2.3.1 (Rambaut & Drummond 2002–2015).

4.2.5. Nuclear assessment of divergent mitochondrial lineages

I produced SNP datasets via the pipeline EAPHY (Blom 2015) with settings to ensure no missing data and random selection of a single biallelic SNP from any non-linked exons, and using these SNPs conducted Principal Components Analyses (PCA) across broad clades. This analysis did not require estimation of relationships between lineages, and instead allowed for assessment of whether molecular signatures in the SNP data could separate clusters of lineages. This form of investigative analysis is particularly useful for detecting distinctness of lineages with shallower divergences which might be otherwise swamped by strong molecular signals of highly divergent lineages in full dataset analyses. I assigned individuals to candidate lineages based on ≥8% Tamura-Nei

(TN) genetic divergence in mtDNA (ND2) which is about the minimum level of divergence between species currently recognised in other gekkotan taxa (Moritz et al.

2015; Oliver et al. 2009). Within the Gehyra occidentalis complex no representatives of

169

Chapter 4: Continental islands support ancient genetic diversity

the lineage multi-North could be included due to lower coverage in the exon-capture dataset reducing ability to call sufficient SNPs. To ensure consistency of results I ran multiple PCAs for each subset of individuals using three different datasets of randomly selected SNPs (1179 for G. occidentalis complex and 1153 for G. xenopus) produced by

EAPHY (Blom 2015). These PCAs I performed in R v3.0.0 (R Development Core

Team, 2011) using the packages “ape” (Paradis et al. 2004) and “adegenet” (Jombart

2008; Jombart & Ahmed 2011).

To further assess tree topology and the present-day independence of divergent mtDNA lineages I also conducted a Bayesian coalescent analysis of SNP datasets for both target groups in the program SNAPP (Bryant et al. 2012). This analysis would either confirm the a priori mtDNA lineages as distinct or collapse them into broader populations. Datasets comprised 1160 and 1135 SNPs for the G. occidentalis complex and G. xenopus, respectively. During preliminary SNAPP analyses I experimented with altering mutation rates, divergence rates, and ancestral population size; however variation across analyses only occurred in the branch lengths and not in relationships or support for lineages which was my main interest here. Final analyses presented involved estimated mutation rates sampled throughout the course of the MCMC chain, a uniform

(0–100000) Yule-prior (λ), and a gamma (2, 1000) prior for ancestral population size

(θ). I ran analyses for two million generations, with sampling every 1000 generations, and after I removed 20% burn-in I assessed convergence using TRACER (Rambaut et al. 2014). Resulting trees I then visualised in DENSITREE v2.2.1 (Bouckaert & Heled). I considered the results from SNAPP in conjunction with Maximum-Likelihood phylogenies of the concatenated nuDNA alignments to assist in assessing nuDNA support for divergent mtDNA lineages.

170

Chapter 4: Continental islands support ancient genetic diversity

4.2.6. Morphometric comparisons

To assess if phenotypic variation was occurring in association with genetic variation within clades I had a suite of morphometric variables (e.g. size measures; see Table

A4.4. for details) measured following standard methods (e.g. Oliver & Doughty 2016) for genotyped individuals where possible (see Tables A4.1.a, A4.5.). In total, measurements were taken for 133 individuals (67 male, 65 female, one unknown) of the

Gehyra occidentalis complex, and 73 individuals (33 male, 39 female, one unknown) from G. xenopus. Analyses did not include measures for juvenile specimens due to potential confounding of non-allometric growth and also comparisons of body-size between groups (Hutchinson et al. 2014).

In assessing for phenotypic variability I chose to consider ecologically relevant metrics such as overall body-size, and measurements for the head-size and -shape.

Maximum body-size was chosen as considerable size variation had already been noted to occur within the species groups concerned (Doughty et al. 2012b). In addition, body- size is one of the most well-known metrics to undergo sometimes even extreme shifts when taxa move from mainland to island habitats (e.g. Lomolino et al. 2005; Thomas et al. 2009; Van Valen 1973). Variation in head-shape and -size is often related to changes ecology, including adaptation to occupy novel niches or take advantage of new diet options which could become available with a shift into an insular environment (e.g.

Herrel et al. 2001; Runemark et al. 2015; Sagonas et al. 2014).

I analysed each focal clade (G. occidentalis complex, and G. xenopus) separately to assess for morphological variation. I also conducted tests for sexual dimorphism using a multivariate analysis of covariance (MANCOVA) in SPSS v23.0 (IBM_Corp. Released

2015) to ensure there was no significant difference in shape traits between males and

171

Chapter 4: Continental islands support ancient genetic diversity

females of the same size. I analysed both raw datasets and with data with measures normalised for size variation using log-shape ratio transformation (Mosimann 1970). To do this, I calculated the geometric mean of all measures for each individual, and then divided each measurement by the geometric mean to get the shape ratios. I then used the logs of the shape ratios for PCAs which I performed in R (R Development Core Team,

2011). In addition, I used box-and-whisker plots plotted in R to compare the discrete counts of pre-cloacal pores in males, and to compare body-size variation across lineages.

4.3. Results

4.3.1. Mitochondrial diversity

Bayesian and Maximum-Likelihood analyses of mtDNA data resolved the same topologies and for each target clade revealed high, geographically structured diversity.

The Gehyra occidentalis complex consists of two currently described taxa—G. multiporosa (central-west coast) and G. occidentalis (south, King Leopold Range).

Mitochondrial data strongly supports (Maximum-Likelihood boostraps ≥75, Bayesian posterior probability ≥0.90) the G. occidentalis complex as monophyletic. However one lineage of G. occidentalis shares close mtDNA affinities with G. multiporosa, which precludes reciprocal monophyly of both recognised taxa (Fig. 4.3.a). In total, the G. occidentalis complex comprises five strongly supported (≥75 BS, ≥0.90 PP) lineages of

7.7–10.3% Tamura-Nei (TN) divergence (see Table 4.1.). Of these five lineages, G. multiporosa comprises two: multi-North (Mitchell Plateau), multi-South (Prince

Regent); whilst G. occidentalis contains three: occi-West (Yampi Islands), and occi-

172

Chapter 4: Continental islands support ancient genetic diversity

East (King Leopold Range), and occi-South (Devonian Reef System [DRS] limestones) which groups with the G. multiporosa lineages (Fig. 4.3.a).

Gehyra xenopus contains 13 mtDNA lineages at 7.5–15.1% TN divergence (Table

4.1.), all of which are strongly supported (≥75 BS, ≥0.90 PP). The complex is supported as monophyletic, and comprises three major clades (North, Mitchell Plateau, and

South+Islands), though exact relationships between these three clades are not resolved

(Fig. 4.3.b). The North clade comprises two mtDNA lineages: North-1, and North-2

(King Edward River); Mitchell Plateau clade contains three lineages: MP1 (Donkins

Hill), MP2 (Capstan Is.), and MP3 (Mitchell Plateau); and the final clade,

South+Islands, covers eight lineages, three of which are highly divergent and endemic to single islands (those not labelled ‘South-’: South-1 (Storr Is.), South-2 (Darcy/Byam

Martin Is.), South-3 (Unnamed Is.), South-4 (Prince Regent), South-5 (King Leopold

Range), Uwins-Is., Boongaree-Is., and Bigge-Is. (see Fig. 4.3.b).

4.3.2. Timeframes of diversification

Age estimates for the crown age of the Gehyra occidentalis complex are difficult to establish from the analysis conducted here. There is a lack of topological support for between-lineage relationships within the broader G. nana species complex of which the

G. occidentalis complex is a part. This is likely due to limited phylogenetic signal within the small number of slowly-evolving nuclear loci used for this analysis. The

BEAST tree failed to even recover G. multiporosa and G. occidentalis as sister-taxa.

Given this result, the best estimate for this complex is given as being less than the age for the first strongly supported ancestral node to both taxa, which happens to be the crown-estimate for the G. nana species complex. Therefore, the G. occidentalis

173

Chapter 4: Continental islands support ancient genetic diversity

complex has to be younger than 8 Ma (mean estimate, highest posterior density 5–10;

Table 4.2.), and is most likely much younger than this estimate. In comparison, the crown-age for G. xenopus is estimated at 9 Ma (HPD 5–13; Table 4.2.), and within clade diversification is estimated for late Miocene/Plio-Pleistocene ages.

4.3.3. Nuclear recovery and relationships of mitochondrial lineages

Exon-captures recovered 1204 loci for the Gehyra occidentalis complex, and 1183 for G. xenopus when 100% individuals/locus stringency settings were implemented for the completeness of datasets. SNP datasets were slightly reduced due to inclusion of only a single biallelic SNP from any non-linked exons, which resulted in datasets of

1180 and 1135 SNPs for the G. occidentalis complex and G. xenopus respectively.

Maximum-Likelihood analysis of the concatenated exon dataset for the G. occidentalis complex recovered two highly supported (≥75 BS) nuDNA clades (Fig.

4.4.a), however clade composition conflicts with the mtDNA phylogeny and instead reflects the current taxonomy (i.e. G. multiporosa versus G.occidentalis). The first clade comprises two highly supported nuDNA lineages—multi-North (Mitchell Plateau) and multi-South (Prince Regent); whilst the second clade consists of three strongly supported lineages—occi-West (Yampi Islands) diverging first before occi-East (King

Leopold Range), and occi-South (DRS limestones; Fig. 4.4.a).

Two highly supported (≥75 BS) nuDNA clades were recovered in the Maximum-

Likelihood phylogenetic analyses for G. xenopus: North+Mitchell Plateau versus

South+Islands (see Fig. 4.4.b). Within the first of these clades, the North and Mitchell

Plateau nuDNA sub-clades were each highly supported (≥75 BS), with limited resolution of further distinct lineages. The second major clade (South+Islands) resolved

174

Chapter 4: Continental islands support ancient genetic diversity

some major nuDNA lineages: Bigge-Island, South-2 (Darcy/Byam Martin Is.), majority of Uwins-Is., and South-5 (King Leopold Rng). In addition, there was a sub-clade grouping the South-3 (Unnamed Is.) and South-1 (Storr Is.) lineages; and a sub-clade grouping remaining samples of South-4 (Prince Regent), Boongaree-Is., and a single

Uwins-Is. (Fig. 4.4.b).

4.3.4. Further nuclear support for distinctness of lineages

PCAs of SNP datasets for the G. occidentalis complex indicate a number of potential

“hybrid” individuals: one occi-South, one occi-West located on the mainland, and one occi-East located on a Yampi island. Once these samples were removed, four discrete clusters (occi -East, -West, and -South; and G. multiporosa) were recovered.

SNP datasets for clades of G. xenopus showed resolution of discrete clusters for both mtDNA lineages in the North clade (Fig. A4.2.c), and some resolution (two of three lineages) within the Mitchell Plateau clade (Fig. A4.2.d). Assessment of the

South+Islands clade showed similar results to the concatenated nuDNA phylogeny, recovering as distinct the most divergent lineages, and compared to clustering of other clades (see Fig. A4.2e–f). In addition, it appeared that a single sample from Uwins Is. may be a hybrid with affinities with the geographically proximate South-4 (Prince

Regent) lineage.

SNAPP analyses consistently converged on a single topology for each focal group.

Analysis of the G. occidentalis complex indicated the earliest divergence was G. multiporosa, with lower resolution for the relationships between the three lineages of G. occidentalis (East, West and South; Fig. 4.4.c).

175

Chapter 4: Continental islands support ancient genetic diversity

The G. xenopus SNAPP analysis resolved three supported (≥0.90 PP) main clades, composition of which deviated from the concatenated nuDNA phylogeny—North,

Mitchell Plateau+Bigge Is./, and South+South Islands (Fig. 4.4.d). Within clade resolution of lineages was again greatest in the North, followed by the Mitchell Plateau which grouped with the genetically divergent but geographically proximate Bigge Is. lineage. Resolution of lineage structure and topology was lowest in the South and southern islands clade, though there was still some recovery of strongly supported

(≥0.90 PP) lineages or sub-clades (Fig. 4.4.d).

4.3.5. Morphometrics

Multivariate analyses of covariance (MANCOVA) revealed no indication of significant sexual dimorphism in either target clade (Gehyra occidentalis complex – F =

1.15, p = 0.338; G. xenopus – F = 1.56, p = 0.341). Therefore males and females were grouped for further morphometric analyses.

Within the G. occidentalis complex variation between taxa was evident in body-size

(snout-vent-length), with lineages of G. multiporosa generally smaller than G. occidentalis; and the occi-South lineage from the DRS limestones of intermediate size

(Fig. A4.1.a). There was also notable distinction across lineages in pre-cloacal pore count with G. occidentalis tending to have fewer pores than G. multiporosa (named for this particular feature), and the occi-South lineage having even fewer (Fig. A4.1.b).

Limited dispersion of lineages was observed in PCAs of the log-shape transformed measurements. Overall variance was highest within the occi-West lineage, with strong signal from a number of samples with highly variable head-length measures (Fig. 4.5.a).

The genetic lineages shared broadly overlapping shape-space in the PCAs suggesting

176

Chapter 4: Continental islands support ancient genetic diversity

particularly high morphological conservatism in the features measured. Only the first and second principal components are shown herein as additional components failed to distinguish any further dispersion of lineages from one another (Fig. 4.5.a–b).

The G. xenopus clade also showed variation in body-size between lineages with individuals from Uwins Island, and some of the South Island lineages being slightly larger overall (Fig. A4.1.c). However, it is important to recognise that these results, especially for the island lineages, are based upon very limited sample sizes (e.g. three individuals). Comparison of pre-cloacal pore counts in males showed fairly low counts across Southern and island populations, slightly higher counts for Mitchell Plateau, and a trend towards higher counts in the North clade (Fig. A4.1.d). PCA of log-shape transformed measures showed fairly limited dispersion of lineages or broader clades, except for the Bigge Island population which showed shorter snout- and head-lengths

(Fig. 4.5.c–d). Comparisons across clades for individual measures suggested slight trends for variation in island lineages: Bigge Is. – trends to shorter head- and snout- length, Boongaree Is. – trends to larger head -length, and -width, and Uwins Is. – trend to wider heads. Higher variation was noted for average trunk-length across clades, with northern populations (North, Mitchell Plateau, and Bigge Is.) trending towards slightly more elongate trunks compared to shorter trunks in southern populations (South,

Boongaree Is., and Uwins Is.).

4.4. Discussion

In this study I tested whether land-bridge island populations within two broadly sympatric congeneric gecko clades showed evidence of genetic and/or morphological differentiation from their mainland counterparts. I found limited evidence of

177

Chapter 4: Continental islands support ancient genetic diversity

morphological divergence despite high genetic diversity and long histories of isolation for multiple island lineages. There were higher levels of deeper and older genetic diversity in the geographically restricted and more ecologically specialised Gehyra xenopus clade, and this diversity was most highly concentrated over short ranges in the central- and north-west Kimberley and multiple islands. In comparison, genetic diversity was younger and shallower within the more widespread G. occidentalis complex, which showed more structuring of diversity in the southern limits of its distribution (three southern G. occidentalis lineages versus two northern G. multiporosa lineages).

In contrast to the high genetic diversity in G. xenopus, there was limited evidence of morphological diversity across lineages, with a slight trend towards variation (body- size, head -length and -width, and trunk-length) in island lineages. However, increased sampling would be needed to better address these patterns. Though previously insular

“dwarf” and “giant” populations had been noted in G. occidentalis (Doughty et al.

2012b), variability within this lineage was so high across its distribution that these populations did not appear particularly anomalous in body-size. Overall, the G. occidentalis complex seemed to show lower morphological divergence than G. xenopus, with no evidence to indicate island-related variability in the measures considered here.

4.4.1. Old endemic lineages in land-bridge islands

The biota of continental islands are generally assumed to be closely related subsets of mainland populations – with either recent or ongoing gene-flow – due to proximity and relatively recent isolation (Holocene, post- Last Glacial Maximum; e.g.

Velo-Antón et al. 2012). However, results from my exon-capture work indicate that the

178

Chapter 4: Continental islands support ancient genetic diversity

island lineages in Gehyra xenopus, and even the predominantly island-restricted lineage of G. occidentalis (Occi-West) have been isolated with limited or no gene-flow from the mainland, likely for millions of years; far pre-dating the most recent isolation of the

Kimberley islands (~8 kya; De Deckker & Yokoyama 2009; Nix & Kalma 1972). These old genetic divergences imply that many of the west Kimberley islands may maintain continuous isolation for particular taxa even when they are connected to the mainland during periods of low sea-levels (e.g. Johnson et al. 2010; Köhler 2011b; Potter et al.

2012). This pattern of long-term persistence of micro-endemic lineages is not universal, however, with studies on other taxa showing limited genetic structuring across these islands and the mainland (e.g. see Chapter 3, although note the Augustus Is. ESU;

Harradine et al. 2015a; Harradine et al. 2015b).

The contrasting signatures of micro-endemism in G. xenopus and higher genetic- connectivity in G. multiporosa across the islands allude to the important role of ecology in shaping insular genetic divergence. Populations of ecologically-specialised (e.g. saxicolous [rock-inhabiting]) or low-dispersing taxa such as G. xenopus are more likely to be isolated by habitat barriers between islands and the mainland than populations of generalist species such as G. multiporosa. This evidence of isolation is consistent with findings in Carlia skinks across the Kimberley islands, which exhibit genetically divergent island lineages in the more specialised lineages compared to low structure across islands in generalist congeners (Silva et al. in review). One hypothesis to explain this isolation is that the intervening landscape between many of the Kimberley islands and mainland regions does not contain the large, open rock-types required by G. xenopus, thus maintaining separation between island and mainland lineages even during low sea-levels.

179

Chapter 4: Continental islands support ancient genetic diversity

Although these island habitats may be maintaining isolation of genetic lineages over long timeframes, the limited evidence for morphological divergence in both Gehyra clades suggests a potential lack of strong ecological-drivers (at least in regards to morphological measures considered here) acting upon lineages restricted to these insular environments. This is an expected result, given that the Kimberley islands are elements of a submerged shoreline and are topographically and climatically similar to adjacent

(<14 km) mainland regions (Palmer et al. 2013). This environmental similarity between proximate island and mainland regions allows these land-bridge islands to potentially act as ecological refugia for mainland taxa. In addition, even if island environments are not driving selection for ecological divergence they are facilitating diversification in some taxa via isolation and maintenance of deeply divergent genetic lineages in spite of a history of geographical connectivity; thereby functioning as important evolutionary refugia.

