perched at the mito-nuclear crossroads: divergent mitochondrial lineages correlate with environment...

Perched at the mito-nuclear crossroads: divergent

mitochondrial lineages correlate with environment in the face of ongoing nuclear gene flow in an Australian bird

Journal: Evolution

Manuscript ID: 12-0536.R2

Manuscript Type: Original Article

Date Submitted by the Author: n/a

Complete List of Authors: Pavlova, Alexandra; Monash University, School of Biological Sciences and

Australian Centre for Biodiversity Amos, Nevil; Monash University, School of Biological Sciences and Australian Centre for Biodiversity Joseph, Leo; CSIRO Australia, Ecosystem Sciences Loynes, Kate; The Australian National University, Division of Evolution, Ecology and Genetics Austin, Jeremy; The University of Adelaide, School of Earth & Environmental Sciences; Museum Victoria, Sciences Department Keogh, Scott; The Australian National University, Evolution, Ecology & Genetics Stone, Graham; University of Edinburgh, Institute of Evolutionary Biology Nicholls, James; University of Edinburgh, Institute of Evolutionary Biology

Sunnucks, Paul; Monash University, School of Biological Sciences and Australian Centre for Biodiversity

Keywords: Gene Flow, Molecular Evolution, Phylogeography, Selection - Natural, Coalescent Theory, Genetic Drift

Evolution: For Review Only

1

PERCHED AT THE MITO-NUCLEAR CROSSROADS: DIVERGENT MITOCHONDRIAL LINEAGES

CORRELATE WITH ENVIRONMENT IN THE FACE OF ONGOING NUCLEAR GENE FLOW IN AN

AUSTRALIAN BIRD

Alexandra Pavlova1, J. Nevil Amos1, Leo Joseph2 , Kate Loynes3, Jeremy J. Austin4,5, J. Scott

Keogh3, Graham N. Stone6, James A. Nicholls6, and Paul Sunnucks1

1School of Biological Sciences and Australian Centre for Biodiversity, Clayton Campus,

Monash University, Wellington Road, Clayton, VIC 3800, Australia

2Australian National Wildlife Collection, CSIRO Ecosystem Sciences, GPO Box 1700,

Canberra, ACT 2601

3 Division of Evolution, Ecology and Genetics, Building 116, Daley Rd, Research School of

Biology, The Australian National University, Canberra, ACT 0200 Australia

4Australian Centre for Ancient DNA (ACAD), School of Earth & Environmental Sciences,

The University of Adelaide, Darling Building, North Terrace Campus, SA 5005, Australia

5Sciences Department, Museum Victoria, Carlton Gardens, Melbourne, Vic, 3001, Australia

6 Institute of Evolutionary Biology, University of Edinburgh, The King’s Buildings, West

Mains Road, Edinburgh EH9 3JT, Scotland

Keywords: evolutionary history, sPCA, dbRDA, female-linked selection, eastern yellow

robin, Eopsaltria australis

Corresponding author: Alexandra Pavlova

School of Biological Sciences and Australian Centre for Biodiversity, Clayton campus,


Page 1 of 101 Evolution: For Review Only

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

2

Phone: (+61) 3 9905 5902; Fax: (+61) 3 9905 5613

Email: [email protected]

Running title: Evolution of the eastern yellow robin

Word count: 6916 words of text (excluding tables, figure captions and literature cited)

This manuscript has 3 tables and 4 figures

We have/will archive our data on sequences, microsatellites and CHD markers, geographic

coordinates, climatic variables, sex, etc in Dryad and GenBank.

Page 2 of 101Evolution: For Review Only

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

3

ABSTRACT

Relationships among multi-locus genetic variation, geography and environment can

reveal how evolutionary processes affect genomes. We examined the evolution of an

Australian bird, the eastern yellow robin Eopsaltria australis, using mitochondrial (mtDNA)

and nuclear (nDNA) genetic markers, and bioclimatic variables. In southeastern Australia,

two divergent mtDNA lineages occur east and west of the Great Dividing Range,

perpendicular to latitudinal nDNA structure. We evaluated alternative scenarios to explain

this striking discordance in landscape genetic patterning. Stochastic mtDNA lineage sorting

can be rejected because the mtDNA lineages are essentially distinct geographically for >1500

km. Vicariance is unlikely: the Great Dividing Range is neither a current barrier nor was it at

the Last Glacial Maximum according to species distribution modeling; nuclear gene flow

inferred from coalescent analysis affirms this. Female philopatry contradicts known female-

biased dispersal. Contrasting mtDNA and nDNA demographies indicate their evolutionary

histories are decoupled. Distance-based redundancy analysis, in which environmental

temperatures explain mtDNA variance above that explained by geographic position and

isolation-by-distance, favors a non-neutral explanation for mitochondrial phylogeographic

patterning. Thus, observed mito-nuclear discordance accords with environmental selection on

a female-linked trait, such as mtDNA, mtDNA-nDNA interactions or genes on W-

chromosome, driving mitochondrial divergence in the presence of nuclear gene flow.


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

4

INTRODUCTION 1

Evolutionary processes (e.g. genetic drift, natural selection and gene flow) operate at 2

different times and places in species’ histories, differentially affecting genomes and phenotypes; 3

thus understanding their roles in speciation remains challenging (Coyne and Orr 2004; Price 4

2008; Butlin et al. 2009; Smadja and Butlin 2011). Discordant geographic patterns within and 5

among genomes in a species could result from coalescent variance (Lohse et al. 2010), 6

genetically localized selective sweeps (Bensch et al. 2006), life history traits (e.g. sex-biased 7

philopatry/dispersal; Turmelle et al. 2011), different modes of marker inheritance (and thus 8

effective population sizes), or differential marker behavior following secondary contact (Petit 9

and Excoffier 2009). A recent review (Toews and Brelsford 2012) found that mito-nuclear 10

discordance, a major difference in the patterns of differentiation between mitochondrial and 11

nuclear DNA, is a common phenomenon in animal systems. In the great majority of reviewed 12

cases, mito-nuclear discordance arose following allopatric divergence and secondary contact 13

(Toews and Brelsford 2012). This is expressed either in more structuring and/or narrower 14

geographic clines in mtDNA compared with nDNA (usually explained by nuclear introgression 15

and/or sex-biased asymmetries including dispersal, mating or offspring survival) or relatively 16

less structuring and/or wider geographic clines of mtDNA, usually explained by adaptive 17

introgression of mtDNA, demographic disparities including genetic drift, or sex-biased 18

asymmetries. Historic isolation and secondary contact can also result in coexistence of deeply 19

divergent mitochondrial lineages in sympatry in one panmictic population (Zink et al. 2008; 20

Webb et al. 2011; Hogner et al. 2012). However, in four cases, all involving avian systems, 21

strong mitochondrial but not nuclear structure has arisen in the absence of obvious geographic 22

isolation (Irwin et al. 2005; Cheviron and Brumfield 2009; Ribeiro et al. 2011; Spottiswoode et 23


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

5

al. 2011). Further, in two bird species, mitochondrial but not nuclear structure correlated with 24

environmental variation (Cheviron and Brumfield 2009; Ribeiro et al. 2011), suggesting that 25

mtDNA haplotypes might be differentially adapted to environmental conditions. Growing 26

evidence of selection on mtDNA (Ballard and Whitlock 2004; Bazin et al. 2006; Meiklejohn et 27

al. 2007) indicates that this process might be more common than recognized, although it is rarely 28

tested for in phylogeographic studies (Toews and Brelsford 2012). Links between mitochondrial 29

variation and energy metabolism (Mishmar et al. 2003; Tieleman et al. 2009) suggest that 30

mtDNA data should be routinely examined in relation to environmental variables in species 31

distributed over wide ranges of climatic conditions. Multilocus investigation of spatial genetic 32

structure in relation to environmental variation should lead to better understanding of 33

evolutionary forces shaping molecular variation. 34

The eastern yellow robin Eopsaltria australis (Passeriformes: Petroicidae) is a common 35

and widespread bird of mesic woodlands and forests of tropical to temperate eastern Australia 36

(Barrett et al. 2003). The species’ distribution spans 20º of latitude, associated with a wide range 37

of climates (Higgins and Peter 2002). Superimposed on this climatic gradient, the Great Dividing 38

Range, a mountain chain of modest elevation peaking at 2228 metres above sea level, runs the 39

entire length of the eastern coastline, and imposes climates that are generally drier and warmer 40

further inland (Bowman et al. 2010; Byrne et al. 2011). Comprehensive, continent-wide bird 41

surveys at 1º grids (1998-2002) showed E. australis to be evenly distributed and breeding 42

throughout the Great Dividing Range (Barrett et al. 2003). Dispersal of E. australis is limited 43

and significantly female-biased (Higgins and Peter 2002; Debus and Ford 2012; Harrisson et al. 44

2012), thus local adaptation is feasible (Schneider et al. 1999). So, too, is neutral divergence in 45

allopatry: aridification and cooling during Pleistocene glacial cycles have promoted divergence 46


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

6

and speciation of some forest and woodland Australian mesic organisms (Sunnucks et al. 2006; 47

Symula et al. 2008; Malekian et al. 2010; Byrne et al. 2011). 48

A preliminary molecular data set (Loynes et al. 2009) showed two highly divergent 49

mitochondrial ND2 haplogroups in E. australis (> 6% divergence), whereas one polymorphic 50

nuclear gene and two nuclear introns were not structured. Using mitochondrial COI Christidis et 51

al. (2011) confirmed this mitochondrial subdivision and suggested that, pending further 52

phylogeographic study, the haplogroups may comprise separate species. Notably, the 53

haplogroups did not correspond to the recognized subspecies, northern E. a. chrysorrhoa and 54

southern E. a. australis (Schodde and Mason 1999), suggesting a possibility of mito-nuclear 55

discordance. Interpreting strong mtDNA divergence as indicative of speciation is based on the 56

usually untested assumption that mitochondrial lineages will behave as if under neutral 57

evolution. Although in some cases this assumption might be valid (Harrison 1989; Zink and 58

Barrowclough 2008), simulations show that even weak selection for local adaptations can result 59

in strong phylogeographic structure of a uniparentally inherited locus, where divergent clades are 60

geographically localized and differently adapted (Irwin 2012). Here we report an intriguing 61

phylogeographic pattern in E. australis, where two major mtDNA lineages were distributed 62

roughly eastwards or westwards of the Great Dividing Range in southeastern Australia, 63

perpendicular to roughly latitudinal isolation-by-distance of some nuclear loci, and we test 64

rigorously whether selection or one or more neutral process better explains this mito-nuclear 65

discordance. We used multiple types of loci (mitochondrial ND2 gene, nuclear microsatellites 66

and introns) having different rates of evolution and patterns of inheritance (autosomal, Z-linked, 67

and female-limited W-linked), and climatic variables, applied over the full range of the species in 68

landscape genetic and genealogical frameworks. First, we tested if mitochondrial haplogroups 69


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

7

diverged by drift in allopatry followed by secondary contact (as in the majority of cases reviewed 70

by Toews and Brelsford 2012). Under this hypothesis, we expected to observe low habitat 71

suitability for E. australis on and along the Great Dividing Range (an agent of vicariance for 72

some taxa along the length of Australia's eastern seaboard; Byrne et al. 2011) during the Last 73

Glacial Maximum (LGM). We used species distribution modeling to test this. Second, we tested 74

if mitochondrial divergence in E. australis occurred in the presence of nuclear gene flow. Under 75

divergence with gene flow (Pinho and Hey 2010), some nuclear gene flow between mtDNA 76

haplogroups is expected. We tested this by fitting an isolation-with-migration (IMA) model to 77

nDNA data. Finally, we tested the hypothesis that female-linked environmental selection 78

(including selection on mtDNA and linked nuclear genes; Rand et al. 2004; Meiklejohn et al. 79

2007; Dowling et al. 2008; Tieleman et al. 2009; Shen et al. 2010; Moghadam et al. 2012) 80

resulted in mtDNA divergence in the presence of nuclear gene flow between populations defined 81

by their mtDNA haplogroups. This hypothesis includes the case where dispersal (which is 82

female-biased in E. australis) between environmentally-divergent areas is lower than dispersal 83

between similar areas. Under the female-linked selection hypothesis, decoupled evolutionary 84

history of mtDNA and nDNA is expected to result in different patterns of geographic structure 85

for these genomes, and mtDNA may show associations with environmental variation, after 86

spatial autocorrelation has been controlled for. We tested these predictions using 87

phylogeographic, landscape genetic and distance-based redundancy analyses. 88

METHODS 89

Modeling present and Last Glacial Maximum species distributions 90

To understand how current and late Pleistocene climatic variation in temperature and 91

humidity might have impacted the spatial distribution of E. australis (and its genetic variation), 92


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

8

and to test for a vicariant effect of the Great Dividing Range, we modeled current and LGM 93

distributions of this species using the machine-learning maximum entropy model implemented in 94

MaxEnt 3.3.3e (Phillips et al. 2006). Nineteen bioclimatic variables for the present and the LGM 95

(Busby 1991) at a resolution of 2.5 arc-minutes (~5 km2 grid) were obtained from the WorldClim 96

1.4 database (Hijmans et al. 2005). Layers were visualized and cropped to span from latitude 97

46°S to 5°S and longitude 130°S to 156°S using DIVA-GIS 7.4.0 (Hijmans 2009). Presence data 98

comprised 454 presence localities of E. australis in the Atlas of Living Australia 99

(http://www.ala.org.au). A set of 341 randomly chosen presence points (75%) was used to train 100

MaxEnt model under current climatic conditions, and a set of 113 points (25%) was used to test 101

the model. The fit of several preliminary MaxEnt models generated with different combinations 102

of variables was compared using the area under the receiver operating curve (AUC). The final 103

model of current species distribution was generated using six variables that maximized AUC: 104

max temperature of warmest month, min temperature of coldest month, mean diurnal range, 105

precipitation of wettest month, precipitation of driest month, and precipitation seasonality. LGM 106

distribution was modeled by projecting the final current model onto the set of these six variables 107

estimated for the LGM (see supplementary material S1 for details). 108

Sampling and molecular methods 109

Genomic DNA was extracted from 63 frozen tissues of E. australis spanning the species’ 110

entire geographic range (Australian National Wildlife Collection, CSIRO Ecosystems Sciences, 111

Canberra) following Kearns et al. (2009), and from 44 blood samples collected in the south and 112

north of the species’ range following Harrisson et al. (2012) (Fig. 1, supplementary material S2). 113

Individuals were screened for eight autosomal nuclear microsatellites, Cpi3, Cpi8 (Doerr 2005), 114

Escmu6 (Hanotte et al. 1994), HrU2 (Primmer et al. 1995), Pocc6 (Bensch et al. 1997), Smm7 115


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

9

(Maguire et al. 2006), Pgm1 and Pgm7 (Dowling et al. 2003) following Harrisson et al. (2012), 116

and for allele-length polymorphism in the female-specific CHD-W locus and its Z-chromosome 117

homolog CHD-Z (Griffiths et al. 1998; supplementary material S3). Pgm1 had suspected null 118

alleles, and Pgm7 was found to be Z-linked and extremely variable, thus neither were used in 119

analyses. The other six microsatellite loci were previously shown to be in Hardy-Weinberg 120

equilibrium and independently segregating in E. australis (Harrisson et al. 2012). 121

The mitochondrial (mtDNA) ND2 gene was amplified using primers L5215 (Hackett 122

1996) and H6313 (Sorenson et al. 1999) following Kearns et al. (2009) and sequenced 123

commercially in both directions using the amplification primers (Macrogen, Korea). Six nuclear 124

introns [aldolase B intron 4 (AB4), glyceraldehyde-3-phosphate dehydrogenase intron 11 125

(GAPDH), phosphenolpyruvate carboxykinase intron 9 (GTP), Z-linked muscle skeletal receptor 126

tyrosine kinase intron 3 (MUSK-I3), rhodopsin intron 2 (RI2) and transforming growth factor-β2 127

intron 5 (TGFβ2)] were amplified in 47 individuals (Table 1, supplementary material S2) using 128

published primers (Waltari and Edwards 2002; Sorenson et al. 2004; McCracken and Sorenson 129

2005; Vallender et al. 2007). All nuclear DNA (nDNA) PCRs contained MgCl2 (1.5 mM), 1x 130

KCl buffer, 0.125 µM each dNTP, 0.4 µM primers, 0.01 units/µl Taq polymerase and 0.2 mg/ml 131

BSA, and the microsatellite touchdown protocol above was applied. Nuclear loci were sequenced 132

using forward primers by the UK NERC Genepool sequencing facility (University of Edinburgh, 133

UK). For a subset of individuals representing all inferred haplotypes (see below), nuclear loci 134

were further sequenced in reverse direction to check that they did not represent sequencing 135

errors. 136

Chromatograms were edited and aligned in Geneious Pro 4.8.5 (Drummond et al. 2010). 137

The mitochondrial origin of ND2 sequences was supported by lack of stop-codons, 138


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

10

insertions/deletions or sequence ambiguities. For nuclear introns containing indels, only the first 139

fragments of sequence, where the majority of sequences were unambiguous, were analyzed 140

(Table 1), and individuals heterozygous for indels within these fragments were removed from 141

analysis (1 for GAPDH, 5 for GTP, 8 for TGFβ2). Gametic phases of heterozygous sequences 142

were reconstructed using the PHASE 2.1 algorithm (Stephens et al. 2001; Stephens and Donnelly 143

2003) implemented in DNASP v5.10.1 (Librado and Rozas 2009). Due to presence of many rare 144

alleles for some loci, haplotypes for small proportions of individuals (0.07 for GTP, 0.19 for 145

MUSK-I3, 0.01 for RI2, 0.18 for TGFβ2 and 0.16 for MC1R) were resolved with <80% 146

probability. We used all best pairs of haplotypes for analyses to avoid systematic bias in 147

estimates of population genetic parameters (Garrick et al. 2010), although incorrectly resolved 148

haplotypes could slightly influence estimates of linkage disequilibrium, recombination and 149

coalescence analyses (Balakrishnan and Edwards 2009). Introns were checked for recombination 150

in DNASP using the four-gamete test (Hudson and Kaplan 1985). 151

Multilocus phylogeography and spatial analysis of genetic variation 152

Relationships among haplotypes for mtDNA and nDNA loci were visualized on a 153

median-joining network (Bandelt et al. 1999) constructed in NETWORK v4.5.1.6 (www.fluxus-154

engineering.com). Owing to recombination, nuclear gene networks may not represent true 155

genealogical relationships, but are nonetheless useful for assessing allele clustering with respect 156

to mtDNA-defined lineages. Support for mitochondrial ND2 haplogroups and their divergence 157

times were estimated in BEAST 1.6.1 and 1.7 (Drummond and Rambaut 2007; Drummond et al. 158

2012) using all 100 ND2 sequences (supplementary material S4). STRUCTURE 2.3.3 (Pritchard et 159

al. 2000) and TESS 2.3.1 (Chen et al. 2007) were used to explore genotype substructuring in 160

microsatellites and autosomal introns (supplementary material S5). Spatial principal component 161


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

11

analysis (sPCA; Jombart et al. 2008) implemented through the functions spca and global.rtest in 162

the package ADEGENET v1.2-8 (Jombart 2008) for R (R Development Core Team 2011) was used 163

to investigate spatial patterns of genetic variation. SPCA maximizes the product of spatial 164

autocorrelation and genetic variance to summarize multivariate data on allele frequencies into a 165

few uncorrelated axes. This approach is powerful for detecting cryptic spatial patterns that are 166

not associated with high genetic variation (Jombart et al. 2008); unlike STRUCTURE or TESS, it 167

does not require Hardy-Weinberg or linkage equilibria, and can be applied to sequence data. We 168

used a Gabriel graph and introduced a small amount of random noise to coordinates of multiple 169

individuals collected at the same location (function jitter, factor=1, amount=0.1; the network was 170

robust to the addition of noise). Analyses were run separately for ND2, each of the sequenced 171

nuclear markers, and the dataset of six microsatellites (removing 9 individuals with missing 172

scores for a locus). 173

Multiple processes, such as isolation-by-distance and local selection (e.g. isolation-by-174

adaptation; Nosil et al. 2008), can simultaneously shape spatial structure of variation of different 175

molecular markers. To separate the effect of geographic position, which may reflect local 176

processes, from any underlying effect of isolation-by-distance, we applied redundancy analysis 177

