genetic consequences of genetic mixing in mammal ... consequences of genetic mixing in mammal...

Genetic consequences of genetic mixing in mammal

translocations in Western Australia using case studies of

Burrowing bettongs and Dibblers

Rujiporn Thavornkanlapachai

Bachelor of Science in Environmental Science (Hons)

This thesis is presented for the degree of

Doctor of Philosophy

School of Animal Biology

Faculty of Natural and Agricultural Science

The University of Western Australia

2016

We are but one thread within it.

Whatever we do to the web, we do to ourselves.

All things are bound together.

All things connect.

– Chief Seattle, 1854

ii

SUMMARY

As more species become threatened, translocation has become an important conservation

tool to prevent species from extinction. Fragmentation and isolation expose populations

to serious genetic problems such as inbreeding, loss of genetic variation and an elevated

risk of extinction. Conservation managers sometimes face a dilemma that mixing

genetically divergent groups of animals may be the only option for establishing viable

new populations, but it comes at the risk of outbreeding depression. Inconsistent

outcomes following translocation of individuals between populations mean that more

studies are needed to understand the implications of genetic mixing. Using case studies

of two Australian mammals, the dibbler (Parantechinus apicalis) and burrowing bettong

(Bettongia lesueur), I demonstrated how fine-scale genetic structure can be used to guide

decisions on selecting individuals from a single source population. Then I examined the

genetic consequences of establishing new populations using individuals from multiple

source populations with different levels of genetic divergence.

Chapter one assessed population genetic structure of the last mainland dibbler population

in the Fitzgerald River National Park. This study revealed two genetic subpopulations

located on western and eastern sides of the park, approximately 60 km apart. The large

geographic distance between the regions and the limited dispersal ability of this species

are likely to be the main factors that restricted gene flow between these subpopulations. I

also found evidence of female philopatry and male-biased dispersal. Females showed

significant positive correlations between estimated levels of relatedness and distance

classes up to 200 m, while males showed no spatial genetic heterogeneity. From this

study, I recommended the western and eastern sides of the park are managed as separate

subpopulations and females should be sampled at least 200 m apart for captive breeding

and translocation programs.

Chapter two demonstrated consequences of genetic mixing between the above

subpopulations in a mainland translocation of the dibbler. Here, I assessed outcomes of

mixing between source populations that had low levels of genetic divergence (FST = 0.05).

We detected evidence of interbreeding between animals from different subpopulations.

Genetic composition of the captive and translocated populations changed as the

contributions from each of the subpopulations varied through time. At least 94% of gene

diversity and 82% of allelic richness were maintained in the captive and translocated

iii

population. Interbreeding between animals from different subpopulations also reduced

genetic relatedness among offspring. Nonetheless, we detected 10 – 16 fold reduction in

the effective population size.

Chapter three illustrated genetic consequences of an island translocation of the dibbler

founded from two geographically adjacent island populations (FST = 0.42). Despite body

size differences between dibblers from different source populations, we found evidence

of genetic mixing both in captivity and in the wild. However, there was a bias towards

larger-sized ancestry because larger males from one ancestry had higher reproductive

success than lighter males from another ancestry. The effective population size of the

translocated population was 18% – 89% lower than the source populations. Even so,

genetic diversity of the translocated population was relatively higher than both source

populations but was not much more than the most diverse source population. Body weight

and pes length of wild-born males in the translocation site was intermediate of those in

the source populations. Females in the source and translocated populations did not show

any differences in body size.

Chapter four investigated the genetic consequences of mixing two geographically isolated

island populations in a translocation of burrowing bettongs to mainland Australia. The

two source populations showed high levels of divergence (FST = 0.42 and ϕST = 0.72 for

nuclear and mitochondrial DNA respectively) as well as clear differences in body size. I

found evidence of reciprocal interbreeding between the two source populations, though

there was a bias towards crosses between males from the smaller-sized source population

and females from the larger-sized source population. Genetic composition of the

translocated population was also influenced by early mortality. The translocated

population showed significant 93% reduction in the effective population size, which is

expected as a result of founder effects. Nonetheless, the translocated population had

higher levels of genetic variation relative to one, but not both source populations. F1

offspring’s body size was more similar to the source population that was larger and

heavier.

My study highlights on the importance of incorporating knowledge of fine-scale genetic

structure when sampling wild individuals for captive breeding and translocation

programs. Outcomes of genetic mixing from this thesis also added to rare empirical

studies of marsupial translocations that were founded from multiple sources. Genetic

mixing has shown to benefit the newly established populations. Increased levels of

iv

genetic diversity and reduced genetic relatedness among offspring were reported in all

three cases. Nevertheless, they also showed a significant reduction of the effective

population size as a result of founder effects. This study shows that population lineages

within newly established populations are prone to changes from mortality and/or

reproductive variance among founders and release strategies. Therefore, genetic

monitoring is important not only for assessing translocation success but also determining

whether animal supplementations are needed.

vi

TABLE OF CONTENTS

Summary ii

Table of contents vi

Acknowledgements x

Statement of candidate contribution xii

CHAPTER ONE: General Introduction

1.1 Introduction 1

1.2 Translocation as a management tool 1

1.3 Role of genetic information in translocation 2

1.4 Common genetic problems in translocation 4

1.4.1 Founder effect 4

1.4.2 Small effective population size 4

1.4.3 Inbreeding 5

1.5 Genetic mixing as a management option in translocation 6

1.5.1 Benefits of genetic mixing 6

1.5.2 Complexity of genetic rescue 7

1.6 Predicting outcomes of genetic mixing 8

1.7 Knowledge gaps in translocation and thesis aims 10

1.8 Study species 13

1.8.1 The dibbler (Parantechinus apicalis) 13

1.8.2 The burrowing bettong (Bettongia lesueur) 18

1.9 Study aims and thesis structure 21

CHAPTER TWO: Fine-scale population genetic structure of the mainland dibbler,

Parantechinus apicalis

2.1 Abstract 27

2.2 Introduction 28

2.3 Materials and methods 30

2.3.1 Study sites and sample collection 30

2.3.2 DNA extraction and microsatellite genotyping 31

2.3.3 Data analysis 32

2.4 Results 36

2.4.1 Genetic variation within sites 36

2.4.2 Population structure 38

2.4.3 Sex-biased dispersal 41

2.5 Discussion 42

2.5.1 Landscape-scale population subdivision 42

2.5.2 Fine-scale population structure and sex-biased dispersal 44

vii

2.5.3 Management implications 45

CHAPTER THREE: Temporal variation in the genetic composition of a newly

established population of dibblers (Parantechinus apicalis) reflects translocation

history

3.1 Abstract 49

3.2 Introduction 50


3.3.1 Study site and sample collection 52

3.3.2 DNA extraction and microsatellite genotyping 54


3.4 Results 57

3.4.1 Effects of translocation on genetic variability 57

3.4.2 Population structure of captive and translocated populations 61

3.4.3 Genetic relatedness comparisons 67

3.5 Discussion 69

3.5.1 Genetic consequences of mixing subpopulations 69

3.5.2 Consequences of admixture on population structure 70

3.5.3 Genetic mixing and relatedness 71

3.5.4 Conservation implications 72

CHAPTER FOUR: Admixture between genetically diverged island populations

bolsters genetic diversity within a newly established island population of the dibbler

(Parantechinus apicalis), but does not prevent subsequent loss of genetic variation

4.1 Abstract 75

4.2 Introduction 76


4.3.1 Sampling and DNA extraction 79

4.3.2 Microsatellite variation 80


4.4 Results 85

4.4.1 Genetic variation within populations 85

4.4.2 Population bottlenecks and estimates of Ne 87

4.4.3 Population structure and genetic mixing within the translocated

population 88

4.4.4 Differences in body size between source populations and factors

influencing mating and reproductive success in captivity 91

4.4.5 Population viability analysis 94

4.5 Discussion 95

4.5.1 Phenotype and genetic differentiation between island populations 95

4.5.2 Genetic composition and influence of male body size on reproductive

success 97

4.5.3 Consequences of genetic mixing on genetic diversity 98

4.5.4 Management implications 100

viii

CHAPTER FIVE: Asymmetrical introgression between genetically distinct

populations of the burrowing bettong (Bettongia lesueur) in a newly established

translocated population

5.1 Abstract 103

5.2 Introduction 104


5.3.1 Studied species 107

5.3.2 Translocation history 1098

5.3.3 Sampling and DNA extraction 1078

5.3.4 Mitochondrial DNA control region sequences 109

5.3.5 Microsatellites 109

5.3.6 Data Analysis 109

5.4 Results 114

5.4.1 Mitochondrial DNA variation 114

5.4.2 Microsatellite variation 118

5.4.3 Population structure and genetic mixing 1189

5.4.4 Effects of introgression on body size, reproductive fitness and

survival probability 122

5.5 Discussion 126

5.5.1 Phenotypic and genetic differentiation between island populations 126

5.5.2 Genetic consequences of mixing geographically isolated island

populations 126

5.5.3 Considerations for conservation and management 129

CHAPTER SIX: General Discussion

6.1 Introduction 130

6.2 Using genetic structure and dispersal pattern to assist founder selection 131

6.3 Outcomes of translocations established using multiple source populations 132

6.3.1 Retention of genetic diversity 132

6.3.2 Genetic similarity and inbreeding coefficient 133

6.3.3 Effective population size 136

6.4 Consequences of genetic mixing on offspring 137

6.5 Factors influencing the genetic contributions of parental lineages 137

6.6 Implications for conservation management 139

6.6.1 Sampling strategies 139

6.6.2 Monitoring 140

6.6.3 Population size and long-term persistence 140

6.7 Study limitations and future research 141

6.8 Conclusion 143

References 144

Appendices 163

x

ACKNOWLEDGEMENTS

Although this thesis has one name on its cover, the work inside could not have been

accomplished without help and support from so many people.

I am extremely grateful for guidance and support from my two main supervisors, Dr Jason

Kennington and Dr Harriet Mills. Jason’s intelligence was inspiring. His advice during

the design, analysis and writing stages helped me tremendously. Harriet’s help with

finding funding enabled me to pursue the project that I wanted to do. Her advice on

various stages of this project and help with writing also improved quality of this thesis. I

would also like to thank my external supervisors, Dr Kym Ottewell and Keith Morris.

Kym’s constructive comments and her never-ending positive attitude had not only helped

improving the thesis but also lifted the heart of its writer. Without Keith, getting funding

for this project would be a difficult task and it would certainly affect the outcome of this

project so I am grateful for his help.

This project would not be completed without funding from the Department of Parks and

Wildlife (DPaW) and the University of Western Australia (UWA). I am also grateful for

UWA for granting the Australian Postgraduate Award, UWA Top-up Scholarship, and a

Travel Award. I also thank DPaW for providing Ad Hoc Scholarship that allowed me to

continue writing for six months.

I appreciate helps with sample collections from following people: Judy Dunlop, Dr

Colleen Sims, Cathy Lambert, Dr Tony Friend, and DPaW people involved in the

trapping of burrowing bettongs at Lorna Glen and dibblers at the Fitzgerald River

National Park. Particularly, I want to thank Colleen for providing an additional mortality

record of the burrowing bettongs at Lorna Glen. Much help with the captive-bred dibblers

came from Cathy for gathering dibbler samples and providing additional data of these

animals. My gratitude is extended to Tim Button from DPaW for providing GPS locations

of the dibblers trapped in the FRNP. Lastly, I am thankful for the burrowing bettong

mitochondrial DNA sequences provided by Felicity Donaldson from her PhD.

I owe many thanks to numerous people at UWA. First, I thank Yvette Hitchen and

Sheralee Lukehurst for laboratory technical supports. Second, I want to thank my fellow

post-graduate students for their supports both technical and emotional. These include

Kaori Yokochi, Veronica Phillips, Gabriella Flacke, Miriam Sullivan, Jamie Tedeschi,

xi

Natalie Rosser, Luke Thomas, and Elizabeth Wiley. Lastly, I thank my one brave

volunteer, Natasha Tay, for helping in the laboratory.

I am blissed with the support from my friends and families both in and outside Australia.

I would not come this far without my parents and their support of my decision to pursuit

a PhD. I thank my dad for his great advice and thoughtful conversations. I am forever

indebted of my mum’s support both before and during the PhD program. I am grateful of

my step-parents’ support and to have them seeing me as their own.