4.4.2. Contrasting diversification responses to environmental change

I found contrasting patterns of diversification between the two Gehyra clades, which could be indicative of varying responses to environmental change dependent on the ecology of the taxa involved. Evidence suggests historic continental aridification and glacial cycling caused broad-scale periodic drying in the AMT with consequent shifts in distribution of more mesic woodlands to be replaced by arid grassland habitat over a north-south cline (Denniston et al. 2013; Proske et al. 2014; Reeves et al. 2013;

Woinarski et al. 2007). This likely had different implications for the two Gehyra clades.

G. xenopus exhibits the highest levels of deep genetic structuring in the most mesic northern region of its distribution. Further south, genetic lineages of G. xenopus are

180

Chapter 4: Continental islands support ancient genetic diversity

shallower and more widespread, with increased genetic connectivity (including introgression), or a more recent common ancestry for southern mainland and island lineages. As arid conditions encroached northwards into the distribution of this mesic- reliant taxon, southern populations were likely extirpated, remaining only in the northern most mesic and climatically buffered micro-refugia afforded by heterogeneous topography. Within these refugia, lineages could then persist and diversify. The distribution of genetic structure within G. xenopus, therefore, is consistent with what we would expect in a particularly vulnerable, specialised clade responding spatially to climatic change. This pattern mirrors those observed in other ancient mesic relicts which are also now restricted to isolated pockets of particularly high rainfall or unique geological regions within the Kimberley (e.g. see Chapter 3; Oliver et al. 2014a; Oliver et al. 2014b). Diversity in the south likely results, then, from younger lineages recolonising aridified areas as environmental conditions became wetter and more favourable for the species. This hypothesized scenario is supported by the estimated crown-age for G. xenopus: the mid- to late-Miocene (9 Ma) which is concurrent with the onset of significant aridification within central Australia ~10 Ma (Byrne et al. 2008;

Fujioka et al. 2009; Pepper & Keogh 2014).

Under these changing climatic conditions, the comparatively widespread and more arid-adapted taxa of the G. occidentalis complex would be less vulnerable than more specialised taxa such as G. xenopus. The ability to use varied microhabitats could likely facilitate greater dispersal over wide ranges and thus potentially increase connectivity between insular and mainland habitats at periods of low sea-level. Consistent with this theory, the G. occidentalis complex exhibits limited genetic structure within the most mesic northern regions of its distribution, including an absence of population structure

181

Chapter 4: Continental islands support ancient genetic diversity

between the north or central islands and mainland. The highest structuring of genetic diversity within this complex occurs instead across the southern range of G. occidentalis, with signatures of isolation and persistence in a collection of south-west islands, and additionally in a newly recognised southern refugium – the Devonian Reef

System (DRS) limestone ranges (see Chapter 3; Oliver et al. in press; Oliver et al.

2014a). The broadly distributed G. occidentalis lineage (Occi-East) across the King

Leopold Range system also potentially represents long-term widespread persistence (see also Chapter 3) on an even larger geographic scale. Furthermore, much like G. xenopus lineages in the south, lineages of G. occidentalis in its southern limits also show some signature of potential gene-flow or introgression, which could be viewed as further evidence of expanding population ranges and break-down of lineage boundaries in a climatically challenging environment at the margins of lineage distributions (see

Chapter 3). These patterns of genetic structure within the G. occidentalis complex are consistent with predictions that a more arid-tolerant taxon would only begin to suffer notable impacts of aridity within the extent of its range which is exposed to the harshest environmental variation, i.e. insular environments only become isolating environments for this taxon at the arid southern edge of the Kimberley.

Finally, though the crown-age for the relict G. xenopus clade is concurrent with the initial onset of Miocene aridification dynamics, the diversification within the G. occidentalis complex is most likely younger (~3 Ma). This age estimate is more concordant with drying associated with Plio-Pleistocene glacial cycling (Bowman et al.

2010; Byrne et al. 2008; Pepper & Keogh 2014), which also fits well with signatures of greater connectivity between island and mainland populations of G. occidentalis

182

Chapter 4: Continental islands support ancient genetic diversity

complex lineages which would have experienced periods of cyclic connectivity of the land-bridge islands throughout this period of varying sea-levels (Hewitt 2000).

4.4.3. Continental islands as evolutionary refugia

The mesic west coast of the Kimberley, and particularly the Mitchell Plateau, has been long recognized as a hotspot for biodiversity and endemism (Bowman et al. 2010;

Crisp et al. 2001; Doughty 2011; Powney et al. 2010; Slatyer et al. 2007). In addition, extremely high patterns of cryptic diversity and micro-endemism have been noted in snails across many of the Kimberley land-bridge islands (e.g. Gibson & Köhler 2012;

Johnson et al. 2010; Köhler 2010, 2011a, b). However, until now there has been limited evidence for similar patterns in vertebrates across these same islands (though see

Chapter 3; Ellis 2016; Oliver et al. 2012; Palmer et al. 2013; Potter et al. 2012; Silva et al. in review). My findings emphasise that the Kimberley islands are also a significant region for old micro-endemic vertebrate diversity (particularly in reptiles).

Many of the divergent lineages discovered are endemic to the larger Kimberley islands such as Augustus and Bigge islands (~17–19 thousand hectares [kHa]), and to a lesser degree Boongaree, Jungulu and Uwins islands (see Fig. 4.6.) (~3–5 kHa; see herein, also Chapter 3; Ellis 2016; Silva et al. in review). Island biogeography theory predicts larger islands will support more species due to higher carrying capacity and the lower risk of extinction for large populations (Brown & Lomolino 1998; MacArthur &

Wilson 1963, 1967). Many of these islands also lie within the highest rainfall zone of the Kimberley region. Molecular evidence suggest that these lineages, particularly in the north (e.g. Bigge Island herein; see also the Montalivet islands in Silva et al. in review), have long remained isolated from their mainland relatives throughout historic island-

183

Chapter 4: Continental islands support ancient genetic diversity

mainland connectivity. These northern islands may therefore represent important evolutionary refugia as the only remaining environments capable of supporting such micro-endemic relicts.

It is worth noting that some of the large islands towards the central-south Kimberley

(e.g. Boongaree and Uwins islands) exhibit evidence of genetically divergent lineages, but also show some levels of genetic connectedness and possible introgression events with geographically proximate mainland populations (e.g. Prince Regent River region; see Fig. 4.6.). Coincidentally, these islands are located particularly close to strong river outflows (i.e. the Prince Regent River) from the mainland. Though higher genetic connectivity within this region could be explained by more recent ancestry of southern island and mainland lineages, alternatively there could be cases of intermittent gene- flow resulting from mainland individuals being transferred to these islands via river systems (e.g. Harradine et al. 2015b).

Finally, there is growing evidence for community turn-over along the Kimberley islands over a north-south gradient, which may be associated with the climatic gradient across this region. Islands of the southern Kimberley off the Yampi Peninsula tend to support different assemblages to islands in the north, with both groups often showing closer relationship affinities to their geographically proximate mainland counterparts than to alternate island populations (see also Chapter 3; Doughty et al. 2012a; Gibson

2014; Gibson & Köhler 2012; Lyons et al. 2014; Palmer et al. 2013). The genetic diversity of G. occidentalis across the southern Kimberley (Yampi) islands suggests there is a higher level of population and geographic connectedness among the land- bridge islands in this region when compared to the single-island endemic lineages of G. xenopus observed further north. Despite this, it is clear that the south islands represent

184

Chapter 4: Continental islands support ancient genetic diversity

an important refuge area for more arid-adapted lineages (e.g. Gibson & Köhler 2012;

Palmer et al. 2013), particularly in the case of the predominantly island-restricted G. occidentalis lineage (Occi-West). It is possible that at times of extreme drying of the southern Kimberley these south-western islands represented a refuge which populations typically widespread across the south Kimberley and King Leopold Range region contracted to, and potentially even a region where in situ lineage diversification occurred. The variation in genetic structuring over this latitudinal and climatic gradient provides further evidence that the tropical savannahs of Kimberley represent a highly heterogeneous landscape, and that topographic and climatic factors interact to shape contrasting diversification patterns throughout this region.

4.4.4. Conclusion

Though the Kimberley has remained one of the most undisturbed regions of Australia

(Woinarski et al. 2010), it is now under increasing threat from agriculture, mining, invasive species and changing fire regimes (Gibson & McKenzie 2012; Legge et al.

2011). The isolation of the Kimberley islands offers protection from many of these mainland threats – particularly invasive species – hence these land-bridge islands represent an area of high conservation significance. In addition to functioning as ecological arks for biodiversity, especially endangered mammals, my findings of ancient genetically divergent micro-endemic lineages indicates these islands also represent significant evolutionary refugia for vertebrate taxa. It would appear that the interaction of historic climatic variation and topography has impacted taxa differently in the south versus the north Kimberley depending on the ecology of the taxa in question, their sensitivity to environmental perturbation and dispersal capabilities. More broadly

185

Chapter 4: Continental islands support ancient genetic diversity

it cannot always be assumed that morphologically divergent taxa in insular environments or on land-bridge islands are actually genetically distinct evolutionary units. Likewise, my findings indicate that morphologically similar taxa in these insular environments may be genetically divergent; especially in regions with highly variable landscape and strong climatic gradients such as the tropical savannahs of northern

Australia.

186

Chapter 4: Continental islands support ancient genetic diversity

Table 4.1. Genetic diversity data for major lineages within the clades of the Gehyra occidentalis complex and Gehyra xenopus.

n mtDNA Clade n (species) Avg Min Max 8% d Gehyra occidentalis complex 291 2 0.085 0.000 0.160 5 Gehyra xenopus 104 ? 0.140 0.000 0.223 13 Gehyra occidentalis complex 1) Multi - North (Mitchell Plateau) 24 0.014 0.000 0.036 0 0.103 2) Multi - South (Prince Regent) 134 0.015 0.000 0.071 0 0.103 3) Occi - East (King Leopold Range) 33 0.008 0.000 0.018 0 0.077 4) Occi - West (Yampi Islands) 61 0.028 0.000 0.051 0 0.078 5) Occi - South (Limestones) 39 0.041 0.000 0.073 0 0.077 Gehyra xenopus 1) North 1 11 0.016 0.000 0.031 0 0.075 2) North 2 - King Edward River 15 0.001 0.000 0.004 0 0.075 3) MP1 - Donkins Hill 1 n.a. n.a. n.a. 0 0.119 4) MP2 - Capstan Island 2 0.015 0.015 0.015 0 0.091 5) MP3 - Mitchell Plateau 20 0.011 0.000 0.026 0 0.091 6) Bigge Island 3 0.000 0.000 0.000 0 0.151 7) Boongaree Island 3 0.002 0.000 0.003 0 0.101 8) Uwins Island 4 0.000 0.000 0.000 0 0.110 9) South 1 - Storr Island 1 n.a. n.a. n.a. 0 0.088 10) South 2 - Darcy/Byam Martin Islands 7 0.003 0.000 0.005 0 0.086 11) South 3 - Unnamed Island 3 0.000 0.000 0.000 0 0.075 12) South 4 - Prince Regent 12 0.019 0.000 0.038 0 0.075 13) South 5 - King Leopold Range 22 0.009 0.000 0.017 0 0.075 n, number of samples; avg, average pairwise Tamura-Nei (TN) divergence between individuals; min, minimum TN divergence between individuals; max, maximum TN divergence between individuals; 8%, number of lineages showing mean divergence over 8% within grouping; d, average TN divergence to nearest relative; n.a., value not calculable from single sample.

187

Chapter 4: Continental islands support ancient genetic diversity

Table 4.2. Prior and posterior probability distributions for the crown-ages of the two west Kimberley clades – Gehyra occidentalis complex and G. xenopus – based on a four loci nuclear (BDNF, Cmos, PDC, and RAG1) dataset, with secondary calibrations.

Crown-age estimates Priors Root (Normal) 30 (SD 4) Aus vs. Pacific Gehyra 22 (SD 3) Posteriors (crown ages) Aus vs. Pacific Gehyra 23 (18–28) Gehyra occidentalis complex <8 (3–10) Gehyra xenopus 9 (5–13) G. xenopus North vs. Mitchell Plateau 4 (3–10) G. xenopus South vs. Islands 6 (3–10)

188

Chapter 4: Continental islands support ancient genetic diversity

Figure 4.1. a) Distribution map for the two focal clades – Gehyra occidentalis complex (including G. multiporosa and G. occidentalis), and G. xenopus. Key regions mentioned throughout the text are indicated; symbols denote different species as labelled in legend. b) Map of rainfall isohyets across the Kimberley as adapted from Palmer et al., (2013).

189

Chapter 4: Continental islands support ancient genetic diversity

Figure 4.2. Maps of genetic sampling for a) the Gehyra occidentalis complex, and b) Gehyra xenopus. Symbol colours denote mitochondrial lineages at ≥8% Tamura-Nei pairwise divergence, and colours and labels (listed in legends) match those for phylogenetic trees henceforth. Circles represent samples sequenced for mitochondrial data only, whilst diamonds indicate samples included in exon-captures (nuclear DNA).

190

Chapter 4: Continental islands support ancient genetic diversity

Figure 4.3. Mitochondrial phylogenies for a) the Gehyra occidentalis complex, and b) Gehyra xenopus. ≥8% Tamura-Nei divergent lineages are labelled and coloured to match the sampling in Fig. 4.2. Bayesian posterior-probabilities are labelled for supported nodes.

191

Chapter 4: Continental islands support ancient genetic diversity

Occi -- South

Occi -- West 1

Occi -- East Occi -- South

0.64 Occi -- East

a) c) Occi -- West

1 0.99 1 1

1 1 0.66 0.99

0.48 1 b) d)

Figure 4.4. Nuclear phylogenies for the Gehyra occidentalis complex a) and c), and G. xenopus b) and d). Maximum-Likelihood phylogenies of full concatenated exon datasets are shown in a) and b), and SNAPP trees of SNP datasets in c) and d). ≥8% Tamura-Nei divergent lineages are labelled and coloured to match lineages in Fig. 4.3. Bootstrap values are labelled for supported nodes in a) and b), and Bayesian posterior- probabilities in c) and d).

192

Chapter 4: Continental islands support ancient genetic diversity

a) b)

c) d)

Figure 4.5. Principal Components Analyses of log-shape transformed body measurements for a–b) Gehyra occidentalis complex, and c–d) G. xenopus. a and c) display the first and second principal components (PCs), whilst b and d) display the second and third PCs. Shape space for each lineage is indicated by coloured convex hulls, the colours of which correspond to the labelled lineages in the legends to the left. Labelled arrows on each axis indicated the variables with the greatest loading for each PC, and the arrow direction indicates the direction of increasing size for that variable.

193

Chapter 4: Continental islands support ancient genetic diversity

Figure 4.6. Maps of key Kimberley continental islands discussed in text; figure adapted from Lyons et al. (2014).

194

Chapter 4: Continental islands support ancient genetic diversity

References

(R Development Core Team, 2011) R: A language and environment for statistical

computing. R Foundation for Statistical Computing. Available at: http://www.R-

project.org, Vienna, Austria.

Anderson BM, Barrett MD, Krauss SL, Thiele K (2016) Untangling a species complex

of arid zone grasses (Triodia) reveals patterns congruent with co-occurring

animals. Molecular Phylogenetics and Evolution 101, 142-162.

Bell RC, Brasileiro CA, Haddad CFB, Zamudio KR (2012) Evolutionary history of

Scinax treefrogs on land-bridge islands in south-eastern Brazil. Journal of

Biogeography 39, 1733-1742.

Bi K, Vanderpool D, Singhal S, et al. (2012) Transcriptome-based exon capture enables

highly cost-effective comparative genomic data collection at moderate

evolutionary scales. BMC Genomics 13, 1.

Bittencourt-Silva, Gabriela B, Silva HR (2014) Effects of fragmentation and sea-level

changes upon frog communities of land-bridge islands off the southeastern coast

of Brazil. PLoS ONE 9, e103522.

Bittkau C, Comes HP (2005) Evolutionary processes in a continental island system:

molecular phylogeography of the Aegean Nigella arvensis alliance

(Ranunculaceae) inferred from chlorplast DNA. Molecular Ecology 14, 4065-

4083.

Blom MPK (2015) EAPhy: a flexible tool for high-throughput quality filtering of exon-

alignments and data processing for phylogenetic methods. PLoS currents 7.

Bouckaert R, Heled J DensiTree 2: Seeing Trees Through the Forest, bioRxiv,

http://dx.doi.org/10.1101/012401.

195

Chapter 4: Continental islands support ancient genetic diversity

Bowman DMJS, Brown GK, Braby MF, et al. (2010) Biogeography of the Australian

monsoon tropics. Journal of Biogeography 37, 201-216.

Bragg JG, Potter S, Bi K, Moritz C (2015) Exon capture phylogenomics: efficacy across

scales of divergence. Molecular Ecology Resources.

Brown JH, Lomolino MV (1998) Island Biogeogrpahy: Patterns in Species Richness.

In: Biogeography, pp. 369-406. Sinauer Associates, Inc. , Sunderland, MA.

Bryant D, Bouckaert R, Felsenstein J, Rosenberg NA, RoyChoudhury A (2012)

Inferring species trees directly from biallelic genetic markers: bypassing gene

trees in a full coalescent analysis. Molecular Biology and Evolution 29, 1917-

1932.

Burbidge AA, McKenzie NL (1978) The islands of the north-west Kimberley: Western

Australia. In: Wildlife Research Bulletin Western Australia No. 7. Department of

Fisheries and Wildlife, Perth.

Byrne M (2008) Evidence for multiple refugia at different time scales during

Pleistocene climatic oscillations in southern Australia inferred from

phylogeography. Quaternary Science Reviews 27, 2576-2585.

Byrne M, Yeates DK, Joseph L, et al. (2008) Birth of a biome: insights into the

assembly and maintenance of the Australian arid zone biota. Molecular Ecology

17, 4398-4417.

Cameron RAD (1992) Land snail faunas of the Napier and Oscar Ranges, Western

Australia; diversity, distribution and speciation. Biological Journal of the

Linnean Society 45, 271-286.

