1 Biogeography and speciation of southwestern Australian frogs Danielle L. Edwards B. Env. Sc. (Hons) Supervisor: Prof. J. Dale Roberts This thesis is presented for the degree of Doctor of Philosophy of The University of Western Australia, School of Animal Biology 2007
222
Embed
Biogeography and speciation of southwestern Australian frogs · Biogeography and speciation of a direct developing frog from the coastal arid zone of Western Australia. In Review
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Biogeography andspeciation of
southwestern Australianfrogs
Danielle L. EdwardsB. Env. Sc. (Hons)
Supervisor: Prof. J. Dale Roberts
This thesis is presented for the degree of Doctor of Philosophy of The University ofWestern Australia, School of Animal Biology
2007
2
3
Thesis Declaration
This thesis contains published work and work prepared for publication, some of whichhas been co-authored. The bibliographic details of the works and where they appear inthe thesis are set out below.
Edwards, DL. Biogeography and speciation of a direct developing frog from thecoastal arid zone of Western Australia. In Review with Molecular Phylogenetics andEvolution. (Data Chapter 1).
Edwards, DL, Roberts, JD and Keogh, SK (In Press). Impact of Plio-Pleistocene aridcycling on the population history of a southwestern Australian frog. Molecular Ecology.(Data Chapter 2).
(This work was primarily conducted by DLE (~90%), JDR provided assistance with projectdesign, and editing and advice on field collection (~5%), JSK provided access to his molecularlab, assistance with editing and advice on analysis techniques (~5%)).
Edwards, DL, Roberts, JD and Keogh, SK. Climatic fluctuations shape thephylogeography of a mesic adapted direct developing frog from the southwesternAustralian biodiversity hotpot. In Prep for Journal of Biogeography. (Data Chapter 3).
(This work was primarily conducted by DLE (~90%), JDR provided assistance with projectdesign, and editing and advice on field collection (~5%), JSK provided access to his molecularlab, assistance with editing and advice on analysis techniques (~5%)).
The fourth data chapter is also to be published; the manuscript is still in preparation.
Signatures…………………………………………………………………………………
4
5
Summary
Southwestern Australia is a global biodiversity hotspot. The region contains a high
number of endemic species, ranging from Gondwanan relicts to more recently evolved
plant and animal species. Biogeographic models developed primarily for plants suggest
a prominent role of Quaternary climatic fluctuations in the rampant speciation of
endemic plants. Those models were not based on explicit spatial analysis of genetic
structure, did not estimate divergence dates and may be a poor predictor of patterns in
endemic vertebrates. Myobatrachid frogs have featured heavily in the limited
investigations of the biogeography of the regions fauna. Myobatrachid frogs are diverse
in southwestern Australia, and while we know they have speciated in situ, we know
little about the temporal and spatial patterning of speciation events.
In order to gain insight into the biogeographic history and potential speciation patterns
of Myobatrachid frogs in the southwest I conducted a comparative phylogeography of
four frog species spanning three life history strategies. I aimed to:
1) assess the biogeographic history of individual species,
2) determine where patterns of regional diversity exist using a comparative framework,
3) determine whether congruent patterns across species enable the development of
explicit biogeographic hypotheses for frogs, and
4) compare patterns of diversity in plants with the models I developed for frogs.
I conducted fine-scale intraspecific phylogeographies on four species. Species were
selected to cover the major biogeographic regions within the southwest, a range of
development modes and potential sensitivities to climatic and associated rainfall
changes.
1) Arenophryne rotunda – a direct developing species endemic to the semi-arid Shark
Bay region covered the plant diversity hotspot on the northwestern coast of
southwestern Australia,
2) Crinia georgiana – an aquatic breeder reliant on predictable seasonal rainfall covered
the forest system (HRZ) into the hotspot region on the southeast coast (SECZ),
3) Metacrina nichollsi – a direct developer endemic to the wettest part of the forest
system, overlapped with C. georgiana and provided a comparison with the habitat
specialists from the Geocrinia rosea species complex, and
6
4) Geocrinia leai – a terrestrial ovipositor with an obligate aquatic and free-swimming
tadpole whose distribution overlaps with that of C. georgiana, M. nichollsi and with the
habitat specialists from the Geocrinia rosea species complex.
Deep intraspecific divergences and marked phylogeographic structure were detected in
all four species with many congruent patterns across species.
Arenophryne rotunda: a deep north-south division was associated with the Late
Miocene uplift of the Victoria Plateau. There was an additional split within the southern
lineage linked to the final incision of the Murchison Gorge during the Pliocene.
Phylogeographic structure within each lineage was shaped by coastal landscape
development and sea level change.
Crinia georgiana: two lineages were identified which largely corresponded to the High
Rainfall and Southeast Coastal Provinces defined by Hopper and Gioia (2004). Lineage
divergence and within lineage phylogeographic structure was been shaped by
Quaternary climate and associated rainfall oscillations.
Metacrinia nichollsi: late Miocene to present climate changes are linked with
divergence and phylogeographic processes in this species. A lineage corresponding to
the isolated Stirling Ranges populations is identified. A second lineage covers the
majority of the remaining range, and shows evidence of recent range expansion. The
third lineage has a disjunct distribution across the southern coast with strong catchment
based patterns of genetic structure.
Geocrinia leai: deep divergences, coincident with late Miocene arid onset, divide this
species into western and southeast coastal lineages, with a third only found within the
Shannon-Gardner River catchments. Phylogeographic history within each lineage has
been shaped by climatic fluctuations from the Pliocene through to the present.
Arenophryne shows the first evidence of geological activity in speciation of a Shark
Bay endemic. Divergence patterns between the High Rainfall and Southeast Coastal
Provinces within C. georgiana are consistent with patterns between Litoria moorei and
L. cyclorhynchus and plant biogeographic regions. Subdivision between drainage
systems along the southern coast (in M. nichollsi, G. leai and the G. rosea species
complex) reflect the relative importance of distinct catchments as refuges during arid
maxima, similarly the northern Darling Escarpment is identified as a potential refugium
(C. georgiana and G. leai).
7
Divergences in Myobatrachid frogs are far older than those inferred for plants with the
late Miocene apparently an important time for speciation of southwestern frogs.
Speciation of Myobatrachids broadly relates to the onset of aridity in Australia in the
late Miocene, with the exception of earlier/contemporaneous geological activity in
Arenophryne. The origins of subsequent intraspecific phylogeographic structure are
coincident with subsequent climatic fluctuations and correlated landscape evolution.
Divergence within frogs in the forest system may be far older than the Pleistocene
models developed for plants because of the heavy reliance on wet systems by relictual
frog species persisting in the southwestern corner of Australia.
8
9
Table of Contents
Summary ……………………………………………………………………….5Table of Contents ……………………………………………………………… 9List of Tables …………………………………………………….………...12List of Figures ………………………………………………………………13Acknowledgements ………………………………………………………………15
Chapter 1: General Introduction ………………………………………………19
1.1 Phylogeography, comparative phylogeography and conservationapplications ………………………………………………………19
1.2 Genetic markers used in phylogeography ………………………201.3 Measures of population structure to infer patterns of gene flow ………211.4 Coalescent theory and Nested Clade Phylogeographic Analysis: A break
through in analytical phylogeography ………………………………211.5 A global view of phylogeography ………………………………221.6 An Australian view of phylogeography ………………………………231.7 Southwestern Australia as a biodiversity hotspot ………………231.8 Speciation and biogeographic hypotheses for southwestern
Australian frogs ………………………………………………………261.9 Climatic and geological history of southwestern Australia ………271.10 Comparative phylogeography of southwestern Australian
frogs ……………………………………………………………....281.11 Study species: Selection rationale and life history ………………29
2.5 Discussion ………………………………………………………532.5.1 Biogeography and speciation in Arenophryne ………………532.5.2 Phylogeography and population structure – Southern Lineage ………………………………………55
10
2.5.3 Phylogeography and population structure – Northern Lineage ………………………………………552.5.4 Conclusions ………………………………………………56
Chapter 3: Phylogeography of Crinia georgiana (The Quacking Frog) ………61
4.5 Discussion ……………………………………………………..1064.5.1 Isolation of the Stirling Ranges Populations ……………..1074.5.2 Biogeography within the southwestern clades of
M. nichollsi ……………………………………………..1084.5.3 Conclusions ……………………………………………..111
Chapter 5: Phylogeography of Geocrinia leai (Lea’s Frog) ……………………..115
5.5 Discussion ……………………………………………………..1315.5.1 Broader phylogenetic pattern within G. leai ……………..1345.5.2 Phylogeographic pattern within G. leai ……………..1355.5.3 Geocrinia leai and the biogeography of southwestern
Australia ……………………………………………..1365.5.4 Conclusions ……………………………………………..137
Chapter 6: General Discussion and Future Directions ……………………..141
6.1 The late Miocene as a time of speciation for southwesternAustralian frogs ……………………………………………………..141
6.2 Plio-Pleistocene climatic fluctuations shape the biogeographyof southwestern Australian frogs ……………………………..145
6.3 Catchments and upland forests as refuges for frogs duringaridity ……………………………………………………………..149
6.4 Biogeography within southwestern Australia ……………………..1506.5 Conservation and Climate – what to expect for the future ……..1516.6 Future Directions ……………………………………………………..151
2.1 Arenophryne rotunda sample sites, sizes and locations ………………………412.2 Biogeographical inferences for A. rotunda ………………………………492.3 Summary of population genetic analyses on A. rotunda lineages ………52
Chapter 3:
3.1 Crinia georgiana sample sites, sizes and locations ………………………653.2 Crinia georgiana ND2 haplotypes ………………………………………723.3 Biogeographical inferences for C. georgiana ………………………………753.4 Summary of population genetic analyses on C. georgiana ………………78
Chapter 4:
4.1 Metacrinia nichollsi sample sites, sizes and locations ………………………934.2 Metacrinia nichollsi ND2 haplotypes ……………………………………..1004.3 Biogeographical inferences for M. nichollsi from NCA ……………………..1034.4 Summary of population genetic analyses on M. nichollsi lineages ……..105
Chapter 5:
5.1 Geocrinia leai sample sites, sizes and locations ……………………………..1195.2 Biogeographical inferences for G. leai from NCA ……………………..1305.3 Summary of population genetic analyses on G. leai lineages ……………..131
13
List of Figures
Chapter 1:
1.1 Southwestern Australian biogeographical provinces ………………………241.2 Distribution of Arenophryne rotunda ………………………………………291.3 Distribution of Crinia georgiana ………………………………………………301.4 Distribution of Metacrinia nichollsi ………………………………………311.5 Distribution of Geocrinia leai ………………………………………………32
Chapter 2:
2.1 Map of A. rotunda sampling locations and phylogenetic results ………………422.2 Haplotype network constructed for the northern A. rotunda lineage ………………………………………………………………………502.3 Haplotype network constructed for the southern A. rotunda lineage ………………………………………………………………………512.4 Biogeographic hypotheses relating to the history of A. rotunda ………………53
Chapter 3:
3.1 Map of C. georgiana sampling locations ………………………………………663.2 Phylogenetic results: phylogram and distribution of major clades within C. georgiana ………………………………………………………733.3 Haplotype network constructed for C. georgiana ………………………763.4 Biogeographic hypotheses relating to the history of C. georgiana ………80
Chapter 4:
4.1 Map of M. nichollsi sampling locations ………………………………………944.2 Phylogenetic results: phylogram and distribution of major clades within M. nichollsi ……………………………………………………..1014.3 Haplotype networks constructed for M. nichollsi ……………………..1044.4 Biogeographic hypotheses relating to the history of M. nichollsi ……..106
Chapter 5:
5.1 Map of G. leai sampling locations ……………………………………..1195.2 Phylogenetic results: phylogram and distribution of major clades within G. leai ……………………………………………………………..1255.3 Haplotype network constructed for a portion of the western G. leai lineage ……………………………………………………………………..1275.4 Haplotype network constructed for remainder of the western G. leai lineage ……………………………………………………………………..1285.5 Overall nesting design for the separate G. leai western lineage haplotypes ……………………………………………………………..1295.6 Biogeographic hypotheses relating to the history of G. leai ……………..133
14
Chapter 6:
6.1 Late Miocene divergence of Arenophryne rotunda lineages ……………...1416.2 Late Miocene divergence of Metacrinia nichollsi lineages ……………...1436.3 Late Miocene divergence of Geocrinia leai lineages ……………………...1446.4 Divergence of Crinia georgiana lineages during the Plio-Pleistocene ……...1466.5 Divergence of Geocrinia leai clades during the Plio-Pleistocene ……...1476.6 Distribution of the Geocrinia rosea species complex ……………………...149
15
Acknowledgements
There are many people, without whom I would never have made it to where I am now.
Dale, you were one of the best supervisors I could ever have asked for. You gave meguidance, overwhelming support, a mentor to look up to and a great colleague to sitaround the joint and chat to over a coffee talking wild and crazy shit about frogs. Yourunwavering faith kept me going and got me out of many the low times. I will alwaysremember my PhD experience with fondness, I didn’t crack no matter what happenedand I largely have you to thank for that.
Scott, even though never official you have been there as a supervisor for me when Ineeded you. I thank you for all the support you have given me throughout the writingprocess and for making me apart of your lab the whole time I was at ANU. You havebeen integral in helping me get the confidence to finally publish some of my work afterso long and get the damn thesis finished.
My Family, well what can I say. Thanks for trying to understand, putting up with my‘occasional’ moods and infrequent visits. You have always provided me with lovingsupport and advice, despite not having a clue what I was doing. I love you all. I shouldalso pay homage to my own matrilineal heritage and environmental conditioning Iguess. I come from a long line of strong, independent women and a family where“tellin’ it like it is barbs ‘n all” is a way of life. I don’t think I would have got throughthings like near death car accidents in the field and debilitating illness to hand this thingin if I hadn’t acquired those qualities from my loving family.
Jane – Thanks for believing in me (that goes for Di and Martin too), I enjoyed my timeat Museum Victoria immensely and hope that we get the opportunity to do lots morework together in future.
There are many more people at both UWA and at the ANU and in Canberra in generalthat have provided drinking partners, councillors and friends.
UWA Crew
Martin – always entertaining, if at times annoying. I don’t think I will ever be the sameafter trying to go shot for shot with you, thanks for being a great mate; Nèe – you arethe biggest bogan I know, make a fine drinking partner and I love you dearly; Vixen –my froggy sister, we should definitely have more jamming in the future, love your way;Kerry – you were always lovely and so friendly; the rest of Happy Hour – thank you.
ANU (and wider Canberra crew)
Dave Rowell – Thanks for reading so many chapter drafts for me and being aninspirational academic; Stu and Jess – you guys are awesome and some of the bestfriends I have ever made, and no body knows how to bring out Drunken Dan like youdo; Matt – I thank you for always challenging me and making a fine coffee partner totalk shit with…when you have finished I expect to celebrate over a beer or three withyou; Mitzy – your bubbly nature is always a pick me up, even if a little loud. Kate &Mel – What can I say!!! Always good for sound advice, sisterhood and crankin jammin’
16
partners; Suzie (Q) – what would I have done without those office beers???; Si – if onlyI looked that good in a skirt and played cranking rock ‘n roll like you do!; Miss Kristy –you have been a wonderful supportive friend; Chris – I would not have been able to dohalf the things in the lab I did, thankyou for being a wonderful teacher and showing methe world of molecular genetics; thanks to all the miscellaneous ANU Happy Hour folk,and finally thanks to the Canberra punks (most of all Christie, Klaus, Laura, Kath,Katie, Ilonka) – Thanks for being mates, drinking partners and opening up a world ofextra-curricular fun for me to enjoy and use as a distraction from things like a thesis.
For funding I would like to thank:
Australian Federal Government – Agriculture, Forestry and Fisheries Australia (AFFA)Awards for Young Scientists 2002, The Western Australian State Government -Department of Conservation and Land Management (C.A.L.M) and The School ofAnimal Biology, The University of Western Australia for funding to DE. Samplecollections and tissue collection procedures were approved by The Department ofConservation and Land Management, Western Australia (Permit No.’s CE000405;SF004276; SF004246) and The University of Western Australia Animal EthicsCommittee (Approval No. 03/100/241).
Acknowledgements for specific chapters:
Chapter 2:Jane Melville, Dave Rowell, Mark True (C.A.L.M - Denham), The Kalbarri C.A.L.Mstaff, The Wardle Family (Dirk Hartog Island), Pam and Paul Dickinson (Steep Point),Lisa Myers, Dr Jane Prince (UWA), Lawrie Poole and Bryan Cane (Shark Bay Salt) forhelp with field collections.
Chapter 3:Beckie Symula, Rachael Heaton and Martin Dziminski for assistance with fieldwork.Thanks also to Ian Scott, Mike Double, Mark Blacket and Michael Kearney for adviceon data analysis, and Dave Rowell for comments on the manuscript. Thanks to PaulDoughty and Brad Maryan from The Western Australian Museum for access to tissues.
Chapter 4:Thanks to Dr. Barbara York Main and Prof. Bert Main for much useful discussion onthe biogeographic history of the southwest and the biology of Metacrinia. Much thanksalso to Jim Lane (C.A.L.M Bunbury), Karlene Bain (C.A.L.M Walpole) and all theC.A.L.M Walpole Staff, Mirelle Edwards, and botany Prof. John Pate for assistancewith field.
Chapter 5:Martin Dziminski and Beckie Symula for assistance with field collections.
17
The story begins…..
18
19
Chapter 1:
Introduction
1.1 Phylogeography, comparative phylogeography and conservation
applications
“Phylogeography”, a term originally coined by Avise et al. (1987), describes the
genealogical relationships between populations within a species in relation to the
landscape. Phylogeographic studies can provide a means to understand the evolutionary
processes shaping the history of species and regions, and a framework for conserving
those evolutionary processes. Phylogeographic studies have also been important in
identifying cryptic species, particularly in continental and morphologically variable taxa
(Riddle et al. 2000; Arbogast, Kenagy 2001), and they have application in species
defended the use of NCPA for phylogeographic data, and with the introduction of
various cross validation techniques (Masta et al. 2003; Templeton 2004) many of these
concerns have been addressed. NCPA remains an important technique in
phylogeography, particularly where poor prior knowledge of the history of species and
regions limits the development of a priori hypotheses, inherent in the development of a
priori hypothesis testing under a statistical phylogeography framework (Knowles 2004).
1.5 A global view of phylogeography
Most phylogeographic studies have focussed on northern hemisphere systems, which
have been heavily affected by Pleistocene glaciation events followed by post-glacial
range expansion. The sheer number of these studies has allowed the synthesis of many
comparative phylogeographic datasets covering entire taxon groups (Zink 1996; Davis,
Shaw 2001; Weir, Schluter 2004; Smith et al. 2005; Macey et al. 2006) and
biogeographic regions, such as Europe, the Americas and various islands (Riddle 1996;
Hewitt 2000; Zink 2002; Lindell et al. 2006; Soltis et al. 2006; Yoder, Nowack 2006).
These studies, combined with the vast body of knowledge on geological and climatic
history, have led to the development of clear biogeographic hypotheses regarding the
history of biota across these specific bioregions. The majority of these studies concern
the effects of Northern Hemisphere glacial cycles. Australia has experienced little or no
23
glaciation but a pattern of arid and mesic conditions coincident with Northern
Hemisphere glacial and interglacial cycles (Galloway, Kemp 1981). Of the
biogeographical studies conducted within Australia, almost all have been conducted in
eastern Australia, with particular focus on the Wet Tropics region (Schneider et al.
1998) and some studies have dealt with continental biogeography of specific taxon
groups (Unmack 2001; Jennings et al. 2003; Baker et al. 2004; Crisp et al. 2004;
Munasinghe et al. 2004; Wuster et al. 2005).
1.6 An Australian view of phylogeography
Many single species phylogeographies have focussed on broadly distributed species that
cover the majority of eastern Australia (James, Moritz 2000; Schäuble, Moritz 2001;
Donnellan, Mahony 2004; Garrick et al. 2004; Wong et al. 2004; Cook et al. 2006;
Sunnucks et al. 2006) and the Wet Tropics (Hughes et al. 1996; McGuigan et al. 1998;
Pope et al. 2000; Hurwood, Hughes 2001; Stuart-Fox et al. 2001; Carini, Hughes 2006;
Dolman, Moritz 2006; Ozeki et al. 2007), with some further studies covering the arid
zone (Strasburg, Kearney 2005; Kearney et al. 2006; Pepper et al. 2006). Studies
focussing on eastern Australia, and the Wet Tropics in particular, have allowed the
development of clear biogeographic hypotheses (Schneider et al. 1998; Moritz et al.
2001) in line with models of climate induced habitat fluctuations (Moussalli et al.
2005). The Wet Tropics region is considered an important region of biodiversity within
Australia (Cincotta et al. 2000; Myers et al. 2000). Southwestern Australia on the other
hand has been described as one of the world’s biodiversity hotspots (Myers et al. 2000),
yet remains severely understudied, with only a rudimentary understanding of the
processes generating its vast diversity.
1.7 Southwestern Australia as a biodiversity hotspotSouthwestern Australia is an iconic region known for its extreme endemicity, high
species diversity and its threatened environments (Cincotta et al. 2000; Myers et al.
2000). According to Hopper (1979) and Hopper & Gioia (2004) the region contains
three rainfall-vegetation zones (Figure 1.1).
1) High Rainfall Zone (800-1400mm/yr) - encompasses the jarrah, marri and karri
forests and woodlands also identified as a distinct biogeographic province of the
same name
24
2) Transitional-rainfall zone 300-800mm/yr) - contains woodland, mallee and
heathland, and covers the Transitional and, Southeast Coastal, biogeographic
provinces:
3) Arid zone (<300mm/yr) – consists of Eucalypt woodland, shrubland and hummock
grassland (Hopper 1979) also identified as a distinct biogeographic province of
the same name.
Figure 1.1: Southwestern Australian biogeographical provinces, as determined by theendemic flora, and rainfall levels. The High Rainfall Province encompasses the forestsystem with rainfall between 800-1400mm/yr. The Transitional Rainfall Zonecollectively includes the Transitional and Southeast Coastal Provinces with rainfallbetween 300-800mm/yr. The Arid Zone is where rainfall falls below 300mm/yr.Adapted from Hopper & Goia (2004).
Southwestern Australia is widely recognised for its extreme diversity and high level of
endemism of plant species (Hopper 1979; Hopper, Gioia 2004). Less known, but
equally spectacular is the high level of faunal diversity, particularly invertebrates (Main
1996), mammals, reptiles (Hopper et al. 1996) and amphibians (Roberts 1993; Slatyer
25
et al. 2007). The region has long been a biogeographical enigma. It lacks obvious
historical geographical barriers arising from events such as glaciation and mountain
building, events that are common in many vicariant speciation models (Hopper, Gioia
2004). Botanical studies have sought to clarify the processes leading to the high levels
of both endemicity and diversity within the southwestern Australian flora.
The late Tertiary and Quaternary have been identified as periods of intense speciation in
southwestern Australian flora (Hopper 1979; Hopper, Gioia 2004). Climatic
fluctuations led to landscape evolution, through differential soil erosional/depositional
histories and coastal dune and sandplain development, which contributed to the high
levels of diversity and endemicity observed in southwestern flora. Extreme levels of
plant diversity are found particularly in the northwestern and southeastern coastal areas
of the region, areas that are more complex topographically than the wider southwestern
forest system (high rainfall zone - HRZ) (Hopper, Gioia 2004). The processes acting on
terrestrial vertebrates might be quite different from those involved in the speciation of
southwestern Australian plants: e.g. range sizes are often higher, and habitat
specializations less marked, so the potential for specialization and isolation on novel
soil types is lower. Comparatively little work has been done to investigate the processes
involved in generating diversity both within and between species of endemic
southwestern Australian fauna, and in the forest system in general.
