Reconstructing the Evolutionary History of the Australian Magpie (Gymnorhina Tibicen): Patterns of Molecular Variation in a Widespread Passerine and Two Species of Obligate Feather Ectoparasites

Author Toon, Alicia

Published 2007

Thesis Type Thesis (PhD Doctorate)

School Australian School of Environmental Studies

DOI https://doi.org/10.25904/1912/96

Copyright Statement The author owns the copyright in this thesis, unless stated otherwise.

Downloaded from http://hdl.handle.net/10072/365874

Griffith Research Online https://research-repository.griffith.edu.au

RECONSTRUCTING THE EVOLUTIONARY HISTORY OF THE AUSTRALIAN MAGPIE ( GYMNORHINA TIBICEN ): PATTERNS OF MOLECULAR VARIATION IN A WIDESPREAD PASSERINE AND TWO SPECIES OF OBLIGATE FEATHER ECTOPARASITES.

Alicia Toon B.Sc. (Hons) Australian School of Environmental Studies, Griffith University

Submitted in fulfilment of the requirements of the degree of Doctor of Philosophy, September 2006

II SUMMARY

During the Pleistocene, fluctuating climates led to cycles of glacial/arid activity interspersed with pluvial periods across continents in both northern and southern hemispheres. Many studies in the northern hemisphere have used genetic analysis to document the important role that glacial activity has played in structuring avian populations at high latitudes. However, few have attempted to study the associated effect of aridification at low latitudes in the southern hemisphere. I investigated the past effects that cyclic aridification may have had on the population structure and history of a widespread endemic Australian bird species, the Australian magpie ( Gymnorhina tibicen) and two species of obligate feather ectoparasites.

1166 samples from across the native range of G. tibicen were analysed for mitochondrial control region sequence variation and variation at six microsatellite loci. Analysis of mitochondrial control region sequence data indicated monophyletic clades that were geographically congruent with an eastern and western region. Analysis of mitochondrial variation at the sites sampled in this study suggested the contemporary distribution of eastern and western clades is non-overlapping but in close proximity. Analysis of microsatellite variation suggested that secondary contact may have occurred between eastern and western clades in north-western Australia. From AMOVA analysis and Bayesian analysis of population structure ( BAPS ) it was indicated that contemporary nuclear gene flow preceded mitochondrial gene flow from eastern populations through to north-western Australia. Most eastern, northern and north-western sites showed little geographic structure for microsatellite variation. BAPS analysis of microsatellite variation however, suggested there was as much structure among south-west populations as there was between eastern and western populations.

For a majority of population comparisons, estimates of gene flow based on coalescent analyses ( LAMARC ) suggested higher gene flow rates for males than predicted by differences in effective population size of nuclear DNA compared to mitochondrial DNA. This result coupled with the spread of nuclear DNA preceding mitochondrial DNA supports earlier studies that suggest dispersal in magpies in male biased.

Using the program IM , eastern and western mainland clades were estimated to have diverged in the Pleistocene around 36, 000 years ago. The island population of Tasmania was even more recent in origin, possibly since sea levels rose 16, 000 years

III

ago, inundating Bass Strait. The putative Carpentarian and possibly Canning barriers in the north and the Nullarbor-Eyrean arid barriers in the south appear to be associated with the divergence between eastern and western mainland populations. Nested clade analysis indicated a signature of range expansion in the eastern region suggesting movement possibly inland and northward subsequent to the last period of aridity. Although not significant, north-eastern and south-eastern populations appeared to show some evidence of a population expansion from mitochondrial DNA. Collectively, phylogeographic analyses suggested that increasing aridity during the Pleistocene played an important role in structuring the Australian magpie.

The east to west pattern of mtDNA divergence that was identified contrasts with the striking north to south pattern in morphological (back colour) variation in magpies. Over a large proportion of northern Australia, magpies are black backed (BB) and over a smaller area in southern Australia, magpies are white backed (WB). Between BB’s and WB’s a contact zone is present where both parental forms and magpies with an intermediate black band occur. The discordance between back colour and mtDNA structure in magpies suggests recent history is not responsible for the morphological variation.

Mitochondrial cytochrome oxidase I sequence variation was analysed for two species of feather lice associated with G. tibicen . Philopterus sp. has greater habitat specificity than Brueelia semiannulata and as predicted showed deeper divergences among populations than B. semiannulata . There was concordance between the distribution of mitochondrial clades for Philopterus sp. and magpies. The overlap of eastern magpie haplotypes and western Philopterus sp. haplotypes at one site suggested secondary contact among eastern and western clades in northern Australia. Two clades were also evident for B. semiannulata . However they were not congruent with geographic structure of the host or Philopterus sp. Rather, the two non-overlapping B. semiannulata clades were distributed in northern and southern Australia. The divergent clades of B. semiannulata may represent populations that diverged on magpies that were isolated prior to the last period of aridity. It was evident that gene flow occurs among populations of northern and southern B. semiannulata ; therefore the contemporary maintenance of divergent clades may be due to selection. One possibility is that selection for thermal tolerance is maintaining current distributions of B. semiannulata.

IV Overall microsatellite variation and mtDNA variation in host and lice suggest that increasing aridity and Pleistocene refugia played a role in structuring populations of the Australian magpie. Since the Pleistocene, the dispersal ability and generalist habitat requirements may have facilitated the movement of magpies into an almost contiguous modern distribution across the continent. This study supports the idea that Pleistocene aridification played an important role in structuring intraspecific variation in low latitudinal southern hemisphere avian species.

V

VI ACKNOWLEDGEMENTS

To my supervisor Professor Jane Hughes for directing me to the path of magpie phylogeography, thank you for your support, advice and encouragement. And thank you to my associate supervisor Peter Mather for your excellent proof reading and helpful suggestions.

Also a special thank you to Andrew Baker, who provided samples and suggestions and read many of the earlier drafts.

For those people that made this study happen, the field assistants, thank you for all your time and patience. Giovanella Carini, Jonathon Marceau, Joel Huey, James Fawcett, Daniel Schmidt, Jemma Sommerville, Kate Masci and Corinna Lange for assistance with collecting samples. Tim Page, Daniel Schmidt, James Fawcett, Joel Huey, Arlene Wheatley for road kill. I am also very grateful to all landowners throughout the Top End and Kimberley who kindly allowed us access to their property.

I am very grateful to all the people that taught me lab techniques, especially Jing Ma, David Gopurenko, Mia Hillyer and Steve Smith. Thanks to the genetics lab and CBBBR for insightful discussions. I appreciate the statistical support, especially from Rod Eastwood, Dan Schmidt and Rachel King. Thanks to the tree building boys Dan, Tim and Rod. James Holman provided a program to calculate great circle distance. Griffith University computing support assisted in running LAMARC . Thank you to Mark Ponniah for comments on the final draft.

A very special thank you to Petney Dickson for help with all my little problems.

Thank you to my parents, family and friends, Ange and Ben for all their support. Also to my old uni buddies, Katie and Jo for all the encouragement. Cheers to my old roomies Kate Durrant and Giovannella Carini and to my new roomies Ryan and Tayner.

Ian Lowndes for your support and understanding, a ‘massive big’ thank you, I couldn’t have done this without you.

This research was funded by Australian Geographic Society and Griffith University. The work was made possible with an APA and CAPRS Scholarship. Thank you kindly to Birds Australia for providing atlas records.

Samples were collected under Western Australian Department of Conservation and Land Management permit SF003903, Queensland Parks and Wildlife permit W4/002670/01/SAA, Northern Territory Parks and Wildlife Commission permit 18145 and animal ethics protocol AES/16/04/AEC. Samples were contributed by Queensland Museum, Western Australian Museum and Museum Victoria.

VII

TABLE OF CONTENTS

CHAPTER 1: GENERAL INTRODUCTION...... 1 1.1 GENETIC VARIATION IN POPULATIONS ...... 1

1.2 GENE FLOW, NEUTRAL EVOLUTION AND THE COALESCENT ...... 1

1.3 MOLECULAR TOOLS IN EVOLUTIONARY STUDIES ...... 2 1.3.1 Allozymes ...... 2 1.3.2 Mitochondrial DNA ...... 3 1.3.3 Microsatellites...... 5

1.4 PHYLOGEOGRAPHY ...... 7 1.4.1 Morphological Variation...... 8 1.4.2 Demographic Change ...... 9 1.4.3 Nested Clade Analysis ...... 10 1.4.4 Avian Phylogeographic Studies ...... 10

1.5 CLIMATE CHANGE IN AUSTRALIA ...... 11

1.6 PHYLOGEOGRAPHY OF AUSTRALIAN BIRDS ...... 13

1.7 HOST -PARASITE PHYLOGEOGRAPHY ...... 14

1.8 STUDY SPECIES ...... 16 1.7.1 Australian Magpie Gymnorhina tibicen...... 16 1.8.2 Ectoparasites of the Australian magpie...... 19

1.9 STUDY AIMS AND HYPOTHESES ...... 19

CHAPTER 2: GENERAL METHODS...... 22 2.1 SAMPLING DESIGN ...... 22

2.2 FIELD SAMPLING TECHNIQUES ...... 23

2.3 MOLECULAR TECHNIQUES ...... 24 2.3.1 G. tibicen...... 24 2.3.1.1 DNA extraction techniques...... 24 2.3.1.2 mtDNA amplification conditions and screening...... 25 2.3.1.3 Sequencing ...... 27 2.3.1.4 Microsatellite amplification and screening ...... 28 2.3.2 Philopterus sp. and B. semiannulata amplification ...... 29 2.4.2.1 Extraction techniques ...... 29 2.3.2.2 Amplification of mtDNA...... 29 2.3.2.3 Sequencing ...... 30

2.5 DATA ANALYSIS ...... 31 2.5.1 MtDNA...... 31 2.5.1.1 Phylogenetics...... 31

VIII 2.5.1.2 Population statistics ...... 31 2.5.2 Microsatellites...... 31 2.5.2.1 Tests for Hardy-Weinberg proportions ...... 31 2.5.2.2 Microsatellite variation...... 32

CHAPTER 3: POPULATION ANALYSIS OF G. TIBICEN SAMPLES ...... 33 3.1 INTRODUCTION ...... 33

3.2 METHODS ...... 35 3.2.1 Population Structure...... 35 3.2.1.1 Hierarchical analysis ...... 35 3.2.1.2 Bayesian analysis ...... 36 3.2.2 Isolation by Distance...... 37 3.2.3 Estimating Levels of Gene Flow ...... 38

3.3 RESULTS ...... 39 3.3.1 MtDNA Sequence Variation and Patterns of Genetic Diversity...... 39 3.3.3 Population Structure...... 45 3.3.3.1 Hierarcichial analysis ...... 45 3.3.3.2 Clustering analysis...... 47 3.3.3 Isolation by Distance...... 48 3.3.4 Estimating Levels of Gene Flow ...... 50

3.4 DISCUSSION ...... 51 3.4.1 Population Structure and Contemporary Barriers to Gene Flow...... 51

CHAPTER 4: PHYLOGEOGRAPHY OF THE AUSTRALIAN MAGPIE...... 57 4.1 INTRODUCTION ...... 57

4.2 SPECIFIC METHODS ...... 58 4.2.1 Phylogeographic Analysis...... 58 4.2.2 Tests for Historical Demographic Change...... 58 4.2.3 Coalescent Analyses...... 60

4.3 RESULTS ...... 61 4.3.1 Phylogenetic Analysis ...... 61 4.3.2 Phylogeography ...... 62 4.3.3 Demographic Expansion...... 66 4.3.4 Coalescent Analyses...... 70

4.4 DISCUSSION ...... 72 4.4.1 Phylogeography of the Australian magpie...... 72

CHAPTER 5: COMPARATIVE ANALYSIS OF TWO SPECIES OF HOST SPECIFIC FEATHER LICE ...... 80 5.1 INTRODUCTION ...... 80

5.2 SPECIFIC METHODS ...... 83 IX

5.2.1 Genetic Variation...... 83 5.2.2 Nuclear Pseudogenes...... 83 5.2.3 Hierarchical Analysis...... 84

5.3 RESULTS ...... 85 5.3.1 Genetic Variation...... 85 5.3.2 Phylogeny of Philopterus sp...... 87 5.3.3 Phylogeny of Brueelia semiannulata ...... 88 5.3.4 Nuclear Pseudogenes...... 89 5.3.5 Distribution of Haplotypes ...... 90 5.3.4 Hierarchical Analysis...... 93

5.4 DISCUSSION ...... 94 5.4.1 Population Structure for two species of feather lice...... 94 5.4.2 Secondary Contact ...... 98 5.4.3 Divergent Lineages ...... 99

CHAPTER 6: GENERAL DISCUSSION...... 102 6.1 A RECONSTRUCTION OF THE RECENT EVOLUTIONARY HISTORY OF THE AUSTRALIAN MAGPIE ...... 102

6.2 SIGNIFICANCE OF BACK COLOUR AND LATITUDINAL PLUMAGE VARIATION .... 107

6.3 THE PLEISTOCENE AND INTRASPECIFIC GENETIC VARIATION OF AVES ...... 109

REFERENCES ...... 110

APPENDICIES ...... 123 APPENDIX I: SITE LOCATIONS AND CONTROL REGION MT DNA HAPLOTYPES ...... 123 APPENDIX II: VARIABLE SITES IN G. TIBICEN CONTROL REGION ...... 126 APPENDIX III: PAIRWISE MT DNA FST VALUES AMONG MAGPIE SITES ...... 127 APPENDIX IV: PAIRWISE MICROSATELLITE FST VALUES AMONG MAGPIE SITES ... 128 APPENDIX V: VARIABLE SITES IN COI FRAGMENT FOR PHILOPTERUS SP ...... 129 APPENDIX VI: VARIABLE SITES IN COI FRAGMENT FOR B SEMIANNULATA ...... 130 APPENDIX VII: AMOVA RESULTS FOR PHILOPTERUS SP . AND B. SEMIANNULATA .132 APPENDIX VIII: FREQUENCY DATA OF SIX MICROSATELLITE LOCI IN G. TIBICEN 133

X LIST OF FIGURES

Figure 1.1 : Hypothetical models of genetic variation in natural populations...... 8 Figure 1.2 : Sea level fluctuations during the last 240, 000 years…………………...…12 Figure 1.3 : Putative refugia and isolating barriers in Australia Barriers……………....14 Figure 1.4 : The distribution of Australian magpie plumage forms…………………....17 Figure 2.1 : Sampling sites for the Australian magpie……………………………...….23 Figure 3.1 : Frequency of mtDNA control region ……………………………………..41 Figure 3.2 : Parsimony network for control region sequence data…………………...... 42 Figure 3.3 : Multidimensional Scaling Plot (MDS) of microsatellite variation……...... 45

Figure 3.4 : BAPS mixture clustering……………………………………………………47

Figure 3.5 : Admixture results from BAPS for the Australian magpie………………….48 Figure 3.6 : Scatterplot of geographic distance against Slatkin’s linearised FST between pairs of eastern magpie populations…………………………………………………....49 Figure 3.7 : Scatterplot of geographic distance against Slatkin’s linearised FST between pairs of western magpie populations…………………………………………………...50 Figure 3.8 : Migration estimates per generation between sites from LAMARC…….....51 Figure 4.1 : Maximum-likelihood tree showing the relationships among 46 magpie mitochondrial control region haplotypes…………………………………………….....63

Figure 4.2 : Mitochondrial parsimony network computed by TCS ……………………..64 Figure 4.3 : Mismatch distribution of magpie haplotypes …………………..……...... 68 Figure 4.4 : Mismatch distribution of magpie haplotypes for each clade………………70 Figure 4.5 : Results from IM for eastern and western Australia………………….……72 Figure 4.6 : Results from IM for eastern Australia and Tasmania……………………..72 Figure 5.1 : Maximum likelihood tree of COI gene for Philopterus sp...……………...88 Figure 5.2 : Maximum likelihood tree of COI gene for B. semiannulata . …………….89 Figure 5.3 : Parsimony network among populations of Philopterus sp………………..91 Figure 5.4 : Parsimony network among B. semiannulata populations…………...…….93 Figure 6.1 : Distribution of variation in the Australian magpie and two species of obligate feather lice………………………………………………….………………..103 Figure 6.2 : Scenario explaining population divergence and range expansion in the Australian magpie……………………………………………………………………..105

XI

LIST OF TABLES

Table 1.1 : Morphology of G. tibicen ………………………………………………...... 18 Table 2.1 : Primer sequence for microsatellite loci isolated from G. tibicen ……....…..28 Table 3.1 : Genetic diversity in Australian magpie populations………………………..40

Table 3.2 : FIS values for magpie populations at six microsatellite loci………………..43 Table 3.3 : AMOVA results for magpie sites……………………………………….….46 Table 4.1 : Nested clade analysis for populations of G. tibicen ………………….....….65 Table 4.2 : Growth (g) simulation results from LAMARC for G. tibicen …………..…66 Table 4.3 : Coalescent parameters and population divergence estimated with IM…..…71 Table 5.1 : Haplotype frequencies for COI variation for Philopterus sp…………...…..86 Table 5.2 : Haplotype frequencies for COI variation for B. semiannulata ………...…...87 Table 5.3 : AMOVA results for Philopterus sp. and B. semiannulata ……………...….94

XII Declaration

This work has not been submitted for a degree or diploma at this or other universities. To the best of my knowledge and belief, this thesis contains no material previously published or written by other persons except as acknowledged in the text.

Alicia Toon

XIII

XIV General Introduction

CHAPTER 1: GENERAL INTRODUCTION

1.1 GENETIC VARIATION IN POPULATIONS Genetic variation within populations is generated by a balance between accumulation and loss of mutations and the transfer of genes among populations (Slatkin). Mutations change in frequency within populations as a result of genetic drift and/or selection where genetic drift is the random sampling of alleles within a population that can lead to fixation at neutral loci in the absence of gene flow (Slatkin 1987). Genetic variation can be maintained at high levels within populations through balancing selection, or natural selection can lead to fixation of traits via directional selection (Spiess 1989). In the absence of natural selection, genetic drift is likely to lead to differentiation among populations while gene flow will tend to homogenise gene frequencies. The impact of genetic drift and time to fixation of alleles will depend on generation time and the effective population size ( Ne) (Page and Holmes 1998). A short generation time and/or a small effective population size will lead to the rapid fixation of neutral alleles. Therefore, genetic variation within and among populations is largely the result of the interplay between natural selection and genetic drift effects.

1.2 GENE FLOW, NEUTRAL EVOLUTION AND THE COALESCENT To explain how genetic variation was exchanged among populations, Wright (1943) developed a simple model termed the ‘island model’ where there was uniform gene flow among all populations. Subsequently a more complex model was proposed that included the idea that genetic differentiation increased with geographical distance in a linear fashion. This model was termed ‘isolation by distance’ (Wright 1943).

With the advent of protein markers (allozymes), a large amount of data on patterns of genetic variation within and among populations could be accessed to test ideas about gene flow among natural populations. In addition, empirical studies of interspecific variation led to the neutral theory of evolution, developed by Kimura in 1968. He proposed that a large proportion of natural genetic variation resulted from genetic drift

1

General Introduction rather than selection (Page and Holmes 1998). Although natural selection was still considered an important mechanism for evolution, proponents of the neutral model of evolution suggested that most mutational changes did not affect an individual’s fitness and therefore will act essentially as silent mutations (Page and Holmes 1998).

F-statistics developed by Wright can be used to measure genetic variation within and among populations (Wright 1978). The inbreeding coefficient ( FST ) compares the variation (heterozygosity) within each population to that within all populations based on the sampled allele frequencies (Spiess 1989). FST is, in practice, an estimate of differentiation between two or more populations and may be used to partition variation at different hierarchical levels (eg. AMOVA). Although new methods that take into account genetic sequence data are increasingly used, F-statistics are still a useful measure of population structure and divergence when assumptions are met (Neigel 2002).

Recently, coalescent theory has been used to develop advanced models to describe population structure. Coalescent theory, originally proposed by Kingman (1982a, 1982b), assumes a neutral model of evolution (Wright-Fisher model) and describes how alleles/haplotypes are related back through time. The additional information provided by the genetic relationships among alleles allows a better understanding of population processes than using traditional diversity approaches (eg. F-statistics) (Fu and Li 1999).

1.3 MOLECULAR TOOLS IN EVOLUTIONARY STUDIES

1.3.1 Allozymes While recent advances in molecular techniques have provided population biologists with powerful tools to expand their knowledge on evolution of species, in the past many studies have used protein markers (allozymes) for this purpose, due to benefits of relatively low cost and time efficiency (Queller et al. 1993). Allozyme loci tend however, to display relatively low levels of population differentiation particularly for

2 General Introduction many avian species (Zink 1997). Low variation may be attributed to conservation of the avian genome due to the high metabolic rate found in birds (Avise 1983, Zink 1991). Variation in non-coding genes therefore, may provide more information about the evolutionary history of a species than in protein coding genes. Recent history of shared ancestry and high dispersal capacity of birds are alternative explanations as to why avian populations generally display low allozyme variability compared with other vertebrates (Barrowclough 1983, Crochet 2000). Crochet et al. (2003) reported comparatively low levels of nuclear differentiation (allozymes and microsatellite loci) compared with mitochondrial gene divergence among species of large white headed gulls ( Larus spp.). The authors suggest the pattern is most likely explained by a recent ancestry for the species and subsequent interspecific gene flow. Although this question has yet to be resolved, other problems may arise from using allozymes. For example, divergent patterns of allozyme variation may be due to selection on proteins rather than historical processes or contemporary gene flow. Allozyme studies may also require fresh tissue which is not always available. Most importantly, evolutionary relationships among alleles cannot be reconstructed with allozyme data because the underlying sequence variation is not known for electromorphs (bands with different electromobility).

1.3.2 Mitochondrial DNA Several attributes of mitochondrial DNA have led to its wide popularity and utility for analysis of genetic variation within a species. mtDNA is maternally inherited and haploid resulting in an effective population size that is one quarter of the size of nuclear DNA (Birky et al. 1989). The reduction in effective population size combined with a high mutation rate for mitochondrial DNA leads to an increased effect of genetic drift which in turn boosts its sensitivity for resolving recently derived phylogenies (Moritz et al. 1987). mtDNA can also be analysed in conjunction with nuclear DNA markers to provide an indication of sex-biased dispersal because of the different modes of inheritance (maternal or biparental) (eg. Crochet et al. 2003).

3

General Introduction

Recent phylogeographic studies have focused on mtDNA due to the relative ease with which sequence information can be attained. Sequence data allows construction of genealogies using relationships among lineages. If information about the molecular divergence rate is also known, a molecular clock can be applied to estimate time to the most recent common ancestor (TMRCA) of a sample of genes, either within a population or between two divergent lineages (Arbogast et al. 2002). This allows the timing of historical events such as fragmentation of populations, due to vicariance or founder events, or the timing of a population expansion. There are several potential problems however, that may arise from assuming a molecular clock. A molecular clock relies on a sequence substitution rate estimated from an independent source (fossil data, geological data, pedigree data) for each DNA fragment used and the species of interest. Independent data are not available for many species so estimates must rely on inferences from those calculated for closely related species. Even for closely related species however, heterogeneity among lineages may be a source of error for estimating divergence when sequence substitution rates are not based directly on the species of interest (Arbogast et al. 2002, Ruokonen and Kvist 2002). Another potential problem is to assume that a molecular clock is linear through time (Ho et al. 2005, Ho and Larson 2006). Rates of mutational change may decrease over time from initially high short term mutation rates, later to much slower and eventually a stable substitution rate (Ho et al. 2005). Ho et al. (2005) suggested that declines in rates of mutational change are due to purifying selection, and saturation. For population level studies that have used slower species-level substitution rates, timing of recent events may be overestimated.

Another issue to be addressed when using mtDNA to reconstruct the history of a species is that it represents a single marker. The stochastic nature of linage sorting, genetic drift in ancestral populations and past selection may cause errors in estimating divergence times when information comes from a single locus (Edwards and Beerli 2000).

One of the potential problems that arises from direct sequencing of mitochondrial genes is the accidental sequencing of a homologue of the target fragment. Nuclear

4 General Introduction homologues of mtDNA, or Numts as they are known, have been detected in a range of invertebrate and vertebrate species including birds (see Sorenson and Fleischer 1996, Sato et al. 2001). Numts are the result of gene transfer of a fragment of mtDNA to the nuclear genome. Due to the loss of function on transfer to the nuclear genome and subsequent independent evolution, Numts are likely to display considerable divergence to the original protein coding mitochondrial gene. Therefore if they are not detected, it may lead to an erroneous conclusion of divergence between individuals sequenced for the Numt and the mitochondrial gene. Studies to date suggest that they are less frequent in birds than other vertebrates, which is possibly due to the smaller conserved mitochondrial genome in birds compared with other taxa (Pereira and Baker 2004). Numts may be detected in sequences via visual inspection of sequence data for the presence of stop codons or frameshift changes in inappropriate places (Bensasson et al. 2001).

1.3.3 Microsatellites Microsatellites are short tandem repeats (STRs) that are abundant across the genome of most species (Jarne and Lagoda 1996). Variation in length allows differences between alleles to be easily detected with electrophoretic gel analysis. Microsatellites are non- coding, and so are assumed to be selectively neutral, and they have been shown to evolve relatively rapidly for nuclear markers resulting in high amounts of genetic variability (Ellegren 2004). High variability of microsatellites make them powerful markers for detecting recent changes in demography, distribution of genetic variation and for estimating levels of gene flow. They are therefore useful for population studies, particularly for addressing questions of contemporary fine scale genetic differentiation and recent population history.

A few problems with microsatellite markers have been identified recently (Balloux et al. 2000, Zhang and Hewitt 2003, Ellegren 2004). The relative cost of developing microsatellite markers per individual remains high when compared with allozymes (Jarne and Lagoda 1996). Also there is continued debate over the mutation model by

5

General Introduction which microsatellites evolve (Ellegren 2004). Choosing an incorrect mutation model can lead to erroneous biological conclusions and as such conservative approaches are often adopted when analysing data that can result in a loss of power that microsatellite markers could otherwise provide. Another consideration for population studies is that mutations at microsatellite loci can result in size homoplasy. Homoplasy refers to alleles that appear to be identical by decent but may have arisen via a different mutational history and in some cases from a different ancestral allele (Jarne and Lagoda 1996). Homoplasy can lead to underestimating variation. However, Estoup et al. (2002) demonstrate that in most situations size homoplasy is offset by the large amount of variation present and therefore should not be a significant problem for most population level studies. Although the above issues need consideration when using microsatellites, they have proven to be powerful markers for a variety of studies, including parentage analysis (Hughes et al. 2003), detecting fine scale structure (Shaw et al. 1999), assigning individuals to populations for conservation purposes (Hansen et al. 2006), for detecting recent demographic change (King et al. 2000) and, when used in combination with mtDNA, for detecting sex biased dispersal (Crochet et al. 2003).

Combining sequence information from mitochondrial markers with multi-locus allelic data available from microsatellite markers provides very powerful independent markers by which many evolutionary hypotheses can be tested. When comparing results from mtDNA and nuclear markers it is necessary to take into consideration that mitochondrial genes are haploid and maternally inherited, whereas nuclear genes are diploid and are biparently inherited. This means that mitochondrial genes will reach reciprocal monophyly on average four times faster than nuclear genes (Moritz et al. 1987). Therefore, we would expect to see a lag in microsatellite allele divergence among isolated populations behind that of mtDNA divergence. In addition, we would expect estimates of gene flow up to 4 times higher for nuclear genes than for mitochondrial genes. These issues will be considered when comparing analyses based on mtDNA and microsatellite loci.

6 General Introduction

1.4 PHYLOGEOGRAPHY In the 1987 landmark paper, ‘Intraspecific Phylogeography: The mitochondrial DNA bridge between population genetics and systematics’, Avise et al. (1987) coined the term phylogeography by linking the previously independent fields of phylogenetics and biogeography. Phylogeography builds on the field of historical biogeography by combining molecular genetics with current geological theory to explain the contemporary distribution of genetic variation within a species. The new discipline arose largely from the new found popularity of mtDNA as an intraspecific genetic marker that could be utilised to assess phylogenetic relationships within species (Avise et al. 1987). By overlaying the phylogeny of a species on to its geography (spatial distribution), hypotheses can be tested about the past distribution of a species across evolutionary time.

Avise et al. (1987) described five general phylogeographic patterns compiled from empirical data. The first two categories described discontinuous patterns of genetic divergence that could arise via (I) isolation of populations due to vicariance or very limited gene flow (Figure 1.1a) or (II) past isolation followed by recent secondary contact or intrinsic barriers among sympatric sister species (Figure 1.1b). Divergence in a neutral gene tree typically indicates category I or II explanations. While the depth of divergence may be used to indicate the time since population isolation, as explained in the previous section, this depends on an appropriate mutation model and rate. The final three categories describe continuous patterns of genetic variation that may result from (III) restricted gene flow, (IV) extensive gene flow (Figure 1.1c) or (V) intermediate levels of gene flow. Extensive gene flow throughout a population will lead to panmixia at neutral loci. However, limited dispersal capacity may restrict gene flow at some scales leading to different patterns of genetic variation among taxa.

7

General Introduction

A.

C.

B.

Figure 1.1: Hypothetical models of genetic variation in natural populations. Each circle represents the proportion of an allele or gene within sampling region 1 (white) or sampling region 2 (black). Connecting lines and cross bars represent the degree of difference among alleles/genes. Pattern A. may be formed through historical isolation among populations and subsequent divergence via genetic drift or natural selection. The gene tree displayed in B. may be formed via historical isolation of two populations with recent secondary contact. A star-like network shown in C. represents high levels of gene flow among populations (panmixia). The ancestral haplotype and many tip haplotypes suggest that this pattern was formed through a recent demographic expansion.

1.4.1 Morphological Variation Phylogeographic patterns can also be assessed in conjunction with morphological variation to consider the relative influence of life-history traits and intraspecific factors (eg. sexual selection). By using a phylogenetic framework, trees based on the morphology of different taxa can be compared with the molecular phylogeny and if tree topologies agree, inferences can be made as to when morphological or behavioural characters first arose (Sheldon and Whittingham 1997). This method is also applicable to intraspecific comparisons and has been particularly useful in avian studies because of the large amount of phenotypic variability in plumage common in many avian taxa. Vicariance and historical isolation has long been suggested as a possible explanations for the highly polytypic nature of many avian species (see Ford 1987, Cracraft 1986, Cracraft and Prum 1988). Support for this hypothesis relies on concordance in divergence patterns for both plumage characters and phylogeny. Concordance has been

8 General Introduction reported for several taxa (eg. Small et al. 2003, Barrowclough et al. 2004). Alternatively, variation in plumage may develop in situ (primary intergradation) via strong selection across an ecological gradient (eg. Hughes et al. 2001). A pattern where morphology is structured geographically with no apparent structure detected in the phylogeny, suggests primary intergradation. However, if the phylogeny is shallow suggesting a recent history, the discordant pattern may also be explained by a recent range expansion as suggested for the swamp sparrow Melospiza georgiana (Greenberg et al. 1998), the red-winged black bird Agelaius phoeniceus , the chipping sparrow Spizella passerina , and song sparrow Melospiza melodia (Zink 1996).

1.4.2 Demographic Change Past demographic events such as population expansions or bottlenecks will leave a signature in the genetic data. A recent population expansion should result in a star-like network (Figure 1.1c.) with few ancestral (central) haplotypes and many recently derived (tip) haplotypes (Schneider and Excoffier 1999). In contrast, populations that are depauperate in genetic variation can indicate that they have been exposed to a recent bottleneck. Rogers and Harpending (1992) introduced a method to test for population expansion using a mismatch distribution of nucleotide differences among individuals. Under a model of sudden population expansion, a peak or ‘wave’ indicative of a past expansion is formed from the distribution of pairwise differences. Testing the significance of mismatch distributions has largely been based on assessing the difference between distributions of a population at equilibrium and those under sudden expansion estimated from the data (SSD, Schneider and Excoffier 1999) or a raggedness index (rg) that tests the ‘smoothness’ of a distribution (Harpending et al. 1993). However, it has been suggested that these methods may not make effective use of the data and are therefore overly conservative (Felsenstein 1992). Moreover, Ramos-Onsins and Rozas (2002) demonstrated via simulation that methods that use mutation information (segregating sites) within a sequence (eg. Tajima’s D, R 2) and methods that use haplotype information (eg. Fu’s FS) are the most powerful methods for detecting recent rapid population growth.