196

Chapter 4: Continental islands support ancient genetic diversity

Carnaval AC, Waltari E, Rodrigues MT, et al. (2014) Prediction of phylogeographic

endemism in an environmentally complex biome. Proceedings of the Royal

Society B: Biological Sciences 281, 20141461.

Couper PJ, Hoskin CJ (2008) Litho-refugia: the importance of rock landscapes for the

long-term persistence of Australian rainforest fanua. Austalian Zoologist 34,

554-560.

Criscione F, Köhler F (2015) On the land snail Damochlora Iredale, 1938 and its

cryptic sibling Nannochlora n. gen. (Stylommatophora: Camaenidae), each

endemic to an island in the Western Australian Kimberley. Molluscan Research

35, 275-286.

Criscione F, Law ML, Köhler F (2012) Land snail diversity in the monsoon tropics of

Northern Australia: revision of the genus Exiligada Iredale, 1939 (Mollusca:

Pulmonata: Camaenidae), with description of 13 new species. Zoological

Journal of the Linnean Society 166, 689-722.

Crisp MD, Laffan S, Linder HP, Monro A (2001) Endemism in the Australian Flora.

Journal of Biogeography 28, 183-198. da Silva JMC, Bates JM (2002) Biogeographic patterns and conservation in the South

American Cerrado: a tropical savanna hotspot. BioScience 52, 225-234.

De Deckker P, Yokoyama Y (2009) Micro-palaeontological evidence for Late

Quaternary sea-level changes in Bonaparte Gulf, Australia. Global and

Planetary Change 66, 85-92.

Denniston RF, Wyrwoll K-H, Asmerom Y, et al. (2013) North Atlantic forcing of

millenial-scale Indo-Australian monsoon dynamics during the Last Glacial

period. Quaternary Science Reviews 72, 159-168.

197

Chapter 4: Continental islands support ancient genetic diversity

Domingos FMCB, Bosque RJ, Cassimiro J, et al. (2014) Out of the deep: cryptic

speciation in a Neotropical gecko (Squamata, Phyllodactylidae) revealed by

species delimitation methods. Molecular Phylogenetics and Evolution 80.

Doughty P (2011) An emerging frog diversity hotspot in the northwest Kimberley of

Western Australia: another new frog species from the high rainfall zone.

Records of the Western Australian Museum 26, 209-216.

Doughty P, Palmer R, Cowan M, Pearson DJ (2012a) Biogeographic patterns of frogs

of the Kimberley islands, Western Australia. Records of the Western Australian

Museum Supplement 81, 109-124.

Doughty P, Palmer R, Sistrom MJ, Bauer AM, Donnellan SC (2012b) Two new species

of Gehyra (Squamata: Gekkonidae) geckos from the north-west Kimberley

region of Western Australia. Records of the Western Australian Museum 27,

117-134.

Drummond AJ, Ashton B, Cheung M, et al. (2008) Geneious v6.0, Available from

http://www.geneious.com/.

Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with

BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29, 1969-1973.

Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high

throughput. Nucleic Acids Research 32, 1792-1797.

Ellis RJ (2016) A new species of blindsnake (Scolecophidia: Typhlopidae: Anilios)

from the Kimberley region of Western Australia. Herpetologica 72, 271-278.

Emerson BC (2002) Evolution on oceanic islands: molecular phylogenetic approaches

to understanding pattern and process. Molecular Ecology 11, 951-966.

198

Chapter 4: Continental islands support ancient genetic diversity

Fujioka T, Chappell J, Fifield LK, Rhodes EJ (2009) Australian desert dune fields

initiated with Pliocene-Pleistocene global climatic shift. Geology 37, 51-54.

Gamble T, Greenbaum E, Jackman TR, Bauer AM (2015) Into the light: diurnality has

evolved multiple times in geckos. Biological Journal of the Linnean Society 115,

896-910.

Gibson LA (2014) Biogeographic patterns on Kimberley islands, Western Australia.

Records of the Western Australian Museum 81, 245-280.

Gibson LA, Köhler F (2012) Determinants of species richness and similarity of species

composition of land snail communities on Kimberley islands Records of the

Western Australian Museum Supplement 81, 40-65.

Gibson LA, McKenzie NL (2012) Identification of biodiversity assets on selected

Kimberley islands: background and implementation. Records of the Western

Australian Museum Supplement 81, 1-14.

Glaw F, Hoegg S, Vences M (2006) Discovery of a new basal relict lineage of

Madagascan frogs and its implications for mantellid evolution. Zootaxa 1334,

27-43.

Glaw F, Köhler J, Townsend TM, Vences M (2012) Rivaling the world's smallest

reptiles: discovery of miniaturized and microendemic new species of leaf

chameleons (Brookesia) from northern Madagascar. PLoS ONE 7, e31314.

Grace J, José SJ, Meir P, Miranda HS, Montes RA (2006) Productivity and carbon

fluxes of tropical savannas. Journal of Biogeography 33, 387-400.

Grazziotin FG, Monzel M, Echeverrigaray S, Bonnatto SL (2006) Phylogeography of

the Bothrops jaracara complex (Serpentes: Viperidae): past fragmentation and

island colonization in the Brazilian Atlantic Forest. Molecular Ecology 15.

199

Chapter 4: Continental islands support ancient genetic diversity

Guarnizo CE, Werneck FP, Giugliano LG, et al. (2016) Cryptic lineages and

diversification of an endemic anole lizard (Squamata, Dactyloidae) of the

Cerrado hotspot. Molecular Phylogenetics and Evolution 94, 279-289.

Harradine E, How RA, Schmitt LH, Spencer PBS (2015a) Island size and remoteness

have major conservation significance for how spatial diversity is partitioned in

skinks. Biodiversity Conservation 24.

Harradine EL, Andrew ME, Thomas JW, et al. (2015b) Importance of dispersal routes

that minimize open-ocean movement to the genetic structure of island

populations. Conservation Biology 29, 1704-1714.

Heinicke MP, Greenbaum E, Jackman TR, Bauer AM (2011) Phylogeny of a trans-

Wallacean radiation (Squamata, Gekkonidae, Gehyra) supports a single early

colonization of Australia. Zoologica Scripta 50.

Herrel A, Meyers JJ, Vanhooydonck B (2001) Correlations between habitat use and

body shape in a phrynosomatid lizard (Urosaurus ornatus): a population-level

analysis. Biological Journal of the Linnean Society 74.

Hewitt GM (2000) The genetic legacy of the Quaternary ice ages. Nature 405.

Huang X, Madan A (1999) CAP3: a DNA sequence assembly program. Genome

Research 9, 868-877.

Hutchinson MN, Sistrom MJ, Donnellan SC, Hutchinson RG (2014) Taxonomic

revision of the Australian arid zone lizards Gehyra variegata and G. montium

(Squamata, Gekkonidae) with description of three new species. Zootaxa 3184,

221-241.

IBM_Corp. (Released 2015) IBM SPSS Statistics for Windows. IBM Corp., Armonk,

NY.

200

Chapter 4: Continental islands support ancient genetic diversity

Johnson MS, O'Brien EK, Fitzpatrick JJ (2010) Deep, hierarchical divergence of

mitochondrial DNA in Amplirhagada land snails (Gastropoda: Camaenidae)

from the Bonaparte Archipelago, Western Australia. Biological Journal of the

Linnean Society 100, 141-153.

Jombart T (2008) adegenet: a R package for the multivariate analysis of genetic

markers. Bioinformatics 24, 1403-1405.

Jombart T, Ahmed I (2011) adegenet 1.3-1: new tools for the analysis of genome-wide

SNP data. Bioinformatics.

Köhler F (2010) Three new species and two new genera of land snails from the

Bonaparte Archipelago in the Kimberley, Western Australia (Pulmonata,

Camaenidae). Molluscan Research 30, 1-16.

Köhler F (2011a) Australocosmica, a new genus of land snails from the Kimberley,

Western Australia (Eupulmonata, Camaenidae). Malacologia 53, 199-216.

Köhler F (2011b) The Camaenid species of the Kimberley islands, Western Australia

(Stylommatophora: Helicoidea). Malacologia 54, 203-406.

Lanfear R, Calcott B, Frandsen P (Forthcoming) PartitionFinder 2: new methods for

selecting partitioning schemes and models of molecular evolution for large

datasets. In preparation.

Legge S, Kennedy MS, Lloyd RAY, Murphy SA, Fisher A (2011) Rapid recovery of

mammal fauna in the central Kimberley, northern Australia, following the

removal of introduced herbivores. Austral Ecology 36, 791-799.

Lomolino M, Riddle D, Brown J (2005) Biogeography, 3rd Edition edn. Sinauer

Associates, Sunderland, MA.

201

Chapter 4: Continental islands support ancient genetic diversity

Lorenzen E, Heller R, Siegismund HR (2012) Comparative phylogeography of African

savannah ungulates. Molecular Ecology 63, 534-542.

Losos JB, Ricklefs RE (2009) Adaptation and diversification on islands. Nature 457.

Lyons MN, Keighery GJ, Gibson LA, Handasyde T (2014) Flora and vegetation

communities of selected islands off the Kimberley coast of Western Australia.

Records of the Western Australian Museum Supplement 81, 205-244.

MacArthur RH, Wilson EO (1963) An equilibrium theory of insular zoogeography.

Evolution 17, 373-387.

MacArthur RH, Wilson EO (1967) The Theory of Island Biogeography. In:

Monographs in Population Biology, no. 1. Princeton University Press,

Princeton, NJ.

Macey JR, Larson A, Ananjeva NB, Fang Z, Papenfuss TJ (1997) Two novel gene

orders and the role of light-strand replication in the rearrangement of the

vertebrate mitochondrial genome. Molecular Biology and Evolution 14, 91-104.

Martin LJ, Blossey B, Ellis E (2012) Mapping where ecologists work: biases in the

global distribution of terrestrial ecological observations. Frontiers in Ecology

and the Environment 10, 195-201.

Mayle FE, Burbridge R, Killeen TJ (2000) Millennial-scale dynamics of southern

Amazonian rainf forests. Science 290, 2291-2294.

Meyer M, Kircher M (2010) Illumina sequencing library preparation for highly

multiplexed target capture and sequencing. Cold Spring Harb Protoc 2010,

t5448.

202

Chapter 4: Continental islands support ancient genetic diversity

Miller MA, Pfeiffer W, Schwartz T (2010) Creating the CIPRES Science Gateway for

inference of large phylogenetic trees. In: Proceedings of the Gateway

Computing Environments Workshop (GCE), pp. 1-8, New Orleans, LA.

Miraldo A, Li S, Borregaard MK, et al. (2016) An Anthropocene map of genetic

diversity. Science 353, 1532-1535.

Moritz C, Fujita MK, Rosauer D, et al. (2015) Multilocus phylogeography reveals

nested endemism in a gecko across the monsoonal tropics of Australia. Mol

Ecol.

Mosimann JE (1970) Size allometry: size and shape variables with characterizations of

the lognormal and generalized gamma distributions. Journal of the American

Statistical Association 65, 930-948.

Nix H, Kalma J (1972) Climate as a dominant control in the biogeography of northern

Australia and New Guinea. In: Bridge and barrier: the natural and cultural

history of Torres Strait (ed. Walker D), pp. 61-91. Research School of Pacific

Publication E6/3. Australian National University Press, Canberra.

Nogueira C, Ribeiro Sr, Costa GC, Colli GR (2011) Vicariance and endemism in a

Neotropical savana hotspot: distribution patterns of Cerrado squamate reptiles.

Journal of Biogeography 38, 1907-1922.

Oliver PM, Adams M, Lee MSY, Hutchinson MN, Doughty P (2009) Cryptic diversity

in vertebrates: molecular data double estimates of species diversity in a radiation

of Australian lizards (Diplodactylus, Gekkota). Proceedings of the Royal Society

B: Biological Sciences 276, 2001-2007.

Oliver PM, Bourke G, Pratt RC, Doughty P, Moritz C (2016) Systematics of small

Gehyra (Squamata: Gekkonidae) of the southern Kimberley, Western Australia:

203

Chapter 4: Continental islands support ancient genetic diversity

redescription of G. kimberleyi Börner & Schüttler, 1983 and description of a

new restricted range species. Zootaxa 4107, 49-64.

Oliver PM, Doughty P (2016) Systematic revision of the marbled velvet geckos

(Oedura marmorata species complex, Diplodactylidae) from the Australian arid

and semi-arid zones. Zootaxa 4088, 151-176.

Oliver PM, Doughty P, Palmer R (2012) Hidden biodiversity in rare northern Australian

vertebrates: the case of the clawless geckos (Crenadactylus, Diplodactylidae) of

the Kimberley. Wildlife Research 39, 429-435.

Oliver PM, Laver RJ, De Mello Martins F, et al. (in press) A novel hotspot of vertebrate

endemism and evolutionary refugium in tropical Australia. Diversity and

Distributions.

Oliver PM, Laver RJ, Melville J, Doughty P (2014a) A new species of Velvet Gecko

(Oedura: Diplodactylidae) from the limestone ranges of the southern Kimberley,

Western Australia. Zootaxa 3873, 49-61.

Oliver PM, Laver RJ, Smith KL, Bauer AM (2014b) Long-term persistence and

vicariance within the Australian Monsoonal Tropics: The case of the giant cave

and tree geckos (Pseudothecadactylus). Australian Journal of Zoology 61, 462-

468.

Palmer R, Pearson DJ, Cowan MA, Doughty P (2013) Islands and scales: a

biogeographic survey of reptiles on Kimberley islands, Western Australia.

Records of the Western Australian Museum Supplement 81, 183-204.

Paradis E, Claude J, Strimmer K (2004) APE: analyses of phylogenetics and evolution

in R language. Bioinformatics 20, 289-290.

204

Chapter 4: Continental islands support ancient genetic diversity

Patiño J, Carine M, Mardulyn P, et al. (2015) Approximate Bayesian Computation

reveals the crucial role of islands for the assembly of continental biodiversity.

Sytematic Biology, syv013.

Paulay G (1994) Biodiversity on oceanic islands: its origin and extinction. American

Zoologist 34, 134-144.

Peñalba JV, Smith LL, Tonione MA, et al. (2014) Sequence capture using PCR-

generated probes: a cost-effective method of targeted high-throughput

sequencing for nonmodel organisms. Molecular Ecology Resources 14, 1000-

1010.

Pepper M, Doughty P, Arculus R, Keogh JS (2008) Landforms predict phylogenetic

structure on one of the world's most ancient surfaces. BMC Evolutionary

Biology 8, 152.

Pepper M, Doughty P, Fujita MK, Moritz C, Keogh JS (2013) Speciation on the rocks:

Integrated systematics of the Heteronotia spelea species complex (Gekkota;

Reptilia) from western and central Australia. PLoS ONE 8, e78110.

Pepper M, Fujita MK, Moritz C, Keogh JS (2011a) Palaeoclimate change drove

diversification among isolated mountain refugia in the Australian arid zone.

Molecular Ecology 20, 1529-1545.

Pepper M, Ho SYW, Fujita MK, Keogh JS (2011b) The genetic legacy of aridification:

Climate cycling fostered lizard diversification in Australian montane refugia and

left low-lying deserts genetically depauperate. Molecular Phylogenetics and

Evolution 61, 750-759.

205

Chapter 4: Continental islands support ancient genetic diversity

Pepper M, Keogh JS (2014) Biogeography of the Kimberley, Western Australia: a

review of landscape evolution and biotic response in an ancient refugium.

Journal of Biogeography 41, 1443-1455.

Pianka ER (1981) Diversity and adaptive radiations of Australian desert lizards. In:

Ecological Biogeography of Australia (ed. Keast A).

Potter S, Bragg JG, Peter BM, Bi K, Moritz C (2016) Phylogenomics at the tips:

inferring lineages and their demographic history in a tropical lizard, Carlia

amax. Molecular Ecology.

Potter S, Eldridge MDB, Taggart DA, Cooper SJB (2012) Multiple biogeographical

barriers identified across the monsoon tropics of northern Australia:

phylogeographic analysis of the brachyotis group of rock-wallabies. Molecular

Ecology 21, 2254-2269.

Powney GD, Grenyer R, Orme CDL, Owens IPF, Meiri S (2010) Hot, dry and different:

Ausralian lizard richness is unlike that of mammals, amphibians and birds.

Global Ecology and Biogeography 19, 386-396.

Proske U, Heslop D, Haberle S (2014) A Holocene record of coastal landscape

dynamics in the eastern Kimberley region, Australia. Journal of Quaternary

Science 29, 163-174.

Rabosky DL, Reid J, Cowan MA, Foulkes J (2007) Overdispersion of body size in

Australian desert lizard communities at local scales only: no evidence for the

Narcissus effect. Oecologia 154, 561-570.

Rambaut A, Drummond AJ (2002–2015) TreeAnnotator v2.3.1,

http://beast.bio.ed.ac.uk/treeannotator.

206

Chapter 4: Continental islands support ancient genetic diversity

Rambaut A, Suchard MA, Xie D, Drummond AJ (2014) Tracer v1.6, Available from

http://beast.bio.ed.ac.uk/Tracer.

Read K, Keogh JS, Scott IAW, Roberts JD, Doughty P (2001) Molecular phylogeny of

the Australian frog genera Crinia and Geocrinia and allied taxa (Anura:

Myobatrachidae). Molecular Phylogenetics and Evolution 21, 294-308.

Reeves JM, Bostock HC, Ayliffe LK, et al. (2013) Palaeoenvironmental change in

tropical Australasia over the last 30,000 years–a synthesis by the OZ-

INTIMATE group. Quaternary Science Reviews 74, 97-114.

Rojas-Soto OR, Westberg M, Navarro-Sigüenza AG, Zink RM (2010) Genetic and

ecological differentiation in the endemic avifauna of Tiburón Island. Journal of

avian biology 41, 398-406.

Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesian phylogenetic inference

under mixed models. Bioinformatics 19, 1572-1574.

Rosauer DF, Blom M, Bourke G, et al. (2016) Phylogeographic hotspots and

conservation priorities: an example from the Top End of Australia. In press.

Runemark A, Sagonas K, Svensson EI (2015) Ecological explanations to island

gigantism: dietary niche divergence, predation, and size in an endemic lizard.

Ecology 96, 2077-2092.