A research focus on the more climatically transitional areas has created the view that the
forest system (high rainfall province) is comparatively species depauperate (Hopper,
Gioia 2004). Others have suggested that diversity of the relictual forest flora may be
fluctuations also were associated with eustatically controlled sea level
transgression/regression cycles, leading to massive changes in the occurrence and area
of coastal sandplain and sand-dune habitats (Hocking et al. 1987; Mory et al. 2003).
Dune building episodes occurred during arid (glacial) cycles intersected with
39
transgressive episodes during interglacial wet periods (240,000yrs ago and 120,000-
130,000yrs ago)(Van de Graaff et al. 1980; Hocking et al. 1987). The most recent
transgressive cycle occurred at the height of the last interglacial and resulted in the final
flooding of Shark Bay, beginning ~10,000yrs ago and reaching its peak ~6,000yrs ago
(Butcher et al. 1984; Hocking et al. 1987).
Pleistocene coastal landscape evolution, driven by climatic fluctuations, has been used
to explain diversity and recent speciation in a number of Shark Bay biota (Storr, Harold
1980; Hopper, Gioia 2004; Rabosky et al. 2004). However older and more fundamental
geological evolution also may play a part in shaping current genetic architecture,
particularly fossorial anurans and reptiles common in the area. While southwestern
Australia in general is considered to have been geologically stable since the Tertiary
(Hopper 1979; Hopper, Gioia 2004), coastal areas of the Shark Bay region have
undergone a complex series of geological processes leading to the evolution of the
current landscape (Van de Graaff et al. 1980; Hocking et al. 1982; Butcher et al. 1984;
Hocking et al. 1987; Mory et al. 2003). After a period of long stability reactivation of
pre-existing faults in the area began in the Miocene and a period of tectonic instability
continued through to the Pleistocene. This tectonic instability has been linked to the
formation and dissection of the Victoria Plateau, the incision of the Murchison Gorge
(Hocking et al. 1982; Hocking et al. 1987), general uplift (Haig, Mory 2003; Mory et
al. 2003) and the gentle folding of anticlines, which are now a controlling factor in
shaping the coastline of the Shark Bay area (Hocking et al. 1987).
Arenophryne rotunda, a highly arid-adapted and fossorial direct-developing frog
endemic to Shark Bay, provides an ideal model species to investigate directly the
influences of both geology and climate change/sea level fluctuations on Shark Bay
fauna. While nothing is known about the history of this species, given the relative age
of the Arenophryne lineage compared to sister taxa, Myobatrachus gouldii and
Metacrinia nichollsi (Read et al. 2001), older climatic and geological events may have
impacted the current genetic architecture of A. rotunda. The distribution of A. rotunda
crosses many significant geological entities within the Shark Bay region, namely the
northern border of the Victoria Plateau and the Murchison Gorge (Fig. 1). The species
also occupies much of the coastal Shark Bay region and Dirk Hartog Island, which
allows for an assessment of the impacts of coastal landscape evolution and the flooding
of Shark Bay. Additionally, given the fossorial habit of the species and its preference
40
for sandplain and dune habitats, Pleistocene dune building episodes may well have
influenced population structure within A. rotunda. Also. I compiled sequence data from
an 1154bp fragment of the mitochondrial gene ND2 from 47 individuals, across 19
localities and covering the whole known range of the species. This study provides the
first comprehensive dataset specific to the Shark Bay region and a comparison for
biogeographic hypotheses developed for plants and herpetofauna of the Shark Bay
region.
2.3 Materials and Methods
2.3.1 Tissue samples
Arenophryne rotunda is a small, fossorial, direct-developing frog endemic to the
southwest (Roberts 1990), from Shark Bay south to Kalbarri – inset Figure 2.1. It
occupies sand dune and sandplain habitats, encompassing several different substrate
types and crossing several climate zones. Its distribution is thought to be continuous
across its range, with some of the highest levels of anuran abundance ever recorded
(Roberts 1985). Forty-seven individuals were sampled (toe-clips) from 13 sites across
the entire species distribution, with 3-4 animals per site (Figure 2.1, Table 2.1). Samples
from EL1, ZU2 and ZU5 were taken from the WA Museum Tissue Collection, WAM
collection numbers 122520-122522, 123493-123495 and 123523-123526 respectively.
Outgroup sequences used in the study were: Metacrinia nichollsi (34°59´38˝
116°39´22˝) and Myobatrachus gouldii (30°01´57˝ 115°49´06˝).
41
Table 2.1: Arenophryne rotunda sampling site location names, abbreviations,sample sizes and GPS co-ordinates in degrees, minutes, seconds. All points are ingeodetic WGS84.
Figure 2.1: Maximum likelihood phylogram of 31 Arenophryne rotunda ND2haplotypes, showing two major lineages, with Metacrinia nichollsi and Myobatrachusgouldii as outgroups. Haplotype numbers are displayed with the sample site fromwhich they came and their frequency in brackets. Bootstrap values for clades above 70are represented by * and were calculated from 100 replicates. The TrN + I model ofDNA evolution was enforced in maximum likelihood analyses as suggested by AICtests in Model Test 3.7. Map of the mid-western Australian coast is shown with mapof the Australian continent inset and with shaded areas representing the distribution ofboth the northern and southern lineages. Tissue collection locations [•] for theArenophryne rotunda phylogeographic study cover the entire known distribution ofthe species.
2.3.2 Molecular genetic methods
Template DNA was extracted from toe samples using a modified CTAB method,
suspended in TE and stored at 0°C. Targeted DNA was amplified using a touch-down
PCR profile (94°C for 5min - 1×; 94°C for 30sec, 70-45°C (decreasing in 5°C
increments) for 20sec, 72°C for 90sec – each of these cycles were repeated 2× for each
extension temperature; 94°C for 30sec, 40°C for 30sec, 72°C for 45sec - 40×; 72°C for
4min - 1× ; 4°C held. Primers used to amplify N D 2 were L4221 (5'-
AAGGRCCTCCTTGATAGGGA-3', modified from (Macey et al. 1998))) & tRNA-
Asn (5'-CTAAAATRTTRCGGGATCGAGGCC-3', (Read et al. 2001))) or Myo tRNA-
43
trp (5'-GGGGTAGYATHCCACAAGTC-3', this paper). Targeted fragments were
amplified in 40µl reactions comprising ~100ng template DNA, 4µl of 10× reaction
buffer, 3 mM MgCl2, 0.5 mM dNTPs, 10 pmol of each primer and 2 units of Platinum
populations north of the Murchison Gorge (Clades 2.1 and 4.1) suggesting limited
dispersal amongst these populations. There is also an inference of allopatric
fragmentation between populations either side of the Murchison Gorge, between the
Clades 4.1 & 4.2, when both networks are joined at the final nesting level.
49
Table 2.2: Biogeographical inferences for nested Arenophryne rotunda clades fromboth the northern and southern lineages with significant phylogeographic structure,specified by a χ2 nested contingency test. P-values are calculated from 10000 randompermutations and are considered significant if permuted expected χ2 values are greaterthan or equal to the observed.
AF – Allopatric Fragmentation; PF – Past Fragmentation; GRE – Gradual Range Expansion; RGF –Restricted Gene Flow; IBD – Isolation by Distance; w/ - with.
Nested χ2 Permuted Inferred Clade P -value Process
1.2 0.024 Int/tip status not determined N/A2.1 <0.001 1-2-3-4-9 AF3.1 <0.001 1-19-20-2-11-12-13-LDC PF w/ GRE
Figure 2.2: Haplotype network for 17 Arenophryne rotunda ND2 haplotypes (includingsite references) from the northern lineage created in TCS 1.21. Each line represents asingle mutational change. Ellipse size is proportional to haplotype frequency with smallopen circles representing missing haplotypes and the square representing the ancestralhaplotype as inferred by TCS using outgroup weights. All connections, up to 11 steps,are within the 95% confidence limits of a parsimonious connection. Clades are nestedaccording to the rules outlined in (Templeton et al. 1987; Crandall 1994; Templeton etal. 1995).
51
Figure 2.3: Haplotype network for 14 Arenophryne rotunda ND2 haplotypes (includingsite references) from the southern lineage created in TCS 1.21. Each line represents asingle mutational change. Ellipse size is proportional to haplotype frequency with smallopen circles representing missing haplotypes and the square representing the ancestralhaplotype as inferred by TCS using outgroup weights. Connections, up to 17 steps, arewithin the 95% confidence limits of a parsimonious connection. A connection betweenclades 4.1 and 4.2 are not within the 95% confidence limits, but are joined by 22mutational steps. Clades are nested according to the rules outlined in (Templeton et al.1987; Crandall 1994; Templeton et al. 1995).
2.4.3 Population genetic analysis
Table 2.3 is a summary of population genetic analyses carried out on the two main
lineages found within the A. rotunda dataset. AMOVA results from the northern A.
rotunda lineage show very little genetic variation (4%) accounted for by the separation
of Dirk Hartog Island populations from the mainland and the majority of genetic
variability distributed amongst populations (68%) with high levels of population
structure (øPT=0.717;P≤0.001). High levels of population divergence amongst
populations of the northern lineage are also indicated by the highly significant Snn
52
value (0.887;P≤0.001). Within the southern A. rotunda lineage most genetic variation is
accounted for by the separation of populations either side of the Murchison Gorge
(78%). Snn (0.648 all populations; 0.548 populations north of the Murchison Gorge)
results also suggest little population divergence, verging on panmixia, within the
southern lineage and particularly in the populations north of the Murchison Gorge.
Table 2.3: Evidence for high levels of population genetic structure and differentiationwithin both the northern and southern Arenophryne lineages. Analyses assessed thepartitioning of variation attributed to the flooding of Shark Bay isolating populations onDirk Hartog Island from mainland populations (Northern Lineage), and the incision ofMurchison Gorge isolating populations to the north and south of this barrier (SouthernLineage). Variation distributed amongst populations within each of these regions andover the whole range of each lineage was also assessed, as was the variation due toindividuals within each population. Hudson’s ‘nearest neighbour’ statistic (Snn)measures the population divergence within each lineage as a whole, and between thepopulations north of the Murchison Gorge only within the southern lineage. P-valueswere calculated via 1000 permutations.
Source df SS MS Est. Var. % Stat ValueNorth Vs South Murchison 1 70.567 70.567 10.411 78% φRT 0.784** Among Pops./Regions 3 16.071 5.357 1.004 8% φPR 0.350* Indiv./Within Pops. 13 24.250 1.865 1.865 14% φPT 0.860***Total Southern Lineage Snn 0.648***North of Murchison Pops. Snn 0.548**
Northern Lineage Population Genetics Analysis
Southern Lineage Population Genetics Analysis
53
2.5 Discussion
I have inferred a molecular phylogeny for a fossorial frog that sheds light on the factors
that have generated population level diversity in this taxon. A major divergence event
has occurred between northern and southern lineages of Arenophryne rotunda (Figures
2.1 & 2.4) approximately ~5.63MYA (±410,000yrs), or in the Late Miocene period.
Within the southern A. rotunda lineage (Figures 2.1, 2.3 & 2.4) divergence of clades has
occurred across the Murchison Gorge ~2.05MYA (± 424,000yrs), or the Late Pliocene
period. Here I first consider the biogeography and speciation of A. rotunda at a broad
level, and then turn to each of the two lineages, with particular reference to examining
how geological and climatic history have influenced current genetic structure.
Figure 2.4: Biogeographic hypotheses regarding the history of the northern andsouthern lineages within Arenophryne rotunda. Hypotheses are synthesized by acombination of phylogenetic, phylogeographic and populations genetic analysistechniques, which were interpreted with the aid of the known geological and climatichistory of the region. PF – Past Fragmentation; IBD – Isolation by Distance; • - sampledpopulations.
2.5.1 Biogeography and speciation in Arenophryne
The major genetic break between the northern and southern Arenophryne lineages is
consistent with the genetic differences observed between sister species in other groups
54
within the Myobatrachidae (Morgan et al.; Read et al. 2001). As a result of this work
clear morphological differences have been measured and a new species description
corresponding to the southern mitochondrial lineage is forthcoming (Doughty et al., in
prep). There has been much discussion of the importance of sea level and climate
fluctuations, predominately occurring during the Plio-Pleistocene, acting as speciation
mechanisms within herpetofauna in the Shark Bay and wider Carnarvon Basin region.
Fluctuating climates and sea levels seem plausible explanations for vicariance in many
of the species with separate populations that have split in the northern Carnarvon Basin,
such as Rankinia adelaidensis (Melville and Doughty, submitted ms) and several other
skink and gecko species (Storr, Harold 1978; 1980). However, divergence estimates
suggest that the major split within Arenophryne predates many of the Plio-Pleistocene
sea level fluctuations resulting in coastal dune evolution in the region (Hocking et al.
1987). While molecular clock estimates are fraught with difficulties (Rambaut,
Bromham 1998), the date obtained in this instance provides an estimate that tightly
links with known climatic and geological changes.
The formation of the Victoria Plateau, in combination with sweeping aridity, is likely to
have led to the Late Miocene divergence between the northern and southern A. rotunda
lineages. Tectonic instability resulted in the reactivation of pre-existing faults and the
uplift and formation of the Victoria Plateau, with in the Kalbarri region the Victoria
Plateau uplifted by as much as 60m (Haig, Mory 2003). The northern border of the
Victoria Plateau roughly corresponds to the position of the genetic break between the
two Arenophryne lineages, and without the presence of the thick coastal sand deposits
of the Edel group (not formed until the Plio-Pleistocene (Hocking et al. 1987)) an
alternative avenue for dispersal was not available. The role of tectonic instability in
inducing vicariance, particularly in fossorial psammophillic species, has largely been
ignored in treatments of the region’s diversity to date in favour of hypotheses relating to
fluctuating sea levels resulting in coastal dune and sandplain development (Hopper,
Gioia 2004; Rabosky et al. 2004); Melville and Doughty, submitted ms).
I hypothesize a scenario that takes into account both geological activity and climatic
shifts (Figure 2.5): I suggest that Arenophryne, a formerly widespread taxon, was split
by geological activity disrupting effective dispersal through sand habitat, a break that
was compounded and reinforced by range contraction westwards with increasingly arid
conditions (Figure 2.4). The onset of aridity also intensified in the Late Miocene,
55
causing a change from a subtropical climate to one that oscillated between arid and
temperate conditions (Dodson, Macphail 2004). Hopper & Gioia (2004) point out that
throughout these climatic fluctuations the Shark Bay region has suffered the most
severe climate change due to the massive differences in rainfall experienced in these
regions during glacial maxima and minima. Arenophryne rotunda is heavily reliant on
soil moisture for dermal rehydration (Cartledge et al. 2006), limited rainfall may have
led to a drop in soil moisture which resulted in a contraction of populations to coastal
areas in the west and reinforcing fragmentation via uplift of the Victoria Plateau.
2.5.2 Phylogeography and population structure – Southern Lineage
Divergence estimates infer a split between the major clades within the southern A.
rotunda lineage (NMG & SMG – Figure 2.1) occurred across the Murchison Gorge
~2.05MYA (Figure 2.4). The Murchison Gorge is the overriding biogeographic feature
within the range of the southern A. rotunda lineage. The final incision of the deep
sandstone gorge in the lower Murchison River is estimated to have occurred between
the late Pliocene and early Pleistocene (Hocking et al. 1987). North of the Murchison
River/Gorge remaining populations in the southern lineage show consistent evidence for
restricted gene flow with isolation by distance, which is suggestive of a species with
limited dispersal over relatively short distances (~50km).
2.5.3 Phylogeography and population structure – Northern Lineage
Allopatric fragmentation was shown (NCPA results) between different prongs (the
north-south oriented finger-like projections seen throughout the Shark Bay coastline)
and between these populations and those on Dirk Hartog Island (Figure 2.4). This
suggests that fragmentation is associated with the flooding of the region and formation
of Shark Bay, as the prong regions and the Island would have been interconnected prior
to Holocene sea level rises, rather than just simply separation of Dirk Hartog Island
from the mainland. Results also suggest that due to the relatively recent flooding of
Shark Bay and subsequent separation of Island vs. Mainland populations, comparatively
little of the genetic variation within this lineage is accounted for by the geographical
separation of Dirk Hartog Island from the Mainland (Table 2.4). The sea level in Shark
Bay is known to have reached its present day levels ~5-6000 years ago (Playford 1990).
This rise led to the separation of Dirk Hartog Island and the formation gulfs between the
56
anticlinal dune ridges which are now the various prongs in the western Shark Bay
region (Butcher et al. 1984).
An overall inference restricted gene flow for the northern lineage (Figure 2.4) is likely
to be the result of a combination of sea level rises, causing both the flooding of Shark
Bay and, when higher than present, isolation of the prong areas along the Shark Bay
coast. Sea level rises, associated with interglacial periods, are likely to be involved in
fragmentation event separating the population around the bay area from those further
south along the coast in the northern lineage (Figure 2.4). Repeated episodes of higher
sea levels than present during the Pleistocene have been shown to have practically
isolated the anticlinal ridges (underlying the several prominent prongs along the coast)
during interglacial maxima. Two specific events during the Pleistocene have been noted
to have dissected the prongs around 240,000yrs and 130-120,000yrs ago, evidenced by
the deposition of distinct limestone formations (Van de Graaff et al. 1980). An
inference of population expansion from coastal sites northwards is likely to have
occurred in response to newly available habitat during the Pleistocene arid maxima
(Figure 2.4). Extensive dune systems formed on the coast during periods of severe
aridity (Hocking et al. 1987). Climatic fluctuations associated with glacial maxima,
which were frequent in the Pleistocene (Dodson, Macphail 2004), also are known to
have resulted in a lowering of sea levels and hence the expansion of coastal dune
complexes in the Shark Bay region (Hocking et al. 1987; Playford 1990).
2.5.4 Conclusions
Historical phylogeography suggests that sea level fluctuations, Pleistocene dune
building episodes and the incision of the Murchison Gorge have led to the development
of major population genetic structure within the northern and southern A. rotunda
lineages. Geological activity during the Miocene resulted in the uplift of the Victoria
Plateau and the reactivation of faults in the area. This geological activity coupled with
the onset of aridity (intensifying ~6MYA) in Australia is likely to have led to the most
prominent phylogenetic break observed within A. rotunda. Additional to the major
mitochondrial lineage split within A. rotunda, morphological evidence suggests species
level differences between the northern and southern lineages and a new species
description for the southern lineage is forthcoming. Overall there seems to be a complex
interaction between geology and climate fluctuations leading to coastal landscape
57
evolution involved in the phylogeographic history of Arenophryne. Such coastal
landscape evolution and climate change have been well recognised in the development
of diversity in the plants and herpetofauna of Shark Bay, however the data from this
study shows that the involvement of geology should not be ignored as a possible
influence in speciation and evolution of the regions biota.
58
59
Group sex is not all its quackedup to be during arid cycles….
60
61
Chapter 3:
The Phylogeography of Crinia georgiana
(The Quacking Frog)
3.1 Abstract
Southwestern Australia is regarded as a global biodiversity hotspot. The region contains
a high number of endemic species, ranging from Gondwanan relicts to much more
recently evolved plant and animal species. Myobatrachid frogs are diverse in
southwestern Australia, and while we know they have speciated in situ in the southwest,
we know little about the temporal and geographic patterning of speciation events.
Crinia georgiana is an ideal subject to test hypotheses concerning the effect of climatic
history on southwestern Australian anurans, as it is an old lineage with a broad
distribution, covering the entire region. I compiled an extensive phylogeographic
dataset based on 1085bp of the mitochondrial gene ND2 for 68 individuals from 18 sites
across the species’ range. Two major genetic clades were identified which were largely
confined to the high rainfall and southeast coastal biogeographic zones respectively.
The clades appear to have diverged around the Plio-Pleistocene border (1.26-
1.72MYA), concordant with increasing intensity and frequency of arid climate cycles.
Subsequent phylogeographic structure appears to have developed primarily during the
Pleistocene climatic fluctuations that also have been integral in generating species
diversity in the endemic southwestern Australian flora. Phylogeographic analyses
identified several dispersal routes, possible refugial areas within the range of the species
and also regions of secondary contact. Dispersal routes identified may now be closed to
the species due to habitat destruction and salinity problems in inland regions, posing
concerns about the evolutionary potential of the species in light of predicted climate
change.
62
3.2 Introduction
Southwestern Australia is an iconic region known for its extreme endemicity, high
species diversity and its threatened environments (Cincotta et al. 2000; Myers et al.
2000). It is widely recognized for its extreme diversity and high level of endemism of
plant species (Hopper 1979; Hopper, Gioia 2004). Less known but equally spectacular
is the high level of faunal diversity, particularly invertebrates (York Main 1996),
mammals, reptiles and amphibians (Hopper et al. 1996). The region has long been a
biogeographical enigma. It lacks obvious historical geographical barriers arising from
events such as glaciation and mountain building, events that are common in many
vicariant speciation models. It has been geologically stable since the Tertiary (Hopper,
Gioia 2004). For animals particularly, our understanding of the processes leading to
speciation and endemism in southwestern Australian fauna is poor. Understanding the
processes generating diversity, both between and within species, is important to the
long-term conservation of conditions that might promote future diversification and
preserve the evolutionary potential of existing species (Moritz 2002).
Processes generating botanical diversity in the southwest are reasonably well
understood. The late Tertiary and Quaternary have been identified as periods of intense
speciation in southwestern Australian flora (Hopper 1979; Hopper, Gioia 2004). During
northern hemisphere glacial cycles of the Quaternary southern Australia experienced
expanding semi-arid conditions with corresponding humid periods during inter-glacial
cycles (Dodson, Ramrath 2001). Studies along the southern margin of Australia also
have shown sea level fluctuations that correspond to interglacial wet and glacial arid
cycling respectively (Galloway, Kemp 1981). Climatic fluctuations led to landscape
evolution, through differential soil erosional/depositional histories and coastal dune and
sandplain development, which contributed to the high levels of diversity and endemicity
observed in southwestern flora (Hopper, Gioia 2004). Extreme levels of plant diversity
are found particularly in the northwestern and southeastern coastal areas of the region,
areas that are more complex topographically than the wider southwestern forest system
(high rainfall zone - HRZ) (Hopper, Gioia 2004). Comparatively little work has been
done to investigate the processes involved in generating diversity both within and
between species of endemic southwestern Australian fauna. The processes acting on
terrestrial vertebrates might be quite different from those involved in the speciation of
63
southwestern Australian plants: e.g. range sizes are often higher, habitat specializations
less marked.
The Myobatrachidae, an ancient anuran family endemic to Australia, show high levels
of diversity and endemism in southwestern Australia (Roberts, Maxson 1985b; a).
There are a number of endemic and relictual anuran species found in the southwest,
particularly in the southern forests, reflecting the ancient history of the region. The
genera Heleioporus, Crinia, Geocrinia and Neobatrachus are highly speciose within
southwestern Australia and this diversity is known to have evolved in situ (Barendse
1984; Roberts, Maxson 1985a; Read et al. 2001; Morgan et al. 2006), but little is
known about the specific speciation mechanisms in most of these genera. Speciation via
polyploidy is known to have occurred within Neobatrachus (Mahony, Robinson 1980;
Mable, Roberts 1997; Roberts 1997), however polyploidy does not occur in other
Myobatrachid genera (Mahony, Robinson 1986). The fragmentation of populations into
drainage systems, associated with periods of drying, may have led to allopatric
speciation in the highly specialized and geographically restricted Geocrinia rosea
species complex (Driscoll 1998a; b). However the same processes seem less likely to
have generated the observed diversity in Crinia or Heleioporus as many species within
these genera have broad distributions that cover semi-arid areas and many congeneric
species are broadly sympatric (Read et al. 2001; Morgan et al. 2006). Thus, it is
important that biogeographic history is assessed in a diversity of species: in particular
those that are widespread across a bioregion (Cracraft 1988; Avise et al. 1998; Riddle et
al. 2000; Zink 2002).