9

General Introduction

1.4.3 Nested Clade Analysis Difficulties in distinguishing historical from alternative contemporary factors has led to the development of nested clade analysis (NCA) (Templeton 1998). NCA uses a permutation test to test for the significance of the geographical association of the distribution of genes/haplotypes based on a gene tree. Where significant associations are detected, an inference key is used to determine between contemporary and historical explanations. This approach however, has been criticised as to the ability of NCA in distinguishing between alternate hypotheses (Knowles and Maddison 2002). Knowles and Maddison (2002) criticised NCA because they did not believe that the method could assess statistical significance among alternate hypotheses. Given this issue, NCA is at least useful for describing the geographical distribution of haplotypes and for inferring possible scenarios using a set of transferable rules. Templeton (2004) has since shown NCA to be effective at interpreting phylogeographic scenarios where a priori hypotheses are available. Future empirical research in this field would benefit from rigorous statistical methods that can assess the statistical error associated with alternate historical and contemporary hypotheses combined with parameter estimators to provide a more comprehensive phylogeographic analysis (Knowles and Maddison 2002).

1.4.4 Avian Phylogeographic Studies Intraspecific avian phylogeography has largely focused on how climate fluctuations during the Pleistocene have contributed to modern patterns of geographic variation within species. The Pleistocene (1.8 million years ago – 10, 000 years ago) was characterised by periods of glaciation (at high latitudes) or aridification (at low latitudes) interspersed with warmer pluvial intervals (Hope 1994). Recent studies have revealed a range of phylogeographic patterns and intraspecific coalescence times indicating that effects of past glaciation have probably varied among avian species (see Zink 1996, Avise and Walker 1998, Qu et al. 2005). Specifically, rapid range expansion from Pleistocene refugia may result in a lack of genetic structure across the modern range of a species as suggested for the swamp sparrow, Melospiza georgiana (Greenberg et al. 1998) and snow finch, Pyrgilauda ruficollis (Qu et al. 2005).

10 General Introduction

Alternatively, some studies have found populations that exhibit distinct phylogenetic lineages, likely to be the result of past isolation (eg. Holder et al. 1999, Peters et al. 2005). Intermittent periods of high gene flow during warmer intervals may have facilitated the mixing of independent lineages that evolved in isolation as reflected in the gene tree for the snow goose, Chen caerulescens (Quinn 1992). The dating of population expansions and population divergence based on mitochondrial genes suggests that these events probably occurred within the last 1.8 million years. It is therefore apparent, that Pleistocene climatic fluctuations have played an important role in determining the genetic and geographic structure of a number of avian species. There is even the suggestion that Plio-Pleistocene climatic fluctuations were a catalyst for major speciation events, however at present this debate has not been resolved (Cracraft and Prum 1988, Klicka and Zink 1997, Johnson and Cicero 2004, Jennings and Edwards 2005).

1.5 CLIMATE CHANGE IN AUSTRALIA Australia is characterised by many of the oldest land formations (Taylor 1994). Over one third of Australia is described as arid or Eremean (Barlow 1981). However, early in the Tertiary (65 – 1.8 MYA ), Australia was virtually covered by forests. The abundance of conifers and Nothofagus forests at this time suggests cool to warm temperate conditions with high rainfall (Greenwood 1994). Towards the end of the Tertiary, an increase in dry adapted species in the interior suggests that the Australian continent had already commenced drying out (Kershaw et al. 1994). The Pliocene and Pleistocene in Australia were characterised by fluctuating climates between arid cool phases (glacial activity at high latitudes) and comparably warmer mesic periods (pluviality) (Kershaw et al. 2003). During this time, the landscape of the interior changed to conditions favourable for the dry adapted species that are present throughout inland Australia today. Many of the animal and plant species adapted to higher rainfall conditions now largely confined to the coast may have been isolated during this time in refuges at peak periods of aridity.

11

General Introduction

Arid phases of the Pleistocene coincided with a eustatic drop in sea level (Figure 1.2).

Sea levels around Australia dropped to ≈ 120 meters below present levels (Chappell 1983, Chappell and Shackleton 1986, Fleming et al. 1998) at the peak of aridity. Rather than simply isolating terrestrial populations, sea level fluctuations may have assisted in formation of land corridors between populations that had been previously isolated by sea. For example, the channel that separates Tasmania from the mainland, Bass Strait, is currently 50-70 meters deep (Evans and Middleton 1998). During the last period of aridity, sea levels dropped below 50 meters and a land bridge was formed between mainland Australia and Tasmania that may have allowed interchange of terrestrial flora and fauna. In addition, land bridges may have been established across southern Australia (south of Nullarbor Plain) and northern Australia (north of Carpentarian divide). As the sea level dropped, eastern and western populations were connected by mesic habitat rather than isolation characteristic of peak periods of aridity (Keast 1961, Voris 2000). Variation in aridity and fluctuating sea levels are likely to have affected the distributions of many avian species. While many studies have reported the impacts of past glaciation cycles, few have considered the associated effect of aridification on terrestrial species during the Pleistocene (but see Bowie et al. 2006, Joseph and Wilke 2006).

LGM GM

0 -50

-100

Sea Sea level -150 0 40 80 120 160 200 240 Time (kyr)

Figure 1.2: Sea level fluctuations during the last 240, 000 years. Low sea levels corresponded with cool dry periods during the Pleistocene in Australia. LGM indicates the last glacial maximum and GM indicates the glacial maximum that preceded the LGM. Sea level curve adapted from Chappell (1983), Chappell and Shackleton (1986), Fleming et al. (1998).

12 General Introduction

1.6 PHYLOGEOGRAPHY OF AUSTRALIAN BIRDS Early workers (eg. Keast 1961, Ford 1987) reported that more than half of Australian avian species exhibit morphological variation that is structured at biogeographical boundaries. They proposed that past vicariance as recent as the Pleistocene has played an important role in determining patterns of intraspecific morphological diversity and speciation events in Australian bird species. Genetic analyses, however, have yet to link Pleistocene climatic oscillations with modern distributions of plumage forms for many polytypic Australian avian species. Moreover, a number of recent studies (Hughes et al. 2001, Toon et al. 2003) have reported a general lack of correlation between neutral mtDNA variation and geographically structured plumage variation, suggesting that local selection pressures may be more significant than historical factors in determining the current distribution of plumage variation for some Australian bird species.

Recent studies have identified a number of possible habitat refugia and barriers to gene flow in Australia as the continent dried out. The Black Mountain Corridor of the Wet Tropics in North Queensland was hypothesised to have been an effective barrier to dispersal for many rainforest bird and other vertebrate species from the late Tertiary to the late Pleistocene (Joseph et al. 1995). Historical barriers have also been identified in south-eastern Australia for satin bowerbirds ( Ptilonorhynchus violaceus , Nicholls and Austin 2005) and some species of invertebrates (Eastwood et al. 2006). For widespread species, the onset of aridity in the Pleistocene led to the contraction of preferred habitat into isolated refugia around the coast and in central Australia (Figure 1.3) due to the formation of extensive sand dune systems (Bowler 1976). At peak aridity during glacial cycles it is expected that populations of birds and other organisms were fragmented across Australia due to loss of preferred habitat. However, it is not well understood which barriers have been important in shaping natural populations. Studies of avian genetic variation have suggested that arid barriers across southern and central Australia, such as the Eyrean, Mallee and Nullarbor barriers (Fig. 1.3), could have been important Pleistocene barriers to dispersal for some avian species (Degnan and Moritz 1992, Driskell et al. 2002, Joseph and Wilke 2006). For those species distributed in northern

13

General Introduction

Australia, the Carpentarian barrier may also have been an important barrier to dispersal during arid phases. Analysis of sequence variation has suggested that divergence between eastern and northern populations of grey crowned babbler Pomatosomus temporalis , concordant with the putative Carpentarian barrier (Edwards and Wilson 1990), may have resulted from this process. Less is known however, of the extent to which putative barriers in the west (Bonaparte, Canning and Murchison barriers) have contributed to modern distributions of avian populations.

Arnhem Land North- East Kimberley 8. 1. 7. 2.

Central Eastern Hamersley Ranges 6.

5.

4.

South-West 3. Eyre Peninsula

Mt. Lofty

Figure 1.3: Putative refugia and isolating barriers during glacial/arid periods in Australia for widespread avifauna adapted from Ford (1987). Barriers: 1. Carpentarian 2. Burdekin 3. Mallee 4. Eyrean 5. Nullarbor 6. Murchison 7. Canning 8. Bonaparte. Highland barriers and rainforest refugia along the east coast are not shown.

1.7 HOST -PARASITE PHYLOGEOGRAPHY Comparative host-parasite studies have highlighted several features of parasite biology, which make them particularly useful in understanding micro-evolutionary processes such as inferring dispersal among populations or for uncovering the recent history of a host. Many species of obligate parasites have a close relationship with their host (host

14 General Introduction specific) due to the dependence upon a host for food and habitat and due to limited dispersal abilities of parasites (Marshall 1981). Host specificity however, varies among species of parasites (Johnson et al. 2002b). Where host specificity is strong, it is expected that the phylogeny of the louse will reflect the phylogeny of the host (Klassen 1992). Parasites with low host specificity are unlikely to share phylogenetic patterns with their host and are therefore unsuitable for population studies. Even for species that display strong host specificity, the occasional dispersal to atypical hosts (host transfer) has been observed (Whiteman et al. 2004). Successful colonisation of atypical hosts (host switching) however, is likely to be constrained by the body size of a new host (Clayton et al. 2003b) thus limiting the frequency of host switching events. Two or more species of parasites may be used as effective replicates on a single host species to compare genetic variation and to infer different dispersal potential for each louse species (eg. Johnson et al. 2002b, Clayton and Johnson 2003).

Some species of parasites such as chewing lice are typically sensitive genetic markers. Studies have shown that the mitochondrial genome of lice evolves up to three times more rapidly than their avian hosts (Page et al. 1998) and up to ten times faster than in mammalian hosts (Hafner et al. 1994). It has been hypothesised that either fast generation time (Hafner et al. 1994), small effective population sizes coupled with founder events (Page et al. 1998) or that lice lineages predate those of the host (Page 2003) may be responsible for the elevated rate of evolution in chewing lice. At present however, it is unclear exactly what is the cause. Due to an increased substitution rate, chewing lice can provide informative markers for inferring recent history or for events that occurred over a short period because signatures of the event may arise faster in lice. In a phylogeographic study of the field mouse ( Apodemus sylvaticus), Nieberding et al. (2004) used genetic variation from a parasitic nematode Heligmosomoides polygyrus to document a history of fragmentation for the host. Previous analysis of genetic variation of the field mouse host had not found any evidence of divergence that suggested historical fragmentation. The authors suggested that similar to chewing lice the parasitic nematodes genealogy was more sensitive than the host to recent historical events

15

General Introduction because it had a faster evolutionary rate. To date few studies have examined variation in avian parasites to infer the population dynamics and phylogeographic structure of an avian host (but see McCoy et al. 2003, McCoy et al. 2005).

1.8 STUDY SPECIES For the current study, I chose a widespread endemic passerine, the Australian magpie Gymnorhina tibicen and two species of host specific lice Philopterus sp . and Brueelia semiannulata to compare the phylogeography of each species.

1.7.1 Australian Magpie Gymnorhina tibicen The Australian magpie, belongs to the Family Artamidae (Order Passeriformes) and is more closely related to other Australian passerines (eg. Cracticus spp.) then northern hemisphere magpies (ie. Pica pica ) (Barker et al. 2004). The endemic distribution of the magpie includes mainland Australia and north-eastern Tasmania (Figure 1.4) as well as a small isolated population in southern Papua New Guinea (Black 1986). The Australian magpie is adapted to dry sclerophyll open woodland and occurs naturally across much of the continent. As a widespread generalist, the species has expanded its natural distribution to many urban and agricultural areas since European colonisation (Campbell 1929). Eight subspecies of the Australian magpie are currently recognised (Schodde and Mason 1999) across mainland Australia and Tasmania (Figure 1.4). Subspecies status is delineated on subtle differences in plumage colour, bill length and body size (Schodde and Mason 1999) with two main plumage forms recognised. The black-back plumage form (BB) is distributed across northern Australia and is easily identified by a broad black dorsal band. The white-back form (WB) has a narrower distribution across southern Australia and Tasmania. Magpies west of the Nullarbor Plain, which I will refer to here as the varied magpie, are a highly sexually dimorphic sub-group of the WB form, G. t. dorsalis . Males of G. t. dorsalis have a white dorsal pattern while females exhibit black feathers edged in white on their back. In mainland populations, body size and bill length vary in a north to south cline. Populations in the north-west of Australia have the smallest body size and the longest bill (Table 1.1), whereas the largest magpies are found in south-east populations of G. t. tyrannica .

16 General Introduction

Paradoxically, magpies from the Tasmanian population, G. t. hypoleuca have a small body size compared with southern mainland populations (Schodde and Mason 1999).

B

A C

D E F

G

H

Figure 1.4: The distribution of Australian magpie plumage forms adapted from Schodde and Mason (1999). Black backed forms: (A) G. t. longirostris (B) G. t. eylandtensis (C) G. t. terraereginae (D) G. t. tibicen White backed forms: (E) G. t. dorsalis (F) G. t. telonocua (G) G. t. tyrannical (H) G. t. hypoleuca Zones of intergradation between back colour forms. Shows the limit to the distribution of G. tibicen .

17

General Introduction

Table 1.1: Back colour description and morphological attributes for each plumage form of G. tibicen based on Schodde and Mason (1999). sub species sex dorsal colour wing length exposed culmen exposed (mm) (mm) culmen/wing ratio eyandtensis ♂ black band 230-255 56-62 0.23-0.25 ♀ black band 225-245 51-57 0.21-0.23 terraereginae ♂ black band 245-265 48-58 0.20-0.23 ♀ black band 235-255 47-53 0.19-0.22 Tibicen ♂ black band 260-285 48-55 0.18-0.21 ♀ black band 255-270 45-50 0.17-0.20 tyrannica ♂ White 270-290 52-57 0.18-0.21 ♀ Grey 260-280 47-53 0.17-0.20 hypoleuca ♂ White 248-258 43-47 0.17-0.18 ♀ Grey 235-245 38-43 0.16-0.18 telonocua ♂ White 255-265 50-56 0.18-0.22 ♀ Grey 245-255 45-50 0.17-0.21 dorsalis ♂ White 258-270 56-60 0.21-0.23 ♀ Black 240-255 48-54 0.20-0.22 longirostris ♂ black band 245-260 60-65 0.23-0.25 ♀ black band 235-250 55-60 0.23-0.25

BB and WB forms intergrade in eastern Australia (zone approximately 100km, Burton and Martin 1976) and the BB and varied forms also intergrade in Western Australia (zone approximately 500km wide, Schodde and Mason 1999). where both parental forms occur as well as intermediates with varying widths of the black dorsal band (Burton and Martin 1976, Hughes 1982). An analysis of mtDNA across the two contact zones revealed that genetic variation is not partitioned on back colour either in the east (Hughes et al. 2001) or the west (Toon et al. 2003). Rather, local selection pressures have been implicated in maintaining the geographical distribution of distinct plumage forms in the face of ongoing gene flow (Hughes et al. 2001). While no significant mitochondrial divergence was observed between northern (BB) and southern (WB, varied) plumage forms, populations across southern Australia are structured east to west, with no maternal gene flow evident across the Nullarbor Plain (Baker et al. 2000, Toon et al. 2003). Whether this pattern of east to west structure is also true among populations across northern Australia is unknown. Hughes et al. (2001) analysed a population of magpies from Tasmania for mtDNA variation and by comparison with mainland populations suggested that the Tasmanian population represented a recently isolated population of the WB form. Additional analysis of mitochondrial and nuclear

18 General Introduction gene diversity in Tasmanian popualtions is required to establish the timing of any population divergence from mainland populations.

1.8.2 Ectoparasites of the Australian magpie The Australian magpie is host to several ectoparasites including species of feather lice, ticks and hippoboscid flies. Two species of feather lice from the Order Ischnocera (Family Philopteridae), Philopterus sp. and Brueelia semiannulata and a third species from the Order Amblycera (Family Menoponidae), Myrsidea sp. are common body ectoparasites of the Australian magpie (Hughes 1984a). This study examined variation in only Philopterus sp. and B. semiannulata as Myrsidea sp. were only found on very few of the sampled magpie individuals.

Philopterus sp. is highly habitat specific and is restricted to the nape feathers. In contrast, Brueelia spp. infest body feathers and are considered to be habitat generalists (Clay 1951). Feather lice feed on dead skin tissue, feathers and blood (Gullan and Cranston 1994). They are obligate ectoparasites and therefore are dependent upon a host for survival and to complete their life cycle (Marshall 1981). This effectively limits dispersal to close contact among magpies. Magpies are highly territorial (Farabaugh et al. 1992) and close contact occurs most often among parents and nestlings or between mates. Contact between birds from different territories may also occur less frequently during territorial defence, extra-territory copulation (see Hughes et al. 2003, Durrant and Hughes 2005) or dispersal of juveniles to a flock (Veltman 1989) or new territory.

1.9 STUDY AIMS AND HYPOTHESES Although phylogeographic patterns of high latitudinal avian species associated with Pleistocene climatic fluctuations are well documented, few studies have focused on the associated effect of aridification on low latitudinal species. Moreover, phylogeographic studies on widespread endemic Australian species are limited due to the difficulties of collecting specimens over a large and in many places remote geographic distribution.

19

General Introduction

The aim of this study therefore was to investigate the relationship between past climate change and genetic differentiation in the Australian magpie and two species of parasitic feather lice. Using a phylogeographic framework, the geographical distribution of variation at six microsatellite loci and a fragment of the control region of mitochondrial DNA for the Australian magpie were examined across the complete Australian distribution of the species. In addition, mitochondrial cytochrome oxidase I (COI) variation was analysed for two host specific species of lice.

Previous studies have undertaken extensive studies of allozyme and control region variation for southern populations of the Australian magpie (see Baker et al. 2000, Baker et al. 2001, Hughes et al. 2001, Toon et al. 2003). In Chapter 3, I will describe microsatellite variation for southern magpie populations as well as both microsatellite and control region variation for populations sampled for the present study in northern Australia. Allelic variation and indirect estimates of gene flow (migration) were estimated and compared with those presented in earlier studies. The general aim is to understand the contemporary patterns of connectivity among all populations of the Australian magpie.

A phylogeographic approach was applied to resolve the recent history of Australian magpies based on microsatellite and control region variation, this analysis is presented in Chapter 4. Nested clade analysis was used to help describe patterns of variation in the magpie gene tree. Various statistical approaches were used to test for recent demographic changes and recently developed coalescent methods were employed to estimate divergence times among populations.

In the final data chapter, patterns of genetic variation were compared in two species of feather lice collected from the Australian magpie. The sensitivity of lice as phylogeographic markers were explored and comparative statistics were used to test the hypothesis that Philopterus sp . and B. semiannulata share a common recent history with the Australian magpie and to answer the question: are the two ectoparasites faithful to

20 General Introduction their host?

Chapter 6 provides a general discussion of the phylogeography of magpies and their louse ectoparasites and places them in a broad overview of the recent biogeographic history of Australia.

21

General Methods

CHAPTER 2: GENERAL METHODS

Laboratory methods and data analysis relevant to more than one chapter are described here. Methods and analysis pertaining to specific questions are provided in the relevant chapters.

2.1 SAMPLING DESIGN A total of 191 magpie samples from localities across mainland Australia were collected specifically for this study (Fig. 2.1, Appendix 1). Samples were collected under Western Australian Department of Conservation and Land Management permit SF003903, Queensland Parks and Wildlife permit W4/002670/01/SAA, Northern Territory Parks and Wildlife Commission permit 18145 and animal ethics protocol AES/16/04/AEC. Regional locations in the north and west of the continent that had not been sampled for previous studies provided the main sampling focus. For north-eastern Australia the natural densities of magpies were high enough to replicate the design of Baker (1999). Between 20 and 30 samples were collected from each of four sites in north-eastern Australia to represent a regional sampling location. This sampling design could not be adhered to northern and western Australia due to difficulties associated with catching live magpies where densities are very low. Therefore, samples were collected where possible across the north and west of the continent and in addition, 14 museum samples from various locations were included in the analysis. Samples were also collected from dead magpies (road kill) in central southern locations and in some parts of western Queensland and New South Wales. Samples available from previous studies and held at Griffith University were used to represent south western and south eastern coastal locations from mainland Australia (Baker et al. 2000, Durrant and Hughes 2005) and Tasmania (Hughes et al. 2001).

Lice infestation rates of magpies are reported to vary seasonally (Hughes 1984a). This earlier work indicated that July had the highest infestation rate of Philopterus sp . whereas B. semiannulata was highest between April and June. Although this was taken

22 General Methods into consideration when louse sampling, it was not always possible to coordinate sampling trips with peak infestation times of lice.

North 37 36 s 3 35 d 34 t e b f 4 5 c g 6 r 33 1 2 h 7 a 31 8 32 q i j 10 9 30 k 11 p l 24 29 12 14 28 o 27 25 13 26 23 22 m 16 15 n Sydney 17 18 20 19

21

Figure 2.1: Sampling sites for the Australian magpie on mainland Australia and Tasmania. Numbered sites represent locations where greater than 3 samples were collected. Letters refer to sites where 1-3 samples where collected and are not used in the population analyses. Refer to Appendix 1 for detailed information about each site.

2.2 FIELD SAMPLING TECHNIQUES Magpies sampled in this study were caught using an open-ended chicken wire cage placed over a caged live decoy magpie. Territorial defence against the decoy along with bait (food) enticed wild magpies into the cage. In some remote sites, magpies did not respond to a decoy alone and in some instances, playing a tape recording of several wild magpies assisted in luring magpies into the cage. A funnel shaped entrance into the cage allowed wild magpies to enter but impeded escape. Trapped wild birds were removed

23

General Methods from traps and samples were collected. Each wild bird was examined visually for the presence of Philopterus sp. and B. semiannulata by picking through the feathers individually for 5-10 minutes. All lice were removed from the feathers of the host with feather-light forceps and stored without food for 24 hours in tubes after which time they were preserved in 95% ethanol. A sample of blood was collected from the toenail of each magpie caught and the blood sample stored in 1 mL of lysis buffer (0.1 M Tris-Hcl pH 8.0, 0.1 M EDTA, 0.5% SDS, 0.01 M NaCl). Blood samples were returned to the lab and stored at –80 °C for later genetic analysis.

2.3 MOLECULAR TECHNIQUES

2.3.1 G. tibicen 2.3.1.1 DNA extraction techniques Two methods of DNA extraction were employed in this study. For analysis of mtDNA variation, total genomic DNA was extracted using standard phenol-chloroform procedures (Doyle and Doyle 1987), following the digestion of 1-2 L of frozen blood in extraction buffer (0.1 M Tris-HCl, 1.4 M NaCl, 0.02 M EDTA, 10g CTAB: hexadecyltrimethylammonium bromide) overnight at 55 ˚C. To remove proteins and lipids, 600 L volumes of chloroform:isoamyl (24:1), was added to each sample and tubes were rotated gently for 10 minutes and centrifuged at 13 500 rpm for 3 minutes. Supernatent was removed to clean 1.5 mL eppendorf tubes. The process was continued with sequential steps of the addition of 600 L phenol:chloroform:isoamyl (25:24:1). The final step involved the addition of 600 L of chloroform:isoamyl (24:1). Template DNA was precipitated with the addition of 700 L of ice cold isopropanol to each tube and the tube left to sit for one hour at -20 ˚C. Tubes were centrifuged for 30 minutes and isopropanol was removed by aspiration with a modified glass pipette. DNA was cleaned with two washes with 70% ethanol and dried in a vacuum. The pellet was re-suspended in 50 L of TE buffer (0.01 M tris-HCl, 1 mM EDTA pH 8.0). 3 L of sample template was electropherised through 1% agarose gels in 1X TAE buffer (0.04 M tris-acetate, 1 mM EDTA pH 8.0). Gels were stained with ethidium bromide and visualised under UV

24 General Methods to quantify DNA template.

It was noted during microsatellite screening that blood samples amplified from chelex- extracted genomic DNA provided a consistently high yield of product. This allowed for greater accuracy in scoring of microsatellite alleles compared with samples extracted using the phenol-chloroform method. Therefore for microsatellite analysis, total genomic DNA was isolated from whole blood following the chelex protocol modified from Singer-Sam et al. (1989). 0.5 L of whole blood was added to 100 L of 5 % chelex solution (Bio-Rad) in 96 well microtiter plates and vortexed before incubation at 55 ˚C for 1 hour. Following incubation, plates were heated to 95 ˚C for 30 minutes and stored at 4 ˚C before use.

2.3.1.2 mtDNA amplification conditions and screening

A 590 base pair fragment of the mtDNA control region, incorporating portions of Domain I and II, was amplified with the magpie specific primers (primer 1: 5’-GGA AAC AGA GGC GCA AAA GAG C-3’, primer 2: 5’ CAA GAT CTG TGG CTT GAA AAG CC-3’) (Baker et al. 2000). Baker et al. (2000) conducted tests for neutrality (Tajima’s D) and found that variation in magpie populations for this fragment did not differ to neutral expectations. Primer 1 was located in Domain II of the control region and Primer 2 was located adjacent to the control region in the tRNA Glu gene. Samples were amplified in 50 L reactions containing 5 L 10X Taq polymerase buffer, 2mM magnesium chloride, 0.2mM dNTPs (Bioline), 0.4 M of each primer, 0.25 U Biotaq Taq polymerase (Bioline). The following program was used to amplify the target fragment, after an initial denaturation of 5 minutes, 40 cycles were completed of: 95 ˚C for 30 seconds, 55 ˚C for 30 seconds, 72 ˚C for 30 seconds followed by a final extension step of 72 ˚C for 5 minutes.

Following Lessa and Applebaum (1993), Temperature Gradiant Gel Electrophoresis (TGGE) combined with heteroduplex analysis was used as a cost efficient method for screening mtDNA variation across a large number of samples. TGGE distinguishes between sequences with nucleotide differences using the fact that the melting curve of

25

General Methods double stranded DNA depends on the DNA sequence. As fragments begin to come apart as they move through the temperature gradient, their electrophoretic mobility will be retarded. Thus, double stranded fragments with different sequence composition can be identified by their position on a gel prior to complete separation. Samples may be heteroduplexed to a known outgroup to maximise differences between electrophoretic bands. The methods described here are adapted from Baker et al. (2000).

A temperature gradient of 30 ˚C to 50 ˚C was tested initially with five trial times separated by ten minutes and using Baker’s (1999) optimum time of 3 hours 55 minutes as the midpoint. The shortest time of 3 hours and 35 minutes produced the clearest banding patterns and this time was further reduced to 3 hours 5 minutes for all runs following optimisation trials. Polyacrylamide gels (7.7 M urea, 4.8% 29:1 acrylamide: bisacrylamide solution, 1.9% glycerol, 1X ME buffer) were cast directly onto GelBond PAG film (Cambrex) and the gel left to set between glass plates overnight. Samples were heteroduplexed to a single reference product prior to loading of the parallel TGGE. TGGE banding patterns were clear and easy to interpret when the target samples were heteroduplexed with a genetically divergent reference product. To ensure a divergent sample was chosen as the reference, all samples were run twice on parallel TGGE, each time heteroduplexed to a different allele identified as belonging to a different clade. Parallel TGGE gels were electrophoresed at 300 volts for 3 hours and 5 minutes and visualised using silver staining. The gel was washed twice with a buffer of 10% ethanol and 0.5% acetic acid for 3 minutes. The gel was then stained with 1% silver solution

(AgNO 3) for 10 minutes while agitating gently. All silver solution was washed off with double distilled H 2O. The gel was incubated for approximately 15 minutes (until the stain appeared) in a sodium hydroxide buffer (1.5% NaOH, 0.01% NaBH 4, 0.015% formaldehyde). The stained gel was fixed for scoring in a 0.75% Na 2CO 3 solution.

Ideally, species and gene specific sequence mutation rates are calculated directly from fossil or geological data. Due to difficulties in obtaining fossil or geological data for many species, sequence mutation rates are usually estimated from closely related

26 General Methods species where the rate has been established previously. Sequence divergence rates for control region in birds can differ among species and among domain I, II and III within a species (Baker and Marshall 1997, Ruokonen and Kvist 2002) and therefore rates may not necessarily be transferable among species. To estimate a sequence mutation rate for coalescent analyses using the control region it was necessary to compare variation and divergence for control region with a protein coding gene where the rate is more likely to be transferable between closely related species. One eastern population (Seymour) and one western population (Perth) of magpies were sequenced for cytochrome b variation using primers L15191 (5'-ATC TGC ATC TAC CTA CAC ATC GG) and H15916 (5'- ATG AAG GGA TGT TCT ACT GGT TG) (Lanyon and Hall 1994). Amplification of the cytochrome b fragment was carried out in 12.5 L reactions containing 1.25 L of 10x Taq polymerase buffer, 2 mM magnesium chloride, 0.2 mM of dNTPs (Bioline), 0.4 mM of each primer, 0.25 U of Biotaq Taq polymerase (Bioline) and 0.5 L of DNA template. Amplification was carried out in a Perkin Elmer GeneAmp thermocycler. The following program was used to amplify the target fragment, after an initial denaturation of 5 minutes, 40 cycles were completed of: 95 ˚C for 30 seconds, 52 ˚C for 30 seconds, 72 ˚C for 30 seconds followed by a final extension step of 72 ˚C for 5 minutes.

2.3.1.3 Sequencing

Two individuals of each unique control region haplotype identified with TGGE/ heteroduplex analysis were sequenced in both forward and reverse directions using an ABI377 automated sequencer (Griffith University Sequencing facility). All samples amplified for cytochrome b were sequenced in forward and reverse directions. PCR product was cleaned prior to sequencing with exoSAP (Promega) following Werle et al. (1994). 5 L of amplified product was gently mixed with a solution containing 0.5 L exonuclease I E. coli and 2 L of shrimp alkaline phosphate (SAP) and incubated at 37 oC for 35 mins, then heated to 80 o C for 20 mins to deactivate the enzyme. Sequencing reactions were carried out in 10 L reactions containing 30 ng of clean DNA template (visualised on 1.6% agarose gel alongside a known marker), 2 L big dye termination mix (ABI), 2 L big dye sequencing buffer and 3.2 pmol of primer.

27

General Methods

2.3.1.4 Microsatellite amplification and screening Microsatellite variation had been described previously for the Rowsley (Rw) magpie population (Durrant 2005). To ensure standardised scoring between Rowsley and all other populations analysed in this study, several samples from Rowsley were rerun against a size ladder of pooled alleles that spanned the allelic range of each microsatellite locus. All samples from the additional 26 populations were analysed for the following six microsatellite loci: Gt43a, Gt112, Gt67c, Gt115a, Gt206b, Gt201a (Table 2.1).

Table 2.1: Primer sequence for six microsatellite loci isolated from the Australian magpie (Hughes et al. 2003). Forward primers were labelled 5’ HEX. Locus Repeat Sequence Primer Sequence AT ( oC)

Gt43a (AAG) 18 F 5' GCT ACC CGT AAA TAA ACA AAC C 55

R 5' GAG ATG GCA GTG TAC AAT AAC

Gt112 (CA) 19 F 5' GAT GCC TGC ATC AGC CAC AAG 52

R 5' AAT CTT TTT GCC CTC CTG ATC

Gt67c (GCA) 2AGACCCA(GCA) 6 F 5' GTC AAA TGT CTA TTT AAA CAG G 52

R 5' CAC AGG AAT ATC TTG TTA CTT C

Gt115a (CA) 2AA(CA) 2AA(CA) 6AA(CA) 8TC(CA) 2 F 5' GTA GTT CTC ACT ATG GAT AAC 48

R 5' CTG CAA TGT TAT CAG TTT GCT

Gt206b (TC)C(TC) 19 CACATTCT(TC) 2C(TC) 2 F 5' CA AGC TCA GCC TAC AAG ATT C 52

R 5' ATC ATT CAG TGC TCGCCG TGG

Gt201a (AG) 4GGGA(AG) 4(GG)(AG) 5TCA(AAG) 3 F 5' CTG AAA TCT CAA GCA TCT TCC 52 R 5' TG TCC TGA TAC CTC TAG CCA A

For each primer set, amplification of nDNA was carried out in 12.5 L reactions containing 1.25 L of 10x Taq polymerase buffer, 2 mM magnesium chloride, 0.2 mM of dNTPs (Bioline), 0.4 mM of each primer, 0.25 U of Biotaq Taq polymerase (Bioline) and 1 L of DNA template in chelex solution (0.5 L of CTAB extracted template). Amplification was carried out on a Perkin Elmer GeneAmp thermocycler. Cycling conditions for each locus were as follows: initial denaturation for 5 minutes at 94˚C, followed by 25 cycles of 94˚C for 30 seconds, annealing temperature (AT) for 30

28 General Methods seconds, 72 ˚C for 30 seconds and a final extension at 72˚C for 5 minutes. See table 2.2 for annealing temperatures for each locus.