(RDA, function rda, R package VEGAN; Dixon 2003; Oksanen et al. 2011). RDA is a constrained 178

ordination technique, coupled with an ANOVA-like permutation test that tests whether a given 179

predictor explains residual variation in response after fitting (conditioning on) other predictors 180

first. We tested whether geographic position can explain genetic variance of nuclear loci beyond 181

what can be explained by the isolation-by-distance effect. As the response, we used data on 182

presence or absence of each identified allele (microsatellites) or variant nucleotide at 183

polymorphic sites (sequences) in each individual, computed using a genind constructor and 184


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

12

centred and scaled using scaleGen (R package ADEGENET). The two predictors were geographic 185

position (latitude and longitude, analyzed together for simplicity of interpretation) and pairwise 186

geographic distance between locations. For RDA (and also distance-based RDA, below) the 187

information about an individual’s pairwise geographic distances to other individuals was reduced 188

to a few columns using the following procedure. First, geographic distances between locations, 189

initially computed with function rdist.earth (R package FIELDS), were expressed as a rectangular 190

matrix using principal coordinates of the neighbourhood matrix procedure (pcnm; Borcard and 191

Legendre 2002) implemented in function pcnm in VEGAN. Then this rectangular matrix was 192

reduced to ≤3 columns best explaining genetic variance. For that, we fitted all pcnm axes as 193

predictors of allele presence using RDA, and chose up to three axes significant at P<0.05 to 194

represent geographic distances in the final RDAs. This approach is more powerful than Mantel 195

tests, which require multivariate environmental data (such as latitude, longitude and 196

environmental variables) to be expressed as distances (Legendre and Fortin 2010). For the final 197

RDAs, we tested for the ability of each predictor to explain significant variance in allele 198

presence-absence data alone (marginal tests), and after fitting the other predictor first 199

(conditional tests). Significance was assessed with 999 permutations (of the rows and columns of 200

the predictor matrix for marginal tests, or of the rows and columns of the multivariate residual 201

matrix for conditional tests). A predictor variable that explained significant (P<0.05) genetic 202

variance in marginal and conditional tests was inferred to have major effect on genetic structure. 203

If both marginal tests were significant but conditional tests were not, both variables were inferred 204

to influence genetic variation. RDAs were performed on each sequenced nuclear marker, and 205

complete genotypes at six microsatellites. For mitochondrial ND2 sequences, effects of 206


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

13

geographic position and distance were explored using distance-based redundancy analysis 207

(dbRDA) as explained below. 208

Genetic diversity, genetic differentiation and selection 209

For each sequenced locus, we calculated haplotype diversity (Hd), nucleotide diversity 210

(π), pairwise Φst between haplogroups, Tajima's D (Tajima 1996) and Fu’s Fs (Fu 1997) tests for 211

selective neutrality using ARLEQUIN 3.11 (Excoffier et al. 2005). DNASP was used to test for 212

selection on ND2 itself using the McDonald-Kreitman (MK) test (McDonald and Kreitman 213

1991), two individuals with missing data for two sites were removed from this analysis (for 214

additional tests for selection see supplementary material S7). ARLEQUIN was used to compute 215

expected heterozygosity (He) and FST between haplogroups for six microsatellites 216

(supplementary material S6). Calculations were performed for all data, and for the two 217

populations defined by mtDNA haplogroups. 218

Isolation-with-migration analysis (IMA2) 219

We fitted the model of isolation-with-migration (Hey and Nielsen 2004) as implemented 220

in IMA2.0 (Hey and Nielsen 2007) to the multilocus dataset comprising the largest non-221

recombining fragments of four autosomal introns that had no signature of selection (AB4, 222

GAPDH, MUSK-I3 and TGFβ2), the size-variable Z-linked intron (CHD-Z) and six autosomal 223

microsatellites for the two populations defined by mtDNA haplogroup membership (A1 and B– 224

see Results, omitting three individuals from an isolated northernmost population). Under this 225

classification, any migration detected between haplogroups would indicate presence of nuclear 226

gene flow inconsistent with vicariant divergence. The fit of a full migration model was compared 227

to that of a no-migration sub-model using a likelihood ratio test. Parameter estimates were 228

converted to demographic units (number of years or individuals) using a generation time of 3.5 229


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

14

years (calculated as average age of reproductive females, assuming age of first reproduction of 1 230

year and lifespan of 6 years) and mutation rates of Lerner et al. (2011) (see supplementary 231

material S8 for details of analyses). 232

Distance-based redundancy analysis (dbRDA) 233

We used dbRDA (Legendre and Anderson 1999) implemented in the capscale function 234

in R-package VEGAN to test whether mitochondrial divergence occurred as a result of selection 235

acting upon a female-linked trait (broadly defined, i.e. mtDNA, mtDNA-nDNA interaction, or 236

genes on the W-chromosome). We examined whether the two bioclimatic variables that were 237

most informative in predicting the E. australis distribution (supplementary material S1) explain 238

mitochondrial variation above that explained by geographic position and/or distance between 239

individuals (Legendre and Fortin 2010). In this analysis, individuals were the units of 240

observation, and pairwise inter-individual TN93+G mitochondrial genetic distances [calculated 241

in MEGA5 (Tamura et al. 2011) excluding two individuals with missing sites; supplementary 242

material S4] were treated as information on multivariate response (and not as a single univariate 243

response variable as in Mantel tests; Geffen et al. 2004). Predictors (standardized to a zero mean 244

and unit variance) were two environmental variables: maximum temperature of warmest month 245

(max.temp) and precipitation of driest month (precipitation), and three geographic variables: 246

latitude, longitude and geographic distance between samples represented by the first significant 247

(P<0.05) pcnm axis (as explained above for RDA). The final dbRDA analyses included (1) 248

marginal tests, where the relationship between mtDNA genetic distances and each of the 249

predictors was analyzed separately, (2) conditional tests, where significance of environmental 250

variables (max.temp or precipitation) as predictors of mtDNA distances was assessed after first 251

fitting other variables one at a time or together, and (3) the reciprocal conditional tests, where 252


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

15

significance of geographic variables was similarly assessed by first fitting environmental 253

variables. Significance was assessed with 999 permutations. Environmental variables were 254

inferred to explain mtDNA variance over that explained by geographic position and distance if 255

marginal as well as conditional tests were significant. Significant association of mitochondrial 256

variation and climatic variables beyond its correlation with geographic position and distance 257

would provide strong evidence against hypotheses assuming neutral behavior of mtDNA. 258

RESULTS 259

Current and LGM species distribution models 260

Current and LGM species distribution models (supplementary material S1) predict the 261

northernmost (Wet Tropics) range of E. australis to be isolated from the remainder. Thus, for 262

this outlying region, vicariant effects might have acted at some time. In contrast, the species is 263

predicted under LGM and current conditions to have a continuous distribution on and along the 264

Great Dividing Range, except for a proportionally small area of habitat unsuitability (localized 265

glaciation) during the LGM in the mountains of southeastern Victoria and southern New South 266

Wales (Barrows et al. 2001). These predicted distributions provide no basis to expect vicariant 267

patterns due to the Great Dividing Range (arguing against allopatric divergence). Two variables 268

describing extremes of climatic conditions, maximum temperature of warmest month and 269

precipitation of driest month, explained the majority of variance in E. australis presence (50% 270

and 38% respectively in the final model). 271

MtDNA phylogeography 272

The 100 individuals sequenced for 1002 base pairs of ND2 revealed 102 polymorphic 273

sites (80 parsimony informative), defining 40 haplotypes (GenBank accession numbers 274

KC466740 - KC466839). All individuals fell into one of haplogroup A or B, between which 275


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

16

there was 6.6% net nucleotide divergence (Bayesian posterior probability=1, MCC BEAST tree, 276

supplementary material S4, and ND2 network, Fig. 2). Assuming neutral evolution with rates 277

similar to ND2 rate of the Hawaiian honeycreepers (Lerner et al. 2011), the haplogroups A and B 278

diverged in the early Pleistocene, ~1.5 (95% highest posterior density [HPD] 0.975-2.147) 279

million years ago. Five of 53 fixed differences between A and B represented non-synonymous 280

but biochemically conservative amino acid substitutions (Grantham 1974; supplementary 281

material S7). Haplogroup A was widespread north to south, and inland of the Great Dividing 282

Range in southeastern Australia (Fig. 1a, red points), while haplogroup B was principally 283

coastwards and restricted to southeastern Australia (Fig. 1a, black points). Two locations, 284

Tenterfield and Shelbourne (Fig. 1a) had individuals of both haplogroups. Haplogroup A 285

comprised haplogroups A1 (widespread except Wet Tropics) and northernmost A2 (three Wet 286

Tropics individuals; Bayesian posterior probability=1; net divergence 1%; MCC tree, 287

supplementary material S4), which diverged in the mid Pleistocene, ~212 (95% HPD 104-341) 288

thousand years ago, assuming neutral evolution. One of 10 fixed differences between A1 and A2 289

was non-synonymous. Consistent with these results, SPCA of mtDNA (Fig. 3) detected two 290

significant global structures (P<0.001), the first (mt-sPC1) reflecting sharp A versus B 291

subdivision, and the second (mt-sPC2) subdivision between A2 (Wet Tropics) and the rest 292

(A1+B). 293

Sex-linked CHD variation and population structure inferred from nuclear microsatellites 294

For 32 females and 64 males sequenced for ND2 and sexed by CHD variation, the 295

distribution of two W-linked (female-limited) alleles exactly matched mtDNA haplogroup 296

variation (allele 366 was restricted to haplogroup A, allele 362 to B), while the two Z-linked 297

alleles were widespread and unstructured with respect to haplogroups (supplementary material 298


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

17

S3). Ninety-two individuals were genotyped for six microsatellites. Genotypic analyses in 299

STRUCTURE and TESS suggested two genetic clusters, with a gradual north to south change of 300

individual cluster probabilities (Fig. 1b, Supplementary material S5). SPCA on microsatellites 301

detected two significant global structures (P<0.001; Fig. 3). The strongest structure (msat-sPC1) 302

showed a roughly north-south cline similar to that detected by STRUCTURE and TESS; the second, 303

subtle structure (msat-sPC2) showed a cline across the Great Dividing Range perpendicular to 304

msat-sPC1, directionally consistent with the major mitochondrial lineage subdivision (Fig. 3). 305

Isolation-by-distance (IBD) was a dominant process shaping microsatellite variation: geographic 306

distances alone could explain the variance in microsatellite frequencies (P=0.005, marginal RDA 307

test; Table 2), and remained significant even after geographic position was fitted first (P=0.017 308

conditional RDA test). In contrast, geographic position, although significant when fitted alone 309

(P=0.005, marginal RDA test), did not explain microsatellite structure after geographic distances 310

were controlled for (P=0.1, the reciprocal conditional RDA test). 311

Geographic variation of sequenced nuclear introns 312

Sequences were obtained from six nuclear introns in 31-47 individuals, which had two to 313

19 alleles (Table 1; GenBank accession numbers KC466694 - KC466739, KC466694 - 314

KC467012). No intron displayed clusters or allele-sharing concordant with mitochondrial 315

haplogroup designations (Fig. 2). Two genetic clusters detected by genotypic analysis of the 316

autosomal introns in TESS were consistent with north-south gradient in cluster membership 317

detected from microsatellites (Supplementary material S5). Significant spatial structure was 318

suggested by sPCA for GAPDH (P=0.053) and for MUSK-I3 (P=0.065; Fig 3) in which 319

southwestern individuals were somewhat differentiated from the rest; no structure was found for 320

the other sequenced introns (P>0.15; RI2 was not tested as only two haplotypes were present). 321


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

18

RDA (Table 2) showed that IBD as well as geographic position explained variation in GTP (both 322

P<0.05 on marginal tests, P>0.05 on conditional tests), geographic position explained variation 323

in AB4, GAPDH and MUSK-I3 beyond that explained by IBD (P<0.05 on marginal and 324

conditional tests); and IBD but not geographic position explained patterns in TFGβ2 (P<0.05 on 325

marginal test). 326

Population genetic variability, differentiation and selection 327

Within-haplogroup nucleotide diversity was lower for ND2 than that for four (of six) 328

nuclear loci (AB4, GAPDH, MUSK-I3 and TGFβ2; Table 1); π for nuclear loci ranged 0-0.0075 329

and 0.0003-0.0102, and for ND2 was 0.0013 and 0.0023 for A1 and B, respectively. Significant 330

differentiation between A1 and B was detected for ND2 (ΦST=0.968; P<0.001), inevitable under 331

reciprocal monophyly. Much lower but significant A1-B differentiation was also found for 332

nuclear introns AB4 (ΦST=0.047; P=0.03), GAPDH (ΦST=0.065; P=0.01), GTP (ΦST=0.034; 333

P=0.02) and for microsatellites (FST=0.029; P<0.001), but not for MUSK-I3, RI2 or TGFβ2 334

(P>0.05; Table 1). Significantly negative Tajima’s D and/or Fu’s Fs, indicative of population 335

expansion, purifying selection or genetic hitchhiking, were found in three loci with the lowest 336

nucleotide diversity (for A1 in ND2 and GTP, and for B in ND2, GTP and RI2). Further, a MK 337

test for selection on ND2 detected significant deficiency of non-synonymous polymorphism 338

between A1 and B (Fisher’s exact test P=0.047) with 50 synonymous and 5 non-synonymous 339

changes fixed between groups, compared to 30 synonymous and 10 non-synonymous 340

polymorphisms present within groups; MK tests between haplogroups A and B or between A1 341

and A2 were not significant (P>0.05). Thus, purifying selection is indicated to have acted upon 342

ND2; positive selection can be rejected (additional tests in supplementary material S7). 343

IMA2 analyses 344


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

19

Coalescent analysis of nuclear data in IMA2 could not definitively estimate some model 345

parameters for two populations defined by their mitochondrial haplogroup membership (details 346

in supplementary material S8), thus, and because assumptions of the model might not be met, 347

IMA2 results should be interpreted with caution. Based on all nuclear loci, populations A and B 348

were estimated to have diverged in the late Pleistocene or Holocene, ~7,000 (95%HPD 700-349

149,000) years ago. Posterior distributions of some Θ parameters did not reach zero, 350

nevertheless, each of them had a distinct peak within the range of prior parameter distributions. 351

Demographic estimates of population sizes from all nuclear data were NA=14 (95%HPD 3-86), 352

NB=13 (2-101) and NANC=186 (122-222) thousand individuals (for A, B and ancestral 353

populations, respectively). Although non-zero probabilities were associated with all prior values 354

of coalescent migration parameters, peak posterior distribution estimates suggested some 355

migration between populations. Migration estimates from all nuclear data converted to 356

population migration rate (2Nm) corresponded to forward-in-time gene movement of 6.2 (95% 357

HPD 0-32) genes per generation from A to B and 3.8 (0-22) genes per generation from B to A. 358

IMA2 was unable to conclusively reject the no-migration submodel based on a likelihood ratio 359

test. 360

Environmental, latitudinal and IBD effects associated with mtDNA diversity 361

DbRDA (Table 3) showed that maximum temperature of warmest month (max.temp), 362

precipitation of driest month, and latitude explain a significant amount of variance in mtDNA 363

genetic distances (marginal tests P<0.005), accounting for 51, 47 and 24% of genetic variance in 364

marginal tests, respectively, whereas longitude and geographic distance do not (P=0.59 and 365

P=0.072). So we excluded longitude from partial tests but retained geographic distance which 366

was close to significance. There was evidence of strong correlation of environmental variation 367


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

20

with mtDNA haplogroups: max.temp remained a significant predictor of genetic distances even 368

when other variables were fitted first alone or together (P=0.005; Table 3). Precipitation seems 369

somewhat less influential than max.temp: it remained significant when the three other variables 370

were fitted individually, or the two geographic variables were fitted first (P=0.005), but not when 371

max.temp plus the two geographic variables were fitted first (P=0.38; Table 3). Latitude 372

remained significant after first fitting the other variables alone or together (P<0.05; Table 3). 373

Overall, these results suggest that max.temp and latitude may play an important role in shaping 374

mitochondrial genetic structure of E. australis. Haplogroup A individuals occupied locations 375

with higher maximum temperatures of the warmest month and lower precipitation of the driest 376

month than those of haplogroup B (Fig. 4). 377

DISCUSSION 378

We attempted to clarify the evolutionary processes shaping a profound mito-nuclear 379

discordance in the Australian bird, the eastern yellow robin Eopsaltria australis, in which the 380

major patterns of spatial variation in mtDNA and nDNA are perpendicular. We addressed several 381

evolutionary scenarios, applying a series of landscape-genetic, spatial-genetic and genealogical 382

tests. 383

Deep mitochondrial divergence can arise by chance in a continuous population, when 384

dispersal and population size are low, due to the stochastic nature of the coalescent process 385

(Irwin 2002). However, such an effect is highly unlikely in E. australis, where the two 386

haplogroups have a sharp spatial division within this continuously distributed, common, 387

widespread species (Barrett et al. 2003). The haplogroups abut over >1500 km with substantial 388

potential for mixing, yet only in two sampled locations did they co-occur (Fig. 1). Random 389

processes also cannot explain correlation of mtDNA lineages with environmental variables. 390


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

21

Under a scenario of allopatric divergence followed by secondary contact, common 391

neutral explanations for more structured mtDNA than nDNA include incompletely sorted nuclear 392

lineages, nuclear introgression, and/or sex-biased asymmetries such as dispersal, mating or 393

offspring production/survival (Toews and Brelsford 2012). While the species distribution models 394

did indicate vicariance as a plausible explanation of the A1-A2 split at or after the Last Glacial 395

Maximum, there is little basis for vicariance in the range of Eopsaltria australis south of the Wet 396

Tropics. The species is common and widespread along the Great Dividing Range (Barrett et al. 397

2003), and its likely presence there during the LGM is inferred from our species distribution 398

models (although the LGM populations on parts of the Range could have been sparse, according 399

to relatively low probability of occurrence; supplementary material S1). Geographic 400

inconsistencies between mtDNA and nDNA structure rule out incomplete lineage sorting as a 401

cause of discordance (Toews and Brelsford 2012). Historical lack of gene flow across the Great 402

Dividing Range but not in other directions would be expected to structure both mitochondrial 403

and microsatellite variation: yet no geographically distinct genetic clusters were detected for 404

microsatellites. Instead, the major spatial pattern for microsatellites is latitudinal isolation-by-405

distance, with only mild IBD across the Great Dividing Range. This suggests that distance, rather 406

than a barrier, has structured nuclear variation. This, and inference of some nuclear gene 407

exchange from coalescent analysis in IMA2 (tempered by parameter uncertainty), suggests that 408

some nuclear gene flow between haplogroups has occurred. Different demographic histories 409

were inferred from mtDNA and nDNA data: significant Tajima’s D and Fu’s Fs neutrality tests, 410

suggestive of population expansion, purifying selection or hitchhiking, were detected for mtDNA 411

and two least variable nuclear loci, not for any of the more variable nuclear markers. Population 412

expansion would be expected to yield generally consistent signals across different markers, thus 413


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

22

non-neutral evolution or hitchhiking of some loci appears more likely, which was also supported 414

by the MK test for ND2. This suggests that evolutionary histories of mitochondrial and nuclear 415

genomes are decoupled. The same conclusion arises from contrasting A-B divergence times 416

estimated from the two genomes while accounting for coalescent variance: the late Pleistocene-417

Holocene divergence of ~7,000 (700-149,000) years ago estimated by IMA2 from nuclear data 418

(Supplementary material S8) is much more recent than the early Pleistocene mitochondrial 419

lineage divergence of ~1.5 (1-2.1) million years ago estimated by BEAST from ND2 420

(Supplementary material S4). Although we note that the divergence time estimates of BEAST and 421

IMA2 are not strictly comparable because IMA2 incorporated migration into the model, the 422

difference represents evidence that mtDNA and nDNA cannot both be responding solely to 423

neutral processes. 424

Very strong female philopatry unrelated to selection, and selection correlated with 425

mtDNA, are both expected to limit female but not male movement and gene flow across 426

populations and thus to impact mtDNA more and nDNA less than expected in the absence of 427

these effects. However, the female philopatry needed to explain observed patterns in the data 428

contrasts strongly with what is known about the biology of E. australis: field observations 429