I extend the limited data on processes generating intraspecific diversity in frogs from
southwestern Australia with a comprehensive phylogeographic dataset for Crinia
georgiana (The Quacking Frog). This species has been the subject of numerous sexual
selection and sperm competition studies (e.g. see (Byrne 2004; Byrne, Roberts 2004;
Hettyey, Roberts 2006)) and its breeding success is highly dependent on a predictable
hydrological regime (Dziminski, Roberts 2006). The distribution of C. georgiana
covers the entire southwest forest system (or the HRZ) and extends into the
topographically complex transitional rainfall zone on the southeastern coast
(southeastern coastal zone - SECZ). This distribution thus covers two botanical
provinces and an important biogeographic track described in Hopper & Gioia (2004) as
a path “along which congruent patterns of speciation have occurred within the
64
southwest”. Crinia georgiana is the sister taxon to four other endemic Crinia species
from southwestern Australia and one from eastern Australia (Read et al. 2001),
suggesting it is an old lineage.
Given the antiquity of this lineage, C. georgiana is likely to have experienced multiple
climate fluctuations during the Miocene and Plio-Pleistocene eras, and given its
geographic range and sensitivity to changes in rainfall, the impacts of past climate
change should be reflected in the phylogeography of this species. Also considering the
sensitivity of this species to predictable hydrological regimes (Dziminski, Roberts
2006) this species also serves as an excellent model for investigating the potential
effects of future climate change (Hughes 2003) on widespread generalist species. These
data will be the first comprehensive data set for fauna to contrast with patterns in
southwestern Australian plants which show higher genetic structure and diversity in the
SECZ compared to the HRZ (Hopper, Gioia, 2004).
3.3 Materials and Methods
3.3.1 Tissue samples
Sixty-eight frogs (toe-clips) were sampled from 18 sites across the species distribution,
2-4 animals per site (Figure 3.1, Table 3.1). There is a large gap in our sampling
between Bremer Bay and Cape Le Grand on the southeastern coast. Despite extensive
fieldwork in the area, I found neither animals nor suitable habitat, so I conclude that this
reflects a real gap in the species’ distribution. Furthermore, there are no historical
records (over the last 150yrs) of the species in this region (WA Museum records), with
far eastern populations apparently disjunct from the main range (Tyler et al. 2000). MIS
samples were from the W.A. Museum Tissue Collection (151200-151201-WAM). The
C. pseudinsignifera outgroup used in phylogenetic analyses was collected as part of
another study (32°43´58˝ 116°6´17˝).
65
Table 3.1: Summary of Crinia georgiana tissue collection sites, sample sizesand locations in degrees, minutes, seconds. All points were geodetic WGS84.
Figure 3.1: Map of southwestern Australia showing all major southwestern drainagesystems with map of the Australian continent inset. Tissue collection locations [•] forthe Crinia georgiana phylogeographic study cover the entire known range of thespecies. The gap between the Bremer Bay (BB) and Cape Le Grand (CLG) sites is aknown gap in the species distribution from both current and historical records. SeeTable 3.1 for details on sample sizes, abbreviations and exact locations.
3.3.2 Molecular genetic methods
Template DNA was extracted from samples using a modified CTAB method, suspended
in TE and stored at 0°C. Targeted DNA was amplified using a touchdown PCR profile
(94°C-5min-1×; followed by a series of touchdown cycles of 94°C-30sec, 70-45°C-
20sec (decreasing in 5°increments) and 72°C-90sec - each of these cycles were repeated
2×; followed by a final cycle-94°C-30sec, 40°C-30sec, 72°C45sec, repeated 40×; then
held at 72°C-4min-1×, finishing at 4°-1min). Primers used to amplify ND2 were L4221
(5'-AAGGRCCTCCTTGATAGGGA-3', modified Macey et al., (1998)) & tRNA-trp
(5'-CTCCTGCTTAGGGSTTTGAAGGC-3' modified Read et al. (2001)). Targeted
fragments were amplified in 40µl reactions comprising of ~100ng template DNA, 4µl
10× reaction buffer, 3 mM MgCl2, 0.5 mM dNTPs, 10 pmol primer and 2units of
The phylogenetic tree (showing the ML phylogram topology - Figure 3.2) shows two
lineages. Lineage 2 has strong support (Bayesian Posterior Probabilities/ML bootstraps
= 100/94) as a monophyletic clade, as do several minor clades within this lineage
(Clade 1.37-99/85; Clade 3.5-100/88 – refer to Figure 3 for Clade names). Lineage 2 is
largely confined to the southeast coastal zone with only one population further west in
71
the HRZ at the Harvey-Waroona population (HW). Lineage 1 occupies the HRZ. In the
Kalgan River population (KAL), a southeast coastal site, 2 of 3 frogs also belonged to
lineage 1. Lineage 1 lacks bootstrap support as a reciprocally monophyletic clade from
Lineage 2 (<50/<50); nevertheless separation of the two lineages is supported in a
network (see below - Figure 3), which is generally a more appropriate way to represent
intraspecific data with low levels of divergence (Templeton et al. 1992). Other clades
which receive strong support within lineage 1 coincide with Clades 2 (100/95), 2.4
(98/88) and 2.8 (97/81) in Figure 3.3.
Pairwise differences in haplotypes between the two major lineages ranged from 1.29%
and 2.49% (uncorrected p – refer to Appendix 2b for complete table). The score of the
likelihood tree without enforcing a molecular clock was –InL=2090.89, the score of the
tree with a molecular clock enforced was –InL=2116.49. The likelihood ratio test
showed that sequences did not depart from a clock-like model of evolution
(n.s;P=0.276). The number of nucleotide substitutions (dA) between Lineages 1 & 2 was
0.01426, giving a divergence time of ~1.49MBP (±2SE of 226,000Y). The first lineage
encompasses the majority of the species range, covering the western and southwestern
populations and encompassing the entire southwest forest system. Sequence
divergences range from 0.09%-1.01% within the southwest forest clade. The second
lineage comprises all populations on the south coast east of Albany. The HW and KAL
populations had only one individual of 4 and 3 respectively from this second lineage.
Sequence divergences within the southeast coastal clade ranged from 0.09%-0.92%.
72
Table 3.2: ND2 haplotypes within the Crinia georgiana phylogeographicdataset. The frequency of haplotypes at each collection location are alsoshown, refer to Table 3.1 for site name abbreviations.
Figure 3.2: Maximum likelihood phylogram of 48 Crinia georgiana ND2 haplotypesshowing two major lineages with Crinia pseudinsignifera as an outgroup. Bootstrapswere calculated from 100 replicates and Bayesian posterior probabilities from 4 millionMCMC generations. ML bootstrap values for clades above 70 are represented by *(refer to text for exact values). TrN + I + G model of DNA evolution was enforced inmaximum likelihood analyses as suggested by AIC tests in Model Test 3.7. Map ofsouthwestern Australia is inset with shaded areas representing the range of the twomajor lineages, for site name references see Table 3.1. Map also shows the distributionof the two biogeographical zones in the range of C. georgiana: the High Rainfall Zoneand the Southeast Coastal Zone (cf. Hopper & Gioia 2004).
3.4.2 Phylogeographic analysis
Tajima’s D showed that the C. georgiana mtDNA dataset was consistent with neutral
evolution (DT=-1.119;P>0.05–n.s). All 48 haplotypes were joined with a 95%
probability of parsimonious connection in TCS 1.21. The total cladogram was nested at
the 5-step level, with a maximum of 14 mutation steps between any two haplotypes
(Figure 3.3). The GeoDis output showed several clades within the nested C. georgiana
haplotype network with significant phylogeographic structure from which
biogeographical inferences could be made (significant χ2 P-value: Table 3.3 – for
complete output refer to Appendix 3b). For clade 2.2, we inferred past gradual range
expansion followed by fragmentation from the northwestern HRZ (MO, SA, SP, MUR
74
& HW) to some south coast forest populations (DF, KH & KAL). Independent tests for
demographic expansion show evidence for range expansion in clade 1.10
(R2=0.1241;P≤0.05), but not for any other clade within the nested group
(R2=0.364;P>0.05-1.22 and R2=0.379-1.11;P>0.05–n.s, R2 could not be calculated for
other clades in the nested group as there were only single haplotypes in these clades).
There is a significant geographic signal within clade 2.6 but inadequate geographic
sampling prevents any viable inference of history.
Significant phylogeographic structure was detected within clade 3.1. Clades 2.1 (SG,
BW & NR), 2.5 (BE & BW) and 2.6 (DW, BW & COL) have ranges that mostly do not
overlap with the rest of the clades in the nested group. Clades 2.1, 2.5 and 2.6 are also
separated from the central ancestral clade by a series of missing haplotypes. Range
expansion was detected in clades 2.2 (R2=0.0707;P<0.001) and 2.9
(R2=0.0843;P<0.001), but not other clades. This suggests gradual range expansion into
southwest coastal areas from the northern high rainfall region, followed by
fragmentation. The supplementary testing for secondary contact shows moderate
distance values for the HW, DF, KH and KAL sites at the 2-step level probably
reflecting the presence of both clades 2.2 and 2.9 at these sites (Figure 3.4). Whilst
clades 3.2 and 2.8 show no significant phylogeographic structure using NCPA, the high
support for clade 2.8 would further add to this inference of range expansion into
southwest coastal areas followed by fragmentation.
Lineage 2 (Clade 4.2), or the SECZ lineage, is characterized by local population
structure and several allopatric fragmentation events. We inferred fragmentation
amongst far southeast coastal zone populations (CLG + CANP & MIS) associated with
the separation of Mondrain Island from the coast by rising sea levels (clade 3.5 – Table
3.3). Fragmentation is also inferred in clade 4.2 between the far southeast coastal
populations from the Esperance region (CLG, CANP & MIS) and the haplotypes from
the western portion of the range of lineage 2 (BB, KAL and HW populations). At the
total cladogram level we made an overall inference of contiguous range expansion.
There is evidence for secondary contact between these two discrete mitochondrial
lineages in the HW and KAL populations (Appendix 4b). Clade 4.1 shows evidence of
range expansion (R2=0.0408;P<0.01). Clade 4.2 does not show evidence of range
expansion (n.s;P>0.05).
75
Table 3.3: Biogeographical inferences for nested Crinia georgiana clades withsignificant phylogeographic structure, specified by a χ2 nested contingency test. P-values are calculated from 10000 random permutations and are considered significant ifpermuted expected χ2 values greater than or equal to the observed.
PF – Past Fragmentation; LDC – Long Distance Colonization; RE – Range Expansion; CRE –Contiguous Range Expansion; PGRE – Past Gradual Range Expansion; AF – Allopatric Fragmentation; F– Fragmentation; IGS – Inadequate Geographic Sampling; w/ - with.* Inference of PF w/ CRE is adopted as the appropriate inference despite simple CRE being inferred bythe NCPA inference key.
Total Cladogram <0.001 1-2-11-RE-12 CRE or PF w/ CRE*
Chain of Inference
76
Figure 3.3: Haplotype network for all 48 Crinia georgiana ND2 haplotypes created inTCS 1.21. Each line represents a single mutational change. Ellipse size is proportionalto haplotype frequency with small open circles representing missing haplotypes and thesquare representing the ancestral haplotype as inferred by TCS using outgroup weights.All connections, up to 14 steps, are within the 95% confidence limits of a parsimoniousconnection. Clades were nested using rules outlined in (Templeton et al. 1987; Crandall1994; Templeton et al. 1995).
77
3.4.3 Population genetic analysis
Analyses of molecular variance across the entire C. georgiana dataset sought to
determine the proportion of genetic variance attributed to Hopper & Gioia’s (2004)
HRZ & SECZ biogeographic regions. Further AMOVA analyses assessed the amount
of genetic variance amongst and within the populations within each of the discrete
lineages (Figure 3.2) within the C. georgiana dataset. As populations of single
individuals cannot be incorporated, for these population analyses the single individuals
from populations KAL & HW that fell out with Lineage 2 were grouped as a single
genetic population unit. This was justified by Principal Components Analysis,
performed in GenAlEx v6 with 1000 permutational steps (Peakall, Smouse 2004),
which indicated that these individuals were from the same genetic population (results
not shown). The network created for NCPA also supports this. Table 3.4 is a summary
table of population genetic analyses. AMOVA results across the entire species range
show that 64% of the genetic variation is accounted for by differences between the HRZ
& SECZ. When calculated for the entire C. georgiana ND2 dataset Snn
(0.322;P>0.001) suggests that total population differentiation is extremely low.
AMOVA concurs with low overall levels of population structure, with more genetic
variation accounted for by individuals within populations (22%) than between (14%).
Low differentiation levels overall are probably reflective of the high levels of dispersal
within the majority of the species range, covered by Lineage 1. Lineage 1, mainly
confined to the HRZ, also exhibits extremely low population differentiation
(Snn=0.120;P>0.001), and this is reflected in the AMOVA results, which show that the
majority of genetic variation is among individuals within populations (76%) rather than
between populations (24%). Lineage 2 on the other hand displays the opposite trend
with highly differentiated populations (Snn=0.844;P>0.001), which also accounts for
85% of the genetic variation within this lineage.
78
Table 3.4: Summary table of population genetic statistics for Crinia georgiana as awhole in addition to results from the two major mitochondrial lineages identified inphylogenetic and phylogeographic analysis. Analysis of Molecular Variance(AMOVA) results are presented for the whole species dividing up the distributioninto two regions (High Rainfall (HR) Zone and Southeastern Coastal (SEC) Zone –sensu Hopper, Gioia, 2004) and amongst populations within each major lineage.Hudson’s Snn ‘nearest neighbour’ statistic is also presented as a measure of geneticdifferentiation amongst populations across the species and within major lineages. P-values for each of these analyses were calculated via 1000 permutations.
*** = P≤0.001
3.5 Discussion
The mtDNA sequence data show two major haplotype lineages within C. georgiana
(Figure 3.2) with between 1.29% and 2.49% sequence divergence with strong bootstrap
support and an estimated divergence date of 1.49MYA (±226,000yrs), or around the
Plio-Pleistocene border (~1.64MYA). Given an initial lineage split at the Plio-
Pleistocene border and the minimum age of isolation of offshore Islands throughout the
southwest ~5000yrs ago, subsequent phylogeographic structure within each lineage
appears to primarily be related to climatic fluctuations throughout the Pleistocene.
Following initial fragmentation both lineages have expanded through inland regions,
coming into secondary contact at two sites, in the central western forest (Harvey-
Waroona) and at the meeting of the high rainfall and southeast coastal zones (Kalgan
River, Figures 3.2 ,3.4 & 3.5A). There is evidence of repeated cycles of fragmentation
followed by range expansion within Lineage 1, the haplotype lineage largely confined
to the HRZ (Figures 3. 2 & 3.5B). Higher levels of genetic structure and signals of
Source df SS MS Est. Var. % Stat Value Hudson's SnnWhole Species 0.322***
HR Zone Vs SEC Zone 1 150.853 150.853 5.666 64% φRT 0.643***Pop's / region 16 104.088 6.506 1.211 14% φPR 0.385***Indiv. / Within Pop's 50 96.500 1.930 1.930 22% φPT 0.781***
Population Genetics Analysis Summary Results Table
79
allopatric fragmentation characterize lineage 2 (Figures 3.5B & 3.5C), which is largely
confined to the more arid SECZ with some obvious patterns of differentiation on
offshore island populations isolated by sea level rises most recently after the last glacial
maximum (Mondrain Island, Figure 3.5C)
80
Figure 3.4: Biogeographic hypotheses regarding the two major lineages within Criniageorgiana and their responses to Plio-Pleistocene climatic fluctuations. Figure 3.5Apresents the biogeographic hypothesis of initial fragmentation of the two major C.georgiana lineages caused by arid conditions followed by recent dispersal across inlandregions during wetter interglacial periods. Figure 3.5B Shows fragmentation ofsoutheast coastal populations and a restriction of dispersal from the north into southwestcoastal populations effected by increasing aridity (the latter may also be compoundedby increasing salinity of coastal wetlands during interglacials). Figure 3.5C showsphylogeographic hypothesis regarding the response of the two major C. georgianalineages to interglacial wet periods, where dispersal is likely to be established acrossinland areas (followed fragmentation by aridity) and populations become isolated onoffshore islands by rising sea levels. - Sampled Populations.
81
3.5.1 Biogeography of Crinia georgiana and southwestern Australia
The main biogeographic hypothesis adopted for Crinia georgiana is that the species
appears to be a formerly widespread lineage fragmented into two lineages, between the
HRZ and SECZ biogeographical regions, each of which then expanded to come into
secondary contact at several sites. Divergence estimates suggest the separation of the
two lineages occurred around the beginning of the Plio-Pleistocene border glacial
cycles, with each lineage subsequently expanding through inland areas during wetter
interglacial periods. An inference of contiguous range expansion was originally given
by NCPA, with no inference of fragmentation despite 14 mutational steps separating the
two major lineages. Additionally range expansion is not detected for Lineage 2, but is
for Lineage 1. Subsequent contraction and fragmentation within lineages may account
for the incorrect inference, alternatively expansion may have been very recent and rapid
leading to a lack of signal may explain both these phenomena (Masta et al. 2003). The
occurrence of divergent lineages with different geographical centres and largely non-
overlapping distributions at the Harvey/Waroona and at the Kalgan sampling sites is
consistent with fragmentation followed by range expansion and subsequent population
mixing. Sampling from populations in intermediate inland areas between the KAL and
HW sites and larger sample sizes, may have yielded more accurate inferences.
Additionally, molecular clock estimates are fraught with difficulties (Rambaut,
Bromham 1998; Gillooly et al. 2004), the date obtained of ~1.49MYA provides an
estimate that is consistent estimated climate change in Australia (Galloway, Kemp
1981; Kendrick et al. 1991) and tightly links with the onset of 100,000 year glacial
cycling at 1.5MYA (Rutherford, D’Hondt 2000), and with dramatic changes seen in
other southwestern Australian biota (Hopper 1979; Rabosky et al. 2004).
The Plio-Pleistocene border (1.64MYA) was a time of immense climatic change in
Australia followed by arid pulses increasing in frequency and intensity during glacial
maxima (Bowler 1976; Kershaw et al. 1991; Macphail 1997). High seas and wet humid
conditions are indicated at the Plio-Pleistocene border, followed by a rapid regression
and reversion back to arid conditions first seen in the late Miocene (Galloway, Kemp
1981; Kendrick et al. 1991). A significant drop in rainfall has been inferred for the
southwest at the Plio-Pleistocene border, falling to below 600mm for the first time on
the southeastern edge of the southwest land division (Macphail 1997). Palynological
evidence also shows rainfall decreasing at both the northwest (<200mm) and
82
southeastern margins (200-400mm) of the higher rainfall zone (Macphail 1997;
Dodson, Macphail 2004) (cf. a current level of 300+mm, up to 600mm in the Esperance
region; Bureau of Meteorology - http://www.bom.gov.au/). Divergence estimates
suggest the separation between C. georgiana lineages occurred around the Plio-
Pleistocene border. Climates today are similar to interglacial wet periods throughout the
Quaternary, however currently it is still slightly drier and less humid than many of the
‘wetter’ interglacial periods (Dodson, Ramrath 2001). Hypotheses closely linking the
historical biogeography of C. georgiana with climate and associated rainfall
fluctuations are plausible as the species relies on seasonably predictable rainfall for
successful recruitment (Dziminski, Roberts 2006). Therefore, any significant change in
rainfall levels and predictability, as has been the case with severe arid pulses, are certain
to disrupt the breeding cycle of this species.
Complex interactions between a changing climate and sea levels has lead to the
diversity observed within the southwestern Australian plant communities, primarily in
the changeable Plio-Pleistocene era. Dramatic fluctuations in rainfall within Hopper’s
(1979) transitional rainfall zone have not only shaped to biogeography of endemic
plants, but have also impacted on endemic fauna. Plant distributions show a similar
pattern to that seen within C. georgiana, with sister species affiliations or disjunct
distributions between the HRZ and scattered throughout the wetter pockets along the
SECZ (Hopper 1979; Hopper, Gioia 2004). The HRZ and SECZ also distinguish much
of the genetic diversity within C. georgiana (Table 4). A scenario where populations are
fragmented into high rainfall and southeast coast lineages followed by expansion during
inter-glacial periods may also explain the distribution patterns of Litoria moorei and L.
cyclorhynchus; a pair of recently diverged anuran species (Roberts, Maxson 1988; Cale
1991; Burns, Crayn 2006) which hybridise in this border region (Cale 1991). With
climate change rainfall patterns within the southwest are beginning to shift and will get
more extreme in the future, a trend of less rainfall during the formerly predictably wet
Autumn/Winter period to more rain in the formerly dry Summer period is predicted to
continue and intensify (Hughes 2003). Given a history so closely tied to climate there is
concern for the ability of C. georgiana and other southwestern Australian endemics to
cope with future climate change. Furthermore, should species be able to cope with the
change in rainfall seasonality and rainfall levels return to normal, the combined effects
of salinity and habitat destruction may alter the ability of biota to move through
historical inland dispersal tracks.
83
3.5.2 Phylogeographic and population genetic patterns
The phylogeographic histories of the two major C. georgiana lineages differ markedly,
with lineage 1, which is largely confined to the HRZ, showing repeated episodes of
range expansion, with subsequent fragmentation in southwest coastal and inland areas.
Range expansion followed by fragmentation occurred between northern and southwest
coastal areas and between the southeastern and northern-forested areas of the HRZ
within this lineage (Figure 3.5B & 3.5C). Population structure within the forest system
(lineage 1) was low, which is also consistent with the repeated dispersal inferences of
NCPA and confirms the results of the only previously available genetic data (from
allozyme studies) for C. georgiana (FST=0.066-over 237km) (McDonald 1998). Similar
levels of population structure are seen in other widespread southwestern amphibian taxa
(FST=0.087 over 100km-Heleioporus albopunctatus (Davis, Roberts 2005); FST=0.088
over 80km-H. psammophilus (Berry 2001). Lineage 2, occupying the arid SECZ,
appears to be characterized by several instances of allopatric fragmentation. This is
reflected in the results of population genetic analyses on this lineage confirming higher
levels of population genetic structure in the southeast coastal zone.
Biogeographic hypotheses within each lineage, following a split around the Plio-
Pleistocene border, are consistent with a fluctuating climate throughout the Pleistocene.
Arid maxima are associated with significant drops in rainfall (Macphail 1997), and
climatic fluctuations in general have been associated with dramatic changes in rainfall
throughout the more inland regions of the southwest (Hopper, Gioia 2004). Arid pulses
are likely to have led to restricted dispersal between the wetter northern and
southeastern coastal regions of the HRZ within lineage 1. Aridity would cause this
primarily by leading to a contraction of the species range to coastal areas and further by
restricting dispersal between the wetter refugial areas along the coast. Strong signatures
of range expansion from the northern and southeastern coastal regions of the HRZ
indicate that these regions have acted as primary refugial areas for C. georgiana lineage
1 haplotypes during arid maxima (Lessa et al. 2003). Northern and southern refugial
areas are also suggested for plant taxa, with a common congruent biogeographic track
observed between northern and southern regions of the HRZ (Hopper, Gioia 2004).
Palynological evidence shows northern highland regions remaining relatively wet even
during arid maxima due to the relief of the Darling Ranges (Macphail 1997). A southern
84
forest refugial area is also well corroborated by the presence of several relictual plants
and animals with Gondwanan affinities (Hopper et al. 1996; Roberts et al. 1997).
Interglacial wet periods appear to have allowed repeated range expansion throughout
the interior regions of the southwest. Wet inter-glacial periods have been noted as times
when the HRZ extended far into the currently semi-arid regions (Hopper, Gioia 2004)
and this would have allowed C. georgiana’s range to expand into inland areas, where
the combined effects of adult movement between catchments during rain and tadpole
movement across catchments during flooding may have resulted in the current dispersal
patterns observed (Figure 3.5C). In inland regions of the southwest many of the upper
catchments of rivers draining towards the coast come into close contact (Figure 3.1).