Microsatellite product was denatured in the presence of formamide and cooled immediately on ice before being run on 80 L poly-acrylamide denaturing gels (7 M urea, 5% 19:1 acrylamide: bisacrylamide solution, 0.6X TBE buffer) on a Gelscan 2000 DNA analyser (Corbett Research). 10 mL of gel mixture was set with the addition of 6 L of TEMED and 60 L of 10% ammonium persulphate and left for one hour to ensure complete polymerisation. Products were sized across gels with a commercial standard (Tamra 350, Applied Biosystems) in four lanes and a control ladder made from PCR product, pooled from a number of individuals, in an additional four lanes. Individuals for the pooled product were chosen from initial screening and had alleles that spanned the complete range for that locus across all populations.

2.3.2 Philopterus sp. and B. semiannulata amplification

2.4.2.1 Extraction techniques Prior to extraction, lice were soaked in hydration buffer (0.02 M tris-HCl, 0.05 M EDTA pH 8.0, 0.05 M NaCl, 0.2% SDS) for 72 hours. Lice were extracted with the DNeasy tissue kit (Qiagen) following the specified protocol for insects. Template was eluted with 50 L of double distilled water. Quantification was not possible due to the low copy DNA template. Samples that failed to amplify were precipitated with 70% ethanol and 0.3M sodium acetate solution, centrifuged for 30 minutes and alcohol removed by aspiration with a modified glass pipette. The pellet was re-suspended in 20 L double distilled water.

2.3.2.2 Amplification of mtDNA A fragment of the protein coding mtDNA gene cytochrome oxidase subunit I (COI) was chosen for mtDNA sequence analysis of the feather lice. Although use of identical gene fragments for both G. tibicen and lice would have been preferred, the non coding region of the mtDNA in lice is AT rich and greatly reduced in length (Covacin et al. 2006). In

29

General Methods contrast, COI has been shown to be a highly variable marker that is useful for intraspecific studies of invertebrate species (see Andersen et al. 2000, Heilveil and Berlocher 2006).

Initially five individuals of each louse species were sequenced taken from two magpie individuals to detect variation present on a single host. No sequence variation was found in the target gene region among the samples from a single host in either Philopterus sp. or B. semiannulata . As a consequence, for all further analyses, only a single louse was sequenced from each magpie host individual. mtDNA was amplified in 50 L reactions containing 5 L 10X Taq polymerase buffer, 2 mM magnesium chloride, 0.2 mM dNTPs (Bioline), 0.4 M of each primer, 0.25 U Biotaq Taq polymerase (Bioline).The following program was used to amplify the target fragment, after an initial denaturation of 5 minutes, 40 cycles were completed of: 95 ˚C for 30 seconds, 50 ˚C for 30 seconds, 72 ˚C for 30 seconds followed by a final extension step of 72 ˚C for 5 minutes.

2.3.2.3 Sequencing The COI fragment for louse samples was sequenced in both forward and reverse directions using an ABI377 automated sequencer (Griffith University Sequencing facility). PCR product was electrophoresised through 1.6 % agarose gels and the resulting band was cut out and the DNA purified with Qiaquick PCR purification (Qiagen) prior to sequencing. Sequencing reactions were carried out in 10 L reactions containing 30 ng of clean DNA template (visualised on 1.6% agarose gel alongside a known marker), 2 L big dye termination mix (ABI), 2 L big dye sequencing buffer and 3.2 pmol of primer.

30 General Methods

2.5 DATA ANALYSIS

2.5.1 MtDNA 2.5.1.1 Phylogenetics Sequences were aligned with Sequencher 4.1 (Gene Codes Corporation). An appropriate substitution model was selected using MODELTEST (Posada and Crandall 1998). Maximum likelihood phylograms were constructed with PAUP version 4.0b 10 (Swofford 2000).

For the magpie analysis, a pied butcherbird Cracticus nigrogularis sequence was selected as the outgroup. While collecting magpie samples I had collected Philopterus sp . from C. nigrogularis and this was used as the outgroup for louse analyses. Unfortunately no outgroups were available for B. semiannulata , therefore trees were rooted with the most divergent clade . Mean divergence among haplotypes within clades and net mean divergence between clades (Nei and Li 1979) were calculated from model-corrected distances generated in PAUP version 4.0b 10 (Swofford 2000).

2.5.1.2 Population statistics Estimates of haplotype diversity, nucleotide diversity and average pair wise differences within populations were calculated in DNASP version 4.0 (Rozas and Rozas 1999).

Pairwise FST estimates were calculated for all population pairs ( ARLEQUIN 3.01, Excoffier et al. 2005). Bonferroni correction for multiple tests was applied to significance levels for pairwise FST comparisons. Relationships among haplotypes were identified by construction of a parsimony network with TCS 1.13 (Clement et al. 2000) using a 0.95 limit to parsimony. Limits to parsimony were relaxed for lice due to the fact that very deep divergences were found between some unique haplotypes.

2.5.2 Microsatellites 2.5.2.1 Tests for Hardy-Weinberg proportions Exact tests were performed to test for significant deviations from Hardy-Weinberg proportions using GENEPOP (Raymond and Rousset 1995) for each locus at each site. Significant deviations may arise from either amplification error (locus specific error),

31

General Methods sampling error or may be related to the dynamics of particular sample populations. Amplification error is due to null or poorly amplified alleles when primers mismatch (null alleles) or from preferential amplification of smaller alleles (allelic dropout). Amplification error tends to result in significant deviations at specific loci within a species, whereas significant deviations at all loci for individual populations suggests either sampling error or a biological population effect. Population deviations may be the result of pooling divergent populations (eg. Wahlund effect) or may indicate a recent bottleneck or a founder event. MICRO -CHECKER (Van Oosterhout et al. 2004) was used to distinguish between potential laboratory produced deviations from Hardy-Weinberg proportions.

2.5.2.2 Microsatellite variation For each population, the mean number of alleles per locus and allelic richness was calculated in FSTAT 2.9.3 (Goudet 1995). Allelic richness takes into account the amount of variation within a population and the sample size of populations to allow for population comparisons. Pairwise FST estimates were calculated for all population pairs

(GENEPOP , Raymond and Rousset 1995). Bonferroni correction for multiple tests was applied to significance levels for pairwise FST comparisons. A multidimensional scaling plot ( MDS ) was constructed in PRIMER version 5 (Primer-E) using Slatkin’s linearised

FST to view the relationship among all population pairs.

32 Population analysis

CHAPTER 3: POPULATION ANALYSIS OF GYMNORHINA TIBICEN SAMPLES

3.1 INTRODUCTION Reconstructing the recent evolutionary history of any organism requires an understanding of microevolutionary (population level) processes. Modern patterns of structure and gene flow are not only a reflection of present ecological and intraspecific factors but are moulded by historical events such as demographic change (recent range expansion) and vicariance. Using genetic variation to estimate levels of population gene flow, some avian studies have reported levels of dispersal higher than expected from life-history traits or social structure alone (see Edwards 1993, Van Treuren et al. 1999, Wright and Wilkinson 2001). These studies highlight the importance of including population analyses and estimates of gene flow in explaining the recent history of a species.

Regional variation in a number of social and life-history traits has been reported for the Australian magpie. Australian magpies are group living. Size of group territories vary however, across the continent with the largest groups found in the south west (3-26) and Tasmania (3-15), and the smallest groups recorded in north eastern Australia (2-4) (Hughes and Mather 1991). Dispersal, inferred indirectly from allozyme and mtDNA data, also varies among regions, with differences attributed to higher levels of natal philopatry in south western populations when compared with eastern BB and WB populations (Baker et al. 2000). A study on south eastern magpie populations using mtDNA and allozyme variation (Baker et al. 2001) estimated high gene flow at geographical scales occasionally exceeding 1000km among populations. The study indicated that long distance dispersal may occur, even though such events may be limited by social structure in some populations. Previous research has also indicated that female magpies tend to remain in their natal territory while male magpies are the primary dispersers (Veltman and Carrick 1990). If this is the case, then estimates based on mtDNA and nuclear DNA may show differentiation due to the different modes of

33

Population analysis inheritance (female versus bi-parental). Baker et al. (2000) however, estimated approximately similar levels of gene flow from both mtDNA and allozyme markers. Allozyme markers tend to display relatively low levels of variation in some birds (Zink 1997) and therefore, may not be suitable for estimates of gene flow. This appeared to be the case for magpies with the most common allele at each allozyme locus greater than 80% in all populations for five of the six loci examined in Baker’s study.

Although potential for long distance dispersal was indicated among south-eastern populations of magpies (Baker et al. 2001), no support was found for effective dispersal between south-eastern and south-western Australian populations (Baker et al. 2000). In a following study, it was found that no gene flow occurs across the Nullarbor Plain (Toon et al. 2003) even though the modern range of magpies extends large distances along the southern coast of Australia. The modern distribution in northern Australia is also continuous, although densities are very low in the north-west region of the continent (Schodde and Mason 1999). This raises the question as to whether the distribution of magpies across northern Australia reflects two or more recently connected but genetically divergent populations. Alternatively, do the two divergent populations that Baker et al. (2000) found in the south-east and south-west represent two ends of a naturally contiguous distribution (a ‘ring’ species)?

Here, I compare patterns of genetic structure among regions for the Australian magpie based on both microsatellite loci and mtDNA. Specifically mtDNA population structure was compared for northern and western populations with southern populations that had been analysed previously (Baker et al. 2000, Hughes et al. 2001, Toon et al. 2003). Six microsatellite loci were assayed for all sample populations and patterns of microsatellite variation were compared among magpie populations. Maximum likelihood coalescent analyses were used to compare estimates of gene flow among populations. Specifically, patterns of microsatellite variation were expected to be more informative about population structure than patterns based on previous allozyme data. It was also expected

34 Population analysis that structure for microsatellite loci along the east coast was likely to be limited as high levels of gene flow have been estimated using mtDNA and allozyme data previously.

3.2 METHODS Molecular techniques used here are outlined in Chapter 2. General population analyses included tests for deviations from Hardy – Weinberg proportions, pairwise FST estimates and diversity statistics following procedures outlined in 2.5. Specific methods for population analyses are outlined below.

3.2.1 Population Structure Two methods were employed here to identify population structure. Firstly, hierarchical analysis was used to test a priori hypotheses based on geographical information. This method uses traditional F-statistics to test for significant structure among populations. A weakness of this method is that it is unable to tease apart structure when it is present at many levels of the hierarchy and among many populations. So, in addition, a recently developed model based approach that complements traditional F-statistics was also used. These methods incorporate the use of additional information such as the presence of Hardy Weinberg proportions in populations and linkage equilibrium among loci to identify non-differentiated populations (Pearse and Crandall 2004). The use of multiple methods when detecting population structure is complementary and adds strength to any conclusions inferred from the results.

3.2.1.1 Hierarchical analysis Potential for significant geographical structure was examined with a hierarchical analysis of molecular variance (AMOVA, Excoffier et al. 1992) separately for mtDNA and microsatellite data. The AMOVA was calculated with different geographical groupings of populations, that could represent possible historical refuge areas, in an attempt to maximise between group differentiation ( FCT ). Geographical position of refuge areas and associated arid barriers were based on contemporary habitat and rainfall patterns (see Keast 1961). The potential barriers considered in the analysis were

35

Population analysis the Murchison in the west, Canning in the north-west, the Carpentarian in the north and the Nullarbor and the Eyrean/Mallee in the south (see Figure 1.3). Preliminary analysis indicated that there was no evidence for historical structure throughout much of the eastern region coast populations and as such several east coast highland barriers such as the Broad Sound barrier and Hunter Valley were not considered further in this analysis. Tasmania was removed from the analysis because it represented a single population from a potential refuge and therefore it was not possible to estimate among populations within groups ( FSC ). AMOVA was used to examine the proportion of variation within all populations, among populations within groups and among groups. F-statistics were calculated for each level of the analysis using ARLEQUIN 3.01 (Excoffier et al. 2005). For mtDNA data, Φ-statistics using pairwise distance estimates were also calculated to incorporate the use of sequence divergence information.

3.2.1.2 Bayesian analysis A Bayesian clustering approach was used to identify population structure across

Australia implemented with Bayesian Analysis of Population Structure software; BAPS

VERSION 4.0 (Corander et al. 2003, 2004). This model based approach uses prior information of geographical sampling locations and population allele frequencies to group non-differentiated populations under assumptions of HWE and linkage equilibrium among loci. BAPS treats allele frequencies and number of populations as the random variables to calculate relative allele frequencies and posterior distribution of population structure.

Initially clusters were assigned at the population level. Priors for the maximum number of populations, k = 2, 3, 4, 5, 6, 7, 8 were entered following initial runs encompassing wider intervals. Posterior probabilities for alternative clustering of populations were calculated using the given output of log marginal likelihood values for each possible clustering partition. Posterior probabilities are generally not very informative about the strength of the result because maximum likelihood parameter space is very large and therefore many solutions are possible (Corander et al. 2006). Therefore Bayes factors

36 Population analysis are presented following Corander et al. (2006) as a measure of strength for the most likely compared with the second most likely, clustering partition. Bayes factors can be considered as a Bayesian analogue to likelihood ratio tests and can give a measure of the relative support for alternative models given the prior and posterior distributions. Bayes factors were calculated as follows.

P (M 1 | D ) / p (M 1 ) = B (D) = Bayes Factor P(M | D) /p (M ) 2 2

Where: P(M 1 / D) = the posterior distribution of model one (M 1) given the data ( χ)

p(M 1) = the prior distribution of model one

P(M 2 / D) = the posterior distribution of model two (M 2) given the data ( χ)

p(M 2) = the prior distribution of model two

Following initial clustering of populations, admixture analysis was performed in BAPS 4.0. Admixture analysis aims to identify the potential number of ancestral source populations present in the contemporary populations that were identified from clustering analysis (Corander and Marttinen 2006). Running conditions were as follows: 100 iterations to estimate admixture for the individuals, 200 reference individuals and 10 iterations for the reference individuals.

3.2.2 Isolation by Distance

Geographic distance was plotted against Slatkin’s linearised FST to test for a signature of isolation by distance. A Mantel’s test (Mantel 1967) was performed to test the significance of the correlation. Separate Mantel’s tests were performed on sites that were identified in earlier analyses to belong to the eastern and western regions, respectively. Mantel’s tests were carried out in ARLEQUIN 3.01 (Excoffier et al. 2005) for microsatellite loci and mtDNA control region with 10, 000 permutations, independently.

37

Population analysis

3.2.3 Estimating Levels of Gene Flow

Using the LAMARC software package (Kuhner et al. 2005) theta ( Θ) and migration estimates were calculated simultaneously. LAMARC uses a maximum likelihood method to estimate parameters based on coalescent simulations (Beerli and Felsenstein 1999, Beerli and Felsenstein 2001). Migration estimates were calculated between adjacent clusters of populations that were identified with BAPS . This allowed the exclusion of estimates among populations that were likely to exhibit high gene flow, thus reducing the number of parameter estimates and hence increasing efficiency of running the analysis. For example, most populations within the eastern region were clustered into a single group and therefore no estimates of gene flow were made between populations within this group. An exception was made for the Kimberley and Pilbara populations because although they clustered with eastern populations for the microsatellite data they were divergent from the eastern populations for the mtDNA analysis. Migration estimates were also calculated between Tasmania and the mainland. Due to the high computational power required for large datasets, I randomly sampled 20 individuals from a randomly selected site within each cluster of populations. Two populations, Charters Towers (Ct) and Ouyen (On), were chosen from the eastern cluster to allow comparison of migration rates between western and eastern Australia via northern and southern routes. Mitochondrial and microsatellite data were analysed separately in

LAMARC . Migration results for each analysis are reported as the average of three runs. For the mitochondrial analysis, the default model F84 was used which allows unequal nucleotide frequencies and unequal rates of transitions and transversions (Kishino and Hasegawa 1989). For the microsatellite analysis, a mixed stepwise model was selected as it was indicated in preliminary runs, that the more basic models (eg. Brownian model) were not suitable for all loci. Watterson’s estimate of theta was used for all starting points. From preliminary runs, a search strategy was formulated consisting of 20 initial chains of 1000 samples and 2 final chains of 20, 000 samples for the final analysis. One advantage of using LAMARC to estimate migration rates over other methods (eg. FST ) is that it allows for unequal population sizes ( Ne) and asymmetrical migration between two populations (Beerli and Felsenstein 2001).

38 Population analysis

3.3 RESULTS 3.3.1 MtDNA Sequence Variation and Patterns of Genetic Diversity 16 new haplotypes were identified following analyses of northern magpie populations, making a total of 46 haplotypes across Australia analysed here (Figure 3.1). 28 of 522 nucleotide sites were variable. Haplotype diversity within populations ranged from 0.428 (± 0.068) to 0.827 (± 0.019) (Table 3.1).

Two clades were identified in the parsimony network (Fig. 3.2), which correspond to an eastern region (As, Ct, Bw, Mb, Rk, Ma, To, Br, Gr, Db, Or, Ck, Gb, Ou, Hm, Sr, Pi, Rw, Sa, Cd, Ts) and a western region (Ep, Al, Bn, Mh, Pe, Pb, Km). There was no evidence of more than a single clade being present at any single sample site (Figure 3.1). Tasmania (Ts) was represented by only three haplotypes, one of which was endemic to the Tasmanian population. Ceduna (Cd), on the eastern side of the Nullarbor Plain, was represented by two haplotypes that occurred at no other site but that were more closely related to haplotypes found in the eastern region. Four haplotypes (2, 3, 4, 6) accounted for 82.9 % of all individuals in the east and two haplotypes (17, 19) represented 82.5 % of all individuals in the west.

Significant FST results were observed for 336 of 378 pairwise comparisons (Appendix

III). Pairwise FST comparisons ranged from 0 indicating panmixia to 0.94 where no sharing of haplotypes was evident and populations had limited haplotype diversity.

39

Population analysis

Table 3.1: Genetic diversity in Australian magpie populations derived from control region mtDNA variation and multilocus microsatellite DNA genotypes. mtDNA diversity microsatellite diversity -3 Site Region n Hap H π(x10 ) n A RS HE HO Fg East 4 2 0.50 ± 0.27 3.82 ± 2.03 - - - - - As East 11 2 0.55 ± 0.07 1.04 ± 0.14 11 9.7 5.09 0.80 0.71 Cc East 4 1 0 ± 0 0 ± 0 - - - - - Ct East 29 6 0.54 ± 0.10 1.44 ± 0.35 30 10.8 4.75 0.81 0.74 Bw East 31 3 0.43 ± 0.10 1.29 ± 0.30 31 10.0 4.58 0.76 0.76 Mb East 21 6 0.76 ± 0.06 2.95 ±0.44 21 10.0 5.19 0.84 0.84 Rk East 29 5 0.71 ± 0.05 2.36 ± 0.44 28 10.2 4.85 0.82 0.74 Ma East 54 5 0.73 ± 0.04 2.54 ± 0.17 48 10.5 4.18 0.72 0.70 To East 55 6 0.74 ± 0.03 2.19 ± 0.24 48 11.7 4.59 0.77 0.75 Br East 50 5 0.66 ± 0.05 2.53 ± 0.44 47 11.8 4.44 0.73 0.71 Gr East 51 6 0.74 ± 0.04 4.79 ± 0.31 48 10.5 4.37 0.76 0.77 Db East 51 6 0.71 ± 0.04 3.04 ± 0.30 48 12.2 4.76 0.80 0.80 Or East 52 7 0.80 ± 0.04 4.64 ± 0.34 48 14.0 4.94 0.80 0.82 Ck East 59 4 0.67 ± 0.04 1.80 ± 0.17 48 11.5 4.53 0.75 0.72 Gb East 50 6 0.83 ± 0.02 3.65 ± 0.33 48 12.7 4.67 0.79 0.77 Ou East 51 7 0.74 ± 0.04 2.55 ± 0.31 48 13.7 4.79 0.77 0.75 Hm East 50 6 0.71 ± 0.05 2.44 ± 0.40 45 12.3 4.53 0.76 0.73 Sr East 50 5 0.75 ± 0.03 2.69 ± 0.29 48 13.2 4.83 0.79 0.79 Pi East 50 5 0.65 ± 0.05 1.71 ±0.23 48 9.7 4.18 0.73 0.69 Rw East 51 11 0.81 ± 0.03 3.40 ± 0.36 54 11.8 4.43 0.76 0.73 Ts East 56 3 0.43 ± 0.07 0.87 ± 0.16 48 7.3 3.51 0.57 0.63 Sa East 4 3 0.83 ± 0.22 2.23 ± 0.70 - - - - - Ki East 7 3 0.71 ± 0.13 1.64 ± 0.42 - - - - - Cd East 8 2 0.57 ± 0.09 4.37 ± 0.72 8 6.7 4.36 0.73 0.62 Ep West 16 3 0.57 ± 0.11 1.66 ± 0.49 16 8.2 4.41 0.76 0.68 Al West 50 3 0.55 ± 0.06 1.26 ± 0.19 48 10.3 4.58 0.76 0.71 Bn West 50 4 0.48 ± 0.06 1.03 ± 0.15 48 8.2 3.90 0.72 0.64 Mh West 50 3 0.52 ± 0.06 1.09 ± 0.17 48 8.5 4.14 0.72 0.73 Pe West 48 4 0.70 ± 0.04 1.87 ± 0.19 48 9.8 4.04 0.70 0.67 Pb West 25 6 0.58± 0.11 1.72 ± 1.46 23 12 5.22 0.82 0.74 Bm West 3 1 0 ± 0 0 ± 0 - - - - - Km West 17 3 0.32 ± 0.14 0.87 ± 0.90 17 7.7 4.04 0.76 0.75 Mt West 4 1 0 ± 0 0 ± 0 - - - - - Total 1141 46 1051 21.5 n: Number of individuals typed for each locality. Hap: Number of haplotypes present. H: Haplotype diversity. π: Nucleotide diversity (x10 -3). A: Mean number of alleles/locus. RS : Allelic richness. HE: Expected heterozygosity. HO: Observed heterozygosity. Only sites with eight or more samples were typed for genetic variation at microsatellite loci and included in subsequent population analyses. Pb includes Pb1-4 from Figure 3.1. Km includes Km1-2 from Figure 3.1.

40 Population analysis

are shown. Refer to TableRefer samplesizes.refer3.1forColours shown. to are Frequency3.1: controlAustraliregion of haplotypesmtDNA in Figure found 2P333.Pb4 32.Pb3 30.Pb1 9.Pe29 North 31.Pb2 28.Mh 27.Bn 26.Al 25. Ep 25. 26.Al 27.Bn 5K136.Km2 35.Km1 34.Bm 27 31 30 29 28 32 26 33 25 37.Mt 34 35 637 36 2 4.C d 23.Ki 2 1 24 3 2.Sa22 23 22 to haplotypesin Figureto 3.2. 17 16 20 18 19 21.Ts 5 4 12 21 13 6 15 14 14 7 anormoreAllsamples4 magpies. sites with 9 16.Ou 17.Hm 18.Sr 19.Pi 19.Pi 18.Sr 17.Hm 16.Ou 11 11 8 1.Fg 10 4.Ct 5.Bw 6.Mb 7.Rk 6.Mb 5.Bw 4.Ct 8.Ma 9.To 10.Br 11.Gr 11.Gr 10.Br 9.To 8.Ma 12.Db 13.Or 14.Ck 15.Gb 14.Ck 13.Or 12.Db 2.A s

3 .Cc 20.Rw

41

Population analysis

9 22 20 33 31 44 50

21 18 17 29 23 16 6 15

30 37 14 10 1 49 19 38 4 43 35 34 45 40 32

8 13 46

2 7 41

36 42 26 48

Frequency

11 3 5 27

39 >100 10-50 5-10 1-5 25

Figure 3.2: Parsimony network computed by TCS using 522 base pairs of control region sequence data for 46 haplotypes. Size of haplotype circles indicates relative frequency of haplotypes. Connecting bars indicate one basepair mutation and filled cicles represent additional basepair mutations. White haplotypes were present at sites where only one- two birds were caught.

3.3.2 Microsatellite Variation Twelve tests out of a possible of 162 were significant for deviation from Hardy

Weinberg proportions (Table 3.2). Positive FIS values indicated heterozygote deficiencies for nine significant tests. Significant tests did not appear to be locus or site specific. MICRO -CHECKER 2.2.3 (Van Oosterhout et al. 2004) indicated that a small number of null alleles were likely to have resulted in the heterozygote deficiencies that were detected.

42 Population analysis

Table 3.2: FIS values (Wier and Cockerham) for magpie populations at six microsatellite loci. Dash indicates monomorphic loci at Tasmania (Ts) and Ceduna (Cd).

POP Gt67c Gt43a Gt201a Gt206b Gt115a Gt112a As 0.636** 0.12 0.059 0.037 -0.057 -0.082 Ct -0.121 -0.018 0.023 0.128 0.101 0.194** Bw 0.183 0.088 -0.027 -0.139 0.123 -0.089 Mb -0.002 -0.057 -0.035 0.024 -0.018 0.021 Rk -0.043 0.185 0.033 0.295** -0.047 -0.038 Ma 0.092 0.034 -0.004 0.088 0.015 -0.036 To -0.096 0.061 0.017 0.100 -0.015 -0.014 Br 0.112 0.079 -0.036** 0.006 0.106 -0.042 Gr -0.044* 0.008 -0.128 0.078 -0.033 -0.023 Db 0.043 -0.066 0.058 -0.053 0.051 -0.048 Or -0.211 0.003 -0.061 0.124 -0.065 -0.021 Ck -0.238 0.101 0.191* 0.124** 0.019 -0.073 Gb -0.032 0.003 0.025 -0.021 0.095 -0.027 Ou 0.136 0.091 -0.051 -0.151 0.076 0.055 Hm -0.079 -0.019** 0.046 0.024 0.161 0.007 Sr -0.009 0.025 -0.020 -0.004 -0.030 -0.020 Pi -0.023 0.034 0.020 0.001 0.140 0.064 Rw -0.087 0.067 0.075 -0.058 0.178 0.020 Ts - 0.052 0.194 0.106 0.091 0.010 Cd - 0.143 0.176 -0.091 0.243 0.067 Ep -0.191 -0.074 -0.15 0.346 0.098 0.224 Al -0.062 0.118* 0.137 0.107 0.048 -0.076 Bn 0.114 0.000 0.417** -0.114 0.255** 0.013 Mh -0.062 -0.023 -0.216 -0.031 0.105 0.049 Pe -0.101 -0.030 -0.015 0.040 0.212** 0.050 Pb -0.052 0.113 0.188 0.199 0.039 -0.033 Km 0.263 0.008 -0.171 0.064 -0.086 -0.093 -0.016 0.035 0.026 0.038 0.067 -0.003 global (0.025) (0.011) (0.024) (0.020) (0.018) (0.011) * Indicates significance P < 0.05 ** Indicates significance P < 0.01

No single locus appeared to be responsible for null alleles and in general, they were spread across all loci indicating that they were most likely due to random PCR failures. A small random loss of alleles spread across a large dataset is unlikely to have a large effect on diversity measures. If the null alleles are the result of primer mismatches however, diversity measures would be expected to be slightly more conservative, due to loss of some variation within populations. It is unlikely that either reason would have

43

Population analysis affected the principle outcomes of this study as highly variable loci were chosen and as such data for all populations were included in the analyses.

Analysis of the six microsatellite loci revealed relatively high heterozygosity for all populations, that ranged from 0.62 to 0.84 (Table 3.1). No populations displayed exceptionally high or low diversity. Allelic richness for mainland populations ranged from 3.90 to 5.22. One locus (Gt67c) was fixed in the Tasmanian population (Ts) and overall, Tasmania had lower allelic richness (3.51) compared with all mainland populations.

Significant FST values ( P < 0.05) were observed for the majority of pairwise population comparisons (Appendix IV); only 7 of 351 comparisons were non-significant. The relationship among populations based on Slatkin’s linearised FST is presented in Figure 3.3. All populations within the eastern region clustered together, while all populations within the western region were as divergent from each other as they were divergent from all populations within the eastern region. The Kimberley (Km) population in north Western Australia and the central south population Ceduna (Cd) were different from all other populations. The Kimberley population however, appeared to be more closely related to populations in the eastern region than to other western populations. The Pilbara (Pb) population, also from the west, appeared to be more similar genetically, to populations in eastern Australia. This result contrasted with mtDNA results where haplotypes found in the north-west (Kimberley) and west (Pilbara) formed part of the western lineage and haplotypes found in the central south (Ceduna) formed part of the eastern lineage (see Figure 3.1 and 3.2). Tasmania (Ts) was divergent from all other populations.

44 Population analysis

2

1

Central (As) North East (Ct, Bw, Mb, Rk) -2 -1 1 2 Central East (Ma, To, Br, Gr) South Central East (Db, Or, Ck, Gb) - South East (Ou, Hm, Sr, Pi, Rw) 1 Tasmania (Ts) Central South (Cd) - South West (Ep, Al, Bn, Mh, Pe) 2 Central West (Pb)

North West (Km)

Figure 3.3: Multidimensional Scaling Plot (MDS) of microsatellite variation using Slatkin’s linearised FST as a measure of differentiation for 27 populations of magpies. Population symbols are assigned according to geographic sampling locality. Population identifier is in parentheses.

3.3.3 Population Structure 3.3.3.1 Hierarcichial analysis

Alternative hierarchical groupings for the mtDNA AMOVA are given in Table 3.3. FCT was maximised (0.242) and significant ( P < 0.001) when populations were divided into eastern and western groups. The subdivision was aligned with known geographical barriers: the Nullarbor Plain in the south and the Carpentarian divide in the north.

Although fixation indices among populations within regions (FSC : 0.151) were also significant, 24% of the variation was attributed to among region variation with less than 12% of the variation allocated to among sites within regions. A similar pattern was observed for Φ− statistic results, although much higher differentiation was evident between eastern and western regions in this analysis. The Φct was maximised (0.807) and highly significant ( P < 0.001) based on the Nullarbor Plain/Carpentarian divide. With the additional information from sequence divergence, most of the variation (80.65 %) was distributed among regions rather than among populations within regions (3.35%).

45

Population analysis

Table 3.3: AMOVA results for a) mtDNA using haplotype frequency differences b) mtDNA using haplotype frequency differences and sequence divergence and c) six microsatellite loci. Different hierarchical groupings are presented. The groups are as follows east (As, Ct, Bw, Mb, Rk, Ma, To, Br, Gr, Db, Or, Ck, Gb, Ou, Hm, Sr, Pi, Rw, Sa, Cd) and west (Ep, Al, Bn, Mh, Pe, Pb, Km). Groupings represent the following potential geographic barriers to dispersal 1. Nullarbor and Carpentarian 2. Eyrean/Mallee and Carpentarian 3. Nullarbor and Canning 4. Eyrean/Mallee and Canning.

a. Percentage Variation among among pop within within FSC P FCT P group group pops 1.(east) (west) 0.151 <0.001 0.242 <0.001 24.19% 11.48% 64.33% 2.(east) (west+Cd) 0.153 <0.001 0.236 <0.001 23.58% 11.66% 64.75% 3.(east+Km) (west) 0.163 <0.001 0.229 <0.001 22.89% 12.54% 64.58% 4. (east+Km+Pb)(west) 0.173 <0.001 0.219 <0.001 21.86% 13.53% 64.61% 5.(east+Km+Pb)(west+Cd) 0.174 <0.001 0.211 <0.001 21.10% 13.76% 65.14%

b. Percentage Variation among pop among within within ΦSC P ΦCT P group group pops 1.(east) (west) 0.173 <0.001 0.807 <0.001 80.65% 3.35% 16.00% 2.(east) (west+Cd) 0.210 <0.001 0.791 <0.001 79.31% 4.39% 16.51% 3.(east+Km) (west) 0.288 <0.001 0.773 <0.001 77.31% 6.54% 16.14% 4. (east+Km+Pb)(west) 0.388 <0.001 0.734 <0.001 73.38% 10.33% 16.29% 5.(east+Km+Pb)(west+Cd) 0.409 <0.001 0.714 <0.001 71.37% 11.70% 16.93%

c. Percentage Variation among pop among within within FSC P FCT P group group pops 1.(east) (west) 0.032 <0.001 0.015 <0.001 1.53% 3.12% 95.35% 2.(east) (west+Cd) 0.032 <0.001 0.015 <0.001 1.51% 3.12% 95.37% 3.(east+Km) (west) 0.031 <0.001 0.018 <0.001 1.83% 3.04% 95.14% 4. (east+Km+Pb)(west) 0.031 <0.001 0.019 <0.001 1.89% 3.06% 95.04% 5.(east+Km+Pb)(west+Cd) 0.031 <0.001 0.018 <0.001 1.84% 3.06% 95.10%

In contrast to the mtDNA results, the AMOVA based on microsatellite variation showed that most variation was present within populations (95.04 - 95.37 %), with marginally less variation evident among groups (1.51 - 1.89 %) rather than among populations within groups (3.04 - 3.12 %) (Table 3.3). The FSC (0.031 - 0.032) was slightly higher than the FCT (0.015 - 0.019) for all alternative arrangements of populations. For the

46 Population analysis microsatellite AMOVA, the grouping that maximized FCT also contrasted with the mtDNA AMOVA results. In the microsatellite AMOVA, the Kimberley (Km) population from the far north-west and the Pilbara (Pb) population from the west grouped with eastern populations to maximize the FCT (0.019, P < 0.001) rather than with the western populations as was the case for the mtDNA AMOVA.