Throughout the years of this project and in the face of stressful times, Ray managed to be

my best friend and a loving husband. I am extremely grateful to have someone like him

to support me through good and difficult times.

xii

STATEMENT OF CANDIDATE CONTRIBUTION

All procedures used in these experiments were approved by the University of Western

Australia Animal Ethics Committee (RA/3/500/@AEC19/07/2012). Vivisection licenses

from the Zoological Parks Authority Animal Ethics Committee (49063 Veterinary

Department – Provision of an animal health care service, 24252 Native Species Breeding

Program – Tissue collection and storage of Dibbler) and the Department of Parks Wildlife

(Permit# DEC AEC: 62/2009 and 66/2009) were held during the duration of the

experiments.

The following people and institutions contributed to the publication of work undertaken

as part of this thesis:

Rujiporn Thavornkanlapachai (UWA) Candidate

Dr. Harriet Mills (UWA) Author 1

Dr. Jason Kennington (UWA) Author 2

Dr. Kym Ottwell (DPaW) Author 3

Dr. Tony Friend (DPaW, Perth Zoo) Author 4

Cathy Lambert (Perth Zoo) Author 5

Judy Dunlop (Murdoch University, DPaW) Author 6

Keith Morris (DPaW) Author 7

Felicity Donaldson (Environmental 360) Author 8

xiii

Author details and their roles:

Thavornkanlapachai, R., W.J. Kennington, K. Ottewell, T. Friend, and H.R.

Mills. In prep. Fine-scale population genetic structure of the mainland dibbler

(Parantechinus apicalis). PLoS ONE.

Sample

and data

collection

Design and

development

Laboratory

work

Analysis Writing Total

contribution

Candidate 75%

Author 1 8%

Author 2 8%

Author 3 5%

Author 4 4%


Mills. In prep. Temporal variation in the genetic composition of a newly

established population of dibbler (Parantechinus apicalis) reflects translocation

history. Conservation Genetics.

Sample

and data

collection

Design and

development

Laboratory

work


contribution

Candidate 75%

Author 1 8%

Author 2 8%

Author 3 5%

Author 4 4%

xiv

Thavornkanlapachai, R., H.R. Mills, K. Ottewell, C. Lambert, T. Friend, and

W.J. Kennington. In prep. Admixture between genetically diverged island

populations bolsters genetic diversity within a newly established island

population of the dibbler (Parantechinus apicalis), but does not prevent

subsequent loss of genetic variation. Biological Conservation.

Sample

and data

collection

Design and

development

Laboratory

work


contribution

Candidate 75%

Author 1 8%

Author 2 8%

Author 3 5%

Author 4 2%

Author 5 2%

Thavornkanlapachai, R., H.R. Mills, J. Dunlop, K. Morris, F. Donaldson, and

W.J. Kennington. Submitted. Asymmetrical introgression between genetically

distinct populations of the boodie (Bettongia lesueur) in a newly established

translocated population. Molecular Ecology.

Sample

and data

collection

Design and

development

Laboratory

work


contribution

Candidate 70%

Author 1 8%

Author 2 10%

Author 3 5%

Author 6 3%

Author 7 2%

Author 8 2%

xv

It is hereby declared that this thesis is a result of my own investigation, except for where

the work of others is acknowledged. It has not been previously submitted for a degree at

any tertiary education.

Signed:

Rujiporn Thavornkanlapachai Harriet Mills

Candidate Coordinating Supervisor

1

CHAPTER ONE

General Introduction

1.1 INTRODUCTION

Australia is home to at least 386 known mammal species (7% of the world total), of which

87% of these are endemic to Australia (Chapman, 2009). Habitat destruction, habitat

fragmentation and introduced predators such as European red foxes (Vulpes vulpes) and

feral cats (Felis catus) have caused dramatic declines in the abundances of many

Australian mammal species over the last 200 years leading to an unprecedented rate of

extinction (Burbidge and McKenzie, 1989, Andersen et al., 2012, Smith and Quin, 1996).

Thirty-four species have been driven to extirpation and a further 56 species are now

threatened with extinction (IUCN, 2015). Many others persist as isolated populations or

on offshore islands (Burbidge et al., 2009). Furthermore, the decline in population sizes

and numbers of species is continuing (Woinarski et al., 2001). To combat these worsening

problems, active management is needed to restore lost populations and ensure species

continuity.

1.2 TRANSLOCATION AS A MANAGEMENT TOOL

Translocation is a popular management tool for species restoration and conservation. It

has been used to increase and expand populations by moving living organisms from one

place to another (IUCN, 1987). There are two main types of translocation: population

restoration and conservation introduction (IUCN/SSC, 2013). Population restoration is

when organisms are moved within their indigenous range, either to existing populations

(Reinforcement) or to the location that the population has disappeared from

(Reintroduction) (IUCN/SSC, 2013). In contrast, conservation introduction is the

movement of organisms outside their indigenous range either because the historical range

is no longer suitable (Assisted colonization) or to perform a specific ecological function

(Ecological replacement) (IUCN/SSC, 2013). Despite their popularity, translocations

have a low success rate. In a series of reviews, success rates (defined as a self-sustaining

population as stated by the authors based on specific objectives) varied from 11% to 46%

(Sheean et al., 2012, Fischer and Lindenmayer, 2000, Beck et al., 1994). Many

translocations fail for unknown reasons even when the causes of decline have been

removed (Fischer and Lindenmayer, 2000). Of those with known failure, a common cause

2

for Australian mammals is predation by foxes and cats, which is possibly exacerbated by

predator-naivety of captive-bred animals (Moseby et al., 2011). In an effort to improve

the success rate of conservation translocations, research has generally focused on factors

affecting population establishment such as habitat quality, size of the release site, the

management of threatening process, and the number of individuals released (Short, 2009,

Moorhouse et al., 2009, Sheean et al., 2012, Wolf et al., 1998). An additional factor

considered to be important is genetic characteristics of the source population (Fischer and

Lindenmayer, 2000, Moseby et al., 2011).

1.3 ROLE OF GENETIC INFORMATION IN TRANSLOCATION

Genetic diversity has been recognized as one of three forms of biodiversity (McNeely et

al., 1990). It gives individuals and populations the capacity to tolerate biological and

environmental stress such as disease outbreak and drought. The level of genetic variation

in the founding individuals determines the short-term evolutionary potential in a

translocated population and therefore plays an important role in the resilience and

persistence of a new population. As such, genetic data provide important information for

translocations at all stages including founder selection, the breeding of animals for release

and population monitoring.

IUCN/SSC guidelines (2013) recommend that selected individuals for any translocations

should provide adequate genetic diversity. Thirty to fifty individuals have been suggested

to capture 90% – 95% of gene diversity (Hedrick, 2000, Ottewell et al., 2014, Allendorf

and Luikart, 2007). These individuals should come from habitats that are geographically

close or are similar environments to the destination habitat (Frankham et al., 2011). If

founders are selected from multiple populations, these populations must be from the

closest race or type to avoid genetic incompatibilities (IUCN/SSC, 2013). However,

getting access to such populations may not be possible, especially when founders come

from offshore islands.

Moritz (1999) suggested using conservation units such as Evolutionary Significant Units

(ESUs) and Management Units (MUs) to guide translocation decisions based on

environmental gradients, historical barriers to gene flow and the analysis of mtDNA

phylogeography and nuclear allele frequencies. He stated that mixing can occur between

MUs if remnant populations show signs of inbreeding and increased fragmentation, but

never between ESUs. Similarly, Weeks et al. (2011) suggested selection of source

3

populations also depends on levels of genetic variability and the nature of environmental

gradients along which populations are being introduced. Therefore, consideration of

phylogeographic variation, population connectivity, spatial genetic structure, and habitat

variation of source populations are important in the founder selection process.

Individuals used to establish translocations are either wild-born or captive-bred animals.

Translocations founded from wild-born animals generally have a better success rate than

those established from captive-bred animals (Fischer and Lindenmayer, 2000, Wolf et al.,

1996, Jule et al., 2008). However, populations of conservation concern may be small in

size. Sourcing individuals for translocation from these populations can lead to over-

harvesting causing a significant reduction in the effective population size and the

associated genetic problems (Allendorf et al., 2008). Captivity provides a benign and

stable environment (i.e. food, health care, removal of predators) that allows high and

constant population growth (Robert, 2009). As a result, many translocation programs use

captive-born animals to establish new populations (Moro, 2003, Nelson et al., 2002, Sigg,

2006). Nonetheless, there are genetic changes during captivity that can jeopardize the

ability of captive populations to reproduce and survive, including loss of genetic diversity

(Neveu et al., 1998), inbreeding depression (Swinnerton et al., 2004), accumulation of

new mildly deleterious mutations (Bryant and Reed, 1999), and genetic adaptation to

captivity (Frankham, 2008). Although there are ways to reduce genetic problems in

captivity, such as minimizing the number of generations in captivity, equalizing family

size, maintaining several small populations with occasional gene flow, immigration from

the wild, and minimize selection (Frankham, 2008, Robert, 2009, Williams and Hoffman,

2009), genetic monitoring of captive populations remains essential.

Genetic monitoring is important for the effective management of reintroduced or

translocated populations. In the short-term, populations can be monitored for changes in

the mating system, hybridization and population structure as these populations become

established (Schwartz et al., 2007). In the long-term, information gathered at different

time points can be used to study larger-scale population dynamics (e.g. De Barba et al.,

2010) and may prompt actions like animal supplementations (if needed) to improve the

probability of population persistence (e.g. Ottewell et al., 2014).

4

1.4 COMMON GENETIC PROBLEMS IN TRANSLOCATION

1.4.1 Founder effect

One of the common genetic problems encountered by translocated populations is the

founder effect. A founder effect occurs when a new population is established from a small

group of individuals (Hartl and Clark, 1997). Factors such as mortality, a biased sex ratio

and skewed breeding success also contributed to further loss of founding individuals

(Jamieson, 2011, Biebach and Keller, 2012). This causes abrupt changes in allele

frequencies and loss of genetic variation (Allendorf et al., 2012). Thus, the new

population may not be genetically representative of the source population. For example,

in reintroductions of four monogamous bird species in New Zealand, only 4 – 25

individuals out of 10 – 58 birds released contributed to the gene pool. This resulted in

increased inbreeding levels and loss of genetic diversity in the translocated populations

after seven breeding seasons (Jamieson, 2011).

1.4.2 Small effective population size

The effective population size (Ne) is the size of an ideal population that would lose

heterozygosity or gene frequency due to drift at a rate equal to the observed population

(Wright, 1939). Genetic drift is a random change of allele frequencies from generation to

generation. In a large population, alleles that benefit survivorship will occur at a high

frequency and deleterious alleles will be present at a low frequency as a result of natural

selection (Binks et al., 2007). As populations become smaller, selection is overwhelmed

by genetic drift and results in random fixation or loss of allelic diversity (Miller and

Lambert, 2004). Ne is associated with the rate at which heterozygosity is lost through

genetic drift at a rate of 1

2𝑁𝑒 per generation (Wright, 1939, Wright, 1922), so the smaller

Ne, the greater loss of heterozygosity. Loss of genetic variation leads to a loss in the

population’s ability to evolve/adapt to changing conditions, which put small populations

at a greater risk of extinction from demographic and environmental fluctuations (Newman

and Pilson, 1997). In small populations, related individuals are more likely to breed with

each other. This increases levels of inbreeding and subsequently reduces fitness of the

population, a phenomenon known as inbreeding depression (e.g. Hemmings et al., 2012,

Eldridge et al., 1999, Grueber et al., 2010). Small Ne is particularly of concern when the

source populations selected for translocation have been isolated and remained small in

size over a period of time. Small island populations, for example, are often viewed as a

poor source of genetic variation, highly inbred, and frequently found to be differentiated

5

from the mainland population (Eldridge et al., 1999, Frankham, 1998, Miller et al.,

2011a). Using these populations as a source of founders is therefore likely to reduce the

probability of new populations successfully establishing and persisting in the

translocation site (but see Moro, 2003, Smith and Hughes, 2008, Gregory et al., 2012).

Small Ne is also a concern for populations during the establishment phase. Slow

population growth could expose the population to further genetic effects of small

population size. The severity depends on the size of the population and how long it takes

to return to a large size (Nei et al., 1975, Maruyama and Fuerst, 1985).

1.4.3 Inbreeding

Inbreeding refers to situations where relatives mate at a higher than expected frequency.