(Debus and Ford 2012) and population genetic data (Harrisson et al. 2012) show that females 430

disperse significantly further than males (at least ~7 km and ~1 km, respectively; Debus and 431

Ford 2012). It is possible that females prefer to settle in habitats similar to their natal habitats, 432

whereas males are less discriminating (Tonnis et al. 2005). However, behavior where females 433

were so unable or unwilling to disperse or settle between adjacent haplogroup ranges over >1500 434

km of contact seems likely to be linked to some major fitness advantage, implying female-linked 435

selection. 436


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

23

If we reject the selectively neutral hypotheses as unparsimonious, it remains to consider 437

possibilities involving female/mtDNA-linked selection. Geographically localized haplogroups 438

and very shallow coalescent times within haplogroups compared to between-haplogroups 439

observed in E. australis are consistent with simulation scenarios under local selection and low to 440

moderate dispersal (Irwin 2012). The correlation of environmental variables with the spatial 441

distribution of mtDNA variation, after controlling for IBD and geographic position (dbRDA 442

results, Table 3) is consistent with environmental selection. The most informative bioclimatic 443

variables with regards to species occurrence (maximum temperature of the warmest month and, 444

to a lesser extent, precipitation of the driest month) explained significant mitochondrial 445

patterning beyond that explained by latitude and IBD, suggesting that these climatic factors may 446

be associated with female-linked selection. 447

These results, coupled with the arguments against selectively neutral explanations, lead 448

us to propose that the most likely solution to the geographically perpendicular mito-nuclear 449

discordance in E. australis is female-linked selection (broadly-defined), taken here to include 450

selection on one or more of mtDNA, nuclear-encoded genes for proteins involved in oxidative 451

phosphorylation, nuclear-encoded genes for mitochondrial proteins, joint mito-nuclear genotype 452

or W-linked parts of the genome (Rand et al. 2004; Meiklejohn et al. 2007; Dowling et al. 2008; 453

Tieleman et al. 2009; Shen et al. 2010; Moghadam et al. 2012), female biology including biased 454

habitat selection (as in Tonnis et al. 2005) and ‘divergence hitchhiking’ (gene flow reduced as a 455

side effect of strong divergent selection on genes involved in local adaptation; Via 2012). 456

Nuclear gene flow within E. australis appears to be structured by isolation-by-distance along the 457

gradients in two perpendicular directions: latitudinal (main microsatellite structure, msat-sPC1; 458

Fig. 3) and across the Great Dividing Range (the cryptic second structure in the microsatellites, 459


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

24

msat-sPC2, directionally consistent with distribution of TESS clusters from analysis of sequenced 460

introns assuming three genetic clusters [supplementary material S5], which could be a signature 461

of gene flow impeded by arrested dispersal or survival of females across the Range). The 462

inference of isolation-by-distance is also supported by congruent geographic structure in color of 463

the upper tail-coverts, which changes from bright yellow to olive-green from north to south and 464

east to west in E. australis (Ford 1979). Keast (1958) and Ford (1979) suggested that this color 465

variation arose due to local ecotypic selection where brighter yellow is preferred in darker 466

environments where it would be more useful for signaling. Thus selective pressures driving 467

evolution of female-linked traits and color variation could be different and their evolution 468

unlinked. Overall, dramatic mito-nuclear discordance in E. australis suggests that natural 469

selection constrains mitochondrial but not selectively neutral nuclear gene flow along 470

environmental gradients. 471

More research is needed to understand the mechanisms through which putative selection 472

on female-linked traits in E. australis operates. Passerine birds from hotter and drier 473

environments have lower basal metabolic rates and evaporative water loss (Tieleman et al. 2003; 474

Williams and Tieleman 2005). Selection on mitochondrial genotypes can drive such adaptive 475

responses because the mitochondrial genome plays an important role in the regulation of energy 476

metabolism in birds (Tieleman et al. 2009). For E. australis, the MK tests suggested purifying 477

selection on ND2, thus positive selection appears not to impact ND2 directly, but could act upon 478

other mtDNA-encoded genes, or indirectly on nDNA-encoded genes that produce structural 479

proteins imported into the mitochondria or on the joint mito-nuclear genotype, mito-nuclear 480

hybrids being selected against in the contact zone between two haplogroups (Dowling et al. 481

2008; Ballard and Melvin 2010). Of all reviewed cases of mito-nuclear discordance (Toews and 482


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

25

Brelsford 2012), only two other studies, both involving birds, report association between mtDNA 483

and environmental variation in the presence of nuclear gene flow when evidence for previous 484

isolation is lacking. In the Rufous-collared Sparrow Zonotrichia capensis, mitochondrial but not 485

nuclear gene flow was significantly reduced along elevational gradients, suggesting that 486

mitochondrial haplotypes could be locally adapted (Cheviron and Brumfield 2009). A 487

relationship between a gradient of aridity and mitochondrial but not nuclear or morphological 488

diversity has been reported in the South African arid-zone endemic Karoo scrub-robin 489

Cercotrichas coryphaeus raising the possibility of common mechanisms for adaptations to 490

extreme environmental conditions (Ribeiro et al. 2011). Due to linked inheritance with the W-491

chromosome, mtDNA may also reflect female-specific selection on W-linked genes, as has been 492

demonstrated for the evolution of gene expression on the W-chromosome (Moghadam et al. 493

2012). Thus it may not be coincidental that all three studies (Cheviron and Brumfield 2009; 494

Ribeiro et al. 2011; the present study) reporting association between mtDNA and environmental 495

variation in the presence of nuclear gene flow are from birds. Whatever the mechanism, any 496

environmental selection acting on female-linked traits in E. australis would strongly inhibit the 497

sympatry of mitochondrial lineages across a climatic gradient associated with the Great Dividing 498

Range. 499

Conclusions 500

Our study of E. australis benefitted from a combination of landscape genetic, spatial 501

genetic and coalescent analyses of multiple locus types and environmental variables in building 502

an integrated understanding of evolutionary processes operating at different scales. It also 503

presents a striking example of how geographically structured mtDNA diversity can unreliably 504

reflect species’ evolutionary history when neutrality is assumed and not thoroughly tested. It is 505


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

26

beneficial to incorporate tests for the effect on genetic variation of geographic position and 506

environmental variation, in addition to IBD, into phylogeographic studies, and to try and tease 507

apart their almost inevitable correlations. Correlation between mitochondrial genetic variation 508

and geographic position, when distance is controlled for, warrants additional exploration of 509

potential selection pressures driving evolution of female-linked traits. Discovery of two highly 510

divergent haplogroups apparently under strong environmental selection within a putatively 511

continuous population of E. australis (possibly a common pattern for species whose range spans 512

environmental gradients) provides another example of divergence with gene flow (Pinho and 513

Hey 2010), which has potentially profound implications for management and taxonomy. The 514

evidence for female-linked environmental selection implies that mtDNA haplogroups are not 515

ecologically exchangeable sensu Crandall et al. (2000). For example, translocation among 516

haplogroup regions may result in negative fitness consequences, and on the evidence here would 517

be an inadvisable management strategy. Our study illustrates how the use of mtDNA as a 518

barcoding tool to define species or lineages (Baker et al. 2009) can be severely compromised 519

without the knowledge to appreciate the role of mtDNA-correlated selection in such units. 520

ACKNOWLEDGEMENTS 521

Funding was provided by the Australian Research Council Linkage Grant (LP0776322), 522

the Victorian Department of Sustainability and Environment (DSE), Museum of Victoria, 523

Victorian Department of Primary Industries, Parks Victoria, North Central Catchment 524

Management Authority, Goulburn Broken Catchment Management Authority, CSIRO 525

Ecosystem Sciences, and the Australian National Wildlife Collection Foundation. NA was 526

funded by a Monash University Faculty of Science Dean’s Postgraduate Scholarship and Birds 527

Australia with additional support from the Holsworth Wildlife Research Endowment. Blood 528


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

27

samples from Victoria were collected under DSE permits (number 10004294 under the Wildlife 529

Act 1975 and the National Parks Act 1975, and number NWF10455 under section 52 of the 530

forest Act 1958). We thank Alan Lill, Naoko Takeuchi and other collectors of specimens and all 531

agencies who granted permission to collect specimens, Robert Palmer for curatorial assistance, 532

Suzanne Metcalfe for help with mitochondrial sequencing, and Katherine Harrisson for help with 533

nuclear sequencing. Computationally intensive analyses (STRUCTURE) were performed on the 534

Monash Sun Grid courtesy of Monash eResearch. We thank the UK NERC Genepool facility for 535

sequencing support. We are grateful to Darren Irwin, John Endler, Evolution editorial input 536

including that of Kenneth Petren, three anonymous reviewers for their comments on the earlier 537

drafts of the manuscript, to Thibaut Jombart for advice on analyses of spatial genetic structures, 538

Gaynor Dolman, Hugh Ford and Peter Teske for helpful discussions. 539

540


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

28

Table 1. Descriptive statistics for all samples and members of distinct mitochondrial clades for sequenced loci (Wet tropics individuals 541

belonging to clade A2 were not sequenced for nDNA). N ind- number of individuals, N alleles- Number of alleles, bp- base pairs, S-542

polymorphic sites, H-number of haplotypes, Hd- haplotype diversity (expected heterozygosity), π- nucleotide diversity, Rm- minimum 543

number of recombinant events, n/s- not significant (P>0.05), n/a- not applicable. Significant ΦST -values are in bold. 544

Locus Data N

ind

N

alleles

Length,

bp

Indels,

bp

S H Hd π Fu's Fs, P Tajima's D,

P

ΦST A1

vs B

Rm

AB4 autosomal

intron

All 44 88 196 1 2 3 0.447 0.0025 0.613 n/s 0.343 n/s 0.047

P=0.027

0

Clade A 20 40 196 1 1 3 0.405 0.0022 -0.078 n/s 0.565 n/s

Clade B 22 44 195 0 2 3 0.519 0.0030 0.609 n/s 0.474 n/s

GAPDH autosomal

intron

All 46 92 317 0 8 10 0.818 0.0047 -2.217 n/s -0.147 n/s 0.065

P=0.009

1

Clade A 21 42 317 0 6 8 0.866 0.0054 -1.263 n/s 0.577 n/s

Clade B 23 46 317 0 6 6 0.723 0.0036 -0.742 n/s -0.432 n/s

GTP autosomal

intron

All 42 84 456 1 8 10 0.410 0.0011 -8.403

P<0.001

-1.942

P<0.001

0.034

P=0.018

0

Clade A 19 38 456 0 4 5 0.292 0.0007 -3.671

P=0.001

-1.630

P=0.016


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

29

Clade B 21 42 456 1 7 8 0.535 0.0018 -4.128

P=0.004

-1.502

P=0.056

MUSK-

I3

Z-linked

intron

All 31 51 519 0 21 15 0.810 0.0078 -2.240 n/s -0.416 n/s 0.061

P=0.072

3

Clade A 20 32 519 0 17 9 0.673 0.0065 -0.114 n/s -0.671 n/s

Clade B 11 17 519 0 18 10 0.919 0.0102 0.516 n/s -0.033 n/s

RI2 autosomal

intron

All 47 94 292 0 2 2 0.021 0.0002 -1.421 n/s -1.386 n/s -0.001

n/s

0

Clade A 22 44 292 0 0 1 0.000 0.0000 0 n/a 0 n/s

Clade B 23 46 292 0 2 2 0.044 0.0003 -0.783 n/s -1.473

P=0.03

TGFβ2 autosomal

intron

All 39 78 595 1 24 19 0.870 0.0070 -3.420 n/s -0.443 n/s 0.007

n/s

4

Clade A 20 40 595 0 23 16 0.868 0.0075 -3.465 n/s -0.599 n/s

Clade B 17 34 594 1 12 9 0.857 0.0066 0.553 n/s 0.935 n/s

ND2 mt coding All 100 100 1002 0 102 40 0.908 0.0334 2.785 n/s 2.303

P=0.01

0.968

P<0.001

n/a

mt coding Clades 97 97 1002 0 94 38 0.902 0.0332 3.382 n/s 2.704


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

30

A1+B P=0.002

Clade A1 53 97 1002 0 26 20 0.738 0.0017 -16.418

P<0.001

-2.276

P<0.001

Clade A2 3 97 1002 0 2 2 0.667 0.0013 - -

Clade B 44 97 1002 0 16 18 0.901 0.0023 -10.728

P<0.001

-1.158 n/s

545


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

31

Table 2. Redundancy (RDA) analysis: separating the effect of IBD and geographic position (latitude and longitude analyzed together) 546

on multivariate distribution of allele frequencies (presence-absence) for six microsatellites and nuclear loci. Distance- geographic 547

distance (represented as a rectangular matrix of pcnm axes), Geography - joint effect of latitude and longitude, N pcnm axes- number 548

of significant pcnm axes representing geographic distance. Marginal tests show the relationship between allele frequencies and either 549

distance or geographic position alone. Conditional tests show the same relationships, but having first fitted the alternative variable 550

(geographic position and distance, respectively) as the covariates in the analyses. Significant P-values (P<0.05) are in bold. 551

Marginal tests Conditional tests

Locus (Loci)

N pcnm

axes

Predictor

Variable tested F P

% variance

explained

Variables

fitted first F P

% variance

explained

Microsatellites 3 Distance 2.62 0.005 9.1 Geography 1.4 0.017 5.2

Geography 3.15 0.005 7.3 Distance 1.33 0.107 3.3

AB4 2 Distance 0.41 0.87 1.9 Geography 1.58 0.19 7.5


GAPDH 3 Distance 1.96 0.015 12.3 Geography 1.58 0.08 10.6


GTP 3 Distance 1.8 0.015 12.4 Geography 1.53 0.08 11.3



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

32

MUSK-I3 2 Distance 2.02 0.037 18.3 Geography 1.69 0.07 16.9


TGFβ2 2 Distance 1.6 0.03 12.0 Geography 1.53 0.07 12.2



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

33

Table 3. Results of dbRDA analysis: tests of relationship between multivariate mtDNA genetic 552

variation (defined by pairwise mtDNA genetic distances) and several predictor variables 553

(geographic and environmental). The first row for each tested predictor is a marginal test for that 554

variable, the next four rows are conditional (partial) tests where variables given in second 555

column are partialed out (fitted before the variable being tested). Last column indicates 556

percentage of multivariate genetic variation explained by variable being tested. Max.temp- 557

maximum temperature of warmest month, Precipitation- precipitation of driest month, All- all 558

variables except the one being tested. Longitude was not a significant predictor of genetic 559

distance in the marginal test (F=0.34, P=0.59), and was not used for partial tests. Significant P-560

values (P<0.05) are in bold. 561

Predictor variable tested Variables fitted first

in partial tests

F P(>F) % variance

explained

Geographic distance 3.3 0.072 3.3

Latitude 26.6 0.005 21.9

Max.temp 0.6 0.42 0.6

Precipitation 1.6 0.24 1.6

All 7.5 0.01 7.5

Latitude 31.1 0.005 24.5


Max.temp 12.3 0.005 11.5


All 11.6 0.005 11.1

Max.temp 100 0.005 51.1



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

34

Latitude 70.6 0.005 42.6

Geog. distance + latitude 57.8 0.005 38.1


All 33.2 0.005 26.3



Latitude 47.6 0.005 33.4


Max.temp 18.3 0.005 16.1

All 0.7 0.38 0.8

562


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

35

LITERATURE CITED 563

Baker, A. J., E. S. Tavares, and R. F. Elbourne. 2009. Countering criticisms of single 564

mitochondrial DNA gene barcoding in birds. Molecular Ecology Resources 9:257-268. 565

Balakrishnan, C. N., and S. V. Edwards. 2009. Nucleotide variation, linkage disequilibrium and 566

founder-facilitated speciation in wild populations of the zebra finch (Taeniopygia 567

guttata). Genetics 181:645. 568

Ballard, J. W. O., and R. G. Melvin. 2010. Linking the mitochondrial genotype to the organismal 569

phenotype. Mol Ecol 19:1523-1539. 570

Ballard, J. W. O., and M. C. Whitlock. 2004. The incomplete natural history of mitochondria. 571

Mol Ecol 13:729-744. 572

Bandelt, H.-J., P. Forster, and A. Roehl. 1999. Median-joining networks for inferring 573

intraspecific phylogenies. Molecular Biology & Evolution 16:37-48. 574

Barrett, G., A. Silcocks, S. Barry, R. Cunningham, and R. Poulter. 2003. The New Atlas of 575

Australian Birds. Royal Australasian Ornithologists Union, Melbourne. 576

Barrows, T. T., J. O. Stone, L. K. Fifield, and R. G. Cresswell. 2001. Late Pleistocene glaciation 577

of the Kosciuszko Massif, Snowy Mountains, Australia. Quat. Res. 55:179-189. 578

Bazin, E., S. Glemin, and N. Galtier. 2006. Population size does not influence mitochondrial 579

genetic diversity in animals. Science 312:570-572. 580

Bensch, S., D. E. Irwin, J. H. Irwin, L. Kvist, and S. Akesson. 2006. Conflicting patterns of 581

mitochondrial and nuclear DNA diversity in Phylloscopus warblers. Mol Ecol 15:161-582

171. 583

Bensch, S., T. Price, and J. Kohn. 1997. Isolation and characterization of microsatellite loci in a 584

Phylloscopus warbler. Mol Ecol 6:91-92. 585


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

36

Borcard, D., and P. Legendre. 2002. All-scale spatial analysis of ecological data by means of 586

principal coordinates of neighbour matrices. Ecol Model 153:51-68. 587

Bowman, D. M. J. S., G. K. Brown, M. F. Braby, J. R. Brown, L. G. Cook, M. D. Crisp, F. Ford, 588

S. Haberle, J. Hughes, Y. Isagi, L. Joseph, J. McBride, G. Nelson, and P. Y. Ladiges. 589

2010. Biogeography of the Australian monsoon tropics. J Biogeogr 37:201-216. 590

Busby, J. R. 1991. BIOCLIM a bioclimatic analysis and prediction system. Pp. 64–68 in C. R. 591

Margules, and M. P. Austin, eds. Nature Conservation: Cost Effective Biological Surveys 592

and Data Analysis. CSIRO, Canberra. 593

Butlin, R., D. Schluter, and J. R. Bridle. 2009. Speciation and patterns of diversity. Cambridge 594

University Press, Cambridge, UK. 595

Byrne, M., D. Steane, L. Joseph, D. Yeates, G. Jordan, D. Crayn, K. Aplin, D. Cantrill, L. G. 596

Cook, M. D. Crisp, J. S. Keogh, J. Melville, C. Moritz, N. Porch, J. M. K. Sniderman, P. 597

Sunnucks, and P. H. Weston. 2011. Decline of a biome: evolution, contraction, 598

fragmentation, extinction and invasion of the Australian mesic zone biota. J Biogeogr In 599

press. 600

Chen, C., E. Durand, F. Forbes, and O. Francois. 2007. Bayesian clustering algorithms 601

ascertaining spatial population structure: a new computer program and a comparison 602

study. Mol Ecol Notes 7:747-756. 603

Cheviron, Z. A., and R. T. Brumfield. 2009. Migration-selection balance and local adaptation of 604

mitochondrial haplotypes in rufous-collared sparrows (Zonotrichia capensis) along an 605

elevational gradient. Evolution 63:1593-1605. 606


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

37

Christidis, L., M. Irestedt, D. Rowe, W. E. Boles, and J. A. Norman. 2011. Mitochondrial and 607

nuclear DNA phylogenies reveal a complex evolutionary history in the Australasian 608

robins (Passeriformes: Petroicidae). Mol Phylogenet Evol 61:726–738. 609

Coyne, J., and H. A. Orr. 2004. Speciation. Sinauer, Sunderland, Massachusetts. 610