3.3.3.2 Clustering analysis From the BAPS analysis, six population clusters were identified among the 26 sampled sites (Figure 3.4). The posterior probability for six clusters was one. The strength of the six cluster partition compared to the five cluster partition (second highest log marginal likelihood) resulted in a Bayes factor of 1.45 x 10 8. A value exceeding 10 indicated that the former clustering solution (six clusters) was strongly supported (Corander et al. 2006). With the exception of the Phillip Island (Pi) and Tasmania (Ts) populations, all eastern populations (those assigned to the eastern mtDNA clade) were assigned to cluster 3. The Kimberley (Km) and Pilbara (Pb) populations clustered with eastern populations. As was the case for the MDS plot, no other populations clustered with Tasmania (Ts). Among the five sites sampled from the south west, three clusters were identified. Overall, no more than two clusters were evident from the entire eastern and north western mainland region, while populations from the south western region alone were assigned to three distinct clusters.

As Ct Bw Mb Rk My To Br Gr DbOr Ck GbOn HmSmPi RwTsEpAl Bn Mh Pe Pb Km

Figure 3.4: BAPS mixture clustering. Each bar represents a sampled site and width of bar represents the number of samples. Six groups (clusters) were identified and are coloured accordingly.

47

Population analysis

From the subsequent admixture analysis, there appeared to be mixing between clusters especially between populations in the eastern cluster (dark blue), Phillip Island (red) and Tasmania (aqua) (Figure 3.5). However, the simulation indicated that only very few individuals within clusters were significant for admixture. Between one and four individuals were significant for admixture ( P < 0.05) from Phillip Island, Tasmania, Albany, Busselton, Mandurah and Perth. This suggests that only a very few individuals in the populations potentially resulted from admixture between ancestral populations (clusters from previous analysis). No apparent admixture was evident within the Kimberley population. The Pilbara population appeared to be mixed between the eastern cluster, Tasmanian and the south-west cluster (Perth and Mandurah) however, no individuals were significant for admixture.

As Ct Bw Mb Rk My To Br Gr DbOr Ck GbOn Hm Sm Pi Rw TsEpAl Bn Mh Pe Pb Km

Figure 3.5: Admixture results from BAPS analysis of six microsatellite loci for the Australian magpie. Each vertical bar represents an individual and is coloured by the proportion that the individual shares with each of six pre-assigned clusters. Populations are labelled below each column that are separated by black bars.

3.3.3 Isolation by Distance A significant relationship was evident between geographic distance and Slatkin’s

2 linearised FST for both microsatellite data (r = 0.15, P < 0.010) and mtDNA control region (r 2 = 0.12, P < 0.010) in the eastern region (Figure 3.6). Although a positive trend was evident for the western region, the Mantel test (Figure 3.7) was not significant for control region variation ( P = 0.098) suggesting that mtDNA variation among western populations is not explained by a strict isolation by distance model. The correlation (r 2 = 0.37) and almost significant relationship found for the microsatellite data ( P = 0.055) among western populations (Figure 3.7), indicates geographic

48 Population analysis association among some populations for nuclear DNA. When the Kimberley (Km) population in the north-west was removed from the analysis however, any relationship ceased to exist (r 2 = 0.01, P = 0.615). It appears that the strong differentiation between the Kimberley population and all other western populations at nuclear loci, determines the relationship between genetic variation and geographic distance.

2 1.8 2 1.6 r = 0.12 1.4 P < 0.010 1.2 1 0.8 0.6 0.4 Slatkins'linearised FST 0.2 0 0 500 1000 1500 2000 2500 A Geographic Distance

0.14 0.12 r2 = 0.15 0.1 P < 0.010

0.08

0.06

0.04

Slatkins' linearisedFST 0.02

0 0 500 1000 1500 2000 2500 B Geographic Distance

Figure 3.6: Scatterplot of geographic distance (great circle distance) against Slatkin’s linearised FST between pairs of eastern magpie populations for mtDNA data (A) and microsatellite data (B). Correlation (r 2) and significance for Mantel’s are presented.

49

Population analysis

1 0.9 0.8 r2 = 0.05 0.7 0.6 P = 0.098 0.5 0.4 0.3 0.2 Slatkins' Linearised FST Linearised Slatkins' 0.1 0 0 500 1000 1500 2000 2500 A Geographic Distance

0.12

0.1

0.08

0.06

0.04

0.02 Slatkins' Linearised FST

0 0 500 1000 1500 2000 2500 Geographic Distance B

Figure 3.7: Scatterplot of geographic distance (great circle distance) against Slatkin’s linearised FST between pairs of western magpie populations for mtDNA data (A) and microsatellite data (B). Correlation (r 2) and significance for Mantel’s are presented.

3.3.4 Estimating Levels of Gene Flow

The migration rate calculated in LAMARC is given as M = m/ where m is the per- generation migration rate and is the per-site mutation rate. For comparison with other studies, the results were converted to Nm (immigrants per generation) by multiplying the recipient population by the calculated theta ( Θ) value ( Nm = M * Θ). Nm estimates for microsatellites were generally large and highly variable, ranging between 1 and 200 immigrants per generation (Figure 3.8). Estimates for mtDNA were much lower with a majority of estimates less than 5 immigrants per generation. Substantial gene flow was indicated however, among some sites such as from Perth (Pe) to Pilbara (Pb) ( Nm = 13). Assymetry in migration was estimated between Perth and Pilbara with higher migration away from Perth. Although microsatellite estimates of gene flow were generally higher than mtDNA estimates, no consistent trend was evident. Microsatellite estimates varied

50 Population analysis between one times to an order of magnitude higher than mtDNA estimates for equivalent population pairs.

North

*/90 0/34 3/30 */200

1.8 /27 13 /41 */132 4.3 /18 2.7 /16 1.2 /19 */89

0.5 /42 1/1 0.5 /11 500 km

Figure 3.8: Migration estimates per generation ( Nm) between sites from LAMARC . mtDNA estimates are in red and microsatellite estimates are in black. * indicate estimates of Θ that did not converge.

3.4 DISCUSSION

3.4.1 Diversity, Population Structure and Contemporary Barriers to Gene Flow Compared to many recent studies of passerine birds, a lower number of control region haplotypes and lower haplotype diversity than expected was found for the Australian magpie. In fact, only 46 haplotypes were recovered from over 1000 samples across the entire continent and haplotype diversity ranged from 0.428-0.827. Although no studies are directly comparable to this study due to the low number of samples in other studies, a study on the black-throated blue warbler ( Dendroica caerulescens , Yang et al. 2006)

51

Population analysis found 48 control region haplotypes from only 125 individuals and haplotype diversity of 0.8710. Moreover, in a study of the white wagtail ( Motacilla alba), Pavlova et al. (2005) found 87 haplotypes and haplotype diversities between 0.45-0.95 from only 232 individuals. Lower diversity than expected in the Australian magpie is possibly due to the generalist behaviour and the recent history of this species. The generalist behaviour of magpies will facilitate gene flow among populations and therefore homogenise populations resulting in overall fewer haplotypes.

Population analysis of the Australian magpie indicated contemporary structure at different geographic scales. On a broad scale for mitochondrial DNA, two discrete groups of populations were identified on mainland Australia, corresponding to eastern region and western region populations. The mtDNA AMOVA indicated high levels of population structuring and a maximum ΦCT (0.807) for the subdivision of populations into eastern and western regions. Moreover, the mitochondrial network showed that the western region and the eastern region represent distinct evolutionary lineages. No sampling locations contained more than a single mitochondrial clade suggesting that secondary contact has not occurred between east and west, at least for sites that were sampled in this study. Previous research indicated the Nullarbor Plains region as the modern location of the southern barrier to gene flow (Toon et al. 2003). In this study, the Carpentarian barrier (Figure 1.3) in northern Australia was identified as the modern location for the boundary between north-eastern and north-western magpie mtDNAs.

Perhaps the most interesting result from the population analysis was that results of mtDNA and microsatellite data for the same populations did not concur as to the geographical location of the subdivision between eastern and western populations. In contrast to the mtDNA, the analysis of microsatellite data suggests the contemporary geographical location for the most likely subdivision between eastern and western clades in northern Australia is either the Murchison or the Canning barrier. The microsatellite AMOVA supported a grouping of the Kimberley (Km) and Pilbara (Pb) populations with eastern populations rather than with western populations, as was the

52 Population analysis case for mtDNA. The MDS plot based on Slatkin’s linearised FST for microsatellite variation also showed that the Kimberley population had a greater genetic similarity to eastern populations than to western populations. In the MDS plot, the Pilbara population was intermediate between eastern populations and western populations. The clustering analysis (BAPS) also assigned the Kimberley and Pilbara populations to the eastern cluster rather than to any of the western clusters. However, there appeared to be mixing for a majority of individuals from the Pilbara population between eastern and western clusters. The association of the Kimberley and Pilbara with eastern populations for nuclear loci and not mtDNA is possibly the result of modern or recent past secondary contact between populations in northern Australia. Since European colonisation, habitat favourable for magpies has increased rapidly across much of the continent (Campbell 1929) and in recent times the building of homesteads and dams in northern Australia may have allowed magpies in the north-east to extend recent range expansion across an otherwise previously impervious barrier. A study in which nestling magpies were colour banded and observed over a decade suggests that dispersal in magpies may be male biased although this pattern is unusual in birds (Veltman and Carrick 1990). Male biased dispersal from east to west after contact between northern populations could account for the spread of nuclear alleles preceding the spread of maternally inherited alleles.

In contrast to the large divergence between east and west for the mitochondrial data,

(FCT = 0.242) the microsatellite analysis revealed only weak geographic structure between east and west ( FCT = 0.018). Differentiation among populations (mtDNA F ST

= 0.356, microsatellite FST = 0.047) was also an order of magnitude lower for nuclear genes. Baker et al. (2001) found a similar disparity in the magnitude of differentiation when they compared mtDNA with protein markers (allozyme) for a subset of populations analysed in this study. The allozyme markers used in their study displayed only very limited allelic variation with the most common allele exceeding frequencies of 80% in all populations for five of the six loci examined. Therefore, the difference they reported could have resulted from differences in the inherent levels of variability in

53

Population analysis the two classes of markers. It is unlikely however, that this could account for the difference between microsatellites and mtDNA as the microsatellite loci chosen for this study exhibited relatively high allelic diversity. Rather, the difference reported here may be accounted for by several nonexclusive hypotheses. Firstly, the disparity between nuclear and mitochondrial markers in the amount of differentiation may be due in part to a smaller effective population size ( Ne) for mtDNA due to its maternal mode of inheritance and haploid structure (Birky et al. 1989). The smaller effective population size for mtDNA will lead to a greater effect of genetic drift and hence time to fixation of alleles will be reduced. Therefore, when dispersal is equal for males and females it would be expected that genetic structure (ie. FST ) will be four times higher for females than males. The FST results indicated a greater disparity between mtDNA and microsatellite data than would be expected from differences in Ne alone. Therefore, male biased dispersal in magpies may also account for the greater level of differentiation than expected among populations observed at the maternally inherited mtDNA locus, compared with nuclear loci.

Biological inferences based solely on FST estimates should be treated with caution however as differences may be an artefact of very high diversity at microsatellite loci. Hedrick (1999) warned that multi locus markers that have a large numbers of alleles will tend to produce high within population heterozygosity estimates resulting in very low FST estimates among populations even when very few alleles are shared. However, patterns of genetic variation inferred from FST analysis were concordant with levels of gene flow estimated from coalescent analyses. Estimates of migration and therefore inferred dispersal events varied between selected populations for mtDNA and microsatellite loci. With the exclusion of Tasmania, all migration estimates based on microsatellite loci were at least 4 times higher than those based on mtDNA, although in some cases were up to 10 times higher. Microsatellite loci should result in an estimate approximately four times higher than mtDNA simply because the estimates are based on

Ne. However, as suggested from other analyses (AMOVA) it is likely that male biased dispersal may occur in magpies. Male biased dispersal would account for the larger

54 Population analysis estimates of migration (greater than 4 times) than would otherwise be expected among some population pairs.

On a local scale, the AMOVA indicated significant within group structure ( FSC and

ΦSC ) for both microsatellites and mtDNA data, suggesting that structuring was present within one or both regions. In particular, microsatellite analysis suggested that contemporary structure was greatest in south-west Australia. Three population clusters were identified with BAPS within the south-west compared with a single cluster encompassing most eastern and north western populations. Low to medium levels of migration were estimated within the south-western populations for mtDNA (0-4.3) and microsatellites (16-42) compared with other pairwise estimates. Migration rate was not estimated among populations within the eastern region because eastern sites formed a single BAPS cluster and thus high gene flow was inferred for at least nuclear DNA. Baker et al. (2000) however, inferred low levels of gene flow for south-west compared to eastern populations based on FST estimates for both mtDNA and allozyme data. They found that group size (size of territory) was greatest for populations in the south-west and suggested that relatively low levels of juvenile dispersal due to social differences may account for the greater observed genetic differentiation among south-western populations (Baker et al. 2000). The results of the microsatellite analysis supports this idea that lower levels of juvenile dispersal in the south-west has led to greater population structure compared with the eastern region.

Comparably high gene flow estimates were reported from Perth (Pe) to the Pilbara (Pb) at both mtDNA ( Nm = 13) and microsatellite loci (Nm = 41). Habitat change occurs abruptly between these sites from open eucalypt woodlands in the south-west to the sparse semi arid/arid vegetation of the mid west and Pilbara regions of Western Australia (Specht 1993). Densities of magpies are also much lower in the arid zone compared with the south-west (Wilson 1946, pers. obs. ). Gene flow is likely to be higher from regions of high density around Perth to low density areas in the Pilbara than in the opposite direction or among populations in the south-west (ie. Perth and

55

Population analysis

Busselton). The pattern of gene flow may indicate recent movement of magpies northward from the south-west. This raises the question, have north-western populations only been founded recently?

With the exception of microsatellite N m into Tasmania ( Nm = 11), only low levels of gene flow were inferred between the mainland and Tasmania. This result was expected as it is unlikely that any contemporary dispersal occurs between the two groups of populations. In addition, this result indicates retention of ancestral alleles/haplotypes due to the recency of isolation between Tasmania and the mainland populations. Tasmania is currently isolated from the mainland by Bass Strait. Bass Strait is on average 50 -70 meters deep (Evans and Middleton 1998) and as sea levels dropped during recent cool dry periods of the Pleistocene a land bridge was formed between Tasmania and the mainland. Sea levels rose to above – 50m by 10, 000 years ago (Fleming et al. 1998) isolating Tasmania again from the mainland. In the following Chapter, the timing of divergence between mainland and Tasmanian populations will be estimated and compared to that expected from known sea level changes.

In conclusion, several lines of evidence suggest higher dispersal for male magpies than female magpies. Disparity between FST estimates for mtDNA and microsatellite data, a different location of the contemporary break between eastern and western groups of populations and higher estimates of gene flow from coalescent analysis support the idea that dispersal in magpies is male biased. Population structure inferred here from genetic data will be used in the following Chapter to reconstruct the recent evolutionary history of the Australian magpie.

56 Phylogeography

CHAPTER 4: PHYLOGEOGRAPHY OF THE AUSTRALIAN MAGPIE

4.1 INTRODUCTION In the previous chapter, three genetic regions were identified for magpies across Australia: an eastern region, a western region and Tasmania. Populations in the south- east and south-west are separated on either side of the Nullarbor Plain region (Toon et al. 2003). Across northern Australia, some nuclear gene flow was evident between eastern and western regions identified from mtDNA.

This chapter focuses on using mtDNA and microsatellite variation to understand the recent evolutionary history of the Australian magpie. Here the hypothesis that the Australian magpie population was fragmented in the past by increasing aridity and fluctuating sea levels during the Pleistocene will be examined. Specifically, 1. I will test if eastern, western and Tasmanian groupings are monophyletic which would support vicariance as the process that has fragmented populations. 2. Coalescent analyses will be used to time the divergence between eastern and western regions and between the mainland and Tasmania, to test if these population groupings are of recent origin (Pleistocene). 3. Finally, signatures of recent demographic change (range expansion or population growth) will be explored using a phylogeographic framework. If eastern and western magpie populations contracted into refugia during periods of aridity, we should expect a signature of range expansion and population growth for each region due to changes as climate became more favourable. Phylogeographic patterns for magpies will be compared with patterns documented in other widespread Australian species. The distribution of magpie plumage variants as a possible result of recent history will also be discussed.

57

Phylogeography

4.2 SPECIFIC METHODS General laboratory and descriptive statistical methods are outlined in Chapter 2. Specific phylogeographic methods are outlined below.

4.2.1 Phylogeographic Analysis

Nested clade analysis ( NCA ) was used to infer population history from the geographic distribution of haplotype variation and to distinguish it from contemporary population structure (Templeton 1998). Using the haplotype network constructed with TCS , clades were nested following the rules of Templeton et al. (1987), Templeton and Sing (1993) and Crandall (1996). Clades were nested using an additional rule reported in Hughes et al. (2004) to utilise the phylogenetic relationship among clades. GEODIS 2.2 (Posada et al. 2000) was used to test the null hypothesis that no relationship was evident between haplotypes/clades and geographic distribution. The latest version of the GEODIS inference key (November 11, 2005) was used to infer the biological explanation when a significant relationship between haplotype/clade variation and geographic distribution was identified in the contingency analysis.

4.2.2 Tests for Historical Demographic Change The following tests for demographic change were conducted at two sampling levels. Demographic analysis compares the observed distribution of sequence polymorphisms to what is expected under neutral evolution (Ramos-Onsins and Rozas 2002). Pooling of populations that are genetically divergent will increase the number of ancestral polymorphisms over that which is expected and therefore will give a significantly different result from that expected under neutral evolution (see Fu 1996). Testing multiple sites however, among which high levels of gene flow occur may also give a false sense of replication, as sites are not independent. Therefore, all tests were conducted at the clade level, with all populations pooled for either the western or eastern clades. This was not ideal however, as subtle structuring among populations was indicated in earlier analyses. For this reason, sites were also pooled at a local sampling

58 Phylogeography level to maximise statistical power, where no indication of genetic structuring was indicated earlier.

Different statistical approaches were employed for control region mtDNA and microsatellite loci to test for a signature of population growth. Point estimates designed for non-recombining DNA are presented for mtDNA. No such point estimates were available to test for population expansion for microsatellite loci. Instead a coalescent model based approached was employed. This method was not used for mtDNA, as even though it is shown to be a good estimator of theta ( Θ), it is a poor estimator of growth (g) when only a single locus is used (Kuhner et al. 1998).

To detect demographic change using the microsatellite dataset, a maximum likelihood method was used in LAMARC version 2.0 (Kuhner et al. 2005) to estimate present day theta ( Θ) and growth (g) for each population under the assumption of exponential growth. Detecting significant demographic change requires the reconstruction of the genealogical relationship from a sample, to test whether estimated coalescence times differ from those expected under neutral evolution (constant population size) (Kuhner et al. 1998). For a population that is expanding, it is expected that the genealogy will be long at the root and compressed at the tips relative to what is expected under the model of neutral evolution. To estimate theta and growth, LAMARC samples genealogies from the posterior distribution and compares the likelihood of these based on different trial values for theta and growth. Due to time constraints (45 days for microsatellite analysis), growth estimates and migration estimates (Chapter 3) were run separately using LAMARC . Twenty individuals were sampled randomly from each group of sites and from each clade. The mixed k-allele/stepwise model was chosen as it was indicated in preliminary runs that the more basic models (eg. Brownian model) were not suitable for all loci. Watterson’s theta was used as a starting point. From preliminary runs, a search strategy was formulated of 20 initial chains of 1000 samples and 2 final chains of 20, 000 samples for the final analysis. Historical theta for each population under an

59

Phylogeography exponential growth model was extrapolated from the results to give an indication of how populations have changed in size over time assuming different mutation rates.

LAMARC has been shown to have an upward bias that may go undetected for a single locus (Kuhner et al. 1998). For mtDNA, statistical tests designed specifically for non- recombining loci were employed that have been shown to be powerful for detecting growth (see Chapter 1). The mismatch distribution was constructed from pairwise differences and plotted for each population against the expected distribution under an expanding and constant population size. However, test statistics for mismatch distribution are generally very conservative. For this study, Fu’s FS (Fu 1997) and Tajima’s D (Tajima 1989) were also calculated to test for a signature of rapid demographic expansion. These tests were designed originally to test for neutral sequence evolution in a population. However they have also proven to be effective tests for assessing rapid demographic expansions or declines (Ramos-Onsins and Rozas 2002). Estimates and tests for significance (10, 000 sampling replicates) were calculated using ARLEQUIN 3.01 (Excoffier et al. 2005).

4.2.3 Coalescent Analyses

Time to most recent common ancestor ( TMRCA ), population divergence times ( t) and effective population size ( Ne) were estimated for the two mainland regions (east and west) and for eastern Australia and Tasmania using the program IM (Nielsen and Wakeley 2001, Hey and Nielsen 2004). Population splitting parameter ( s) were also estimated from the data which allowed for populations to change in size (Hey 2005).

Prior distribution settings for IM were based on posterior estimates from preliminary runs using wide parameter intervals following Nielsen and Wakeley (2001). Metropolis coupling was initiated to swap between five heated chains to ensure parameter space was sampled effectively. Each run was continued for 3, 000, 000 updates and the resulting distribution peaks were used as parameter estimates of population divergence time ( t) and ancestral theta ( ΘA) and contemporary theta for each geographic region. To convert divergence time in units into years, a sequence divergence rate and generation

60 Phylogeography time were required. Sequence divergence rates for control region in birds can differ among species and among domain I, II and III within a species (Baker and Marshall 1997, Ruokonen and Kvist 2002). Thus, I calculated a species-specific divergence rate for my control region fragment using the divergence between clades for cytochrome b. One eastern population (Seymour) and one western population (Perth) were sequenced for cytochrome b variation using the following primers L15191 (5'-ATC TGC ATC TAC CTA CAC ATC GG) and H15916 (5'-ATG AAG GGA TGT TCT ACT GGT TG)

(Lanyon and Hall 1994) (GENBANK accession no.s EF173681-EF173693). Mean net divergence was calculated between clades and compared with divergence calculated for control region for the two populations. Assuming a 2%/ MYR sequence divergence rate for cytochrome b which was estimated for avian mtDNA coding region (Shields and Wilson 1987, Weir 2006) a sequence divergence rate for the control region fragment of

1.9%/ MYR was estimated. This estimate is unusually low for the control region. However, Ruokonen and Kvist (2002) have demonstrated that control region is not always more variable than that of cytochrome b. A generation time of 4 years was used when converting parameter estimates into years based on data from a long term parentage study on magpies (J. M. Hughes data unpublished). IM was run repeated times for each pair wise comparison to ensure consistency of parameter estimates.

4.3 RESULTS

4.3.1 Phylogenetic Analysis

The hierarchical likelihood ratio test employed in MODELTEST supported the use of the TrN+I+G ( i = 0.8871; α = 0.7333) model for evolution of the magpie control region fragment. The maximum likelihood phylogeny (Figure 4.1) was constructed from 522 base pairs of control region mtDNA sequence and rooted with a pied butcherbird (Cracticus nigrogularis ) sequence. Two clades were supported by bootstrapping that corresponded to an eastern (clade A) and a western clade (clade B). Clades A and B were monophyletic.

61

Phylogeography

Model corrected mean sequence divergence between clades was 1.64 % and mean pairwise difference within eastern and western clades was 0.76 % and 0.57 % respectively.

4.3.2 Phylogeography Nesting of the haplotype network is presented in Figure 4.2. Nesting of the haplotypes identified 26 clades over four levels including the total cladogram. A significant interaction was found between geographic location and clade (significant permutation test) for 14 of the 26 clades (Table 4.1). Haplotypes within one clade (2-5) could be nested two ways and both nesting arrangements were significant. One nesting arrangement was inferred as restricted gene flow, however the other was inconclusive. The total cladogram was significant and long distance colonisation with fragmentation or past fragmentation with range expansion was indicated for the pattern of genetic divergence observed between eastern and western regions. Within the western region, all 1-step and 2-step significant clades were interpreted as restricted gene flow with either isolation by distance (IBD) (clade 1-1 and 1-2) or long distance colonisation (clade 2-2). This result suggests that the contemporary mtDNA population structure of magpies in the west is limited by restricted gene flow among certain populations. Within the eastern region; with the exception of clade 1-11 (restricted gene flow with IBD) and clade 2-5 (inconclusive or restricted gene flow with IBD), all significant 1- step (1-7, 1-10, 1-11, 1-12, 1-15, 1-17) and 2-step (2-4, 2-7) clades were consistent with a signature of contiguous range expansion. The only significant geographic association at the 3-step level corresponds with populations in the east. This association was interpreted as indicating restricted gene flow with IBD.

62 Phylogeography

gt1 gt10 gt 41 gt8 gt13 gt14 gt26 gt36 gt2 gt3 gt5 gt11 gt25 gt39 gt7 Clade A gt42 gt46 gt49 gt48 gt4 gt35 gt40 gt45 gt9 gt6 gt15 gt27 gt44 gt50 gt16 gt17 gt18 gt20 gt32 gt21 gt31 gt19 gt30 Clade B gt34 gt37 gt43 gt38 62 gt22 gt33 gt23 gt29 Cracticus nigrogularis 0.005 substitutions/site

Figure 4.1: Maximum-likelihood tree (-ln = 1203.41) constructed in PAUP showing the relationships among 46 magpie mitochondrial control region haplotypes. The tree was rooted with a pied butcherbird C. nigrogularis sequence. Branch lengths are proportional to reconstructed distances. Numbers next to branches indicate nodes supported in > 60% of 100 bootstrap replicates.

63

Phylogeography

22 9 20 33 31 2-2 44 50 2-3 1-3 1-2 1-7 21 18 17 29 23 16 6 15 1-5 1-6 1-8

30 37 3-2 14 10 1 49 19 38 4 1-11 43 3-1 35 1-10 1-4 34 45 40 32 1-9 2-1 1-1 2-4

2-6 8 46 13 1-14 1-12 2 7 41 36 1-13 1-15 42 26 48 2-5

11 3 5 27 3-3 1-17 39 25 2-7 1-16

Figure 4.2: Parsimony network computed by TCS using control region sequence data for 46 haplotypes. Associated nested design for the nested clade analysis is also shown. Size of haplotype circles indicates comparative frequency of haplotypes.

64 Phylogeography

Table 4.1: Nested clade analysis for populations of G. tibicen . Only clades where a significant permutation test ( χ2) was indicated are shown. Significant values for the S clade ( Dc), nested clade ( Dn) and interior-tip (I-T) distances are designated by (small) and L (large). Alternative nesting and results are given for clade 2-5. 2 Clade χ ( P) Clade no. Position Dc Dn Inference (key) 1-1 265.1 19 I 829.0 822.0 restricted gene flow with IBD west (0.002) 30 T 184.0 678.0 (1-2-3-4) 32 T 0.0 S 1162.0 34 T 0.0 858.0 38 T 0.0 274.0 S I-T 774.0 L -45.0 1-2 221.8 17 I 764.0 758.0 restricted gene flow with IBD west (0.000) 22 T 0.0 761.0 (1-19-20-2-11-17-4) 31 T 0.0 376.0 S 33 T 0.0 765.0 I-T 764.0 L 92.0 1-7 186.4 6 I 601.0 S 660.0 S contiguous range expansion east (0.016) 9 T 328.0 831.0 (1-2-11-12) 44 T 0.0 719.0 50 T 0.0 1312.0 L I-T 355.0 -217.0 S 1-10 421.0 4 I 763.0 S 774.0 S contiguous range expansion east (0.001) 35 T 0.0 667.0 (1-19-20-2-11-12) 49 T 0.0 1398.0 L I-T 763.0 S -258.0 S 1-11 56.0 1 T 225.0 S 321.0 S restricted gene flow with IBD east (0.000) 10 I 124.0 930.0 L (1-19-20-2-3-4) 14 I 0.0 773.0 I-T -126.0 578.0 L 1-12 29.8 8 T 553.0 L 930.0 contiguous range expansion east (0.000) 13 I 799.0 797.0 (1-2-11-12) I-T 246.0 -132.0 1-15 45.9 7 I 760.0 750.0 contiguous range expansion east (0.019) 41 T 907.0 L 1208.0 L (1-2-11-12) 42 T 0.0 682.0 46 T 0.0 1307.0 I-T 111.0 -398.0 1-17 18.0 5 I 573.0 S 704.0 S contiguous range expansion east (0.001) 27 T 0.0 S 1022.0 (1-19-20-2-11-12) I-T 573.0 -319.0 2-2 88.3 1-2 I 725.0 705.0 restricted gene flow with some long west (0.004) 1-3 T 139.0 S 569.0 distance colonisation 1-5 I 25.0 337.0 S (1-2-3-5-6-7) I-T 573.0 L 130.0 2-4 218.1 1-9 T 847.0 L 928.0 contiguous range expansion east (0.000) 1-10 I 719.0 724.0 (1-2-11-12) 1-11 T 427.0 850.0 I-T 252.0 -133.0 2-5 45.0 1-12 T 807.0 835.0 restricted gene flow with IBD east (0.000) 1-13 I 482.0 1009.0 (1-19-20-2-11-17-4)

65

Phylogeography

2 Clade χ ( P) Clade no. Position Dc Dn Inference (key) I-T -325.0 174.0 L 2-5 45.0 1-12 I 806.7 834.9 Inconclusive east (0.000) 1-13 T 482.1 1009.3 (1-19-20-2-11-17) I-T 324.6 -174.4 S 2-7 136.2 1-16 I 842.0 S 858.0 S contiguous range expansion east (0.000) 1-17 T 784.0 978.0 (1-2-11-12) I-T 57.0 -120.0 3-3 307.2 2-5 T 872.0 881.0 restricted gene flow with IBD east (0.000) 2-6 I 1004.0 1018.0 L (1-2-3-4) 2-7 T 876.0 S 917.0 I-T 129.0 L 107.0 L Total 1497.5 3-1 T 777.0 S 1713.0 L long distance colonisation or past (0.000) 3-2 I 737.0 S 1189.0 fragmentation and range expansion 3-3 T 956.0 S 1229.0 S (1-2-11-12-13-21) I-T -157.0 S -210.0 S

4.3.3 Demographic Expansion

LAMARC simulation results are shown in Table 4.2. Large positive values of g indicate an expanding population (Kuhner et al. 1998). There was no evidence supporting population growth for any of the locations tested here. All values for growth were very close to zero. An ANOVA was used to test for differences among regions. No significant differences were found among any populations ( F = 1.67, df = 29, P = 0.16) suggesting all regions have undergone a similar demographic history.

Table 4.2: Growth (g) simulation results from LAMARC for populations of magpies. Estimation of theta and g and standard deviations of three runs are given. Randomly sampled populations that were used for analysis are in bold. Location Populations Θ growth (g)

north east Ct , Bw, Mb, Rk 4.68 ± 1.19 -0.07 ± 0.26 central east Ma, To , Br, Gr 4.10 ± 0.71 -0.12 ± 0.11 south-central east Db , Or, Ck, Gb 4.11 ± 0.80 -0.05 ± 0.17 south east Ou , Hm, Sr, Pi 4.62 ± 1.39 -0.09 ± 0.42 Tasmania Ts 1.92 ± 0.30 -0.07 ± 0.06 south west Ep, Al 3.34 ± 0.51 -0.06 ± 0.04 Busselton Bn 2.45 ± 0.52 -0.06 ± 0.54 south-central west Mh, Pe 3.32 ± 0.58 -0.05 ± 0.29 Pilbara Pb 6.37 ± 1.30 -0.06 ± 0.22 Kimberley Km 7.28 ± 4.08 2.01 ± 2.66

66 Phylogeography

Mismatch distributions for pooled sites shown in Figure 4.3, are uni-modal, which is indicative of an expanding population. Interestingly there appears to be little difference between the observed distribution and simulated distribution under a constant population size for most populations. No populations at the local level or clade level were significant for Tajima’s D. Fu’s FS was significant ( P < 0.02) and mismatch distributions were uni-modal when populations were pooled at the clade level for both east and west (Figure 4.4). No population pooled at the local level however, was significant for the Fu’s FS test. The significance at the total clade level may result from either pooling of populations that are genetically differentiated across the region or increased statistical power gained from a larger sample size. Although no population at the local level was significant, two pooled populations (north east P = 0.08 and south east P = 0.03) at opposite ends of the eastern range were close to significance, which may suggest that a population expansion has occurred at the outer range of the distribution for magpies in eastern Australia. However, the western clade was also significant for Fu’s FS and in no cases was a significant result found at the local level in the western region.