Whilst these upper regions of catchments are currently thought of as more
palaeodrainage systems (Beard 1999), significant increases in rainfall during wetter
interglacials would have expanded suitable available breeding habitat for the species
throughout inland reaches of the southwest.
Dramatic sea level fluctuations were also associated with climatic fluctuations of the
Pleistocene; high sea levels stands correspond to interglacial maxima and low sea level
stands correspond to glacial/arid maxima (Galloway, Kemp 1981). Cenozoic
transgressions during high sea level stands have been shown to have consistently
affected the area east of Augusta and Geographe Bay areas in the extreme southwest, in
addition to vast sections of the western coastline (Sircombe, Freeman 1999).
Subsequently, southwestern coastal plain vegetation communities did not fully develop
to their current positions until the mid-late Pleistocene (Kendrick et al. 1991). During
lower sea levels the species could move into and occupy newly available coastal
habitats on the Swan Coastal Plain and extreme southwest corner. Higher sea levels
than present are known to have lead to severe and rapid change in coastal plant
communities (Sircombe, Freeman 1999; Hageman et al. 2003) and coastal wetlands and
estuarine systems (Hodgkin, Hesp 1998). These processes are very likely to have lead to
the restricted gene flow between coastal populations (lineage 1-Figure 3.5B) and
isolated populations on offshore Islands (lineage 2-Figure 3.5C). Arid cycles were also
noted to impact on the extreme southwestern flora and fauna (Dortch 2004), therefore
the combined effects of dramatic sea level and salinity changes and pulses of aridity are
likely to be responsible for the signal of restricted gene flow among coastal populations
and between coastal populations and those in more stable refugial areas (Figure 3.5B).
85
Predominant inferences within lineage 2 are of fragmentation of populations in the
Esperance region from populations further west within the range of lineage 2. This is
most likely to be due to the increasingly frequent and intense arid pulses of the
Pleistocene (Figure 3.5B). Crinia georgiana has never been collected in the area
between these two regions, and has been noted as extremely rare in the Fitzgerald
region (Chapman, Newbey 1995), 30-40km east of Bremer Bay. Rainfall maps
(Hopper, Gioia 2004) show that between these regions rainfall declines to below
600mm, which appears to be the limit of the species’ distribution from known records.
Dispersal may have occurred through-now flooded coastal habitats during low sea
levels along the southeastern coastline, to be fragmented by rapidly rising seas
throughout the late Pleistocene (Hageman et al. 2003), as has been the case with
populations known from offshore Islands. Alternatively, the area between Esperance
and Bremer Bay may still have been extremely arid during recent interglacials. Ever
increasing aridity would therefore prevent significant dispersal of C. georgiana through
newly created coastal habitats, resulting in a pattern of isolated refugial populations
often seen in the plants of this region (Wright, Ladiges 1997; Hopper, Gioia 2004).
Hence it is likely that the combined influence of sea level and climatic fluctuations have
contributed to the fragmentation of the Esperance populations from the rest of lineage 2.
3.5.3 Conclusions
On the basis of this study, I propose the following scenario to explain the current
haplotype distributions of C. georgiana. The clear phylogenetic break within C.
georgiana, which separates lineages from the HRZ (lineage 1) & SECZ (lineage 2),
resulted from aridification around the Plio-Pleistocene border. This fragmented
populations from the HRZ and those from the southeastern coast, isolating the latter into
more mesic pockets along the predominately arid and hostile southeastern coast. With
ameliorating conditions during Pleistocene interglacials the two now-divergent lineages
expanded through inland areas to reclaim much of the species’ former range.
Subsequent intense Pleistocene aridification cycling would then have resulted in
repeated fragmentation within both lineages. Refugia existed in the northern and
southeastern portions of the HRZ (lineage 1) and the species has persisted in the Bremer
Bay-Fitzgerald River, and Esperance regions along the semi-arid southeast coast
(lineage 2). Wetter interglacial climates during the Quaternary allowed for repeated
86
dispersal through inland areas between refugial areas within the HRZ. The compounded
effects of high seas, leading to isolation of C. georgiana populations on offshore Islands
off the southeastern coast, and arid conditions probably effected fragmentation of
southwest coastal populations of the HRZ lineage. Together these results imply a
remarkably similar biogeographical history to that seen in relictual plants and other
endemic frogs of southwestern Australia, confirming the biogeographical zones outlined
by Hopper & Gioia (2004). Given these historical patterns and the human mediated
modification of habitats throughout inland regions, there is some concern for the
evolutionary potential of the species in light of predicted climate change.
87
The elusive Nicholl’s Toadlet….
I am here, you can hear me, butcan you find me?!!!
88
89
Chapter 4:
The Phylogeography of Metacrinia nichollsi
(Nicholl’s Toadlet)
4.1 Abstract
Southwestern Australia is a biodiversity hotspot of intense evolutionary interest due to
the large number of endemic and relictual plant and animal species, long-term
geological stability and what appears to be rampant in situ speciation. Southwestern
Australian distributed myobatrachid frogs have featured heavily in the testing of
biogeographic hypotheses for the region. Increasing evidence suggests that historical
arid periods have played a critical role in initiating divergence of the group in the
southwest, with isolation on major drainage systems a recurring pattern along the
southern coast. Metacrinia nichollsi provides an excellent contrast to other frogs in the
region because it is an abundant, continuously distributed species with direct developing
eggs deposited on land not necessarily associated with drainage systems. We have
compiled an extensive phylogeographic dataset comprising sequences of ND2 for 69
animals from 16 sites, representing the entire distribution of the species. Late Miocene-
Pliocene aridity appears to have isolated the Stirling Ranges populations, which are of
serious conservation concern due to impending climate change. Similarly this period of
aridity is also likely to have resulted in the formation of two major lineages within the
remaining range of the species in a primarily north-south orientation. One of these
lineages has strong levels of drainage-based population structure, while the other shows
a strong signature of recent expansion. Our results confirm that climatic fluctuations in
the region have impacted this species, further adding to the increasing body of
knowledge on the impacts of climate change on the biogeographic history of the poorly
studied southwestern Australian fauna.
90
4.2 Introduction
The southwestern corner of Australia provides an interesting biogeographical and
evolutionary conundrum. The region is a centre of endemism and a biodiversity hotspot
of global significance due to its high species diversity and highly threatened
environments (Cincotta et al. 2000; Myers et al. 2000). Yet the southwest of Australia
has lacked obvious vicariant forces typically involved in speciation, such as glaciation
or major tectonic or volcanic activity, as the extreme southwest has been geologically
stable since the Tertiary (Hopper, Gioia 2004). Southwestern Australia is world-famous
for its extreme diversity of plant species (Hopper 1979; Hopper, Gioia 2004), but it is
also home to a great diversity of endemic invertebrates (Main 1996), mammals, reptiles
and amphibians (Hopper et al. 1996). While phylogenetic and phylogeographic
investigations into speciation mechanisms in the plants of the southwest have rapidly
accumulated over the last 30 years (Hopper, Gioia 2004), our understanding of the
processes resulting in speciation and genetic diversity within southwestern faunal
assemblages remains comparatively poor. Given the levels of human habitat
modification in the region, understanding speciation processes and the distribution of
genetic diversity is paramount for competent conservation efforts, in addition to their
inherent evolutionary interests (Moritz, Faith 1998; Moritz et al. 2001; Moritz 2002).
Hopper & Gioia (2004) have analysed patterns/mechanisms of speciation in
southwestern Australian flora, providing a significant foundation for investigations into
faunal speciation. They focussed on processes and diversity in the transitional climatic
zone between the wet and arid zones. The flora forest system and wet rainfall areas on
the south and lower western coasts are less rich floristically and more relictual in nature.
Concurrent with the increased climatic fluctuations of the late Tertiary and Quaternary
(Dodson, Ramrath 2001) are periods of intense speciation in southwestern flora,
primarily in Hopper’s Transitional Rainfall Zone (Hopper 1979; Hopper, Gioia 2004).
Climatic fluctuations lead to landscape evolution, primarily due to soil erosion and
deposition processes, and cyclical population fragmentation and expansion that resulted
in explosive speciation (Hopper, Gioia 2004). Disjunct distributions within the
southwestern flora are also more the rule rather than the exception due to fragmentation
of environments during long-term geological stability (Dirnböck et al. 2002). However,
relative to plants, our understanding of speciation processes has been limited in endemic
91
southwestern Australian fauna, which must have been subject to many of the same
climatic and geological processes.
The Myobatrachidae are an ancient anuran family endemic to Australia and Papua and
have long been recognised as being particularly diverse in the southwest (Roberts,
Maxson 1985b; a). There are a number of endemic, monotypic and relictual
Myobatrachids in the southwest (Roberts et al. 1997), particularly in the more mesic
southern forest, signalling the ancient history of the region. The genera Heleioporus,
Crinia, Geocrinia and Neobatrachus are of particular note with a diversity of endemics
known to have speciated in situ in the southwest (Morgan et al.; Barendse 1984;
Roberts, Maxson 1985a; Read et al. 2001). Some Neobatrachus species have speciated
via polyploidy (Mahony, Robinson 1980; Mable, Roberts 1997; Roberts 1997), but
polyploid evolution has not occurred in other Myobatrachid genera (Mahony, Robinson
1986).
While broader mechanisms of speciation in southwestern myobatrachids are yet to be
clarified, there is increasing evidence that climatic fluctuations, such as those acting to
generate diversity in the plants, may have been particularly important in shaping the
distributions of southwestern Australian frogs. Peripheral isolation and fragmentation of
populations via fluctuating climate is thought to be involved in speciation within the
Geocrinia rosea species complex, a series of allopatric, highly restricted and specialised
species across the relictual wet forested southern coast of Western Australia (Wardell-
Johnson, Roberts 1993; Driscoll 1998a; b). Pleistocene climatic fluctuations appear to
have been important in shaping the historical and current distribution in the widespread
frog Crinia georgiana (Edwards et al., submitted ms). Generation of a comprehensive
view of the historical biogeography of frogs in the region requires data from multiple
species with varying life histories and habitat relationships (Cracraft 1988; Avise et al.
1998; Riddle et al. 2000; Moritz et al. 2001; Zink 2002). Data so far are from
conventional aquatic breeders (Edwards et al. submitted ms) or direct developers with
very specialised wet forest requirements for spring and summer breeding (Driscoll
1998a; b).
To compile a phylogeographic dataset for Metacrinia nichollsi we sequenced a 1125bp
fragment of the mitochondrial ND2 gene from sixty-nine individuals from 16 sites
across the entire species range. Metacrinia nichollsi is a direct developer with non-
92
specific breeding site requirements which occurs widely across a range of landscapes
from relatively dry coastal heaths to the wettest karri and tingle forest systems in the
high rainfall zone of southwestern Australia. There are also populations in the eastern
Stirling Range, which are geographically isolated from the forest systems to the
southwest (Tyler et al. 2000). The distribution of M. nichollsi is not obviously tied to
drainage systems and therefore may not show the same extreme fragmenting effects of
climate as seen in the G. rosea species complex (Driscoll 1998a; b). However, M.
nichollsi is also an old lineage (Read et al. 2001), which still retains a summer breeding
regime. Given its apparent abundance, continuous distribution and relictual
characteristics the species is likely to display the general effects of long-term climate
change across the ‘relictual’ wet forests along the southwestern Australian coast
providing a contrast to studies so far conducted on restricted specialist species.
4.3 Materials and Methods
4.3.1 Tissue samples
Sixty-nine individuals were sampled (toe-clips) from 16 sites across the entire species
distribution with 2-10 animals per site (Figure 4.1, Table 4.1). The gap that exists
between the Stirling Ranges population and the main range of the species is real; both
current and historical surveys have failed to find the species in intervening areas. Past
and present surveys in The Porongurup Mountains (34°40'46" 117°52'23") have
recovered no records of the species (Past 50 years of trapping - B. York-Main, pers.
comm.; Current surveys – D. Edwards pers. obs.), despite what is apparently ideal
habitat for the species. Two sites, approximately 15km apart, were sampled within the
Stirling Ranges, with 5 animals from each. Due to lack of genetic diversity they are
considered together below. Arenophryne rotunda (27°49'59" 114°21'53") and
Myobatrachus gouldii (30°01'57" 115º49'06") sequences were used as outgroups for
this study.
93
Table 4.1: Metacrinia nichollsi sampling site location names, abbreviations, samplesizes and exact GPS coordinates in degrees, minutes, seconds. All points are inGeodetic WGS 84.
Figure 4.1: Map of the southwestern Australian coastline with map of the Australiancontinent inset. Tissue collection locations [•] for the Metacrinia nichollsiphylogeographic study cover the entire known distribution of the species. Refer toTable 1 for further information on sample sizes, abbreviations and exact locations.
4.3.2 Molecular genetic methods
Template DNA was extracted from toe samples using a modified CTAB method,
suspended in TE and stored at 0°C. Targeted DNA fragments were amplified using a
touch-down PCR profile (94°C for 5min - 1×; 94°C for 30sec, 70-45°C (decreasing in
5°C increments) for 20sec, 72°C for 90sec - 2×; 94°C for 30sec, 40°C for 30sec, 72°C
for 45sec - 40× ; 72°C for 4min - 1× ; 4°C held. Primers used to amplify the
mitochondrial gene ND2 were L4221 (5'-AAGGRCCTCCTTGATAGGGA-3', modified
from Macey et al., 1998) & tRNA-trp (5'-CTCCTGCTTAGGGSTTTGAAGGC-3'
modified from Read et al. (2001)). Targeted fragments were amplified in 40µl reactions
comprising of ~100ng template DNA, 4µl of 10× reaction buffer, 3 mM MgCl2, 0.5 mM
dNTPs, 10 pmol of each primer and 2 units of Platinum Taq polymerase (Life
Technologies, Gaithersburg, MD).
95
Samples were run out on a 2% Agarose gel and cleaned up using a Mo Bio UltraClean
DNA Purification Kit (Mo Bio Laboratories, Inc). Approximately 100ng of PCR
product was added to sequence reactions using either DYEnamic ET Terminator
(Amersham Pharmica Biotech) or Big Dye Terminator 3.1 (Applied Biosystems)
sequence mix and run according to manufacturers specifications. Internal primers,
L4437 (5'-AAGCTTTCGGGGCCCATACC-3', Macey et al., 1998) and H4980 (5'-
ATTTTTCGTAGTTGGGTTTGRTT-3' Macey et al. (1998)), were used for sequencing
in addition to PCR primers to obtain reliable sequence across the entire gene. Cleaned
reactions were then resuspended in a loading dye/formamide mix. Sequences were
visualised on an ABI 377 Automated Sequencer or an ABI 3010 Capillary sequencer
(Applied Biosystems). DNA sequence data were then edited using Sequencher 3.0
(Gene Codes Corporation).
Sequences were aligned individually using ClustalX (Thompson et al., 1997).
Alignments were then checked by eye. Sequences were translated using the mammalian
genetic code option in Sequencher 3.0, and a clear reading frame was observed in all
sequences. Thus sequences were assumed to be genuine mitochondrial copies and not
nuclear paralogues.
4.3.3 Phylogenetic analysis
We have used phylogenetic analysis techniques in conjunction with sequence
divergence estimates and a rough molecular clock to assess overall phylogenetic
structure and approximate timing of major splits within M. nichollsi. Maximum
likelihood (ML), Maximum Parsimony (MP) analyses (both using PAUP*4.0b10
(Swofford 2002)) and Bayesian MCMC analyses (using MrBayes v3.1.2 (Huelsenbeck,
Ronquist 2001; Ronquist, Huelsenbeck 2003)) of haplotypes were carried out to resolve
and assess support for relationships between the major clades and overall phylogenetic
structure. Akaike Information Criteria (AIC) were used to select the best-fit model of
evolution from the data for ML analyses using Modeltest 3.7 (Posada, Crandall 1998),
and to calculate the nucleotide frequencies, substitution rates, gamma distribution and
proportion of invariant sites for the data under the selected model. Branch support for
the ML and MP trees is provided in the form of likelihood bootstrap values calculated
from 100 bootstrap replicates. For ML and MP analyses starting trees were obtained by
step-wise addition and the TBR method of branch swapping was employed in each
96
heuristic search. Bayesian analyses were conducted using the GTR model with a
proportion of invariable sites and the remaining variable sites having a gamma
distribution using default priors for MCMC analyses in MrBayes v3.1.2. Four
independent runs of 4 chains each were run for 4×106 generations sampling every 100
generations, burnin was set at 400,000 generations. Convergence of posterior
probabilities and stationarity of likelihood scores between the two runs was assessed in
Tracer v1.3 (Rambaut, Drummond 2005). Other descriptive statistics such as haplotype
diversity (Hd) and nucleotide diversity (π) were calculated in DnaSP v4.10.3 (Rozas,
Rozas 1999).
Divergence between major M. nichollsi lineages was calculated using the formula of
Nei and Li for dA (the average number of nucleotide substitutions per site between
clades/lineages (Nei 1987). The dA parameter estimates and their standard errors were
calculated using DnaSP v4.10.8 (Rozas, Rozas 1999). There are no appropriate external
calibration points/fossils with which to calibrate a molecular clock rate for any
southwestern frog genera, despite the existence of some fossils found in recent to
Pleistocene cave deposits (Roberts, Watson 1993; Price et al. 2005). Therefore, we
adopted the molecular clock rate of 0.957%/million years, calibrated for ND2 in
Eleutherodactylid frogs (Crawford 2003). To ensure that the M. nichollsi ND2
sequences were evolving in a clock like manner, a maximum likelihood search was
conducted in PAUP*4.0b10 (Swofford 2002) enforcing a molecular clock. A likelihood
ratio test was then performed to assess if there was any significant difference between
the likelihood scores of trees with and without a molecular clock enforced (Felsenstein
1981) in Modeltest 3.7 (Posada, Crandall 1998).
4.3.4 Phylogeographic analysis
Aims of the phylogeographic and intraspecific analyses were to provide a measure of
geographical significance of genetic pattern and to attain an inference of the
evolutionary history of M. nichollsi. These results were then directly compared to the
known climatic history to determine the impact, if any, of climate fluctuations on M.
nichollsi and as a direct comparison to other species studied across the southern
Western Australian coast. Nested Clade Phylogeographic Analysis (NCPA) tests for
significant geographic clustering of haplotype variation and is one method of
97
distinguishing between the historical factors responsible for the associations between
gene trees and geography (Templeton 1998).
Unrooted statistical parsimony haplotype networks or gene trees were created using
TCS 1.21 (Clement et al. 2000), the network was then nested according to the nesting
rules outlined in Templeton & Sing (1993), Templeton et al. (1995) and (Crandall et al.
1994). Where interior/tip status was ambiguous, particularly at the final nesting level of
the separate networks, clade outgroup probability (Castelloe, Templeton 1994) and
position in relation to outgroups in the phylogenetic tree (Figure 4.2) were used to
determine the interior clade. Tests for geographical association were carried out on the
nested haplotype network in GeoDis v2.4 (Posada et al. 2000) using the latitude and
longitude coordinates for each sampling location. Clades with significant
phylogeographic structure were specified by a significant χ2 value from contingency
tests calculated over 1000 random permutations. The distance values (DC & DN) from
the clades with significant phylogeographic structure were then used in conjunction
with the NCPA inference key (http://darwin.uvigo.es/software/geodis.html) to
reconstruct population histories.
Recent criticism of NCPA, based on the lack of separation of biological interpretation
from statistical testing (Knowles, Maddison 2002), was successfully defended by
(Templeton 2004). Therefore, NCPA remains a powerful phylogeographic analysis
technique, particularly where the events and processes affecting species evolutionary
histories are not known a priori (Templeton 2004). We employed several analytical
techniques to complement the NCPA analyses. Initially Tajima’s D (DT) was calculated
to ensure sequence data fitted the assumption of neutral evolution (Tajima 1989), using
DnaSP v4.10.8 (Rozas, Rozas 1999). Where NCPA requires confirmation of recent
population expansion in certain clades (e.g. step 21 of the current key) R2 tests (Ramos-
Onsins, Rozas 2002) were conducted to test the hypothesis of constant population size
versus population growth using the coalescent simulations and permuted 1000 times in
DnaSP v4.10.3 (Rozas, Rozas 1999). R2 tests for population growth based on the
difference between the number of singleton mutations and the average number of
nucleotide differences between sequences and is a powerful test, especially with limited
sample sizes (Ramos-Onsins, Rozas 2002). Where secondary contact between distinct
haplotype lineages was suspected the supplementary tests described in Templeton
(2001) were carried out. This involves the calculation of pairwise distances between the
98
geographical centres of each haplotype/clade (provided by the GeoDis v2.4 output)
found at each sampling site, this is calculated for every nesting level of the cladogram.
Secondary contact can be inferred if haplotypes/clades with divergent geographical
centres are found at the one location (Templeton 2001; 2004).
4.3.5 Population genetic analysis
Population genetic statistics were used to investigate and describe genetic structure
within the southcoastal and main range lineages of M. nichollsi. DnaSP v4.10.8 (Rozas,
Rozas 1999) was used to calculate Hudson’s Snn ‘nearest neighbour’ statistic with 1000
permutations via the coalescent, to provide a quantitative measure of population genetic
structure both for the entire species data and the major lineages specified above.
Hudson’s Snn ‘nearest neighbour’ statistic is specifically designed for haplotype
sequence data and has been shown to outperform a range of other statistics used to
estimate genetic differentiation (Hudson 2000). Values of Snn are expected to be close
to 0.5 if populations are panmictic, and closer to 1 if populations are highly
differentiated (Hudson 2000). Analysis of Molecular Variance (AMOVA) was
calculated in GenAlEx v6 (Peakall, Smouse 2004) with 1000 permutations. AMOVA’s
were calculated between and among populations across the major lineages specified to
assess genetic variation amongst populations.
4.4 Results
4.4.1 Phylogenetic Analysis
The 1125bp sequence fragment of ND2 from 69 individuals recovered 26 haplotypes
(Table 2) with 93 variable sites, 49 of which were parsimony informative (Appendix
1c). Total haplotype diversity (Hd) was 0.861 ± 0.034 and total nucleotide diversity (π)
was 0.02316 ± 0.0021. For phylogenetic analysis the TIM + I + G model of DNA
evolution was selected using AIC tests in Modeltest. The following parameters, Base =
87/92/100; Clade 3.2 – 69/70/99; 2.3 – 84/80/97; see Figure 4.3 for reference clade
names).
Pairwise differences in haplotypes between the SRL and MRL range from 4.36% to
4.71% sequence divergence (uncorrected p), and between the SRL and SCL from
4.62% to 5.42% sequence divergence. Differences between haplotypes between the two
lineages present in the bulk of the M. nichollsi range, MRL and SCL, range between
2.76% to 3.56% sequence divergence (Appendix 2c). The Hd for the SCL was 0.906 ±
0.04 and π = 0.00654 ± 0.00067. All haplotypes from the SRL were the same (Hd & π =
0). The remaining MRL had Hd = 0.655 ± 0.088 & π = 0.00179 ± 0.0004. Divergences
within the MRL vary between 0.09-0.8%, with divergences within the SCL ranging
from 0.09-1.33%. The score of the likelihood tree without enforcing a molecular clock
was –InL = 2200.1561, the score for the tree enforcing a molecular clock was –InL =
2180.8659. The likelihood ratio tests showed that sequences did not depart from a clock
like model of evolution (P=0.03021; n.s using default and conservative α=0.01). When
this same test is run on all samples excluding the SRL haplotype, the molecular clock
assumption is accepted much more strongly (P=0.07771; -InL[clock]=2010.1618; -
InL[no clock]=1993.5639). The average number of nucleotide substitutions per site (dA)
between SRL and MRL was 0.0454 ± 0.00322, providing a divergence estimate of
4.74MYA ± 330,000yrs between these two lineages. Between SRL and SCL was dA =
0.04963 ± 0.00527 and therefore divergence between these two lineages is estimated at
5.19MYA ± 551,000yrs. The estimates of divergence placed on the separation of the
SRL clade are taken more as a guide rather than an exact measure due to potential
confounding factors associated with this clade. Finally a more recent divergence is
obtained between SCL and MRL of 2.89MYA ± 177,000yrs (dA=0.02764±0.00169).