Crandall, K. A., O. R. P. Bininda-Emonds, G. M. Mace, and R. K. Wayne. 2000. Considering 611

evolutionary processes in conservation biology. Trends Ecol Evol 15:290-295. 612

Debus, S. J. S., and H. A. Ford. 2012. Responses of Eastern Yellow Robins Eopsaltria australis 613

to translocation into vegetation remnants in a fragmented landscape. Pac Conserv Biol in 614

press. 615

Dixon, P. 2003. VEGAN, a package of R functions for community ecology. J Veg Sci 14:927-616

930. 617

Doerr, E. D. 2005. Characterization of microsatellite loci in the brown treecreeper (Climacteris 618

picumnus) and cross-species amplification in the white-throated treecreeper (Cormobates 619

leucophaeus). Mol Ecol Notes 5:654-656. 620

Dowling, D., U. Friberg, and J. Lindell. 2008. Evolutionary implications of non-neutral 621

mitochondrial genetic variation. Trends Ecol Evol 23:546-554. 622

Dowling, D. K., G. J. Adcock, and R. A. Mulder. 2003. Novel polymorphic microsatellite 623

markers for paternity analysis in the red-capped robin (Petroica goodenovii: Aves). Mol 624

Ecol Notes 3:517-519. 625

Drummond, A., and A. Rambaut. 2007. BEAST: Bayesian evolutionary analysis by sampling 626

trees. BMC Evol Biol 7:214. 627


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

38

Drummond, A. J., B. Ashton, S. Buxton, M. Cheung, A. Cooper, J. Heled, M. Kearse, R. Moir, 628

S. Stones-Havas, S. Sturrock, T. Thierer, and A. Wilson. 2010. Geneious v5.1. Available 629

from http://www.geneious.com. 630

Drummond, A. J., M. A. Suchard, D. Xie, and A. Rambaut. 2012. Bayesian phylogenetics with 631

BEAUti and the BEAST 1.7. Mol Biol Evol 29:1969-1973. 632

Excoffier, L., G. Laval, and S. S. 2005. Arlequin ver. 3.0: An integrated software package for 633

population genetics data analysis. Evolutionary Bioinformatics Online 1:47-50. 634

Ford, J. C. 1979. Speciation or Subspeciation in the Yellow Robins? Emu 79:103-106. 635

Fu, Y.-X. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking 636

and backgroud selection. Genetics 147:915-925. 637

Garrick, R. C., P. Sunnucks, and R. J. Dyer. 2010. Nuclear gene phylogeography using PHASE: 638

dealing with unresolved genotypes, lost alleles, and systematic bias in parameter 639

estimation. BMC Evol Biol 10:118. 640

Geffen, E. L. I., M. J. Anderson, and R. K. Wayne. 2004. Climate and habitat barriers to 641

dispersal in the highly mobile grey wolf. Mol Ecol 13:2481-2490. 642

Grantham, R. 1974. Amino acid difference formula to help explain protein evolution. Science 643

185:862-864. 644

Griffiths, R., M. C. Double, K. Orr, and R. J. G. Dawson. 1998. A DNA test to sex most birds. 645

Mol Ecol 7:1071-1075. 646

Hackett, S. J. 1996. Molecular Phylogenetics and Biogeography of Tanagers in the Genus 647

Ramphocelus (Aves). Mol Phylogenet Evol 5:368-382. 648


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

39

Hanotte, O., C. Zanon, A. Pugh, C. Greig, A. Dixon, and T. Burke. 1994. Isolation and 649

characterization of microsatellite loci in a passerine bird: The reed bunting Emberiza 650

schoeniclus. Mol Ecol 3:529-530. 651

Harrison, R. G. 1989. Animal mitochondrial DNA as a genetic marker in population and 652

evolutionary biology. Trends Ecol. Evol. 4(1):6-11. 653

Harrisson, K., A. Pavlova, J. N. Amos, N. Takeuchi, A. Lill, J. Q. Radford, and P. Sunnucks. 654

2012. Fine-scale effects of habitat loss and fragmentation despite large-scale gene flow 655

for some regionally declining woodland bird species. Landsc Ecol 27:813-827. 656

Hey, J., and R. Nielsen. 2004. Multilocus methods for estimating population sizes, migration 657

rates and divergence time, with applications to the divergence of Drosophila 658

pseudoobscura and D-persimilis. Genetics 167:747-760. 659

Hey, J., and R. Nielsen. 2007. Integration within the Felsenstein equation for improved Markov 660

chain Monte Carlo methods in population genetics. Proceedings of the National Academy 661

of Sciences 104:2785-2790. 662

Higgins, P. J., and J. M. Peter, eds. 2002. Handbook of Australian, New Zealand and Antarctic 663

Birds. Volume 6: Pardalotes to Shrike-thrushes. Oxford University Press, Melbourne. 664

Hijmans, R. J., S. E. Cameron, J. L. Parra, P. G. Jones, and A. Jarvis. 2005. Very high resolution 665

interpolated climate surfaces for global land areas. International Journal of Climatology 666

25:1965–1978. 667

Hijmans, R. J. 2009. DIVA-GIS, version 7. A geographic information system for the analysis of 668

biodiversity data. Available at: http://www.diva-gis.org. 669

Hogner, S., T. Laskemoen, J. T. Lifjeld, J. Porkert, O. Kleven, T. Albayrak, B. Kabasakal, and 670

A. Johnsen. 2012. Deep sympatric mitochondrial divergence without reproductive 671


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

40

isolation in the common redstart Phoenicurus phoenicurus. Ecology and Evolution doi: 672

10.1002/ece3.398. 673

Hudson, R. R., and N. L. Kaplan. 1985. Statistical properties of the number of recombination 674

events in the history of a sample of DNA sequences. Genetics 111:147. 675

Irwin, D. E. 2002. Phylogeographic breaks without geographic barriers to gene flow. Evolution 676

56:2383-2394. 677

Irwin, D. E., S. Bensch, J. H. Irwin, and T. D. Price. 2005. Speciation by distance in a ring 678

species. Science 307:414-416. 679

Irwin, D. E. 2012. Local adaptation along smooth ecological gradients causes phylogeographic 680

breaks and phenotypic clustering. The American Naturalist 180:35-49. 681

Jombart, T. 2008. Adegenet: a R package for the multivariate analysis of genetic markers. 682

Bioinformatics 24:1403-1405. 683

Jombart, T., S. Devillard, A. B. Dufour, and D. Pontier. 2008. Revealing cryptic spatial patterns 684

in genetic variability by a new multivariate method. Heredity 101:92-103. 685

Kearns, A. M., L. Joseph, S. V. Edwards, and M. C. Double. 2009. Inferring the phylogeography 686

and evolutionary history of the splendid fairy-wren Malurus splendens from 687

mitochondrial DNA and spectrophotometry. J Avian Biol 40:7-17. 688

Legendre, P., and M. J. Anderson. 1999. Distance-based redundancy analysis: Testing 689

multispecies responses in multifactorial ecological experiments. Ecol Monogr 69:1-24. 690

Legendre, P., and M.-J. Fortin. 2010. Comparison of the Mantel test and alternative approaches 691

for detecting complex multivariate relationships in the spatial analysis of genetic data. 692

Molecular ecology resources 10:831-844. 693


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

41

Lerner, H. R. L., M. Meyer, H. F. James, M. Hofreiter, and R. C. Fleischer. 2011. Multilocus 694

Resolution of Phylogeny and Timescale in the Extant Adaptive Radiation of Hawaiian 695

Honeycreepers. Curr Biol 21:1838-1844. 696

Librado, P., and J. Rozas. 2009. DnaSP v5: a software for comprehensive analysis of DNA 697

polymorphism data. Bioinformatics 25:1451. 698

Lohse, K., B. Sharanowski, and G. N. Stone. 2010. Quantifying the Pleistocene history of the 699

oak gall parasitoid Cecidostiba fungosa using twenty intron loci. Evolution 64:2664-700

2681. 701

Loynes, K., L. Joseph, and J. S. Keogh. 2009. Multi-locus phylogeny clarifies the systematics of 702

the Australo-Papuan robins (Family Petroicidae, Passeriformes). Mol Phylogenet Evol 703

53:212-219. 704

Maguire, G. S., P. J. Guay, and R. A. Mulder. 2006. Isolation and characterization of 705

microsatellite markers in the southern emu-wren (Stipiturus malachurus: Aves). Mol 706

Ecol Notes 6:422-424. 707

Malekian, M., S. J. B. Cooper, J. A. Norman, L. Christidis, and S. M. Carthew. 2010. Molecular 708

systematics and evolutionary origins of the genus Petaurus (Marsupialia: Petauridae) in 709

Australia and New Guinea. Mol Phylogenet Evol 54:122-135. 710

McCracken, K. G., and M. D. Sorenson. 2005. Is homoplasy or lineage sorting the source of 711

incongruent mtDNA and nuclear gene trees in the stiff-tailed ducks (Nomonyx-Oxyura)? 712

Syst Biol 54:35. 713

McDonald, J. H., and M. Kreitman. 1991. Adaptive protein evolution at the Adh locus in 714

Drosophila. Nature 351:652-654. 715


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

42

Meiklejohn, C. D., K. L. Montooth, and D. M. Rand. 2007. Positive and negative selection on 716

the mitochondrial genome. Trends Genet 23:259-263. 717

Mishmar, D., E. Ruiz-Pesini, P. Golik, V. Macaulay, A. G. Clark, S. Hosseini, M. Brandon, K. 718

Easley, E. Chen, and M. D. Brown. 2003. Natural selection shaped regional mtDNA 719

variation in humans. Proceedings of the National Academy of Sciences 100:171. 720

Moghadam, H. K., M. A. Pointer, A. E. Wright, S. Berlin, and J. E. Mank. 2012. W chromosome 721

expression responds to female-specific selection. Proceedings of the National Academy 722

of Sciences 109:8207-8211. 723

Nosil, P., S. P. Egan, and D. J. Funk. 2008. Heterogeneous genomic differentiation between 724

walking-stick ecotypes: ‘‘Isolation by adaptation’’ and multiple roles for divergent 725

selection. Evolution 62:316–336. 726

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, R. B. O'Hara, G. L. Simpson, P. Solymos, M. 727

Henry, H. Stevens, and H. Wagner. 2011. vegan: Community Ecology Package. R 728

package version 1.17-11. 729

Petit, R. J., and L. Excoffier. 2009. Gene flow and species delimitation. Trends Ecol Evol 730

24:386-393. 731

Phillips, S., R. Anderson, and R. Schapire. 2006. Maximum entropy modeling of species 732

geographic distributions. Ecol Model 190:231-259. 733

Pinho, C., and J. Hey. 2010. Divergence with Gene Flow: Models and Data. Annual Review of 734

Ecology, Evolution, and Systematics 41:215-230. 735

Price, T. 2008. Speciation in birds. Roberts and Company Publishers., Greenwood Village, 736

Colorado. 737


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

43

Primmer, C. R., A. P. Moller, and H. Ellegren. 1995. Resolving genetic relationships with 738

microsatellite markers: A parentage testing system for the swallow Hirundo rustica. Mol 739

Ecol 4:493-498. 740

Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using 741

multilocus genotype data. Genetics 155:945-959. 742

R Development Core Team. 2011. R: A Language and Environment for Statistical Computing. R 743

Foundation for Statistical Computing, Vienna, Austria. 744

Rand, D. M., R. A. Haney, and A. J. Fry. 2004. Cytonuclear coevolution: the genomics of 745

cooperation. Trends Ecol Evol 19:645-653. 746

Ribeiro, Â. M., P. Lloyd, and R. C. K. Bowie. 2011. A tight balance between natural selection 747

and gene flow in a southern African arid-zone endemic bird. Evolution DOI: 748

10.1111/j.1558-5646.2011.01397.x. 749

Schneider, C. J., T. B. Smith, B. Larison, and C. Moritz. 1999. A test of alternative models of 750

diversification in tropical rainforests: Ecological gradients vs. rainforest refugia. Proc 751

Natl Acad Sci U S A 96:13869-13873. 752

Schodde, R., and I. J. Mason. 1999. The directory of Australian birds. A taxonomic and 753

zoogeographic atlas of the biodiversity of birds in Australia and its Territories. 754

Passerines. CSIRO Publishing Collingwood. 755

Shen, Y. Y., L. Liang, Z. H. Zhu, W. P. Zhou, D. M. Irwin, and Y. P. Zhang. 2010. Adaptive 756

evolution of energy metabolism genes and the origin of flight in bats. Proceedings of the 757

National Academy of Sciences 107:8666-8671. 758

Smadja, C. M., and R. K. Butlin. 2011. A framework for comparing processes of speciation in 759

the presence of gene flow. Mol Ecol 20:5123–5140. 760


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

44

Sorenson, M. D., J. C. Ast, D. E. Dimcheff, T. Yuri, and D. P. Mindell. 1999. Primers for a PCR-761

Based Approach to Mitochondrial Genome Sequencing in Birds and Other Vertebrates. 762

Mol Phylogenet Evol 12:105-114. 763

Sorenson, M. D., C. N. Balakrishnan, and R. B. Payne. 2004. Clade-limited colonization in 764

brood parasitic finches (Vidua spp.). Syst Biol 53:140. 765

Spottiswoode, C. N., K. F. Stryjewski, S. Quader, J. F. R. Colebrook-Robjent, and M. D. 766

Sorenson. 2011. Ancient host specificity within a single species of brood parasitic bird. 767

Proceedings of the National Academy of Sciences 108:17738-17742. 768

Stephens, M., and P. Donnelly. 2003. A comparison of bayesian methods for haplotype 769

reconstruction from population genotype data. The American Journal of Human Genetics 770

73:1162-1169. 771

Stephens, M., N. J. Smith, and P. Donnelly. 2001. A new statistical method for haplotype 772

reconstruction from population data. The American Journal of Human Genetics 68:978-773

989. 774

Sunnucks, P., M. J. Blacket, J. M. Taylor, C. J. Sands, S. A. Ciavaglia, R. C. Garrick, N. N. Tait, 775

D. M. Rowell, and A. Pavlova. 2006. A tale of two flatties: different responses of two 776

terrestrial flatworms to past environmental climatic fluctuations at Tallaganda in montane 777

southeastern Australia. Mol Ecol 15:4513-4531. 778

Symula, R., J. S. Keogh, and D. C. Cannatella. 2008. Ancient phylogeographic divergence in 779

southeastern Australia among populations of the widespread common froglet, Crinia 780

signifera. Mol Phylogenet Evol 47:569-580. 781

Tajima, F. 1996. The amount of DNA polymorphism maintained in a finite population when the 782

neutral mutation rate varies among sites. Genetics 143:1457-1465. 783


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

45

Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and S. Kumar. 2011. MEGA5: 784

Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary 785

Distance, and Maximum Parsimony Methods. Mol Biol Evol (submitted). 786

Tieleman, B. I., M. A. Versteegh, A. Fries, B. Helm, N. J. Dingemanse, H. L. Gibbs, and J. B. 787

Williams. 2009. Genetic modulation of energy metabolism in birds through 788

mitochondrial function. Proceedings of the Royal Society B: Biological Sciences 789

276:1685. 790

Tieleman, B. I., J. B. Williams, and P. Bloomer. 2003. Adaptation of metabolism and 791

evaporative water loss along an aridity gradient. Proceedings of the Royal Society of 792

London. Series B: Biological Sciences 270:207. 793

Toews, D. P. L., and A. Brelsford. 2012. The biogeography of mitochondrial and nuclear 794

discordance in animals. Mol Ecol 21:3907-3930. 795

Tonnis, B., P. R. Grant, B. R. Grant, and K. Petren. 2005. Habitat selection and ecological 796

speciation in Galápagos warbler finches (Certhidea olivacea and Certhidea fusca). 797

Proceedings of the Royal Society B: Biological Sciences 272:819-826. 798

Turmelle, A. S., T. H. Kunz, and M. D. Sorenson. 2011. A tale of two genomes: contrasting 799

patterns of phylogeographic structure in a widely distributed bat. Mol Ecol. 800

Vallender, R., R. J. Robertson, V. L. Friesen, and I. J. Lovette. 2007. Complex hybridization 801

dynamics between golden-winged and blue-winged warblers (Vermivora chrysoptera and 802

Vermivora pinus) revealed by AFLP, microsatellite, intron and mtDNA markers. Mol 803

Ecol 16:2017-2029. 804


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

46

Via, S. 2012. Divergence hitchhiking and the spread of genomic isolation during ecological 805

speciation-with-gene-flow. Philosophical Transactions of the Royal Society B: Biological 806

Sciences 367:451-460. 807

Waltari, E., and S. V. Edwards. 2002. Evolutionary dynamics of intron size, genome size, and 808

physiological correlates in archosaurs. Am Nat:539-552. 809

Webb, W. C., J. M. Marzluff, and K. E. Omland. 2011. Random interbreeding between cryptic 810

lineages of the Common Raven: evidence for speciation in reverse. Mol Ecol 20:2390-811

2402. 812

Williams, J. B., and B. I. Tieleman. 2005. Physiological adaptation in desert birds. Bioscience 813

55:416-425. 814

Zink, R. M., and G. F. Barrowclough. 2008. Mitochondrial DNA under siege in avian 815

phylogeography. Mol Ecol 17:2107-2121. 816

Zink, R. M., A. Pavlova, S. Drovetski, and S. Rohwer. 2008. Mitochondrial phylogeographies of 817

five widespread Eurasian bird species. Journal of Ornithology 149:399-413. 818


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

47

FIGURE CAPTIONS 819

Figure 1. (a) Sampling localities (dots) and geographic distribution of mitochondrial haplogroups 820

(colors): light red dots - A1, dark red dots - A2, black dots - B; numbers show distribution of 821

haplotypes (labeled as on Fig. 3). Insets show two locations where both haplogroups were 822

sampled: in Tenterfield, a haplogroup A male was collected on the same day from the same 823

group of birds as a haplogroup B male and female; in Shelbourne, a haplogroup B male and 824

haplogroup A male were caught in May and October 2008, respectively. (b) Distribution of Q-825

values (pie charts within circles) for two microsatellite genetic clusters (‘red’ and ‘black’), 826

detected by STRUCTURE; each circle’s outer border color shows the mitochondrial haplogroup for 827

each individual: light red - A1, dark red - A2, black - B; samples are spaced apart for easier 828

viewing. Altitudinal layers of the Great Dividing Range are: light grey - 200 to 500 m, grey - 500 829

to 1000 m, dark grey - above 1000 m. 830

831

Figure 2. Haplotype networks for mitochondrial ND2 and six nuclear introns. Circles indicate 832

unique haplotypes, with area proportional to haplotype frequency. Connections between circles 833

are a single mutation unless indicated otherwise (italic). Colors correspond to the haplogroups: 834

grey- A (light grey - A1, corresponding to light red dots on Fig. 1A, dark grey - A2 [two 835

haplotypes, H9 and H12, on mtDNA network], dark red dot on Fig. 1A), white - B (black dots on 836

Fig. 1A); black pies on nuclear networks indicate individuals sequenced for nDNA but not for 837

mtDNA (thus, mtDNA haplogroup for these individuals was unknown). 838

839

Figure 3. Geographic distribution of spatial principal component (sPCA) scores for 840

mitochondrial ND2 (mt-sPC1 and mt-sPC2), microsatellites (msat-sPC1 and msat-sPC2), 841


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

48

GAPDH (GAPDH-sPC1) and MUSK-I3 (MUSK-I3-sPC1). Large black squares indicate samples 842

well differentiated from those denoted by large white squares, small squares are less 843

differentiated. Lines indicate values of sPCA scores for interpolated surface (lags). Narrowly 844

spread lines and a lack of small squares indicate discontinuity or sharp geographic transition 845

between two groups (as for mtDNA). Presence of smaller squares and broadly spread lines 846

indicate a cline (as for microsatellites). 847

848

Figure 4. Distribution of two major mtDNA lineages of Eopsaltria australis in environmental 849

space, as defined by two BIOCLIM variables, max temperature of warmest month and 850

precipitation of driest month. Two localities where both haplogroups occurred together are 851

marked with arrows.852

853


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

A1-mtDNA

B- mtDNA

A2-mtDNA

‘Red’- nuclear

‘Black’- nuclear

South Australia

South Australia

a b


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

H30

H31

H35H37

H24

H25

H38

H40

H36

H34

H33

H39

H27

H26

H32

H23

H29

H28

H16H8

H20

53 mutations

10 mutations

H4

H12H9

H19

H3 H7

H2

H22

H14

H15

H13

H17

H10

H11

H5

H21

H1

H18

H6

2

42

2

2

2

2

2

4

3

2

2

3

2

3

2

2

2

2

ND2 AB4 GAPDH

GTP MUSK-I4

TGFβ2

RI2


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Mt-sPC1 Mt-sPC2

Msat-sPC2Msat-sPC1

GAPDH-sPC1

MUSK-I4-sPC1I3


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

of 101Evolution: For Review Only

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

1

PERCHED AT THE MITO-NUCLEAR CROSSROADS: DIVERGENT MITOCHONDRIAL LINEAGES

CORRELATE WITH ENVIRONMENT IN THE FACE OF ONGOING NUCLEAR GENE FLOW IN AN

AUSTRALIAN BIRD

Alexandra Pavlova1, J. Nevil Amos1, Leo Joseph2 , Kate Loynes3, Jeremy J. Austin4,5, J. Scott

Keogh3, Graham N. Stone6, James A. Nicholls6, and Paul Sunnucks1

1School of Biological Sciences and Australian Centre for Biodiversity, Clayton Campus,


2Australian National Wildlife Collection, CSIRO Ecosystem Sciences, GPO Box 1700,

Canberra, ACT 2601

3 Division of Evolution, Ecology and Genetics, Building 116, Daley Rd, Research School of

Biology, The Australian National University, Canberra, ACT 0200 Australia

4Australian Centre for Ancient DNA (ACAD), School of Earth & Environmental Sciences,

The University of Adelaide, Darling Building, North Terrace Campus, SA 5005, Australia

5Sciences Department, Museum Victoria, Carlton Gardens, Melbourne, Vic, 3001, Australia

6 Institute of Evolutionary Biology, University of Edinburgh, The King’s Buildings, West

Mains Road, Edinburgh EH9 3JT, Scotland

Keywords: evolutionary history, sPCA, dbRDA, female-linked selection, eastern yellow

robin, Eopsaltria australis

Corresponding author: Alexandra Pavlova

School of Biological Sciences and Australian Centre for Biodiversity, Clayton campus,



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

2

Phone: (+61) 3 9905 5902; Fax: (+61) 3 9905 5613

Email: [email protected]

Running title: Evolution of the eastern yellow robin

Word count: 6950 6916 words of text (excluding tables, figure captions and literature cited)

This manuscript has 3 tables and 4 figures

We have/intend towill archive our sequencing data in GenBank, and other data (on

sequences, microsatellites and CHD markers, geographic coordinates, climatic variables, sex,

etc) in Dryad and GenBank.