67

Phylogeography

Alice Springs

0.6 D = 0.20 ( P = 0.68) FS = 0.11 ( P = 0.45) 0.4

0.2

0 0 1 2 3 4 5 6 7 8

north east

0.6 D = -0.89 ( P = 0.20) F = -3.39 ( P = 0.08) 0.4 S

0.2

0 0 1 2 3 4 5 6 7 8

central east south-central east

0.4 D = -0.04 ( P = 0.55) 0.4 D = 0.11 ( P = 0.62) FS = 0.63 ( P = 0.65) FS = 0.25 ( P = 0.60) 0.2 0.2

0 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

south east Phillip Island

0.6 D = -0.30 ( P = 0.45) 0.6 D = -0.48 ( P = 0.39) F = -5.94 ( P = 0.03) F = -0.42 ( P = 0.40) 0.4 S 0.4 S

0.2 0.2

0 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

Tasmania

0.8 D = 0.08 ( P = 0.65) observed 0.6 FS = 0.25 ( P = 0.47) simulated constant 0.4

0.2 simulated expansion 0 0 1 2 3 4 5 6 7 8

Figure 4.3: Mismatch distribution of magpie haplotypes in 14 populations. Observed pairwise differences are plotted along with the simulated distribution expected under a constant size population and an expanding population. Sites were pooled and coloured circles indicate location of sites on the map. Tajima’s D and Fu’s FS tests are presented for each population. Significance values are in parentheses.

68 Phylogeography

central south

0.6 D = -0.44 ( P = 0.34) 0.4 FS = 0.17 ( P = 0.10) 0.2

0 0 1 2 3 4 5 6 7 8

Ceduna

0.8 D = 2.10 ( P = 0.99) 0.6 FS = 3.93 ( P = 0.97) 0.4 0.2 0 0 1 2 3 4 5 6 7 8

south west Busselton

0.6 D = 0.69 ( P = 0.79) 0.8 D = -0.41 ( P = 0.38) 0.6 0.4 FS = -1.23 ( P = 0.25) FS = -0.66 ( P = 0.32) 0.4 0.2 0.2 0 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

south-central west Pilbarra

0.6 D = 0.62 ( P =0.77) 0.6 D = -1.15 ( P = 0.14) 0.4 FS = 0.73 ( P = 0.66) 0.4 FS = -1.44 ( P = 0.12)

0.2 0.2

0 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

Kimberley

observed 0.8 D = -1.41 ( P = 0.07) 0.6 FS = 0.07 ( P = 0.41) simulated constant 0.4 simulated expansion 0.2 0 0 1 2 3 4 5 6 7 8

Figure 4.3 continued: Mismatch distribution of magpie haplotypes in 14 populations. Observed pairwise differences are plotted along with the simulated distribution expected under a constant size population and an expanding population. Sites were pooled and coloured circles indicate location of sites on the map. Tajima’s D and Fu’s FS tests are presented for each population. Significance values are in parentheses.

69

Phylogeography

eastern region

0.4 0.35 D = -0.77 ( P = 0.243) 0.3 FS = -15.38 ( P = 0.003) 0.25 0.2 0.15 0.1 0.05 0 0 2 4 6 8 10 12 14 16 18 20

western region

0.6 D = -1.40 ( P = 0.056) 0.5 FS = -10.45 ( P = 0.001) 0.4

0.3

0.2

0.1

0 0 2 4 6 8 10 12 14 16

observed simulated constant

simulated expansion

Figure 4.4: Mismatch distribution of magpie haplotypes for the east clade and west clade. Observed pairwise differences are plotted along with the simulated distribution expected under a constant size population and an expanding population. Tajima’s D and Fu’s FS tests are presented for each clade. Significance values are in parentheses.

4.3.4 Coalescent Analyses Estimates of divergence time between the eastern and western regions indicated a recent split during the late Pleistocene. The peak of the distribution (Figure 4.5) corresponded to 36, 533 years bp (90% HPD interval 22, 645 - 51, 027 years bp) (Table 4.3). The split between the eastern region and the Tasmanian region (Figure 4.6) was even more recent and was estimated at 15, 399 years bp (90% HPD interval 6, 944 - 25, 664 years bp). As expected TMRCA estimates (Table 4.3) were greater than population divergence estimates for all pairwise comparisons. Estimates of the effective population size ( Ne)

70 Phylogeography for the eastern region varied from 828, 217 to 1, 160, 429 among runs and was consistently far greater than estimates for the western region ( Ne: 20, 887 - 115, 579) or the Tasmania region ( Ne: 5, 671 - 36, 458). The population splitting parameter ( s) was estimated at 0.89 (90% HPD 0.87 - 0.91) for the east-west comparison, indicating only a small proportion (< 0.15) of the ancestral population apparently founded the western region.

Table 4.3: Coalescent parameters and population divergence estimated with IM . Two comparisons are made: divergence of east and west and the divergence of east and Tasmania. Population size estimates ( Ne) are given for each population and ancestral population. ( NeA). Maximum likelihood estimation (MLE) refers to the peak of the distribution. Lower and upper highest posterior density levels (HPD) are given for each estimate.

t TMRCA Θ Θ2 Θ Ne1 Ne2 NeA A (years) (years) 1.east x 2.west

MLE 92.2 5.2 11.4 1160 429 65 447 143 429 36 534 442 230

Lower 90% 70.1 1.7 6.5 881 926 20 887 82 158 22 644 HPD Upper 90% 115.6 9.2 16.7 1453 785 115 578 210 269 51 027 HPD

1.east x 2. Tasmania

MLE 65.8 0.6 8.4 828 017 8 102 105 973 15 399 331 119 Lower 90% 44.7 0.2 4.3 562 274 2 268 53 472 6 944 HPD Upper 90% 87.7 1.4 13.2 1103 481 17 824 166 252 25 664 HPD

71

Phylogeography

0.08 0.06 0.07 Western Population 0.05 0.06 Eastern Population 0.04 0.05 Ancestral Population 0.04 0.03 0.03

P robability P robability0.02 0.02 0.01 0.01 0 0 0 50 100 150 200 250 300 0 0.5 1 1.5 2 2.5 3 A. Population Size B. Time

Figure 4.5: Posterior probability results from IM . A. Estimated population size for the east and the west and for the ancestral population. B. Estimated time of population split. Results given as IM output. Actual population size and time are given in Table 4.3.

0.09 0.08 0.08 Tasmania 0.07 0.07 Eastern Population 0.06 0.06 0.05 0.05 Ancestral population 0.04 0.04 0.03

P robability0.03 P robability 0.02 0.02 0.01 0.01 0 0 0 50 100 150 200 0 0.5 1 1.5 2 2.5 3 Population Size B. Time A.

Figure 4.6: Posterior probability results from IM . A. Estimated population size for the east and Tasmania and for the ancestral population. B. Estimated time of population split. Results given as IM output. Actual population size and time are given in Table 4.3.

4.4 DISCUSSION

4.4.1 Phylogeography of the Australian magpie The shallow divergence of monophyletic clades shown in the maximum likelihood tree (mtDNA control region) indicates a relatively recent independent history for eastern and western populations of magpies. As reported earlier, eastern and western regions of mainland Australia are separated by the Nullarbor Plain in the south and the Carpentarian divide in northern Australia. The central Australian population from Alice

72 Phylogeography

Springs (As) formed part of the eastern clade. The Nullarbor and Carpentarian divide have been proposed as important barriers to dispersal for many species during periods of aridity during the Pleistocene (Cracraft 1986, Ford 1987). The formation of arid barriers began at a much earlier date during the tertiary between 65-1.8 million years ago, however intensity of arid periods continued to increase to present day levels (Bowler 1976). Previous genetic studies have suggested that the Nullarbor Plain continues to provide a significant barrier to gene flow for white-winged fairy wrens (Driskell et al. 2002) and other vertebrate species (Spencer et al. 2001). From mtDNA control region data, I estimated that all western populations have been isolated from eastern populations for at least 36, 000 years. Thus, the divergence may have begun early in the most recent period of aridity approximately 100, 000 to 16, 000 bp (Miller et al. 1997). The credibility intervals around these estimates were relatively small (eg. east-west estimate = 22 - 50 kya) and each run resulted in consistent estimates. It is important to consider however, that the sequence divergence rate (1.9%/million years) used to calculate the estimate was low compared with that for many control region studies. Estimates recently used for avian control region range from 5% (Freeland and Boag 1999) to 9.7% (Peters et al. 2005) based on a higher mutation rate in control region when compared with mtDNA coding regions. Other studies (Ho et al. 2005, Weir 2006) have shown that the rate of change in avian coding mtDNA may also be faster than 2%/million years over recent time scales (< 2 million years). Using a faster rate to calculate the divergence estimates however, would give an even more recent east-west divergence time than 36, 000 years reinforcing the suggestion that divergence among magpie populations is a very recent phenomenon.

Baker (1999) in an earlier study showed that south-eastern populations and south- western magpie populations were monophyletic. However, without samples from northern Australia, Baker could not rule out the possibility that eastern and western magpie populations may be separated by isolation by distance across the north with the southern populations at the two extremes of the distribution. My study showed that eastern and western mtDNA clades are in fact reciprocally monophyletic. This could

73

Phylogeography indicate that divergence between mainland populations was the result of vicariance rather than isolation by distance. There was insufficient evidence from nested clade analysis to decide between long distance colonisation and past gradual range expansion followed by fragmentation. The population splitting parameter s indicated only a small proportion of the ancestral population founded the western population, which would be expected if only a small population of eastern birds colonised the western region. This may also be due however, to either a colonisation event or fragmentation of an ancestral population that shows a pattern of isolation by distance from east to west. Although it is difficult to distinguish between alternative hypotheses about the formation of separate populations, the timing of the divergence coincides with other reported events occurring in Australia, indicating that vicariance due to habitat change is the most likely explanation. Firstly, the timing of divergence coincides with evidence for an increase in aridity suggested by analysis of sand dune formation in central Australia (Bowler 1976). Secondly, the timing of the divergence falls within the last period of aridity (120, 000 – 18, 000 years ago) when sea levels fell significantly (see Figure 1.2) (Fleming et al. 1998).

Several avian (white-winged fairy wren, Malurus leucopterus Driskell et al. 2002; ringneck parrot, Barnardius zonarius Joseph and Wilke 2006) and other vertebrate species (dunnart, Sminthopsis crassicaudata Cooper et al. 2000; brush-tailed phascogale, Phascogale tapoatafa Spencer et al. 2001) also demonstrate a pattern of east to west divergence. Using mtDNA sequence divergence between lineages, several of these studies have dated divergence between eastern and western lineages ( TMRCA ) to the Plio-Pleistocene boundary around 1.8 million years ago (eg. ringneck parrot and Australian dunnart) or within the last 200, 000 years (eg. white-winged fairy wren). As it is generally accepted that an estimate of lineage divergence will pre-date actual population divergence times (Arbogast et al. 2002), these studies are consistent with a proposed Pleistocene or even a late Pleistocene east-west population split for magpies. Other studies however, suggest much older divergence times between eastern and western lineages (eg. 5-8 million years for populations of the brush-tailed phascogale)

74 Phylogeography indicating that increasing levels of aridity may have fragmented some species before the Pleistocene. The difference among species in the timing of an east-west split, is likely to be the result of variability in habitat specificity and relative vagility as is reported for differences among some species across the Black Mountain Corridor in the Wet Tropics in north eastern Australia (Joseph et al. 1995).

For the ringneck parrot, B. zonarius , Joseph and Wilke (2006) have identified the location of the east-west split to be east of the Nullarbor Plain, at a location that is consistent with the putative Eyrean barrier. The Eyrean and Mallee barriers are considered to have been the most severe of the arid barriers during the most recent period of aridity (Ford 1987). In this study a single site, Ceduna (Cd), was sampled between the Nullarbor Plain and the Eyrean barrier and although the two haplotypes found at Ceduna were nested within the eastern clade, they were not shared with any other eastern population. The sample size at Ceduna however, was small (n = 8). This result indicates however, that isolation of Ceduna from other eastern populations may have occurred, possibly in very recent times, as only one base pair difference was observed from other eastern haplotypes. Moreover, Ceduna samples fall within the eastern clade, suggesting that if in fact they represent an isolated population, there has not been enough time for lineage sorting to occur. Alternatively, it is quite plausible that magpies recolonised southern Australia from refugia located to the east of the Mallee barrier and to the west of the Nullarbor Plain. Further intensive sampling across these putative barriers will be required to work out the historical limits to eastern and western populations across the south.

Although the above mentioned species show concordance with the phylogeographic structure of the Australian magpie, they are mostly distributed throughout the lower half of Australia. Few studies have tested for phylogeographic breaks in northern Australian birds. The grey-crowned babbler Pomatostomus temporalis is distributed across much of northern and eastern Australia (Edwards and Wilson 1990). A study using restriction analysis and direct sequencing to construct a phylogeny for the grey-crowned babbler

75

Phylogeography found divergence between western and eastern populations that were geographically concordant with the refugia on either side of the putative Carpentarian barrier (Edwards and Wilson 1990). Analysis of control region variation (sub unit I) indicated a far deeper divergence (5.3%) for the grey-crowned babbler (Edwards and Kot 1995, Edwards 1997) than I have estimated for control region (I and II) for eastern and western magpie populations (1.6%). Assuming that control region substitution rates for babblers and magpies (Australian songbirds) are similar, it is plausible that the different depth of divergence is due to the rise of different barriers, the Carpentarian (north) for the babblers and the Nullarbor/Eyrean (south) for the magpies. Australian magpies are far more densely populated in southern Australia whereas babblers are not present over much of central and western regions of southern Australia. Densities of magpies today suggest that they were fragmented to south-eastern and south-western Australia on either side of the Nullarbor during the late Pleistocene and have since recolonised northern Australia as habitat has become more favourable.

The east to west pattern of divergence that has been identified for neutral DNA sequences contrasts with the striking north to south pattern in morphological (back colour) variation in this species. Using the morphological classification in Schodde and Mason (1999) the eastern clade (inclusive of Tasmania) incorporates five ultrataxa ( G. t. terraereginae , G. t. tibicen , G. t. tyrannica , G. t. hypoleuca , G. t. telonocua ) and shares one ( G. t. eylandtensis ) with the western clade. The western clade includes two additional ultrataxa ( G. t. dorsalis and G. t. longirostris ). Even when simplified to differences based on dorsal colour, both BB and WB forms occur in the eastern region and BB and varied (a sub group of WB) forms occur in the western region. Hughes et al. (2001) have hypothesised that persistence of variation in dorsal pattern (BB and WB) in south eastern Australia is not due to recent history, but may result from a combination of sexual selection favouring brightly coloured WB males and selection favouring camouflage (BB) against predation. Data presented here also indicates that there is no evidence that north-south morphological variation in back colour in magpies is due to

76 Phylogeography recent history but rather that recent history has played a role in isolating magpies independent of back colour to eastern and western sides of the continent.

Climate change during the Pleistocene coincided with periods of significant sea level fluctuations, allowing periodical connectivity between some island and mainland populations. During the last glacial cycle, the sea level is believed to have dropped by as much as 150 meters (Chappell 1983, Chappell and Shackleton 1986) establishing a land bridge between Tasmania and the mainland and potentially allowing dispersal of magpies across Bass Strait. However, sea levels rose during the inter-glacial/arid periods facilitating isolation of Tasmania. My estimate of time since population divergence from the mainland of 15, 000 years suggests that the Tasmanian population was isolated from mainland populations after the Last Glacial Maximum, 18, 000 years ago. As predicted from the sharing of haplotypes between eastern mainland populations and Tasmania, this split was of very recent origin and more recent than the timing of the mainland east to west split. The IM results also suggested a relatively small population size for Tasmania, between 5, 000 and 37, 000 individuals, compared with estimates for the eastern region of around one million individuals. Furthermore, Tasmania had lower allelic richness and heterozygosity than all other populations sampled on the mainland. Reduced variation at neutral loci has been reported in several island populations when compared with their mainland counterparts for Australian marsupial species (Eldridge et al. 2004). Although Tasmania is a relatively large island, suitable habitat for magpies is confined to limited small areas in the north and east (Schodde and Mason 1999) and therefore modern populations on Tasmania may be affected by loss of genetic diversity in a similar way to that of many small island populations.

4.4.2 Demographic Growth and Range Expansion Ford (1987) identified up to nine possible coastal and inland refugia that may have been present during the last period of cyclic aridity in Australia. From my analysis it appears that only two refugia (east and west) are required to explain the modern variation and population structure in the Australian magpie. Visual inspection of the haplotype

77

Phylogeography network indicates that lineages from populations on either side of the Nullarbor Plain are more closely related (fewer parsimony steps) to each other, than equivalent populations on either side of the Carpentarian divide. This pattern supports a southern origin for the east and west clades and their subsequent expansion northward. In the east, more variation was observed in mtDNA in some southern populations than northern populations which also suggests that the southern populations are ancestral. Recent contiguous range expansion for the Australian magpie throughout the eastern region was also inferred from the NCA analysis. Support was not found however, for a northward expansion from a south-eastern refuge as significant clades were found across the entire eastern range. Rather, the east coast may have been a refuge for magpies with subsequent expansion directed inland towards central Australia. Proposed geographical barriers along the east coast were generally highland barriers (eg. Burdekin, McPherson range, Blue Mountains/Snowy Mountains) and were unlikely to be effective barriers to dispersal for a moderately vagile species with low habitat specificity such as the magpie. Thus, a continuous refuge along the east coast for some avian species may have remained throughout the last period of aridity. Central Australia however, was isolated from coastal refuges at the last peak of aridity (18, 000 years bp) due to formation of extensive sand dune systems. The central Australian site (As) shows no genetic disjunction from other eastern sites suggesting that it has probably been colonised from the east in recent times. No differences in diversity were detected among populations in the western region and no signature of range expansion was indicated in the NCA suggesting that the western region has been relatively stable, at least in recent times.

Historical inferences of range expansion using NCA can be corroborated with methods that detect a signature of population expansion (eg. Fu’s FS, Tajima’s D). Considering habitat would most likely have been reduced for magpies during times of aridity, and the last period of aridity was relatively recent (maximum 18, 000 years bp), it was expected that a signature of population expansion would be present in the dataset. Interestingly, the tests used to detect demographic expansion returned conflicting results

78 Phylogeography depending on whether the sampling level was a clade or a group of local sites. Tests for population growth may be significant if there is population structure in the data set (Fu 1996). This is a difficult assumption to satisfy as most natural populations demonstrate some contemporary structure over large spatial scales. In the current study, the assumption of population structure at the clade level was violated and therefore only results at the local level will be discussed. Two eastern populations (north-east and south-east) were almost significant for Fu’s FS, indicating that population growth may be restricted to north and south regions in the east. This pattern may be the result of the northern and southern populations extending their range into less populated adjacent regions of northern and southern Australia. There was however, no evidence for population growth for the western region for mtDNA or for either region using microsatellite data. In fact, no difference was detected among any of the populations simulated for growth using microsatellite variation. Due to the high mutation rate and larger effective population size of microsatellite loci (Jarne and Lagoda 1996), signatures of population growth may be eroded more quickly in microsatellite data than for mtDNA. If this is the case, microsatellite data may only be useful when detecting very recent demographic changes. There are several hypotheses as to why little evidence was found for population expansion. It is possible that populations of magpies may not have increased significantly in size as they expanded their range out of refugia. This has been suggested as a possibility for the lack of evidence for population growth of North American clades of barn swallows Hirundo rustica (Zink et al. 2006). Generalist species such as magpies, are less likely to be affected by habitat reduction than other species due to their ability to access a wide range of resources and therefore may be able to retain relatively high densities even in small habitat patches. Alternatively, the model for population growth may not be adequate for detecting more complex demographic changes that could evolve during many cyclic periods of habitat expansion and contraction in the Pleistocene.

79

Feather Lice

CHAPTER 5: COMPARATIVE ANALYSIS OF TWO SPECIES OF HOST SPECIFIC FEATHER LICE

5.1 INTRODUCTION Comparative analysis of genetic variation among host - parasite systems can provide insights into the ecology and history of both the parasite and the host. Obligate ectoparasites spend their entire life cycle on a host and dispersal to a new host relies on direct contact between two hosts (Marshall 1981). Because of this close relationship, obligate parasites share a common history with their host, therefore host dispersal and population structure may be inferred from genetic variation in the parasite. McCoy et al. (2003) looked at genetic variation in a species tick ( Ixodes uriae ) that occurred on two hosts, the black-legged kittiwake ( Rissa tridactyla ) and the Atlantic puffin ( Fratercula arctica ) and found greater genetic structure for kittiwakes than for puffins. They inferred that inter-colony movements (dispersal without gene flow) were more common for puffins than for kittiwakes.

Some parasites such as feather lice have been shown to have an elevated rate of evolution compared to their host (Page et al. 1998). This elevated rate of evolution may be utilised to uncover the recent history of the host. In the current study, genetic variation in two species of obligate feather lice will be analysed to test the idea that lice mtDNA is more sensitive to host history than host mtDNA due to the elevated rate of evolution in feather lice. For the magpie, analysis of microsatellite variation suggested there was gene flow from eastern Australia through to north-western Australia, although there was no indication of secondary contact between east and west for the mtDNA. Analysis of genetic variation for host specific ectoparasites of magpies could verify if there has been secondary contact between eastern and western populations of magpies.

Two species of chewing lice: Philopterus sp. and Brueelia semiannulata are obligate ectoparasites of the Australian magpie, G. tibicen . They are distantly related species and therefore act as independent taxa for evolutionary questions. Clayton et al. (2003a)

80 Feather lice suggested that the distribution and abundance of parasites on a host may influence the genetic variation in parasite populations. For example, those species that are found at low densities on a host will have an increased chance of extinction thereby reducing the total number of sub-populations that exchange genes. Philopterus sp. are highly habitat specific being found only on the head and nape feathers of the magpie. B. semiannulata are found on the body feathers of the magpie. In contrast, the habitat for B. semiannulata is much larger and supports a range of feather sizes suggesting that it is a generalist species (Clay 1951). Higher habitat specificity and lower habitat availability for Philopterus sp. may result in lower dispersal potential than for B. semiannulata . This is because there are probably less chances for Philopterus sp. to gain access to a new host through host to host contact. Other species of lice that share a host but live in different habitats also display differences in dispersal potential. Wing lice ( Columbicola spp.) have a higher incidence of host transfer than do body lice ( Physconelloides spp.) in pigeons and doves resulting in greater levels of local population structure in body lice (Johnson et al. 2002b).

Hughes (1984a) examined the distribution and density of feather lice on the Australian magpie. She found significantly more birds infested with B. semiannulata than Philopterus sp. and generally, although the difference did not reach significance, B. semiannulata were found at higher densities than were Philopterus sp. Infestations of Philopterus sp. were reported to be significantly higher on mainland birds than on Tasmanian birds, although no differences were found among mainland populations. Percent of birds infested with B. semiannulata also differed among mainland magpie populations with significantly more BB magpies infested than either WB magpies (mainland) or varied magpies. Hughes (1984a) postulated that variation in percent of infestation among contiguous populations of magpies for B. semiannulata but not Philopterus sp. may be due to variable habitat (plumage forms) that B. semiannulata occupies. Philopterus sp . only occurs on the white nape feathers in all magpie populations. In contrast B. semiannulata occurs on black dorsal feathers in northern populations, white dorsal feathers in south eastern populations and black (females and

81

Feather Lice juveniles) or white (males) dorsal feathers in south western populations. Selection for conspicuousness of lice on different background or thermal attributes of feather colours may explain relative differences in percentage and density of infestations among populations (Hughes 1984a).

In a morphometric analysis based on head characters, Hughes (1984b) reported little difference among mainland populations of B. semiannulata or between mainland and Tasmanian lice. The Tasmanian population of Philopterus sp. however, was found to be distinct from mainland populations. Moreover, morphological variation was found between some mainland populations of Philopterus sp., although no consistent geographical pattern (eg. Isolation by Distance) was evident.

Phylogeographic analysis of the magpie presented earlier in Chapter 4 based on mitochondrial control region variation indicated shallow east-west structuring. It was estimated that populations had been isolated on either side of the continent in the late Pleistocene (approximately 36, 000 years ago). Tasmania however, represented a population that had been isolated much more recently (approximately 15, 000 years ago), possibly since the last glacial maximum. Lice morphological variation is not associated however, with the east-west divergence among magpie host populations for either species of lice. This discordance between lice morphology and genetic variation in the host, suggests that either variation in louse morphology does not reflect population structure of their host or that morphometric analysis of head characteristics may not be reflective of past lice-host evolution.

Here molecular phylogenies and gene networks will be used to investigate the population structure of two species of ectoparasitic feather lice and their avian host. Feather lice were collected from a subset of mainland magpie populations used in the genetic analysis of magpies. The objective where possible, was to sample populations from both eastern and western magpie mtDNA clades.

82 Feather lice

The aim of the study was to address the following 1. Do the phylogenies of Philopterus sp. and B. semiannulata reflect the phylogeny of the avian host? 2. Does genetic variation for either louse species support the inference of secondary contact between north-eastern and north-western populations of the magpie host as suggested from the microsatellite analysis? 3. Does the higher habitat specificity and lower densities of Philopterus sp. result in deeper mean genetic divergence between populations of Philopterus sp. than between equivalent populations of B. semiannulata ?

5.2 SPECIFIC METHODS General molecular methods and statistics are described in Chapter 2. Specific methods and statistical analyses are presented in detail below.

5.2.1 Genetic Variation Sequence variation is described for each louse species. Two sequence data sets were used for analysis of each species. Construction of phylograms included the total sequence length inclusive of base ambiguities (672 bp for Philopterus sp. and 600 bp for B. semiannulata ) to maximise sequence data for tree building. All population analyses including parsimony networks and AMOVA’s were based on the trimmed sequence dataset to minimise base ambiguities (464 bp for Philopterus sp. and 433 bp for B. semiannulata ) and to maximise the number of samples for analysis.

5.2.2 Nuclear Pseudogenes To rule out the possibility that genetic patterns in the louse phylogeny were due to sequencing nuclear mitochondrial pseudogenes (Numts) several tests were employed following Bensasson et al. (2001, see www.pseudogene.net). A sample of nine B. semiannulata selected randomly from three populations (Km, Mb, Sr) were extracted with a sequential method that isolates mtDNA (extract 2) from nuclear DNA (extract 1). Sequences were then compared from the two extracts and if nuclear copies of mitochondrial genes were present, different sequences were expected from purified mtDNA extract and the concentrated nuclear extract. A whole louse was homogenised

83

Feather Lice with 500 L of extraction buffer (0.25 M sucrose, 30 mM Tris-HCl ph 7.5, 10mM EDTA) using a small pestle. Homogenised lice were centrifuged at 3000 rpm for 1 minute at 4ºC and supernatant transferred to a clean tube. Supernatant was further centrifuged at 12 500 rpm for 10 minutes at 4ºC to isolate mitochondria from nuclear DNA. Supernatant was removed to a clean tube and the extraction of nuclear DNA (extract 1) proceeded following the phenol-chloroform method outlined in section 2.4.1.1. The pelleted mitochondria from the previous step were resuspended in hydration buffer (10 mM Tris-EDTA pH 8.0, 0.15 M NaCl) to a total volume of 50 L. To the extraction, 100 L of 0.18 M NaOH with 1% SDS was added and the solution was set on ice for 5 minutes. 150 L of ice-cold potassium acetate was added and mixed and set on ice for a further 5 minutes. The extraction was centrifuged at 12 500 rpm for 5 minutes, following which equal volumes of phenol-chloroform and chloroform: isoamyl 24:1 were added and mixed gently on a rotating wheel for 3 minutes. Samples were centrifuged at 12 500 rpm for 3 minutes and the aqueous phase was transferred to a clean tube. To each sample, 100 % ethanol was added and DNA was precipitated for one hour at -20ºC. Samples were centrifuged at 12 500 rpm for 3 minutes to pellet the mtDNA followed by a final wash with 70 % ethanol. mtDNA (extract 2) was resuspended in 30 L of TE buffer (0.01 M Tris-HCl, 1 mM EDTA pH 8.0). Due to the low numbers of samples collected, the experiment was not repeated for Philopterus sp.

Visual inspection of sequences were undertaken for both species as an indication of

Numts. SEQUENCHER 4.1 (Gene Codes Corporation) was used to align sequences, check for the occurrence of stop codons and compare between sequences and clades for amino acid changes.

5.2.3 Hierarchical Analysis An analysis of molecular variance (AMOVA, Excoffier et al. 1992) was used to investigate whether genetic variation among lice populations could be explained by patterns of variation in the host. For each species, populations were grouped based on magpie mtDNA clades such that variation was partitioned east to west. Populations of

84 Feather lice each species were also grouped based on magpie back colour (BB or WB/varied) so that groups represented a north-south partition. A hierarchical partition based on microsatellite data groupings was not tested because there were not enough lice sites that corresponded with different microsatellite groups. Genetic variation was compared within all populations, among populations within groups ( ΦSC ) and between groups

(ΦCT ). Φ-statistics were calculated for each level of analysis using ARLEQUIN 3.01

(Excoffier et al. 2005).

5.3 RESULTS

5.3.1 Genetic Variation 38 Philopterus sp. were collected from sites in northern and eastern Australia. Unfortunately, no Philopterus sp. samples were collected for south-western magpie populations, even though more than 30 birds were captured and examined during sampling. Louse samples could not be obtained from Western Australian Museum as the museum was being relocated.

Philopterus sp. individuals were sequenced for Cytochrome Oxidase I (COI) from 11 sites. The fragment sequenced was highly variable. There were 85 variable sites and a total of 93 mutations. 23 unique haplotypes were identified (see table 5.1) for both total (672 bp) and trimmed (464 bp) sequence. No haplotype was found at more than a single site. Nucleotide sequences are deposited in GENBANK ; accession numbers TBA.

From the 62 B. semiannulata samples sequenced for 600 base pairs of COI, 22 unique haplotypes were identified. However, when the sequence was trimmed to remove ambiguous bases, the number of haplotypes was reduced to 17 (Table 5.2). 85 variable sites were also found for B. semiannulata representing a total of 99 mutations. 16 haplotypes were unique to a single site. The remaining 6 haplotypes represented 65% of all sampled individuals. Nucleotide sequences are deposited in GENBANK ; accession numbers TBA.

85

Feather Lice

While comparing sequence variation within species, it became apparent that some sequences at low frequency were highly divergent to all other haplotypes. For Philopterus sp., a single highly divergent haplotype (ps20) was found in two lice from the Kimberley1 population. This haplotype was more similar genetically to the outgroup collected from a pied butcherbird. Seven very divergent haplotypes forming two clades were identified from B. semiannulata . Divergent haplotypes were found in populations in north-eastern Australia. The discovery of divergent haplotypes in both louse species led to further analysis to rule out the possibility that I had sequenced nuclear homologs of the mitochondrial COI fragment.

Table 5.1: Haplotype frequencies for COI variation for Philopterus sp. using 464 bp As Cc Ct Mb Rk Sr Rw Bm Km1 Km2 Mt total ps1 3 3 ps2 4 4 ps3 1 1 ps4 1 1 ps5 1 1 ps6 1 1 ps7 3 3 ps8 2 2 ps9 1 1 ps10 1 1 ps11 1 1 ps12 1 1 ps13 1 1 ps14 1 1 ps15 1 1 ps16 3 3 ps17 3 3 ps18 2 2 ps19 2 2 ps20 2 2 ps21 1 1 ps22 1 1 ps23 1 1 4 1 5 5 1 6 1 1 7 4 3 38

86 Feather lice

Table 5.2: Haplotype frequencies for 433 base pairs of COI for B. semiannulata Cc Ct Mb Rk Br Sr Rw Pe Bm Ht Mt bs 1 +3 3 1 3 7 3 2 2 21 bs 2 1 1 bs 4 + 5 1 1 2 bs 6 +10 3 4 2 9 bs 7 1 1 bs 8 1 1 bs 9 + 13 1 3 4 bs 11 +14 3 3 bs 12 3 3 bs 15 6 6 bs 16 4 4 bs 17 1 1 bs 18 1 1 bs 19 1 1 bs 20 1 1 bs 21 1 1 2 bs 22 1 1 2 2 4 8 8 15 2 10 3 2 3 62

5.3.2 Phylogeny of Philopterus sp. The hierarchical likelihood ratio test (hLRT) employed in MODELTEST supported the use of the TVM+I+G ( i = 0.5028; α = 0.7288) model for the COI fragment. Maximum likelihood searches produced a single tree (Figure 5.1). Five clades were supported by 70% or greater bootstrap values in the maximum likelihood tree reconstruction. Clade 3 was represented by only a single individual at Rowsley (Rw) and Clade 5 by two individuals from the Kimberley1 (Km1) site. Clade 2 was also restricted to a single site (Alice Springs, As). Clade 1 and Clade 4 accounted for the majority of all samples and both clades were present at multiple sites.