100
Table 4.2: ND2 haplotypes within Metacrinia nichollsi. The frequencyof haplotypes at each location within each lineage are also shown, referto Table 1 for site name abbreviations.
Figure 4.2: Maximum Likelihood phylogram of 26 Metacrinia nichollsi ND2haplotypes showing three major lineages with Arenophryne rotunda and Myobatrachusgould i i as outgroups. Clade support is provided by MP bootstrap/MLbootstrap/Bayesian Posterior Probabilities. The TIM + I + G model of DNA evolutionwas enforced in ML analyses selected by AIC tests in Model Test 3.7. Map of thesouthwestern Australian coastline is shown inset with shaded areas representing thedistribution of the main range, southcoastal and Stirling Ranges lineages, for site namereferences see Table 1.
4.4.2 Phylogeographic Analysis
Intraspecific analysis techniques were used to provide information on the biogeographic
and historical inferences contained within the data. Tajima’s D for the M. nichollsi
dataset showed that sequences were evolving neutrally (DT=1.02547;n.s-P>0.1). Three
separate networks were joined at the 95% probability of a parsimonious connection.
The first contained the haplotypes from the Stirling Ranges Lineage (SRL - Figure 3A).
The second contained haplotypes from the majority of the species range, excluding
some of the southern catchment areas and termed the Main Range Lineage (MRL),
haplotypes in this network were connected by a maximum of 10 mutational steps
(Figure 4.3B). Lastly haplotypes from the NRS, DFS, KHS & KALS sites were all
102
joined in the South Coastal Range Lineage network (SCL - Figure 4.3) by up to 16
mutational steps. The SCL network differs from the MRL and SCL networks by 49 and
52 mutational steps respectively. The MRL and SCL networks differ by 31 mutational
steps. Due to the large divergence between each of the separate networks they were not
joined for nested clade analyses.
The GeoDis 2.4 output showed several clades with significant distance values (for
complete GeoDis output refer to Appendix 3c). A summary of the clades with
significant phylogeographic signal and the subsequent biological inferences obtained is
outlined in Table 4.4. Clade 2.1 shows evidence of restricted gene flow with isolation
by distance amongst all sites represented by the MRL network, except the NRM and
NRN sites. Also in the MRL network an inference of either long distance colonisation
with fragmentation or fragmentation followed by range expansion in Clade 3.1 is
obtained. Clades 2.1 and 2.3 show evidence of range expansion using independent tests
(R2 = 0.13531; P≤0.01 and R2 = 0.30728; P≤0.05 respectively), where clade 2.2 does
not (R2 = 0.34418; n.s). Using the tests for secondary contact outlined in Templeton
(2001) the BS site shows strong evidence of contact between divergent clades, with
some slight signal for the NRM site (Appendix 4c). Long distance colonisation is not a
realistic expectation for an animal of this size (up to 25mm); contiguous range
expansion is a more biologically realistic conclusion. Therefore the most likely
inference is past fragmentation across the Naturaliste Ridge and southern Blackwood
area followed by range expansion with secondary contact at BS and NRM.
Inferences for the SCL network include evidence for contiguous range expansion from
DF across the southern coast to KALS in clade 3.2. At the final nesting level for the
SCL network is an inference of either long distance colonisation with fragmentation or
past fragmentation with range expansion. Using independent tests for demographic
expansion clade 3.3 shows evidence for range expansion (R2 = 0.2662; P≤0.05) whereas
clade 3.2 does not (R2 = 0.14472; n.s), there is also no evidence for secondary contact
(Appendix 4c). Long distance colonisation is not considered feasible the inference due
to reasons outlined above; therefore past gradual expansion across the southern coast
followed by fragmentation is adopted as the appropriate biological inference.
103
Table 4.3: Biogeographical inferences for nested Metacrinia nichollsi clades from themain range and south coastal lineages with significant phylogeographic structure,specified by a χ2 nested contingency test. P-values are calculated from 1000 randompermutations and are considered significant if permuted expected χ2 values are greaterthan or equal to the observed. RGF – Restricted Gene Flow; IBD – Isolation byDistance; PF – Past Fragmentation; RE – Range Expansion; CRE – Contiguous RangeExpansion; PGRE – Past Gradual Range Expansion; F – Fragmentation; w/ - with.
Nested χ2 Permuted InferredClade P - value Process
3.2 <0.001 1-2-11-12 CRETotal Cladogram <0.001 1-19-20-2-11-12-13-21 PGRE w/ F
MR
LSC
L
Lineage Chain of Inference
104
Figure 4.3: Haplotype networks for 26 Metacrinia nichollsi ND2 haplotypes, created inTCS 1.21. Three distinct networks were created corresponding to the Stirling RangesLineage – SRL (A), the Main Range Lineage – MRL (B), and the South CoastalLineage – SCL (C). Each line represents a single mutational change. Ellipse size isproportional to haplotypes frequency with small open circles representing missinghaplotypes and the square representing the ancestral haplotype as inferred by TCS usingoutgroup weights. Connections up to 10 and 16 steps are within the 95% confidencelimits of a parsimonious connection for the SCL and MRL networks respectively. TheSRL differs from the MCL and SCL by 49 & 52 mutational steps respectively, while theMRL and SCL differ by 31 mutational steps. Clades are nested according to the rulesoutlined in (Templeton et al. 1987; Crandall 1994; Templeton et al. 1995).
4.4.3 Population genetic analysis
Table 4 is a summary of the population genetic analyses carried out on the MRL and
SCL within M. nichollsi separately, SRL data was not analysed in this manner due to a
lack of polymorphism. AMOVA results from the SCL of M. nichollsi show extremely
high levels of population structure, with 86% of genetic variation accounted for
between populations/catchments (each population of this lineage is in a different
catchment). Hudson’s Snn also corroborate these results, suggesting
populations/catchments are completely differentiated (Snn=1.000). The AMOVA
results for the MR Lineage of M. nichollsi show lower levels of population genetic
structure (accounting for 56% of the variation), with Snn suggesting populations are
panmictic (Snn=0.238). AMOVA analyses considering catchment groups within the
105
MRL show that there is much more variation among populations within catchments
(34%) compared to between catchment groups (10%).
Table 4.4: Summary table of population genetic statistics for the south coastal and mainrange Metacrinia nichollsi lineages observed in phylogenetic and phylogeographicanalyses. Analysis of Molecular Variance (AMOVA) results for each lineage arepresented separately. Within the main range lineage the distribution is divided up intocatchment regions, within the south coastal lineage discrete populations are in separatecatchment regions already. Hudson’s ‘nearest neighbour’ statistic (Snn) is also shownfor each lineage as a whole. P-values were calculated via 1000 permutations.
n.s = P>0.05; *** = P≤0.001
Source df SS MS Est. Var. % Stat ValueAmong Pops. 3 56.861 18.954 3.870 86%Within Pops. 15 9.350 0.623 0.623 14% φPT 0.861***Total South Coastal Lineage Snn 1.000***
Source df SS MS Est. Var. % Stat ValueAmong Catchments 6 14.472 2.412 0.103 10% φRT 0.099n.s
Among Pops./Catchment 4 7.233 1.808 0.353 34% φPR 0.374***Indiv./Within Pops. 28 16.500 0.589 0.589 56% φPT 0.436***Total Main Range Lineage Snn 0.238***
Main Range Lineage Population Genetic Analysis
South Coastal Lineage Population Genetics Analysis
106
4.5 Discussion
My study has clarified the historical factors that have lead to population diversity in this
mesic-adapted direct-developing frog. There are two major phylogenetic divergence
events within the M. nichollsi lineage that account for the majority of genetic diversity
observed. The first is the separation of the relictual Stirling Ranges populations (SRL)
from the remainder of the species range during the Late Miocene – Pliocene (Figure 2
& 4). The second splits the remainder of the species distribution into a lineage covering
the majority of the species range (MRL) and another with a disjunct distribution across
the south coast (SCL), with divergence estimates dating this split during the mid-late
Pliocene (Figure 4.2 & 4.4). The biogeographic history of these major divergence
events, followed by the phylogeographic history within the various lineages, is
examined with reference to how the climatic history of the southwest corner has
impacted on the current genetic structure of this species.
Figure 4.4: Biogeographic hypotheses relating to the Metacrinia nichollsiphylogeographic dataset. Hypotheses are generated from both the nested cladephylogeographic analysis results and the known geological and climatic history of theregion. MRL – Main Range Lineage; SRL – Stirling Ranges Lineage; SCL – SouthCoastal Lineage; PF – Past Fragmentation; RE – Range Expansion; • - sampledpopulations.
107
4.5.1 Isolation of the Stirling Ranges Populations
The isolated Stirling Ranges M. nichollsi population appears to have been separated
from the main range of the species in the Late Miocene – Pliocene, based on divergence
estimates. However, such estimates may be confounded by the lack of resolution within
the phylogenetic tree and the probable repeated bottlenecking of this population over
time, possibly leading to genetic drift and an overestimate in divergence and mutation
rates (Bromham, Penny 2003; Welch, Bromham 2005). A past severe bottlenecking
event is indicated by the complete lack of genetic diversity in the two Stirling Ranges
populations sampled (Nei et al. 1975). Despite these difficulties, a divergence beginning
around the late Miocene for separation of the Stirling Ranges lineage fits well with a
shift from a subtropical climate to one of semi-arid conditions throughout many inland
regions in the southwest at this time (Hopper, Gioia 2004). Arid conditions on the
Australian continent began approximately 10MYA in the northwest increasing in
intensity over time and reaching the southwest approximately 6MYA (Macphail 1997;
Dodson, Macphail 2004).
The Stirling Ranges is a subregion within southwestern Australia, with an extreme
diversity of plant species, many recently evolved (Dirnböck et al. 2002; Hopper, Gioia
2004). However, the Stirling Ranges is also home to many ancient species, as the high
topographical relief and ‘wet, moist’ upland regions and creek/gully systems provide an
island refuge of microhabitats for formerly widespread Gondwanan relicts (Main 1999;
2001). The distribution of one such Gondwanan relict group, myglamorph spiders,
exactly matches the distribution of M. nichollsi, with many species found in the Stirling
Ranges and others isolated on the extreme southwestern coast (Main 1999). The two
taxa are also often found in the same microhabitats (Main, B.Y. – pers. comm.). Such
persistence in these habitats in the southwest has been suggested to be a result of
contraction of higher rainfall to the southwestern coast, leading to a loss of rainforest
taxa, from the Late Miocene onwards (Archer 1996; Main 1999; 2001).
Given the history of the area, the distribution of M. nichollsi and associations with other
‘relictual’ taxa, we suggest that M. nichollsi was a formerly widespread species, with
isolation of the Stirling Ranges populations during the late Miocene-Early Pliocene
onset of aridity on the Australian continent. Today the species survives in a few gully
systems and mountaintops on the eastern side of the range (pers. obs.). The Stirling
108
Ranges lineage should be recognized as a distinct Evolutionary Significant Unit (ESU)
for conservation purposes because these populations are genetically distinct and
geographically isolated (which fits the definition of an ESU - (Moritz 2002)), and there
is also no genetic diversity within the Stirling Ranges lineage. Forecast climate change
predicts higher temperatures similar to those seen at the Plio-Pleistocene border
(Cronin, Dowsett 1993). However current trends suggest that warmer temperatures will
be associated with reduced rainfall (Bureau of Meteorology - http://www.bom.gov.au/),
which is likely to lead to increased fire frequency and intensity in the Stirling Ranges.
Such increases in fire frequency and intensity in the area, primarily human-induced,
have already been linked to population bottlenecks and local extinctions in other
relictual taxa occupying the same microhabitats as M. nichollsi (Main 1999). Therefore,
impending climate change is likely to seriously threaten the viability of not only this
relictual and distinct population of M. nichollsi, but also many of the other relictual
species currently found in the Stirling Ranges today.
4.5.2 Biogeography within the southwestern clades of M. nichollsi
Within the southwestern clades of M. nichollsi, our data reveal a complex distribution
of two relatively divergent haplotype lineages (2.64-3.41MY separation). The first
lineage covers the majority of the species main range, the second has a disjunct
distribution along the southern coast with the disjunct southern populations separated by
the Main Range lineage. No sharing of haplotypes from the two divergent lineages was
observed at any of the sampling locations in this study. There are no obvious
morphological differences between the two groups. There are several known significant
arid pulses from the mid-late Pliocene period in southwestern climate history, notably
palynological evidence points to two specific events at 2.6 and 2.9 MYA (Dodson,
Ramrath 2001; Dodson, Macphail 2004). These dates match our divergence estimates
for the two clades within the southwestern range of M. nichollsi closely, however there
are several scenario’s that are consistent with the biogeographic history and current
distribution of the SCL and MRL clades.
Pliocene arid events are likely to have led to isolated populations in the north and south
of the species range. In the south (SCL) the species is likely to have contracted towards
the coast, where rainfall remained high, albeit reduced, during arid cycles, as indicated
by a large number of relictual animals and plants (Hopper et al. 1996). The northern
109
(MRL) clade may have had individuals persisting in a variety of sites, namely Lake
Muir or parts of the upper Blackwood catchment (see Figure 4.1), which may have
remained wet enough during severe arid pulses in the Pliocene to preserve this northern
clade. The biogeographic pattern observed in each of these clades appears to be vastly
different. Within the SCL a pattern of restricted dispersal between catchment groups
exists and fragmentation between the Blackwood and Deep Rivers is inferred. Animals
in the MRL, on the other hand, show evidence of recent widespread dispersal across its
range with some restriction of gene flow across the Naturaliste Ridge and more
currently across the extent of the species range. A biogeographic interpretation of the
history of these two lineages is dealt with in turn.
Along the southern coast the SCL haplotypes are restricted to discrete groups based on
catchment, highlighting a probable role for catchments as important refugial areas
during periods of reduced rainfall. Finer-scale genetic studies within this region support
this showing localised genetic groups within catchments with limited dispersal between
groups (D. Edwards unpubl. data). Phylogeographic analyses suggest that there was
initial expansion from west to east in this lineage, initially from NRS east and then more
recently from DFS and KHS east to KALS, followed by fragmentation across the area
between the Scott River Coastal Plain (east of the Blackwood River) and the Pingerup
Plains (west of the Deep River). The Pingerup Plains also define the geographic break
between the ranges of Geocrinia rosea and G. lutea. It is thought that the Pingerup
Plains, with extremely waterlogged, swampy ground during winter drying rapidly in
spring into summer is incompatible with survival and reproductive success in these
wetter adapted species (Wardell-Johnson, Roberts 1993). The Scott Coastal Plain also
has a similar pattern of surface water levels in relation to seasonal rainfall (Strategen –
Information Series Report No. 1 2005) and intrudes between the ranges of G. alba / G.
vitellina and G. rosea (Wardell-Johnson, Roberts 1993). Metacrinia nichollsi is a direct
developer with less reliance on moisture in drainage systems than species in the G.
rosea complex but is dependant on available soil moisture in autumn, the driest season,
for breeding. Metacrinia nichollsi is more widespread in the forest system that species
in the G. rosea complex but is likely to be affected by similar soil moisture conditions.
One scenario that may explain the current distribution within the main range of M.
nichollsi is that SCL populations became extinct between the Naturaliste Ridge and the
Deep River during a climatic extreme, with expanding northern populations then
110
quickly moving to occupy available habitat. Extinction of the southern coastal
haplotype lineage may have been caused by severe arid pulses during the Pleistocene
(Galloway, Kemp 1981), or been caused by flooding by rising water table during high
sea level stands during the interglacials of this period (Sircombe, Freeman 1999).
Alternatively the northern haplotype group may have a selective advantage with
hybridisation only occurring between the northern females and southern males.
Regardless of the mechanism the MRL has a strong signal of recent expansion across all
of its current range. Assessing the extent of the species’ distribution and genetic testing
using both mitochondrial and nuclear markers in the potential ‘hybrid zone’ areas
should be done to rule out selective forces before any conclusions can be drawn about
the historical reasons for the current distribution of these two divergent lineages.
Phylogeographic inferences within the MRL suggest that dispersal is restricted between
Naturaliste Ridge Populations and those within the remaining range of the MRL, with
some secondary contact mainly at the BS site, and a smaller signal at the NRM site. It is
most likely that wetter periods throughout the Quaternary have lead to this trend
through higher rainfall and higher sea levels (Hodgkin, Hesp 1998; Sircombe, Freeman
1999), with dispersal and secondary contact occurring during drier times. The lower
Blackwood River and Scott River Plain swamps would have been unfavourable for the
species isolating the BS site from the Naturaliste Ridge Populations in the southern part
of the MRL range. These barriers may have also contributed to the long break between
the MRL and SCL in this region discussed above. In the north-western range of the
MRL Naturaliste Ridge populations were most likely isolated from the remainder of the
range of this lineage by higher sea levels are known to have lead to dramatic changes in
the coastline between these two areas (Sircombe, Freeman 1999; Hageman et al. 2003).
This dispersal route is unlikely to be open to the species regardless of climatic
conditions in the future due to the vast amount of agricultural clearing that has been
conducted in the area between these sites (Wardell-Johnson, Roberts 1993). The most
recent phylogeographic inference for this lineage is of restricted dispersal with isolation
by distance across the majority of the range of the MRL (excluding the Naturalist Ridge
sites). An inference of restricted dispersal, suggests that despite a relatively recent
dramatic expansion of the range of this lineage, that a trend of more restricted dispersal
in current times. This may also suggest that once established there is little impetus for
movement and that restricted dispersal may be more the rule than extensive movement.
111
4.5.3 Conclusions
Metacrinia nichollsi presents an important case study of the biogeography of the
southwestern Australian wet forest system. As expected the Stirling Ranges populations
appear to be relicts from a formerly widespread range, which has been dramatically
altered by climatic shifts from a tropical to more temperate/semi-arid climate in the late
Miocene-Pliocene. Arid pulses from the Pliocene to present are most likely associated
with the separation of the two lineages within the main range of M. nichollsi, a
contraction of the southern lineage to the coast, and a restriction of dispersal within and
between catchments. Potential refugial areas available to preserve the species in the
north are likely to have been in the vicinity of Lake Muir or along the Blackwood River
catchment. The processes that have led to the current distribution of these two disparate
lineages is less clear as a strong signature of extensive range expansion within MRL is
indicated and a distinction between a hypothesis of extinction (of SCL) followed by
colonisation (of MRL) vs. competitive exclusion, or one-way hybridisation requires
more extensive sampling and analysis using mitochondrial and nuclear markers
combined. Despite appearing to have disparate biogeographical histories, restricted
dispersal appears to be more the rule in this direct developing species, a phenomenon
common in other direct developers in the southwest (Driscoll 1997; 1998a; b) and
abroad (Crawford 2003). Metacrinia nichollsi appears to be dramatically affected by
climate, in particular rainfall levels, and the importance of drainage systems (and other
wet areas) as refugia along the southern coast of southwestern Australia. This is in spite
of the view that climate has not varied as much in coastal regions, as has been suggested
for the transitional rainfall areas to the north and east.
112
113
Lea’s Frog
Small frog with a big past….
114
115
Chapter 5:
The Phylogeography of Geocrinia leai
(Lea’s Frog)
5.1 Abstract
The diverse, endemic southwestern Australian biota, combined with high levels of
human disturbance, have made southwestern Australia a biodiversity hotspot of global
importance. To conserve this regional diversity, there needs to be an understanding of
how it has evolved and consequently how it might react to future pressures.
Phylogeographic studies on endemic myobatrachid frogs with direct and conventional
aquatic development have shown that arid periods are critical drivers of divergence,
with isolation on major drainage systems a recurring pattern along the southern coast of
Western Australia. Geocrinia leai deposits eggs on land above water, but has an aquatic
free-swimming tadpole: the third life history pattern for frogs in this region. A
comprehensive phylogeographic dataset comprising 50 ND2 sequences across the range
of the species uncovered three deeply divergent lineages (3.8-5.3MYA), one large
lineage along the western coast and two others along the southern coast. Divergences
are consistent with species-level breaks seen in other endemic myobatrachids, and
estimates suggest that lineage separation may be associated with the Late Miocene onset
of aridity in Australia. Subsequent within-lineage structure appears to have developed
from the Plio-Pleistocene to present, and is likely to be associated with intense climatic
fluctuations between arid and mesic climates during this time. There is consistent
evidence of dispersal between northern Darling Escarpment and southern coastal
refugia, and strong catchment based genetic structure along the southern coast. These
data suggest that diversity within southwestern Australian forests may be severely
underestimated, and in view of habitat destruction levels and predicted climate change,
there is a need to conduct more research into the biogeographic history of the forest
biota of southwestern Australia.
116
5.2 Introduction
Southwestern Australia provides an important system for the study of speciation and
biogeography. It is unusual because it has been geologically stable since the Tertiary,
and it has limited topographical relief (Hopper et al. 1996; Hopper, Gioia 2004).
Despite the subdued topography and geological stability, the region is a hotspot of
biodiversity and endemism of global significance due to the extreme diversity of plants
and highly threatened ecosystems (Cincotta et al. 2000; Myers et al. 2000). Regional
patterns of diversity and species richness of southwestern Australian plants have, to a
certain extent, been described (Gioia, Piggott 2000; Dirnböck et al. 2002). However, it
is essential to understand processes that have generated diversity and endemicity, to
ensure ongoing conservation of pattern as well as process (Moritz, Faith 1998; Moritz et
al. 2001; Moritz 2002). Hypotheses regarding the speciation of the highly diverse
endemic flora have focussed on more transitional climatic zones, in inland and
northwetern and southeast coastal regions, rather than on the forested areas of
southwestern Australia (Hopper 1979; Hopper, Gioia 2004). A complex interaction
between Pleistocene climatic fluctuations and landscape evolution is thought to have led
to the explosive speciation of southwestern Australian endemic plant species in these
regions (Hopper 1979; Hopper, Gioia 2004). However, little work has been conducted
on the processes generating diversity and endemicity in the fauna of the southwest, and
studies of plant and animal taxa covering the forest system in the southwest are in
general lacking.
The Myobatrachidae, an Australo-Papuan endemic frog family, are particularly diverse
in the southwest, and some species have featured heavily in the generation of
biogeographic hypotheses between the east and west of Australia (Roberts, Maxson
1985b; a). There are many endemic, monotypic and relictual myobatrachids within the
southwest (Roberts et al. 1997), largely concentrated along the mesic southwestern
coast and reflecting the ancient history of this area in particular. Investigations generally
have focussed on the large amount of diversity and endemicity observed in the
Heleioporus, Crinia, Geocrinia and Neobatrachus genera. Speciation within most of
these genera is thought to have occurred in situ (Morgan et al.; Main et al. 1958;
where polyploidy has played a role in speciation of southwestern Australian endemics
(Roberts 1997).
117
Geocrinia comprises seven species, five of which are endemic to southwestern
Australia, with another two species in southeastern Australia. Systematic studies on
most of the species in the genus have shown two sister lineages, which vary in life
history strategy. The G. rosea species complex comprises four allopatric, direct
developing species distributed across the extreme southwestern Australian coast all of
which are geographically restricted and have highly specialised habitat preferences
(Wardell-Johnson, Roberts 1993). The sister lineage comprises species with terrestrial
oviposition, with an obligate free-swimming aquatic tadpole. Species in this lineage
include G. leai, a southwestern endemic, and the G. victoriana/G. laevis complex,
which is endemic to southeastern Australia (Read et al. 2001). Geocrinia is the only
genus to have received any comprehensive treatment to clarify speciation mechanisms
within southwestern Australia, and only species within the Geocrinia rosea species
complex have been considered.