Formatted: English (Australia)

Field Code Changed




123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

3

ABSTRACT

Relationships among multi-locus genetic variation, geography and environmental

variation can reveal how evolutionary processes affect genomes. We examined the evolution

of an Australian bird, the eastern yellow robin Eopsaltria australis, using mitochondrial

(mtDNA) and nuclear (nDNA) genetic markers, and bioclimatic variables. In southeastern

Australia, two divergent mtDNA lineages occur east and west of the Great Dividing Range,

perpendicular to latitudinal nDNA structure. We evaluated alternative scenarios to explain

this striking discordance in landscape genetic patterning. Stochastic mtDNA lineage sorting

implausibly explains this pattern can be rejected because the mtDNA lineages are essentially

distinct geographically for >1500 km. Vicariance is unlikely: the Great Dividing Range itself

is currently notis neither a current barrier nor was it at the Last Glacial Maximum according

to species distribution modeling; nuclear gene flow inferred from coalescent analysis affirms

this too. Female philopatry contradicts known female-biased dispersal. Contrasting mtDNA

and nDNA demographies for mtDNA and nDNA suggestindicate that their evolutionary

histories are decoupled have been decoupled. A non-neutral explanation for geographic

patterning of mtDNA variation is supported by dDistance-based redundancy analysis, in

which shows environmental temperatures to explain significant mtDNA variance variation

above impacts of that explained by geographic position and isolation-by-distance, favors a

non-neutral explanation for mitochondrial phylogeographic patterning. Thus, observed mito-

nuclear discordance is consistentaccords with environmental selection on a female-linked

trait, such as mtDNA, mtDNA-nDNA interactions or genes on W-chromosome, driving

mitochondrial divergence in the presence of nuclear gene flow.


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

4

INTRODUCTION 1

Evolutionary processes (e.g. genetic drift, natural selection and gene flow) operate at 2

different times and places in species’ histories, differentially affecting genomes and phenotypes; 3

thus understanding their roles in speciation remains challenging (Coyne and Orr 2004; Price 4

2008; Butlin et al. 2009; Smadja and Butlin 2011). Discordant geographic patterns within and 5

among genomes in a species could result from coalescent variance (Lohse et al. 2010), 6

genetically localized selective sweeps (Bensch et al. 2006), life history traits (e.g. sex-biased 7

philopatry/dispersal; Turmelle et al. 2011), different modes of marker inheritance (and thus 8

effective population sizes), or differential marker behavior following secondary contact (Petit 9

and Excoffier 2009). A recent review (Toews and Brelsford 2012) found that mito-nuclear 10

discordance, a major difference in the patterns of differentiation between mitochondrial and 11

nuclear DNA, is a common phenomenon in animal systems. In the great majority of reviewed 12

cases, mito-nuclear discordance arose following allopatric divergence and secondary contact 13

(Toews and Brelsford 2012). This is expressed either in more structuring and/or narrower 14

geographic clines in mtDNA compared with nDNA (usually explained by nuclear introgression 15

and/or sex-biased asymmetries including dispersal, mating or offspring survival) or relatively 16

less structuring and/or wider geographic clines of mtDNA, usually explained by adaptive 17

introgression of mtDNA, demographic disparities including genetic drift, or sex-biased 18

asymmetries. Historic isolation and secondary contact can also result in coexistence of deeply 19

divergent mitochondrial lineages in sympatry in one panmictic population (Zink et al. 2008; 20

Webb et al. 2011; Hogner et al. 2012). However, in four cases, all involving avian systems, 21

strong mitochondrial but not nuclear structure has arisen in the absence of obvious geographic 22

isolation (Irwin et al. 2005; Cheviron and Brumfield 2009; Ribeiro et al. 2011; Spottiswoode et 23


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

5

al. 2011). Further, in two bird species, mitochondrial but not nuclear structure correlated with 24

environmental variation (Cheviron and Brumfield 2009; Ribeiro et al. 2011), suggesting that 25

mtDNA haplotypes might be differentially adapted to environmental conditions. Growing 26

evidence of selection on mtDNA (Ballard and Whitlock 2004; Bazin et al. 2006; Meiklejohn et 27

al. 2007) indicates that this process might be more common than recognized, although it is rarely 28

tested for in phylogeographic studies (Toews and Brelsford 2012). Links between mitochondrial 29

variation and energy metabolism (Mishmar et al. 2003; Tieleman et al. 2009) suggest that 30

mtDNA data should be routinely examined in relation to environmental variables in species 31

distributed over wide ranges of climatic conditions. Multilocus investigation of spatial genetic 32

structure in relation to environmental variation should lead to better understanding of 33

evolutionary forces shaping molecular variation. 34

Here we report an intriguing phylogeographic pattern in an Australian bird, where two 35

major mtDNA lineages were distributed roughly eastwards or westwards of the Great Dividing 36

Range in southeastern Australia, perpendicular to roughly latitudinal isolation-by-distance of 37

some nuclear loci, and we assess the processes shaping this molecular pattern. The eastern 38

yellow robin Eopsaltria australis (Passeriformes: Petroicidae) is a common and widespread bird 39

of mesic woodlands and forests of tropical to temperate eastern Australia (Barrett et al. 2003). 40

The species’ distribution spans 20º of latitude, associated with a wide range of climates (Higgins 41

and Peter 2002). Superimposed on this climatic gradient, the Great Dividing Range, a mountain 42

chain of modest elevation peaking at 2228 metres above sea level, runs the entire length of the 43

eastern coastline, and imposes climates that are generally drier and warmer further inland 44

(Bowman et al. 2010; Byrne et al. 2011). Comprehensive, continent-wide bird surveys at 1º grids 45

(1998-2002) showed E. australis to be evenly distributed and breeding throughout the Great 46


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

6

Dividing Range (Barrett et al. 2003). Dispersal of E. australis is limited and significantly 47

female-biased (Higgins and Peter 2002; Debus and Ford 2012; Harrisson et al. 2012), thus local 48

adaptation is feasible (Schneider et al. 1999). So, too, is neutral divergence in allopatry: 49

aridification and cooling during Pleistocene glacial cycles have promoted divergence and 50

speciation of some forest and woodland Australian mesic organisms (Sunnucks et al. 2006; 51

Symula et al. 2008; Malekian et al. 2010; Byrne et al. 2011). 52

A preliminary molecular data set (Loynes et al. 2009) showed two highly divergent 53

mitochondrial ND2 haplogroups in E. australis (> 6% divergence), whereas one polymorphic 54

nuclear gene and two nuclear introns were not structured. Using mitochondrial COI Christidis et 55

al. (2011) confirmed this mitochondrial subdivision and suggested that, pending further 56

phylogeographic study, the haplogroups may comprise separate species. Notably, the 57

haplogroups did not correspond to the recognized subspecies, northern E. a. chrysorrhoa and 58

southern E. a. australis (Schodde and Mason 1999), suggesting a possibility of mito-nuclear 59

discordance. Interpreting strong mtDNA divergence as indicative of speciation is based on the 60

usually untested assumption that mitochondrial lineages will behave as if under neutral 61

evolution. Although in some cases this assumption might be valid (Harrison 1989; Zink and 62

Barrowclough 2008), simulations show that even weak selection for local adaptations can result 63

in strong phylogeographic structure of a uniparentally inherited locus, where divergent clades are 64

geographically localized and differently adapted (Irwin 2012). Here we report an intriguing 65

phylogeographic pattern in E. australisan Australian bird, where two major mtDNA lineages 66

were distributed roughly eastwards or westwards of the Great Dividing Range in southeastern 67

Australia, perpendicular to roughly latitudinal isolation-by-distance of some nuclear loci, and we 68

assess the processes shaping this molecular pattern., The aim of this study was toand we test 69


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

7

rigorously whether selection or one or more neutral process better explains this the extreme 70

mtDNA divergence and mito-nuclear discordance observed in E. australis. We used multiple 71

types of loci (mitochondrial ND2 gene, nuclear microsatellites and introns) having different rates 72

of evolution and patterns of inheritance (autosomal, Z-linked, and female-limited W-linked), and 73

climatic variables, applied over the full range of the species in landscape genetic and 74

genealogical frameworks. First, we tested if mitochondrial haplogroups diverged by drift in 75

allopatry followed by secondary contact (as in the majority of cases reviewed by Toews and 76

Brelsford 2012). Under this hypothesis, we expected to observe low habitat suitability for E. 77

australis on and along the Great Dividing Range (an agent of vicariance for some taxa along the 78

length of Australia's eastern seaboard; Byrne et al. 2011) during the Last Glacial Maximum 79

(LGM). We used species distribution modeling to test this. Second, we tested if mitochondrial 80

divergence in E. australis occurred in the presence of nuclear gene flow. Under divergence with 81

gene flow (Pinho and Hey 2010), some nuclear gene flow between mtDNA haplogroups is 82

expected. We tested this by fitting an isolation-with-migration (IMA) model to nDNA data. 83

Finally, we tested the hypothesis that female-linked environmental selection (including selection 84

on mtDNA and linked nuclear genes; Rand et al. 2004; Meiklejohn et al. 2007; Dowling et al. 85

2008; Tieleman et al. 2009; Shen et al. 2010; Moghadam et al. 2012) resulted in mtDNA 86

divergence in the presence of nuclear gene flow between populations defined by their mtDNA 87

haplogroups. This hypothesis includes the case where dispersal (which is female-biased in E. 88

australis) between environmentally-divergent areas is lower than dispersal between similar 89

areas, because such bias demands a selective explanation. Under the female-linked selection 90

hypothesis, decoupled evolutionary history of mtDNA and nDNA is expected to result in 91

different patterns of geographic structure for these genomes, and mtDNA may show associations 92


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

8

with environmental variation, after spatial autocorrelation has been controlled for. We tested 93

these predictions using phylogeographic, landscape genetic and distance-based redundancy 94

analyses. 95

METHODS 96

Modeling present and Last Glacial Maximum species distributions 97

To understand how current and late Pleistocene climatic variation in temperature and 98

humidity might have impacted the spatial distribution of E. australis (and its genetic variation), 99

and to test for a vicariant effect of the Great Dividing Range, we modeled current and LGM 100

distributions of this species using the machine-learning maximum entropy model implemented in 101

MaxEnt 3.3.3e (Phillips et al. 2006). Nineteen bioclimatic variables for the present and the LGM 102

(Busby 1991) at a resolution of 2.5 arc-minutes (~5 km2 grid) were obtained from the WorldClim 103

1.4 database (Hijmans et al. 2005). Layers were visualized and cropped to span from latitude 104

46°S to 5°S and longitude 130°S to 156°S using DIVA-GIS 7.4.0 (Hijmans 2009). Presence data 105

comprised 454 presence localities of E. australis in the Atlas of Living Australia 106

(http://www.ala.org.au). A set of 341 randomly chosen presence points (75%) was used to train 107

MaxEnt model under current climatic conditions, and a set of 113 points (25%) was used to test 108

the model. The fit of several preliminary MaxEnt models generated with different combinations 109

of variables was compared using the area under the receiver operating curve (AUC). The final 110

model of current species distribution was generated using six variables that maximized AUC: 111

max temperature of warmest month, min temperature of coldest month, mean diurnal range, 112

precipitation of wettest month, precipitation of driest month, and precipitation seasonality. LGM 113

distribution was modeled by projecting the final current model onto the set of these six variables 114

estimated for the LGM (see supplementary material S1 for details). 115


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

9

Sampling and molecular methods 116

Genomic DNA was extracted from 63 frozen tissues of E. australis spanning the species’ 117

entire geographic range (Australian National Wildlife Collection, CSIRO Ecosystems Sciences, 118

Canberra) following Kearns et al. (2009), and from 44 blood samples collected in the south and 119

north of the species’ range following Harrisson et al. (2012) (Fig. 1, supplementary material S2). 120

Individuals were screened for eight autosomal nuclear microsatellites, Cpi3, Cpi8 (Doerr 2005), 121

Escmu6 (Hanotte et al. 1994), HrU2 (Primmer et al. 1995), Pocc6 (Bensch et al. 1997), Smm7 122

(Maguire et al. 2006), Pgm1 and Pgm7 (Dowling et al. 2003) following Harrisson et al. (2012), 123

and for allele-length polymorphism in the female-specific CHD-W locus and its Z-chromosome 124

homolog CHD-Z (Griffiths et al. 1998; supplementary material S3). Pgm1 had suspected null 125

alleles, and Pgm7 was found to be Z-linked and extremely variable, thus neither were used in 126

analyses. The other six microsatellite loci were previously shown to be in Hardy-Weinberg 127

equilibrium and independently segregating in E. australis (Harrisson et al. 2012). 128

The mitochondrial (mtDNA) ND2 gene was amplified using primers L5215 (Hackett 129

1996) and H6313 (Sorenson et al. 1999) following Kearns et al. (2009) and sequenced 130

commercially in both directions using the amplification primers (Macrogen, Korea). Six nuclear 131

introns [aldolase B intron 4 (AB4), glyceraldehyde-3-phosphate dehydrogenase intron 11 132

(GAPDH), phosphenolpyruvate carboxykinase intron 9 (GTP), Z-linked muscle- skeletal 133

receptorspecific tyrosine kinase intron 4 3 (MUSK-I4MUSK-I3), rhodopsin intron 2 (RI2) and 134

Transforming transforming Growth growth Factorfactor-β2 intron 5 (TGFβ2)] were amplified in 135

47 individuals (Table 1, supplementary material S2) using published primers (Waltari and 136

Edwards 2002; Sorenson et al. 2004; McCracken and Sorenson 2005; Vallender et al. 2007). All 137

nuclear DNA (nDNA) PCRs contained MgCl2 (1.5 mM), 1x KCl buffer, 0.125 µM each dNTP, 138


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

10

0.4 µM primers, 0.01 units/µl Taq polymerase and 0.2 mg/ml BSA, and the microsatellite 139

touchdown protocol above was applied. Nuclear loci were sequenced using forward primers by 140

the UK NERC Genepool sequencing facility (University of Edinburgh, UK). For a subset of 141

individuals representing all inferred haplotypes (see below), nuclear loci were further sequenced 142

in reverse direction to check that they did not represent sequencing errors. 143

Chromatograms were edited and aligned in Geneious Pro 4.8.5 (Drummond et al. 2010). 144

The mitochondrial origin of ND2 sequences was supported by lack of stop-codons, 145

insertions/deletions or sequence ambiguities. For nuclear introns containing indels, only the first 146

fragments of sequence, where the majority of sequences were unambiguous, were analyzed 147

(Table 1), and individuals heterozygous for indels within these fragments were removed from 148

analysis (1 for GAPDH, 5 for GTP, 8 for TGFβ2). Gametic phases of heterozygous sequences 149

were reconstructed using the PHASE 2.1 algorithm (Stephens et al. 2001; Stephens and Donnelly 150

2003) implemented in DNASP v5.10.1 (Librado and Rozas 2009). Due to presence of many rare 151

alleles for some loci, haplotypes for small proportions of individuals (0.07 for GTP, 0.19 for 152

MUSK-I34, 0.01 for RI2, 0.18 for TGFβ2 and 0.16 for MC1R) were resolved with <80% 153

probability. We used all best pairs of haplotypes for analyses to avoid systematic bias in 154

estimates of population genetic parameters (Garrick et al. 2010), although incorrectly resolved 155

haplotypes could slightly influence estimates of linkage disequilibrium, recombination and 156

coalescence analyses (Balakrishnan and Edwards 2009). Introns were checked for recombination 157

in DNASP using the four-gamete test (Hudson and Kaplan 1985). 158

Multilocus phylogeography and spatial analysis of genetic variation 159

Relationships among haplotypes for mtDNA and nDNA loci were visualized on a 160

median-joining network (Bandelt et al. 1999) constructed in NETWORK v4.5.1.6 (www.fluxus-161


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

11

engineering.com). Owing to recombination, nuclear gene networks may not represent true 162

genealogical relationships, but are nonetheless useful for assessing allele clustering with respect 163

to mtDNA-defined lineages. Support for mitochondrial ND2 haplogroups and their divergence 164

times were estimated in BEAST 1.6.1 and 1.7 (Drummond and Rambaut 2007; Drummond et al. 165

2012) using all 100 ND2 sequences (supplementary material S4). STRUCTURE 2.3.3 (Pritchard et 166

al. 2000) and TESS 2.3.1 (Chen et al. 2007) were used to explore genotype substructuring in 167

microsatellites and autosomal introns (supplementary material S5). Spatial principal component 168

analysis (sPCA; Jombart et al. 2008) implemented through the functions spca and global.rtest in 169

the package ADEGENET v1.2-8 (Jombart 2008) for R (R Development Core Team 2011) was used 170

to investigate spatial patterns of genetic variation. SPCA maximizes the product of spatial 171

autocorrelation and genetic variance to summarize multivariate data on allele frequencies into a 172

few uncorrelated axes. This approach is powerful for detecting cryptic spatial patterns that are 173

not associated with high genetic variation (Jombart et al. 2008); unlike STRUCTURE or TESS, it 174

does not require Hardy-Weinberg or linkage equilibria, and can be applied to sequence data. We 175

used a Gabriel graph and introduced a small amount of random noise to coordinates of multiple 176

individuals collected at the same location (function jitter, factor=1, amount=0.1; the network was 177

robust to the addition of noise). Analyses were run separately for ND2, each of the sequenced 178

nuclear markers, and the dataset of six microsatellites (removing 9 individuals with missing 179

scores for a locus). 180

Multiple processes, such as isolation-by-distance and local selection (e.g. isolation-by-181

adaptation; Nosil et al. 2008), can simultaneously shape spatial structure of variation of different 182

molecular markers. To separate the effect of geographic position, which may reflect local 183

processes, from any underlying effect of isolation-by-distance, we applied redundancy analysis 184