Inbreeding increases homozygosity and decreases observed heterozygosity lower than the

level expected under Hardy–Weinberg Equilibrium. This leads to expression of recessive

deleterious alleles and/or reduces heterozygosity at loci where heterozygotes have a

selective advantage (Allendorf and Luikart, 2007). Deleterious recessive alleles are

introduced to the genome by mutation. They are present in all species because natural

selection is insufficient in removing them as most copies are hidden in heterozygotes that

do not reduce fitness unless expressed in homozygotes (Allendorf and Luikart, 2007).

Rare and highly deleterious recessive mutations may be ‘purged’ to reduce genetic load

as a result of selection (Charlesworth and Willis, 2009). However, the ability of

populations to purge their genetic load is reduced if the populations become small and

genetic drift results in fixation of these deleterious mutations (Kimura, 1962). Once the

mutations become fixed, they cannot be removed by purging. The expression of recessive,

deleterious alleles and/or reduction of heterozygosity at loci where heterozygosity is

advantageous results in inbreeding depression, a loss of fitness in the progeny of related

individuals in comparison to unrelated individuals (Charlesworth and Willis, 2009). Some

inbreeding depression is expected in all species (Hedrick and Kalinowski, 2000) and can

be accumulated across different life-stages (Grueber et al., 2010). There are many

examples in the literature illustrating associations between inbreeding and its effects on

fitness such as declines of species’ viability, survival, reproduction, and longevity

(Madsen et al., 1996, Grueber et al., 2010, Hemmings et al., 2012, Nielsen et al., 2012).

For example, a high level of inbreeding has been shown to reduce breeding success in red

deer (Slate et al., 2000), caused early mortality in passerine birds (Hemmings et al., 2012),

increased susceptibility to parasites in Soya sheep (Coltman et al., 1999), and reduced

germination success in White champion (Richards, 2000).

6

1.5 GENETIC MIXING AS A MANAGEMENT OPTION IN TRANSLOCATION

1.5.1 Benefits of genetic mixing

Mixing founders from multiple populations may benefit translocated populations and

overcome many problems caused by founder effects and small population effects

(Crnokrak and Roff, 1999, Keller and Waller, 2002, Hedrick and Kalinowski, 2000,

Madsen et al., 1996, Grueber et al., 2010, Slate et al., 2000). In addition, hybridization

between diverged populations can reverse deleterious effects of inbreeding by masking

deleterious recessives (dominance) or increasing heterozygosity at loci where

heterozygotes have a selective advantage (overdominance) (Edmands and Timmerman,

2003). Long-isolated populations often carry different subsets of alleles as a result of lack

of gene flow, genetic drift, and local selection (Eldridge et al., 1999). By mixing these

populations, a translocated population would receive both sets of alleles, provided that

individuals from both source populations interbreed. For example, translocated

populations of Bembicium vittatum founded from multiple sources showed higher levels

of genetic diversity than source populations (Kennington et al., 2012).

Genetic rescue is a recovery of fitness in outbred offspring relative to the parents, which

occurs through increased heterozygosity or reduced homozygote deleterious allele

frequency (Allendorf and Luikart, 2007). Therefore, genetic rescue is more apparent if

the source populations are highly inbred (Tallmon et al., 2004). Genetic rescue has been

reported in several translocations of both animal (Hedrick and Fredrickson, 2010, Miller

et al., 2012) and plant species (Pickup et al., 2013, Willi et al., 2007).

Heterosis is similar to genetic rescue, but instead of crossing between diverged

populations resulting in recovery of fitness, it leads to enhanced fitness by sheltering

deleterious recessive alleles and increasing heterozygosity where the heterozygotes have

selective advantage over the homozygotes (Allendorf and Luikart, 2007). Often heterosis

is detected in the F1 generation and it is lost or decreased in subsequent generations as

more backcrosses are produced (Edmands, 1999). One additional benefit of genetic

mixing is that traits from both parents may be inherited by offspring, which increases

evolutionary potential and may enable offspring to survive in a wider range of habitats

(e.g. Taylor et al., 2006, Binks et al., 2007).

7

1.5.2 Complexity of genetic rescue

Genetic mixing does not always have a favourable outcome. There are three potential

disadvantages associated with mixed source populations. First, pre-zygotic isolation,

which includes differences in morphology, behaviour, ecology, reproductive biology and

gametic compatibility, prevents individuals from different source populations from

interbreeding (Alexandrino et al., 2005, Latch et al., 2006, Coyne and Orr, 2004). This

may reduce the effective population size and induce genetic problems associated with a

small population size. Second, hybridization between individuals from different

populations may produce offspring that have lower fitness as a result of outbreeding

depression. Outbreeding depression can be intrinsic (environment independent), which

occurs at chromosomal and genic levels. At the chromosomal level, differences in number

or structure may result in the production of gametes with abnormal number of

chromosomes (Allendorf and Luikart, 2007, Fishman and Willis, 2001). At the genic

level, fitness of progeny may be lower due to heterozygote disadvantage, harmful

epistatic interaction between alleles of the parents, or disrupting of co-adapted gene

complexes (Charlesworth and Willis, 2009). Common signs of intrinsic outbreeding

depression include reduction in fertility and viability of hybrid offspring such as sterility

(Fishman and Willis, 2001), low survival rate (Gharrett et al., 1999), slow growth rate

(Huff et al., 2011), decreased reproductive success (Lancaster et al., 2007), and high

susceptibility to diseases (Goldberg et al., 2005). Extrinsic (environment dependent)

outbreeding depression occurs when hybrid offspring are maladapted to either parental

environment due to an intermediate phenotype. For example, a hybridization of two garter

snakes populations (Thamnophis ordinoides) produced a mismatch in body pattern and

behaviour of hybrid snakes that led to higher mortality from predation in comparison to

purebred snakes (Brodie, 1992). Third, hybridization can also threaten a local population

by disrupting local adaptation, which subsequently reduces an overall fitness of the local

population (Lenormand, 2002, Roberts et al., 2010). This is also known as genetic

swamping. For example, hybridization between native cutthroat trout (Oncorhynchus

clarkia) and rainbow trout (O. mykiss) posed a threat to two native cutthroat populations,

causing population declines that have been attributed to asymmetric introgression from

the rainbow trout (Metcalf et al., 2008).

8

1.6 PREDICTING OUTCOMES OF GENETIC MIXING

Predicting the outcome of genetic mixing is not simple. As previously described

interbreeding between different populations may produce various effects on the genetics

and fitness of offspring. The consequence of interbreeding depends on the species and the

distance (genetic or geographic) between the source populations (Figure 1.1) (Lynch,

1991, Allendorf and Luikart, 2007). To further complicate matters, the consequences of

hybridization can change between generations (e.g. Edmands, 1999, Fenster and

Galloway, 2000). For example, outbreeding depression may occur in the first generation

hybrids (F1) from disruptions in local adaptation or heterozygote disadvantage (Goto et

al., 2011, Hufford et al., 2012). Alternatively, fitness declines may not occur until the

second (F2) or later generations from disruption of the original parental co-adapted gene

complexes by recombination (Huff et al., 2011). It is also possible to have both increased

fitness in the F1 generation due to heterosis or heterozygote advantage and decreased

fitness in the F2 generation from disruption of co-adapted gene complexes (Edmands et

al., 2005).

Figure 1.1 Different relative fitness of offspring in response to average genetic distance

between breeders of species A, B, and C which result in breeding system continuum (from

Allendorf and Luikart, 2007).

9

Genetic divergence is only roughly correlated with outbreeding depression, and the

relationship is not strong enough to guide management decision (Edmands, 2002). The

risk of outbreeding depression is complicated by many factors. These include

chromosome structure or number, the level of genetic divergence, geographical distance,

the number of generations since isolation, population size, breeding system, the adaptive

differentiation among populations, and the environment (Hathaway et al., 2009, Hufford

et al., 2012, Edmands, 2007, Hendry et al., 2007, Frankham et al., 2011). Allendorf et al.

(2001) and Edmands (2007) have recommended that augmenting gene flow between

fragmented populations should only be carried out if the populations have lost substantial

genetic variation and effects of inbreeding depression are apparent. An opposing view

was promoted by Frankham et al. (2011) who argued that inbred populations may have

undocumented inbreeding depression, and while these populations are waiting for data

on the effects of inbreeding to be collected, this may put these populations at risk of

extinction. Weeks et al. (2011) agreed with this view suggesting that if the risk of

outbreeding depression is overestimated, it can prevent rational use of gene flow for

genetic rescue. Frankham et al. (2011), Weeks et al. (2011) and Hedrick and Fredrickson

(2010) have all developed guidelines to evaluate when genetic rescue is a good

management option. Frankham et al. (2011) proposed a framework for evaluating the

probability of outbreeding depression based on taxonomic status, fixed chromosome

differences, historical gene flow, environmental differences between populations and the

number of generations since separation (Figure 1.2). Weeks et al. (2011) suggested a

decision tree based on the purpose of translocation (inside or outside historical range),

genetic structure and genetic isolation among populations and the likelihood of

hybridization to other species/subspecies (if the translocation is outside historical range).

Hedrick and Fredrickson (2010) recommended outcrossing if the fragmented populations

show evidence of low fitness, but the donor population must be closely related to the

recipient population and experimental data from a captive population is required to

support validity of genetic rescue. All authors agreed that population monitoring is

required for several generations.

10

Figure 1.2 Decision tree for determining the probability of outbreeding depression

between two populations developed by Frankham and others (from Frankham et al.,

2011).

1.7 KNOWLEDGE GAPS IN TRANSLOCATION AND THESIS AIMS

As more and more species become threatened with extinction, translocation is a

conservation tool that has the potential to create insurance populations. However, since

species selected for translocations often persist as small and isolated populations or on

offshore islands, this limits choice of source populations suitable for translocation.

Although preserving the unique characteristics of each population is preferable to protect

ecological and genetic processes of a species (Wayne et al., 1994, Moritz, 1999, Huff et

al., 2010), keeping populations separated can prevent crossing between historically

outbred populations which further increases levels of inbreeding and the likelihood of

extinction (Bijlsma et al., 2000). Genetic mixing has potential benefits in increasing

genetic diversity and reducing inbreeding depression, but it could induce undesired

effects on outbred individuals as a result of outbreeding depression (Allendorf et al.,

2001). This uncertainty over the outcomes of genetic mixing discourages conservation

managers from attempting to mix source populations for threatened species

translocations.

11

In Australia, there are only a limited number of case studies from intentionally

outcrossing between mammal populations for conservation purposes (e.g. Mansergh et

al., 2013, Weeks et al., 2015). So far, a meta-analysis of intentional outcrossing of inbred

populations of vertebrates, invertebrate and plants with a low outbreeding depression risk

(evaluated using Frankham et al. 2011 decision tree, Figure 1.2) has shown to increase

composite fitness (combined fecundity and survival) (Frankham, 2015). This suggests

that outcrossing has the potential to become a management option if it is carried out under

appropriate circumstances (Frankham et al., 2011). However, due to the limited number

of case studies, the consequences of intentional outbreeding are unclear, especially with

respect to mammal translocations in Australia. Moreover, there is a need for more long-

term studies, which are rarely carried out, often due to time and financial constraints

(Schwartz et al., 2007). Without a long-term monitoring, the effects of mixing may go

undetected as outbreeding depression may not become apparent until F2 or later

generations (Huff et al., 2011, Edmands et al., 2005). Besides monitoring for genetic and

phenotypic consequences of mixing source populations, factors influencing the rate and

direction of admixture between source populations also need to be investigated.

Mortality, biased reproductive success and the type of release strategy can result in loss

or maintenance of lineages that shape the genetic composition of a translocated

population (e.g. Biebach and Keller, 2012) and ultimately determine a long-term

persistence of the translocated population. Another factor that affects the resilience and

long-term persistence of a new population is how well the founders represent evolutionary

potential of their source populations. In translocation, it is as important to understand the

spatial genetic structure of the source populations, so that appropriate sampling strategies

can be carried out to maximize genetic diversity (Miller et al., 2010b).