87

Feather Lice

ps14 Population key ps5 As 90 ps1 east Cc Ct ps3 Mb 97 ps4 Rk Clade 1 Sr ps7 Rw ps6 west Bm Km 94 ps2 Mt ps8 0.005 substitutions/site ps10 82 86 ps11 80 Clade 2 ps9 ps12 ps15 Clade 3 100 ps13 ps18 ps19 ps22 Clade 4 99 ps21 86 ps23 ps17 ps16 ps20 Clade 5 Pied butcherbird 0.005 substitutions/site

Figure 5.1: Maximum likelihood tree (-ln =1781.80) estimated for the 672 bps of COI gene for Philopterus sp. Branch lengths are proportional to reconstructed distances. Numbers on branch nodes indicates > 70% support for 500 bootstrap replicates. Magpie ML tree is inset.

5.3.3 Phylogeny of Brueelia semiannulata The GTR+G model (i = 0.0; α = 0.3455) was suggested as the most likely model using the likelihood ratio test in Modeltest for the B. semiannulata dataset. Maximum likelihood searches produced a single tree (Figure 5.2). Four clades were supported by 70% or greater bootstrap values. Clades 1 and 2 accounted for the majority of haplotypes and represented the entire sampling distribution. Net mean divergence among clades ranged from 0.042 to 0.172. Divergence within clades ranged between 0.003 and 0.017.

88 Feather lice

bs14 Population key bs11 Cc east Ct bs7 Mb Rk bs8 Br bs6 Sr Rw 86 bs12 Clade 1 west Pe 95 Bm bs15 Km bs16 0.005 substitutions/site Mt bs10 bs9 98 bs13 bs1 76 bs5 100 bs2 Clade 2 77 bs3 bs4 100 bs18 Clade 3 bs22

99 bs17 bs21 bs20 Clade 4 bs19 0.01 substitutions/site

Figure 5.2: Maximum likelihood tree (-ln = 1685.41) estimated for the 600 base pairs of COI gene for B. semiannulata. Branch lengths are proportional to reconstructed distances. Numbers on branch nodes indicates > 70% support for bootstrap replicates. Magpie ML tree is inset.

5.3.4 Nuclear Pseudogenes The success of the nuclear insertion experiment was limited. Only three samples amplified from both methods of extraction. High amplification failure was possibly the result of the phenol – chloroform method used to extract DNA from B. semiannulata as it had been noted in early attempts to extract DNA from lice that the kit method was far

89

Feather Lice more successful. All three samples that worked were from Moranbah (Mb) and were identical (bs1 and bs3) for both methods. Although there was no evidence of Numts from either extraction method, I was not able to reject the possibility that the divergent lineages were of nuclear origin.

On visual inspection of all sequences, no base ambiguities were noted between complimentary strands or within a sequence. Also there was no indication of any frameshift mutations or stop codons in either louse species sequences, suggesting that the fragment sequenced were in fact from coding genes.

Analysis of amino acid changes in B. semiannulata indicated a single amino acid change and 22 synonymous changes between the two common clades. There were 7 amino acid changes and 65 synonymous changes between the common clades (Clade 1 and Clade 2) and Clade 3. Up to 14 amino acid changes and 79 synonymous changes were identified that differentiated the common clades from Clade 4. Such a high number of amino acid changes suggests either that the divergent lineages may be a different species of louse or that pseudogenes are present. Only one amino acid change (40 synonymous substitutions) differentiated haplotype ps20 from all other Philopterus sp . haplotypes.

As the sequences of concern were low in frequency (7 for B. semiannulata and 2 for Philopterus sp.) and restricted in distribution, these sequences were excluded from all further analyses of the geographic distribution of variation for each louse species. Possible origins for these sequences will be discussed later (discussion).

5.3.5 Distribution of Haplotypes The parsimony network presented in Figure 5.3 shows the relationships and distribution of haplotypes within clades for Philopterus sp. Nine haplotypes were found within Clade 1, which were restricted to eastern Australia. Within Clade 1, no haplotypes were shared among sites and a large number of missing haplotypes were indicated in the

90 Feather lice parsimony analysis. Clade 2 consisted of four haplotypes that were present at a single site only, Alice Springs (As). Eight haplotypes were present in Clade 4, which was restricted to north-western Australia. Haplotype 13, within Clade 4 was present only at the Cape Crawford (Cc) site. Net mean divergence among clades ranged between 3.7% and 9.1%. Divergence within clades was 1.6% for Clade 1, 0.7% for Clade 2 and 1.0 % for Clade 4.

15 17 11 Clade 3 21 3.8% 23 10 6.0% 9 12 3 Clade 2 3.7% 16 22 1 18 4 Clade 4 19 7.1% 14 13

7 5 6 2 3-4 20 1-2 Clade 5 Clade 1 No. of individuals 7.4-9.1% 8

Figure 5.3: Parsimony network ( TCS ) for COI variation among populations of Philopterus sp . Net mean divergence ( p distance) is shown between Clade 1, Clade 2, Clade 3 and Clade 4. The range of net mean divergence for Clade 5 from all other clades is shown. Haplotype size is proportional to the frequency of sampled haplotypes. 91

Feather Lice

The geographical distribution of B. semiannulata mtDNA haplotypes produced a very different pattern to that inferred for Philopterus sp. and the Australian magpie. Two non-overlapping widespread clades (Clade 1 and2) were found in B. semiannulata . These corresponded to northern and southern regions (Figure 5.4). Fewer base pair differences were found between Perth and south – eastern populations (2 base pairs) than between northern and southern clades (4.2% divergence). Within Clade 1, haplotype 6 was present in both south-eastern sites and in Brisbane (Br). Within Clade 2, haplotype 1 was present in all northern sites. Clade 3 and Clade 4 were found in eastern (Brisbane, Rockhampton, Moranbah) and northern sites (Mataranka) at low frequencies and were highly divergent (>10%) from Clade 1 and Clade 2.

92 Feather lice

18

Clade 3 12.5-16.9%

2 4 22

1

Clade 2

4.2% 17

11 12 21

15 16 6 Clade 4 19 7-21 4-6 8 7 14.5%-17.2% 9 20 1-3 Clade 1 No. of individuals

Figure 5.4: Parsimony network for COI variation among B. semiannulata populations. Net mean divergence ( p distance) is shown between Clade 1 and Clade 2. The range of net mean divergence from all other clades is shown for Clade 3 and Clade 4. Haplotype size is proportional to frequency of sampled haplotype.

5.3.4 Hierarchical Analysis Initially AMOVA’s were calculated based on all the data and later with all divergent lineages removed. As removal of the divergent lineages did not change the overall results, I have only presented these AMOVA’s here. Analysis with all data included is in Appendix VII.

93

Feather Lice

The ΦSC indicated a high level of between population within group variation for both species of lice for each AMOVA grouping (Table 5.3). The between group analysis revealed however, remarkably different results, depending of which grouping was used.

Based on host mtDNA grouping, the ΦCT was highly significantly for Philopterus sp. (0.75, P < 0.05), yet was not different for the B. semiannulata. Based on host back- colour grouping, the ΦCT was highly significant for B. semiannulata (0.76, P < 0.05) but not for Philopterus sp.

Table 5.3: AMOVA results for a) Philopterus sp . b) B. semiannulata a. Percentage Variation

among among populations within ΦSC P ΦCT P groups within groups populations 1. host mtDNA clades 0.74 <0.001 0.75 <0.05 74.6% 18.9% 6.5% 2. host back colour 0.90 <0.001 -0.16 0.75 -15.5% 104.3% 11.2%

b. Percentage Variation

among among populations within ΦSC P ΦCT P groups within groups populations 1. host mtDNA clades 0.95 <0.001 -0.37 0.89 -37.2% 130.4% 6.8% 2. host back colour 0.82 <0.001 0.76 <0.05 75.6% 20.2% 4.2%

5.4 DISCUSSION

5.4.1 Population Structure for two species of feather lice The two species of feather lice studied here showed two very distinct patterns of geographic variation. Philopterus sp . showed deep east-west divergence congruent with magpie mtDNA sequence data. No haplotypes were shared between eastern and western populations and the AMOVA based on magpie mtDNA clades estimated a very large

ΦCT (0.75) that was highly significant ( P < 0.05). In contrast, B. semiannulata shows a non-overlapping north south distribution of divergent clades. Haplotypes were shared between east and west populations and the AMOVA based on mtDNA clades was not significant although a north-south partition of variation (host back colour) was

94 Feather lice significant ( ΦCT = 0.76, P < 0.05). Due to the fact that feather lice are obligate parasites and are unable to survive for long away from their host, a concordant pattern of variation was expected with their host. Recent interspecific studies however, suggest this is not always the case and that often lice do not share a strict phylogenetic pattern with the host due to frequent host switching events (Johnson et al. 2002a, Weckstein 2004). Few previous studies have documented the phylogeographic pattern within a host species so it was interesting to demonstrate that Philopterus sp. showed a distinctly concordant pattern in relation to magpie variation.

In fact, Philopterus sp. phylogeographic patterns may provide an insight into very recent history of magpies. Apart from the eastern and western clades found in roughly the same geographic distribution as magpie clades, a third clade was found at the Alice Springs site. In Chapters 3 and 4, no evidence was found to suggest magpies from Alice Springs (As) were differentiated from other eastern sites using magpie control region or microsatellite data. The presence however, of a divergent clade for Philopterus sp. at the Alice Springs site but not for magpies may suggest that the Alice Springs population has been isolated historically for a short period of time but not for sufficient time to show divergence in the host. Nieberding et al. (2004) used the divergence in the phylogeny of the nematode Heligmosomoides polygyrus to argue for historical phylogeographic structuring of its host Apodemus sylvaticus that had not been detected in host genetic variation patterns. However, an alternative explanation is that dispersal of magpies among eastern populations may not necessarily facilitate dispersal of Philopterus sp. over large distances. Dispersal of lice requires direct host to host contact and therefore lice may possibly be limited to dispersal largely within family groups (territories). Furthermore, Philopterus sp. are restricted to the nape of the magpie and therefore contact between hosts at the nape is possibly less frequent than body contact suggesting limited potential for Philopterus sp. to disperse. In a study that compared gene flow among populations of seabirds and their ticks, McCoy et al. (2005) found a greater degree of structuring among populations of the ectoparasite than of its host. It was inferred that dispersal for the parasite was linked to movements during the breeding

95

Feather Lice season, whereas a high amount of dispersal occurred year round for the host. Within the eastern Philopterus sp. clade there appears to be considerable divergence within and among populations (within divergence = 0.7 – 1.6%) suggesting that dispersal of Philopterus sp. may be restricted. If dispersal is limited, divergent lineages may arise within the host population suggesting that magpie hosts do not have to be completely isolated for divergence to evolve in their parasites. To test these hypotheses, population analysis comparing gene flow over a small versus a large scale to estimate dispersal potential of Philopterus sp. was undertaken.

It is difficult to make a direct comparison of substitution rates between louse and host because different mtDNA fragments were analysed here. Nevertheless, by comparing percentage divergence between eastern and western clades Philopterus sp. (7.1%) were at least 4 times more divergent than between eastern and western magpie clades (1.6 %). Considering that the non-coding control region (magpie) is likely to have a higher rate of evolution than protein coding regions, the difference in divergence is actually the opposite of what might be expected due to different mtDNA regions alone. In a study that examined mtDNA cytochrome b variation in bird lice, Page et al. (1998) found that parasitic lice of swiftlets evolved at two to three times the rate of their host. An elevated rate of evolution for lice up to 5.5 times that of the host has also been found for some seabirds (Orders Procellariiformes and Sphenisciformes) and their parasitic lice (Paterson et al. 2000). Possibly the apparent elevated rates of evolution in parasitic lice is due to the lice lineages pre-dating those of the host (ie. evolved on another host that was more divergent). Although this may explain differences in rates of evolution for an interspecific study, it does not explain the difference between magpie and Philopterus sp. mtDNA because the phylogeographic concordance in the data suggest both host and parasite have a shared evolutionary history. More likely the small effective population sizes with limited dispersal potential and recurring founder events has led to greater levels of divergence among populations and the appearance of an elevated mutation rate (Page et al. 1998).

96 Feather lice

While Philopterus sp. showed a congruent pattern of east-west phylogeographic structuring that was congruent with the host, B. semiannulata in contrast was structured north to south. One common haplotype (bs1) within the northern clade (Clade 2) was found in all northern populations distributed from the Kimberley (Km) region across to Rockhampton (Rk) in eastern Australia, is a distance greater than 3000km. This result was surprising as maternal gene flow is absent in the host between north-east and north- west Australia. It is possible that B. semiannulata disperse primarily on male magpies across northern Australia. Microsatellite data suggested that male magpie dispersal did occur between north-east and north-west Australia

Dispersal via male magpies however, does not account for the large divergence (4.2%) observed between north-eastern and south-eastern populations. Moreover, north to south divergence is greater than divergence between Perth (Pe) and south-eastern populations (2 base pairs). Although I have suggested that recent gene flow may have occurred across northern Australia possibly via a range expansion from the east, no evidence exists that populations on either side of the Nullarbor Plain have been in contact for approximately 36, 000 years. In contrast, only a few base pair differences separate B. semiannulata on either side of the Nullarbor Plain while as many as 22 bp differences were identified between non-overlapping clades in the north and south for this species.

Two possible scenarios may account for this striking pattern. Firstly, the lack of differentiation over such wide host ranges may result from transfer among multiple hosts facilitated by phoresis. If B. semiannulata individuals were using multiple hosts that largely overlapped with magpie distribution, it would be expected that clades would be mixed over the entire distribution of the host rather than showing the distinct geographical pattern evident here. Another possibility is that this north-south pattern may be accounted for by selection. Selection of the mitochondrial genome due to thermal tolerance would possibly result in a north to south pattern of divergence. Mitochondria are the powerhouse of the cell and are responsible for cellular respiration. Cellular respiration occurs with strict thermal limits, outside of which respiration breaks

97

Feather Lice down (Portner 2002). As a result, organisms may be geographically restricted, not because of extrinsic barriers to dispersal but due to selection against dispersal. Several studies (Klock et al. 2004, Hicks and McMahon 2005) have demonstrated limits to thermal tolerance possibly due to cellular respiration for aquatic and terrestrial organisms. The magpie is distributed over a large latitudinal scale and as such ectoparasites of the magpie may experience a very wide range of temperatures. Temperatures that parasitic lice experience could also be exaggerated further by individual plumage colour of the host they infest. As BB magpies are broadly distributed over northern Australia where it is warmest, the black plumage colour may intensify an already warm habitat. Dark coloured feathers absorb more visible light than white feathers (Lustick et al. 1980) thus increasing the temperature of the habitat of B. semiannulata that live on BB magpies. White feathers (WB form) however, may reflect heat away from lice habitat in southern populations of B. semiannulata . The AMOVA based on host plumage colour showed a significant ΦCT (0.76, P < 0.05) suggesting that host plumage colour rather than host history may explain the patterns of genetic variation in B. semiannulata . Further analysis of mtDNA in conjunction with nuclear sequence analysis will shed light on whether nuclear gene flow occurs between northern and southern clades or if they are reproductively isolated. To test for selection, cage experiments for thermal tolerance of B. semiannulata could provide an insight into the factors responsible for the north-south divergence of lineages. Previous research on Brueelia spp. suggests that short-term survival away from the host is possible (Dumbacher 1999). Therefore, cage experiments would entail removing live lice from magpies and treating them to different temperature ranges.

5.4.2 Secondary Contact The position of the break between eastern and western lineages in Philopterus sp. differed slightly from their magpie host. Sequence data suggested that Philopterus sp. collected from Cape Crawford (Cc) belonged to a clade restricted to north-western populations, while magpies from the same site belonged to the eastern mtDNA magpie clade. Although no mixing among host mtDNA clades was found, gene flow from east

98 Feather lice to north-west was inferred from microsatellite data. Presence of western clade Philopterus sp. on eastern clade magpie individuals provides additional support for recent secondary contact among northern magpie populations.

5.4.3 Divergent Lineages Divergent lineages of both species of lice, were found to overlap geographically with widespread clades. Two divergent lineages of B. semiannulata (Clade 3 and 4) were found in the following magpie populations: Brisbane (Br), Rockhampton (Rk), Moranbah (Mb) and Mataranka (Mt). Between 7 and 14 amino acid changes were found between the common widespread lineages (Clades 1 and 2) and Clades 3 and 4 whereas only a single amino acid change differentiated Clades 1 and 2. The high number of amino acid substitutions between the common widespread clades and the divergent clades suggests that these lineages have not formed simply via genetic drift in an isolated host population. Two hypotheses may account for the divergent lineages. Firstly, the presence of Numts can not be excluded as a possible origin for the highly differentiated sequences. However, there was no indication of double sequence peaks, stop codons or frameshift changes in the sequences suggesting that the fragment sequenced was in fact coding DNA.

The other possible explanation for divergent lice lineages is as a result of host transfer or host switching. Host transfer refers to the dispersal of lice between different hosts whereas host switching is the permanent movement of a parasitic species to a new host (Clayton and Johnson 2003). Two Philopterus sp. individuals (ps20) from a single magpie population (Kimberley1) were closer genetically to the pied butcherbird louse outgroup (3.6 %) than to any other Philopterus sp. (7.4 -9.1%) found on magpies suggesting that they may have originated on another avian host, possibly pied butcherbirds. Rates of host transfer may differ among species because habitat specificity of some feather lice varies. In a study that compared two genera of feather lice that infested pigeons and doves, wing lice were shown to be less host specific and have a higher incidence of host transfer than were highly host specific body lice that showed

99

Feather Lice evidence of cospeciation with their host (Clayton et al. 2003b). Thus relative dispersal potential may be linked to habitat specificity of lice species. This may explain why the occurrence of divergent lineages was more common in B. semiannulata than for Philopterus sp. Philopterus sp . inhabits only nape feathers and in relative terms is more habitat specific than with B. semiannulata which are found all over body feathers of magpies. Even if habitat specificity can account for this pattern, a major question that needs attention is to determine how lice physically move between different host species?

Lice are wingless insects (apterous) that require a host for providing food and microhabitat requirements, they are therefore unable to disperse directly to atypical hosts. Dispersal may be possible however, via phoresis. Phoresis is the relationship that some species of lice have with more mobile organisms that allow lice to ‘hitch a ride’, thereby mediating dispersal among hosts (Gullan and Cranston 1994). Hippoboscid flies are known to have a phoretic relationship with some species of feather lice (Clayton et al. 2003a). Keirans (1975) reported that Brueelia spp. accounted for over 80% of all lice transported by hippoboscid flies. Johnson et al. (2002a) found that lice of the genus Brueelia did not necessarily reflect host phylogeny and in some cases louse species were shared between hosts suggesting a high potential for gene flow. During the course of this study hippoboscid flies were commonly observed on captured wild magpies. Brueelia spp. have been observed on hippoboscid flies collected from dead magpies (Hughes 1980). B. semiannulata Clades 3 and 4 could be the result of dispersal between magpie hosts and another host and thus may be evidence for multiple host switching events in magpie lice.

In conclusion, the two species of obligate feather lice of the Australian magpie examined here showed very different phylogeographic patterns. Philopterus sp. are strongly congruent with their magpie host. In contrast, phylogeographic patterns in B. semiannulata are much more complex and are possibly influenced by ecology of the host (selection for different back colour) rather than reflecting the recent history of the

100 Feather lice host. Furthermore, genetic diversity and lineage divergence of magpie feather lice may be affected by host transfer and/or host switching events.

101

General discussion

CHAPTER 6: GENERAL DISCUSSION

To understand the processes that are important in shaping a species distribution it is important to consider not only modern environmental factors but also the evolutionary history of the organism.

6.1 A RECONSTRUCTION OF THE RECENT EVOLUTIONARY HISTORY OF THE

AUSTRALIAN MAGPIE Plumage variation (dorsal colour pattern) of the Australian magpie is distributed essentially north to south (Figure 6.1b). Previous research showed that BB and WB plumage form distributions can not be explained by patterns mtDNA variation (Hughes et al. 2001, Toon et al. 2003). In this current study different geographic patterns of genetic variation in the Australian magpie were associated with varying levels of concordance among markers. The mtDNA magpie phylogeny indicates two mainland discrete clades of magpies (Figure 6.1c) and no evidence of secondary contact suggesting historical isolation. Analysis of microsatellite variation suggested that there are six groups of modern magpie populations (Figure 6.1d) among which gene flow was evident. In addition nuclear DNA data suggests recent magpie expansion from east to west via a northern route. The genetic variation assayed for the two species of parasitic feather lice showed divergent patterns. Philopterus sp. mtDNA variation was mostly concordant with magpie mtDNA variation (6.1e); however a greater depth of structure was evident in the lice phylogenies. The distribution of variation for B. semiannulata did not accord with that of the host (6.1f). Sites where divergent northern and southern B. semiannulata clades were found in the east were separated by less than 650 km. These sites showed little difference for mtDNA or nDNA in magpies however, suggesting there were no impediments to gene flow for the host at least.

102 General discussion

A B

C D

? ?

E F ? ?

Figure 6.1: Distribution of variation in the Australian magpie and two species of obligate feather lice. A. Birds Australia Atlas records for 1998-2002. B. Distribution of plumage variation and inferred relationships among plumage forms based on Schodde and Mason (1999). C. mtDNA control region haplotype variation for magpies shows an east-west divergence. D. magpie nuclear variation for six microsatellite loci shows six groups of populations, however most eastern, northern and north-western populations form one cluster. E. Four clades were detected for Philopterus sp. All eastern clades grouped together and were very divergent from the north-west clade. F. Two clades were reported for B. semiannulata . They were found non-overlapping but in close proximity and were distributed north-south. Question marks show areas sampled for magpies but not lice. Trees are not drawn to scale.

103

General discussion

The results of the current study suggest a complex recent evolutionary history for the Australian magpie that included Pleistocene isolation of populations and subsequent range expansion to form the current contiguous distributions around Australia. A model that describes the potential history of fragmentation and expansion of the magpie population during the Pleistocene is outlined in Figure 6.2. The mtDNA phylogeny indicated that the two mainland clades of magpies were monophyletic suggesting that a once widespread ancestral population was divided in the past into eastern and western populations. Coalescent analysis ( IM ) suggests that only a small proportion of the hypothetical ancestral population contributed to modern western magpie populations. This argues for colonisation of Western Australia by a subset of the ancestral population (founder event) rather than resulting from population fragmentation. However, fragmentation of an ancestral population that was distributed in a ‘stepping stone’ fashion could equally explain the data. The model of magpie evolution presented above is congruent with predicted past climatic events and putative biogeographical regions suggesting that past isolation of magpie populations was due to vicariance as a result of formation of climate induced physical barriers (aridity) rather than a long distance dispersal event. The mtDNA data indicate that independent eastern and western magpie populations were most likely formed during the last period of aridity and have been isolated for approximately 36, 000 years ago (Chapter 4). Climates during the Tertiary were considerably warmer and wetter than those present today (Kershaw et al. 1994). Moreover, investigations of sand dune formation suggest that towards the end of the Pleistocene, periods of aridity increased in severity (Bowler 1976). Earlier peaks of aridity could have influenced genetic structure of less vagile species than magpies (see Chapter 4). There was no evidence however, of any mtDNA haplotypes older than the most recent glacial/arid period for the Australian magpie suggesting that only a single ancestral population was present prior to the last period of aridity.

104 General discussion

Time 1: Inferred ancestral magpie population distributed widely across Australian mainland.

Time 2: At the peak of aridity during the last dry, cool period magpie populations were fragmented into two populations. At this time the eastern population was connected with the Tasmanian population.

Time 3: The Australian magpie is contiguously distributed through recent range expansion.

Inferred female expansion

Inferred male expansion

Figure 6.2: Scenario explaining population divergence and range expansion in the Australian magpie. Distributions at time one and two are hypothetical based on putative barriers (Keast 1961, Ford 1987) to dispersal formed through the increase in aridity throughout central Australia.

Although Australia’s climate is presently in an interglacial phase, much of inland and northern Australia today is covered in arid or semi-arid vegetation (Jones and Bowler 1980). Since their formation, arid barriers to dispersal such as the Nullarbor or Carpentarian barriers have not been completely eroded. It is apparent from Birds Australia Atlas records (see Figure 6.1a) that magpies have most likely re-colonised parts of Australia to form a now contiguous distribution. However, genetic data presented here suggests that contact between eastern and western magpies is of relatively recent origin. At present there is no evidence of mixing of populations found 105

General discussion south-east and south-west of the Nullarbor Plain. In addition there has apparently not been mixing of mtDNA clades among populations sampled in northern Australia. Furthermore, only a single population in northern Australia (Cape Crawford) contains louse clades that apparently have been transferred between eastern and western lineages. There was however, evidence for nuclear gene flow between eastern populations and the Kimberley and Pilbara in the north-west. I accounted for this difference through male biased dispersal of magpies. Moreover, it supports the inference of recent range expansion via a northern route. Range expansion was indicated for the eastern region in the Nested Clade Analysis (NCA). Although tests were not significant, there was evidence for population growth in north-eastern and south-eastern populations indicating movement from high density regions in eastern Australia towards low density regions in central Australia. The coalescent analysis also suggested an increase from an estimated ancestral population size of approximately 80, 000 – 200, 000 to a current estimated eastern population size of 900, 000 – 1, 500, 000.

Timing of contact between east and west populations in the north is of interest because there could be potential for human facilitation. Magpies may have expanded across the northern extent of their range, at low densities during the Holocene (last 10, 000 years) as climates became more favourable. However, a search of early orthnithological literature (Whitlock 1909, Hill 1913, Campbell 1929) reveals few records of Australian magpies in northern Australia (Barclay Tablelands, north-east of Alice Springs) and north-western Australia (Kimberley region) even when actively searching for them. Early surveys suggest that a break was present in the magpie distribution between the Barclay Tablelands in the Northern Territory across to the Pilbara. Thus the possibility exists that magpies may have extended their range only in modern times (post-European colonisation) and certainly densities have increased since. Evidence of a recent population expansion suggested from genetic data supports the idea that barn swallows H. rustica have increased in population size in Europe and Asia due to a close association with human settlement (Zink et al. 2006). The association of magpies with human settlement may explain the recent range expansion. Magpies are generalist

106 General discussion species and are very successful at establishing populations within highly disturbed habitats such as urban and agricultural land (Luck et al. 1999). However, successful establishment of populations is limited by access to water (Fisher et al. 1972). Therefore, the pattern of recent range expansion from eastern across to north-western Australia could have been facilitated by the establishment of pastoral land and homesteads with associated permanent water for stock across the north of Australia. Anecdotal evidence supports the idea that magpies may have followed the settling of pastoral land (Carter 1924, Campbell 1929).

6.2 SIGNIFICANCE OF BACK COLOUR AND LATITUDINAL PLUMAGE VARIATION The main findings from this study support the suggestion from previous work that plumage variation is most likely uncoupled from neutral genetic variation in Australian magpies. The mtDNA data show clearly that the pattern of east-west divergence was discordant with the latitudinal divide in dorsal colour (BB, WB) shown in Figure 6.1b. In an analysis of the south eastern contact zone, Hughes et al. (2001) found no evidence that the distribution of back colour plumage forms could be explained by a recent vicariant history. Instead the authors favoured a theory that variation arose in situ and has been maintained via natural selection. They suggested that sexual selection for brightly coloured males was possibly balanced by natural selection for a dark less conspicuous back colour in open areas that dominate large areas of northern Australia. In a subsequent study, Toon et al. (2003) found a similar lack of mtDNA evidence that the western contact zone was formed through secondary contact of diverged plumage forms. This amounts to two independent lines of evidence supporting discordance between neutral genetic variation and distinct plumage forms in magpies. Results in the present study show that neutral genetic variation representing the mtDNA clades is also of very recent origin. Therefore, it is difficult to infer how plumage forms first arose as they possibly predate the east-west divergence. Even so, plumage forms are currently being maintained in their present distribution even where range expansion and high gene flow is possible among populations (migration estimates). An alternative explanation for the origin of the two plumage forms is that a Torressian (BB) form and a

107

General discussion

Bassian (WB) form arose in isolation prior to the late Pleistocene. This explanation requires that populations came into secondary contact, and subsequently, evidence of divergence at neutral loci has been eroded through contraction of populations during the Pleistocene. B. semiannulata showed a deep divergence (4.2%) between northern and southern populations, in contrast only 2 base pairs differentiated south-eastern and south-western populations. In Chapter 5, I suggested this contrasting pattern in the lice to be evidence for the effects of selection. However, such divergent lineages in B. semiannulata possibly evolved on isolated northern and southern populations of magpies that have since come back into contact and the divergence is currently maintained through selection. Further analysis is required to test whether gene flow occurs between the divergent lineages of B. semiannulata .

Apart from the obvious north-south divide in magpie plumage forms, body size and bill size in Australian magpies also vary on a latitudinal gradient. Mainland populations follow Bergman’s rule with the smallest body sizes reported for northern populations and largest body sizes found in white back populations in south-eastern Australia. Interestingly, the Tasmanian magpie is significantly smaller (Schodde and Mason 1999) than both WB and BB mainland forms a pattern which contrasts with what is expected under Bergman’s rule (ie. populations at higher latitudes, cold climates are expected to have a larger body size). Moreover, the Tasmanian population does not follow the ‘island rule’, in which passerines (small body birds) are expected to have a larger body size on islands than their mainland counterparts (Clegg and Owens 2002, Scott et al. 2003). Clegg and Owens (2002) have suggested that a shift towards medium body size may be a shift towards generalist foraging behaviour in the absence of interspecific competition. All magpies are generalist feeders and may be considered as medium sized birds and as such isolation on an island associated with a reduction in interspecific competition may provide little extra benefit for a species already adapted to a wide feeding niche. The apparent shift in magpies towards a smaller body size in Tasmania remains however a complex question that can only be fully understood with analysis of feeding and other ecological data for multiple island populations.

108 General discussion

6.3 THE PLEISTOCENE AND INTRASPECIFIC GENETIC VARIATION OF AVES In general, studies of phylogeographic variation have suggested that climate fluctuations during the Pleistocene have had a significant impact on contemporary distributions of many bird species. Two main patterns are evident among northern hemisphere species distributed in previously glaciated areas. 1. Geographic structuring of mtDNA variation that is concordant with Pleistocene refugia (see Avise and Walker 1998) and 2. Evidence of range expansion/ population expansion since glaciers receded (see Qu et al. 2005). The current study of genetic variation in Australian magpies suggests that increasing aridity in low latitudinal areas may also have had an impact on geographic variation of a widespread passerine in Australia. East-west fragmentation of magpie populations during the last period of aridity was indicated by the data. It is possible however, that European settlement and recent habitat change may have facilitated range expansion of the Australian magpie to produce its modern contiguous distribution. Empirical research on other wide spread Australian species is required to see if this pattern may be repeated in other native species.

109

References

REFERENCES

Andersen NM, Cheng L, Damgaard J and Sperling FAH (2000) Mitochondrial DNA sequence variation and phylogeography of oceanic insects (Hemiptera: Gerridae: Halobates spp.). Marine Biology, 136, 421-430.

Arbogast BS, Edwards SV, Wakeley J, Beerli P and Slowinski JB (2002) Estimating divergence times from molecular data on phylogenetic and population genetic timescales. Annual Review of Ecological Systematics, 33, 707-740.

Avise JC (1983) Commentary. In Brush AH and Clark Jr. GA (Eds.) Perspectives in Ornithology. New York, Cambridge University Press.

Avise JC, Arnold J, Ball RM, Bermingham E, Lamb T, Neigel JE, Reeb CA and Saunders NC (1987) Intraspecific Phylogeography: The mitochondrial DNA bridge between population genetics and systematics. Annual Review of Ecology and Systematics, 18, 489-522.

Avise JC and Walker D (1998) Pleistocene phylogeographic effects on avian populations and the speciation process. Proceedings of the Royal Society B: Biological Sciences, 265, 457-463.