Species in the G. rosea complex are thought to have formed through peripheral isolation
of allopatric populations over a 200km range across the mesic, relictual southwest
coastal forest system (Wardell-Johnson, Roberts 1993; Roberts, Wardell-Johnson
1995). Allozyme studies show genetic groups within each species are associated with
drainage systems, these data also provide evidence for multiple range expansions,
contractions and shifts possibly in response to historic climatic fluctuations (Driscoll
1998a; b). Species in the G. rosea complex have direct-developing eggs and very
restricted dispersal (Driscoll 1997; 1998a; b), contrasting with other southwest frog
species with aquatic tadpoles and much more extensive dispersal capabilities (Berry
2001; Davis, Roberts 2005). The majority of the genetic studies on southwestern
Australian frogs point to climatic change as an important factor in shaping the
biogeographic history of the region and genetic diversity within each of these species.
Biogeographic history and mechanisms of speciation should be assessed in a diversity
of species with varying life history strategies and habitat requirements, so general rather
than species-specific patterns emerge (Cracraft 1988{Riddle, 2000 #491; Moritz et al.
2001; Zink 2002; Wiens, Donoghue 2004).
Most southwestern endemic myobatrachid species are heavily reliant on seasonal
rainfall regimes, yet are varied in the ecological and life history strategies that have
developed in response to this rainfall regime. There are three life history strategies
118
within the endemic southwestern Australian myobatrachid fauna: direct development
with endotrophic eggs, terrestrial oviposition with an obligate free-swimming tadpole
and conventional aquatic oviposition and tadpole development (Roberts, Watson 1993).
Biogeographic studies on southwestern Australian endemic myobatrachid species with
both direct and aquatic development have shown varied responses to primarily climatic
and associated rainfall fluctuations. Geocrinia leai is an old lineage within the
southwest (Read et al. 2001) whose range covers the entire forest system, thus
overlapping with that of most of the species already studied. Given the age of the
lineage, G. leai is likely to have experienced multiple climatic changes occurring from
the Miocene to present. Also within G. leai there is great potential for downstream
tadpole dispersal, which contrasts with other catchment-based species with little
dispersal (e.g. G. rosea complex). Movement between catchments may be limited in G.
leai, therefore catchment based patterns of population genetic structure are likely to be
more prominent in this species compared to other species with aquatic larvae already
studied. In order to allow for a potential diversity of responses related to life history and
sensitivity to climatic change I compiled a phylogeographic dataset for Geocrinia leai.
For this study I sequenced a 1120bp fragment of the mitochondrial ND2 gene from 50
animals from fourteen sites across the range of the species.
5.3 Materials and Methods
5.3.1 Tissue samples
A total of 50 individuals were sampled (toe-clips) from 14 sites across the entire species
distribution with 3-4 animals per site (Figure 5.1, Table 5.1). Geocrinia victoriana
(37°49´ 146°10´) and G. laevis (37º36´ 140º28´) were used as outgroup taxa for the
intraspecific phylogenetic analysis of G. leai, as previous phylogenetic studies have
suggested these two species are the sister group to G. leai (Read et al. 2001).
119
Table 5.1: Geocrinia leai phylogeography sampling location names, abbreviations andexact GPS coordinates in degrees minutes, seconds. All points are in Geodetic WGS 84.
Figure 5.1: Map of the southwestern Australian coastline with map of Australia inset.Tissue collection locations [•] for the Geocrinia leai phylogeographic study cover theentire known distribution of the species. Refer to Table 5.1 for further information onsample sizes, abbreviations and exact locations.
0.6679, were enforced in a likelihood analysis with 100 bootstrap replicates to assess
branch support. The maximum likelihood tree (Figure 5.2) shows three major lineages
within G. leai corresponding to a western lineage (WL), a lineage isolated to the
Shannon-Gardner catchment group (SGL) and a southeastern coastal lineage (SGL)
(Figure 5.2).
125
Figure 5.2: Maximum likelihood phylogram of Geocrinia leai ND2 haplotypesshowing three major lineages and with G. victoriana and G. laevis as outgroup taxa.The sites each haplotype was found at are indicated, # indicated a haplotype with afrequency of two. Support for clades is given by ML bootstrap/Bayesian PosteriorProbability values. Map of southwestern Australia is shown with shaded areasrepresenting the distribution of the Western, Southeast Coastal and Shannon/GardnerLineages, for site name references refer to Table 1. * support values = 86/100, yet forpresentation reasons values are not shown. // indicates that branch lengths have beenshortened for presentation purposes.
Pairwise sequence divergences between the WCL and SGL ranged from 5.36-6.25%
and between the WL and SCL ranged from 4.46-5.53% (for complete uncorrected p
sequence divergence table see Appendix 2d). Similarly sequence divergence estimates
between the SGL and SCL ranged from 4.82-6.34%. Pairwise sequence divergence
between the outgroup taxa, G. laevis and G. victoriana, are 4.4-4.5% (Figure 5.2). The
WL includes all populations from the southwest corner (west of DW – Figure 5.2) and
all populations from the Darling Escarpment, within this lineage sequence divergences
ranged from 0.09-2.68%. The WL contained 36 haplotypes, with Hd = 0.998 ± 0.007
and π = 0.01601 ± 0.00081. The SGL is represented only by the individuals sampled
126
from that site with low sequence divergence, between 0.18-0.36%. The SGL contained
3 haplotypes and Hd = 1 ± 0.272 and π = 0.00238 ± 0.00079. The SGL includes
populations from DF east to KAL and contains haplotypes that have divergences
between 0.09-1.07%. The SCL contained 9 haplotypes, Hd = 0.978 ± 0.054 and π =
0.03121 ± 0.00137.
The score of the likelihood tree without enforcing a molecular clock was –InL =
2968.3700, the score for the tree enforcing a molecular clock was –InL = 2995.8845.
The likelihood ratio tests showed that sequences did not depart from a clock like model
of evolution (P=0.169927; n.s). The average number of nucleotide substitutions per site
(dA) between WL and SGL was 0.05104 ± 0.00739, providing a divergence estimate of
5.3MYA ± 772,000yrs between these two lineages. Between SCL and SGL was dA =
0.04977 ± 0.01336 and therefore divergence between these two lineages is estimated at
5.2MYA ± 1.4MYA. Divergence between the WL and SCL was estimated to have
occurred 3.8MYA ± 393,000yrs (dA=0.03658±0.00377). Divergences between several
minor clades within the WL several also had strong support corresponding to distinct
haplotype networks - WL main (Figure 5.3) vs. DW/BN (Figure 5.4A) = 1.11MYA ±
270,000yrs (dA=0.01064±0.00258); WL main vs. BS (Figure 5.4B) = 1.41MYA ±
304,000yrs (dA=0.01349±0.00291); DW/BN vs. BS = 1.51MYA ± 503,000yrs
(dA=0.01442±0.00507). As did the minor clades within the SCL – DF/KH vs. KAL =
Intraspecific analysis techniques were used to provide further detailed information on
the biogeographic and historical inferences contained in the data. The TCS 1.21 output
for the whole G. leai dataset showed 6 separate networks at the 95% probability of
parsimonious connection. One network corresponded to the SGL and two networks
were specific to the SCL, biogeographic interpretations were not assessed due to lack of
geographical variation for the SGL network and inadequate sampling for the SCL
networks. Within the WL there were three separate networks, the first corresponded to
the majority of the range of the WL lineage (Figure 5.3) and the two others were
specific to catchments along the extreme southwestern coast (Figures 5.4A and 5.4B).
The three networks were joined at the highest nesting level (Figure 5.5) to attain overall
127
biogeographic inferences for the whole WL. Several significant inferences were
identified using the Templeton (2004) inference key, for significant geographic analysis
results and the chain of inference for each of these clades see Table 5.2 (for complete
GeoDis output refer to Appendix 3d).
Figure 5.3: Haplotype network for the majority of the Western Lineage Geocrinia leaiND2 haplotypes. Twenty-six haplotypes (comprising clade 6.1) and the site they weresampled from are included. See Figure 4 for remaining haplotype networks and Figure 5for overall nesting design of all G. leai WL haplotype networks. Each line represents asingle mutational change. Ellipse size is proportional to haplotype frequency; with smallopen circles representing missing haplotypes and the square representing the ancestralhaplotype inferred by TCS using outgroup weights. Connections up to 17 steps arewithin the 95% confidence limits of a parsimonious connection. Clades are nestedaccording to the rules outlined in (Templeton et al. 1987; Crandall 1994; Templeton etal. 1995).
128
Figure 5.4: Remaining haplotype networks for the Western Geocrinia leai Lineage. SeeFigure 5 for overall nesting design for the entire WL. Haplotypes are shown along withthe site they were sampled from; Figure 4a includes the network of 6 haplotypes forclade 6.2 and Figure 4b includes the network of 4 haplotypes for clade 6.3. Connectionsup to 17 steps are within the 95% confidence limits of a parsimonious connection. Eachline represents a single mutational change. Ellipse size is proportional to haplotypefrequency; with small open circles representing missing haplotypes and the squarerepresenting the ancestral haplotype inferred by TCS using outgroup weights. Cladesare nested according to the rules outlined in (Templeton et al. 1987; Crandall 1994;Templeton et al. 1995).
129
Figure 5.5: General overview of the nested design for the three individual haplotypenetworks created for all 36 haplotypes from the Geocrinia leai Western Lineage. Clade6.1 differs from clades 6.2 and 6.3 by 15 and 17 mutational steps respectively. Cladesare nested according to the rules outlined in (Templeton et al. 1987; Crandall 1994;Templeton et al. 1995).
Despite relatively fine-scale sampling within G. leai, several inferences could not be
resolved for clades (Clades 4.1 and 5.2 – refer to Figures 5.3 and 5.5 respectively) due
to inadequate geographical sampling to differentiate between allopatric fragmentation
and other biogeographic scenarios. Clade 5.1 shows evidence for either past
fragmentation or long distance colonisation. Supplementary testing shows evidence for
demographic range expansion of clade 4.1 (R2=0.16052; P≤0.05), but not for clade 4.2
(R2–n.s; P>0.05). Long distance movement of up to 200km is unlikely for this small and
presumably short-lived species, other Geocrinia species are also short-lived (Conroy,
Brook 2003). A more biologically realistic inference is gradual range expansion from
the northern Darling Escarpment (SA, SP & MUR) region into the southwestern coastal
catchments (Upper Blackwood & Collie systems) with subsequent fragmentation.
130
Table 5.2: Biogeographic inferences for Geocrinia leai western lineage clades fromthe Western G. leai Lineage (WL) with significant phylogeographic structure,specified by a χ2 nested contingency test. P-values are calculated from 1000 randompermutations and are considered significant if permuted χ2 values are greater than orequal to the observed.
IGS – Inadequate Geographic Sampling; GRE – Gradual Range Expansion; F – Fragmentation;
AF – Allopatric Fragmentation; w/ - with
Clade 6.1 (Figure 5.5 & 5.6), containing haplotypes from the majority of the WL range,
shows evidence of either long distance colonisation with fragmentation or past
fragmentation followed by range expansion. Using the tests for secondary contact
outlined in Templeton (2001), large clade distances within the SA, COL, BM and to a
lesser extent BN can be observed, suggesting secondary contact of divergent lineages
within these sites (refer to Appendix 4d). The entire clade itself does not show any
independent evidence of demographic range expansion (R2–n.s), and neither clade 5.1
nor clade 5.2 show significant expansion (R2–n.s). As argued above an individual long
distance movement of this species is unlikely. Therefore, gradual range expansion from
the northern Darling Escarpment into southwestern catchment areas (including further
expansion into the Naturaliste Ridge (NR) area on the Margaret River) is inferred with
subsequent fragmentation. At the total cladogram level there is evidence of allopatric
fragmentation between all the separate networks within the western lineage of G. leai
and a very strong signature of secondary contact in the Upper Blackwood (BN site).
5.4.3 Population genetic analysis
Table 5.3 presents a summary of population genetic analyses carried out on two of the
three lineages within the G. leai dataset. AMOVA results in the western lineage show
that dividing up the range of this lineage into a Darling Escarpment region and distinct
Nested Permutation Chain of InferredClade P -value Inference Process
4.1 <0.05 1-19-20 IGS5.1 <0.05 1-2-3-5-15-21 GRE w/ F5.2 <0.01 1-19-20 IGS6.1 <0.01 1-2-11-12-13-21 GRE w/ F
Total Cladogram <0.001 1-2-3-4-9 AF
131
southern coastal catchment regions accounts for a moderate amount (30%) of the
genetic variation within this lineage. Estimates suggest that population divergence
within this lineage is relatively high, but suggestive of some dispersal among sites
(Snn=0.743). AMOVA results for the SCL shows much more marked patterns of
catchment based population genetic structure (89% of genetic variation) with high
levels of population divergence (Snn=1). Individuals within populations of the SCL
account for very little of the genetic variation within this lineage (11%).
Table 5.3: Summary table of population genetic statistics for the western and southeastcoastal Geocrinia leai lineages observed in phylogenetic and phylogeographic analyses.Analysis of Molecular Variance (AMOVA) results for the western lineage are dividedup into the following regions: the Darling Escarpment (SA, SP, MUR & HW), and theninto each of the separate catchments of the southwestern coast (COL, NR, Blackwood(BS, BM, BN) & DW) within the western lineage. AMOVA analyses of the southeastcoastal lineage were already sampled from each individual catchment therefore regionswere not defined. Hudson’s ‘nearest neighbour’ statistic (Snn) is also shown from eachlineage as a whole. P-values were calculated via 1000 permutations.
Three deeply divergent lineages were uncovered, one broadly distributed throughout the
western portion of the species’ range (WL) and another two along the southern coast
(SGL & SCL) (Figure 5.2). Divergence estimates place the separation of these three
major lineages in the Late Miocene Early Pliocene (3.8-5.3MYA), while structure
Source df SS MS Est. Var. % Stat ValueAmong Regions 4 154.939 38.735 2.996 30% φRT 0.300***Among Pops./Regions 5 84.788 16.958 3.923 39% φPR 0.560***Indiv./Within Pops. 27 83.083 3.077 3.077 31% φPT 0.692***Total Western Lineage Snn 0.743***
Source df SS MS Est. Var. % Stat ValueAmong Pops./Regions 2 59.433 29.717 8.695 89%Indiv./Within Pops. 7 7.167 1.024 1.024 11% φPT 0.895***Total Southeast Coastal Lineage Snn 1.000***
Southeast Coastal Lineage Analysis of Molecular Variance
Western Lineage Analysis of Molecular Variance
132
within lineages appears to have developed throughout the Plio-Pleistocene period (1.1-
1.73MYA). These periods are characterised by an initial onset of aridity and climatic
fluctuations between arid and mesic condition in the southwest respectively (Figure
5.6A). Phylogeographic analyses, able to be conducted only on the WL, suggest a series
of fragmentation and range expansion events have occurred within this lineage,
particularly between the upper Darling Escarpment and southern catchment regions
(Figure 5.6B). The results of various analysis techniques suggest there is a pattern of
catchment based genetic structure along the southern coast across the whole range of G.
leai. The biogeographic history of major divergence events and phylogeographic
structure within lineages is examined below with reference to the known climatic
history of southwestern Australia and the reconstructed biogeographic history of other
taxa endemic to the region.
133
Figure 5.6: Biogeographic hypotheses relating to the Geocrinia leai phylogeographicdataset. Figure 6A present specific hypotheses relating to the processes leading to thedevelopment of the major G. leai lineages and several minor clades within theselineages from the Late Miocene through to the Plio-Pleistocene Border. Figure 6Bpresents specific biogeographical hypotheses relating to the phylogeographic structuredevelopment within G. leai during the Pleistocene. Hypotheses are generated from boththe nested clade phylogeographic analysis and the known geological history of theregion. WL – Western Lineage; SGL – Shannon/Gardner Lineage; SCL – SoutheastCoastal Lineage; 2º - Secondary. • - sampled populations.
134
5.5.1 Broader phylogenetic pattern within G. leai
The three major and apparently allopatric lineages within the range of G. leai differ by
between 4.5 and 6.3% uncorrected p sequence divergence. This level of sequence
divergence is the same as between other sister species of endemic southwestern
Australian myobatrachids in both the Heleioporus genus (Morgan et al.) and the G.
rosea species complex (Read et al. 2001). Geocrinia laevis and G. victoriana, the
outgroup taxa used in this study, also show divergences just below those observed
between the G. leai lineages and are known to hybridise (Gollmann 1991; Scroggie,
Littlejohn 2005). There are currently no known morphological differences between the
G. leai lineages and nothing is known about whether the distinct lineages overlap
geographically or whether they hybridise. Geographically the split between the WL and
SGL matches up with a known dichotomy between specific genetic groups within G.
rosea from the Shannon/Gardner and Warren/Dombakup/Donnelly catchment areas’,
characterised by fixed differences in many allozyme loci (Driscoll 1998b). Divergence
between the SCL and the SGL also corresponds geographically to a species level split
between G. rosea and G. lutea, which are thought to be sister species (Wardell-Johnson,
Roberts 1993).
Divergence estimates suggest that the three-way spilt within G. leai occurred during the
Late Miocene-Early Pliocene period, which is considerably earlier than major
divergences estimated within M. nichollsi (Chapter 4) and C. georgiana (Chapter 3).
While there are issues with molecular clock estimates (Rambaut, Bromham 1998),
evidence suggests that the gene region used in the current study evolves in a clock-like
manner. The dates obtained for divergence between the major G. leai lineages also link
to known climatic changes within southwestern Australia. The late Miocene / early
Pliocene was a period of dramatic climate change throughout the Australian continent.
During the late Miocene arid condition intensified and rapid drying occurred throughout
the southwest between 3 and 5MYA (Dodson, Macphail 2004). Geocrinia leai is a
relatively old lineage (Read et al. 2001), which today is generally associated with wet
drainage system areas during breeding (Tyler et al. 2000) and non-breeding times (D.
Edwards, pers. obs). A shift towards an increasingly arid climate in the region is likely
to have isolated and fragmented populations of G. leai to refugial wetter riverine
catchments and aquifer fed springs along the southern coast, similar to those still
135
housing many Gondwanan plant and animal species (Wardell-Johnson, Roberts 1993;
Hopper et al. 1996; Wardell-Johnson, Coates 1996).
Allozyme data were used to assess the within G. rosea split (Driscoll 1998b), and G.
lutea was not included in the taxonomic assessment of the Geocrinia genus (Read et al.
2001), making comparable timing estimates of congruent splits seen in G. leai
impossible. However, investigations within the G. rosea complex have identified
processes that may be relevant in explaining the biogeographic history of G. leai.
Potential edaphic barriers were identified between both genetic groups within G. rosea
and between G. rosea and G. lutea, in the form of swampy soils that have great fluxes
in soil moisture in association with seasonal rainfall and fluctuating climates (Wardell-
Johnson, Roberts 1993). Dramatic fluxes in the moisture levels of these soils are not
congruent with successful recruitment in direct developing Geocrinia species (Wardell-
Johnson, Roberts 1993) and may also play a role in fragmenting G. leai. Models
developed to explain speciation within the G. rosea complex suggest allopatric
speciation has occurred through climatic peripheral isolation of populations across
edaphic barriers (Wardell-Johnson, Roberts 1993). The biogeographic histories of
individual species also show evidence of climatic induced range expansion/contraction
and range shifts (Driscoll 1998a; b). The same processes are likely to be involved in the
development and maintenance of major phylogenetic structure within the G. leai
lineage.
5.5.2 Phylogeographic pattern within G. leai
Across the range of G. leai catchments along the southern coast are in general discrete
genetic entities, there is strong catchment based genetic structure in the SCL and
allopatric fragmentation of southern catchment regions within the WL (Figure 5.6A).
Divergence estimates suggest Plio-Pleistocene timing for many of these fragmentation
events. Repeated episodes of range expansion between the northern Darling Escarpment
(SA, SP & MUR) region and the southwestern catchments (Upper Blackwood (BN &
BM), Collie (COL) & Margaret (NR) River systems) with subsequent fragmentation
were inferred within the WL also. A similar pattern of range expansion and subsequent
fragmentation between the northern Darling Escarpment and the southwestern rivers has
been observed within Crinia georgiana (Chapter 3). Such tracks of dispersal and
patterns of fragmentation are most likely related to the severe fluctuating climates of the
136
Plio-Pleistocene to present leading to dramatic changes in rainfall, particularly through
inland regions (Macphail 1997; Hopper, Gioia 2004). Palynological studies have shown
that even during arid maxima mesic plants were able to survive in the northern Darling
Escarpment area, supporting the notion that this area is an ideal refuge area during dry
climates (Macphail 1997).
The Plio-Pleistocene period was characterized by the commencement of increasing
frequency and intensity of arid pulses separated by mesic interglacial periods, and
associated rainfall fluctuations in Australia (Bowler 1976; Kershaw et al. 1991;
Macphail 1997). The divergences between clades within the Western and Southeast
Coastal G. leai lineages fall within this Plio-Pleistocene period, and are consistent with
dramatic changes seen in other southwestern Australian biota (Hopper 1979; Rabosky et
al. 2004), including frogs (M. nichollsi – Chapter 4; C. georgiana – Chapter 3), and
may be linked to known climatic changes (Galloway, Kemp 1981). Arid periods are
likely to have fragmented populations of G. leai into more mesic pockets both along the
southern coast and upper Darling Escarpment. During interglacial periods increases in
rainfall levels shifted the inland border of Hopper & Gioia’s (2004) High Rainfall Zone
far into currently semi-arid parts of the southwest. More mesic interglacial conditions
have probably allowed for the repeated episodes of dispersal between northern and
southwestern refugium. Wetter climates would have allowed G. leai to extend its
distribution inland to where the upper reaches of southwestern catchments (see Figure
5.1) meet (Beard 1999). Adult movement across catchments and tadpole movement
within catchments during interglacials could easily have been responsible for the
dispersal patterns observed.
5.5.3 Geocrinia leai and the biogeography of southwestern Australia
The results of this study suggest that climate driven processes throughout the Late
Miocene and Plio-Pleistocene may be heavily involved in the phylogeographic history
of G. leai. Biogeographic studies on the endemic flora have mainly focussed more on
the Pleistocene explosive speciation in more marginal climatic zones in response to
climatic fluctuations (Hopper 1979; Hopper, Gioia 2004). However, minimal
investigation within the high rainfall zone on the southern coast of Western Australia
has shown that many Gondwanan relict species still persist (Hopper et al. 1996). Many
ancient wetland monocotyledon, wet Eucalypt and mycorrhizal species (Wardell-
137
Johnson, Coates 1996), Myglamorph spiders (Main 1996), Onychophorans (Hopper et
al. 1996) as well as relictual (Roberts et al. 1997) and wet restricted frogs (Driscoll
1998a; b) survive in perched swamps, aquifer fed springs and permanently wet riverine
areas along the southern coast. Yet few have utilised genetic data. Those that have are
generally based on allozyme markers, and Pleistocene models of speciation often are
invoked to explain diversity amongst these relict species (Wardell-Johnson, Coates
1996; Coates, Hamley 1999). The data in this chapter suggest that divergence may be
considerably earlier in many of these species, and they also highlights the importance of
obtaining divergence estimates to consider approximate timing of events.
5.5.4 Conclusions
Climatic cycles appear to have played an important role in shaping the biogeography of
G. leai and other southwestern Australian endemic amphibians. The age of the G. leai
lineage and heavy reliance on sufficient rainfall for adult and juvenile survival success
lends support to the notion of biogeographic structure developing in response to
climatic change from the Late Miocene to the present. Divergence estimates for G. leai
are considerably earlier than for other amphibians for which data exist, highlighting the
potential sensitivity of G. leai to changes in rainfall. More extensive sampling is
required to uncover the exact details of the complex biogeographic history of G. leai,
however evidence does point to the northern Darling Escarpment and individual
southern catchment areas as important refugia for G. leai, and other taxa, during the arid
maxima of the Late Miocene - Pliocene and Quaternary. While biogeographic studies to
date have focused on the more transitional climatic zones in the southwest, diversity
both within and between taxa in the higher rainfall areas may be severely
underestimated. The impacts of future climate change (Hughes 2003) also may be felt
more keenly within these taxa. This, combined with high levels of habitat destruction
limiting the ability of species to shift ranges in response to such change (Wardell-
Johnson, Roberts 1993), creates concern for the persistence of the higher rainfall
regions taxa. The sensitivity to changing climates, combined with the impact of human
occupancy and habitat destruction in the southwest, highlights the need for greater
understanding of the historical responses of many more species in the southwestern
Australian forests and highlights the extreme southwestern coast as a biodiversity
hotspot for endemic amphibians.