This project sets out to investigate questions about genetic mixing, genetic

representativeness and persistence of genetic variation within translocated populations

using case studies of marsupial translocations established from multiple source

populations with different levels of genetic divergence between them. My research takes

advantage of long-term genetic and demographic sampling from three major translocation

programs conducted by the Western Australian Department of Parks and Wildlife

(DPaW), in collaboration with Perth Zoo, on the dibbler (Parantechinus apicalis) and

burrowing bettong (Bettongia lesueur). The Dibbler translocations provide two long-term

(12 – 13 years) case studies with low and medium risk of outbreeding depression. With

complete pedigrees from the captive breeding colonies and regular post-translocation

12

monitoring, this species provides an excellent opportunity to study genetic mixing

outcomes in the short- and long-term, as well as admixture dynamics. In addition, I

highlight the importance of understanding spatial genetic structure in source populations

using a case study of mainland dibblers and demonstrate how it can be incorporated into

translocation planning and practice to achieve improved genetic representativeness. The

translocation of two island populations of burrowing bettongs to the mainland provides a

rare case study with a high risk outbreeding depression and an opportunity to further

investigate admixture dynamics. In each case study, extensive information of the

populations such as sample availability, demographic information, and samples from

follow-up monitoring were readily available.

13

1.8 STUDY SPECIES

1.8.1 The dibbler (Parantechinus apicalis)

Figure 1.3 The Dibbler (Parantechinus apicalis) in a native species breeding program at

Perth Zoo (from Perth Zoo, 2013).

Parantechinus apicalis is commonly known as the dibbler. The dibbler is a marsupial

with a generalist and opportunistic diet, consisting mainly of insects (Miller et al., 2003).

They are semi-arboreal and crepuscular (i.e. they are most active at dawn and dusk). This

small size dasyurid (40 – 125 g) is readily distinguished by the white rings around the

eyes, a tapering hairy tail, and the flecked appearance of its coarse fur (Woolley, 1995,

Woolley, 2008). The fur colour is brownish grey changing to greyish-white tinged with

yellow on the ventral surface (Figure 1.3).

Dibblers used to be widely distributed across a large proportion of the coastal region of

southwest Western Australia and some part of South Australia (Moro, 2003). Their

current distribution is in a small part of southwest of Western Australia and on two islands

in Jurien Bay, Boullanger and Whitlock Islands (Figure 1.4). Although they have been

recorded over a diverse range of habitats, they seem to prefer vegetation with a dense

canopy > 1 metre high that has been unburnt for at least 10 years (Start and Baczocha,

1997).

The mating system of dibblers is polygynandrous, with both males and females pairing

with several mates (Lambert and Mills, 2006). There is strong sexual dimorphism, with

males being larger than females. Dibblers breed once a year during autumn (February to

April) (Mills et al., 2012). A female can produce as many as eight young per breeding

14

season. These young reach sexual maturity after 10 – 11 months (Woolley, 1995). On the

islands, males die after the first mating season in some years but not in all years (Mills

and Bencini, 2000). This happens more frequently on Boullanger Island than Whitlock

Island (Mills and Bencini, 2000). This phenomenon has been termed facultative male die-

off which is thought to be associated with nutrient inputs from seabirds (Wolfe et al.,

2004). On the mainland, males survive well into their second year (Friend and Collins,

2005). The average lifespan is estimated to be 1.5 years (Woinarski et al., 2014), though

longevity appears to vary between island and mainland populations (Lambert and Mills,

2006, Wolfe et al., 2004, Friend and Collins, 2005). In captivity, island dibblers can live

up to 3.5 years (Lambert pers. comm.), but most don’t live pass two years and no

facultative male die-off has been observed in captivity (Lambert and Mills, 2006, Wolfe

et al., 2004). Longevity of mainland dibblers in captivity is up to 5.5 years (Lambert pers.

comm.) but in the wild, males and females live up to three and four years respectively

(Friend and Collins, 2005).

15

Figure 1.4 Past and present distribution of the Dibbler (Parantechinus apicalis) adapted

from Moro (2003). Past distribution is shown in the smaller map above. Present

distribution is shown in the larger map. Normal font represents natural populations and

italic font represents translocated populations.

Mainland and island populations may represent taxonomically distinct populations as the

islands have been separated from the mainland for at least 500 years (Chalmers and

Davies, 1984) and a genetic study by Mills et al. (2004) revealed significant genetic

differentiation between the mainland and island populations. Genetic diversity and body

size of mainland dibblers are also much higher and larger than island dibblers (Mills et

al., 2004). While there was no significant differences between the two island populations,

but dibblers on Boullanger Island have slightly higher levels of genetic diversity and are

larger than dibblers on Whitlock Island (Mills et al., 2004, Mills and Bencini, 2000).

The dibbler is listed as Endangered under the Environment Protection and Biodiversity

Conservation Act (1999) and IUCN Red List of Threatened Species (Friend et al., 2008).

It is threatened by introduced predators, loss of habitat due to fire and disease, human

disturbance, and resource competition by the house mouse (Mus musculus) (Friend,

2003). Several translocations have been carried out as a part of recovery actions (Table

Boullanger Island

Whitlock Island

Escape Island

Peniup Nature Reserve Fitzgerald River National Park

Perth

Jurien

16

1.1). However, only translocations to Peniup Nature Reserve and Escape Island have been

successfully established (> 5 years) with reported recruitment of second and later

generations, and approximated known-to-be-alive numbers similar to those in source

populations (Friend, 2003, Moro, 2003).

Table 1.1 Translocation records of captive bred dibblers at Perth Zoo. Only Escape Island

and Peniup National Reserve translocations have been successful.

Species Source

population

Release Site Year of

release

Total No.

Adults

Released

Dibblers

(Island)

Whitlock Island

Boullanger

Island

Escape Island

1998 26

1999 41

2000 19

2001 2

Total 88

Dibblers

(Mainland)

Fitzgerald River

National Park

Peniup

Nature

Reserve

2001 41

2002 46

2003 43

2006 6

2008 24

2009 34

2010 41

Total 235

Stirling

Range

National Park

2004 54

2005 62

2006 38

2007 40

Total 194

Waychinicup

National Park

2010 20

2011 74

2012 70

2013 12

Total 176

The first translocation to Peniup Nature Reserve took place from 2001 to 2003 (Table

1.1). Follow-up releases at this site occurred from 2006 to 2010. The dibblers used to

establish this translocated population (N = 223) were raised in captivity at Perth Zoo as a

part of the Dibbler recovery program. The captive population was established from twenty

six individuals collected from the Fitzgerald River National Park (FRNP) (Friend, 2003).

The Peniup translocation and FRNP population provide unique opportunities to

investigate the genetic outcomes of a translocation with a low risk of outbreeding

depression and, given availability of trapping locations, to examine fine-scale spatial

genetic structure in a ‘natural’ population (FRNP). So far there has been no genetic study

17

on the Peniup population despite regular trapping sessions and no such investigations

conducted on the fine-scale population structure of this species.

The Escape Island translocation was established from 1998 to 2000 using captive-bred

individuals (N = 83) raised at Perth Zoo. The captive population used in this translocation

was established from two breeding pairs from Boullanger Island, two breeding pairs from

Whitlock Island, and a later addition of three males from Whitlock Island (Friend, 2003,

Moro, 2003). Population monitoring on Escape Island was carried out from 1998 to 2001

by Moro (2003). He recorded at least 121 new born dibblers on Escape Island. Based on

the population dynamics, demographics, and establishment of the surviving progeny,

Moro’s (2003) results suggested the translocation was successful, at least in a short-term.

This was confirmed with follow-up monitoring up until 2012. Facultative male die-off

was not observed in the Escape Island population (Moro, 2003). A small scale genetic

study of the Escape Island population showed it had a higher level of observed

heterozygosity (0.232) than the Whitlock Island source population (0.114), but lower than

the level found in the other source population from Boullanger Island (0.382) (Wilcox,

2003). However, this study did not investigate temporal variation in the captive or

translocated populations or the extent of interbreeding between the source population

lineages (Mills et al., 2004).

18

1.8.2 The burrowing bettong (Bettongia lesueur)

Figure 1.5 The burrowing bettong (Bettongia lesueur) in the Operation Rangeland

Restoration translocation at Lorna Glen. Photo credits: Judy Dunlop.

Bettongia lesueur is most commonly known as the burrowing bettong or boodie. This

medium-sized marsupial is characterised by a short blunt head, with small rounded and

erect ears (Figure 1.5). They are yellowy grey with a light grey underside. The legs, feet,

and tail are more yellow in colour. Their fat tails are lightly haired and some have a

distinctive white tip (Burbidge and Short, 1995). They are omnivorous, nocturnal, and the

only macropod that shelters in burrows on a regular basis (Burbidge and Short, 1995).

Burrowing bettongs are social animals, forming strong social groups of one male and one

female or one male to many females centred on a cluster of warrens (Short and Turner,

2000). There is no significant dimorphism between sexes (Short and Turner, 1999). They

appear to have a polygynous mating system (Sander et al., 1997). Breeding can occur

throughout the year, but is generally broken by a period of anoestrous (Short and Turner,

1999). A female normally gives birth to one young per breeding season and up to three

young may be produced per year (Tyndale Biscoe, 1968). These young reach sexual

maturity after 7 – 8 months of age (Short and Turner, 1999). Burrowing bettongs normally

live to at least 3 years of age (Short and Turner, 1999).

Prior to European settlement their distribution covered the middle and western half of

Australia (Short and Turner, 1993), but they disappeared from mainland Australia around

the early 1960s (Finlayson, 1961, Burbidge et al., 1988). At present, the only known

natural populations are found on Bernier, Dorre, and Barrow Islands off the coast of

Western Australia (Figure 1.6). The physical isolation of the islands from the mainland is

estimated to have occurred at least 8,000 years ago (Dortch and Morse, 1984). The

19

population size on Bernier Island is estimated in 2010 to be 842 individuals, while the

Dorre Island population has approximately 3291 individuals and the Barrow Island

population has around 3,000 individuals (Richards, 2012). Burrowing bettongs found on

Bernier and Dorre Islands (Bettongia lesueur lesueur) are suspected to be a different

subspecies to bettongs on Barrow Island because due to morphological differences and

the geographical isolation of this population (Donaldson, unpublished data). The

morphological differences include body size, numbers of nipples, and the timing of the

peak breeding season. In term of body size, burrowing bettongs on Barrow Island are

smaller than bettongs on Bernier and Dorre Islands with average weights of 750 g and

1,300 g respectively (Short and Turner, 1999, Richards, 2012). Most females on Barrow

Island have four nipples, while those on Bernier and Dorre Islands have only two

(Richards, 2007). The breeding season on Bernier and Dorre Islands peaks over the austral

winter, whereas breeding on Barrow Island peaks over the austral summer (Richards,

2012). Nevertheless, both breeding cycles seem to be coincided with the first major

rainfall (Richards, 2012). No taxonomic studies have yet confirmed that these populations

are different subspecies.

Figure 1.6 Past and present distribution of the Burrowing bettong (Bettongia lesueur)

adapted from Richards (2012). Red dots represent translocated populations.

20

The burrowing bettong is listed as near threatened under IUCN Red List of Threatened

Species (Richards et al., 2008b) and as vulnerable under the Environmental Protection

and Biodiversity Conservation Act (1999). The primary threats to this species are exotic

predators, introduced herbivores, fire and climate change (Richards, 2007, Richards,

2012). Several reintroductions have been conducted, mostly around nearby islands, and

some to the mainland, using founders from a single subspecies (Richards, 2007).

Burrowing bettongs from the Dryandra Field Breeding Facility (N = 109, originally

stocked from Dorre Island) and Barrow Island (N = 67) where reintroduced to Lorna Glen

in 2010 as part of Operation Rangeland Restoration conducted by DPaW. The Operation

Rangeland Restoration project aims to restore natural ecosystem function and biodiversity

by reintroducing 11 arid zone mammal species, including burrowing bettongs, by 2020

(DEC, no date). So far no empirical study has been carried out on the burrowing bettong

population at Lorna Glen. Furthermore, no translocation involving burrowing bettongs

has previously attempted to mix the different subspecies. This translocated population,

therefore, provides an opportunity to investigate the process of genetic mixing and the

associated fitness consequences when there is a high level of divergence between the

source populations.

21

1.9 STUDY AIMS AND THESIS STRUCTURE

The aim of this project is to advance the current understanding of genetic mixing between

source populations with different levels of genetic divergence, using cases studies of two

mammals - dibblers and burrowing bettongs translocated within Western Australia. The

questions underpinning this thesis are:

How can we use knowledge of fine-scale spatial population structure to assist in

the selection of individuals for translocation? In particular how can we use this

information to minimize inbreeding and improve genetic representativeness?