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

12

(RDA, function rda, R package VEGAN; Dixon 2003; Oksanen et al. 2011). RDA is a constrained 185

ordination technique, coupled with an ANOVA-like permutation test that tests whether a given 186

predictor explains residual variation in response after fitting (conditioning on) other predictors 187

first. We tested whether geographic position can explain genetic variance of nuclear loci beyond 188

what can be explained by the isolation-by-distance effect. As the response, we used data on 189

presence or absence of each identified allele (microsatellites) or variant nucleotide at 190

polymorphic sites (sequences) in each individual, computed using a genind constructor and 191

centred and scaled using scaleGen (R package ADEGENET). The two predictors were geographic 192

position (latitude and longitude, analyzed together for simplicity of interpretation) and pairwise 193

geographic distance between locations. For RDA (and also distance-based RDA, below) the 194

information about an individual’s pairwise geographic distances to other individuals was reduced 195

to a few columns using the following procedure. First, geographic distances between locations, 196

initially computed with function rdist.earth (R package FIELDS), were expressed as a rectangular 197

matrix using principal coordinates of the neighbourhood matrix procedure (pcnm; Borcard and 198

Legendre 2002) implemented in function pcnm in VEGAN. Then this rectangular matrix was 199

reduced to ≤3 columns best explaining genetic variance. For that, we fitted all pcnm axes as 200

predictors of allele presence using RDA, and chose up to three axes significant at P<0.05 to 201

represent geographic distances in the final RDAs. This approach is more powerful than Mantel 202

tests, which require multivariate environmental data (such as latitude, longitude and 203

environmental variables) to be expressed as distances (Legendre and Fortin 2010). For the final 204

RDAs, we tested for the ability of each predictor to explain significant variance in allele 205

presence-absence data alone (marginal tests), and after fitting the other predictor first 206

(conditional tests). Significance was assessed with 999 permutations (of the rows and columns of 207


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

13

the predictor matrix for marginal tests, or of the rows and columns of the multivariate residual 208

matrix for conditional tests). A predictor variable that explained significant (P<0.05) genetic 209

variance in marginal and conditional tests was inferred to have major effect on genetic structure. 210

If both marginal tests were significant but conditional tests were not, both variables were inferred 211

to influence genetic variation. RDAs were performed on each sequenced nuclear marker, and 212

complete genotypes at six microsatellites. For mitochondrial ND2 sequences, effects of 213

geographic position and distance were explored using distance-based redundancy analysis 214

(dbRDA) as explained below. 215

Genetic diversity, genetic differentiation and selection 216

For each sequenced locus, we calculated haplotype diversity (Hd), nucleotide diversity 217

(π), pairwise Φst between haplogroups, Tajima's D (Tajima 1996) and Fu’s Fs (Fu 1997) tests for 218

selective neutrality using ARLEQUIN 3.11 (Excoffier et al. 2005). DNASP was used to test for 219

selection on ND2 itself using the McDonald-Kreitman (MK) test (McDonald and Kreitman 220

1991), two individuals with missing data for two sites were removed from this analysis (for 221

additional tests for selection see supplementary material S7). ARLEQUIN was used to compute 222

expected heterozygosity (He) and FST between haplogroups for six microsatellites 223

(supplementary material S6). Calculations were performed for all data, and for the two 224

populations defined by mtDNA haplogroups. 225

Isolation-with-migration analysis (IMA2) 226

We fitted the model of isolation-with-migration (Hey and Nielsen 2004) as implemented 227

in IMA2.0 (Hey and Nielsen 2007) to the multilocus dataset comprising the largest non-228

recombining fragments of four autosomal introns that had no signature of selection (AB4, 229

GAPDH, MUSK-I34 and TGFβ2), the size-variable Z-linked intron (CHD-Z) and six autosomal 230


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

14

microsatellites for the two populations defined by mtDNA haplogroup membership (A1 and B– 231

see Results, omitting three individuals from an isolated northernmost population). Under this 232

classification, any migration detected between haplogroups would indicate presence of nuclear 233

gene flow inconsistent with vicariant divergence. The fit of a full migration model was compared 234

to that of a no-migration sub-model using a likelihood ratio test. Parameter estimates were 235

converted to demographic units (number of years or individuals) using a generation time of 3.5 236

years (calculated as average age of reproductive females, assuming age of first reproduction of 1 237

year and lifespan of 6 years) and mutation rates of Lerner et al. (2011) (see supplementary 238

material S8 for details of analyses). 239

Distance-based redundancy analysis (dbRDA) 240

We used dbRDA (Legendre and Anderson 1999) implemented in the capscale function 241

in R-package VEGAN to test whether mitochondrial divergence occurred as a result of selection 242

acting upon a female-linked trait (broadly defined, i.e. mtDNA, mtDNA-nDNA interaction, or 243

genes on the W-chromosome). We examined whether the two bioclimatic variables that were 244

most informative in predicting the E. australis distribution (supplementary material S1) explain 245

mitochondrial variation above that explained by geographic position and/or distance between 246

individuals (Legendre and Fortin 2010). In this analysis, individuals were the units of 247

observation, and pairwise inter-individual TN93+G mitochondrial genetic distances [calculated 248

in MEGA5 (Tamura et al. 2011) excluding two individuals with missing sites; supplementary 249

material S4] were treated as information on multivariate response (and not as a single univariate 250

response variable as in Mantel tests; Geffen et al. 2004). Predictors (standardized to a zero mean 251

and unit variance) were two environmental variables: maximum temperature of warmest month 252

(max.temp) and precipitation of driest month (precipitation), and three geographic variables: 253


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

15

latitude, longitude and geographic distance between samples represented by the first significant 254

(P<0.05) pcnm axis (as explained above for RDA). The final dbRDA analyses included (1) 255

marginal tests, where the relationship between mtDNA genetic distances and each of the 256

predictors was analyzed separately, (2) conditional tests, where significance of environmental 257

variables (max.temp or precipitation) as predictors of mtDNA distances was assessed after first 258

fitting other variables one at a time or together, and (3) the reciprocal conditional tests, where 259

significance of geographic variables was similarly assessed by first fitting environmental 260

variables. Significance was assessed with 999 permutations. Environmental variables were 261

inferred to explain mtDNA variance over that explained by geographic position and distance if 262

marginal as well as conditional tests were significant. Significant association of mitochondrial 263

variation and climatic variables beyond its correlation with geographic position and distance 264

would provide strong evidence against hypotheses assuming neutral behavior of mtDNA. 265

RESULTS 266

Current and LGM species distribution models 267

Current and LGM species distribution models (supplementary material S1) predict the 268

northernmost (Wet Tropics) range of E. australis to be isolated from the remainder. Thus, for 269

this outlying region, vicariant effects might have acted at some time. In contrast, the species is 270

predicted under LGM and current conditions to have a continuous distribution on and along the 271

Great Dividing Range, except for a proportionally small area of habitat unsuitability (localized 272

glaciation) during the LGM in the mountains of southeastern Victoria and southern New South 273

Wales (Barrows et al. 2001). These predicted distributions provide no basis to expect vicariant 274

patterns due to the Great Dividing Range (arguing against allopatric divergence). Two variables 275

describing extremes of climatic conditions, maximum temperature of warmest month and 276


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

16

precipitation of driest month, explained the majority of variance in E. australis presence (50% 277

and 38% respectively in the final model). 278

MtDNA phylogeography 279

The 100 individuals sequenced for 1002 base pairs of ND2 revealed 102 polymorphic 280

sites (80 parsimony informative), defining 40 haplotypes (GenBank accession numbers 281

KC466740 - KC466839xxxx-yyyy). All individuals fell into one of haplogroup A or B, between 282

which there was 6.6% net nucleotide divergence (Bayesian posterior probability=1, MCC BEAST 283

tree, supplementary material S4, and ND2 network, Fig. 2). Assuming neutral evolution with 284

rates similar to ND2 rate of the Hawaiian honeycreepers (Lerner et al. 2011), the haplogroups A 285

and B diverged in the early Pleistocene, ~1.5 (95% highest posterior density [HPD] 0.975-2.147) 286

million years ago. Five of 53 fixed differences between A and B represented non-synonymous 287

but biochemically conservative amino acid substitutions (Grantham 1974; supplementary 288

material S7). Haplogroup A was widespread north to south, and inland of the Great Dividing 289

Range in southeastern Australia (Fig. 1a, red points), while haplogroup B was principally 290

coastwards and restricted to southeastern Australia (Fig. 1a, black points). Two locations, 291

Tenterfield and Shelbourne (Fig. 1a) had individuals of both haplogroups. Haplogroup A 292

comprised haplogroups A1 (widespread except Wet Tropics) and northernmost A2 (three Wet 293

Tropics individuals; Bayesian posterior probability=1; net divergence 1%; MCC tree, 294

supplementary material S4), which diverged in the mid Pleistocene, ~212 (95% HPD 104-341) 295

thousand years ago, assuming neutral evolution. One of 10 fixed differences between A1 and A2 296

was non-synonymous. Consistent with these results, SPCA of mtDNA (Fig. 3) detected two 297

significant global structures (P<0.001), the first (mt-sPC1) reflecting sharp A versus B 298


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

17

subdivision, and the second (mt-sPC2) subdivision between A2 (Wet Tropics) and the rest 299

(A1+B). 300

Sex-linked CHD variation and population structure inferred from nuclear microsatellites 301

For 32 females and 64 males sequenced for ND2 and sexed by CHD variation, the 302

distribution of two W-linked (female-limited) alleles exactly matched mtDNA haplogroup 303

variation (allele 366 was restricted to haplogroup A, allele 362 to B), while the two Z-linked 304

alleles were widespread and unstructured with respect to haplogroups (supplementary material 305

S3). Ninety-two individuals were genotyped for six microsatellites. Genotypic analyses in 306

STRUCTURE and TESS suggested two genetic clusters, with a gradual north to south change of 307

individual cluster probabilities (Fig. 1b, Supplementary material S5). SPCA on microsatellites 308

detected two significant global structures (P<0.001; Fig. 3). The strongest structure (msat-sPC1) 309

showed a roughly north-south cline similar to that detected by STRUCTURE and TESS; the second, 310

subtle structure (msat-sPC2) showed a cline across the Great Dividing Range perpendicular to 311

msat-sPC1, directionally consistent with the major mitochondrial lineage subdivision (Fig. 3). 312

Isolation-by-distance (IBD) was a dominant process shaping microsatellite variation: geographic 313

distances alone could explain the variance in microsatellite frequencies (P=0.005, marginal RDA 314

test; Table 2), and remained significant even after geographic position was fitted first (P=0.017 315

conditional RDA test). In contrast, geographic position, although significant when fitted alone 316

(P=0.005, marginal RDA test), did not explain microsatellite structure after geographic distances 317

were controlled for (P=0.1, the reciprocal conditional RDA test). 318

Geographic variation of sequenced nuclear introns 319

Sequences were obtained from six nuclear introns in 31-47 individuals, which had two to 320

19 alleles (Table 1; GenBank accession numbers KC466694 - KC466739, KC466694 - 321


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

18

KC467012xxxx-yyyy). No intron displayed clusters or allele-sharing concordant with 322

mitochondrial haplogroup designations (Fig. 2). Two genetic clusters detected by genotypic 323

analysis of the autosomal introns in TESS were consistent with north-south gradient in cluster 324

membership detected from microsatellites (Supplementary material S5). Significant spatial 325

structure was suggested by sPCA for GAPDH (P=0.053) and for MUSK-I4MUSK-I3 (P=0.065; 326

Fig 3) in which southwestern individuals were somewhat differentiated from the rest; no 327

structure was found for the other sequenced introns (P>0.15; RI2 was not tested as only two 328

haplotypes were present). RDA (Table 2) showed that IBD as well as geographic position 329

explained variation in GTP (both P<0.05 on marginal tests, P>0.05 on conditional tests), 330

geographic position explained variation in AB4, GAPDH and MUSK-I4MUSK-I3 beyond that 331

explained by IBD (P<0.05 on marginal and conditional tests); and IBD but not geographic 332

position explained patterns in TFGβ2 (P<0.05 on marginal test). 333

Population genetic variability, differentiation and selection 334

Within-haplogroup nucleotide diversity was lower for ND2 than that for four (of six) 335

nuclear loci (AB4, GAPDH, MUSK-I4MUSK-I3 and TGFβ2; Table 1); π for nuclear loci ranged 336

0-0.0075 and 0.0003-0.0102, and for ND2 was 0.0013 and 0.0023 for A1 and B, respectively. 337

Significant differentiation between A1 and B was detected for ND2 (ΦST=0.968; P<0.001), 338

inevitable under reciprocal monophyly. Much lower but significant A1-B differentiation was also 339

found for nuclear introns AB4 (ΦST=0.047; P=0.03), GAPDH (ΦST=0.065; P=0.01), GTP 340

(ΦST=0.034; P=0.02) and for microsatellites (FST=0.029; P<0.001), but not for MUSK-I4MUSK-341

I3, RI2 or TGFβ2 (P>0.05; Table 1). Significantly negative Tajima’s D and/or Fu’s Fs, indicative 342

of population expansion, purifying selection or genetic hitchhiking, were found in three loci with 343

the lowest nucleotide diversity (for A1 in ND2 and GTP, and for B in ND2, GTP and RI2). 344


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

19

Further, a MK test for selection on ND2 detected significant deficiency of non-synonymous 345

polymorphism between A1 and B (Fisher’s exact test P=0.047) with 50 synonymous and 5 non-346

synonymous changes fixed between groups, compared to 30 synonymous and 10 non-347

synonymous polymorphisms present within groups; MK tests between haplogroups A and B or 348

between A1 and A2 were not significant (P>0.05). Thus, purifying selection is indicated to have 349

acted upon ND2; positive selection can be rejected (additional tests in supplementary material 350

S7). 351

IMA2 analyses 352

Coalescent analysis of nuclear data in IMA2 could not definitively estimate some model 353

parameters for two populations defined by their mitochondrial haplogroup membership (details 354

in supplementary material S8), thus, and because assumptions of the model might not be met, 355

IMA2 results should be interpreted with caution. Based on all nuclear loci, populations A and B 356

were estimated to have diverged in the late Pleistocene or Holocene, ~7,000 (95%HPD 700-357

149,000) years ago. Posterior distributions of some Θ parameters did not reach zero, 358

nevertheless, each of them had a distinct peak within the range of prior parameter distributions. 359

Demographic estimates of population sizes from all nuclear data were NA=14 (95%HPD 3-86), 360

NB=13 (2-101) and NANC=186 (122-222) thousand individuals (for A, B and ancestral 361

populations, respectively). Although non-zero probabilities were associated with all prior values 362

of coalescent migration parameters, peak posterior distribution estimates suggested some 363

migration between populations. Migration estimates from all nuclear data converted to 364

population migration rate (2Nm) corresponded to forward-in-time gene movement of 6.2 (95% 365

HPD 0-32) genes per generation from A to B and 3.8 (0-22) genes per generation from B to A. 366


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

20

IMA2 was unable to conclusively reject the no-migration submodel based on a likelihood ratio 367

test. 368

Environmental, latitudinal and IBD effects associated with mtDNA diversity 369

DbRDA (Table 3) showed that maximum temperature of warmest month (max.temp), 370

precipitation of driest month, and latitude explain a significant amount of variance in mtDNA 371

genetic distances (marginal tests P<0.005), accounting for 51, 47 and 24% of genetic variance in 372

marginal tests, respectively, whereas longitude and geographic distance do not (P=0.59 and 373

P=0.072). So we excluded longitude from partial tests but retained geographic distance which 374

was close to significance. There was evidence of strong correlation of environmental variation 375

with mtDNA haplogroups: max.temp remained a significant predictor of genetic distances even 376

when other variables were fitted first alone or together (P=0.005; Table 3). Precipitation seems 377

somewhat less influential than max.temp: it remained significant when the three other variables 378

were fitted individually, or the two geographic variables were fitted first (P=0.005), but not when 379

max.temp plus the two geographic variables were fitted first (P=0.38; Table 3). Latitude 380

remained significant after first fitting the other variables alone or together (P<0.05; Table 3). 381

Overall, these results suggest that max.temp and latitude may play an important role in shaping 382

mitochondrial genetic structure of E. australis. Haplogroup A individuals occupied locations 383

with higher maximum temperatures of the warmest month and lower precipitation of the driest 384

month than those of haplogroup B (Fig. 4). 385

DISCUSSION 386

We attempted to clarify the evolutionary processes shaping a profound mito-nuclear 387

discordance in the Australian bird, the eastern yellow robin Eopsaltria australis, in which the 388

major patterns of spatial variation in mtDNA and nDNA are perpendicular. We addressed several 389


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

21

evolutionary scenarios, applying a series of landscape-genetic, spatial-genetic and genealogical 390

tests. 391

Deep mitochondrial divergence can arise by chance in a continuous population, when 392

dispersal and population size are low, due to the stochastic nature of the coalescent process 393

(Irwin 2002). However, such an effect is highly unlikely in E. australis, where the two 394

haplogroups have a sharp spatial division within this continuously distributed, common, 395

widespread species (Barrett et al. 2003). The haplogroups abut over >1500 km with substantial 396

potential for mixing, yet only in two sampled locations did they co-occur (Fig. 1). Random 397

processes also cannot explain correlation of mtDNA lineages with environmental variables. 398

Under a scenario of allopatric divergence followed by secondary contact, common 399

neutral explanations for more structured mtDNA than nDNA include incompletely sorted nuclear 400

lineages, nuclear introgression, and/or sex-biased asymmetries such as dispersal, mating or 401

offspring production/survival (Toews and Brelsford 2012). While the species distribution models 402

did indicate vicariance as a plausible explanation of the A1-A2 split at or after the Last Glacial 403

Maximum, there is little basis for vicariance in the range of Eopsaltria australis south of the Wet 404

Tropics. The species is common and widespread along the Great Dividing Range (Barrett et al. 405

2003), and its likely presence there during the LGM is inferred from our species distribution 406

models (although the LGM populations on parts of the Range could have been sparse, according 407

to relatively low probability of occurrence; supplementary material S1). Geographic 408

inconsistencies between mtDNA and nDNA structure rule out incomplete lineage sorting as a 409

cause of discordance (Toews and Brelsford 2012). Historical lack of gene flow across the Great 410

Dividing Range but not in other directions would be expected to structure both mitochondrial 411

and microsatellite variation: yet no geographically distinct genetic clusters were detected for 412


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

22

microsatellites. Instead, the major spatial pattern for microsatellites is latitudinal isolation-by-413

distance, with only mild IBD across the Great Dividing Range. This suggests that distance, rather 414

than a barrier, has structured nuclear variation. This, and inference of some nuclear gene 415

exchange from coalescent analysis in IMA2 (tempered by parameter uncertainty), suggests that 416

some nuclear gene flow between haplogroups has occurred. Different demographic histories 417

were inferred from mtDNA and nDNA data: significant Tajima’s D and Fu’s Fs neutrality tests, 418

suggestive of population expansion, purifying selection or hitchhiking, were detected for mtDNA 419

and two least variable nuclear loci, not for any of the more variable nuclear markers. Population 420

expansion would be expected to yield generally consistent signals across different markers, thus 421

non-neutral evolution or hitchhiking of some loci appears more likely, which was also supported 422

by the MK test for ND2. This suggests that evolutionary histories of mitochondrial and nuclear 423

genomes are decoupled. The same conclusion arises from contrasting A-B divergence times 424

estimated from the two genomes while accounting for coalescent variance: the late Pleistocene-425

Holocene divergence of ~7,000 (700-149,000) years ago estimated by IMA2 from nuclear data 426

(Supplementary material S8) is much more recent than the early Pleistocene mitochondrial 427

lineage divergence of ~1.5 (1-2.1) million years ago estimated by BEAST from ND2 428

(Supplementary material S4). Although we note that the divergence time estimates of BEAST and 429

IMA2 are not strictly comparable because IMA2 incorporated migration into the model, the 430

difference represents evidence that mtDNA and nDNA cannot both be responding solely to 431

neutral processes. 432

Very strong female philopatry unrelated to selection, and selection correlated with 433

mtDNA, are both expected to limit female but not male movement and gene flow across 434

populations and thus to impact mtDNA more and nDNA less than expected in the absence of 435