138
139
A few closing remarks….
140
141
Chapter 6:
General Discussion & Future Directions
6.1 The late Miocene as a time of speciation for southwesternAustralian frogs
Three out of the four species studied showed significant phylogenetic breaks with
divergence estimates dated to the Late Miocene period. The locations of these breaks
and their divergences are outlined below, with a brief description of the biogeographic
hypothesis generated to explain them. The use of molecular clock rates can be
controversial (Rambaut, Bromham 1998; Pulquerio, Nichols 2007), however each
individual dataset was shown to be evolving in a clock-like manner.
1) Arenophryne rotunda has diverged into a northern and a southern lineage
(Figure 6.1), which subsequent morphological analysis has shown is a
species-level split, and descriptions are forthcoming. The geographic
position of the break between the two species corresponds to the northern
border of the Victoria Plateau. The divergence between the two (~5.63MYA
± 0.41MYA) is most probably due to a combination of increasing aridity and
tectonic activity resulting in uplift of the Victoria Plateau (Hocking et al.
1987), forcing a distributional contraction westwards and disrupting
sandplain habitats respectively.
142
Figure 6.1: Late Miocene divergence of Arenophryne rotunda northern andsouthern lineages.
2) Metacrinia nichollsi has diverged into the three major lineages, the main
range lineage (MRL), southeast coastal lineage (SCL), and Stirling Ranges
lineage (SRL). Divergences between the SRL and the MRL (4.74MYA ±
0.33MYA) and SCL (5.19MYA ± 0.55MYA) were dated to the late Miocene
– early Pliocene period (Figure 6.2). Stirling Ranges populations (SRL) are
likely to have been isolated from the remainder of the species range by the
onset of aridity sweeping in from the northwest and beginning ~6MYA
(Dodson, Macphail 2004).
143
Figure 6.2: Late Miocene isolation of Stirling Ranges Metacrinia nichollsipopulations from the remainder of the species range.
3) Geocrinia leai has diverged into three major lineages, along the western
coast (WL), in the Shannon/Gardner catchment (SGL) and along the rest of
the southeast coast (SCL). Divergences between the SGL and the WL
(5.3MYA ± 0.77MYA) and SCL (5.2MYA ± 1.4MYA), are dated to the late
Miocene period (Figure 6.3). Divergence between the WL and SCL is dated
a little later in the early Pliocene (3.8MYA ± 0.39MYA). Lineages probably
were initially fragmented by the onset of aridity. Later divergence estimates
between the WL and SCL may be accounted for by reestablishment of
dispersal during the mid-Pliocene mesic period between the upper
Blackwood and Frankland river catchments (Dodson, Macphail 2004) as the
tops of these catchments are relatively close together (Figure 6.3).
144
Figure 6.3: Fragmentation of Geocrinia leai lineages during the lateMiocene - early Pliocene.
The Late Miocene was a period of climatic change in southwestern Australia, with an
initial onset of aridity and an associated decline in rainfall levels and predictability
(Galloway, Kemp 1981; Macphail 1997; Dodson, Macphail 2004). This change began
in the northwestern region of Western Australia and then swept southwards,
intensifying in the upper southwest ~6MYA. The relative timing of each of these
divergences could be accounted for easily, given that aridity was moving in a
southwards direction (Dodson, Macphail 2004). Furthermore, G. leai is likely to be
more sensitive to changes in rainfall levels and predictability than M. nichollsi, due to
its requirements for wet egg deposition sites associated with surface water for tadpole
development. Later divergence estimates between the WL and SCL within G. leai may
be accounted for by re-establishment of dispersal between the two clades on the upper
Blackwood and Frankland River catchments during the mid-Pliocene mesic period
(Dodson, Macphail 2004) as the tops of these catchments are relatively close together
(Figure 6.3). The relative order of events is necessarily speculative. Without accurate
calibration of a molecular clock, we may never know the exact timing of divergences
within these taxa.
My preferred interpretation of the data in this thesis is that the Late Miocene was
probably a key period for the divergence of major changes within southwestern
Australian frog genera. My data are consistent with estimated divergence dates in the
145
genus Heleioporus (Morgan et al.), but contrary to some of those patterns in that the
species studied here show local divergence has not led to overt speciation and sympatry
of lineages.
6.2 Plio-Pleistocene climatic fluctuations shape the biogeography ofsouthwestern Australian frogs
There are a number of divergences either within or between the major lineages of each
of the species used in this study that occur during the Pliocene (~5-1.65MYA) and after
the Plio-Pleistocene border (~1.64MYA).
1) Divergence between the southern Arenophryne lineage clades (2.05MYA ±
0.42MYA) either side of the Murchison Gorge (North Murchison Gorge –
NMG & South Murchison Gorge – SMG – Figure 6.1) is probably more
related to tectonic activity and the finial incision of the Murchison Gorge
(Hocking et al. 1987), than any arid pulses. Phylogeographic analysis of the
northern lineage data does show evidence of range expansion and
fragmentation events, which are most likely related to coastal dune building
and sea level transgressions coincident with climatic fluctuations of the
Pleistocene (Hocking et al. 1987).
2) Divergences between major lineages within C. georgiana (1.49MYA ±
0.23MYA – see Figure 6.4) are largely congruent with Hopper and Gioia’s
(2004) distinction between the High Rainfall Zone (HRZ) and the Southeast
Coastal Zone (SECZ). Lineages are likely to have been fragmented across
the southeast coast during a particularly intense arid pulse just after the Plio-
Pleistocene border (Galloway, Kemp 1981; Kendrick et al. 1991) and the
beginning of 100,000 year arid cycles (Rutherford, D'Hondt 2000), with
subsequent range expansion through inland areas during interglacial wet
periods thereafter. Repeated fragmentation and dispersal events are indicated
between populations from the upper Darling Escarpment (MO, SA, SP &
MUR – Figure 6.4) and populations on the southern coast within Lineage 1.
Restricted dispersal amongst southern populations may be associated with
fluctuating climate/sea levels and associated landscape changes.
146
Figure 6.4: Divergence of Crinia georgiana lineages during the Plio-Pleistocene.
3) The estimated divergence between the two lineages (MRL & SCL see Figure
6.2 for geographic location) within the southwestern range of M. nichollsi
(2.89MYA ± 0.18MYA) tightly links it with a known and particularly severe
arid pulse ~2.9MYA (Dodson, Macphail 2004). The Pliocene arid pulse is
likely to have led to the separation of the MRL and SCL to the northern and
southern parts of the southern coast, with the former populations likely to
have found refuge in the wetter Lake Muir or upper Blackwood regions.
Phylogeographic analyses show high levels of genetic structure between
distinct catchment areas along the southern coast within the SCL, with a
strong signal of recent range expansion through western and southern
regions occupied by the MRL (see Figure 6.2). Data analyses could not
distinguish between hypotheses regarding the biogeographic history of the
MRL and SCL. One hypothesis is extinction of the SCL in catchments
between the Blackwood and Deep Rivers (associated with changing climates
and water tables), followed by range expansion of the MRL into these areas.
Alternatively, competitive exclusion/one way hybridisation between the
MRL & SCL may be responsible.
4) Divergences between clades within the WL and SCL G. leai lineages (Figure
6.5) occur from the Plio-Pleistocene border (~1.64MYA) up until ~1MYA
(Main WL clade & DW/BN-1.11MYA±0.27MYA; Main WL & BS-
Wardell-Johnson G, Roberts JD (1993) Biogeographical barriers in a subdued
landscape: The distribution of the Geocrinia rosea (Anura: Myobatrachidae)
complex in south-western Australia. Journal of Biogeography 20, 95-108.
Weir JT, Schluter D (2004) Ice sheets promote speciation in boreal birds. Proceedings
of The Royal Society of London Series B 271, 1881-1887.
Welch JJ, Bromham L (2005) Molecular dating when rates vary. Trends in Ecology and
Evolution 20, 320-327.
Whitlock MC, McCauley DE (1999) Indirect measures of gene flow and migration: Fst
≠ 1/(4nm + 1). Heredity 82, 117-125.
Wiens JJ, Donoghue MJ (2004) Historical biogeography, ecology and species richness.
Trends in Ecology and Evolution 19, 639-644.
Wiens JJ, Penkrot TA (2002) Delimiting species using DNA and morphological
variation and discordant species limits in spiny lizards (Sceloporus). Systematic
Biology 51, 69-91.
Wong BBM, Keogh JS, McGlashan DJ (2004) Current and historical patterns of
drainage connectivity in eastern Australia inferred from population genetic
structuring in a widespread freshwater fish Pseudomugil signifer
(Pseudomugilidae). Molecular Ecology 13, 391-401.
Wright IJ, Ladiges PY (1997) Geographic variation in Eucalyptus diversifolia
(Myrtaceae) and the recognition of new subspecies E. diversifolia subsp.
170
hesperia and E. diversifolia subsp. megacarpa. Australian Systematic Botany
10, 651-680.
Wright S (1931) Evolution in Medelian populations. Genetics 16, 97-159.
Wuster W, Dumbrell AJ, Hay C, et al. (2005) Snakes across the strait: Trans-Torresian
phylogeographic relationships in three genera of Australasian snakes (Serpentes:
Elapidae: Acanthophis, Oxyuranus, and Pseudechis). Molecular Phylogenetics
and Evolution 34, 1-14.
Yoder AD, Nowack MD (2006) Has vicariance or dispersal been the predominant
biogeographic force in Madagascar? Only time will tell. Annual Review of
Ecology and Systematics 37, 405-431.
York Main B (1996) Terrestrial invertebrates in south-west Australian forests: The role
of relict species and habitats in reserve design. Journal of the Royal Society of
Western Australia 79, 277-280.
Zhang D-X, Hewitt GM (2003) Nuclear DNA analysis in genetic studies of populations:
Practice, problems and prospects. Molecular Ecology 12, 563-584.
Zink RM (1996) Comparative phylogeography in North American birds. Evolution 50,
308-317.
Zink RM (2002) Methods in comparative phylogeography, and their application to
studying evolution in the North American aridlands. Integrative and
Comparative Biology 42, 953-959.
171
Appendix 1:
Complete Tables of Polymorphic sitesfor each Data Chapter
172
173
Appendix 1a: Table of polymorphic sites for each of the unique Arenophryne rotunda ND2 haplotypes.8 1
7
19
89
10
8
12
0
13
9
16
3
16
7
17
5
17
8
18
1
18
4
19
0
22
6
24
7
25
0
25
3
25
9
27
4
28
0
29
2
32
5
33
4
33
7
35
2
38
8
39
7
40
6
42
2
43
3
43
6
44
8
47
2
48
4
49
0
50
8
51
8
52
0
52
9
54
7
54
8
55
0
57
5
57
7
58
1
58
3
60
0
60
1
61
0
62
0
63
1
63
4
64
6
Haplotype C A T G T C G C C A A T T T T C C T T G A A A T C C T T A C C G A A A A A C A A A G T A C T A T C C C A A A1 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -2 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -3 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -4 - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -5 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - G - - - G - - - - - - - - - -6 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -7 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -8 - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - G - - - - - - - - - -9 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - A - G - - - - - - - - - -
10 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - G11 - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - G12 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - G - - - - - - - - - - - - G13 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - G - - - - - - - - - - - - G14 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - G15 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C G - - - - - - - - - G16 - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - G17 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - G18 - G C - - - A A - G G C C - C T T C - A - G G - - - C - T T - A G - T G - T G - G - - - T C G - T A T G G -19 - G C - - - A A - G G C C - C T T C - A - G G - - - C - T T - A G - T G - T G - G - C - T C G - T A T G G -20 - G C - G - A A - G G C C - C T T C - A - G G - - - C - T T - A G - T G - T G - G - C - T C G - T A T G G -21 - G C - - - A A - G G C C - C T T C - A - G G - - - C - T T - - G - T G - T G - G - - - T C G - T A T G G -22 - G C - - - A A - G G - - - C T T C - A - G G - - - C - T T - A G - T G - T G - G - - - T C G - T A T G G -23 - G C - - - A A - G G C C - C T T C - A - G G - - T C - T T - A G - T G - T G - G - - - T C G - T A T G G -24 - G C - - - A A - G G C C - C T T C - A G G G - - T C C T T - A G - T G - T G - G - - - - C G - T A T G G -25 - G C - - - A A - G G C C - C T T C - A - G G - - - C - T T - A G - T G - T G - G - - - T C G - T A T G G -26 - G C - - - A A - G G C C - C T T C - A - G G - - - C - T T - A G - T G - T G - G - - - T C G - T A T G G -27 - G - - - - A A - G G C C - C T T C - - - G G C - - C - T T - A G G T G - T G - G - - - T C G C T A T G G -28 G G - - - - A A - G G C C - C T T C C - - G G C - - C - T T - A G G T G - T G - G - - - T C G - T A T G G -29 - G C - - T A A - G G - C C C T T - - A - G G - T - - - T T T A - - T - - - - - G - C - T C G - T A T G G -30 - G C - - T A A - G G - C C C T T - - A - G G - T - - - T T T A - - T - - - - - G - C - T C G - T A T G G -31 - G C - - T A A - - G - C C C T T - - A - G G - T - - - T T T A - - T - - - - - G - C - T C G - T A T G G -
Position
174
175
Appendix 1a Cont.: Table of polymorphic sites for each of the unique Arenophryne rotunda ND2 haplotypes continued.6
55
68
2
68
9
69
7
70
1
71
0
71
2
71
8
72
4
72
8
73
0
75
5
75
7
79
0
79
6
79
7
81
5
82
6
83
0
83
6
84
5
86
8
88
6
88
9
89
8
90
4
91
0
91
1
91
2
92
8
94
8
95
0
95
3
96
7
96
8
98
5
10
00
10
06
10
18
10
27
10
31
10
36
10
51
10
83
10
90
10
91
10
94
10
96
10
99
11
05
11
08
11
18
11
21
11
24
11
36
11
38
11
44
11
47
11
54
Haplotype A T A G G T G C A T A T A T G G G A T A A C C T C A T G C C A A C T A C G C C C G G T A A A T A C T G G A G A C C C A1 - - - - - - - - - - - - - - - - A - - - - - - - - - C - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -2 - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -3 - - - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -4 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -5 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A - - - - - - -6 - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A - - - - - - -7 - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A - - - - - - -8 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - G - - - - - - - - - - - - - -9 - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
10 - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -11 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - -12 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - -13 - - - - - - - - - - - - - - - - - - G C - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - A - - - - -14 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -15 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - G - - - -16 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - G - - - -17 - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - T G - - - - - - - - - G - - - -18 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - - - G T C - - - - - - - T -19 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - - - G T C - - - - - - - T -20 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - - - G T C - - - - - - - T -21 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - - - G T C - - T - - - - T -22 C C G A A C A - C - G C - - - - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - - - G T C - - - - - - - T -23 C C G A A C A - C - G C - G A - - G - - G - T A T G - - T T - - T C - T A T - T - A C T - - C G T C - - - - - - - T -24 C C G A A C A - C C G C - - A - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - - C G T C - - - - - - - T G25 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - - A T - T - A C T - - - G T C A - - - - - - T -26 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - T A T - T - A C T - G - G T C A - - - - - - T -27 C C G A A C A - C - G C - - A - - G - - G - - A T G - - T T - - T C - T A T - - - A C T G - - G T C - - - - - - - T -28 C C G A A C A - C - G C - - A - - - - - G - - A T G - - T T - - T C G T A T - - - A C T - - - G T C - - - - - - - - -29 C C G A A C T - C - G C G - A - - G - - - - - A T - - - T T T T T C - - A - - T A A C T - - - G T C - - - - - T T - -30 C C G A A C A - C - G C G - A - - G - - - - - A T - - - T T T T T C - - A - - T A A C T - - - G T C - - - - - T T - -31 C C G A A C A - C - G C G - A - - G - - - - - A T - - - T T T T T C - - A - - T A A C T - - - G T - - - - - - T T - -
Position
176
177
Appendix 1b: Table of polymorphic sites for each of the unique Crinia georgiana ND2 haplotypes.1 123
156
189
240
306
312
318
322
338
342
345
348
363
366
372
390
396
402
432
435
438
450
465
477
495
531
543
558
562
565
570
600
603
696
702
711
714
724
733
735
747
763
764
780
810
813
834
879
903
909
918
921
930
933
939
951
957
966
978
981
982
992
993
1002
1014
1033
1043
1063
1083
Hap # T C C G A C C C A C G A G A A T A C C C A G C T T A C C C G G C G C A G C A A T T T G C A A T T A A G T T C T T C A A C T C T C T A G C A T1 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -2 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - -3 - - - A - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -4 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -5 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - A - - - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -6 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - A - - - - - - - - G - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - C7 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - A - - - - C - - - - - - - - - - - - - - T - - - - - - - - - - T - -8 - - - - - - - - - - A - - - - - G - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -9 - - - - - - - - T - - - - - T - - - - - - A - - - - T - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - -10 - - - - - - - - T - - - - - T - - - - - - A - - - G T - - - - - - - G - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - -11 - - - - - - - - T - - - - - T - - - - - - A - - - - T - - - - - - - - - - - - - - - A - - - - - - - A - - - - - - - - - - - - - - - - - - -12 - - - - - - - - T - - - - - T - - - - - - A - - - - T - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - C - - - - - - - - -13 - - - - - - - - - - - - - - - C - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -14 - - - - - - - - - - - - - - - - - T - - - - - - - - T - - - A - - - - - - - - - - - - - - - - - - G - - - - C - - - - - - - - - - - - - - -15 - - - - - - T - - - - - - - - - G T - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - G - - - - C - - - - - - - - - - - - - - -16 - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - -17 - - - - - - - - - - - - - - - - - - - T - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - -18 - - - - - - - - - - - - - - - - - - - - - - - - - - T T - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -19 - - - - - - - - - - - - - - - - - - - - - - - - - - T T - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -20 - - - - - - - - - - - - - - - - - - - - - - - - - - T T - - - - - T - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - -21 - - - - - T - - - - - - - - - - - - - - - - - - - - T - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A - - -22 - - - - - - - - - - - - A - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - T - - - - - - - - C - - - - - - - - - - - - - - - - -23 - - - - G - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -24 - - - - G - - - - - - - - G - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -25 - - - - G - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -26 - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -27 - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -28 - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -29 - - - - G - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -30 - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - -31 - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - -32 - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - -33 - - - - G - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - -34 - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - -35 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - C36 - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - C37 - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - C38 - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - A - - C39 - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - G - - - - - - - - - - - - - - - - - - - - C40 - - T - G - - - - - - - - T G C - - - - - - T - - - - - T - - - A - - A - G - - - - - - - G - - - - A C - - - C - - - A - - - - C - - - - -41 - - T - G - - - - T - - - T G C - - - - - - T - - - - - T - - - A - - A - G - - - - - - - G - - - - A C - - - C - - - A - - - - C - - - - -42 - - T - - - - - - - - - - T - C - - - - - - T - - - - - T - - - A - - A - G - - - - - - - G - - - - A C - - - C - - - A - - - - C - - - - -43 - - T - - - - - - - - - - T - C - - - - - - T - - - - - T - - - A - - A - G - - - - - - - G - C - - A C - - - C - - - A - - - - C - - - - -44 - - T - - - - - - - - - - T - C - - - - - - T - - - - - T - - - A - - A - G - - - - - - - G - - G - A C - T - C - - - A - - - - C - - - G -45 - - T - - - - - - - - - - T - C - - - - G - T - - - - - T - - - A - - A - G - - - - A - - G - - G - A C - T - C - - - A - - - - C - - - G -46 - - T - - - - - - - - - - T - C - - - - G - T - - - - - T - - - A - - A - G - - - - A - - G C - G - A C - T - C - - - A - - - - C - - - G -47 - - T - - - - - - - - - - T - C - - - - G - T - - - - - T - - - A - - A - - - - - - A - - G - - G - A C - T - C - - - A - - - - C - - - G -48 C T T - - - - - - - - - - T - C - - T - - - T - - - - - T - - - A - - A - G - - - - - - - G - - G - A C - T - C - - - A - T C - C - - - - -
Position
178
179
Appendix 1c: Table of polymorphic sites for each of the unique Metacrinia nichollsi ND2 haplotypes.
Hap. 11
18
85
15
5
16
1
16
4
17
1
17
3
18
3
18
5
18
8
19
7
22
4
23
9
24
5
24
8
26
0
26
9
27
2
27
5
29
3
34
4
35
0
36
8
37
1
37
4
38
0
38
3
40
7
41
6
43
7
43
8
44
6
46
2
47
3
48
2
48
8
50
3
53
9
54
6
54
9
56
4
56
9
58
0
59
3
60
1
61
7
G T T C C C G T G T G C T C G C T A G C A A C A C A A A A C G G A C A G C C C C G A A A C T T1 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - G -2 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - G -3 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -4 - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -5 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -6 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -7 - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - G - - - -8 - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -9 T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -
10 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - -11 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - G - - - -12 - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - A - - - - - - - - - - - - - - -13 - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -14 - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -15 - A - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -16 - - - T - T - C A C A - - - A T C G A T G C - G - C - - G - - - - T - A T T - A A G G G T - C17 - - C - - - - - - - A - - T A - - - A - - C T - - - G G - - - - - - - - - T A - - - - - - - -18 - - C - T - - - - - A - - T A - - - A - - C T - - - G G - - - - - - - - - T A - - - - - - - -19 - - C - A - - - - - A - - - A - - - A - - C - - - - G G - - A - G - - - - T A - - - - - - - -20 - - C - A - - - - - A - - - A - - - A - - C - - - - G G - - A - G - - - - T A - - - - - - - -21 - - C - A - - - - - A - C - A - - - A - - C - - - - G G - - A - G - - - - T A - - - - - - - -22 - - C - A - - - - - A - C - A - - - A - - C - - - - G G - - A - G - - - - T A - - - - - - - -23 - - C - - - - - - - A - - - A - - - A - - C - - - - G G - - A - - - - - - T A - - - - - - - -24 - - C - - - - - - - A - - - A - - - A - - C - - T - G G - - A - - - - - - T A - - - - - - - -25 - - C - - - - - - - A - - - A - - - A - - C - - - - G G - - A - - - - - - T A - - - - - - - -26 - - C - - - - - - - A T - - A - - - A - - C - - - - G G - - A - - - - - - T A - - - - - - - -
Position
180
181
Appendix 1c Cont.: Table of polymorphic sites for each of the unique Metacrinia nichollsi ND2 haplotypes continued.