Sagonas K, Pafilis P, Lymberakis P, et al. (2014) Insularity affects head morphology,

bite force and diet in a Mediterranean lizard. Biological Journal of the Linnean

Society 112, 469-484.

Si X, Baselga A, Ding P (2015) Revealing beta-diversity patterns of breeding bird and

lizard communities on inundated land-bridge islands by separating the turnover

and nestedness components. PLoS ONE 10, e0127692.

207

Chapter 4: Continental islands support ancient genetic diversity

Silva ACA, Bragg JG, Potter S, et al. (in review) Tropical specialist versus climate

generalist: diversification and demographic history of sister species of Carlia

skinks from northwestern Australia. Molecular Ecology.

Sistrom MJ, Donnellan SC, Hutchinson MN (2013) Delimiting species in recent

radiations with low levels of morphological divergence: a case study in

Australian Gehyra geckos. Molecular Phylogenetics and Evolution 68, 135-143.

Sistrom MJ, Edwards DL, Donnellan SC, Hutchinson MN (2012) Morphological

differentiation correlates with ecological but not with genetic divergence in a

Gehyra gecko. Journal of Evolutionary Biology 25, 647-660.

Sistrom MJ, Hutchinson MN, Hutchinson RG, Donnellan SC (2009) Molecular

phylogeny of Australian Gehyra (Squamata: Gekkonidae) and taxonomic

revision of Gehyra variegata in south-eastern Australia. Zootaxa 2277, 14-32.

Slater GS, Birney E (2005) Automated generation of heuristics for biological sequence

comparison. BMC Bioinformatics 6, 31.

Slatyer C, Rosauer D, Lemckert F (2007) An assessment of endemism and species

richness patterns in the Australian Anura. Journal of Biogeography 34, 583-596.

Stamatakis A (2006) RAxML-VI-HPC: Maximum likelihood-based phylogenetic

analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688-

2690.

Stamatakis A, Hoover P, Rougemont J (2008) A rapid bootstrap algorithm for the

RAxML web servers. Systematic Biology 57, 758-771.

Thomas GH, Meiri S, Phillimore AB (2009) Body size diversification in Anolis: Novel

environment and island effects. Evolution 63, 2017-2030.

208

Chapter 4: Continental islands support ancient genetic diversity

Trifonov VA, Vorobieva NN, Rens W (2009) FISH with and without COT1 DNA. In:

Fluorescence In Situ Hybridization (FISH): Application Guide (ed. Liehr T), pp.

99-109. Springer, Berlin.

Van Valen L (1973) A new evolutionary law. Evolutionary Theory 1, 1-33.

Velo-Antón G, Zamudio KR, Cordero-Rivera A (2012) Genetic drift and rapid

evolution of viviparity in insular fire salamanders (Salamandra salamandra).

Heredity 108.

Vieites DR, Wollenberg KC, Andreone F, et al. (2009) Vast underestimation of

Madagascar's biodiversity evidenced by an integrative amphibian inventory.

Proceedings of the National Academy of Science USA 106, 8267-8272.

Vitousek P, Looope LL, Adsersen H (2013) Islands: biological diversity and ecosystem

function Springer Science & Business Media.

Waeber PO, Wilmé L, Ramamonjisoa B, et al. (2015) Dry forests in Madagascar:

neglected and under pressure. International Forestry Review 17, 127-148.

Wallis LA (2001) Environmental history of northwest Australia based on phytolith

analysis at Carpenter's Gap 1. Quaternary International 82, 103-117.

Wegener JE, Gartner GE, Losos JB (2014) Lizard scales in an adaptive radiation:

variation in scale number follows cliamtic and structrual habitat diversity in

Anolis lizards. Biological Journal of the Linnean Society 113, 570-579.

Werneck FP, Gamble T, Colli GR, Rodrigues MT, Sites Jr. JW (2012) Deep

diversification and long-term persistence in the South American 'dry diagonal':

integrating continent-wide phylogeography and distribution modeling of geckos.

Evolution 66, 3014-3034.

209

Chapter 4: Continental islands support ancient genetic diversity

Whittaker RJ, Fernández-Palacios JM (2007) Island biogeography: ecology, evolution,

and conservation Oxford University Press, Oxford.

Wilcox BA (1978) Supersaturated island faunas: a species-age relationship for lizards

on post-Pleistocene land-bridge islands. Science 199, 996-998.

Wilgenbusch JC, Warren DL, Swofford DL (2004) AWTY: A system for graphical

exploration of MCMC convergence in Bayesian phylogenetic inference.

Woinarski J, Mackey B, Nix H, Traill B (2007) The Nature of Northern Australia:

Natural Values, Ecological Processes and Future Prospects. The Australian

National University E Press, Canberra.

Woinarski JCZ, Armstrong M, Brennan K, et al. (2010) Monitoring indicates rapid and

severe decline of native small mammals in Kakadu National Park, northern

Australia. Wildlife Research 37, 116-126.

Zerbino DR, Birney E (2008) Velvet: algorithms for de novo short read assembly using

de Bruijn graphs. Genome Research 18, 821-829.

210

Chapter 4: Continental islands support ancient genetic diversity

Appendix

Table A4.1. Specimen numbers, localities, & GenBank accession and Dryad details for a) ingroup samples included in molecular and morphological analyses, and b) outgroup samples included in molecular dating analysis.

Table A4.1. a)

GenBank accession Dryad Species Specimen number Locality Lat Long Morphology ND2 BDNF CMOS PDC RAG1 Exons Gehrya multi1 ABTC28145 WA: Mitchell Plateau -14.82 125.71 √ – – – – – – Gehrya multi1 CCM1182 WA: Mitchell Plateau -14.82 125.72 √ – – – – – – Gehrya multi1 CCM1185 WA: Mitchell Plateau -14.83 125.72 √ – – – – – – Gehrya multi1 CCM1186 WA: Mitchell Plateau -14.83 125.72 √ – – – – – – Gehrya multi1 CCM1211 WA: Surveyors Pool, Mitchell Plateau -14.67 125.73 √ – – – – – – Gehrya multi1 CCM1215 WA: Mitchell Plateau -14.83 125.72 √ – – – – – – Gehrya multi1 CCM1217 WA: Mitchell Plateau -14.83 125.72 √ – – – – √ – Gehrya multi1 CCM1218 WA: Mitchell Plateau Rngr Stn -14.82 125.72 √ – – – – – – Gehrya multi1 CCM1223 WA: Mitchell Plateau Rngr Stn -14.82 125.72 √ – – – – – – Gehrya multi1 CCM1224 WA: Mitchell Plateau Rngr Stn -14.82 125.72 √ – – – – – – Gehrya multi1 SAMAR53963 WA: Cape Voltaire -14.35 125.58 JX524068 – – – – – – Gehrya multi1 WAMR158819 WA: Bigge Is -14.60 125.20 √ – – – – – √ Gehrya multi1 WAMR158824 WA: Bigge Is -14.60 125.20 √ – – – – – √ Gehrya multi1 WAMR167804 WA: Surveyors Pool, Mitchell Plateau -14.67 125.73 JX524070 – – – – – – Gehrya multi1 WAMR167855 WA: Mitchell Plateau -14.83 125.72 JX524075 – – – – – √ Gehrya multi1 WAMR167890 WA: Mitchell Plateau -14.82 125.72 JX524074 – – – – – √ Gehrya multi1 WAMR168652 WA: Katers Is -14.47 125.53 √ – – – – – Juvenile Gehrya multi1 WAMR168711 WA: Katers Is -14.47 125.53 JX524071 – – – – – √ Gehrya multi1 WAMR168732 WA: Katers Is -14.47 125.53 JX524072 – – – – – √ Gehrya multi1 WAMR168902 WA: Bigge Is -14.60 125.12 JX524073 – – – – – √ Gehrya multi1 WAMR168905 WA: Bigge Is -14.60 125.12 √ – – – – – √ Gehrya multi2 WAMR117901 WA: Mitchell Plateau -14.91 126.05 JX524077 – – – – – Juvenile Gehrya multi2 WAMR168148 WA: Prince Regent Rvr Nature Rsrv -15.59 125.19 JX524083 – – – – √ √ Gehrya multi2 WAMR168149 WA: Prince Regent Rvr Nature Rsrv -15.59 125.19 JX524084 – – – – – Juvenile Gehrya multi2 WAMR168177 WA: Boongaree Is -15.08 125.18 JX524078 – – – – – √ Gehrya multi2 WAMR168575 WA: Augustus Is -15.35 124.53 √ – – – – – Juvenile Gehrya multi2 WAMR168576 WA: Augustus Is -15.35 124.53 √ – – – – – √ Gehrya multi2 WAMR168577 WA: Augustus Is -15.35 124.53 JX524088 – – – – – √ Gehrya multi2 WAMR168579 WA: Darcy Is -15.34 124.39 √ √ √ √ √ √ Juvenile Gehrya multi2 WAMR168580 WA: Darcy Is -15.34 124.39 √ – – – – – √ Gehrya multi2 WAMR168582 WA: Coronation Is -14.98 124.92 JX524079 – – – – – √ Gehrya multi2 WAMR168584 WA: Byam Martin Is -15.38 124.35 JX524089 – – – – – √ Gehrya multi2 WAMR168585 WA: St Andrew Is -15.35 125.00 √ – – – – – √ Gehrya multi2 WAMR168587 WA: St Andrew Is -15.35 125.00 JX524080 – – – – – √ Gehrya multi2 WAMR171078 WA: Coronation Is -15.03 124.95 √ – – – – – √ Gehrya multi2 WAMR171491 WA: Prince Regent Rvr Nature Rsrv -15.29 125.50 JX524081 – – – – – √ Gehrya multi2 WAMR171501 WA: Harding Rng, Prince Regent Rvr Nature Rsrv -16.32 124.76 JX524086 – – – – – √ Gehrya multi2 WAMR171547 WA: Prince Regent Rvr Nature Rsrv -15.76 125.26 JX524085 – – – – – √ Gehrya multi2 WAMR172066 WA: Storr Is -15.95 124.56 JX524087 – – – – – √ Gehrya multi2 WAMR96949 WA: Mt Trafalger -15.28 125.07 JX524082 – – – – – √ Gehrya occi3 AMS140350 WA: Tunnel Crk -17.86 125.34 JN393920 – – – – – – Gehrya occi3 NMVZ29020 WA: Oscar Rng, W Leopold Downs Rd -17.91 125.28 √ – – – – – √ Gehrya occi3 NMVZ29021 WA: Oscar Rng, W Leopold Downs Rd -17.91 125.28 √ – – – – – √ Gehrya occi3 NMVZ29022 WA: Oscar Rng, W Leopold Downs Rd -17.91 125.28 √ – – – – – √ Gehrya occi3 NMVZ29023 WA: Oscar Rng, W Leopold Downs Rd -17.91 125.28 √ – – – – – √ Gehrya occi3 NMVZ29027 WA: W Leopold Downs Rd -17.91 125.29 √ – – – – – √ Gehrya occi3 NMVZ29031 WA: Oscar Rng, E Leopold Downs Rd -17.92 125.30 √ – – – – √ √ Gehrya occi3 NMVZ29032 WA: Oscar Rng, E Leopold Downs Rd -17.92 125.30 √ – – – – – √ Gehrya occi3 NMVZ29110 WA: McSherry Gap -17.56 125.10 √ – – – – √ √ Gehrya occi3 NMVZ29111 WA: McSherry Gap -17.56 125.10 √ – – – – – √ Gehrya occi3 PMO153 WA: Wire Springs -17.69 125.14 √ – – – – – – Gehrya occi3 PMO156 WA: Wire Springs -17.69 125.14 √ – – – – – – Gehrya occi3 PMO158 WA: Wire Springs -17.69 125.14 √ – – – – – – Gehrya occi3 PMO159 WA: Wire Springs -17.70 125.14 √ – – – – – – Gehrya occi3 PMO166 WA: Wire Springs -17.70 125.14 √ – – – – – – Gehrya occi3 PMO171 WA: Wire Springs -17.69 125.14 √ – – – – – – Gehrya occi3 PMO181 WA: Wire Springs -17.69 125.14 √ – – – – – – Gehrya occi3 PMO186 WA: Wire Springs -17.67 125.07 √ √ √ √ √ √ – Gehrya occi3 PMO193 WA: Brooking Grg -18.03 125.54 √ – – – – √ – Gehrya occi3 PMO195 WA: Brooking Grg -18.03 125.54 √ – – – – – – Gehrya occi3 PMO196 WA: Brooking Grg -18.03 125.54 √ – – – – – –

211

Chapter 4: Continental islands support ancient genetic diversity

Table A4.1. a) continued

GenBank accession Dryad Species Specimen number Locality Lat Long Morphology ND2 BDNF CMOS PDC RAG1 Exons Gehrya occi3 PMO198 WA: Brooking Grg -18.03 125.54 √ – – – – – – Gehrya occi3 PMO202 WA: Geikie Grg -18.11 125.69 √ – – – – – – Gehrya occi3 PMO203 WA: Geikie Grg -18.11 125.69 √ – – – – √ – Gehrya occi3 PMO206 WA: Brooking Grg -18.03 125.54 √ – – – – – – Gehrya occi3 PMO207 WA: Brooking Grg -18.03 125.54 √ – – – – – – Gehrya occi3 PMO212 WA: Brooking Grg -18.03 125.54 √ – – – – – – Gehrya occi3 WAMR172074 WA: Geikie Grg -18.10 125.70 √ – – – – – √ Gehrya occi3 WAMR172077 WA: Geikie Grg -18.10 125.70 √ – – – – – √ Gehrya occi3 WAMR172718 WA: Tunnel Crk -17.61 125.15 √ – – – – – √ Gehrya occi3 WAMR172719 WA: Tunnel Crk -17.61 125.15 √ – – – – – √ Gehrya occi3 WAMR172720 WA: Tunnel Crk -17.61 125.15 √ – – – – – √ Gehrya occi3 WAMR172728 WA: Tunnel Crk -17.61 125.15 √ – – – – – √ Gehyra occi1 BP00762 WA: Sunday Is -16.43 123.18 √ √ √ √ √ √ – Gehyra occi1 WAMR114451 WA: Koolan Is -16.15 123.78 √ – – – – – Juvenile Gehyra occi1 WAMR114452 WA: Koolan Is -16.15 123.78 √ – – – – √ Juvenile Gehyra occi1 WAMR114453 WA: Koolan Is -16.15 123.75 √ – – – – – √ Gehyra occi1 WAMR133190 WA: Koolan Is -16.12 123.72 √ – – – – – √ Gehyra occi1 WAMR146018 WA: Kimbolton HS -16.68 123.83 JX524091 – – – – √ √ Gehyra occi1 WAMR158009 WA: Koolan Is -16.12 123.73 √ – – – – – √ Gehyra occi1 WAMR165444 WA: Irvine Is -16.08 123.54 √ – – – – – √ Gehyra occi1 WAMR165445 WA: Irvine Is -16.08 123.54 √ – – – – √ √ Gehyra occi1 WAMR165551 WA: Koolan Is -16.14 123.75 √ – – – – – √ Gehyra occi1 WAMR168249 WA: Bathurst Is -16.04 123.54 √ – – – – – √ Gehyra occi1 WAMR168303 WA: Bathurst Is -16.02 123.52 √ – – – – – √ Gehyra occi1 WAMR172075 WA: Lachlan Is -16.62 123.47 √ – – – – √ √ Gehyra occi1 WAMR172076 WA: NW Molema Is -16.25 123.82 JX524092 – – – – – √ Gehyra occi1 WAMR172078 WA: Lachlan Is -16.62 123.48 √ – – – – – √ Gehyra occi1 WAMR172082 WA: Sunday Is -16.43 123.18 √ – – – – – – Gehyra occi1 WAMR172083 WA: Lachlan Is -16.62 123.47 √ – – – – – Juvenile Gehyra occi1 WAMR172084 WA: Long Is -16.56 123.35 √ – – – – – Juvenile Gehyra occi1 WAMR172086 WA: Long Is -16.56 123.36 JX524093 – – – – – √ Gehyra occi1 WAMR172097 WA: Lachlan Is -16.62 123.47 JX524094 – – – – – √ Gehyra occi1 WAMR172100 WA: Lachlan Is -16.62 123.47 √ – – – – – √ Gehyra occi1 WAMR172104 WA: Sunday Is -16.42 123.18 √ – – – – – √ Gehyra occi1 WAMR172109 WA: Long Is -16.56 123.36 √ – – – – – Juvenile Gehyra occi1 WAMR172708 WA: Kathleen Is -16.06 123.55 √ – – – – – √ Gehyra occi2 CCM1232 WA: Bell Grg -17.04 125.23 √ – – – – √ – Gehyra occi2 CCM1246 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1250 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1251 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1253 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1254 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1255 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1259 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1260 WA: Silent Grv Crk -17.07 125.25 √ – – – – – – Gehyra occi2 CCM1296 WA: Mt Matthew Grg Crk -16.79 124.92 √ – – – – – – Gehyra occi2 CCM1299 WA: Mt Hart Rd Basalt -16.97 125.03 √ – – – – – – Gehyra occi2 CCM1300 WA: Mt Hart Rd Basalt -16.97 125.03 √ – – – – – – Gehyra occi2 CCM1329 WA: Gibb Rvr Rd -17.12 125.13 √ – – – – – – Gehyra occi2 CCM1331 WA: Gibb Rvr Rd -17.12 125.34 √ – – – – – – Gehyra occi2 CCM1351 WA: Glemont Lookout -16.67 125.43 √ – – – – – – Gehyra occi2 CCM1374 WA: Grevillea Grg -16.50 125.34 √ – – – – – – Gehyra occi2 CCM1450 WA: Teronis Grg -17.30 127.26 √ – – – – – – Gehyra occi2 CCM1477 WA: Gibb Rvr Rd -17.17 125.35 √ – – – – – – Gehyra occi2 CCM1478 WA: Gibb Rvr Rd -17.17 125.35 √ – – – – – – Gehyra occi2 CCM1479 WA: Bell Grg -16.99 125.20 √ – – – – – – Gehyra occi2 CCM1565 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1566 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1567 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1568 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1569 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1577 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1578 WA: Barnett Rvr Grg -16.54 126.13 √ – – – – – – Gehyra occi2 CCM1595 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1596 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1597 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1598 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1599 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1606 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1607 WA: Mt Elizabeth Rd Crk -16.49 126.21 √ – – – – – – Gehyra occi2 CCM1646 WA: Warla Grg -16.34 126.24 √ – – – – – – Gehyra occi2 CCM1670 WA: Warla Grg -16.34 126.24 √ – – – – – – Gehyra occi2 CCM1672 WA: Warla Grg -16.34 126.24 √ – – – – – – Gehyra occi2 CMWA07 WA: Mornington -17.51 126.11 √ – – – – – – Gehyra occi2 CMWA08 WA: Mornington -17.51 126.11 √ – – – – – – Gehyra occi2 NMVZ29036 WA: Silent Grv Crk -17.07 125.25 √ – – – – – √