Baker AJ and Marshall HD (1997) Mitochondrial control region sequences as tools for understanding evolution. In Mindell DP (Ed.) Avian molecular evolution and systematics. San Diego, Academic Press.

Baker AM (1999) Dispersal patterns and group size variation among mainland populations of the Australian magpie. Queensland, Queensland University of Technology.

Baker AM, Mather PB and Hughes JM (2000) Population genetic structure of Australian magpies: evidence for regional differences in juvenile dispersal behaviour. Heredity, 85, 167-176.

Baker AM, Mather PB and Hughes JM (2001) Evidence for long-distance dispersal in a sedentary passerine, Gymnorhina tibicen (Artamidae). Biological Journal of the Linnean Society 72, 333-343.

Balloux F, Brünner H, Lugon-Moulin N, Hausser J and Goudet J (2000) Microsatellites can be misleading: an empirical and simulation study. Evolution, 54, 1414-1422.

Barker FK, Cibois A, Schikler P, Feinstein J and Cracraft J (2004) Phylogeny and diversification of the largest avian radiation. Proceedings of the National Academy of Sciences, 101, 11040-11045.

Barlow BA (1981) The Australian flora: its origins and evolution. Flora of Australia. Canberra, Australian Government Publishing Service.

Barrowclough GF (1983) Biochemical studies of microevolutionary processes. In Bush AH and Clark Jr. GA (Eds.) Perspectives in ornithology. New York, Cambridge University Press.

Barrowclough GF, Groth JG, Mertz LA and Gutterrez RJ (2004) Phylogeographic

110 References structure, gene flow and species status in blue grouse ( Dendragapus obscurus ). Molecular Ecology, 13, 1911-1922.

Beerli P and Felsenstein J (1999) Maximum-likelihood estimation of migration rates and effective population numbers in two populations using a coalescent approach. Genetics, 152, 763-773.

Beerli P and Felsenstein J (2001) Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences, 98, 4563-4568.

Bensasson D, Zhang D-X, Hartl D and GM H (2001) Mitochondrial pseudogenes: evolution's misplaced witnesses. Trends in Ecology & Evolution, 16, 314-321.

Birky JCW, Fuerst P and Maruyama T (1989) Organelle gene diversity under migration, mutation, and drift: equilibrium expectations, approach to equilibrium, effects of heteroplasmic cells, and comparison to nuclear genes. Genetics, 121, 613-627.

Black A (1986) The Taxonomic Affinity of the New Guinean Magpie Gymnorhina- tibicen -papuana . Emu, 86, 65-70.

Bowie RCK, Fjeldsa J, Hackett SJ, Bates JM and Crowe TM (2006) Coalescent models reveal the relative roles of ancestral polymorphism, vicariance, and dispersal in shaping phylogeographical structure of an African montane forest robin. Molecular Phylogenetics and Evolution, 38, 171-188.

Bowler JM (1976) Aridity in Australia: Age, origins and expression in Aeolian landforms and sediments. Earth Science Reviews, 12, 279-310.

Burton TC and Martin AA (1976) Analysis of hybridization between black-backed and white-backed magpies in south-eastern Australia. EMU, 76, 30-36.

Campbell AG (1929) Australian magpies of the genus Gymnorhina . The Emu, 28, 165- 176.

Carter (1924) Birds of Broome Hill. The Emu, 23, 317-?

Chappell J (1983) A revised sea-level record for the last 300,000 years from Papua New Guinea. Search, 14, 99-101.

Chappell J and Shackleton NJ (1986) Oxygen isotopes and sea level. Nature, 324, 137- 140.

Clay T (1951) An introduction to a classification of the avian Ischnocera (Mallophaga): Part I. Transactions of the Royal Entomological Society of London, 102, 171-194.

Clayton DH, Al-Tamimi S and Johnson KP (2003a) The ecological basis of coevolutionary history. In Page RDM (Ed.) Tangled Trees. Chicago, The University of Chicago Press.

Clayton DH, Bush SE, Goates BM and Johnson KP (2003b) Host defense reinforces host-parasite cospeciation. Proceedings of the National Academy of Sciences, 100, 15694-15699.

111

References

Clayton DH and Johnson KP (2003) Linking Coevolutionary History to Ecological Process: Doves and Lice. Evolution, 57, 2335-2342.

Clegg SM and Owens IPF (2002) The 'island rule' in birds: medium body size and its ecological explanation. Proceedings of the Royal Society of London Series B-Biological Sciences, 269, 1359-1365.

Clement M, Posada D and Crandall KA (2000) TCS: a computer program to estimate gene genealogies. Molecular Ecology, 9, 1657-1659.

Cooper SJB, Adams M and Labrinidis A (2000) Phylogeography of the Australian dunnart Sminthopsis crassicaudata (Marsupialia: Dasyuridae). Australian Journal of Zoology, 48, 461-473.

Corander J and Marttinen P (2006) Bayesian identification of admixture events using multilocus molecular markers. Molecular Ecology, 15, 2833-2843.

Corander J, Marttinen P and Mäntyniemi S (2006) A Bayesian method for identification of stock mixtures from molecular marker data. Fishery Bulletin, 104, 550-558.

Corander J, Waldmann P, Marttinen P and Sillanpää MJ (2004) BAPS 2: enhanced possibilities for the analysis of genetic population structure. Bioinformatics, 20, 2363- 2369.

Corander J, Waldmann P and Sillanpää MJ (2003) Bayesian analysis of genetic differentiation between populations. Genetics, 163, 367-374.

Covacin C, Shao R, Cameron S and Barker SC (2006) Extraordinary number of gene rearrangements in the mitochondrial genomes of lice (Phthiraptera: Insecta). Insect Molecular Biology, 15, 63-68.

Cracraft J (1986) Origin and evolution of continental biotas: speciation and historical congruence within the Australian avifauna. Evolution, 40, 977-996.

Cracraft J and Prum RO (1988) Patterns and processes of diversification: speciation and historical congruence in some neotropical birds. Evolution, 42, 603-620.

Crandall KA (1996) Multiple interspecies transmissions of human and simian t-cell leukemia/lymphoma virus type I sequences. Molecular Biological Evolution, 13, 115- 131.

Crochet PA (2000) Genetic structure of avian populations - allozymes revisited. Molecular Ecology, 9, 1463-1469.

Crochet PA, Chen JZ, Pons JM, Lebreton JD, Hebert PDN and Bonhomme F (2003) Genetic Differentiation at nuclear and mitochondrial loci among large white-headed gulls: sex-biased interspecific gene flow? Evolution, 57 .

Degnan SM and Moritz C (1992) Phylogeography of mitochondrial DNA in two species of white-eyes in Australia. The Auk, 109, 800-811.

Doyle JJ and Doyle JL (1987) A rapid DNA isolation procedure from small quantities of fresh leaf tissues. Phytochemistry Bulletin, 19, 11-15.

112 References

Driskell AC, Pruett-Jones S, Tarvin KA and Hagevik S (2002) Evolutionary relationships among blue- and black-plumaged populations of the white-winged fairy- wren ( Malurus leucopterus ). Australian Journal of Zoology, 50, 581-595.

Dumbacher JP (1999) Evolution of toxixity in pitohuis: I. Effects of Homobactrachotoxin on chewing lice (order Phthiraptera) The Auk, 116, 957-963.

Durrant KL and Hughes JM (2005) Differing rates of extra-group paternity between two populations of the Australian magpie ( Gymnorhina tibicen ). Behavioral Ecology and Sociobiology, 57, 536-545.

Eastwood R, Pierce NE, Kitching RL and Hughes JM (2006) Do ants enhance diversification in lycaenid butterflies? Phylogeographic evidence from a model myrmecophile, Jalmenus evagoras. Evolution, 60, 315-327.

Edwards SV (1993) Long-distance gene flow in a cooperative breeder detected in genealogies of mitochondrial DNA sequences. Proceedings of the Royal Society B: Biological Sciences, 252, 177-185.

Edwards SV (1997) Relevance of microevolutionary processes to higher level molecular systematics. In Mindell DP (Ed.) Avian Molecular evolution and systematics. San Diego, Academic Press.

Edwards SV and Beerli P (2000) Perspective: Gene divergence, population divergence, and the variance in coalescence time in phylogeographic studies. Evolution, 54, 1839- 1854.

Edwards SV and Kot M (1995) Comparative methods at the species level: Geographic variation in morphology and group size in grey-crowned babblers ( Pomatostomus temporalis ). Evolution, 49, 1134-1146.

Edwards SV and Wilson AC (1990) Phylogenetically informative length polymorphism and sequence variability in mitochondrial DNA of Australian songbirds (Pomatostomus ). Genetics, 126, 695-711.

Eldridge MDB, Kinnear JE, Zenger KR, McKenzie LM and Spencer PBS (2004) Genetic diversity in remnant mainland and "pristine" island populations of three endemic Australian macropodids (Marsupialia): Macropus eugenii , Lagorchestes hirsutus and Petrogale lateralis . Conservation Genetics, 5, 325-338.

Ellegren H (2004) Microsatellites: Simple sequences with complex evolution. Nature Reviews Genetics, 5, 435-445.

Estoup A, Jarne P and Cornuet J-M (2002) Homoplasy and mutation model at microsatellite loci and their consequences for population genetics analysis. Molecular Ecology, 11, 1591-1604.

Evans SR and Middleton JF (1998) A Regional Model of Shelf Circulation near Bass Strait: A New Upwelling Mechanism. Journal of Physical Oceanography, 28, 1439- 1457.

Excoffier L, Laval F and Schneider S (2005) Arlequin ver 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online, 1,

113

References

47-50.

Excoffier L, Smouse PE and Quattro JM (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics, 131, 479-491.

Farabaugh SM, Brown ED and Hughes JM (1992) Cooperative territorial defence in the Australian magpie. Gymnorhina tibicen (Passeriformes, Cracticidae), a group-living songbird. Ethology, 92, 283-292.

Felsenstein J (1992) Estimating effective population-size from samples of sequences - inefficiency of pairwise and segregating sites as compared to phylogenetic estimates. Genetical research, 59, 139-147.

Fisher CD, Lindgren E and Dawson WR (1972) Drinking patterns and behavior of Austalian desert birds in relation to their ecology and abundance. The Condor, 74, 111- 136.

Fleming K, Johnston P, Zwartz D, Yokoyama Y, Lambeck K and Chappell J (1998) Refining the eustatic sea-level curve since the Last Glacial Maximum using far- and intermediate-field sites. Earth and Planetary Science Letters, 163, 327-342.

Ford J (1987) Hybrid zones in Australian birds. Emu, 87, 158-178.

Freeland JR and Boag PT (1999) Phylogenetics of darwin's finches: paraphyly in the tree-finches, and two divergent lineages in the warbler finch. The Auk, 116, 577-588.

Fu Y-X (1996) New statistical tests for neutrality for DNA samples from a population. Genetics, 143, 557-570.

Fu Y-X (1997) Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics, 147, 915-925.

Fu Y-X and Li W-H (1999) Coalescing into the 21st century: An overview and prospects of coalescent theory. Theoretical Population Biology, 56, 1-10.

Goudet J (1995) Fstat version 1.2: a computer program to calculate Fstatistics. Journal of Heredity, 86, 485-486.

Greenberg R, Cordero PJ, Droege S and Fleischer RC (1998) Morphological adaptation with no mitochondrial DNA differentiation in the coastal plain swamp sparrow. The Auk, 115, 706-712.

Greenwood DR (1994) Palaeobotanical evidence for Tertiary climates. In Hill RS (Ed.) History of the Australian vegetation. Cambridge, Cambridge Universtiy Press.

Gullan PJ and Cranston PS (1994) The insects: an outline of entomology, London, Chapman and Hall.

Hafner MS, Sudman PD, Villablanca FX, Spradling TA, Demastes JW and Nadler SA (1994) Disparate rates of molecular evolution in cospeciating hosts and parasites. Science, 265, 1087-1090.

Hansen MM, Bekkevold D, Jensen LF, Mensberg K-LD and Nielsen EE (2006) Genetic

114 References restoration of a stocked brown trout Salmo trutta population using microsatellite DNA analysis of historical and contemporary samples. Journal of Applied Ecology, 43, 669- 679.

Harpending HC, Sherry ST, Rogers AR and Stoneking M (1993) The Genetic-Structure of Ancient Human-Populations. Current Anthropology, 34, 483-496.

Hedrick W, Philip (1999) Highly variable loci and their interpretation in evolution and conservation. Evolution, 53, 313.

Heilveil JS and Berlocher SH (2006) Phylogeography of postglacial range expansion in Nigronia serricornis Say (Megaloptera: Corydalidae). Molecular Ecology, 15, 1627- 1641.

Hey J (2005) On the number of new world founders: A populations genetic portrait of the peopling of the Americas. Public Library of Science, 3, 1-11.

Hey J and Nielsen R (2004) Multilocus methods for estimating population sizes, migration rates and divergence time, with applications to the divergence of Drosopila pseudoobscura and D. persimilis . Genetics, 167, 747-760.

Hicks DW and McMahon RF (2005) Effects of temperature on chronic hypoxia tolerance in the non-indigenous brown mussel, Perna perna (Bivalvia : Mytilidae) from the Texas Gulf of Mexico. Journal of Molluscan Studies, 71, 401-408.

Hill GF (1913) Ornithological notes, Barclay expedition. Emu, 12, 238-261.

Ho SYW and Larson G (2006) Molecular clocks: when times are a-changin'. Trends in Genetics, 22, 79-83.

Ho SYW, Phillips MJ, Cooper A and Drummond AJ (2005) Time Dependency of Molecular Rate Estimates and Systematic Overestimation of Recent Divergence Times. Molecular Biology and Evolution, 22, 1561-1568.

Holder K, Montgomerie R and Friesen VL (1999) A test of the glacial refugium hypothesis using patterns of mitochondrial and nuclear DNA sequence variation in rock ptarmigan ( Lagopus mutus ). Evolution, 53, 1936-1950.

Hope GS (1994) Quaternary vegetation. In Hill RS (Ed.) History of vegetation of Australia. Cambridge, Cambridge University Press.

Hughes J, Baker AM, Bartlett C, Bunn S, Goudkamp K and Somerville J (2004) Past and present patterns of connectivity among populations of four cryptic species of freshwater mussels Velesunio spp. (Hyriidae) in central Australia. Molecular Ecology, 13, 3197-3212.

Hughes JM (1980) Geographic variation in the Australian magpie and its mallophagan parasites. Melbourne, La Trobe University.

Hughes JM (1982) An explanation for the asymmetrical 'hybrid' zone between white- backed and black-backed magpies. EMU, 82, 50-53.

Hughes JM (1984a) Distribution of Mallophaga on the Australian magpie ( Gymnorhina

115

References tibicen Latham) (Family: Cracticidae). The Australian Zoologist, 21, 459-465.

Hughes JM (1984b) Morphometric variation in the Mallophaga of the Australian magpie ( Gymnorhina tibicen Latham). The Australian Zoologist, 21, 467-477.

Hughes JM, Baker AM, de Zylva G and Mather PB (2001) A phylogeographic analysis of southern and eastern populations of the Australian magpie: evidence for selection in maintenance of the distribution of two plumage morphs. Biological Journal of the Linnean Society, 74, 25-34.

Hughes JM and Mather PB (1991) Variation in the size of territorial groups in the Australian magpie, Gymnorhina tibicen . Proceedings of the Royal Society of Queensland, 101, 13-19.

Hughes JM, Mather PB, Toon A, Ma J, Rowley I and Russell E (2003) High levels of extra-group paternity in a population of Australian magpies Gymnorhina tibicen : evidence from microsatellite analysis. Molecular Ecology, 12, 3441-3450.

Jarne P and Lagoda PJL (1996) Microsatellites, from molecules to populations and back. Trends in Ecology and Evolution, 11, 424-429.

Jennings WB and Edwards SV (2005) Speciational history of Australian grass finches (Poephila) inferred from thirty gene trees. Evolution, 59, 2033-2047.

Johnson KP, Adama RJ and Clayton DH (2002a) The phylogeny of the louse genus Brueelia does not reflect host phylogeny. Biological Journal of the Linnean Society, 77, 233-247.

Johnson KP, Williams BL, Drown DM, Adams RJ and Clayton DH (2002b) The population genetics of host specificity: genetic differentiation in dove lice (Insecta: Phthiraptera). Molecular Ecology, 11, 25-38.

Johnson NK and Cicero C (2004) New mitochondrial DNA data affirm the importance of Pleistocene speciation in North American birds. Evolution, 58, 1122-1130.

Jones R and Bowler J (1980) Struggle for the savanna: northern Australia in ecological and prehistoric perspective. In Jones R (Ed.) Northern Australia: options and implications. Canberra, SOCPAC.

Joseph L, Moritz C and Hugall A (1995) Molecular support for vicariance as a source of diversity in rain-forest. Proceedings of the Royal Society of London Series B-Biological Sciences, 260, 177-182.

Joseph L and Wilke T (2006) Molecular resolution of population history, systematics and historical biogeography of the Australian ringneck parrots Barnardius: are we there yet? EMU.

Keast A (1961) Bird speciation on the Australian continent. Bulletin of the Museum of Comparative Zoology, 123, 307-495.

Keirans J (1975) A review of the phoretic relationsip between Mallophaga (Phthiraptera: Insecta) and Hippoboscidae (Diptera: Insecta). Journal of Medical Entomology, 12, 71-76.

116 References

Kershaw AP, Martin HA and McEwen Mason JRC (1994) The Neogene: A period of transistion. In Hill RS (Ed.) History of the Australian vegetation. Cambridge, University Press.

Kershaw P, Moss P and Van Der Kaars S (2003) Causes and consequences of long-term climatic variability on the Australian continent. Freshwater Biology, 48, 1274-1283.

King JP, Kimmel M and Chakraborty R (2000) A Power Analysis of Microsatellite- Based Statistics for Inferring Past Population Growth. Molecular Biology and Evolution, 17, 1859-1868.

Kingman JFC (1982a) The Coalescent. Stochastic Processes and their Applications, 13, 235-248.

Kingman JFC (1982b) On the genealogy of large populations. Journal of Applied Probability, 19A, 27-43.

Kishino H and Hasegawa M (1989) Evaluation of the Maximum-Likelihood Estimate of the Evolutionary Tree Topologies from DNA-Sequence Data, and the Branching Order in Hominoidea. Journal of Molecular Evolution, 29, 170-179.

Klassen GJ (1992) Coevolution - a History of the Macroevolutionary Approach to Studying Host-Parasite Associations. Journal of Parasitology, 78, 573-587.

Klicka J and Zink RM (1997) The importance of recent ice ages in speciation: A failed paradigm. Science, 277, 1666-1669.

Klock CJ, Sinclair BJ and Chown SL (2004) Upper thermal tolerance and oxygen limitation in terrestrial arthropods. The Journal of Experimental Biology, 207, 2361- 2370.

Knowles LL and Maddison WP (2002) Statistical phylogeography. Molecular Ecology, 11, 2623-2635.

Kuhner M, Yamato J, Beerli P, Smith L, Rynes E, Walkup E, Li C, Sloan J, Colacurcio P and Felsenstein J (2005) LAMARC. v 2.0. ed., University of Washington.

Kuhner MK, Yamato J and Felsenstein J (1998) Maximum likelihood estimation of population growth rates based on the coalescent. Genetics, 149, 429-434.

Lanyon SM and Hall JG (1994) Reexamination of barbet monophyly using mitonchondiral-DNA sequence data. The Auk, 111, 389-397.

Lessa EP and Applebaum G (1993) Screening techniques for detecting allelic variation in DNA-sequences. Molecular Ecology, 2, 119-129.

Luck GW, Possingham HP and Paton DC (1999) Bird responses at inherent and induced edges in the Murray Mallee, South Australia. 1. Differences in abundance and diversity. Emu, 99, 157-169.

Lustick S, Adam M and Hinko A (1980) Interaction between posture, color, and the radiative heat load in birds. Science, 208, 1052-1053.

Marshall AG (1981) The ecology of ectoparasite insects, London, Academic Press. 117

References

McCoy KD, Boulinier T and Tirard C (2005) Comparative host-parasite population structures: disentangling prospecting and dispersal in the black-legged kittiwake Rissa tridactyla . Molecular Ecology, 14, 2825-2838.

McCoy KD, Boulinier T, Tirard C and Michalakis Y (2003) Host-dependent genetic structure of parasite populations: Differential dispersal of seabird tick host races. Evolution, 57, 288-296.

Miller GH, Margee JW and Jull AJT (1997) Low-latitude glacial cooling in the Southern Hemisphere from amino-acid racemization in emu eggshells. Nature, 385, 241-244.

Moritz C, Dowling TE and Brown WM (1987) Evolution of animal mitochondrial DNA: relevance for population biology and systematics. Annual Review of Ecological Systematics, 18, 269-292.

Nei M and Li W-H (1979) Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences, 76, 5269-5273.

Neigel JE (2002) Is Fst obselete? Conservation Genetics, 3, 167-173.

Nicholls JA and Austin JJ (2005) Phylogeography of an east Australian wet-forest bird, the satin bowerbird ( Ptilonorhynchus violaceus ), derived from mtDNA, and its relationship to morphology. Molecular Ecology, 14, 1485-1496.

Nieberding C, Morand S, Libois R and Michaux JR (2004) A parasite reveals cryptic phylogeographic history of its host. Proceedings of the Royal Society of London Series B-Biological Sciences, 271, 2559-2568.

Nielsen R and Wakeley J (2001) Distinguishing migration from isolation: a markov chain monte carlo approach. Genetics, 158, 885-896.

Page RDM (2003) Introduction. In Page RDM (Ed.) Tangled Trees. Chicago, The University of Chicago Press.

Page RDM and Holmes EC (1998) Molecular Evolution, Cambridge, University Press.

Page RDM, Lee PLM, Becher AA, Griffiths R and Clayton DH (1998) A different tempo of mitochondrial DNA evolution in birds and their parasitic lice. Molecular Phylogenetics and Evolution, 9, 276-293.

Paterson AM, Wallis GP, Wallis LJ and Gray RD (2000) Seabird and louse coevolution: Complex histories revealed by 12S rRNA sequences and reconciliation analyses. Systematic Biology, 49, 383-399.

Pavlova A, Zink RM, Rohwer S, Koblik EA, Red'kin YA, Fadeev IV and Nesterov EV (2005) Mitochondrial DNA and plumage evolution in the white wagtail Motacilla alba. Journal of Avian Biology, 36, 322-336.

Pearse DE and Crandall KA (2004) Beyond Fst: Analysis of population genetic data for conservation. Conservation Genetics, 5, 585-602.

118 References

Pereira SL and Baker AJ (2004) Low number of mitochondrial pseudogenes in the chicken (Gallus gallus) nuclear genome: implications for molecular inference of population history and phylogenetics. Bmc Evolutionary Biology, 4.

Peters JL, Gretes W and Omland KE (2005) Late pleistocene divergence between eastern and western populations of wood ducks (Aix sponsa) inferred by 'isolation with migration' coalescent method. Molecular Ecology, 14, 3407-3418.

Portner HO (2002) Climate variations and the physiological basis of temperature dependent biogeography: systemic to molecular hierarchy of thermal tolerance in animals. Comparative Biochemistry and Physiology a-Molecular and Integrative Physiology, 132, 739-761.

Posada D and Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics, 14, 817-818.

Posada D, Crandall KA and Templeton AR (2000) GeoDis: a program for the cladistic nested analysis of the geographical distribution of genetic haplotypes. Molecular Ecology, 9, 487-488.

Qu YH, Ericson PGP, Lei FM and Li SH (2005) Postglacial colonization of the Tibetan plateau inferred from the matrilineal genetic structure of the endemic red-necked snow finch, Pyrgilauda ruficollis . Molecular Ecology, 14, 1767-1781.

Queller DC, Strassman JE and Hughes CR (1993) Microsatellites and kinship. Trends in Ecology and Evolution, 8, 285-288.

Quinn TW (1992) The genetic legacy of Mother Goose - phylogeographic patterns of lesser snow goose Chen caerulescens caerulescens maternal lineages. Molecular Ecology, 1, 105-117.

Ramos-Onsins SE and Rozas J (2002) Statistical properties of new neutrality tests against population growth. Molecular Biology and Evolution, 19, 2092-2100.

Raymond F and Rousset M (1995) Population genetics software for exact tests and ecumenicism. Journal of Heredity, 86, 248-249.

Rogers AR and Harpending H (1992) Population growth makes waves in the distribution of pairwise genetic differences. Molecular Biological Evolution, 9, 552- 569.

Rozas J and Rozas R (1999) DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics, 15, 174-175.

Ruokonen M and Kvist L (2002) Structure and evolution of the avian mitochondrial control region. Molecular Phylogenetics and Evolution, 23, 422-432.

Sato A, Tichy H, O'HUigin C, Grant PR, Grant BR and Klein J (2001) On the origin of Darwin's finches. Molecular Biology and Evolution, 18, 299-311.

Schneider S and Excoffier L (1999) Estimation of Past Demographic Parameters From the Distribution of Pairwise Differences When the Mutation Rates Vary Among Sites: Application to Human Mitochondrial DNA. Genetics, 152, 1079-1089.

119

References

Schodde R and Mason IJ (1999) The Directory of Australian Birds: Passerines, Melbourne, CSIRO Publishing.

Scott SN, Clegg SM, Blomberg SP, Kikkawa J and Owens IPF (2003) Morphological shifts in island-dwelling birds: The roles of generalist foraging and niche expansion. Evolution, 57, 2147-2156.

Shaw PW, Pierce GJ and Boyle PR (1999) Subtle population structuring within a highly vagile marine invertebrate, the viened squid Loligo forbesi, demonstrated with micorsatellite DNA markers. Molecular Ecology, 8, 407-417.

Sheldon FH and Whittingham LA (1997) Phylogeny in studies of bird ecology, behavior and morphology. In Mindell DP (Ed.) Avian molecular evolution and systematics. London, Academic Press.

Shields GF and Wilson AC (1987) Calibration of mitochondrial DNA evolution in geese. Journal of Molecular Evolution, 24, 212-217.

Singer-Sam J, Tanguay RL and Riggs AD (1989) Use of Chelex to improve The PCR signal from a small number. Amplifications: A forum for PCR Users of cells. Applied Biosystems, 3, 11.

Slatkin M (1981) Estimating Levels of Gene Flow in Natural-Populations. Genetics, 99, 323-335.

Slatkin M (1987) Gene flow and the geographic structure of natural populations. Science, 236, 787-792.

Small MP, Stone KD and Cook JA (2003) American marten ( Martes americana ) in the Pacific Northwest: population differentiation across a landscape fragmented in time and space. Molecular Ecology, 12, 89-103.

Sorenson MD and Fleischer RC (1996) Multiple independent transpositions of mitochondrial DNA control region sequences to the nucleus. Proceedings of the National Academy of Sciences of the United States of America, 93, 15239-15243.

Specht RL (1993) Vegetation. In McTainsh G and Broughton WC (Eds.) Land degradation processes in Australia. Melbourne, Longman Cheshire.

Spencer PBS, Rhind SG and Eldridge MDB (2001) Phylogeographic structure within Phascogale (Marsupialia: Dasyuridae) based on partial cytochrome b sequence. Australian Journal of Zoology, 49, 369-377.

Spiess EB (1989) Genes in Populations, New York, Wiley and Sons.

Swofford D (2000) PAUP: Phylogenetic Analysis Using Pasrsimony, Version 4.0b4a. Sinauer, Sunderland, MA.

Tajima F (1989) Statistical Method for Testing the Neutral Mutation Hypothesis by DNA Polymorphism. Genetics, 123, 585-595.

Taylor G (1994) Landscapes of Australia: their nature and evolution. In Hill RS (Ed.) History of the Australian Vegetation. Cambridge, Cambridge University Press.

120 References

Templeton AR (1998) Nested clade analysis of phylogeographic data: testing hypotheses about gene flow and population history. Molecular Ecology, 7, 381-397.

Templeton AR, Boerwinkle E and Sing CF (1987) A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping. I. Basic theory and as analysis of alcohol dehydrogenase activity in Drosophila. Genetics, 117, 343-351.

Templeton AR and Sing CF (1993) A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping. IV. Nested analyses with cladogram uncertainty and recombination. Genetics, 134, 659-669.

Toon A, Hughes J, Baker A and Mather P (2003) Discordance between morphology and genetic structure among three plumage forms of the Australian magpie. EMU, 103, 337- 343.

Van Oosterhout C, Hutchinson WF, Wills DPM and Shipley P (2004) MICRO- CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes, 4, 535-538.

Van Treuren R, Bijlsma R, Tinbergen JM, Heg D and Van de Zande L (1999) Genetic analysis of the population structure of socially organized oystercatchers ( Haematopus ostralegus ) using microsatellites. Molecular Ecology, 8, 181-187.

Veltman CJ (1989) Flock, pair and group living lifestyles without cooperative breeding by Australian magpies Gymnorhina tibicen. Ibis, 131, 601-608.

Veltman CJ and Carrick R (1990) Male-biased dispersal in Australian magpies. Animal Behaviour, 40, 190-192.

Voris HK (2000) Maps of Pleistocene sea levels in Southeast Asia: shorelines, river systems and time durations. Journal of Biogeography, 27, 1153-1167.

Weckstein JD (2004) Biogeography explains cophylogenetic patterns in Toucan chewing lice. Systematic Biology, 53, 154-164.

Weir JT (2006) Divergent timing and patterns of species accumulation in lowland and highland neotropical birds. Evolution, 60, 842-855.

Werle E, Schneider C, Renner M, Volker M and Fiehn W (1994) Convenient Single- Step, One Tube Purification of PCR Products for Direct Sequencing. Nucleic Acids Research, 22, 4354-4355.

Whiteman NK, Santiago-Alarcon D, Johnson KP and Parker PG (2004) Differences in straggling rates between two genera of dove lice (Insecta: Phthirapters) reinforce population genetic and cophylogenetic patterns. International Journal of Parasitology, 34, 1113-1119.

Whitlock FL (1909) Notes on birds observed on the Pilbarra goldfield, North-Western Australia. The Emu, 13, 173-195.

Wilson H (1946) The life history of the western magpie ( Gymnorhina dorsalis ). Emu, 45, 233-285.

121

References

Wright S (1943) Isolation by distance. Genetics, 28, 114-138.

Wright S (1978) Evolution and genetics of populations. Vol. 4. Variability within and among populations. , Chicago, University of Chicago Press.

Wright TF and Wilkinson GS (2001) Population genetic structure and vocal dialects in an amazon parrot. Proceedings of the Royal Society of London Series B-Biological Sciences, 268, 609-616.

Yang S-J, Yin Z-H, Ma X-M and Lei F-M (2006) Phylogeography of ground tit (Pseudopodoces humilis) based on mtDNA: Evidence of past fragmentation on the Tibetan Plateau. Molecular Phylogenetics and Evolution, 41, 257-265.

Zhang D-X and Hewitt GM (2003) Nuclear DNA analysis in genetic studies of populations: practice, problems and prospects. Molecular Ecology, 12, 563-584.

Zink R (1996) Comparative phylogeography in North American birds. Evolution, 50, 308-317.

Zink RM (1991) The geography of mitochondrial DNA variation in two sympatric sparrows. Evolution, 45, 329-339.

Zink RM (1997) Phylogeographic Studies of North American Birds. In Mindell DP (Ed.) Avian molecular evolution and systematics. London, Academic Press.

Zink RM, Pavlova A, Rohwer S and Drovetski SV (2006) Barn swallows before barns: population histories and intercontinental colonization. Proceedings of the Royal Society B-Biological Sciences, 273, 1245-1251.