How effective are translocations established using more than one source

population in maintaining genetic variation and do the outcomes vary according

to the levels of divergence between them?

What are the phenotypic consequences of genetic mixing between source

populations with differences in body size and breeding seasons?

In mixed translocations, what are the factors affecting genetic contribution from

each source population and are changes in the genetic contributions from source

populations common?

The thesis consists of six chapters as follows:

Chapter one: Project overview

This chapter is an introduction to the project and presents a literature review. It contains

the definition of translocation, genetic issues associated with translocation and

advantages and disadvantages of genetic mixing. It also identifies the gaps in knowledge

that need to be addressed. Lastly, this chapter provides background descriptions of the

two study species and describes the project aims and overall thesis structure.

Chapter two: A fine-scale population structure of mainland dibblers in the

Fitzgerald River National Park

In most species, populations are often subdivided into subpopulations, which are

interconnected by migration and gene flow (Allendorf and Luikart, 2007). Gene flow is

an important mechanism for maintaining genetic variation within populations and reduces

genetic differentiation among subpopulations. Landscape features such as rivers,

mountains or human-mediated habitat alternations such as farms or roads can become

barriers to gene flow, leading to genetic heterogeneity (Holderegger and Wagner, 2008).

22

Even in the absence of landscape barriers, gene flow can be limited by dispersal ability,

which can give rise to isolation-by-distance, where genetic differentiation between

populations increases with geographical separation (Wright, 1943). In small mammals,

males and females may also exhibit different fine-scale genetic structures as a result of

different dispersal strategies e.g. philopatry (e.g. Banks and Peakall, 2012). In chapter

two, I assess fine-scale genetic structure of the mainland dibbler population in the FRNP.

I also look for evidence of dispersal differences between different sexes and different age

classes. I then evaluate options for more effective sampling of individuals from this

population for future captive breeding and translocation programs. Chapter two will be

published as:


Mills. In prep. Fine-scale population genetic structure of the mainland dibbler

(Parantechinus apicalis). PLoS ONE.

Chapter three: Genetic outcomes of the dibbler mainland translocation to Peniup

Nature Reserve

Chapter three examines genetic outcomes of a dibbler translocation involving individuals

selected from different subpopulations identified in the previous chapter. Interbreeding

individuals from genetically divergent populations has been shown to increase genetic

diversity and reduce inbreeding levels in recipient or translocated populations (e.g.

Kennington et al., 2012, Miller et al., 2012, Hedrick and Fredrickson, 2010). However, it

is still unknown whether genetic mixing between source populations with a low level of

genetic divergence between them will have any genetic benefits. In chapter three, I

compare levels of genetic diversity and genetic relatedness of the captive (Perth Zoo) and

translocated populations (Peniup NR) to the source subpopulations and assess whether

they changed over the first 10 years since the translocation. Chapter three will be

published as:


Mills. In prep. Temporal variation in the genetic composition of a newly

established population of dibbler (Parantechinus apicalis) reflects

translocation history. Conservation Genetics.

23

Chapter four: Genetic consequences of admixture between genetically diverged

populations in a dibbler translocation to Escape Island

The probability of outbreeding depression increases when source populations have been

separated longer than 500 years (Frankham et al., 2011). In different environments, this

amount of time may be enough to allow adaptive differentiation to be developed. In

similar environments, it would require a minimum of thousands of generations to evolve

reproductive isolation mechanisms (Coyne and Orr, 2004). A translocation of dibblers to

Escape Island was established using dibblers originally sourced from Boullanger and

Whitlock Islands. These islands have separated from the mainland for at least 500 years.

Dibblers on these islands exhibit differences in body size and life-history (i.e. facultative

male die-off). Since the islands are exposed to a similar environment, due to their

proximate location, reproductive isolation mechanisms are less likely to develop.

Therefore, this chapter predicts a medium probability of outbreeding depression. To

investigate this possibility, this chapter examines the extent of genetic mixing between

the source populations in both captive-born and wild-born dibblers. Then, I measure and

compare levels of genetic variation between captive-born and wild-born to dibblers from

the source populations. Chapter four will be published as:

Thavornkanlapachai, R., H.R. Mills, K. Ottewell, C. Lambert, T. Friend, and W.J.

Kennington. In prep. Admixture between genetically diverged island populations

bolsters genetic diversity within a newly established island population of the dibbler

(Parantechinus apicalis), but does not prevent subsequent loss of genetic variation.

Biological Conservation.

Chapter five: Genetic consequences of admixture between genetically diverged

populations of burrowing bettongs translocated to Lorna Glen

Although genetic incompatibilities are not expected to develop between populations

separated for thousands of generations, populations in different environments can show

the first signs of outbreeding depression within a few dozens of generations (Coyne and

Orr, 2004). The burrowing bettong population established at Lorna Glen was established

using individuals originally sourced from two island populations, Barrow Island and

Dorre Island. These islands are 531 kilometres apart and have been separated from the

mainland for at least 8,000 years (Dortch and Morse, 1984). Burrowing bettongs on these

islands differ in body size and in the timing of their peak breeding season. When the

population was established, it was uncertain whether individuals from the different source

24

populations would interbreed, and if they did, whether there would be outbreeding

depression. This chapter focuses on the consequences of genetic mixing between the

highly diverged source populations by assessing changes in population structure using

microsatellites and mitochondrial DNA gene sequencing. I assess how genetic

introgression affects both genetic and morphological characteristics of the offspring.

Lastly, I compare overall genetic diversity between the translocated and source

populations to see how well genetic diversity is maintained within the first few

generations after the translocation began. Chapter five has been submitted to Molecular

Ecology as:

Thavornkanlapachai, R., H.R. Mills, K. Ottewell, J. Dunlop, C. Sims, K.

Morris, F. Donaldson, and W.J. Kennington. Submitted. Asymmetrical

introgression between genetically distinct populations of the boodie

(Bettongia lesueur) in a newly established translocated population. Molecular

Ecology.

Chapter six: Project conclusion and application

This chapter identifies common major findings shared between chapter two to five and

their significance in the current knowledge of translocation relative to other literatures.

Possible conservation strategies are formulated to improve translocation planning and

practice. Research limitations and potential future research directions are also discussed.

26

Photo credit: Perth Zoo

27

CHAPTER TWO

Fine-scale population genetic structure of the mainland

dibbler, Parantechinus apicalis

2.1 ABSTRACT

Dispersal plays an important role in the population structure and resilience of species. To

gain a better understanding of dispersal in the endangered Australian marsupial, the

dibbler (Parantechinus apicalis), we screened 199 individuals from seven locations

within the Fitzgerald River National Park for genetic variation at 17 microsatellite DNA

loci. There were high levels of genetic variation within all sites (gene diversity ranged

from 0.68 to 0.71) as well as significant genetic differentiation between sites less than 19

km apart that were consistent over multiple years (FST = 0.021 – 0.073). A Bayesian

clustering analysis revealed the presence of two genetic clusters separating P. apicalis in

the western side from the central-eastern side of the National Park. There was also

evidence of fine-scale population structure with a positive correlation between genetic

structure and distances up to 200 m in females. By contrast males did not exhibit

significant fine-scale population structure, thus suggesting P. apicalis exhibits female

philopatry and male-biased dispersal. We recommended that management should take

into account the existence of two subpopulations within the National Park and manage

accordingly. Individuals selected for captive breeding and translocation programs,

especially females, should be sampled at least 200 m apart to reduce the likelihood of

selecting related individuals.

28

2.2 INTRODUCTION

Dispersal strongly influences the dynamics and persistence of populations and determines

the level of genetic differentiation between populations (Dieckmann et al., 1999).

Nevertheless, relatively little is known about the spatial or temporal extent of dispersal in

most species. This largely reflects the difficultly in carrying out intensive, large scale, and

long-term demographic studies to track individuals and monitor dispersal outcomes

(Koenig et al., 1996). The application of molecular genetic techniques provides an

alternate approach for investigating dispersal across the landscape (e.g. Banks and

Peakall, 2012, Peakall et al., 2003). Information on dispersal is particularly important for

populations of threatened species occupying fragmented landscapes, which are highly

vulnerable to loss of genetic variation and inbreeding.

Numerous studies have demonstrated how landscape features such as rivers and land

clearing can limit dispersal patterns and subsequent gene flow between populations

(Banks et al., 2005, Potter et al., 2012, Levy et al., 2013). For example, black vole

(Clethrionomys glareolus) populations separated by approximately 100 m of open water

were documented to share only a small proportion of rare haplotypes, with most

haplotypes restricted to a single bank (Aars et al., 1998). Land clearing for agriculture

had shown to reduce gene flow between outcrops of the granite outcrop-dwelling lizard

(Ctenophorus ornatus) and subsequently enhanced genetic structuring, increased genetic

differentiation between outcrops and reduced genetic diversity relative to individuals

found in an adjacent nature reserve (Levy et al., 2010). Even in the absence of landscape

barriers, gene flow may be limited by the dispersal ability of the species. Under isolation-

by-distance (Wright, 1943), pairs of populations closer to each other will be more

genetically similar due to a higher likelihood of genetic flow than more distant

populations (e.g. Forbes and Hogg, 1999, Neaves et al., 2009).

Other than physical barriers, dispersal is also influenced by behaviour (Croteau et al.,

2010, Double et al., 2005, Lampert et al., 2003, Hazlitt et al., 2006). For example,

different dispersal strategies can lead to different patterns of dispersal in males and

females. These dispersal strategies evolve in response to resource availability, kin

competition, and inbreeding avoidance (Lawson Handley and Perrin, 2007). Greenwood

(1980) hypothesized that direction of sex-biased dispersal is tightly linked to the mating

system, and its intensity is driven by the complexity of social system (Lawson Handley

and Perrin, 2007). In mammals, male-biased dispersal and female philopatry are common

29

in species that have polygynous and promiscuous mating systems (Greenwood, 1980).

That is, females remain close to their natal habitats throughout their lives while males

disperse (Cockburn et al., 1985, Fisher, 2005). For example, juvenile males of Antechinus

spp. disperse shortly after they are weaned while females stay in their natal habitats

(Fisher, 2005, Cockburn et al., 1985). This generates a strong positive spatial genetic

autocorrelation over a short geographical distance in the resident females relative to the

dispersing males (Banks and Peakall, 2012).

The dibbler (Parantechinus apicalis) is a small (approximately 40 – 125 g) dasyurid

marsupial, endemic to the southwest of Australia (Miller et al., 2003, Mills et al., 2004,

Mills and Bencini, 2000, Woolley, 2008). Its current distribution is restricted to the

Fitzgerald River National Park (FRNP, ~ 3000 km2), 180 km north-east of Albany, and

on two small islands, Boullanger and Whitlock Islands, off the coast from Jurien Bay with

translocated populations at Peniup nature reserve and Escape Island (Morcombe, 1967,

Fuller and Burbidge, 1987). It is listed as Endangered under the Environment Protection

and Biodiversity Conservation Act (1999) and 2014 IUCN Red List of Threatened

Species (Friend et al., 2008). The main threats to this species include introduced predators

such as foxes (Vulpes vulpes) and feral cats (Felis catus), inappropriate fire regimes,

habitat degradation by dieback (Phytophathora cinnamomi), and competition with house

mice (Mus musculus) on islands (Friend, 2003).

Parantechinus apicalis have a polygynandrous mating system, in which both males and

females pair with several mates (Lambert and Mills, 2006). In FRNP, females produce

up to eight young per breeding season from mid-April to May (Mills et al., 2004). In

captivity, 94.5% of these young survive to weaning (unpublished data). These young

become independent in September, three to four months after birth, and reach sexual

maturity when they are ten to 11 months old (Woolley, 1995, Woolley, 2008). In FRNP,

female P. apicalis can live up to four years, while males live up to three years (Friend and

Collins, 2005). In this population, dibblers occupy distinct, but overlapping home ranges

and males appear to occupy larger home ranges than females (Friend, 2003). However,

the actual size of their home range and dispersal distance are still unknown. Since island

populations are often viewed as poor source populations for captive breeding and

translocation (but see Abbott, 2000, Moro, 2003, Cardoso et al., 2009), the mainland

population has become the most heavily exploited for translocations involving P. apicalis

to mainland sites (Friend, 2003). The aim of this study is to investigate the population

genetic structure across the Fitzgerald River National Park to gain a better understanding

30

of the dispersal distances in P. apicalis and the consequences for population genetic

structure. We also look for evidence of differences in dispersal between sexes to evaluate

the role of sex-biased dispersal in this species with a polygamous mating system.