Hap. 62
3
65
3
66
8
68
0
68
3
69
2
70
1
74
0
75
0
78
5
79
4
79
8
81
0
83
3
84
0
84
5
84
8
85
2
85
7
86
6
88
2
89
3
89
9
91
4
91
7
92
6
92
7
92
9
96
1
96
5
96
8
97
2
97
7
98
9
10
65
10
76
10
84
10
91
10
98
11
06
11
07
11
16
11
18
11
19
11
24
11
25
G A G G T C A T G A T G G A A T C A T A C G T G A C G C C T T G T C A A A A A C C C A G A G1 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -2 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A3 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A4 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A5 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - A6 - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - -7 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A8 - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A9 - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A
10 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A11 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - G - - - - - - A12 - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A13 - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - G -14 - G - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A15 - G - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - A16 - G - - C T G - A G - - A G - C T G - G - - A A G T A T - C C A - T - - - - - - - - C - - A17 A - A - - T - - A - - - - - G C - G C - - A G - G T A - T C - - - - C G - G G - A T - - - -18 A - A - - T - - A - - - - - G C - G C - - A G - G T A - T C - - - - C G G G G - A T - - - -19 A - A - - T - - A - - - - - G C - G C - - A G - G T A - T C - - - - C - G G - T A T - - - -20 A - A - - T - - A - - - - - G C - G C - - A A - G T A - T C - - - - C - G G - T A T - - - -21 A - A - - T - - A - - - - - G C - - C - - A G - G T A - T C - - - - C - - G - T A T - - - -22 A - A - - T - - A - - - - - G C - - C - T A G - G T A - T C - - - - C - - G - T A T - A - -23 A - A - - T - - A - C - - - G C - G C - - A G - G T A T T C - - C - C - G - - - A T - - - -24 A - A - - T - - A - - - - - G C - G C - - A G - G T A T T C - - C - C - G - - - T - C - - -25 A - A - - T - - A - - - - - G C - G C - - A G - G T A T T C - - C - C - - - - - T - - - - -26 A - A - - T - - A - - - - - G C - G C - - A G - G T A T T C - - C - C - - - - - A - - - - -
Position
182
183
Appendix 1d: Table of polymorphic sites for each of the unique Geocrinia leai ND2 haplotypes.
19
45
46
10
5
12
0
12
6
13
2
13
6
13
8
14
1
14
2
15
4
16
8
17
7
18
4
19
2
19
3
19
5
21
0
22
0
22
1
22
9
23
4
25
2
25
8
26
4
26
7
27
9
28
5
29
1
30
0
30
9
31
8
32
4
32
9
34
4
35
1
35
4
36
3
36
7
36
9
37
3
37
8
38
7
40
2
40
5
40
6
42
0
42
3
44
7
45
3
45
6
47
1
Hap. A C T C A C C A C C G A C T T A G G C A T T A A T G G T A A T T C T T C A A T T A C T C T T C C G T A C C1 - T - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -2 - T - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -3 - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -4 - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -5 - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - T - - - - G - - - - - - - - - - - - - - -6 - - C - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -7 - - C - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -8 - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -9 - - - - - - - - - - - - - - - - - - - G - C G - - - - - - - - - - - - - - - - - - - - - - C - - - - - - -
10 - - - - - - - - - - - - - - - - - - - G - C G - - - - - - - - - - - - - - - - - - - - - - C - - - - - - -11 - - - - - - - - - - - - - - - - - - - G - C G - - - - - - - - - - - - - - - - - - - - - - C - - - - - - -12 - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -13 - - - T - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -14 - - - T - - - - - - - - - - - - - - - G - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -15 - - - T - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - -16 - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -17 - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -18 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -19 - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - -20 - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - C - - - G - C - - - - - - - - A - - - -21 - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - C - - - - - C - - - - - - - - - - - - -22 - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - C - - - - - C - - - - - - - - - - - - -23 - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - C - - - - - C - - - - - - - - - - - - -24 - - - T - - - - - - - - - - - - - - - G C C - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - -25 - - - - - - - - - - - - - - - - - - - G C C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -26 - - - - - - - - - - - - - - - - - - - G C C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -27 - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -28 - - - - - - - - - - A - - - C - - - - - - - - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -29 - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -30 - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - C - - - - - C - - - - - - - - - - - - -31 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - C - - T32 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - C - - T33 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - G - - - - - - - - - - - - C - - T34 - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - T - - C - - T35 - - - T - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - C - - T36 - - - T - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - C - - T37 - - - - - T - - T - - G - - - G A T - - - - - - - A A C - - - C - - - - - - C C - - C T C - - - - C - T T38 - - - - - T - - T - - G - - - G A T - - - - - - - A A C - - - C - - - - - - C C - - C T C - - - - C - T T39 - - - - - T - - T - - G - - - G A T - - - - - - - A A C - - - C - - - - - - C C - - C - C - - T - C - T T40 G - - - - T - - T - - G - - - G A T - - - - - - - A A C - - - C - - - - - - C C - - C - C - - T - C - T T41 - - - - - T - G T - - G - - - G A T - - - - - - - A A C - - - C - - - - - - C C - T C - C - - T - C - T T42 - - C - - T T - T - - G - C - G A C - - - - - G - A A C - - - C - - - - - - C C T - C - C - - - - C G T T43 - - - - - T T - T - - G - C - G A C - - - - - G - A A C - - - C - - - - - - C C T - C - C - - - - C G T T44 - - - - - T T - T - - G - C - G A C - - - - - G - A A C - - - C - - - - - - C C T - C - C - - - - C G T T45 - - - - - T T - T - - G - C - G A C - - - - - G - A A C - - - C - - - - - - C C T - C - C - - - - C G T T46 G - - T G T - - - T - - - - - G A C T - - - G - - A - - G G C - - - - T - - C C - - C - - - - - - C - T T47 G - - T G T - - - T - - - - - G A C T - - - G - - - - - G G C - - - - T - - C C - - C - - - - - - C - T T48 G - - T G T - - - T - - - - - G A C T - - - G - - - - - G G C - - - - T - - C C - - C - - - - - - C - T T
Position
184
185
Appendix 1d Cont.: Table of polymorphic sites for each of the unique Geocrinia leai ND2 haplotypes continued.
47
5
48
3
49
2
49
5
49
8
50
1
50
4
50
8
51
0
51
7
52
8
53
4
54
0
54
1
54
2
54
3
55
5
55
8
56
1
58
8
59
1
59
4
59
7
60
0
60
3
60
6
61
7
61
8
62
6
63
3
63
6
66
6
66
9
69
0
69
6
69
7
70
8
71
1
71
3
72
1
72
3
72
6
72
7
73
0
73
2
73
4
73
5
74
1
74
7
75
3
75
5
75
7
77
1
77
4
Hap. C G C C G G A C C A C A C G A C C T A C C G T G A A A G T A T T A T T C C C A A T T C T A T G T A A T C G C1 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -2 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -3 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -4 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -5 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -6 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -7 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - G G - C - - - - - - - - - - - -8 - - - - A - - - - - - - - - - - - - - - - A G - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -9 - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - C - - - - A - - - - - - -
10 - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - C - - - - A - - - - - - -11 - - - - - - - A - - - - - - G - - - - - - - - - - - G A - - - - - - - - - - - G - C - - - - A - - - - - - -12 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -13 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -14 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -15 - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - G - C - - - - - - - - - - - -16 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -17 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -18 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -19 - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - -20 - - - - - - G - T - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -21 - - - T - - G - T - - - - A - - - - - - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -22 - - - T - - G - T - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -23 - - - T - - G - T - - - - A - - - - - - - - G - - - - - - - - - - C - - - - - - - - - - - - - - - - - - - -24 - - - - C - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - C - - - - - - - - - - - -25 - - - - C - - - - - - - - - - - - - - - - - - - G G - - - - - - - - - - - - - - - C - - - - - - - - - - - -26 - - - - C - - - - - - - - - - - - - - - - - - - G G - - - - - C - - - - - - - - - C - - - - - - - - - - - -27 - A - - A - G - T - - - - - - - - - - - - - - - G - - - - - - - - - G - - - - - - - T - - - - - - - - - - -28 - A - - A - G - T - - - - - - - - - - - - - - - G - - - - - - - - - G - - - - - - - - - - - - - - - - - - -29 - A - - A - G - T - - - - - - - - - - - - - - - G - - - - - - - - - G - - - - - - - - - - G - - - - - - - -30 - A - - A - G - T - - - - - - - - - - - - - - - G - - - - - - - - - G - - - - - - - - - - - - - - - - - - -31 - A - - - - G - T - - - T - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -32 - A - - - - G - T - - - T - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -33 - A - - - - G - T - - - T - - - - - - - - - C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -34 - A - - - - G - T G - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -35 - A - - - - G - T - - - T A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - -36 - A - - - - G - T - - - T A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - A -37 T A - - - - - - - - T - - A - - - C G - - - G - G G - A - G C - - - C A - - - - - - - C - - A C - - - - - -38 - A - - - - - - - - T - - A - - - C G - - - G - G G - A - G C - - - C A - - - - - - - C - - A C - - - - - -39 - A - - - A - - - - T - - A - - - C G - - - - - G G - A C G C - - - C A - - - - - - - C - - A - - - - T - -40 - A - - - A - - - - T G - A - - - C G - - - - - G G - A - G C - - - C A - - - - - - - C - - A - - - - T - -41 - A - - - A - - - - T G - A - - - C G - - - - - G G - A - G C - - - C A - - - - - - - C - - A - - - - T - -42 - A T - - - - - - - T - T - - - - C - - - - - - G - - A - G C - - - C A - - - G - - - C - - A - G - - - - -43 - A T - - - - - - - T - T - - - - C - - - - - - G - - A - G C - - - C A - - - G - - - C - - A - G - - - - -44 - A T - - - - - - - T - T - - - - C - - - - - - G - - A - G C - - - C A - - - G - - - C - - A - G - - - - -45 - A T - - A - - - - T - T - - - - C - - - - - - G - - A - G C - - - C A - - - G - - - C - - A - G - - - - -46 T A - - - - - - - - - - - A - T T C - T T - G - G G - A - - - C G - C - A T - - C - - C G - A - - G - - A A47 T A - - - - - - - - - - - A - T T C - T T - G A G G - A - - - C G - C - A T - - C - - C G - A - - G - - A A48 T A - - - - - - - - - - - A - T T C - T T - G A G G - A C - - C G - C - A T - - C - - C G - A - - G - - A A
Position
186
187
Appendix 1d Cont.: Table of polymorphic sites for each of the unique Geocrinia leai ND2 haplotypes continued.7
77
78
3
78
6
79
3
81
0
81
6
81
7
82
5
82
6
82
9
83
1
83
7
84
9
85
2
85
5
85
8
86
2
86
4
86
7
87
3
87
6
87
9
88
0
90
0
91
5
92
9
93
0
93
6
93
7
94
8
95
1
96
9
97
2
99
3
99
6
99
8
10
02
10
14
10
20
10
39
10
42
10
53
10
55
10
56
10
61
10
66
10
68
10
93
10
95
10
96
11
01
11
17
11
19
11
20
Hap. A C T G G T A T G C A A A C C A C A C A A C G C C T G C G T C C C T C T A C T A G C T C A T C C A G A A C C1 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - C T - - - G - - -2 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - A G - - -3 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - - G - - -4 - - - - A C - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - - G - - -5 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - - G - - -6 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - - G - - -7 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - - G - - -8 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C T - - - - - T -9 - - - - A - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - G - - -10 - - - - A - - - - - - - - - - G - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - T - - - G - - -11 - - - - A - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - G - - -12 - - - - A - - - - - - - - - - - - - - - G - - - - - - - - C - - - - - - - - - - - - - - - C T - - - G - - -13 - - - - A - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - C T - - - G - - T14 - - - - A - - - - - - - - - - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - C T - G A - - - -15 - - - - A C - - - - - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -16 - - - - A - - - - - - - G - - - - - - - - A - - - - - - - C - - - - - - - - - - - - - - - - T - - - G - - -17 - - - - - - - - - - - - G - - - - - - - - A - - - - - - - C - - - - - - - - - - - - - - - - T - - - G - - -18 - - - - - - - - - - - - G - - - - - - - - A - - - - - - - C - - - - - - - - - - - - - - - - T - - - G - - -19 - - - - - - - - - - - - G - - - T - - - - - - - - - - - - C - - - - - - - T - - - - - - - - T - - - - - - -20 - - - - - - - - - - - - G - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - - - - - - - -21 - - - - - - - - - - G - - T - - - - - - - - - - - - - T - - - - - - - - - - - - - - C - - - - - - - - - - -22 - - - - - - - - - - G - - T - - - - - - - - - - - - - T A - - - - - - - - - - - - - C - - - - - - - - - - -23 - - - A - - - - - - G - - T - - - - - - - - - - - - - T A - - - - - - - - - - - - - C - - - - - - - - - - -24 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - T - - - G - - -25 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - T T - - G - - -26 - - - - A - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C - - - - - - - - - - T - - - - C - T27 - - - - - - - - - - - - G - - - G - T G - - - - - - - - - - - T - - - - - - - G - - - - - - - - - - - - - -28 - - - - - - - - - - - - G - - - - - T G - - - - - - - - - - - T - - - - - - - G - - - - - - - - - - - - - -29 - - - - - - - - - - - - G - G - - - T G - - - - - - - - - - - T - - - - - - - G - - - - - - - - - - - - - -30 - - - - - - - - - - - - G - - G - - T G - - - - - - - - - - - T - - - - - - - G - - - - - - - - - - - - - -31 - - - - - - G - - - - - G T - G - - - - - - - - - - - - - - - - - - - - G - - - - - - - - - - - - - G - - -32 - - - - - - G - - - - - G - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G - - -33 - - - - - - G - - T - - G - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - G A G - - -34 - - - - - - G - - - - - G - - G - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -35 - - - - - - G - - - - - G - - G - - - - - - - - - - - - - - T - - - - - - T - - - - - - - - - - - - - - - -36 - - - - A - G - - - - - G - - G - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - -37 - - C - - C - - - - - G - - A G T - - - - - A T - C A - - - - - - C - - - - - - A - - T - - - - - - - - - -38 - - C - - C - - - - - G - - A G T - - - - - A T - C A - - - - - - C - - - - - - A - - T - - - - - - - - - -39 - - - - - C - - - - - G - - - G T - - - - - A T - C A - - - - - - C - - - - - - A - - T - - - - - - - - - -40 - - - - - - - - - - - G - - - G T - - - - - A T - C A - - - - - - C - - - - - - A - - T - - - - - - - - - -41 - - - - - C - - - - - G - - - G T - - - - - A T - C A - - - - - - C - - - - - - A - - T - - - - - - - - - -42 - - - - - - - - - - - G - - - G T - - - - - A - T C A T - - - - - C - - - - - - A - - T - C - - - - - - - -43 - - - - - - - - - - - G - - - G T - - - - - A - T C A T - - - - - C - - - - - - A - - T - C - - - - - - - -44 - - - - - - - - - - - G - - - G T - - - - - A - T C A - - - - - - C - - - - - - A - - T - C - - - - - - - -45 - - - - - - - - - - - G - - - G T - - - - - A - T C A - - - - - - C - - - - - - A - - T - C - - - - - - - -46 G T G - - - - - A - - - - - G G - G - G - - - T - C - - - - - - - - - - - - C - - T - - - - - - - - - - - -47 G T G - - - - - A - - - - - G G - G - G - - - T - C - - - - - - - - - - - - C - - T - - - - - - - - - - - -48 G T G - - - - C A - - - - - G G - G - G - - - T - C - - - - - - - - - - - - C - - T - - - - - - - - - - - -
Position
189
Appendix 2:
Complete Table of Pairwise Genetic Distances(uncorrected p sequence divergence)
1-Y; 19-Y; 20-Y; 2-N; 11-Y-RE; 12-Y; 13-Y-LDC w/ F or PF w/ RE; 21-?
212
213
Appendix 3d: Complete GeoDis v 2.4 output for the Geocrinia leai (Western Lineage Only) Nested Clade Phylogeographic Analysis.
Haplotypes 1-step clades 2-step clades 3-step clades 4-step clades 5-step clades 6-step cladesNo. Dc Dn No. Dc Dn No. Dc Dn No. Dc Dn No. Dc Dn No. Dc Dn No. Dc Dn1 t 0 02 t 0 0 1.1 t 0 4.803 I 0 04 t 0 0 1.3 I 0 19.365 t 0 0 1.2 t 0 4.86 I 0 07 t 0 1.5 t 23.91 23.718 t 0 0 1.4 t 0 29.5
IT -7.97 4.14 2.1 I 0 0 3.1 I 15.57< 22.04<
9 I 0 010 t 0 0 1.10 I 0 011 t 0 0 1.11 t 0 0 2.4 t 0 0 3.2 t 0< 50.87>
IT 15.57 -28.84<1-Y; 19-Y; 20-N - IGS 4.1 t 30.28< 40.13<
12 t 0 0 1.12 I 0 108.9313 t 0 0 1.13 I 0 108.98 2.5 I 108.95 109.16
24 t 0 0 1.42 I 0 0 2.6 I 0 39.07 3.3 I 85.80 80.91
25 t 0 0 1.44 I 0 0 2.7 I 0 0
26 t 0 0 1.46 t 0 0 2.8 t 0 0 3.4 t 0 23.44IT 85.80 57.47 4.2 I 57.92 99.15>
IT 27.65 -59.02<1 - N; 2 - Y; 3 - Y; 5 - Y; 15 - N - PF &/OR LDC - 21 - 5.1 t 52.26< 71.11<
14 t 0 0 1.15 t 0 0 2.9 I 0 82.81
15 t 0 0 1.19 t 0 0 2.11 t 0 82.74I-T 0 0.07 3.5 I 82.78 96.96
16 I 0 017 I 0 0 1.23 I 0 0 2.13 I 0 14.80
18 t 0 0 1.26 t 0 0 2.14 t 0 29.60IT 0 -14.80 3.6 I 19.73 48.72
19 t 0 0 1.27 I 0 0 2.15 I 0 0 3.7 I 0 47.82 4.3 I 64.65 63.84
20 t 0 0 1.33 t 0 0 2.18 t 0 0 3.8 I 0 0
21 t 0 0 1.37 I 0 0 2.19 I 0 0
22 I 0 0 1.38 I 0 023 t 0 0 1.38 t 0 0 2.20 t 0 0 3.9 t 0 0 4.4 t 0< 46.40
IT 64.65> 17.431-Y; 19-Y; 20-N - IGS 5.2 t 54.53< 99.96>
I-T 2.26 28.01<1 - N; 2 - N; 11 - Y - RE; 12 - Y; 13 - Y - LDC w/ F OR PF w/ RE; 21 - 6.1 I 82.06< 86.05
27 t 0 0 1.47 t 0 028 t 0 030 t 0 0 1.48 I 0 029 t 0 0 1.49 t 0 0 2.21 t 0 0 3.10 t 0 0 4.5 t 0 0 5.3 t 0 0 6.2 t 0< 99.69
31 t 0 0 1.50 t 0 032 I 0 0 1.51 I 0 0 2.22 I 0 11.5
33 t 0 0 1.53 t 0 0 2.23 t 0 11.5
34 t 0 0 1.55 t 0 0 2.24 I 0 45.99I-T 0 8.62 3.11 t 18.4 26.28
35 t 0 0 1.58 I 0 0 3.12 t 0 32.8536 t 0 0 1.59 t 0 0 2.26 t 0 0 I-T 18.4 -6.57 4.6 t 0 0 5.4 t 0 0 6.3 t 28.16< 99.6
I-T 64.14> -13.581 - N; 2 - Y; 3 - N; 4 - Y; 9 - Y - AF
215
Appendix 4:
Supplementary Tests for Nested Clade PhylogeographicAnalysis to assess Secondary Contact
for each Data Chapter
216
217
Appendix 4a: Mean pairwise distances (km) between geographical clade centres found ateach Arenophryne rotunda (Northern Lineage) sampling location at various nesting levels.Sites where geographically divergent clades (i.e. high distance values) are present relativeto the distribution of the lineage represent sites of secondary contact between divergentlineages. For principles and methodology behind this supplementary test for NCPA seeTempleton (2001).
1 Step Clades
0
5
10
15
20
25
30
35
40
DHN DHM DHS SP FE2 PC EL1 WW
Collection Sites
Mean
Pair
wis
e D
ista
nce
b
/w
Cla
des
(km
)
2 Step Clades
0
10
20
30
40
50
60
DHN DHM DHS SP FE2 PC EL1 WW
Sampling Location
Mean
Pair
wis
e D
ista
nce
b
/w
Cla
des
(km
)
3 Step Clades
0
5
10
15
20
25
30
35
40
45
50
DHN DHM DHS SP FE2 PC EL1 WW
Sampling Location
Mean
Pair
wis
e D
ista
nce
b
/w
Cla
des
(km
)
218
219
Appendix 4b: Mean pairwise distances (km) between geographical clade centres found ateach Crinia georgiana sampling location at various nesting levels. Sites wheregeographically divergent clades (i.e. high distance values) are present relative to thedistribution of the species represent sites of secondary contact between divergent lineages.For principles and methodology behind this supplementary test for NCPA see Templeton(2001). Only clade levels with a possible inference of secondary contact are shown.
2 Step Clades
0
50
100
150
200
250
MO SA
SP
MU
R
HW
CO
L
NR
BW BE
DW SG DF
KH
KA
L
BB
CLG MIS
CA
NP
Sampling Location
Mean
Pair
wis
e D
ista
nce
sb
/w
Cla
des
(km
)
3 Step Clades
0
50
100
150
200
250
MO SA
SP
MU
R
HW
CO
L
NR
BW BE
DW SG DF
KH
KA
L
BB
CLG MIS
CA
NP
Sampling Location
Mean
Pair
wis
e D
ista
nce
s b
/w
Cla
des
(km
)
Total Cladogram
0
50
100
150
200
250
300
350
400
450
500
MO SA
SP
MU
R
HW
CO
L
NR
BW BE
DW SG DF
KH
KA
L
BB
CLG MIS
CA
NP
Sampling Locations
Mean
Pair
wis
e D
ista
nce
s b
/w
Cla
des
(km
)
220
221
Appendix 4c: Mean pairwise distances (km) between geographical clade centres found ateach Metacrinia nichollsi (Main Range Lineage) sampling location at various nestinglevels. Sites where geographically divergent clades (i.e. high distance values) are presentrelative to the distribution of the lineage represent sites of secondary contact betweendivergent lineages. For principles and methodology behind this supplementary test forNCPA see Templeton (2001).
1 Step Clades
0
20
40
60
80
100
120
NR
N
NR
M
NR
S
BS
SA
B
BN
DW
N
DW
S
SG
N
SG
S
DFN
DFS
KH
N
KH
S
KA
LS
Sampling Location
Mean
Pair
wis
e D
ista
nce
b
/w
Cla
des
(km
)
2 Step Clades
0
20
40
60
80
100
120
NR
N
NR
M
NR
S
BS
SA
B
BN
DW
N
DW
S
SG
N
SG
S
DFN
DFS
KH
N
KH
S
KA
LS
Samplng Location
Mean
Pair
wis
e D
ista
nce
b
/w
Cla
des
(km
)
222
223
Appendix 4d: Mean pairwise distances (km) between geographical clade centres found ateach Geocrinia leai (Western Lineage) sampling location at various nesting levels. Siteswhere geographically divergent clades (i.e. high distance values) are present relative to thedistribution of the lineage represent sites of secondary contact between divergent lineages.For principles and methodology behind this supplementary test for NCPA see Templeton(2001).
1 Step Clades
0
5
10
15
20
25
30
SA
SE
RP
MU
R
HW
CO
L
NR
BS
BM BN
DW
Sampling Location
Mean
Pair
wis
eD
ista
nce
s b
/w
Cla
des
(km
)
2 Step Clades
0
10
20
30
40
50
60
70
80
90
SA
SE
RP
MU
R
HW
CO
L
NR
BS
BM BN
DW
Sampling Location
Mean
Pair
wis
e D
ista
nce
s b
/w
Cla
des
(km
)
3 Step Clades
0
10
20
30
40
50
60
70
80
90
100
SA
SE
RP
MU
R
HW
CO
L
NR
BS
BM BN
DW
Sampling Location
Mean
Pair
wis
e D
ista
nce
s b
/w
Cla
des
(km
)
224
225
Appendix 4d Cont.: Mean pairwise distances (km) between geographical clade centresfound at each Geocrinia leai (Western Lineage) sampling location at various nesting levels.Sites where geographically divergent clades (i.e. high distance values) are present relativeto the distribution of the lineage represent sites of secondary contact between divergentlineages. For principles and methodology behind this supplementary test for NCPA seeTempleton (2001).