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

23

these effects. However, the female philopatry needed to explain observed patterns in the data 436

contrasts strongly with what is known about the biology of E. australis: field observations 437

(Debus and Ford 2012) and population genetic data (Harrisson et al. 2012) show that females 438

disperse significantly further than males (at least ~7 km and ~1 km, respectively; Debus and 439

Ford 2012). It is possible that females prefer to settle in habitats similar to their natal habitats, 440

whereas males are less discriminating (Tonnis et al. 2005). However, behavior where if females 441

are were so unable or unwilling to disperse or settle between adjacent haplogroup ranges over 442

>1500 km of contact, isseems likely to be linked the observed pattern demands explanation 443

bytoto some major fitness advantage, implying female-linked selection. Unlike a scenario 444

involving mtDNA-linked selection, neither female philopatry nor female habitat preferences 445

explain the origin of mtDNA divergence, given that there was no evidence for vicariance, and 446

chance split is highly unlikely. 447

If we reject the selectively neutral hypotheses as unparsimonious, it remains to consider 448

possibilities involving female/mtDNA-linked selection. Geographically localized haplogroups 449

and very shallow coalescent times within haplogroups compared to between-haplogroups 450

observed in E. australis are consistent with simulation scenarios under local selection and low to 451

moderate dispersal (Irwin 2012). The correlation of environmental variables with the spatial 452

distribution of mtDNA variation, after controlling for IBD and geographic position (dbRDA 453

results, Table 3) is consistent with environmental selection. The most informative bioclimatic 454

variables with regards to species occurrence (maximum temperature of the warmest month and, 455

to a lesser extent, precipitation of the driest month) explained significant mitochondrial 456

patterning beyond that explained by latitude and IBD, suggesting that these climatic factors may 457

be associated with female-linked selection. 458


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

24

These results, coupled with the arguments against selectively neutral explanations, lead 459

us to propose that the most likely solution to the geographically perpendicular mito-nuclear 460

discordance in E. australis is female-linked selection (broadly-defined), taken here to include 461

selection on one or more of mtDNA, nuclear-encoded genes for proteins involved in oxidative 462

phosphorylation, nuclear-encoded genes for mitochondrial proteins, joint mito-nuclear genotype 463

or W-linked parts of the genome (Rand et al. 2004; Meiklejohn et al. 2007; Dowling et al. 2008; 464

Tieleman et al. 2009; Shen et al. 2010; Moghadam et al. 2012), female biology including biased 465

habitat selection (as in Tonnis et al. 2005) and ‘divergence hitchhiking’ (gene flow reduced as a 466

side effect of strong divergent selection on genes involved in local adaptation; Via 2012). 467

Nuclear gene flow within E. australis appears to be structured by isolation-by-distance along the 468

gradients in two perpendicular directions: latitudinal (main microsatellite structure, msat-sPC1; 469

Fig. 3) and across the Great Dividing Range (the cryptic second structure in the microsatellites, 470

msat-sPC2, directionally consistent with distribution of TESS clusters from analysis of sequenced 471

introns assuming three genetic clusters [supplementary material S5], which could be a signature 472

of gene flow impeded by arrested dispersal or survival of females across the Range). The 473

inference of isolation-by-distance is also supported by congruent geographic structure in color of 474

the upper tail-coverts, which changes from bright yellow to olive-green from north to south and 475

east to west in E. australis (Ford 1979). Keast (1958) and Ford (1979) suggested that this color 476

variation arose due to local ecotypic selection where brighter yellow is preferred in darker 477

environments where it would be more useful for signaling. Thus selective pressures driving 478

evolution of female-linked traits and color variation could be different and their evolution 479

unlinked. Overall, dramatic mito-nuclear discordance in E. australis suggests that natural 480


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

25

selection constrains mitochondrial but not selectively neutral nuclear gene flow along 481

environmental gradients. 482

More research is needed to understand the mechanisms through which putative selection 483

on female-linked traits in E. australis operates. Passerine birds from hotter and drier 484

environments have lower basal metabolic rates and evaporative water loss (Tieleman et al. 2003; 485

Williams and Tieleman 2005). Selection on mitochondrial genotypes can drive such adaptive 486

responses because the mitochondrial genome plays an important role in the regulation of energy 487

metabolism in birds (Tieleman et al. 2009). For E. australis, the MK tests suggested purifying 488

selection on ND2, thus positive selection appears not to impact ND2 directly, but could act upon 489

other mtDNA-encoded genes, or indirectly on nDNA-encoded genes that produce structural 490

proteins imported into the mitochondria or on the joint mito-nuclear genotype, mito-nuclear 491

hybrids being selected against in the contact zone between two haplogroups (Dowling et al. 492

2008; Ballard and Melvin 2010). Of all reviewed cases of mito-nuclear discordance (Toews and 493

Brelsford 2012), only two other studies, both involving birds, report association between mtDNA 494

and environmental variation in the presence of nuclear gene flow when evidence for previous 495

isolation is lacking. In the Rufous-collared Sparrow Zonotrichia capensis, mitochondrial but not 496

nuclear gene flow was significantly reduced along elevational gradients, suggesting that 497

mitochondrial haplotypes could be locally adapted (Cheviron and Brumfield 2009). A 498

relationship between a gradient of aridity and mitochondrial but not nuclear or morphological 499

diversity has been reported in the South African arid-zone endemic Karoo scrub-robin 500

Cercotrichas coryphaeus raising the possibility of common mechanisms for adaptations to 501

extreme environmental conditions (Ribeiro et al. 2011). Due to linked inheritance with the W-502

chromosome, mtDNA may also reflect female-specific selection on W-linked genes, as has been 503


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

26

demonstrated for the evolution of gene expression on the W-chromosome (Moghadam et al. 504

2012). Thus it may not be coincidental that all three studies (Cheviron and Brumfield 2009; 505

Ribeiro et al. 2011; the present study) reporting association between mtDNA and environmental 506

variation in the presence of nuclear gene flow are from birds. Whatever the mechanism, any 507

environmental selection acting on female-linked traits in E. australis would strongly inhibit the 508

sympatry of mitochondrial lineages across a climatic gradient associated with the Great Dividing 509

Range. 510

Conclusions 511

Our study of E. australis benefitted from a combination of landscape genetic, spatial 512

genetic and coalescent analyses of multiple locus types and environmental variables in building 513

an integrated understanding of evolutionary processes operating at different scales. It also 514

presents a striking example of how geographically structured mtDNA diversity can unreliably 515

reflect species’ evolutionary history when neutrality is assumed and not thoroughly tested. It is 516

beneficial to incorporate tests for the effect on genetic variation of geographic position and 517

environmental variation, in addition to IBD, into phylogeographic studies, and to try and tease 518

apart their almost inevitable correlations. Correlation between mitochondrial genetic variation 519

and geographic position, when distance is controlled for, warrants additional exploration of 520

potential selection pressures driving evolution of female-linked traits. Discovery of two highly 521

divergent haplogroups apparently under strong environmental selection within a putatively 522

continuous population of E. australis (possibly a common pattern for species whose range spans 523

environmental gradients) provides another example of divergence with gene flow (Pinho and 524

Hey 2010), which has potentially profound implications for management and taxonomy. The 525

evidence for female-linked environmental selection implies that mtDNA haplogroups are not 526


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

27

ecologically exchangeable sensu Crandall et al. (2000). For example, translocation among 527

haplogroup regions may result in negative fitness consequences, and on the evidence here would 528

be an inadvisable management strategy. Our study illustrates how the use of mtDNA as a 529

barcoding tool to define species or lineages (Baker et al. 2009) can be severely compromised 530

without the knowledge to appreciate the role of mtDNA-correlated selection in such units. 531

ACKNOWLEDGEMENTS 532

Funding was provided by the Australian Research Council Linkage Grant (LP0776322), 533

the Victorian Department of Sustainability and Environment (DSE), Museum of Victoria, 534

Victorian Department of Primary Industries, Parks Victoria, North Central Catchment 535

Management Authority, Goulburn Broken Catchment Management Authority, CSIRO 536

Ecosystem Sciences, and the Australian National Wildlife Collection Foundation. NA was 537

funded by a Monash University Faculty of Science Dean’s Postgraduate Scholarship and Birds 538

Australia with additional support from the Holsworth Wildlife Research Endowment. Blood 539

samples from Victoria were collected under DSE permits (number 10004294 under the Wildlife 540

Act 1975 and the National Parks Act 1975, and number NWF10455 under section 52 of the 541

forest Act 1958). We thank Alan Lill, Naoko Takeuchi and other collectors of specimens and all 542

agencies who granted permission to collect specimens, Robert Palmer for curatorial assistance, 543

Suzanne Metcalfe for help with mitochondrial sequencing, and Katherine Harrisson for help with 544

nuclear sequencing. Computationally intensive analyses (STRUCTURE) were performed on the 545

Monash Sun Grid courtesy of Monash eResearch. We thank the UK NERC Genepool facility for 546

sequencing support. We are grateful to Darren Irwin, John Endler, Evolution editorial input 547

including that of Kenneth Petren, three anonymous reviewers for their comments on the earlier 548


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

28

drafts of the manuscript, to Thibaut Jombart for advice on analyses of spatial genetic structures, 549

Gaynor Dolman, Hugh Ford and Peter Teske for helpful discussions. 550

551


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

29

Table 1. Descriptive statistics for all samples and members of distinct mitochondrial clades for sequenced loci (Wet tropics individuals 552

belonging to clade A2 were not sequenced for nDNA). N ind- number of individuals, N alleles- Number of alleles, bp- base pairs, S-553

polymorphic sites, H-number of haplotypes, Hd- haplotype diversity (expected heterozygosity), π- nucleotide diversity, Rm- minimum 554

number of recombinant events, n/s- not significant (P>0.05), n/a- not applicable. Significant ΦST -values are in bold. 555

Locus Data N

ind

N

alleles

Length,

bp

Indels,

bp

S H Hd π Fu's Fs, P Tajima's D,

P

ΦST A1

vs B

Rm

AB4 autosomal

intron

All 44 88 196 1 2 3 0.447 0.0025 0.613 n/s 0.343 n/s 0.047

P=0.027

0

Clade A 20 40 196 1 1 3 0.405 0.0022 -0.078 n/s 0.565 n/s

Clade B 22 44 195 0 2 3 0.519 0.0030 0.609 n/s 0.474 n/s

GAPDH autosomal

intron

All 46 92 317 0 8 10 0.818 0.0047 -2.217 n/s -0.147 n/s 0.065

P=0.009

1

Clade A 21 42 317 0 6 8 0.866 0.0054 -1.263 n/s 0.577 n/s

Clade B 23 46 317 0 6 6 0.723 0.0036 -0.742 n/s -0.432 n/s

GTP autosomal

intron

All 42 84 456 1 8 10 0.410 0.0011 -8.403

P<0.001

-1.942

P<0.001

0.034

P=0.018

0

Clade A 19 38 456 0 4 5 0.292 0.0007 -3.671

P=0.001

-1.630

P=0.016


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

30

Clade B 21 42 456 1 7 8 0.535 0.0018 -4.128

P=0.004

-1.502

P=0.056

MUSK-

I4MUSK-

I3

Z-linked

intron

All 31 51 519 0 21 15 0.810 0.0078 -2.240 n/s -0.416 n/s 0.061

P=0.072

3

Clade A 20 32 519 0 17 9 0.673 0.0065 -0.114 n/s -0.671 n/s

Clade B 11 17 519 0 18 10 0.919 0.0102 0.516 n/s -0.033 n/s

RI2 autosomal

intron

All 47 94 292 0 2 2 0.021 0.0002 -1.421 n/s -1.386 n/s -0.001

n/s

0

Clade A 22 44 292 0 0 1 0.000 0.0000 0 n/a 0 n/s

Clade B 23 46 292 0 2 2 0.044 0.0003 -0.783 n/s -1.473

P=0.03

TGFβ2 autosomal

intron

All 39 78 595 1 24 19 0.870 0.0070 -3.420 n/s -0.443 n/s 0.007

n/s

4

Clade A 20 40 595 0 23 16 0.868 0.0075 -3.465 n/s -0.599 n/s

Clade B 17 34 594 1 12 9 0.857 0.0066 0.553 n/s 0.935 n/s

ND2 mt coding All 100 100 1002 0 102 40 0.908 0.0334 2.785 n/s 2.303

P=0.01

0.968

P<0.001

n/a


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

31

mt coding Clades

A1+B

97 97 1002 0 94 38 0.902 0.0332 3.382 n/s 2.704

P=0.002

Clade A1 53 97 1002 0 26 20 0.738 0.0017 -16.418

P<0.001

-2.276

P<0.001

Clade A2 3 97 1002 0 2 2 0.667 0.0013 - -

Clade B 44 97 1002 0 16 18 0.901 0.0023 -10.728

P<0.001

-1.158 n/s

556


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

32

Table 2. Redundancy (RDA) analysis: separating the effect of IBD and geographic position (latitude and longitude analyzed together) 557

on multivariate distribution of allele frequencies (presence-absence) for six microsatellites and nuclear loci. Distance- geographic 558

distance (represented as a rectangular matrix of pcnm axes), Geography - joint effect of latitude and longitude, N pcnm axes- number 559

of significant pcnm axes representing geographic distance. Marginal tests show the relationship between allele frequencies and either 560

distance or geographic position alone. Conditional tests show the same relationships, but having first fitted the alternative variable 561

(geographic position and distance, respectively) as the covariates in the analyses. Significant P-values (P<0.05) are in bold. 562

Marginal tests Conditional tests

Locus (Loci)

N pcnm

axes

Predictor

Variable tested F P

% variance

explained

Variables

fitted first F P

% variance

explained

Microsatellites 3 Distance 2.62 0.005 9.1 Geography 1.4 0.017 5.2


AB4 2 Distance 0.41 0.87 1.9 Geography 1.58 0.19 7.5


GAPDH 3 Distance 1.96 0.015 12.3 Geography 1.58 0.08 10.6


GTP 3 Distance 1.8 0.015 12.4 Geography 1.53 0.08 11.3



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

33

MUSK-

I4MUSK-I3 2 Distance 2.02 0.037 18.3 Geography 1.69 0.07 16.9


TGFβ2 2 Distance 1.6 0.03 12.0 Geography 1.53 0.07 12.2



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

34

Table 3. Results of dbRDA analysis: tests of relationship between multivariate mtDNA genetic 563

variation (defined by pairwise mtDNA genetic distances) and several predictor variables 564

(geographic and environmental). The first row for each tested predictor is a marginal test for that 565

variable, the next four rows are conditional (partial) tests where variables given in second 566

column are partialed out (fitted before the variable being tested). Last column indicates 567

percentage of multivariate genetic variation explained by variable being tested. Max.temp- 568

maximum temperature of warmest month, Precipitation- precipitation of driest month, All- all 569

variables except the one being tested. Longitude was not a significant predictor of genetic 570

distance in the marginal test (F=0.34, P=0.59), and was not used for partial tests. Significant P-571

values (P<0.05) are in bold. 572

Predictor variable tested Variables fitted first

in partial tests

F P(>F) % variance

explained


Latitude 26.6 0.005 21.9

Max.temp 0.6 0.42 0.6


All 7.5 0.01 7.5

Latitude 31.1 0.005 24.5


Max.temp 12.3 0.005 11.5


All 11.6 0.005 11.1

Max.temp 100 0.005 51.1



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

35

Latitude 70.6 0.005 42.6



All 33.2 0.005 26.3



Latitude 47.6 0.005 33.4


Max.temp 18.3 0.005 16.1

All 0.7 0.38 0.8

573


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

36

LITERATURE CITED 574

Baker, A. J., E. S. Tavares, and R. F. Elbourne. 2009. Countering criticisms of single 575

mitochondrial DNA gene barcoding in birds. Molecular Ecology Resources 9:257-268. 576

Balakrishnan, C. N., and S. V. Edwards. 2009. Nucleotide variation, linkage disequilibrium and 577

founder-facilitated speciation in wild populations of the zebra finch (Taeniopygia 578

guttata). Genetics 181:645. 579

Ballard, J. W. O., and R. G. Melvin. 2010. Linking the mitochondrial genotype to the organismal 580

phenotype. Mol Ecol 19:1523-1539. 581

Ballard, J. W. O., and M. C. Whitlock. 2004. The incomplete natural history of mitochondria. 582

Mol Ecol 13:729-744. 583

Bandelt, H.-J., P. Forster, and A. Roehl. 1999. Median-joining networks for inferring 584

intraspecific phylogenies. Molecular Biology & Evolution 16:37-48. 585

Barrett, G., A. Silcocks, S. Barry, R. Cunningham, and R. Poulter. 2003. The New Atlas of 586

Australian Birds. Royal Australasian Ornithologists Union, Melbourne. 587

Barrows, T. T., J. O. Stone, L. K. Fifield, and R. G. Cresswell. 2001. Late Pleistocene glaciation 588

of the Kosciuszko Massif, Snowy Mountains, Australia. Quat. Res. 55:179-189. 589

Bazin, E., S. Glemin, and N. Galtier. 2006. Population size does not influence mitochondrial 590

genetic diversity in animals. Science 312:570-572. 591

Bensch, S., D. E. Irwin, J. H. Irwin, L. Kvist, and S. Akesson. 2006. Conflicting patterns of 592

mitochondrial and nuclear DNA diversity in Phylloscopus warblers. Mol Ecol 15:161-593

171. 594

Bensch, S., T. Price, and J. Kohn. 1997. Isolation and characterization of microsatellite loci in a 595

Phylloscopus warbler. Mol Ecol 6:91-92. 596


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

37

Borcard, D., and P. Legendre. 2002. All-scale spatial analysis of ecological data by means of 597

principal coordinates of neighbour matrices. Ecol Model 153:51-68. 598

Bowman, D. M. J. S., G. K. Brown, M. F. Braby, J. R. Brown, L. G. Cook, M. D. Crisp, F. Ford, 599

S. Haberle, J. Hughes, Y. Isagi, L. Joseph, J. McBride, G. Nelson, and P. Y. Ladiges. 600

2010. Biogeography of the Australian monsoon tropics. J Biogeogr 37:201-216. 601

Busby, J. R. 1991. BIOCLIM a bioclimatic analysis and prediction system. Pp. 64–68 in C. R. 602

Margules, and M. P. Austin, eds. Nature Conservation: Cost Effective Biological Surveys 603

and Data Analysis. CSIRO, Canberra. 604

Butlin, R., D. Schluter, and J. R. Bridle. 2009. Speciation and patterns of diversity. Cambridge 605

University Press, Cambridge, UK. 606

Byrne, M., D. Steane, L. Joseph, D. Yeates, G. Jordan, D. Crayn, K. Aplin, D. Cantrill, L. G. 607

Cook, M. D. Crisp, J. S. Keogh, J. Melville, C. Moritz, N. Porch, J. M. K. Sniderman, P. 608

Sunnucks, and P. H. Weston. 2011. Decline of a biome: evolution, contraction, 609

fragmentation, extinction and invasion of the Australian mesic zone biota. J Biogeogr In 610

press. 611

Chen, C., E. Durand, F. Forbes, and O. Francois. 2007. Bayesian clustering algorithms 612

ascertaining spatial population structure: a new computer program and a comparison 613

study. Mol Ecol Notes 7:747-756. 614

Cheviron, Z. A., and R. T. Brumfield. 2009. Migration-selection balance and local adaptation of 615

mitochondrial haplotypes in rufous-collared sparrows (Zonotrichia capensis) along an 616

elevational gradient. Evolution 63:1593-1605. 617


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

38

Christidis, L., M. Irestedt, D. Rowe, W. E. Boles, and J. A. Norman. 2011. Mitochondrial and 618

nuclear DNA phylogenies reveal a complex evolutionary history in the Australasian 619

robins (Passeriformes: Petroicidae). Mol Phylogenet Evol 61:726–738. 620

Coyne, J., and H. A. Orr. 2004. Speciation. Sinauer, Sunderland, Massachusetts. 621