212

Chapter 4: Continental islands support ancient genetic diversity

Table A4.1. a) continued

GenBank accession Dryad Species Specimen number Locality Lat Long Morphology ND2 BDNF CMOS PDC RAG1 Exons Gehyra occi2 NMVZ29037 WA: Silent Grv Crk -17.07 125.25 √ – – – – – √ Gehyra occi2 NMVZ29055 WA: Bell Grg -17.07 125.25 √ – – – – – √ Gehyra occi2 NMVZ29094 WA: Gibb Rvr Rd, King Leopold Rngs -17.14 125.24 √ – – – – – √ Gehyra occi2 NMVZ29095 WA: Gibb Rvr Rd, King Leopold Rngs -17.14 125.24 √ – – – – – √ Gehyra occi2 NMVZ29104 WA: Napier Rng -17.48 125.03 √ – – – – √ √ Gehyra occi2 NMVZ29105 WA: Napier Rng -17.48 125.03 √ – – – – – √ Gehyra occi2 NMVZ29106 WA: Napier Rng -17.48 125.03 √ – – – – – √ Gehyra occi2 NMVZ29107 WA: Napier Rng -17.48 125.03 √ – – – – – √ Gehyra occi2 NMVZ29108 WA: Napier Rng -17.48 125.03 √ – – – – – √ Gehyra occi2 NMVZ29145 WA: Matthew Grg, King Leopold Rngs -16.78 124.92 √ – – – – – √ Gehyra occi2 PMO121 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – √ Gehyra occi2 PMO135 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – √ Gehyra occi2 PMO136 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – √ Gehyra occi2 PMO174 WA: S Napier Rng -17.48 125.03 √ – – – – – √ Gehyra occi2 PMO95 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra occi2 PMO96 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra occi2 PMO97 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra occi2 PMO98 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra occi2 PMO99 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra occi2 WAMR114458 WA: King Hall Is -16.07 123.41 √ – – – – √ √ Gehyra occi2 WAMR114459 WA: King Hall Is -16.07 123.41 √ – – – – – √ Gehyra occi2 WAMR164773 WA: Mt Nyulasy -16.75 128.28 JX524095 √ √ √ √ √ √ Gehyra occi2 WAMR164774 WA: Mt Nyulasy -16.75 128.28 KJ025272 – – – – – √ Gehyra occi2 WAMR164775 WA: Mt Nyulasy -16.75 128.28 KJ025273 – – – – √ √ Gehyra occi2 WAMR168079 WA: Prince Regent Rvr Nature Rsrv -15.99 125.33 KJ025275 – – – – – √ Gehyra occi2 WAMR168193 WA: Prince Regent Nature Rsrv -15.99 125.33 √ – – – – – Juvenile Gehyra occi2 WAMR168194 WA: Bachsten Crk, Prince Regent Nature Rsrv -15.99 125.33 KJ025276 – – – – – Juvenile Gehyra occi2 WAMR171401 WA: Bachsten Crk, Prince Regent Nature Rsrv -15.99 125.33 √ – – – – – Juvenile Gehyra occi2 WAMR171433 WA: Bachsten Crk, Prince Regent Nature Rsrv -15.99 125.33 √ – – – – – √ Gehyra occi2 WAMR171566 WA: Old Beverley Springs Rd, Prince Regent Nature Rsrv -16.18 125.44 √ – – – – – √ Gehyra occi2 WAMR171572 WA: Old Beverley Springs Rd, Prince Regent Nature Rsrv -16.18 125.44 √ – – – – – √ Gehyra occi2 WAMR171593 WA: Lower Monjon Rocks, Prince Regent Nature Rsrv -15.98 125.37 √ – – – – – √ Gehyra occi2 WAMR172070 WA: Manning Grg -16.66 125.93 JX524100 – – – – – √ Gehyra occi2 WAMR172071 WA: Windjana Grg -17.41 124.95 √ – – – – – √ Gehyra occi2 WAMR172073 WA: Sunday Is -16.43 123.18 √ – – – – √ √ Gehyra occi2 WAMR172081 WA: Hidden Is -16.22 123.45 √ – – – – – Juvenile Gehyra occi2 WAMR172087 WA: Wulalam Is -16.37 124.23 √ – – – – – √ Gehyra occi2 WAMR172089 WA: Manning Grg -16.66 125.93 JX524103 – – – – – √ Gehyra occi2 WAMR172090 WA: Windjana Grg -17.41 124.95 √ – – – – – √ Gehyra occi2 WAMR172092 WA: Manning Grg -16.66 125.93 KJ025274 – – – – – Juvenile Gehyra occi2 WAMR172093 WA: Hidden Is -16.22 123.45 √ – – – – – Juvenile Gehyra occi2 WAMR172094 WA: Wulalam Is -16.37 124.23 JX524104 – – – – – √ Gehyra occi2 WAMR172098 WA: Wulalam Is -16.36 124.22 √ – – – – – √ Gehyra occi2 WAMR172101 WA: Hidden Is -16.22 123.45 √ – – – – – √ Gehyra occi2 WAMR172103 WA: Chambers Is -16.27 123.52 √ – – – – – Juvenile Gehyra occi2 WAMR172106 WA: Manning Grg -16.66 125.93 JX524101 – – – – – Juvenile Gehyra occi2 WAMR172112 WA: Adcock Grg -16.89 125.80 JX524105 – – – – – √ Gehyra occi2 WAMR172114 WA: Unnamed Is -16.34 124.23 √ – – – – – Juvenile Gehyra occi2 WAMR172140 WA: Kingfisher Is -16.08 124.08 JX524106 – – – – – √ Gehyra occi2 WAMR172143 WA: Kingfisher Is -16.08 124.07 JX524108 – – – – √ √ Gehyra occi2 WAMR172144 WA: Kingfisher Is -16.09 124.09 JX524107 – – – – √ √ Gehyra occi2 WAMR172722 WA: Windjana Grg -17.41 124.95 √ – – – – √ √ Gehyra occi2 WAMR172723 WA: Windjana Grg -17.41 124.95 √ – – – – – √ Gehyra occi2 WAMR172724 WA: Windjana Grg -17.41 124.95 √ – – – – – √ Gehyra occi2 WAMR172730 WA: Napier Rng -17.43 124.98 √ – – – – – √ Gehyra occi2 WAMR172731 WA: Napier Rng -17.43 124.98 √ – – – – – √ Gehyra occi2 WAMR172732 WA: Napier Rng -17.43 124.98 √ – – – – – √ Gehyra occi2 WAMR172733 WA: Napier Rng -17.43 124.98 √ – – – – – √ Gehyra occi2 WAMR172768 WA: March Fly Glen, King Leopold Rngs -17.16 125.31 √ – – – – – √ Gehyra occi2 WAMR172772 WA: March Fly Glen, King Leopold Rngs -17.16 125.31 √ – – – – – – Gehyra occi2 WAMR172773 WA: March Fly Glen, King Leopold Rngs -17.16 125.31 √ – – – – – √ Gehyra occi2 WAMR172777 WA: King Leopold Rngs -17.14 125.27 √ – – – – – Juvenile Gehyra occi2 WAMR172778 WA: King Leopold Rngs -17.14 125.27 √ – – – – – √ Gehyra occi2 WAMR172785 WA: March Fly Glen, King Leopold Rngs -17.16 125.31 √ – – – – – √ Gehyra occi2 WAMR172786 WA: March Fly Glen, King Leopold Rngs -17.16 125.31 √ – – – – – √ Gehyra occi2 WAMR172796 WA: Napier Rng -17.20 124.92 √ – – – – – √ Gehyra occi2 WAMR172797 WA: Napier Rng -17.20 124.92 √ – – – – – √ Gehyra occi2 WAMR172798 WA: Napier Rng -17.34 124.82 √ – – – – – √ Gehyra occi2 WAMR172799 WA: Manning Grg, Mt Barnett Stn -16.66 125.93 √ – – – – – √ Gehyra occi2 WAMR172826 WA: Barnett Rvr Grg, Mt Barnett Stn -16.54 126.13 √ – – – – – √ Gehyra occi2 WAMR175479 WA: Sir John Grg -17.53 126.21 √ – – – – – √ Gehyra occi2 WAMR175481 WA: Glemont Lookout -16.67 125.43 √ – – – – – √ Gehyra occi2 WAMR175482 WA: Silent Grv Crk -17.07 125.25 √ – – – – – √ Gehyra occi2 WAMR175483 WA: Mt Hart Grg -16.79 124.92 √ – – – – – √ Gehyra occi2 WAMR175486 WA: Grevillea Grg -16.50 125.34 √ – – – – – Juvenile Gehyra occi2 WAMR175487 WA: Mornington -17.51 126.11 √ – – – – – √

213

Chapter 4: Continental islands support ancient genetic diversity

Table A4.1. a) continued

GenBank accession Dryad Species Specimen number Locality Lat Long Morphology ND2 BDNF CMOS PDC RAG1 Exons Gehyra occi2 WAMR175488 WA: Bell Grg -16.99 125.20 √ – – – – – √ Gehyra occi2 WAMR175490 WA: Bell Grg -17.04 125.23 √ – – – – – √ Gehyra occi2 WAMR175491 WA: Mt Hart Grg -16.79 124.92 √ – – – – – √ Gehyra occi2 WAMR175494 WA: Gibb Rvr Rd -17.12 125.13 √ – – – – – √ Gehyra occi2 WAMR175497 WA: Bell Grg -16.99 125.20 √ – – – – – √ Gehyra occi2 WAMR175498 WA: Police Valley -16.82 126.22 √ – – – – √ √ Gehyra occi2 WAMR175499 WA: Mt Hart Rd Basalt -16.97 125.03 √ – – – – – √ Gehyra occi2 WAMR175752 WA: Mt Hart Rd Basalt -16.97 125.03 √ – – – – – √ Gehyra occi2 WAMR175753 WA: Bell Grg -16.99 125.20 √ – – – – – √ Gehyra occi2 WAMR175754 WA: Glemont Lookout -16.67 125.43 √ – – – – – √ Gehyra occi2 WAMR175755 WA: Grevillea Grg -16.50 125.34 √ – – – – – √ Gehyra occi2 WAMR175756 WA: Mornington -17.51 126.11 √ – – – – – √ Gehyra occi2 WAMR175757 WA: Gibb Rvr Rd -17.12 125.13 √ – – – – – √ Gehyra occi2 WAMR175758 WA: Sir John Grg -17.53 126.21 √ – – – – – √ Gehyra occi2 WAMR175760 WA: Silent Grv Crk -17.07 125.25 √ – – – – – √ Gehyra occi2 WAMR175761 WA: Silent Grv Crk -17.07 125.25 √ – – – – – √ Gehyra occi2 WAMR175762 WA: Gibb Rvr Rd -17.17 125.35 √ – – – – – √ Gehyra occi2 WAMR175764 WA: Gibb Rvr Rd -17.12 125.13 √ – – – – – √ Gehyra xenopus BiggeIs WAMR164925 WA: Bigge Is -14.58 125.10 √ √ √ √ √ √ √ Gehyra xenopus BiggeIs WAMR168217 WA: Bigge Is -14.60 125.12 √ – – – – √ √ Gehyra xenopus BiggeIs WAMR168218 WA: Bigge Is -14.60 125.12 √ – – – – √ √ Gehyra xenopus BoongareeIs WAMR168550 WA: Boongaree Is -15.10 125.20 √ – – – – √ Juvenile Gehyra xenopus BoongareeIs WAMR168551 WA: Boongaree Is -15.10 125.20 √ – – – – √ √ Gehyra xenopus BoongareeIs WAMR168552 WA: Boongaree Is -15.10 125.20 √ √ √ √ √ √ √ Gehyra xenopus MP1 WAMR138892 WA: Donkins Hill -14.99 125.51 KJ025235 √ √ √ √ √ √ Gehyra xenopus MP2 WAMR158844 WA: Capstan Is -14.58 125.26 √ – – – – – √ Gehyra xenopus MP2 WAMR158910 WA: Capstan Is -14.58 125.26 √ √ √ √ √ √ √ Gehyra xenopus MP3 AMS140173 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.70 JN393932 – – – – – – Gehyra xenopus MP3 CCM1216 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 √ – – – – – – Gehyra xenopus MP3 WAMR164892 WA: Katers Is -14.45 125.52 √ – – – – – √ Gehyra xenopus MP3 WAMR164893 WA: Katers Is -14.45 125.52 √ – – – – – √ Gehyra xenopus MP3 WAMR164911 WA: Katers Is -14.45 125.52 √ – – – – – √ Gehyra xenopus MP3 WAMR167763 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 JX524121 – – – – – √ Gehyra xenopus MP3 WAMR167764 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 √ – – – – – √ Gehyra xenopus MP3 WAMR167765 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 JX524123 – – – – – √ Gehyra xenopus MP3 WAMR167807 WA: Surveyors Pool, Mitchell Plateau -14.67 125.73 JX524125 – – – – – √ Gehyra xenopus MP3 WAMR167808 WA: Surveyors Pool, Mitchell Plateau -14.67 125.73 JX524120 – – – – – √ Gehyra xenopus MP3 WAMR167809 WA: Surveyors Pool, Mitchell Plateau -14.67 125.73 JX524124 – – – – – √ Gehyra xenopus MP3 WAMR167866 WA: Mitchell Plateau -14.82 125.72 JX524122 – – – – – √ Gehyra xenopus MP3 WAMR168712 WA: Katers Is -14.47 125.53 √ √ √ √ √ √ Juvenile Gehyra xenopus MP3 WAMR168748 WA: Katers Is -14.47 125.53 √ – – – – √ √ Gehyra xenopus MP3 WAMR171587 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 √ – – – – – Juvenile Gehyra xenopus MP3 WAMR171588 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 √ – – – – – √ Gehyra xenopus MP3 WAMR171589 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 √ – – – – √ √ Gehyra xenopus MP3 WAMR175484 WA: Mitchell Plateau -14.83 125.72 √ – – – – – Juvenile Gehyra xenopus MP3 WAMR175492 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.71 √ – – – – √ √ Gehyra xenopus MP3 WAMR175763 WA: Little Mertens Falls, Mitchell Plateau -14.82 125.72 √ – – – – √ √ Gehyra xenopus Nth1 WAMR113993 WA: King Edward Rvr -14.92 126.20 √ √ √ √ √ √ √ Gehyra xenopus Nth1 WAMR151976 WA: SW Osborn Is -14.35 125.95 √ – – – – – √ Gehyra xenopus Nth1 WAMR151985 WA: SW Osborn Is -14.35 125.95 √ – – – – – √ Gehyra xenopus Nth1 WAMR164854 WA: Middle Osborn Is -14.31 126.03 √ – – – – √ Juvenile Gehyra xenopus Nth1 WAMR168791 WA: Middle Osborn Is -14.32 126.00 √ – – – – – – Gehyra xenopus Nth1 WAMR168792 WA: Middle Osborn Is -14.32 126.00 √ – – – – – – Gehyra xenopus Nth1 WAMR172064 WA: King Edward Rvr Xing -14.89 126.20 KJ025233 – – – – – √ Gehyra xenopus Nth2 CCM0914 WA: King Edward Rvr, Theda -14.52 126.46 √ – – – – – – Gehyra xenopus Nth2 CCM0942 WA: King Edward Rvr, Theda -14.52 126.46 √ – – – – – – Gehyra xenopus Nth2 CCM0943 WA: King Edward Rvr, Theda -14.52 126.46 √ – – – – – – Gehyra xenopus Nth2 CCM0944 WA: King Edward Rvr, Theda -14.52 126.46 √ – – – – – – Gehyra xenopus Nth2 WAMR174021 WA: King Edward Rvr -14.75 126.21 √ – – – – – Juvenile Gehyra xenopus Nth2 WAMR174022 WA: King Edward Rvr -14.75 126.21 √ – – – – – √ Gehyra xenopus Nth2 WAMR174027 WA: King Edward Rvr -14.75 126.21 √ – – – – – √ Gehyra xenopus Nth2 WAMR174029 WA: King Edward Rvr -14.75 126.21 √ – – – – – √ Gehyra xenopus Nth2 WAMR174180 WA: N King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR174181 WA: N King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR174182 WA: N King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR174183 WA: N King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR174184 WA: N King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR174185 WA: N King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR174187 WA: N King Edward Rvr -14.52 126.45 √ – – – – – Juvenile Gehyra xenopus Nth2 WAMR174190 WA: King Edward Rvr -14.52 126.45 √ – – – – – √ Gehyra xenopus Nth2 WAMR175485 WA: King Edward Rvr, Theda -14.52 126.46 √ – – – – √ Juvenile Gehyra xenopus Nth2 WAMR175489 WA: King Edward Rvr, Theda -14.52 126.46 √ – – – – – Juvenile Gehyra xenopus Nth2 WAMR175759 WA: King Edward Rvr, Theda -14.52 126.46 √ √ √ √ √ √ √ Gehyra xenopus Sth1 WAMR172052 WA: Storr Is -15.95 124.56 √ √ √ √ √ √ √ Gehyra xenopus Sth2 WAMR168544 WA: Byam Martin Is -15.38 124.35 √ – – – – – √ Gehyra xenopus Sth2 WAMR168548 WA: Darcy Is -15.34 124.39 √ – – – – √ Juvenile