122 Appendix

APPENDICIES

Appendix I: Site locations and control region mtDNA haplotypes present at each site. Frequency of haplotypes is in parentheses. Site ID ID Sample Haplotypes Site Location Collector Accession numbers Fig. 2.1 size Fg 1 4 3(3)42(1) 23°40'S 132°39'E This study EF156252-EF156274 As 2 11 2(5)3(6) 23°42'S 133°52'E This study EF156252-EF156274 Cp A 2 2(1)7(1) 24°52'S 133°49'E This study EF156252-EF156274 Cc 3 4 2(4) 16°41'S 135°43'E This study EF156252-EF156274 Qm7 B 1 5(1) 19°56'S 138°07'E Qld Museum EF156252-EF156274 Qm9 C 1 3(1) 20°53'S 140°21'E Qld Museum EF156252-EF156274 Qm3 D 1 3(1) 17°52'S 140°49'E Qld Museum EF156252-EF156274 Qm2 E 2 13(1)50(1) 19°29'S 143°06'E Qld Museum EF156252-EF156274 Qm1 F 2 3(1)4(1) 20°02'S 144°18'E Qld Museum EF156252-EF156274 Qm10 G 1 4(1) 20°52'S 144°32'E Qld Museum EF156252-EF156274 Ct 4 29 2(19)3(6)4(1)5(1)39(1)41(1) 20°04'S 146°15'E This study EF156252-EF156274 Bw 5 31 2(3)3(23)4(5) 25°11'S 151°39'E This study EF156252-EF156274 Mb 6 21 2(8)3(2)4(1)5(7)36(2)41(1) 22°00'S 148°03'E This study EF156252-EF156274 Rk 7 29 2(12)3(9)6(3)7(3)40(1) 23°22'S 150°30'E This study EF156252-EF156274 Qm4 H 1 6(1) 23°56'S 143°53'E Qld Museum EF156252-EF156274 Qp I 1 4(1) 26°37'S 144°16'E This study EF156252-EF156274 Rm J 1 3(1) 26°34'S 148°47'E This study EF156252-EF156274 Ma 8 54 2(7)3(23)4(14)6(8)11(2) 25°32'S 152°42'E A M Baker AF198500-AF198524 To 9 55 1(1)2(18)3(14)4(17)5(1)6(4) 27°33'S 151°57'E A M Baker AF198500-AF198524 Br 10 50 1(2)2(11)3(26)4(9)8(2) 27°28'S 153°01'E A M Baker AF198500-AF198524 Iw K 1 2(1) 28°23'S 151°19'E This study EF156252-EF156274 Gr 11 51 1(14)2(4)3(21)4(5)6(6)9(1) 29°40'S 152°56'E A M Baker AF198500-AF198524 Bk L 2 3(1)13(1) 31°11'S 146°50'E This study EF156252-EF156274 Db 12 51 2(14)3(7)4(3)6(23)8(1)13(3) 32°14'S 148°36'E A M Baker AF198500-AF198524 Or 13 52 1(6)2(6)3(7)4(3)6(19)13(9)16(2) 33°16'S 149°06'E A M Baker AF198500-AF198524 Ck 14 59 2(15)3(10)4(29)6(5) 32°49'S 151°21'E A M Baker AF198500-AF198524

123 Appendix

Site ID ID Sample Haplotypes Site Location Collector Accession numbers Fig. 2.1 size Gb 15 50 2(11)3(4)4(12)6(11)13(6)15(6) 34°44'S 149°44'E A M Baker AF198500-AF198524 Ou 16 51 2(9)3(3)4(22)5(1)6(2)7(12)8(2) 35°04'S 142°21'E A M Baker AF198500-AF198524 Hm 17 50 2(7)3(2)4(23)6(13)8(2)10(3) 36°42'S 142°13'E A M Baker AF198500-AF198524 Sr 18 50 2(11)4(19)6(11)13(8)14(1) 37°01'S 145°09'E J M Hughes AF198500-AF198524 Pi 19 50 2(4)4(26)6(14)9(5)10(1) 38°27'S 145°15'E A M Baker AF198500-AF198524 Rw 20 51 2(8)4(12)6(17)7(1)8(1)13(3)41(3)4 37°43'S 144°24'E K Durrant EF156252-EF156274 4(1)45(2)46(1)48(2) Ts 21 56 3(41)5(4)25(11) 41°26'S 147°08'E J M Hughes AF198500-AF198524 Mv5 M 1 6(1) 35°16'S 140°52'E Museum Victoria EF156252-EF156274 Mv3 N 1 35(1) 35°53'S 140°04'E Museum Victoria EF156252-EF156274 Sa1 O 1 4(1) 32°50'S 139°10'E This study EF156252-EF156274 Sa 22 4 2(1)4(1)6(2) 34°34'S 138°50'E This study EF156252-EF156274 Ki 23 7 2(3)4(3)49(1) 35°44'S 137°30'E This study EF156252-EF156274 Cd 24 8 26(4)27(4) 32°07'S 133°40'E J M Hughes EF156252-EF156274 Ep 25 16 19(10)32(4)33(2) 33°51'S 121°53'E J M Hughes EF156252-EF156274 Al 26 50 17(31)18(6)19(12)20(1) 35°00'S 117°53'E A M Baker AF198500-AF198524 Bn 27 50 17(15)19(33)22(1)23(1) 33°39'S 115°21'E A M Baker AF198500-AF198524 Mh 28 50 17(33)18(7)19(10) 32°31'S 115°44'E A M Baker AF198500-AF198524 Pe 29 48 17(21)18(11)19(12)21(4) 31°53'S 115°51'E A M Baker AF198500-AF198524 Ch P 2 19(1)31(1) 29°41'S 115°53'E A Toon EF156252-EF156274 Pb1 30 6 17(1)19(5) 24°48'S 114°25'E A Toon EF156252-EF156274 Pb2 31 7 17(2)19(2)29(1)30(2) 27°49'S 114°41'E A Toon EF156252-EF156274 Pb3 32 8 19(7)30(1) 27°05'S 116°09'E A Toon EF156252-EF156274 Mwa3 Q 1 19(1) 25°54'S 122°21'E WA Museum EF156252-EF156274 Pb4 33 4 19(2)37(1)38(1) 22°41'S 117°47'E A Toon EF156252-EF156274 Mwa1 R 1 19(1) 22°51'S 114°56'E WA Museum EF156252-EF156274 Bm 34 3 19(3) 17°57'S 122°14'E This study EF156252-EF156274 Km1 3 7 19(4)34(2)43(1) 16°49'S 124°55'E This study EF156252-EF156274 Km2 36 10 19(10) 15°42'S 126°22'E This study EF156252-EF156274 Kr S 1 17(1) 15°49'S 129°02'E This study EF156252-EF156274

124 Appendix

Site ID ID Sample Haplotypes Site Location Collector Accession numbers Fig. 2.1 size Mt 37 4 19(4) 14°55'S 133°04'E This study EF156252-EF156274 Bh T 1 19(1) 19°42'S 135°49'E This study EF156252-EF156274 Total 1166

125 Appendix

Appendix II: 28 variable sites for 46 G. tibicen control region haplotypes

[ 1111111112 2222222223 33333344] [ 0557789992 3367777782 23358834] [ 7785870121 0254678986 85981496] Gt1 TCACTCTACA CCCCTAACAT GACAAAGG Gt2 ...... TG ...... GG.. Gt3 .....T..TG ...... GG.. Gt4 ...... TG ...... G... Gt5 .....T..TG ...... G...

Gt6 ...... TG ...... A...... G... Gt7 ...... T...... GG.. Gt8 ...T.....G .T...... GG.. Gt9 ...... GTG ...... A...... G... Gt10 ...... G... Gt11 .....TC.TG ...... GG.. Gt13 ...T.....G ...... GG.. Gt14 ...... G ...... G... Gt15 ...TC...TG ...... A...... G... Gt16 ...... TG ....C..A...... G... Gt17 ..TT.....G .T.TC..A...... G... Gt18 ..TT.....G .T.TC..AG. ....G... Gt19 ..TT.....G .T.TC..A...... GG.. Gt20 ..TT.....G .T.TC..GG. ....G... Gt21 ..TT.....G .T.TCG.AG. ....G... Gt22 C.TT.....G .T.TC..A...... G... Gt23 ..T...... G .T.TC..A...... G... Gt25 .....T..TG ...... G...... GG.. Gt26 ...... G ...... A.. A...GG.. Gt27 .....T..TG ...... A...... G... Gt29 ..T...... G ...TC..A...... G... Gt30 ..TT.....G .T.TC.GA...... GG.. Gt31 ..TT.....G .T.TCG.A...... G... Gt32 ..TT.....G .T.TC..AG. ....GG.. Gt33 ..TT.....G .T.TC..G...... G... Gt34 ..TT.....G .T.TC..A.. ..T.GG.. Gt35 ...... TG .....T...... G... Gt36 ...... G ...... A...GG.. Gt37 ..TTC....G .T.TC..A...... GG.C Gt38 ..TT.....G .T.TC..A...... GGA. Gt39 .....T..TG .....G...... GG.. Gt40 ...... GTG T...... G... Gt41 ...... GG.. Gt42 ...... T. ..T...... TGG.. Gt43 .TTTC....G .T.TC..A...... GG.. Gt44 ...... T...... A...... G... Gt45 ...... GTG ...... G... Gt46 ...... T. T...... GG.. Gt48 ...... TG ...... C ....GG.. Gt49 ...... T...... G... Gt50 ...... TG ...... A.. .C..G...

126 Appendix

Appendix III: Pairwise mtDNA FST values among magpie sites. Non-significant values in bold. P< 0.02 As Mb Ct Bw Rk Ma To Br Gr Db Or Ck Gb On Hm Sm Pi Rw Sa Cd Ts Ep Al Bn Mh Pe Pb As 0 Mb 0.1 0 Ct 0.07 0.08 0 Bw 0.04 0.12 0.22 0 Rk 0.03 0.04 0 0.14 0 Ma 0.12 0.01 0.11 0.11 0.05 0 To 0.18 0.01 0.09 0.21 0.05 0.02 0 Br 0.01 0 0.05 0.03 0.02 0.04 0.07 0 Gr 0.17 0.05 0.17 0.19 0.12 0.1 0.1 0.11 0 Db 0.31 0.15 0.23 0.35 0.16 0.12 0.1 0.22 0.14 0 Or 0.26 0.12 0.22 0.31 0.17 0.14 0.12 0.2 0.07 0.02 0 Ck 0.34 0.09 0.23 0.36 0.17 0.08 0.02 0.18 0.14 0.09 0.12 0 Gb 0.31 0.14 0.23 0.35 0.18 0.14 0.11 0.22 0.14 0.02 0.02 0.09 0 On 0.28 0.1 0.17 0.35 0.13 0.13 0.05 0.18 0.1 0.12 0.1 0.05 0.1 0 Hm 0.45 0.2 0.36 0.47 0.29 0.2 0.15 0.31 0.15 0.06 0.05 0.07 0.05 0.1 0 Sm 0.38 0.16 0.28 0.43 0.23 0.18 0.11 0.26 0.15 0.05 0.03 0.07 0.01 0.08 0.02 0 Pi 0.61 0.36 0.53 0.61 0.44 0.32 0.29 0.45 0.25 0.13 0.12 0.2 0.12 0.24 0.03 0.12 0 Rw 0.35 0.16 0.26 0.4 0.2 0.17 0.12 0.26 0.13 0.02 0.02 0.08 0.02 0.07 0.01 0.01 0.07 0 Sa 0.49 0.18 0.37 0.52 0.26 0.17 0.11 0.29 0.12 0.05 0.05 0.03 0.03 0.05 -0.02 0.01 0.05 -0.01 0 Cd 0.58 0.4 0.6 0.62 0.48 0.44 0.51 0.49 0.29 0.31 0.22 0.56 0.33 0.51 0.45 0.43 0.53 0.34 0.46 0 Ts 0.31 0.39 0.5 0.16 0.39 0.32 0.45 0.25 0.35 0.52 0.47 0.58 0.52 0.56 0.64 0.61 0.73 0.57 0.71 0.72 0 Ep 0.91 0.85 0.9 0.91 0.86 0.85 0.86 0.85 0.76 0.81 0.73 0.88 0.78 0.85 0.85 0.83 0.89 0.8 0.88 0.82 0.93 0 Al 0.92 0.89 0.92 0.92 0.89 0.88 0.88 0.88 0.81 0.84 0.78 0.89 0.82 0.87 0.86 0.85 0.89 0.83 0.9 0.87 0.94 0.36 0 Bn 0.93 0.89 0.92 0.93 0.9 0.88 0.89 0.88 0.81 0.85 0.79 0.9 0.83 0.88 0.87 0.86 0.9 0.84 0.91 0.88 0.94 0.14 0.24 0 Mh 0.93 0.89 0.92 0.93 0.9 0.88 0.89 0.89 0.81 0.85 0.79 0.9 0.83 0.88 0.87 0.86 0.9 0.84 0.91 0.88 0.94 0.42 - 0.28 0 0.02 Pe 0.9 0.86 0.89 0.9 0.87 0.86 0.87 0.86 0.79 0.83 0.76 0.88 0.8 0.86 0.85 0.84 0.87 0.81 0.87 0.84 0.92 0.28 0.02 0.26 0.03 0 Pb 0.91 0.86 0.9 0.91 0.87 0.85 0.87 0.86 0.77 0.82 0.74 0.88 0.79 0.85 0.85 0.83 0.89 0.8 0.88 0.83 0.93 0.07 0.32 0.03 0.37 0.3 0 Km 0.92 0.86 0.91 0.92 0.88 0.86 0.87 0.86 0.77 0.82 0.75 0.89 0.79 0.86 0.86 0.84 0.9 0.81 0.9 0.85 0.94 0.04 0.41 0.09 0.47 0.37 0.02

127 Appendix

Appendix IV: Pairwise microsatellite FST values among magpie sites. Non-significant values in bold. P< 0.02 Km Ep Al Bn Mh Pe Pb As Ct Bw Mb Rk Ma Br To Gr Ck Db Or Gn On Hm Sm Pi Rw Cd Km 0 Ep 0.08 0 Al 0.07 0.04 0 Bn 0.1 0.09 0.05 0 Mh 0.08 0.07 0.05 0.04 0 Pe 0.09 0.05 0.04 0.06 0.02 0 Pb 0.06 0.02 0.03 0.05 0.05 0.03 0 As 0.03 0.07 0.04 0.06 0.06 0.07 0.02 0 Ct 0.06 0.06 0.05 0.08 0.07 0.05 0.03 0.04 0 Bw 0.06 0.05 0.04 0.07 0.07 0.05 0.02 0.04 0.04 0 Mb 0.03 0.03 0.02 0.06 0.05 0.05 0.01 0.01 0.01 0.02 0 Rk 0.06 0.06 0.05 0.09 0.1 0.09 0.03 0.04 0.03 0.03 0.02 0 Ma 0.07 0.07 0.03 0.07 0.07 0.05 0.05 0.04 0.04 0.03 0.04 0.05 0 Br 0.07 0.08 0.04 0.07 0.07 0.07 0.04 0.03 0.04 0.04 0.02 0.05 0.01 0 To 0.06 0.07 0.04 0.05 0.05 0.05 0.04 0.02 0.03 0.04 0.01 0.04 0.01 0.01 0 Gr 0.05 0.07 0.04 0.08 0.05 0.06 0.04 0.03 0.03 0.04 0.03 0.05 0.02 0.02 0.02 0 Ck 0.06 0.06 0.03 0.07 0.07 0.06 0.03 0.03 0.03 0.03 0.02 0.04 0.02 0.03 0.02 0.02 0 Db 0.04 0.05 0.03 0.05 0.05 0.04 0.02 0.02 0.02 0.03 0.01 0.04 0.02 0.03 0.01 0.02 0.02 0 Or 0.04 0.04 0.02 0.05 0.05 0.04 0.02 0.02 0.02 0.01 0.01 0.02 0.01 0.02 0.01 0.02 0.01 0.01 0 Gn 0.06 0.04 0.03 0.06 0.07 0.05 0.03 0.03 0.03 0.03 0.02 0.03 0.01 0.02 0.02 0.03 0.02 0.02 0 0 On 0.04 0.04 0.02 0.04 0.04 0.03 0.02 0.02 0.03 0.02 0.02 0.04 0.01 0.02 0.01 0.03 0.02 0.01 0 0.01 0 Hm 0.04 0.07 0.03 0.05 0.04 0.05 0.03 0.01 0.03 0.03 0.02 0.04 0.01 0.01 0.01 0.02 0.02 0.02 0 0.01 0 0 Sm 0.05 0.05 0.03 0.05 0.05 0.05 0.02 0.02 0.03 0.03 0.02 0.04 0.02 0.02 0.02 0.02 0.02 0.02 0.01 0.02 0.01 -0.01 0 Pi 0.05 0.09 0.05 0.07 0.05 0.06 0.06 0.04 0.06 0.05 0.05 0.06 0.04 0.06 0.04 0.04 0.04 0.04 0.02 0.04 0.02 0.01 0.03 0 Rw 0.03 0.08 0.04 0.06 0.06 0.05 0.04 0.03 0.03 0.04 0.02 0.04 0.03 0.03 0.02 0.03 0.03 0.02 0.01 0.03 0.01 0.01 0.02 0.02 0 Cd 0.1 0.07 0.05 0.09 0.07 0.08 0.03 0.05 0.07 0.07 0.04 0.09 0.07 0.06 0.05 0.06 0.05 0.05 0.04 0.06 0.04 0.06 0.03 0.08 0.07 0 Ts 0.13 0.16 0.08 0.08 0.09 0.09 0.11 0.1 0.1 0.08 0.11 0.12 0.05 0.06 0.05 0.08 0.07 0.07 0.06 0.07 0.05 0.05 0.06 0.06 0.06 0.15

128 Appendix

Appendix V: 118 variable sites in 672bp of COI fragment for Philopterus sp. Ambiguous sites are coded with a ?.

[ 1111111111 1111111111 1222222222 2222222222 2333333333 33333] [ 1345568889 0011222233 4555557889 9011224455 6777788999 9001223334 55566] [ 8981431373 5817034935 7046797062 8403020958 4067925147 8368173695 17803] Ps_1 ??GATATCGG GATTATTCTA TTCAACTATT AGAATACTGG CAATGTATCT CCCGGATAGT CCTAA Ps_2 TT....G.A...... T...... T...... AG.. T.... Ps_3 ???...... G...... G Ps_4 ?????????? ...... T...... T...... A...... Ps_5 ?????????? ???...... T...... T...... A...... Ps_6 ??....GT.A ...... T...... T. ...A..AG...... Ps_7 ?????????? ...... T...... T. ...A..AG...... Ps_8 ??....G.A...... T...... T...... TAG.. T.... Ps_9 ?????????? ..C...... A.T.G.CG.. G..G.G...A T...A..C.. ...A...... Ps_10 ?????????? ..C...... A.T.G.CG.. G..G.G...A T...A..C.. ...A...... Ps_11 ?????????? ..C...... A.T.G.CG.. G..G.G...A T...A..C.. ...A...... Ps_12 ?????????? ..C...... A.T.G..... G..G.G..AA T...A..C...... Ps_13 ?????????? ..C..C.TC. .CT.G.C... G.GG..TC.A T.G.ACC... T.T..G..AC .T.G. Ps_14 ?????????? ...... TG...... T...... A...... Ps_15 TT..CGG... .G...... T.G.G... G.GG....AA ....A.T..C ...... Ps_16 CCA..GG... ..C..C.TC. .CT.G.C... G.GG..T..A T.G.ACC... T.T..G..AC .T.G. Ps_17 CCA..GG... ..C..C.TC. .CT.GTC... G.GG..T..A T...ACC... T.T..G..AC .T.G. Ps_18 CCA..GG... ..C..C.TC. .CT.G.C... G.GG..TC.A T.G.ACC... T.T..G..AC .T.G. Ps_19 CCA..GG... ..C..C..C. .CT.G.C... G.GG..TC.A ....ACC... T.T..G..AC .T.G. Ps_20 ?????????? ???GG.CT.G ..TG....C. G..G..T..A T.G.A.CC.C .ATCAGAT.C ..C.. Ps_21 ??A..GG... ..C..C.TC. .CT.GTC... G.GG..T..A T...ACC... T.T..G..A. .T.G. Ps_22 CCAG.GG... ..C..C.TC. .CT.G.C..C G.GG..TC.A T.G.ACC... T.T..G..AC .T.G. Ps_23 CTA..GG... ..C..C.TC. .CT.GTC... G.GG..T..A T...ACC... T.T..G..AC .T.G. BBCC1 ?????????? A.CGG.CT.. ..TG....C. GA.GC.T..A T.GCT.T..C .A.C.GAT......

[ 3333334444 4444444444 4444444445 5555555555 5556666666 666] [ 6688890011 1333444455 6677888990 1234566788 8990022233 444] [ 8914702612 4268014709 2817046274 3873514025 8173614736 568] Ps_1 ACTAAGATAC TGATTACGAT TTGAAGAACG TCCCCGAGGG AAGTATTATG GCT Ps_2 ....G...... T... .C....GG...... A. ....G..G.. ... Ps_3 ...... G...... Ps_4 ....G...... T... .C...... ?? ?????????? ??? Ps_5 ....GA...... C.T...... ?? ?????????? ??? Ps_6 ....G...... T...... G.. .TT..AT...... G..G.. ... Ps_7 ....G...... T...... G...... ?? ?????????? ??? Ps_8 ....G...... T... .C....GG...... A..A. ....G..G.. ... Ps_9 ....G..C...... T.G. C.....TG...... T...?? ?????????? ??? Ps_10 ....G..C...... T.G. C.....TG.. ...TT...?? ?????????? ??? Ps_11 ....G..C...... T.G. C.....TG.. ..TTT...?? ?????????? ??? Ps_12 ....G..C...... T.G. C.....TG.. ...TT...?? ?????????? ??? Ps_13 .T.TG..CGT AA....T...... GG.TGTA C.TTT...?? ?????????? ??? Ps_14 ....GA...... T...... ?? ?????????? ??? Ps_15 T.C.G..C.. ..GC..T.G. ..A...CG.. ..TT...... GA.G..G.A A.. Ps_16 .T.TG..CGT A.....T...... GG.TGTA C.TTT....A G..C...GG. ATA Ps_17 .T.TGA.CGT A.....T...... G.TGTA C.TTT..A.A G..C...GG. ATA Ps_18 .T.TG..CGT A.....T...... GG.TGTA C.TTT....A G..C...GG. ATA Ps_19 .T.TG..CGT A.....T...... GG.TGTA C.TTT....A G..C...GG. ATA Ps_20 ...TGA.... .A...GTA.. .C.G..TG.. C..T.....A G...GCAGG. A.. Ps_21 .T.TGA.CGT A.....T...... GG.TGTA C.TTT..A.A G..C...GG. ATA Ps_22 .T.TGA.CGT A.....T...... GGATGTA C.TTT....A G..C...GG. ATA Ps_23 .T.TGA.CGT A.....T...... GG.TGTA C.TTT..A.A G..C...GG. ATA BBCC1 ...TG.GC.. .A...GTA.C .C.G..TG.. C.TT....?? ?????????? ???

129 Appendix

Appendix VI: 154 variable sites in 600bp of COI fragment for B semiannulata . Ambiguous sites are coded with a ?.

[ 11111 1111111111 1111112222 2222222222 22222] [ 111122233 4556677777 8888900111 2234455667 7889990123 4445556667 77788] [ 6258912436 2176925789 1789058147 3921739584 7032587687 0692581273 47925] Bs_1 TAATTGTGAA AAAAAATTTA TTTCTTATTT TAGTTAAAGT TGAAGATAAG GGACCTTTAG CGGAG Bs_2 ?????????? ????...... A...... Bs_3 ?????????? ...... Bs_4 ????????...... T.... Bs_5 ?????????? ?...... T.... Bs_6 ????????.G G..G...... AC.G..A. ....A....A ..GT.C.C...... Bs_7 ?????????? ?..G...... AC.G..A. ....A....A ..GT.C.C...... Bs_8 ????????.G G..G...... AC.G..A. ....A....A ..GT.C.C...... Bs_9 ...... A.G G..G...... AC.G..A. ....A....A ..GT.C.C...... Bs_10 ?????????? G..G...... AC.G..A. ....A....A ..GT.C.C...... Bs_11 ????????.G G..G...... C...... AC.G..A. ....A....A ..GT.C.C...... Bs_12 ????????.G G..G...... AC.G..A. C...A....A ..GT.C.C...... Bs_13 ????????.G G..G...... AC.G..A. ....A....A ..GT.C.C...... Bs_14 ?????????? G..G...... C...... AC.G..A. ....A....A ..GT.C.C...... Bs_15 ...... A.. G..G...... GAC.G..A. ....A....A ..GT.C.C...... Bs_16 ...... A.. G..G...... GAC.G..A. ....A....A ..GT.C.C...... Bs_17 AGGGC.CC.. TT...GAACT A.GGG.GAA. .G...T.TA. .A.G.GGT.A .A.TT.A.TA .T.T. Bs_18 A..A.A..G. T.GGT....G ....A.G... C.A.A.G..C ..GGA...G. A..TT.C.GA ..A.A Bs_19 ?????????? ?...... A.T A.G.G.G.A. .G...T.TA. .A.G.GGT.A .A.T..G.TA .T.T. Bs_20 ?????????? ?...... A.T A.G.G.G.A...... T.TA. .A.G.GGT.A .A.TT.A.TA .T.T. Bs_21 AGGGC.CC.. TT...GAACT A.GGG.GAA. .G...T.TA. .A.G.GGT.A .A.TT.A.TA .T.T. Bs_22 ?????????? ?.GGT....G ....AGG..G C.A...G..C ...GA...G. A..TT.C.GA ..A.A

[ 2222222233 3333333333 3333333333 3333333333 4444444444 4444444444 44444] [ 8888999900 0001113333 3444455556 6778899999 0011123444 5555556677 88899] [ 6789156702 3692581367 9058901470 8121803679 2514735147 0146892517 03425] Bs_1 ATATAAATTC AAGTATCTTA ACAATCGAAT TTATCTGTTA TTGATCTGTA GAATTGAGAT GCAAT Bs_2 ...... Bs_3 ...... Bs_4 ...... Bs_5 ...... Bs_6 T.G...... G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_7 T.G...... G...... G...... A.C. ..A..T.A.G ...... TGG. Bs_8 T.G...... G...... G...... A... ..A..T.A.G ...... TGG. Bs_9 T.G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_10 T.G...... G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_11 T.G...... G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_12 T.G...... G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_13 T.G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_14 T.G...... G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_15 T.G...... G...... G...... A.C. ..A..T.A.G ...... T.G. Bs_16 T.G...... G...... G...... A.C. ..A..T.A.G ...... A.. .T.G. Bs_17 TATAGGTA.T TG..G.TG.G TTT.C.TG.A AC.ATGTA.G A..T.TAACG .G.ACAGAT. AT.G. Bs_18 T.GAT...GT G.TA.CTAA. .TGG.TATG. .CT.TAT... .A.TATA... .GT....AT. .T.GA Bs_19 TATAGGTA.T TG..G.TG.G TTT.C.TG.A AC.ATGTA.G A..T.TAACG .G.ACAGAT. AT.T. Bs_20 TATAGGTA.T TG..G.TG.G TTT.C.TG.A AC.ATGTA.G A..T.TAACG .G.ACAGAT. AT.T. Bs_21 TATAGGTA.T TG..G.TG.G TTT.C.TG.A AC.ATGTA.G A..T.TAACG .G.ACAGAT. AT.T. Bs_22 T.GAT...GT G.T..CTAA. .TGG.TATG. .CT.TAT... .A.TATA... AGT....ATG .T..A

130 Appendix

[ 4555555555 5555555555 5556] [ 9000111223 3445556677 8990] [ 8478013581 7362581406 5170] Bs_1 ATGGTCTTAC ATGAAAGGGA CTTT Bs_2 ...... Bs_3 ....C....T ..A...A...... Bs_4 ....C....T ..A...A...... Bs_5 ...... Bs_6 ....C....T ..A...A...... Bs_7 ....C....T ..A...A...... Bs_8 ....C....T ..A...A...... Bs_9 ...... T ..A...A...... Bs_10 ...... T ..A??????? ???? Bs_11 ....C....T ..A...A...... Bs_12 ....C....T ..A...A...... Bs_13 ....C....T ..A...A...... Bs_14 ...... T ..A...A...... Bs_15 G...... T ..A...A...... Bs_16 G...... T ..A...A...... Bs_17 .G.....G.T ..A.G...AG TCAA Bs_18 ..TACTA.GT GA.TGG.A.. TCCG Bs_19 .G.....G.T ..A.G...AG TCAA Bs_20 .G.....G.T ..A.G...AG TCAA Bs_21 .G.....G.T ..A.G...AG TCAA Bs_22 .????????? ?????????? ????

131 Appendix

Appendix VII: AMOVA results inclusive of all samples for a) Philopterus sp . b) B. semiannulata. a. Percentage Variation among among pop within ΦSC P ΦCT P group within group pops 1. host mtDNA clades 0.74 <0.001 0.75 <0.05 75.0% 18.5% 6.5% 2. host back colour 0.90 <0.001 -0.01 0.47 -0.6% 90.5% 10.1%

b. Percentage Variation among among pop within ΦSC P ΦCT P group within group pops 1. host mtDNA clades 0.52 <0.001 -0.17 0.96 -16.6% 61.2% 55.5% 2. host back colour 0.26 <0.001 0.44 <0.01 43.7% 14.6% 41.7%

132 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt201a

Alice Springs

0.8 0.7

0.6 0.5

0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Cape Crawford

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Charters Towers

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Bowen

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Moranbah

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Rockhampton

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

133 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt201a

Maryborough

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Toowoomba

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Brisbane

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Grafton

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180

Dubbo

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

134 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt201a

Orange

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Cessnock

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Golbourne

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Ouyen

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Horsham

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Seymour

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

135 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt201a

Phillip Island

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Rowsely

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Tasmania

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Ceduna

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Esperance

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Albany

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

136 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt201a

Busselton

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Mandurah

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Perth

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Pilbara

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Kimberley

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

Mataranka

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 146 148 150 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182

137 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt206b

Alice Springs

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Cape Crawford

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Charters Towers

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Bowen

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Moranbah

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Rockhampton

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

138 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt206b

Maryborough

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Toowoomba

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Brisbane

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Grafton

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Dubbo

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Orange

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

139 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt206b

Cessnock

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Golbourne

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Horsham

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Ouyen

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Seymour

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Phillip Island

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

140 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt206b

Rowsely

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Tasmania

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Ceduna

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Esperance

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Albany

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Busselton

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

141 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt206b

Mandurah

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Perth

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Kimberley

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Pilbara

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

Mataranka

0.6

0.5

0.4

0.3

0.2

0.1

0 124 126 128 130 132 134 136 138 140 142 144 146 148 150 152 154 156 158 160

142 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt115a

Alice Springs

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Cape Crawford

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Charters Towers

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Bowen

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Moranbah

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Rockhampton

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

143 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt115a

Maryborough

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Toowoomba

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Brisbane

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Grafton

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Dubbo

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Orange

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

144 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt115a

Cessnock

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Golbourne

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Ouyen

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Horsham

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Seymour

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Phillip Island

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

145 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt115a

Rowsely

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Tasmania

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Ceduna

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Esperance

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Albany

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Busselton

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

146 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt115a

Mandurah

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Perth

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Pilbara

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Kimberley

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

Mataranka

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 174 176 178 180 182 184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 216 218 220 222 224 226

147 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt112a

Alice Springs

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Cape Crawford

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Charters Towers

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Bowen

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Moranbah

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

148 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt112a

Rockhampton

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Maryborough

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Toowoomba

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Brisbane

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Grafton

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

149 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt112a

Dubbo

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Orange

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Cessnock

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Golbourne

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Ouyen

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

150 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt112a

Horsham

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Seymour

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Phillip Island

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Rowsely

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Tasmania

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

151 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt112a

Ceduna

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Esperance

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Albany

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Busselton

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Mandurah

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

152 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt112a

Perth

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Pilbara

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Kimberley

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

Mataranka

0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 183

153 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt67c

Alice Springs

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Cape Crawford

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Charters Towers

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Bowen

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Moranbah

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Rockhampton

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

154 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt67c

Maryborough

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Toowoomba

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Brisbane

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Grafton

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Dubbo

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Orange

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

155 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt67c

Cessnock

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Golbourne

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Ouyen

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Horsham

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Seymour

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Phillip Island

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

156 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt67c

Rowsely

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Tasmania

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Ceduna

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Esperance

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Albany

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Busselton

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

157 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus: Gt67c

Mandurah

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Perth

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Pilbara

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Kimberley

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

Mataranka

1

0.8

0.6

0.4

0.2

0 118 121 124 127 130 133

158 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus:Gt43a Alice Springs 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247 Cape Crawford 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Charters Towers 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Bowen 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Moranbah 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Rockhampton 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

159 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus:Gt43a Maryborough 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Toowoomba 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Brisbane 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Grafton 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247 Dubbo 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247 Orange 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

160 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus:Gt43a Cessnock 0.5 0.4 0.3 0.2 0.1 0 175 17 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 24 247 8 4 Golbourne 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247 Ouyen 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Horsham 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Seymour 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Phillip Island 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

161 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus:Gt43a Rowsley 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247 Tasmania 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Ceduna 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Esperance 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Albany 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Busselton 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

162 Appendix

Appendix VIII: Frequency data of six microsatellite loci in G. tibicen populations Locus:Gt43a Mandurah 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Perth 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Pilbara 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Kimberley 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

Mataranka 0.5 0.4 0.3 0.2 0.1 0 175 178 181 184 187 190 193 196 199 202 205 208 211 214 217 220 223 226 229 232 235 238 241 244 247

163