Crandall, K. A., O. R. P. Bininda-Emonds, G. M. Mace, and R. K. Wayne. 2000. Considering 622

evolutionary processes in conservation biology. Trends Ecol Evol 15:290-295. 623

Debus, S. J. S., and H. A. Ford. 2012. Responses of Eastern Yellow Robins Eopsaltria australis 624

to translocation into vegetation remnants in a fragmented landscape. Pac Conserv Biol in 625

press. 626

Dixon, P. 2003. VEGAN, a package of R functions for community ecology. J Veg Sci 14:927-627

930. 628

Doerr, E. D. 2005. Characterization of microsatellite loci in the brown treecreeper (Climacteris 629

picumnus) and cross-species amplification in the white-throated treecreeper (Cormobates 630

leucophaeus). Mol Ecol Notes 5:654-656. 631

Dowling, D., U. Friberg, and J. Lindell. 2008. Evolutionary implications of non-neutral 632

mitochondrial genetic variation. Trends Ecol Evol 23:546-554. 633

Dowling, D. K., G. J. Adcock, and R. A. Mulder. 2003. Novel polymorphic microsatellite 634

markers for paternity analysis in the red-capped robin (Petroica goodenovii: Aves). Mol 635

Ecol Notes 3:517-519. 636

Drummond, A., and A. Rambaut. 2007. BEAST: Bayesian evolutionary analysis by sampling 637

trees. BMC Evol Biol 7:214. 638


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

39

Drummond, A. J., B. Ashton, S. Buxton, M. Cheung, A. Cooper, J. Heled, M. Kearse, R. Moir, 639

S. Stones-Havas, S. Sturrock, T. Thierer, and A. Wilson. 2010. Geneious v5.1. Available 640

from http://www.geneious.com. 641

Drummond, A. J., M. A. Suchard, D. Xie, and A. Rambaut. 2012. Bayesian phylogenetics with 642

BEAUti and the BEAST 1.7. Mol Biol Evol 29:1969-1973. 643

Excoffier, L., G. Laval, and S. S. 2005. Arlequin ver. 3.0: An integrated software package for 644

population genetics data analysis. Evolutionary Bioinformatics Online 1:47-50. 645

Ford, J. C. 1979. Speciation or Subspeciation in the Yellow Robins? Emu 79:103-106. 646

Fu, Y.-X. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking 647

and backgroud selection. Genetics 147:915-925. 648

Garrick, R. C., P. Sunnucks, and R. J. Dyer. 2010. Nuclear gene phylogeography using PHASE: 649

dealing with unresolved genotypes, lost alleles, and systematic bias in parameter 650

estimation. BMC Evol Biol 10:118. 651

Geffen, E. L. I., M. J. Anderson, and R. K. Wayne. 2004. Climate and habitat barriers to 652

dispersal in the highly mobile grey wolf. Mol Ecol 13:2481-2490. 653

Grantham, R. 1974. Amino acid difference formula to help explain protein evolution. Science 654

185:862-864. 655

Griffiths, R., M. C. Double, K. Orr, and R. J. G. Dawson. 1998. A DNA test to sex most birds. 656

Mol Ecol 7:1071-1075. 657

Hackett, S. J. 1996. Molecular Phylogenetics and Biogeography of Tanagers in the Genus 658

Ramphocelus (Aves). Mol Phylogenet Evol 5:368-382. 659


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

40

Hanotte, O., C. Zanon, A. Pugh, C. Greig, A. Dixon, and T. Burke. 1994. Isolation and 660

characterization of microsatellite loci in a passerine bird: The reed bunting Emberiza 661

schoeniclus. Mol Ecol 3:529-530. 662

Harrison, R. G. 1989. Animal mitochondrial DNA as a genetic marker in population and 663

evolutionary biology. Trends Ecol. Evol. 4(1):6-11. 664

Harrisson, K., A. Pavlova, J. N. Amos, N. Takeuchi, A. Lill, J. Q. Radford, and P. Sunnucks. 665

2012. Fine-scale effects of habitat loss and fragmentation despite large-scale gene flow 666

for some regionally declining woodland bird species. Landsc Ecol 27:813-827. 667

Hey, J., and R. Nielsen. 2004. Multilocus methods for estimating population sizes, migration 668

rates and divergence time, with applications to the divergence of Drosophila 669

pseudoobscura and D-persimilis. Genetics 167:747-760. 670

Hey, J., and R. Nielsen. 2007. Integration within the Felsenstein equation for improved Markov 671

chain Monte Carlo methods in population genetics. Proceedings of the National Academy 672

of Sciences 104:2785-2790. 673

Higgins, P. J., and J. M. Peter, eds. 2002. Handbook of Australian, New Zealand and Antarctic 674

Birds. Volume 6: Pardalotes to Shrike-thrushes. Oxford University Press, Melbourne. 675

Hijmans, R. J., S. E. Cameron, J. L. Parra, P. G. Jones, and A. Jarvis. 2005. Very high resolution 676

interpolated climate surfaces for global land areas. International Journal of Climatology 677

25:1965–1978. 678

Hijmans, R. J. 2009. DIVA-GIS, version 7. A geographic information system for the analysis of 679

biodiversity data. Available at: http://www.diva-gis.org. 680

Hogner, S., T. Laskemoen, J. T. Lifjeld, J. Porkert, O. Kleven, T. Albayrak, B. Kabasakal, and 681

A. Johnsen. 2012. Deep sympatric mitochondrial divergence without reproductive 682


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

41

isolation in the common redstart Phoenicurus phoenicurus. Ecology and Evolution doi: 683

10.1002/ece3.398. 684

Hudson, R. R., and N. L. Kaplan. 1985. Statistical properties of the number of recombination 685

events in the history of a sample of DNA sequences. Genetics 111:147. 686

Irwin, D. E. 2002. Phylogeographic breaks without geographic barriers to gene flow. Evolution 687

56:2383-2394. 688

Irwin, D. E., S. Bensch, J. H. Irwin, and T. D. Price. 2005. Speciation by distance in a ring 689

species. Science 307:414-416. 690

Irwin, D. E. 2012. Local adaptation along smooth ecological gradients causes phylogeographic 691

breaks and phenotypic clustering. The American Naturalist 180:35-49. 692

Jombart, T. 2008. Adegenet: a R package for the multivariate analysis of genetic markers. 693

Bioinformatics 24:1403-1405. 694

Jombart, T., S. Devillard, A. B. Dufour, and D. Pontier. 2008. Revealing cryptic spatial patterns 695

in genetic variability by a new multivariate method. Heredity 101:92-103. 696

Kearns, A. M., L. Joseph, S. V. Edwards, and M. C. Double. 2009. Inferring the phylogeography 697

and evolutionary history of the splendid fairy-wren Malurus splendens from 698

mitochondrial DNA and spectrophotometry. J Avian Biol 40:7-17. 699

Legendre, P., and M. J. Anderson. 1999. Distance-based redundancy analysis: Testing 700

multispecies responses in multifactorial ecological experiments. Ecol Monogr 69:1-24. 701

Legendre, P., and M.-J. Fortin. 2010. Comparison of the Mantel test and alternative approaches 702

for detecting complex multivariate relationships in the spatial analysis of genetic data. 703

Molecular ecology resources 10:831-844. 704


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

42

Lerner, H. R. L., M. Meyer, H. F. James, M. Hofreiter, and R. C. Fleischer. 2011. Multilocus 705

Resolution of Phylogeny and Timescale in the Extant Adaptive Radiation of Hawaiian 706

Honeycreepers. Curr Biol 21:1838-1844. 707

Librado, P., and J. Rozas. 2009. DnaSP v5: a software for comprehensive analysis of DNA 708

polymorphism data. Bioinformatics 25:1451. 709

Lohse, K., B. Sharanowski, and G. N. Stone. 2010. Quantifying the Pleistocene history of the 710

oak gall parasitoid Cecidostiba fungosa using twenty intron loci. Evolution 64:2664-711

2681. 712

Loynes, K., L. Joseph, and J. S. Keogh. 2009. Multi-locus phylogeny clarifies the systematics of 713

the Australo-Papuan robins (Family Petroicidae, Passeriformes). Mol Phylogenet Evol 714

53:212-219. 715

Maguire, G. S., P. J. Guay, and R. A. Mulder. 2006. Isolation and characterization of 716

microsatellite markers in the southern emu-wren (Stipiturus malachurus: Aves). Mol 717

Ecol Notes 6:422-424. 718

Malekian, M., S. J. B. Cooper, J. A. Norman, L. Christidis, and S. M. Carthew. 2010. Molecular 719

systematics and evolutionary origins of the genus Petaurus (Marsupialia: Petauridae) in 720

Australia and New Guinea. Mol Phylogenet Evol 54:122-135. 721

McCracken, K. G., and M. D. Sorenson. 2005. Is homoplasy or lineage sorting the source of 722

incongruent mtDNA and nuclear gene trees in the stiff-tailed ducks (Nomonyx-Oxyura)? 723

Syst Biol 54:35. 724

McDonald, J. H., and M. Kreitman. 1991. Adaptive protein evolution at the Adh locus in 725

Drosophila. Nature 351:652-654. 726


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

43

Meiklejohn, C. D., K. L. Montooth, and D. M. Rand. 2007. Positive and negative selection on 727

the mitochondrial genome. Trends Genet 23:259-263. 728

Mishmar, D., E. Ruiz-Pesini, P. Golik, V. Macaulay, A. G. Clark, S. Hosseini, M. Brandon, K. 729

Easley, E. Chen, and M. D. Brown. 2003. Natural selection shaped regional mtDNA 730

variation in humans. Proceedings of the National Academy of Sciences 100:171. 731

Moghadam, H. K., M. A. Pointer, A. E. Wright, S. Berlin, and J. E. Mank. 2012. W chromosome 732

expression responds to female-specific selection. Proceedings of the National Academy 733

of Sciences 109:8207-8211. 734

Nosil, P., S. P. Egan, and D. J. Funk. 2008. Heterogeneous genomic differentiation between 735

walking-stick ecotypes: ‘‘Isolation by adaptation’’ and multiple roles for divergent 736

selection. Evolution 62:316–336. 737

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, R. B. O'Hara, G. L. Simpson, P. Solymos, M. 738

Henry, H. Stevens, and H. Wagner. 2011. vegan: Community Ecology Package. R 739

package version 1.17-11. 740

Petit, R. J., and L. Excoffier. 2009. Gene flow and species delimitation. Trends Ecol Evol 741

24:386-393. 742

Phillips, S., R. Anderson, and R. Schapire. 2006. Maximum entropy modeling of species 743

geographic distributions. Ecol Model 190:231-259. 744

Pinho, C., and J. Hey. 2010. Divergence with Gene Flow: Models and Data. Annual Review of 745

Ecology, Evolution, and Systematics 41:215-230. 746

Price, T. 2008. Speciation in birds. Roberts and Company Publishers., Greenwood Village, 747

Colorado. 748


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

44

Primmer, C. R., A. P. Moller, and H. Ellegren. 1995. Resolving genetic relationships with 749

microsatellite markers: A parentage testing system for the swallow Hirundo rustica. Mol 750

Ecol 4:493-498. 751

Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using 752

multilocus genotype data. Genetics 155:945-959. 753

R Development Core Team. 2011. R: A Language and Environment for Statistical Computing. R 754

Foundation for Statistical Computing, Vienna, Austria. 755

Rand, D. M., R. A. Haney, and A. J. Fry. 2004. Cytonuclear coevolution: the genomics of 756

cooperation. Trends Ecol Evol 19:645-653. 757

Ribeiro, Â. M., P. Lloyd, and R. C. K. Bowie. 2011. A tight balance between natural selection 758

and gene flow in a southern African arid-zone endemic bird. Evolution DOI: 759

10.1111/j.1558-5646.2011.01397.x. 760

Schneider, C. J., T. B. Smith, B. Larison, and C. Moritz. 1999. A test of alternative models of 761

diversification in tropical rainforests: Ecological gradients vs. rainforest refugia. Proc 762

Natl Acad Sci U S A 96:13869-13873. 763

Schodde, R., and I. J. Mason. 1999. The directory of Australian birds. A taxonomic and 764

zoogeographic atlas of the biodiversity of birds in Australia and its Territories. 765

Passerines. CSIRO Publishing Collingwood. 766

Shen, Y. Y., L. Liang, Z. H. Zhu, W. P. Zhou, D. M. Irwin, and Y. P. Zhang. 2010. Adaptive 767

evolution of energy metabolism genes and the origin of flight in bats. Proceedings of the 768

National Academy of Sciences 107:8666-8671. 769

Smadja, C. M., and R. K. Butlin. 2011. A framework for comparing processes of speciation in 770

the presence of gene flow. Mol Ecol 20:5123–5140. 771


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

45

Sorenson, M. D., J. C. Ast, D. E. Dimcheff, T. Yuri, and D. P. Mindell. 1999. Primers for a PCR-772

Based Approach to Mitochondrial Genome Sequencing in Birds and Other Vertebrates. 773

Mol Phylogenet Evol 12:105-114. 774

Sorenson, M. D., C. N. Balakrishnan, and R. B. Payne. 2004. Clade-limited colonization in 775

brood parasitic finches (Vidua spp.). Syst Biol 53:140. 776

Spottiswoode, C. N., K. F. Stryjewski, S. Quader, J. F. R. Colebrook-Robjent, and M. D. 777

Sorenson. 2011. Ancient host specificity within a single species of brood parasitic bird. 778

Proceedings of the National Academy of Sciences 108:17738-17742. 779

Stephens, M., and P. Donnelly. 2003. A comparison of bayesian methods for haplotype 780

reconstruction from population genotype data. The American Journal of Human Genetics 781

73:1162-1169. 782

Stephens, M., N. J. Smith, and P. Donnelly. 2001. A new statistical method for haplotype 783

reconstruction from population data. The American Journal of Human Genetics 68:978-784

989. 785

Sunnucks, P., M. J. Blacket, J. M. Taylor, C. J. Sands, S. A. Ciavaglia, R. C. Garrick, N. N. Tait, 786

D. M. Rowell, and A. Pavlova. 2006. A tale of two flatties: different responses of two 787

terrestrial flatworms to past environmental climatic fluctuations at Tallaganda in montane 788

southeastern Australia. Mol Ecol 15:4513-4531. 789

Symula, R., J. S. Keogh, and D. C. Cannatella. 2008. Ancient phylogeographic divergence in 790

southeastern Australia among populations of the widespread common froglet, Crinia 791

signifera. Mol Phylogenet Evol 47:569-580. 792

Tajima, F. 1996. The amount of DNA polymorphism maintained in a finite population when the 793

neutral mutation rate varies among sites. Genetics 143:1457-1465. 794


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

46

Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and S. Kumar. 2011. MEGA5: 795

Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary 796

Distance, and Maximum Parsimony Methods. Mol Biol Evol (submitted). 797

Tieleman, B. I., M. A. Versteegh, A. Fries, B. Helm, N. J. Dingemanse, H. L. Gibbs, and J. B. 798

Williams. 2009. Genetic modulation of energy metabolism in birds through 799

mitochondrial function. Proceedings of the Royal Society B: Biological Sciences 800

276:1685. 801

Tieleman, B. I., J. B. Williams, and P. Bloomer. 2003. Adaptation of metabolism and 802

evaporative water loss along an aridity gradient. Proceedings of the Royal Society of 803

London. Series B: Biological Sciences 270:207. 804

Toews, D. P. L., and A. Brelsford. 2012. The biogeography of mitochondrial and nuclear 805

discordance in animals. Mol Ecol 21:3907-3930. 806

Tonnis, B., P. R. Grant, B. R. Grant, and K. Petren. 2005. Habitat selection and ecological 807

speciation in Galápagos warbler finches (Certhidea olivacea and Certhidea fusca). 808

Proceedings of the Royal Society B: Biological Sciences 272:819-826. 809

Turmelle, A. S., T. H. Kunz, and M. D. Sorenson. 2011. A tale of two genomes: contrasting 810

patterns of phylogeographic structure in a widely distributed bat. Mol Ecol. 811

Vallender, R., R. J. Robertson, V. L. Friesen, and I. J. Lovette. 2007. Complex hybridization 812

dynamics between golden-winged and blue-winged warblers (Vermivora chrysoptera and 813

Vermivora pinus) revealed by AFLP, microsatellite, intron and mtDNA markers. Mol 814

Ecol 16:2017-2029. 815


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

47

Via, S. 2012. Divergence hitchhiking and the spread of genomic isolation during ecological 816

speciation-with-gene-flow. Philosophical Transactions of the Royal Society B: Biological 817

Sciences 367:451-460. 818

Waltari, E., and S. V. Edwards. 2002. Evolutionary dynamics of intron size, genome size, and 819

physiological correlates in archosaurs. Am Nat:539-552. 820

Webb, W. C., J. M. Marzluff, and K. E. Omland. 2011. Random interbreeding between cryptic 821

lineages of the Common Raven: evidence for speciation in reverse. Mol Ecol 20:2390-822

2402. 823

Williams, J. B., and B. I. Tieleman. 2005. Physiological adaptation in desert birds. Bioscience 824

55:416-425. 825

Zink, R. M., and G. F. Barrowclough. 2008. Mitochondrial DNA under siege in avian 826

phylogeography. Mol Ecol 17:2107-2121. 827

Zink, R. M., A. Pavlova, S. Drovetski, and S. Rohwer. 2008. Mitochondrial phylogeographies of 828

five widespread Eurasian bird species. Journal of Ornithology 149:399-413. 829


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

48

FIGURE CAPTIONS 830

Figure 1. (a) Sampling localities (dots) and geographic distribution of mitochondrial haplogroups 831

(colors): light red dots - A1, dark red dots - A2, black dots - B; numbers show distribution of 832

haplotypes (labeled as on Fig. 3). Insets show two locations where both haplogroups were 833

sampled: in Tenterfield, a haplogroup A male was collected on the same day from the same 834

group of birds as a haplogroup B male and female; in Shelbourne, a haplogroup B male and 835

haplogroup A male were caught in May and October 2008, respectively. (b) Distribution of Q-836

values (pie charts within circles) for two microsatellite genetic clusters (‘red’ and ‘black’), 837

detected by STRUCTURE; each circle’s outer border color shows the mitochondrial haplogroup for 838

each individual: light red - A1, dark red - A2, black - B; samples are spaced apart for easier 839

viewing. Altitudinal layers of the Great Dividing Range are: light grey - 200 to 500 m, grey - 500 840

to 1000 m, dark grey - above 1000 m. 841

842

Figure 2. Haplotype networks for mitochondrial ND2 and six nuclear introns. Circles indicate 843

unique haplotypes, with area proportional to haplotype frequency. Connections between circles 844

are a single mutation unless indicated otherwise (italic). Colors correspond to the haplogroups: 845

grey- A (light grey - A1, corresponding to light red dots on Fig. 1A, dark grey - A2 [two 846

haplotypes, H9 and H12, on mtDNA network], dark red dot on Fig. 1A), white - B (black dots on 847

Fig. 1A); black pies on nuclear networks indicate individuals sequenced for nDNA but not for 848

mtDNA (thus, mtDNA haplogroup for these individuals was unknown). 849

850

Figure 3. Geographic distribution of spatial principal component (sPCA) scores for 851

mitochondrial ND2 (mt-sPC1 and mt-sPC2), microsatellites (msat-sPC1 and msat-sPC2), 852


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

49

GAPDH (GAPDH-sPC1) and MUSK-I4MUSK-I3 (MUSK-I4MUSK-I3-sPC1). Large black 853

squares indicate samples well differentiated from those denoted by large white squares, small 854

squares are less differentiated. Lines indicate values of sPCA scores for interpolated surface 855

(lags). Narrowly spread lines and a lack of small squares indicate discontinuity or sharp 856

geographic transition between two groups (as for mtDNA). Presence of smaller squares and 857

broadly spread lines indicate a cline (as for microsatellites). 858

859

Figure 4. Distribution of two major mtDNA lineages of Eopsaltria australis in environmental 860

space, as defined by two BIOCLIM variables, max temperature of warmest month and 861

precipitation of driest month. Two localities where both haplogroups occurred together are 862

marked with arrows.863

864


123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

perched at the mito-nuclear crossroads: divergent mitochondrial lineages correlate with environment...

Documents