214

Chapter 4: Continental islands support ancient genetic diversity

Table A4.1. a) continued

GenBank accession Dryad Species Specimen number Locality Lat Long Morphology ND2 BDNF CMOS PDC RAG1 Exons Gehyra xenopus Sth2 WAMR168549 WA: Darcy Is -15.34 124.39 √ – – – – – √ Gehyra xenopus Sth2 WAMR171037 WA: Byam Martin Is -15.39 124.36 √ – – – – – √ Gehyra xenopus Sth2 WAMR171038 WA: Byam Martin Is -15.39 124.36 √ √ √ √ √ √ √ Gehyra xenopus Sth2 WAMR171040 WA: Darcy Is -15.34 124.39 √ – – – – – √ Gehyra xenopus Sth2 WAMR171041 WA: Darcy Is -15.34 124.39 √ – – – – – √ Gehyra xenopus Sth3 WAMR172051 WA: Unnamed Is -15.91 124.46 √ √ √ √ √ √ √ Gehyra xenopus Sth3 WAMR172053 WA: Unnamed Is -15.91 124.46 √ – – – – √ √ Gehyra xenopus Sth3 WAMR172065 WA: Unnamed Is -15.91 124.46 √ – – – – – Juvenile Gehyra xenopus Sth4 WAMR168150 WA: Prince Regent Nature Reserve -15.59 125.19 √ – – – – – √ Gehyra xenopus Sth4 WAMR168151 WA: Camp Crk, Prince Regent -15.59 125.19 √ – – – – – √ Gehyra xenopus Sth4 WAMR168152 WA: Camp Crk, Prince Regent -15.59 125.19 JX524128 – – – – – √ Gehyra xenopus Sth4 WAMR171400 WA: Bachsten Crk, Prince Regent Nature Reserve -15.99 125.33 √ – – – – √ √ Gehyra xenopus Sth4 WAMR171428 WA: Bachsten Crk, Prince Regent Nature Reserve -15.99 125.32 √ – – – – – √ Gehyra xenopus Sth4 WAMR171431 WA: Monjon Rocks, Prince Regent Nature Reserve -15.98 125.37 √ – – – – – √ Gehyra xenopus Sth4 WAMR171432 WA: Monjon Rocks, Prince Regent Nature Reserve -15.98 125.37 √ – – – – – √ Gehyra xenopus Sth4 WAMR171492 WA: Prince Regent Nature Reserve -15.29 125.50 √ √ √ √ √ √ √ Gehyra xenopus Sth4 WAMR171493 WA: Prince Regent Nature Reserve -15.29 125.50 √ – – – – – √ Gehyra xenopus Sth4 WAMR171507 WA: Prince Regent Nature Reserve -15.76 125.26 √ – – – – – √ Gehyra xenopus Sth4 WAMR171546 WA: Prince Regent Nature Reserve -15.76 125.26 √ – – – – – √ Gehyra xenopus Sth5 CCM1286 WA: Mt Hart Grg -16.79 124.92 √ – – – – √ – Gehyra xenopus Sth5 NMVZ29050 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ – – – – – – Gehyra xenopus Sth5 NMVZ29144 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ – – – – – – Gehyra xenopus Sth5 NMVZ29146 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ – – – – – – Gehyra xenopus Sth5 NMVZ29147 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ – – – – – – Gehyra xenopus Sth5 NMVZ29148 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ √ √ √ √ √ – Gehyra xenopus Sth5 NMVZ29149 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ – – – – – – Gehyra xenopus Sth5 PMO105 WA: W Artesian Rng -16.55 125.00 √ – – – – √ – Gehyra xenopus Sth5 PMO114 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra xenopus Sth5 PMO118 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – – Gehyra xenopus Sth5 PMO119 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – – Gehyra xenopus Sth5 PMO120 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – – Gehyra xenopus Sth5 PMO139 WA: Charnley Rvr, E Artesian Rng -16.35 125.23 √ – – – – – – Gehyra xenopus Sth5 PMO59 WA: Matthew Grg, King Leopold Rng -16.78 124.92 √ – – – – – – Gehyra xenopus Sth5 PMO94 WA: W Artesian Rng -16.55 125.00 √ – – – – – – Gehyra xenopus Sth5 WAMR168051 WA: Quail Falls -15.75 125.37 KJ025234 – – – – – – Gehyra xenopus Sth5 WAMR172776 WA: King Leopold Rng -17.14 125.27 √ – – – – – √ Gehyra xenopus Sth5 WAMR172780 WA: King Leopold Rng -17.14 125.27 √ – – – – – Juvenile Gehyra xenopus Sth5 WAMR175480 WA: Bell Grg -16.99 125.20 √ – – – – √ √ Gehyra xenopus Sth5 WAMR175493 WA: Mt Hart Grg -16.79 124.92 √ – – – – – √ Gehyra xenopus Sth5 WAMR175495 WA: Bell Grg -16.99 125.20 √ – – – – – √ Gehyra xenopus Sth5 WAMR175496 WA: Munja Trk, Mt Elizabeth -16.18 125.99 √ – – – – – √ Gehyra xenopus Sth5 WAMR175751 WA: Bell Grg -17.04 125.23 √ – – – – – √ Gehyra xenopus UwinsIs WAMR168545 WA: Uwins Is -15.27 124.82 √ – – – – √ Juvenile Gehyra xenopus UwinsIs WAMR168546 WA: Uwins Is -15.27 124.82 √ – – – – √ √ Gehyra xenopus UwinsIs WAMR168547 WA: Uwins Is -15.27 124.82 √ √ √ √ √ √ √ Gehyra xenopus UwinsIs WAMR171039 WA: Uwins Is -15.26 124.80 √ – – – – √ √

215

Chapter 4: Continental islands support ancient genetic diversity

Table A4.1. b)

GenBank accession Species Specimen number Locality Lat Long BDNF Cmos PDC RAG1 Gehrya australis1 WAMR172043 WA: Sunday Is -16.43 123.18 √ √ √ √ Gehrya australis2 NTMR21022 NT: Black Point -11.15 132.17 √ √ √ √ Gehrya australis3 4W02 √ √ √ √ Gehrya australis4 CCM0322 QLD: Bowthorn Stn Tip -18.09 138.30 √ √ √ √ Gehrya borroorloola1 CCM2577 NT: Limmen Rvr -15.27 135.50 √ √ √ √ Gehrya borroorloola2 CCM2688 NT: Yiyinti Rng, Lorella Springs -15.72 135.64 √ √ √ √ Gehrya "Cape York sp" NTMR34839 QLD: Peninsula Development Rd, 9k S Archer Rvr -13.48 142.97 √ √ √ √ Gehrya catenata SAMAR55893 QLD: N Tambo -24.65 146.38 √ √ √ √ Gehrya dubia1 CCM0346 QLD: Hell's Gate RH -17.45 138.36 √ √ √ √ Gehrya dubia2 NTMR86676 QLD: Pallarenda, Townsville -19.21 146.78 √ √ √ √ Gehrya girloorloo CCM3330 WA: Gogo Stn, Virgin Hills, 1km Bob's Bore -18.29 125.58 √ √ √ √ Gehrya "Groote" CCM3733 NT: Groote Eyland -13.97 136.58 √ √ √ √ Gehrya ipsa CCM3210 WA: Whipsnake Grg, Bungle Bungles -17.40 128.41 √ √ √ √ Gehrya kimberleyi CCM1332 WA: Meda Stn -17.37 124.00 √ √ √ √ Gehrya koira1 CCM0974 √ √ √ √ Gehrya koira2 NTMR35099 NT: Dorat Road, Adelaide Rvr -13.24 131.11 √ √ √ √ Gehrya koira3 CCM3333 WA: Bob's Bore, Gogo Stn -18.29 125.58 √ √ √ √ Gehrya manus RNF9050 PNG: Manus Is -2.08 147.00 √ √ √ √ Gehrya membranicuralis BPBM34738 PNG: Papua New Guinea -9.45 147.18 √ √ √ √ Gehrya "multihybrid" WAMR172824 WA: Manning Grg, Mt Barnett Stn -16.66 125.93 √ √ √ √ Gehrya mutilata SAMAR35716 IND: Krakatau Is -6.10 105.43 √ √ √ √ Gehrya nana1.1 CCM1433 WA: Sir John Grg, Mornington -17.53 126.21 √ √ √ √ Gehrya nana1.2 CCM0658 NT: Keep Rvr, Jarnem Loop Wlk -15.75 129.09 √ √ √ √ Gehrya nana2.1 CCM2896 NT: Enviro Camp, Litchfield -13.06 130.97 √ √ √ √ Gehrya nana2.2 NTMR37066 NT: 50k SW Katherine -14.80 131.93 √ √ √ √ Gehrya nana3.1 CCM0651 NT: Bullo Rvr Stn -15.66 129.66 √ √ √ √ Gehrya nana3.2 CCM2936 NT: Florence Falls TO -13.10 130.78 √ √ √ √ Gehrya nana4.1 CCM0985 WA: Mud Springs, 20k SW Theda HS -14.88 126.36 √ √ √ – Gehrya nana4.2 TR759 WA: Coucal Falls -15.01 126.84 √ √ √ √ Gehrya nana4.3 WAMR172134 WA: Wargul Wargul -13.94 126.17 √ √ √ √ Gehrya nana4.4 WAMR173130 WA: Cockburn Rng -15.82 128.10 √ √ √ √ Gehrya nana5 CCM1441 WA: Sir John Grg, Mornington -17.53 126.21 √ √ √ √ Gehrya nana6 WAMR165439 WA: Irvine Is -16.08 123.55 √ √ √ √ Gehrya nana7 CCM2411 NT: Ngukurr -14.73 134.73 √ √ √ √ Gehrya nana8 WAMR174026 WA: King Edward Rvr -14.75 126.21 √ √ √ √ Gehrya nanaEIU SAMAR55786 QLD: Einasleigh Uplands, Mt Surprise Rd -18.29 144.08 √ √ √ √ Gehrya pamela1 CCM0473 NT: Gunlom Falls Trk, Kakadu -13.43 132.42 √ √ √ √ Gehrya pamela2 NMVZ218866 √ √ √ √ Gehrya pamela3 NTMR20373 NT: Jabiluka -12.55 132.92 √ √ √ – Gehrya "pseudoaustralis" CCM1240 WA: Silent Grove -17.07 125.25 √ √ √ √ Gehrya robusta1 CCM0330 QLD: Kingfisher Camp -17.81 138.28 √ √ √ √ Gehrya robusta2 CCM0366 QLD: Calvert Hills -17.20 137.43 √ √ √ √ Gehrya robusta3 CCM0271 QLD: 40k S Riversleigh Rd -18.99 138.69 √ – √ √ Gehrya robusta4 CCM0280 QLD: Lawn Hill Stn -18.58 138.58 √ – √ √ Gehrya robusta5 CCM0239 NT: Bushy Park -21.19 139.76 √ √ √ √

216

Chapter 4: Continental islands support ancient genetic diversity

Table A4.2. Details of primers and protocols used in this study.

217

Chapter 4: Continental islands support ancient genetic diversity

Table A4.3. Optimal partitioning and model schemes for Maximum-Likelihood and Bayesian phylogenetic analyses

Optimal partition and model scheme Dataset RAxML MrBayes/BEAST ND2 ND2_1 (GTR+G); ND2_2 (GTR+G); ND2_3 (GTR+G) ND2_1 (GTR+G); ND2_2 (GTR+G); ND2_3 (GTR+G) Gehyra occidentalis complex 1204 exons, 35 partitions (100% samples/exon) Gehyra xenopus 1183 exons, 35 partitions (100% samples/exon)

218

Chapter 4: Continental islands support ancient genetic diversity

Table A4.4. Morphometric variables measured for analyses of phenotypic variation

Measurement Definition Snout-vent-length (SVL) Length from tip of snout to anterior edge of vent Head width (HW) Maximum width of head Head depth (HD) Maximum depth of head just posterior to orbitals Head length (HL) Length from anterior edge of ear to tip of snout Snout length (SnL) Length of snout as measured from anterior edge of eye to anterior most point of rostrum on a diagonal measure Axilla to groin (trunk) distance (Trk) Length from posterior edge of forelimb insertion to anterior edge of hindlimb insertion, averaged from left and right Number of pre-cloacal pores (PP) Total number of expressed pre-cloacal pores on males

219

Chapter 4: Continental islands support ancient genetic diversity

Table A4.5. Raw morphometric measures recorded for specimens of a) Gehyra occidentalis complex, and b) Gehyra xenopus used in assessment of phenotypic variation.

Table A4.5. a)

Museum Reg# CCM reg# Lineage Sex SVL HL HW HD TrkLR TrkLL TrkLAv SnL PCP WAM 158819 multi1 F 47.2 13.6 8.9 4.8 21.0 21.4 21.2 5.2 WAM 158824 multi1 F 37.5 11.0 7.2 3.7 18.0 8.7 18.0 4.4 WAM 167855 multi1 M 51.3 13.9 9.7 5.0 23.8 23.8 23.8 5.4 45 WAM 167890 multi1 M 46.3 13.3 9.3 4.5 22.6 23.4 23.0 5.0 17 WAM 168652 multi1 JUV 36.2 10.5 7.2 3.4 15.8 16.2 16.0 4.4 WAM 168711 multi1 F 54.5 15.1 10.2 4.9 24.5 23.8 24.2 6.0 WAM 168732 multi1 F 51.1 14.3 10.0 4.5 24.7 23.1 23.9 5.6 WAM 168902 multi1 M 51.3 13.8 9.2 4.9 25.1 23.8 24.5 5.6 WAM 168905 multi1 M 48.8 14.8 9.6 5.3 21.1 22.5 21.8 6.0 50 WAM 96949 multi2 F 54.3 15.2 10.0 4.9 24.5 24.5 24.5 6.2 WAM 117901 multi2 JUV 37.9 11.6 7.5 4.2 16.4 17.4 16.9 4.6 WAM 168148 multi2 F 43.0 12.5 7.9 4.6 21.7 20.1 20.9 5.3 WAM 168149 multi2 JUV 34.4 9.8 6.4 3.5 15.5 16.0 15.8 4.1 WAM 168177 multi2 F 55.0 15.9 9.9 5.4 27.5 26.6 27.1 6.2 WAM 168575 multi2 JUV 44.7 13.8 9.0 4.7 20.7 19.9 20.3 5.6 WAM 168576 multi2 F 49.9 14.1 9.5 5.1 25.1 22.7 23.9 5.4 WAM 168577 multi2 M 52.1 14.5 9.9 5.0 24.7 24.6 24.7 5.8 45 WAM 168579 multi2 JUV 27.1 8.4 5.6 3.1 12.4 12.8 12.6 3.4 WAM 168580 multi2 M 48.2 13.7 9.2 4.7 24.3 23.1 23.7 5.7 26 WAM 168582 multi2 M 51.0 14.5 9.6 4.8 27.0 25.5 26.3 5.8 18 WAM 168584 multi2 F 53.4 14.9 10.0 5.5 27.9 26.2 27.1 6.0 WAM 168585 multi2 M 49.7 14.9 9.3 4.9 21.0 21.8 21.4 5.9 50 WAM 168587 multi2 M 51.8 15.3 10.4 5.4 23.8 23.2 23.5 6.4 47 WAM 171078 multi2 F 51.9 16.8 10.2 5.2 24.8 24.6 24.7 5.5 WAM 171491 multi2 M 54.7 15.4 10.3 5.8 26.3 24.8 25.6 5.8 49 WAM 171501 multi2 M 39.9 11.8 7.9 3.7 19.4 16.6 18.0 4.7 WAM 171547 multi2 M 47.7 13.7 9.4 4.5 20.2 21.7 21.0 5.4 20 WAM 172066 multi2 M 55.0 14.9 10.3 5.3 25.5 25.0 25.3 6.2 46 WAM 114451 occi1 JUV 41.0 13.4 7.9 4.7 17.7 18.8 18.3 5.2 WAM 114452 occi1 JUV 41.9 13.6 8.2 4.3 18.9 17.2 18.1 5.2 WAM 114453 occi1 F 58.8 17.7 10.8 6.3 25.8 22.6 24.2 7.2 WAM 133190 occi1 M 61.4 19.0 12.1 7.1 30.9 26.3 28.6 7.1 25 WAM 146018 occi1 F 55.2 16.3 10.8 6.5 26.2 27.7 27.0 6.7 37 WAM 158009 occi1 F 60.0 16.9 11.8 6.7 26.5 28.3 27.4 6.9 WAM 165444 occi1 F 46.8 14.4 9.4 4.2 21.0 22.0 21.5 5.9 WAM 165445 occi1 F 57.3 18.0 11.3 5.1 27.1 27.6 27.4 6.7 WAM 165551 occi1 M 57.9 17.5 11.5 5.5 26.4 25.9 26.2 7.0 27 WAM 168249 occi1 F 60.2 18.7 11.9 6.6 30.2 27.9 29.1 7.2 WAM 168303 occi1 F 62.3 18.0 11.6 6.4 32.5 32.2 32.4 7.4 WAM 172075 occi1 F 61.5 18.0 12.0 6.3 31.3 28.7 30.0 7.0 WAM 172076 occi1 M 64.4 18.3 12.5 7.2 31.9 31.6 31.8 7.3 27