2.3 MATERIALS AND METHODS

2.3.1 Study sites and sample collection

The Fitzgerald River National Park (FRNP) is dominated by open to very open mallee

and shrubland. Heath is common throughout, with woodlands occurring along the rivers

and in swamps (Moore et al., 1991). Although P. apicalis have been recorded in a diverse

range of habitat types within the FRNP, they seem to prefer vegetation with a dense

canopy more than one metre high in long-unburnt heathland (Friend, 2003). Abundance

estimates of mainland populations are currently not available, but the total number of

mature individuals is estimated to be less than 1000 (Woinarski et al., 2014).

Samples were collected from seven sites up to 76 km apart within the FRNP (Table 2.1;

Figure 2.1). Trapping is based either on grids of limited size (450m x 600m) or 5 km

linear transects. Recapture of individuals is therefore limited to dispersal movements

within these distances. We only used post-dispersal individuals (as defined below) in

population genetic estimates because use of pre-dispersal individuals can result in an

overestimate of a fine-scale spatial structure (e.g. Peakall et al., 2003). Post-dispersal

individuals were classified using trapping time, signs of sexual maturity and body weight.

First, all adults were classified as post-dispersal individuals. These were the individuals

trapped from February to August because it was the period when only adults were present

in the population (i.e. mating season to the time before pouch young became

independent). Outside this period, there was a mixture of individuals from different age

classes, so adults were classified as individuals that had a fully developed pouch, nipples,

or fully developed testes. Post-dispersal sub-adults were classified as individuals that

lacked signs of sexual maturity and had body weight greater than 50 g in females and 58

g in males. These weight limits were calculated from the average weight minus one

standard deviation of males (75.5 ± 16.8 g) and females (62.1 ± 11.4 g). Pre-dispersal

sub-adults were individuals that lacked signs of sexual maturity and were lighter than the

body weight limit. A total of 199 post-dispersal individuals (98 females and 101 males)

and 27 pre-dispersal individuals (16 females and 11 males) were trapped using cage and

Elliott traps between 2000 and 2013. All of the trapped animals had a small piece of ear

31

tissue (~ 1 mm2) taken, a microchip implanted, and had their trapping location, sex and

age class recorded. The collected ear tissues were stored in 20% DMSO2 saturated with

NaCl at room temperature.

Figure 2.1 Map showing the sampling sites: Twertup (TW), Wilderness gate (WG),

Quoin Head (QH), Hamersley Moir (HAM), Gravel pit (GP), Moir track (MT), East Mt

Barren (EMB) in the Fitzgerald River National Park (shaded area).

2.3.2 DNA extraction and microsatellite genotyping

DNA was extracted using the ‘salting-out’ method (Sunnucks and Hales, 1996) with a

modification of a 56 °C incubation and the addition of 10 mg/mL Proteinase K to 300 µL

TNES. After the DNA was extracted, each animal was genotyped using the following 21

microsatellite loci: pPa2D4, pPa2A12, pPa2B10, pPa7A1, pPa7H9, pPa9D2, pPa1B10,

pPa4B3, pPa8F10 from a previous study (P. apicalis, Mills and Spencer, 2003) ;

pDG1A1, pDG1H3, pDG6D5 (Dasyurus geoffroii, Spencer et al., 2007) ; 3.1.2, 3.3.1,

3.3.2, 4.4.2, 4.4.10 (Dasyurus spp., Firestone, 1999) ; Sh3o, Sh6e (Sarcophilus laniarius,

Jones et al., 2003) ; Aa4A (Antechinus agilis, Banks et al., 2005) and Aa4J (A. agilis,

TW

WG

EMBQH

HAM GP

MT

32

Kraaijeveld-Smit et al., 2002b). Primer concentrations were from 0.04 to 1.5 µM and 10

– 20 ng of DNA were added to PCR reactions (volume 10 µL) using a QIAGEN Multiplex

PCR Kit (Table S2.1, Appendices). Amplifications were performed in an Eppendorf

Mastercycler epgradientS Thermocycler using the following parameters: 15 min at 95 °C,

a total of 35 or 40 cycles of 30 s at 94 °C, 90 s at annealing temperatures ranging from 46

°C to 58 °C and 60 s at 72 °C, and concluding with 30 min at 60 °C (Table S2.1). PCR

products were analysed with a ABI 3730 sequencer using a GeneScan-600 LIZ internal

size standard and scored using GENEMARKER version 1.90 (SoftGenetics).

2.3.3 Data analysis

Prior to any analysis of the microsatellite data, we assessed genotyping quality by

calculating the allele-specific and locus-specific genotypic error rates (Pompanon et al.,

2005). We tested for the presence of null alleles in the source populations at each locus

using MICROCHECKER (Van Oosterhout et al., 2004). LOSITAN was used to test for

natural selection acting on loci based on the FST outlier approach (Beaumont and Nichols,

1996, Antao et al., 2008). The outlier FST values were identified by plotting FST against

heterozygosity to generate a null distribution using an island model of migration. Loci

with extremely high or low FST values would indicate directional or balancing selection

respectively. For this analysis, individuals were grouped according to the year and site of

collection. Each group was analysed separately using both stepwise and infinite allele

mutation models with 100,000 simulations, 99% confidence interval and the neutral and

forced mean options. For these analyses, and elsewhere, we excluded any sites or

collection years that had a sample size less than ten.

Estimates of genetic variability such as allelic richness (an estimate of allele number per

locus corrected for sample size) and gene diversity were calculated using FSTAT version

2.9.3.2 (Goudet, 2001). The inbreeding coefficient (FIS) was calculated to assess

deviations from Hardy-Weinberg Equilibrium with the significance of the deviations

determined using randomization tests. Genotypic disequilibrium between each pair of loci

within each population sample was assessed by testing the significance of association

between genotypes. For these tests, a sequential Bonferroni correction (Rice, 1989) was

applied to control for type I statistical error. Genetic differentiation among sites and

collection years was assessed by calculating Weir & Cockerham’s (1984) estimator of

FST (θ) using FSTAT version 2.9.3.2 (Goudet, 2001) and Shannon’s mutual information

index using GENALEX version 6.5 (Peakall and Smouse, 2012). For Shannon’s mutual

33

information index, we used the Shannon Partition option (1000 permutations) to test for

significant population differentiations. Friedman’s ANOVA and Wilcoxon’s signed-rank

tests were used to test for differences in genetic variation between population samples

using the R (version 3.0.1) statistical package (R Core Team, 2014).

We estimated effective population sizes for each collection year and site using the single-

sample estimator of Ne as implemented in the software package LDNE (Waples and Do,

2008). We used a random mating model and estimated linkage disequilibrium amongst

alleles using only alleles with frequencies > 5% to avoid bias from rare alleles (Waples

and Do, 2010). The 95% confidence interval for Ne was calculated by jackknifing

disequilibrium values among pairs of loci (Waples and Do, 2008).

The occurrence of recent population bottlenecks was investigated by testing for an excess

in heterozygosity, under both stepwise mutation models (SMM) and two-phase model

(TPM). Variance for TPM was set to 12 and the proportion of SMM in TPM was 95%

with 1000 iterations using the program BOTTLENECK (Piry et al., 1999). If a population

has experienced a recent bottleneck, the program will generate allele frequency

distributions that are under-represented by (rare) alleles at low frequency (< 10%) and

over-represented by intermediate frequency (> 10%) classes (Luikart et al., 1998).

Population structure was analysed using two Bayesian clustering methods, implemented

using STRUCTURE 2.3.4 (Pritchard et al., 2000) and GENELAND (Guillot et al., 2005).

Both methods determine the most likely number of genetic clusters (K) by maximising

the within-cluster Hardy-Weinberg and linkage equilibria. They also assign each

individual to a genetic cluster according to its membership coefficient, which is the

fraction of the genome associated with a particular genetic cluster. GENELAND differs

from STRUCTURE in that geographical information can be incorporated to assess the

population structure based on the spatial distribution of individuals. The STRUCTURE

analysis was performed using an admixture ancestry model with the collecting site

locations set as prior information, correlated allele frequencies, a burn-in length of 50,000

iterations, a run length of 500,000 Markov chain Monte Carlo (MCMC) repetitions, and

carried out 10 iterations for each value of K (1 – 10). To estimate K, we compared the

likelihood values for each K and used the ∆K method of Evanno et al. (2005b) to choose

K. We re-ran the analysis for each identified genetic cluster separately to confirm that

there was no additional population structure within each cluster. The STRUCTURE

estimated cluster membership coefficient over multiple runs were permuted using the

34

Greedy option to obtain a mean across replicates in CLUMPP (Jakobsson and Rosenberg,

2007). The output from CLUMPP was depicted in Distruct1.1 (Rosenberg, 2004).

For GENELAND analysis, we included spatial coordinates of each individual to run the

spatial model. The parameters we used were zero coordinate uncertainty, uncorrelated

and null allele models, 500,000 MCMC iterations with a thinning of 500 and 200 burn-

ins. Thinning of 500 means only each 500th iteration will be saved on the disk (5000

iterations will be saved in total). A burn-in of 200 would discard the first 200 saved

iterations. Values of K from one to ten were tested and ten independent runs were

performed for each K. The most likely number of clusters was determined from the modal

K from each independent run with the highest posterior probability. To assess the extent

of population structuring, an analysis of molecular variance (AMOVA) was carried out

in ARLEQUIN version 3.0 (Excoffier et al., 2005). The locus by locus variance

component was estimated from 16,000 permutations. We analysed the data by grouping

collection years with a sample size greater than ten and had them nested within their

sampling sites.

Spatial population structure was also assessed by using spatial autocorrelation (SA)

analysis to evaluate genetic similarity between individuals over varying distance classes.

The analysis were conducted using GENALEX version 6.5 (Peakall and Smouse, 2012)

with the results presented by plotting the spatial autocorrelation coefficient (r) as a

function of distance class to produce a spatial autocorrelogram. Two spatial scales were

investigated. Firstly, we looked at broad-scale patterns by using seven distance classes

covering from 0 to 19 km (the maximum distance between trapping sites within a single

genetic cluster as determined in the STRUCTURE analysis). Separate SA analyses were

conducted for each genetic cluster and each sex. Secondly, we looked at fine-scale

patterns by using six 100 m distance class intervals (0 to 600 m). For the fine-scale

analyses, we excluded individuals from Twertup because trap lines at this site were set

one km apart. We analysed both with and without pre-dispersal individuals to investigate

the effects of these individuals on the fine-scale patterns. For both broad- and fine-scale

patterns, we combined distance classes if the number of pairwise comparisons (n) of each

distance class was less than 100. If both sexes were compared, the distance classes were

combined to accommodate the sex with the minimum n. Significance testing for spatial

autocorrelation (H0: r = 0, H1: r < > 0) was carried out by permutation testing and

bootstrapping calculating the 95% CI around r. We also calculated a correlogram wide

‘Omega’ (ω) to test for heterogeneity between groups. ω is computed from the t-value

35

obtained from t-tests comparing r values between groups at each distance class. If

observed omega is larger than expected, the null hypothesis of homogeneous correlogram

between groups is rejected. It is deemed significant when the P-value is less than 0.01 as

recommended by Banks and Peakall (2012). We also calculated pairwise relatedness,

which estimates the fraction of alleles that is Identical By Descent (IBD) among pairs of

individuals. GENALEX was used to estimate the mean pairwise relatedness within each

site relative to the total using the Pops Mean option with 999 permutations and the 95%

confidence interval around pairwise relatedness estimated using 1000 bootstraps.

Pairwise genetic relatedness for female-female, male-male, and female-male pairs within

the same site, between sites within the same genetic cluster, and between genetic clusters

were estimated using the method of Lynch and Ritland (1999). Wilcoxon’s signed-rank

tests were used to detect any significant differences in pairwise genetic relatedness

between sites and sexes using the R version 3.0.1 statistical package (R Core Team,

2014).

Finally, the relationship between genetic distance and geographical distance was assessed

using Mantel tests (1000 permutations) to test for an ‘isolation-by-distance’ effect,

performed using GENALEX (Peakall and Smouse, 2012).