220

Chapter 4: Continental islands support ancient genetic diversity

Table A4.5. a) continued

Museum Reg# CCM reg# Lineage Sex SVL HL HW HD TrkLR TrkLL TrkLAv SnL PCP WAM 172078 occi1 M 61.2 17.6 12.2 6.9 28.2 30.3 29.3 6.7 33 WAM 172082 occi1 F 61.3 17.3 11.1 5.7 31.0 31.1 31.1 6.9 WAM 172083 occi1 JUV 28.1 9.7 5.7 3.2 12.4 12.4 12.4 3.8 WAM 172084 occi1 JUV 40.3 12.1 7.8 4.0 20.1 20.3 20.2 5.1 WAM 172086 occi1 F 62.0 19.1 12.2 6.5 29.2 30.0 29.6 7.1 WAM 172097 occi1 F 56.7 16.2 10.6 5.6 27.0 26.9 27.0 6.4 WAM 172100 occi1 M 59.0 17.5 11.7 6.6 26.4 29.1 27.8 6.6 29 WAM 172104 occi1 M 56.1 16.8 10.8 6.2 25.6 26.1 25.9 6.3 25 WAM 172109 occi1 JUV 43.5 13.4 8.1 4.4 20.4 20.3 20.4 5.6 WAM 172708 occi1 M 69.0 19.8 14.0 6.9 30.7 30.7 30.7 7.5 21 NMV 75775 occi2 F 59.6 16.6 11.9 6.1 30.1 28.9 29.5 6.8 NMV 75776 occi2 F 60.3 17.1 11.0 5.6 28.6 29.1 28.9 7.3 NMV 75780 occi2 M 60.6 17.1 12.1 6.4 28.0 29.0 28.5 6.7 18 NMV 75781 occi2 F 59.2 16.8 12.4 6.2 27.7 28.4 28.1 7.0 NMV 75782 occi2 M 62.9 17.3 12.6 7.2 28.8 29.3 29.1 7.4 27 NMV 75790 occi2 JUV 34.3 10.9 6.9 3.5 17.5 17.6 17.6 4.0 NMV 76958 occi2 F 57.2 16.2 11.1 5.5 26.9 25.2 26.1 6.8 NMV 76974 occi2 F 58.7 16.9 10.5 5.7 26.5 24.0 25.3 6.7 NMV 77008 occi2 M 56.6 15.8 11.4 5.9 27.2 27.0 27.1 6.6 28 NMV 77009 occi2 M 66.5 18.7 12.9 7.0 33.7 31.2 32.5 7.8 24 NMV 77016 occi2 F 52.6 15.0 10.2 5.7 25.2 24.5 24.9 6.3 NMV 77017 occi2 M 53.5 15.7 10.3 5.5 25.3 25.8 25.6 6.6 19 NMV 77018 occi2 M 66.4 18.2 13.1 6.8 32.2 32.5 32.4 7.4 25 NMV 77019 occi2 M 55.8 16.3 10.5 5.6 28.2 28.0 28.1 6.6 15 NMV 77020 occi2 F 59.7 18.0 11.4 5.9 28.5 27.5 28.0 7.2 WAM 114458 occi2 F 42.2 13.6 7.9 4.5 17.8 18.1 18.0 5.5 WAM 114459 occi2 F 46.6 13.2 9.0 4.4 20.9 20.9 20.9 5.9 WAM 164773 occi2 F 63.2 17.5 12.0 6.3 28.8 28.8 28.8 8.1 WAM 164774 occi2 M 68.6 19.1 13.2 6.9 32.6 32.6 32.6 8.3 25 WAM 164775 occi2 M 75.7 20.4 13.3 7.6 37.2 36.3 36.8 8.7 WAM 168079 occi2 F 27.8 9.5 5.5 2.6 14.0 12.4 13.2 3.7 WAM 168193 occi2 JUV 29.2 9.3 6.1 3.4 12.6 12.2 12.4 3.6 WAM 168194 occi2 JUV 60.6 17.6 11.4 5.9 28.7 26.3 27.5 6.9 WAM 171401 occi2 JUV 32.9 10.5 6.3 3.5 16.4 14.2 15.3 3.9 WAM 171433 occi2 M 62.2 17.2 11.8 5.6 26.2 27.5 26.9 7.0 27 WAM 171566 occi2 F 53.7 15.7 9.8 5.1 25.4 25.4 25.4 5.8 WAM 171572 occi2 M 59.1 17.4 12.0 6.3 30.1 25.3 27.7 6.9 25 WAM 171593 occi2 M 52.0 15.1 9.8 5.1 25.8 24.8 25.3 6.2 25 WAM 172070 occi2 M 61.8 17.4 11.9 5.8 29.7 27.1 28.4 7.9 28 WAM 172071 occi2 M 66.0 19.5 13.0 6.7 33.7 32.5 33.1 8.0 23 WAM 172073 occi2 M 57.7 15.9 11.2 6.5 28.5 27.2 27.9 6.5 29 WAM 172081 occi2 JUV 34.0 11.6 7.5 3.9 14.2 14.6 14.4 4.6 WAM 172087 occi2 M 62.7 18.9 13.2 6.8 29.7 29.4 29.6 7.6 28 WAM 172089 occi2 F 48.6 12.8 8.8 4.1 24.8 23.0 23.9 5.7 WAM 172090 occi2 M 65.8 19.1 11.8 6.5 33.1 29.7 31.4 7.8 WAM 172092 occi2 JUV 27.8 9.8 6.2 2.7 11.9 12.1 12.0 3.7 WAM 172093 occi2 JUV 26.7 9.5 5.5 3.3 10.6 9.6 10.1 3.6 WAM 172094 occi2 M 62.1 17.4 11.9 6.0 29.8 30.4 30.1 6.7 30

221

Chapter 4: Continental islands support ancient genetic diversity

Table A4.5. a) continued

Museum Reg# CCM reg# Lineage Sex SVL HL HW HD TrkLR TrkLL TrkLAv SnL PCP WAM 172098 occi2 M 60.5 17.8 12.1 5.6 27.3 27.3 27.3 7.1 26 WAM 172101 occi2 F 61.3 17.7 12.1 6.4 30.4 27.4 28.9 7.8 WAM 172103 occi2 Juvi 40.2 12.5 8.3 4.1 19.1 16.9 18.0 5.2 WAM 172106 occi2 JUV 36.4 11.6 6.7 3.6 17.1 16.4 16.8 4.7 WAM 172112 occi2 M 62.2 16.5 11.5 5.8 29.4 30.8 30.1 7.1 24 WAM 172114 occi2 JUV 34.8 11.6 6.9 3.8 15.4 14.4 14.9 4.8 WAM 172140 occi2 M 53.1 15.6 10.3 5.4 23.7 21.7 22.7 6.4 20 WAM 172143 occi2 F 52.9 16.0 10.2 5.8 24.4 24.4 24.4 6.7 WAM 172144 occi2 M 49.6 14.5 10.0 5.3 21.0 20.0 20.5 5.9 16 WAM 172722 occi2 F 65.5 18.3 12.7 6.2 30.2 29.6 29.9 7.9 WAM 172723 occi2 F 60.6 17.2 10.8 6.1 29.0 27.8 28.4 7.3 WAM 172724 occi2 F 60.6 17.5 11.5 6.6 29.6 28.6 29.1 6.7 WAM 172730 occi2 F 65.8 18.5 11.7 6.7 31.2 29.3 30.3 7.5 WAM 172731 occi2 F 60.8 16.7 11.9 6.1 28.8 30.4 29.6 6.8 WAM 172732 occi2 M 65.7 18.6 12.6 6.6 32.1 29.0 30.6 7.3 25 WAM 172733 occi2 F 43.7 13.1 8.4 4.7 20.5 20.3 20.4 5.5 WAM 172768 occi2 M 62.2 17.8 11.7 6.6 30.5 29.1 29.8 7.3 WAM 172772 occi2 JUV 31.2 11.1 7.3 3.6 18.0 17.4 17.7 4.8 WAM 172773 occi2 M 35.5 10.6 6.9 3.6 16.4 16.3 16.4 4.7 WAM 172777 occi2 JUV 32.9 10.2 6.5 3.2 15.1 14.5 14.8 4.3 WAM 172778 occi2 F 64.0 10.6 12.4 6.4 32.2 31.5 31.9 7.3 WAM 172785 occi2 F 67.0 18.3 12.4 6.5 30.2 31.3 30.8 7.4 WAM 172786 occi2 M 58.6 17.6 11.0 5.8 26.9 25.5 26.2 6.8 28 WAM 172796 occi2 M 68.5 18.5 12.0 6.7 32.7 33.3 33.0 7.7 24 WAM 172797 occi2 F 67.3 18.6 12.5 6.1 30.7 31.8 31.3 7.7 WAM 172798 occi2 F 58.5 17.3 10.6 6.2 25.9 23.5 24.7 6.8 WAM 172799 occi2 M 65.3 16.5 11.8 6.0 31.4 31.8 31.6 7.0 30 WAM 172826 occi2 F 61.6 15.9 11.1 5.4 30.6 31.4 31.0 6.9 WAM 175479 CCM1442 occi2 M 67.1 18.3 12.5 6.5 29.3 28.2 28.8 7.5 28 WAM 175481 CCM1358 occi2 F 61.2 16.6 11.7 6.4 29.6 29.2 29.4 6.9 WAM 175482 CCM1257 occi2 M 65.5 11.9 12.1 6.3 30.8 30.0 30.4 7.4 27 WAM 175483 CCM1294 occi2 M 62.0 17.5 12.3 5.8 37.1 28.8 33.0 7.6 30 WAM 175486 CCM1405 occi2 JUV 38.4 12.0 7.5 3.8 19.8 18.6 19.2 4.9 WAM 175487 CMWA06 occi2 F 47.9 14.0 8.8 4.3 21.7 20.9 21.3 5.8 WAM 175488 CCM1488 occi2 F 59.7 15.9 11.5 5.1 25.2 25.6 25.4 6.9 WAM 175490 CCM1495 occi2 F 61.2 16.3 11.7 5.5 28.2 26.8 27.5 6.8 WAM 175491 CCM1293 occi2 M 64.0 17.7 12.4 6.2 30.3 30.6 30.5 7.5 28 WAM 175494 CCM1340 occi2 F 64.0 17.6 12.0 6.3 31.9 29.7 30.8 7.3 WAM 175497 CCM1272 occi2 F 57.2 15.2 10.8 5.4 27.4 26.2 26.8 6.5 WAM 175498 CCM1548 occi2 M 54.4 15.4 10.7 5.2 24.7 23.9 24.3 6.3 25 WAM 175499 CCM1303 occi2 F 62.5 16.1 11.6 5.8 31.6 32.0 31.8 6.8 WAM 175752 CCM1305 occi2 M 64.7 17.8 12.3 5.9 32.5 29.9 31.2 7.5 29 WAM 175753 CCM1494 occi2 F 55.5 16.2 11.1 5.1 24.0 24.1 24.1 6.7 WAM 175754 CCM1350 occi2 F 61.9 16.0 12.5 5.9 30.6 28.0 29.3 6.4 WAM 175755 CCM1393 occi2 M 60.6 17.1 11.3 5.9 26.4 26.0 26.2 6.8 28 WAM 175756 CMWA41 occi2 M 67.6 18.9 12.6 7.0 34.0 29.4 31.7 7.7 26 WAM 175757 CCM1341 occi2 M 66.2 18.6 11.9 6.7 27.7 30.6 29.2 7.7 30 WAM 175758 CCM1439 occi2 F 55.8 16.7 10.9 5.3 25.2 24.8 25.0 6.9

222

Chapter 4: Continental islands support ancient genetic diversity

Table A4.5. a) continued

Museum Reg# CCM reg# Lineage Sex SVL HL HW HD TrkLR TrkLL TrkLAv SnL PCP WAM 175760 CCM1252 occi2 M 61.6 17.8 11.6 6.0 27.7 26.3 27.0 6.8 26 WAM 175761 CCM1258 occi2 M 61.4 18.7 12.7 6.3 26.7 26.0 26.4 7.6 21 WAM 175762 CCM1493 occi2 F 61.4 16.5 12.6 5.8 28.2 29.2 28.7 7.3 WAM 175764 CCM1337 occi2 F 50.3 13.2 9.4 4.5 25.2 24.7 25.0 6.0 NMV 76943 occi3 F 46.9 13.8 8.9 4.9 22.9 22.4 22.7 5.4 NMV 76944 occi3 M 45.4 13.6 8.4 4.5 20.6 22.9 21.8 5.5 7 NMV 76945 occi3 M 57.1 16.2 11.1 6.4 27.4 26.6 27.0 6.8 25 NMV 76946 occi3 F 52.2 15.2 9.8 5.4 24.8 25.1 25.0 5.9 NMV 76949 occi3 M 52.2 15.2 10.7 5.0 24.8 23.6 24.2 6.0 21 NMV 76952 occi3 JUV 26.7 8.9 5.3 2.8 10.6 10.5 10.6 3.5 NMV 76953 occi3 F 52.7 14.7 10.0 5.7 27.1 26.1 26.6 6.5 NMV 77022 occi3 F 48.2 13.8 9.0 4.8 22.1 22.1 22.1 5.9 NMV 77023 occi3 M 49.6 14.8 9.1 4.8 22.5 22.9 22.7 5.8 20 WAM 172074 occi3 M 64.2 18.8 11.7 6.7 34.7 30.3 32.5 7.2 WAM 172077 occi3 F 55.8 16.3 10.3 5.4 25.2 29.9 27.6 6.6 WAM 172718 occi3 F 57.2 16.2 11.0 5.7 25.4 25.3 25.4 6.2 WAM 172719 occi3 F 61.3 17.7 11.5 6.3 29.3 27.4 28.4 6.8 21 WAM 172720 occi3 M 40.6 12.4 8.2 4.7 19.1 20.1 19.6 5.3 WAM 172728 occi3 F 56.5 16.3 10.4 6.2 28.1 25.3 26.7 6.2

223

Chapter 4: Continental islands support ancient genetic diversity

Table A4.5. b)

Museum Reg# CCM reg# Lineage Sex SVL HL HW HD TrkLR TrkLL TrkLAv SnL PCP WAM 164925 Bigge_Is F 67.4 19.0 13.0 6.9 34.6 31.4 33.0 8.7 WAM 168217 Bigge_Is F 69.8 20.0 13.8 7.4 34.7 32.4 33.6 8.4 WAM 168218 Bigge_Is F 71.6 19.4 13.4 7.2 34.1 38.8 36.5 8.9 WAM 168897 Bigge_Is M 68.4 20.1 13.6 7.3 34.7 36.6 35.7 8.8 10 WAM 168550 Boongaree_Is JUV 40.4 13.7 8.5 3.8 18.0 19.6 18.8 5.9 WAM 168551 Boongaree_Is F 71.7 22.5 13.9 7.2 31.3 28.8 30.1 9.3 WAM 168552 Boongaree_Is F 68.8 22.0 14.7 7.4 33.8 30.3 32.1 9.3 WAM 138876 MP1_Donkins_Hill F 70.9 20.7 13.0 7.9 31.8 35.6 33.7 9.1 WAM 138892 MP1_Donkins_Hill M 71.7 20.7 13.2 7.3 35.5 32.1 33.8 8.9 13 WAM 158844 MP2_Capstan_Is M 66.5 19.6 12.2 6.9 32.9 28.9 30.9 8.4 10 WAM 158910 MP2_Capstan_Is F 61.8 18.4 11.8 6.0 28.4 28.4 28.4 7.8 WAM 164892 MP3_Mitchell_Plateau F 66.4 19.7 13.3 7.0 32.1 30.1 31.1 9.1 WAM 164893 MP3_Mitchell_Plateau M 68.2 19.5 13.3 7.8 29.8 32.2 31.0 8.9 10 WAM 164911 MP3_Mitchell_Plateau F 64.8 18.9 12.6 7.3 29.0 31.0 30.0 8.6 WAM 167763 MP3_Mitchell_Plateau F 67.5 19.2 13.1 7.0 30.6 30.8 30.7 8.9 WAM 167764 MP3_Mitchell_Plateau F 73.9 20.3 13.6 7.3 33.6 33.4 33.5 8.6 WAM 167765 MP3_Mitchell_Plateau M 73.0 20.9 13.2 7.1 34.0 33.9 34.0 8.9 12 WAM 167807 MP3_Mitchell_Plateau M 67.4 20.1 13.4 6.7 32.3 32.1 32.2 9.2 15 WAM 167808 MP3_Mitchell_Plateau F 72.9 20.8 14.4 7.0 35.4 35.0 35.2 8.9 WAM 167809 MP3_Mitchell_Plateau F 74.1 20.9 14.0 7.1 33.4 36.4 34.9 9.7 WAM 167866 MP3_Mitchell_Plateau F 61.9 17.9 12.0 5.6 29.9 30.0 30.0 7.9 WAM 168491 MP3_Mitchell_Plateau – 78.0 23.4 14.4 8.0 37.5 38.2 37.9 9.6 WAM 168712 MP3_Mitchell_Plateau JUV 45.4 14.0 8.6 4.5 22.5 20.5 21.5 6.0 WAM 168748 MP3_Mitchell_Plateau M 66.9 18.4 12.4 6.5 32.6 30.1 31.4 8.3 12 WAM 171587 MP3_Mitchell_Plateau JUV 38.7 12.6 7.6 4.0 16.5 17.2 16.9 5.4 WAM 171588 MP3_Mitchell_Plateau F 68.1 19.9 13.0 6.9 30.9 31.3 31.1 8.7 WAM 171589 MP3_Mitchell_Plateau M 68.2 20.0 12.6 6.5 28.2 32.6 30.4 8.8 14 WAM 175484 CCM1222 MP3_Mitchell_Plateau JUV 39.9 12.1 7.2 4.2 18.7 17.6 18.2 5.1 WAM 175492 CCM1208 MP3_Mitchell_Plateau F 70.9 19.0 12.8 6.3 33.0 32.8 32.9 8.8 WAM 175763 CCM1723 MP3_Mitchell_Plateau F 67.7 19.7 12.6 5.7 31.7 32.5 32.1 8.6 WAM 113993 North1 M 68.9 19.3 13.8 5.1 32.1 32.6 32.4 8.5 14 WAM 151976 North1 F 55.7 16.7 10.3 5.3 25.8 28.9 27.4 7.4 WAM 151985 North1 F 73.8 21.4 13.7 7.6 34.4 36.1 35.3 9.9 WAM 164854 North1 JUV 44.6 13.5 8.9 4.1 22.0 21.4 21.7 6.2 WAM 172064 North1 F 74.4 21.7 13.4 6.7 33.3 34.0 33.7 9.1 WAM 174021 North1 JUV 47.8 13.8 8.8 4.6 24.2 34.6 29.4 6.3 WAM 174022 North1 M 71.2 20.1 12.4 7.2 32.2 31.3 31.8 8.7 20 WAM 174027 North1 M 70.9 20.6 13.4 7.2 31.6 33.8 32.7 9.1 17 WAM 174029 North1 F 59.8 17.0 11.5 5.5 28.8 30.1 29.5 7.7 WAM 174180 North2_King_Edward_Rvr M 68.0 19.7 13.8 7.5 32.6 33.2 32.9 8.7 18 WAM 174181 North2_King_Edward_Rvr M 74.4 21.0 13.7 6.7 37.2 34.8 36.0 8.6 21 WAM 174182 North2_King_Edward_Rvr M 65.2 18.9 11.8 5.7 29.7 31.1 30.4 8.3 10 WAM 174183 North2_King_Edward_Rvr F 72.3 20.8 13.4 6.8 34.2 34.9 34.6 8.9 WAM 174184 North2_King_Edward_Rvr M 71.2 20.2 13.2 6.9 31.5 35.9 33.7 8.8 39 WAM 174185 North2_King_Edward_Rvr F 63.0 18.7 11.4 5.8 28.2 33.6 30.9 7.7 WAM 174187 North2_King_Edward_Rvr JUV 30.2 10.3 6.2 2.9 15.4 15.0 15.2 3.9 WAM 174190 North2_King_Edward_Rvr F 57.2 16.9 10.3 5.2 27.0 27.9 27.5 7.2 WAM 175485 CCM0941 North2_King_Edward_Rvr JUV 50.9 13.8 9.3 4.6 24.4 24.8 24.6 6.6 WAM 175489 CCM0913 North2_King_Edward_Rvr JUV 40.6 12.0 7.4 3.9 18.2 17.1 17.7 5.3 WAM 175759 CCM0940 North2_King_Edward_Rvr M 71.7 20.5 13.3 6.7 32.5 33.9 33.2 8.7 20 WAM 172052 South1_Storr_Is M 73.7 22.5 14.8 7.2 31.6 31.4 31.5 10.2 10 WAM 168544 South2_Darcy/Byam_Martin_Is F 78.9 23.6 14.9 8.2 35.5 35.4 35.5 10.4 WAM 168548 South2_Darcy/Byam_Martin_Is JUV 51.3 16.5 10.3 5.5 26.2 23.4 24.8 7.2 WAM 168549 South2_Darcy/Byam_Martin_Is M 78.0 23.7 15.2 8.1 34.9 33.7 34.3 10.4 11 WAM 171037 South2_Darcy/Byam_Martin_Is M 74.0 21.2 14.8 7.5 34.0 30.8 32.4 9.6 10 WAM 171038 South2_Darcy/Byam_Martin_Is F 81.2 23.5 15.7 8.3 37.0 40.8 38.9 11.0 WAM 171040 South2_Darcy/Byam_Martin_Is M 74.2 22.6 15.0 8.1 31.7 33.4 32.6 10.0 11

224

Chapter 4: Continental islands support ancient genetic diversity

Table A4.5. b) continued

Museum Reg# CCM reg# Lineage Sex SVL HL HW HD TrkLR TrkLL TrkLAv SnL PCP WAM 171041 South2_Darcy/Byam_Martin_Is M 79.7 24.2 15.1 8.2 38.0 35.2 36.6 10.4 11 WAM 172051 South3_Unnamed_Is F 77.0 23.1 14.0 7.4 35.3 36.6 36.0 9.9 WAM 172053 South3_Unnamed_Is M 77.0 22.8 14.7 7.6 34.4 35.6 35.0 10.1 10 WAM 172065 South3_Unnamed_Is JUV 38.2 11.4 7.3 3.5 16.3 16.6 16.5 5.0 WAM 99055 South4_Prince_Regent M 69.7 21.7 13.6 7.5 31.9 27.7 29.8 9.7 9 WAM 127324 South4_Prince_Regent F 71.6 23.1 14.5 7.4 34.3 30.2 32.3 10.7 WAM 168150 South4_Prince_Regent F 70.9 21.9 13.7 7.0 31.0 30.0 30.5 9.8 WAM 168151 South4_Prince_Regent M 64.4 19.4 12.7 6.6 28.8 29.0 28.9 8.9 11 WAM 168152 South4_Prince_Regent F 71.4 20.8 13.6 6.9 32.0 32.0 32.0 9.1 WAM 168153 South4_Prince_Regent M 71.5 21.4 13.8 7.1 33.0 33.1 33.1 9.8 18 WAM 171400 South4_Prince_Regent M 73.3 22.5 14.2 6.8 33.2 29.7 31.5 10.1 8 WAM 171428 South4_Prince_Regent F 74.7 22.7 14.3 7.4 36.0 32.0 34.0 9.8 WAM 171431 South4_Prince_Regent M 72.1 20.9 13.1 7.1 33.1 33.2 33.2 9.4 9 WAM 171432 South4_Prince_Regent F 74.5 22.0 13.8 7.5 39.0 32.5 35.8 9.4 WAM 171492 South4_Prince_Regent F 75.2 21.8 13.5 7.0 33.9 37.5 35.7 9.5 WAM 171493 South4_Prince_Regent M 75.9 21.8 14.0 6.9 34.2 33.7 34.0 10.1 11 WAM 171507 South4_Prince_Regent M 67.5 21.3 13.4 6.4 27.1 31.7 29.4 9.3 10 WAM 171546 South4_Prince_Regent M 61.0 18.9 11.1 5.9 30.5 28.1 29.3 8.3 10 WAM 175496 CCM1638 South4_Prince_Regent F 74.4 20.1 13.2 6.5 35.5 34.0 34.8 9.2 WAM 172776 South5_King_Leopold_Rng F 59.8 18.8 11.7 5.8 27.7 25.4 26.6 8.3 WAM 172780 South5_King_Leopold_Rng JUV 38.1 13.2 7.4 3.8 18.1 17.9 18.0 5.5 WAM 175480 CCM1275 South5_King_Leopold_Rng F 71.7 21.2 13.9 6.9 35.0 33.4 34.2 9.7 WAM 175493 CCM1292 South5_King_Leopold_Rng F 78.8 21.7 15.0 7.9 37.8 36.8 37.3 9.9 WAM 175495 CCM1273 South5_King_Leopold_Rng M 78.4 23.2 14.3 7.2 35.2 32.3 33.8 9.8 9 WAM 175751 CCM1234 South5_King_Leopold_Rng M 66.0 18.6 12.7 6.3 30.5 30.1 30.3 7.5 30 WAM 168545 Uwins_Is JUV 38.4 12.9 7.5 4.0 15.4 16.3 15.9 5.6 WAM 168546 Uwins_Is F 79.6 24.5 15.8 8.3 32.3 35.9 34.1 10.6 WAM 168547 Uwins_Is M 77.0 23.1 15.5 7.7 32.6 32.7 32.7 9.9 11 WAM 171039 Uwins_Is F 75.2 22.1 14.7 7.6 37.2 36.7 37.0 9.8

225

Chapter 4: Continental islands support ancient genetic diversity

a) c)

b) d) Figure A4.1. Box-and-whisker plots indicating variation in a and c) body size, and b and d) male pre-cloacal pore count, across lineages for a–b) Gehyra occidentalis complex, and c–d) G. xenopus. Box colours match broad lineage or clade colours from maps and phylogenetic trees in Figs. 4.2. and 4.3. respectively.

226

Chapter 5: General discussion

General discussion

227

Chapter 5: General discussion

Accurately recognising and describing patterns of diversity is crucial for biological research and the ability to understand ecological and evolutionary processes. Advances in genetic technologies improve our capability to assemble increasingly large datasets, which simultaneously reveal the challenges posed by morphologically cryptic lineages and introgression whilst also providing the power to better resolve such complex systems. It is important we continue to accumulate robust datasets and use integrative methods to more precisely assess patterns of diversity within environments. This is particularly relevant to understanding the processes that drive the accumulation of biodiversity in different regions, many of which are subject to significant environmental degradation or change. The aim of my research was to accurately document patterns of diversity within the understudied tropical savannah region of the Kimberley so as to better understand how the landscape and environmental history of this region has shaped the current distribution of biodiversity. I used comparative phylogeographic and integrative taxonomic methods, including next-generation sequencing techniques, to assess and compare diversity patterns across the Kimberley and related regions and biomes. I investigated these phylogeographic patterns using geckos as a model system and compared ecologically variable taxa with broadly similar distributions to infer the role of ecology in shaping contrasting diversification responses to environmental change. Here I summarise the major findings of this work and their broader evolutionary significance, and finally provide some recommendations for future avenues of research which would further improve our understanding of diversification processes and the factors shaping biodiversity.

228

Chapter 5: General discussion

5.1. Key findings

5.1.1. High diversity

The Kimberley region supports high levels of deep genetic diversity in geckos, much of which is cryptic and geographically-structured. Previous studies had revealed higher than expected levels of mtDNA diversity in various reptiles, amphibians, and mammals throughout the Kimberley, however the pervasiveness and taxonomic relevance of these patterns was not clearly understood (e.g. Catullo et al. 2014; Pepper et al. 2011; Potter et al. 2012; Smith et al. 2011). In each of the taxonomic groups I investigated I used multi-locus nuclear datasets to confirm which patterns are indicative of presently isolated lineages. Results indicated high levels of cryptic diversity within the Kimberley in each of the taxa I considered, much of which was supported by additional molecular data. Larger nuclear datasets provided a better ability to recover genetic structure, although some deep mtDNA lineages still tended to collapse under further scrutiny. My findings support the recognition of 15+ new candidate species in Strophurus (Chapter

2), Oedura (Chapter 3; Oliver et al. 2014), and Gehyra (Chapter 4) geckos endemic to the Kimberley. These are deeply divergent and independently evolving lineages, recoverable and supported by different molecular markers, and in some cases also morphological traits. Though not all diversity identified may represent unrecognised species, it is still indicative of important population structure, and can help improve inference and understanding of evolutionary processes of diversification and responses of taxa to environmental change.

229

Chapter 5: General discussion

5.1.2. Scaling of lineage depth and range-size with aridity

Phylogeographic patterns varied across each taxonomic group I investigated, however I observed a common trend for the depth of genetic divergence and range-size of lineages to scale with aridity. More widespread distributions of shallower genetic structure were observed in arid regions compared to deep, short-range endemics in mesic areas. This pattern within the Kimberley appears to reflect at a micro-scale a similar scaling of diversity as has been observed at a continental scale across the AMT and AAZ biomes (Chapters 2 and 3). In addition, the ability to delimit species or independently evolving lineages tended to degrade as distributions extended into more climatically harsh environments, suggesting a possible break-down of lineage boundaries. Probable introgression events were observed in the same general region

(central and south Kimberley) regardless of taxonomic group. This suggests that climatically stable areas, such as highly mesic regions, may facilitate greater persistence of fine-scale diversity and potentially in situ diversification of lineages. In contrast, climatically harsh regions may be more vulnerable to extreme climatic and habitat change, which may result in more dynamic extirpation-re-colonisation processes or shifts in the distributions of taxa tracking more suitable habitat conditions.

5.1.3. Refugia

Concordance of phylogeographic patterns were observed in all taxa in terms of consistency in the locations of high genetic diversity and short-range endemic signatures. Three key regions in particular stood out: the mesic west-coast and continental islands of the Kimberley, the lowland Ord Region dividing the Kimberley and Top End, and the southern Devonian Reef System (DRS) limestone ranges (Fig.

230

Chapter 5: General discussion

5.1.). Evidence of ancient and deeply genetically-divergent micro-endemics within multiple continental islands within the Kimberley is particularly interesting. Contrary to the general assumption that land-bridge island biota reflect a subset of mainland taxa – due to relatively recent isolation and proximity to mainland populations – my results indicate, at least within tropical savannah regions continental islands can support highly genetically divergent lineages, similar to mainland rock refugia, beyond timeframes of island-mainland isolation (Chapter 4). In addition, the more restricted distributions of highly-specialised taxa without other closely related congeners within the Kimberley

(e.g. Gehyra xenopus; see Chapter 4) could be considered relictual in nature, with the suggestion that these lineages are all that remains after extirpation and fragmentation of populations from a once much more widespread common ancestor. The regions to which these lineages are now restricted corresponds with the same areas highlighted for particularly short-range endemism in other widespread taxa (Chapter 3), implying that the habitat or climate within these environments is unique. This evidence suggests these key regions are refugia within the Kimberley that allow greater isolated persistence of diversity compared to the rest of the landscape, and may even facilitate in situ diversification over long timescales.

5.1.4. The role of ecology in environmentally-driven diversification

Regardless of the form of ecological specialisation, genetic diversity appears to be higher for taxa within the Kimberley compared to related taxa in the AAZ or even other regions of the AMT such as the Top End (Chapters 2 and 3). However, in addition to higher patterns of diversity within the Kimberley in general, I observed that ecology plays an important role in how different taxa respond to environmental change and the

231

Chapter 5: General discussion

consequent diversification process they undergo can result in contrasting patterns of genetic structuring across a landscape. This is most clearly indicated in the starkly contrasting distribution patterns (relictual versus widespread) of the two rock-specialist

Kimberley Oedura which, when considered in conjunction with quite varied morphologies, imply quite distinct ecologies (Chapter 3). Furthermore, the more specialised the ecology, the more sensitive the taxa to environmental perturbation. The higher degree of genetic diversity over a more restricted distribution range exhibited in

Gehyra xenopus compared to G. occidentalis could be attributed to the greater specialisation of this clade for very particular environmental and habitat conditions

(Chapter 4). Greater sensitivity to environmental change can result either in extremely high genetic structuring (e.g. G. xenopus; Chapter 4), whereby lineages become heavily sundered into micro-refugia throughout a landscape, or in large-scale extirpation leaving only remnants of a lineage persisting in highly-restricted distributions in remaining suitable habitat (e.g. Oedura filicipoda, O. murrumanu; Chapter 3).

5.2. Future considerations

5.2.1. Introgression

Incidence of possible introgression events or break-down of lineage boundaries was highest within the southern and central Kimberley regions and was observed in some form for all taxa investigated (least noticeable in Strophurus, except in the paraphyly S. mcmillani and S. robinsoni in the southern Kimberley; Chapter 2). In addition, this pattern is not unique to the taxa included herein, but has been noted to occur in the same region for skinks and other gecko species (C. Moritz, pers. comm.). More detailed investigation with increased sampling and next-generation genetic datasets would

232

Chapter 5: General discussion

greatly enhance understanding of the type, age, and frequency of hybridisation occurring within these areas. Furthermore, such work could allow us to more clearly elucidate what it is in particular about the environment or history of these regions of the

Kimberley that is resulting in a higher incidence of lineage boundary break-down compared to other areas.

5.2.2. Additional trait data

My findings highlight the inherent challenges of recognising and understanding diversity patterns in highly morphologically cryptic species complexes, especially in the possible presence of reticulate population processes and introgression. In certain cases the molecular data or morphological characteristics available may not provide the power to assist resolution if by chance these elements are not being selectively acted upon as part of distinction between lineages. In many cases however, the deep lineages detected in mtDNA may be highly distinct in alternative traits not measured in this work, but highly relevant to occurrence in particular habitats or regions. For this reason, future work would benefit from consideration of ecological or physiological trait data, including tolerance for different climatic conditions. Additionally, investigating variation in traits such as pheromones, which could be important for species recognition and potentially reproductive isolation between lineages, may help establish how parapatric cryptic lineages remain evolutionarily independent.

5.3. Theoretical directions

In general, these findings only go a small way towards improving our understanding of the evolutionary processes that generate diversity. Comparative studies would be

233

Chapter 5: General discussion

useful to determine if the patterns observed for vertebrate herpetofauna of the

Kimberley region are universal within tropical savannah biomes. With the increased use of large-scale genomic datasets we can begin to better understand speciation histories, unravel cryptic species complexes and paint a clearer picture of the true level of biodiversity on earth. Furthermore, technological advances in ecological and geological modelling methods will enable us to put such genetic patterns into the context of environmental change with more accuracy than ever before. With these tools, we can therefore contribute to integrative conservation efforts that can protect the most important areas for generating and maintaining biodiversity.

5.4. Summary

This thesis builds upon previous phylogeographic understanding of diversity patterns within the Kimberley region of the Australian Monsoonal Tropics with major increases in geographic and molecular sampling and consideration of ecologically variable taxa.

My findings confirmed the presence of high geographically-structured biodiversity and established that much of this cryptic diversity represents important evolutionarily independent lineages. I also added significant evidence to support that the Mitchell

Plateau represents a biodiversity hotspot within the Kimberley, in addition to recognition of multiple extra hotspots for localised endemism in refugia including the west Kimberley islands, southern DRS limestone ranges, and the Ord Region to the east

(Fig. 5.1.). This thesis indicates there is a complex interaction of climatic variation, heterogeneous topology, and ecology shaping the diversification processes of taxa within the Kimberley resulting in the production and persistence of particularly high diversity compared to related regions of the AMT and surrounding biomes (AAZ).

234

Chapter 5: General discussion

Figure 5.1. Key biodiversity hotspots with deeply divergent localised endemism: west Kimberley islands, mesic west-coast (Mitchell Plateau), southern Devonian Reef System (DRS) limestone ranges, and the lowland east Kimberley (Ord Region, delineated by dark blue dashed line). Brown dashed line denotes the broader Kimberley Plateau region.

235

Chapter 5: General discussion

References

Catullo RA, Lanfear R, Doughty P, Keogh JS (2014) The biogeographical boundaries of

northern Australia: evidence from ecological niche models and a multi-locus

phylogeney of Uperoleia toadlets (Anura: Myobatrachidae). Journal of

Biogeography 41, 659-672.

Oliver PM, Laver RJ, Melville J, Doughty P (2014) A new species of Velvet Gecko

(Oedura: Diplodactylidae) from the limestone ranges of the southern Kimberley,

Western Australia. Zootaxa 3873, 49-61.

Pepper M, Fujita MK, Moritz C, Keogh JS (2011) Palaeoclimate change drove

diversification among isolated mountain refugia in the Australian arid zone.

Molecular Ecology 20, 1529-1545.

Potter S, Eldridge MDB, Taggart DA, Cooper SJB (2012) Multiple biogeographical

barriers identified across the monsoon tropics of northern Australia:

phylogeographic analysis of the brachyotis group of rock-wallabies. Molecular

Ecology 21, 2254-2269.

Smith KL, Harmon LJ, Shoo LP, Melville J (2011) Evidence of constrained phenotypic

evolution in a cryptic species complex of agamid lizards. Evolution 65, 976-992.

236

Minerva Access is the Institutional Repository of The University of Melbourne

Author/s: Laver, Rebecca Jan

Title: Comparative phylogeography and diversity of Australian Monsoonal Tropics lizards

Date: 2016

Persistent Link: http://hdl.handle.net/11343/123383

File Description: PhD Thesis

Terms and Conditions: Terms and Conditions: Copyright in works deposited in Minerva Access is retained by the copyright owner. The work may not be altered without permission from the copyright owner. Readers may only download, print and save electronic copies of whole works for their own personal non-commercial use. Any use that exceeds these limits requires permission from the copyright owner. Attribution is essential when quoting or paraphrasing from these works.