36

2.4 RESULTS

2.4.1 Genetic variation within sites

Across all samples, there was an amplification rate of 0.901 per locus. The allele-specific

and locus-specific genotypic error rates were 0.008 and 0.014 respectively. Out of the 21

loci genotyped, MICROCHECKER suggested null alleles were present at locus pPa1B10

in most collection years at the Hamersley Moir and Twertup sites, so it was removed from

further analysis. Both Sh3o and pDG6D5 were almost monomorphic at all sites with

frequencies of the most common allele >95%. Locus pPa7Al showed a signal of

directional selection with an expected heterozygosity almost one in a pooled estimate.

This suggested that this locus may be linked to a genic region that might be subjected to

selection for heterozygosity. However, there were no clear patterns of allele distribution

and frequency across sampled populations. These loci were removed from further

analysis. The remaining loci showed high levels of genetic variation in all sampling sites

(Table 2.1). There were no significant differences in allelic richness or gene diversity

among sites or among collection years, except for one pairwise comparison between a

2005 sample from Twertup and a 2006 sample from Hamersley Moir (Wilcoxon rank

sum test, P = 0.030). A significant positive FIS value was identified in a 2005 sample from

Hamersley Moir as well as the pooled samples from this site (Table 2.1). However, FIS

values of most sites were not significantly different from each other except for pooled

Hamersley Moir and Quoin Head samples (Wilcoxon rank sum tests, P = 0.019).

Genotypic disequilibrium was detected in only one pair of loci across all sites.

37

Table 2.1 Sample size and estimates of genetic variation of post-dispersal Parantechinus

apicalis within sites for each collection year. FIS is the inbreeding coefficient. AR is allelic

richness. H is gene diversity. Standard errors are given after mean values. Asterisks show

FIS values significantly different to zero (P < 0.05).

Site/year Sample size

FIS AR H Bottleneck

Female Male Total

East Mt Barren

2005 4 0 4 2007 2 0 2 2008 1 0 1

Moir track

2003 0 3 3

2004 3 0 3

2005 3 2 5

Pooled 6 5 11 0.04 3.9±0.2 0.71±0.02 N

Gravel pit

2007 3 2 5

Hamersley Moir

2005 23 16 39 0.08* 3.8±0.2 0.68±0.02 N

2006 2 8 10 -0.01 4.1±0.2 0.71±0.03 Y

2007 9 9 18 0.01 3.9±0.2 0.70±0.03 N

2008 10 8 18 0.03 3.8±0.2 0.68±0.03 Y

2009 0 4 4 2010 8 9 17 0.01 3.9±0.2 0.70±0.03 N

2011 0 2 2 2012 6 9 15 -0.05 3.8±0.2 0.71±0.02 N

Pooled 58 65 123 0.03* 3.9±0.2 0.70±0.02 N

Quoin Head

2008 3 1 4 2010 4 4 8

Pooled 7 5 12 -0.08 3.9±0.2 0.69±0.03 N

Wilderness gate

2006 1 2 3

Twertup

2000 3 3 6

2001 2 0 2 2002 1 2 3 2003 1 3 4 2004 4 9 13 -0.02 3.8±0.2 0.68±0.03 N

2005 5 5 10 0.03 3.4±0.2 0.65±0.03 Y

Pooled 16 22 38 0.02 3.7±0.2 0.68±0.03 N

38

For most samples, estimates of effective population size (Ne) were negative or had an

upper confidence limit of infinity. However, we were able to obtain estimates of Ne for

the pooled samples from Hamersley Moir and Twertup, which were 70.1 (47.6 – 112.8,

95% CI) and 48.1 (30.2 – 95.4, 95% CI) respectively. Three samples (Hamersley Moir

2006, 2008 and Twertup 2006) displayed shifted modes in their allelic frequency

distributions, consistent with a recent genetic bottleneck event. However, a significant

heterozygosity excess (Wilcoxon test, P < 0.001) relative to the long-term evolutionary

expectation of heterozygosity due to the number of different allelic types in the population

was only found in one sample, the 2005 collection from Twertup.

2.4.2 Population structure

There was strong evidence of genetic structure within the FRNP. Both the STRUCTURE

and GENELAND analyses indicated that K = 2 as the most likely number of genetic

clusters (Table S2.2 and S2.3). Furthermore, under the K = 2 model, in both analyses the

proportion of individuals assigned to each genetic cluster varied according to the

geographical location of the sampling site with individuals from the western site

(Twertup) assigned to one cluster and those from the central and eastern sites assigned to

the other (Figure 2.2). No further sub-structuring was found within each genetic cluster

when individuals assigned to a particular cluster were analysed separately. Pairwise tests

for genetic differentiation also revealed divergences between Twertup and the other sites

across collection years (Table 2.2). By comparison, significant genetic divergences

between samples collected from the same site on different collection years were relatively

uncommon, with only one instance in 16 FST pairwise comparisons and six instances in

16 pairwise Shannon’s mutual information index comparisons (Table 2.2). The AMOVA

revealed 3.2% of the total genetic variation was between genetic clusters, with 1.2% of

the total genetic variation among collection years within genetic cluster (Table 2.3).

39

Figure 2.2 Summary of the Bayesian clustering analysis assuming two admixed

populations (K = 2). Each individual is represented by a bar showing its estimated

membership to a particular cluster (represent by different colours). Black lines separate

individuals collected from different sites. The sampling sites include East Mt Barren

(EMB), Moir track (MT), Gravel pit (GP), Hamersley Moir (HAM), Quoin Head (QH),

Wilderness gate (WG), and Twertup (TW).

Table 2.2 Pairwise FST values (above diagonal) and Shannon’s mutual information index

(below diagonal) between population samples with N ≥ 10. Significant FST values after

correction for multiple comparisons and significant Shannon’s mutual information index

based on 1000 permutation tests are highlighted in bold text.

Hamersley Moir Twertup

Population Year 2005 2006 2007 2008 2010 2012 2004 2005

Hamersley 2005 - 0.002 0.008 0.022 0.012 0.020 0.021 0.050

Moir 2006 0.073 - -0.004 0.000 -0.007 0.007 0.035 0.031

2007 0.058 0.063 - -0.002 -0.001 0.015 0.040 0.071

2008 0.065 0.080 0.046 - -0.004 0.022 0.045 0.073

2010 0.082 0.071 0.061 0.060 - 0.009 0.042 0.060

2012 0.071 0.088 0.068 0.074 0.074 - 0.044 0.043

Twertup 2004 0.097 0.166 0.136 0.135 0.150 0.139 - 0.031

2005 0.117 0.177 0.163 0.162 0.170 0.132 0.097 -

EMB MT HAM QH TW

GP WG

1.0

0.8

0.6

0.4

0.2

0.0

Cluster 1 Cluster 2

40

Table 2.3 Analysis of Molecular Variance (AMOVA) of population samples with N ≥ 10

using 17 microsatellite loci.

Variance % total P ϕ-statistics

Between genetic clusters 0.19 3.2 <0.0001 ϕCT = 0.032

Between collection

years/genetic cluster 0.07 1.2 <0.0001 ϕSC = 0.012

Within collection years 5.85 95.7 <0.0001 ϕST = 0.043

There was no relationship between genetic and geographic distance between sites when

all sites were used (Mantel test, r = 0.074, P = 0.232). However, when the genetically

distinct Twertup site was excluded, a significant positive relationship was evident (r =

0.901, P = 0.037). The SA analyses also provided evidence of spatial genetic structure.

An analysis using all samples from the main genetic cluster (i.e. all samples except those

from Twertup) revealed a significantly positive r value at the first distance class (0 to 1

km) based on the permutation test (P = 0.009). The values of r in the remaining distance

classes were lower and non-significantly different from zero (Table S2.4). These fine-

scale patterns were more apparent when SA analyses were conducted using shorter

distance classes, with significantly positive r values evident in distance classes up to 200

m (Figure 2.3, Table S2.5).

Figure 2.3 Correlogram plot of the genetic correlation coefficient (r) as a function of

distance based on 80 females (black line) and 77 males (grey line), post-dispersal P.

apicalis. The bootstrapped 95% confidence limits are shown for each distance class.

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

100 200 300 400 500 600

r

Distance Class (m)

Females

Males

41

2.4.3 Sex-biased dispersal

We did not detect any differences between male and female spatial autocorrelation

profiles when the full range of distances were used (ω = 3.808, P = 0.697). However,

significant heterogeneity between males and females was evident at fine-spatial scales (ω

= 27.732, P = 0.007, Figure 2.3). While no spatial genetic structure was evident in males,

significantly positive r values were found over the first two distance classes (1 – 100 m

and 101 – 200 m) in females (Figure 2.3). The differences between males and females

were only significant in the first 1 – 100 m distance class (t = 4.855, P = 0.027). The

differences between sexes became more apparent when pre-dispersal individuals were

included in the fine-scale analysis (ω = 41.44, P = 0.001, Figure 2.4). Positive r values of

the first two distance classes in females were significantly higher than males (Distance

class of 1 – 100 m: t = 9.27, P = 0.005; Distance class of 101 – 200 m: t = 7.23, P =

0.007).

Pairwise relatedness of both sexes collected from the same site were higher than between

sites and only significantly higher than pairs collected from different genetic clusters

(Paired wilcoxon rank sum test, females: V = 516314, P < 0.001; males: V = 860767, P <

0.001, Figure 2.5). There was no significant difference in pairwise relatedness between

pairs of females and pairs of males (Wilcoxon rank sum tests, P > 0.05 in all cases, Figure

2.5).

Figure 2.4 Correlogram plot of the genetic correlation coefficient (r) as a function of

distance based on 96 females (black line) and 88 males (grey line), pre- and post-dispersal

P. apicalis. The bootstrapped 95% confidence error bars are shown.

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

100 200 300 400 500 600

r

Distance Class (m)

Females

Males

42

Figure 2.5 Pairwise relatedness between females (colored bars) and males (open bars)

taken from the same sampling site, between sites within the same genetic cluster and

between genetic clusters. The error bars are 95% confidence limits determined with

bootstrapping.

2.5 DISCUSSION

2.5.1 Landscape-scale population subdivision

Parantechinus apicalis within the FRNP exhibited significant population genetic

structuring over relatively small spatial scales (30 – 76 km). Population structure was

most evident with the Bayesian clustering analysis that identified two clear genetic

clusters separating individuals collected at Twertup on the western side of the FRNP from

those collected in the centre and eastern regions of the park. Pairwise FST values also

revealed significant genetic divergences between sites from opposite sides of the FRNP,

which were consistent across collection years. In addition, the clustering analysis

provided no evidence of recent dispersal between the western and central-eastern regions

of the FRNP, suggesting gene flow between these regions is rare.

The genetically distinct Twertup site is located near Twertup Creek and Sussetta Rivers,

the major tributaries of the Fitzgerald River (Figure 2.1), so it is possible that these

waterways act as a barrier to gene flow. Landscape features such as rivers have been

shown to restrict gene flow and cause genetic differentiation between populations in

mammals (Aars et al., 1998, Pfau et al., 2001, Goossens et al., 2005, Eriksson et al., 2004,

Quemere et al., 2010). For example, historical gene flow of the yellow-footed antechinus

(Antechinus flavipes) between different sides of Murray River was disrupted by the

-0.025

-0.020

-0.015

-0.010

-0.005

0.000

0.005

0.010

Within site Between sites

within the same

genetic cluster

Between genetic

clusters

Pai

rwis

e re

late

dnes

s

43

construction of dams and weirs which kept summer river flows consistently high (Lada

et al., 2008). This generated genetic structuring on different sides of the river in just 50

generations (Lada et al., 2008). However, the river systems in the FRNP are seasonal

(Water and Rivers Commission, 2003). Winter rainfall is less intense than in summer.

Flooding in the FRNP occurs mostly in summer when post-cyclonic rain events occur.

While floods and strong river currents can act as barriers after heavy or prolonged rainfall,

riverbeds are mostly dry in summer and less likely to act as a barrier to gene flow. The

Twertup area is on a ridge of upland located between the Fitzgerald River and Twertup

Creek (Chapman and Newbey, 1995). The Fitzgerald River valley has spongolite cliffs

as its distinctive feature. These landscape features which could isolate the Twertup area

from surrounding populations. Further, the western and central regions have a similar

vegetation type (open mallee) so it is unlikely that vegetation patterns would result in

genetic heterogeneity (Moore et al., 1991).

There was evidence of isolation-by-distance (IBD) within one of the genetic clusters. This

was evident with the Mantel tests and also the SA analyses, which indicated dispersal was

restricted to distances up to 1 km when males and females were analysed together. The

dispersal distance of P. apicalis may be influenced by the small body size which could

restrict the dispersal ability to a relatively short distance (Jenkins et al., 2007). Our

trapping records showed an average maximum movement of both sexes to be

approximately 413 metres. One male was reported to move up to 900 m which suggested

males may move further than females. However, the maximum distance a male can travel

is still uncertain because of the limitations of the two trapping regimes used when

collecting tissue samples for this study. In a transect, an animal could have been

recaptured anywhere along the 5 km transect but within the 450m x 500m grid, an animal

could only have been recaptured a maximum of 600m away. Nevertheless, the observed

movement distance in this study is consistent with many dasyurid species of a similar size

(Kraaijeveld-Smit et al., 2007, Banks and Peakall, 2012, Fisher, 2005). Alternatively, the

restricted movement distance would be influenced by resource availability within the

habitat. For instance, the movement of honey possums (1,277 m2 in males and 701 m2 in

females) in the FRNP was suggested to be influenced by the variation in density of nectar

producing plants, and for individual animals the knowledge of the resource availability

may outweigh the benefits of moving elsewhere (Garavanta et al., 2000). Due to these

factors, population structuring of P. apicalis demonstrates a pattern of IBD (Forbes and

Hogg, 1999, Levy et al., 2010, Neaves et al., 2009, Hazlitt et al., 2006). Therefore, the

44

main factor contributing to the observed genetic structure is likely to be the large

geographic distance between the regions and the limited dispersal ability of P. apicalis.

2.5.2 Fine-scale population structure and sex-biased dispersal

In addition to the genetic differentiation between sampling sites within the FRNP, we also

detected spatial genetic structure at fine spatial scales. This was most clearly evident with

the SA analyses, which showed positive genetic structure in distance classes up to 200 m

in females and no significant genetic structure in males over distances up to 600 m. These

results suggest females of P. apicalis are philopatric and males are the dispersing sex, and

are consistent with previous studies on small marsupials (Peakall et al., 2003, Cockburn

et al., 1985, Hazlitt et al., 2004, Soderquist and Lill, 1995). For example, in Antechinus

agilis females exhibit positive spatial genetic structure over distances up to 300 m, while

the spatial genetic structure in males was weak (Banks and Peakall, 2012). In addition,

males in many small dasyurid species have been reported to move greater distances than

females: A. agilis 1,181 m vs 87 m (Banks and Peakall, 2012), A. stuartii 1,230 m vs 270

m (Fisher, 2005), Pseudantechinus macdonnellensis 180 m vs 86 m (Pavey et al., 2003),

Ningaui yvonneae 160 m vs 84 m (Bos and Carthew, 2008), Dasycercus blythi 571 m vs

446 m (Körtner et al., 2007), and D. cristicauda 582 m vs 317 m (Masters, 2003).

There are two mechanisms explaining the evolution of the philopatry and sex-biased

dispersal pattern. Firstly, philopatry promotes genetic heterogeneity among populations

by providing opportunities for inbreeding (Piertney et al., 1998) and thereby maintaining

high frequencies of alleles that are locally advantageous (e.g. Stiebens et al., 2013).

Secondly, male-biased dispersal prevents breeding between related individuals (Lawson

Handley and Perrin, 2007, Greenwood, 1980), therefore providing a mechanism to

prevent the deleterious effects of inbreeding caused by philopatry (Piertney et al., 1998).

Other mechanisms of inbreeding avoidance in small mammals include female mating

preference for non-related males (Parrott et al., 2015) and mothers enforcing dispersal of

juvenile males shortly after they are weaned (Cockburn et al., 1985, Fisher, 2005).

45

The level of fine scale population genetic structure of P. apicalis also varied temporally

due to changes associated with the timing of dispersal. There was much stronger fine-

scale genetic structure when pre-dispersal individuals were included in the analyses,

especially in females. This likely reflects the sampling of individuals within family

groups, which can increase the signal of positive spatial structure (Peakall et al., 2003).

These patterns can subsequently disappear after dispersal has occurred (Scribner and

Chesser, 1993).

2.5.3 Management implications

Our findings have management implications not only for P. apicalis but also for other

small mammals exhibiting restricted dispersal. Firstly, landscape features and long

distances act as barriers to dispersal in small mammals leading to genetic structure as the

result of reduced gene flow. Conservation management should therefore recognize the

presence of population structuring and manage subpopulations accordingly. For example,

our study revealed that P. apicalis from eastern and western sides of the FRNP are

genetically distinct, and should therefore be managed as separate subpopulations. Hence,

fire management and predation control, an important part of conservation management in

the southwest Australian Mediterranean ecosystems, needs to be designed in a way that

ensures the persistence of both subpopulations. Parantechinus apicalis in particular may

be more sensitive to stochastic events due to its short-dispersal distance. Replenishment

of populations depleted by stochastic events may not occur, leaving parts of the park

unpopulated and the remaining populations vulnerable to population size declines and

loss of genetic diversity. Indeed, the estimates of effective population size we obtained

were consistently low, highlighting the vulnerability of the subpopulations to inbreeding

and genetic drift.

Secondly, if one of the objectives for captive breeding and translocation programs is to

maximise genetic variation, individuals should be selected from multiple subpopulations.

Because genetic differentiation between these subpopulations was very low, the risk of

outbreeding depression is also low. It has been recommended that at least 30 individuals

should be selected to retain 90 – 95 % of the genetic diversity in the source population

(Hedrick, 2000, Ottewell et al., 2014, Allendorf and Luikart, 2007). However, without

assessing genetic structure, the intended level of genetic diversity may not be adequately

captured and representative of a small genetic subpopulation may be lost from the new

population (e.g. Raisin et al., 2012). For example, sampling individuals from one

46

subpopulation within the FRNP only would lose at least 3.2% of the total genetic variation

and unique alleles from the other subpopulation would be lost. Selecting individuals from

both subpopulations would not only capture unique alleles occurring in each

subpopulation, but also significantly reduce the average genetic relatedness between

individuals. To reduce the potential for inbreeding within captive or translocated

populations (e.g. Swinnerton et al., 2004), we also recommend sampling post-dispersal

males i.e. sexually mature males and females from sites that are at least 200 m apart.

48


49

CHAPTER THREE

Temporal variation in the genetic composition of a newly

established population of dibblers (Parantechinus apicalis)

reflects translocation history

3.1 ABSTRACT

Loss of genetic variation and increased population differentiation from source

populations are common problems for translocations that use captive animals or a small

number of founders to establish a new population. This study evaluated the genetic

changes occurring in a captive and a translocated population of the dibbler

(Parantechinus apicalis) that were established from multiple source populations over a

ten year period. While the levels of genetic variation within the captive and translocated

populations were relatively stable and did not differ significantly from the source

populations, their effective population size reduced 10 – 16 fold over the duration of this

study. Evidence of genetic bottlenecks was detected only after the translocated population

was established. There were also marked changes in the genetic composition of both

populations that were strongly associated with the origins of individuals introduced to the

populations. Interbreeding between individuals from different source populations

lowered genetic relatedness among offspring, but this was short-lived. These results

highlight the importance of the origins and the timing of release of founding individuals

in determining the genetic composition of a newly established population.

50

3.2 INTRODUCTION

Many species have experienced declines in their abundance, distribution, or have become

extinct as a result of human activities and introduced species (Burbidge and McKenzie,

1989). As these threats continue to extirpate populations, translocations, a conservation

tool, which involves moving individuals from one location to another, have been

implemented to restore populations and facilitate species continuity (Fischer and

Lindenmayer, 2000, Wolf et al., 1996). The success of a translocation is influenced by

various factors including the efficiency in the removal of threat(s), habitat quality, size of

released area, and the number of individuals released (Wolf et al., 1998, Fischer and

Lindenmayer, 2000, Short, 2009). Recently, genetic approaches have become important

in determining appropriate source populations for release as well as for ongoing

monitoring to assess whether there has been loss of genetic diversity or inbreeding

(Moritz, 1999, Schwartz et al., 2007, Ottewell et al., 2014, Kennington et al., 2012).

Translocated and captive populations are prone to loss of genetic diversity and inbreeding

because they are often established with a small number of individuals (Earnhardt, 1999,

Robert, 2009, Jamieson, 2011). Small numbers of founders often leads to small effective

population sizes that can result in fluctuations of allele frequencies and divergences from

source populations (Hundertmark and Van Daele, 2010, Broders et al., 1999, Biebach and

Keller, 2009). Further, establishing new populations using individuals selected from

inbred wild populations (e.g. Slate et al., 2000, Nielsen et al., 2012, Grueber et al., 2010,

Madsen et al., 1996) or captive populations (e.g. Laikre and Ryman, 1991, Bilski et al.,

2013) has the potential to further increase inbreeding and reduce fitness (e.g. Swinnerton

et al., 2004).

The dibbler (Parantechinus apicalis) is a small (40 – 100 g) insectivorous marsupial

(Miller et al., 2003, Bencini et al., 2001) endemic to Western Australia. It is found on two

islands, Boullanger and Whitlock Islands, off the coast from Jurien Bay and in the

Fitzgerald River National Park (~ 3000 km2) (Morcombe, 1967, Fuller and Burbidge,

1987). Dibblers are currently listed as Endangered under the Environment Protection and

Biodiversity Conservation Act (1999) and 2014 IUCN Red List of Threatened Species

(Friend et al., 2008). They have a polygynandrous mating system (Strahan, 1983) and

exhibit male-biased dispersal and female philopatry (Thavornkanlapachai et al., in prep).

They are seasonal breeders, breeding once a year around February to early April (Mills

et al., 2012). A female produces up to eight young per breeding season (Mills et al., 2004)

51

with the young reaching sexual maturity after 10 to 11 months (Woolley, 1995). Female

dibblers can live up to four years and males up to three years (Friend and Collins, 2005).

While island dibblers exhibit facultative male die-off after the first breeding season,

mainland male dibblers have been reported to survive well into their second year (Friend

and Collins, 2005).

The main threats to dibblers include introduced predators such as foxes and feral cats,

inappropriate fire regimes, habitat degradation and competition with the introduced house

mouse, Mus musculus (Friend, 2003). These threats have resulted in a steady decline in

dibbler population sizes. There are currently less than 1000 mature individuals estimated

to be on the mainland (Woinarski et al., 2014) and approximately 100 individuals known

to be alive on both islands (Moro, 2003). In a bid to bolster declining population sizes, a

captive breeding and several translocated populations have been established. The captive

breeding program at Perth Zoo commenced in 2000 using wild dibblers from the

Fitzgerald River NP mainland population. In 2001, captive born dibblers were released

to a translocation site at Peniup Nature Reserve (~ 30 km West of the FRNP). A further

six releases from the captive population to this site followed over the next nine years.

A previous study on the mainland population confirmed two distinct genetic clusters on

the western and eastern sides of the Fitzgerald River NP (Thavornkanlapachai et al., in

prep). Both populations were used in the captive breeding program, but it is unknown if

both lineages were successfully established. The objective of this study is to determine

the relative success of the Peniup translocation in maintaining genetic variation from its

sources and to examine the extent of admixture within captive and translocated

populations over a ten year period.

52


3.3.1 Study site and sample collection

Samples were collected from the translocation site at Peniup Nature Reserve (34°10’S,

118°49’E) at the time of its establishment and during follow-up population monitoring

between 2001 to 2013 (Figure 3.1, Table 3.1). The captive population was established

using 26 individuals collected over multiple years from several sites in the Fitzgerald

River National Park, Western Australia (33°52’S, 119°54’E) (see Table 3.1). Pairing

selections were based on a minimum kinship design. Each sex was ordered according to

their minimum kinship estimates, and the males and females with the lowest estimate

were paired together, and on down the list until all have been allocated a partner. From

this captive population 219 captive born dibblers as well as 16 of the original founders

were released to the translocation site once a year in 2001 – 2003, 2006, and 2008 – 2010

(Table 3.1). In addition to samples from the captive and translocated populations, samples

were collected from each of the source populations during regular monitoring of these

populations between 2000 and 2012. A total of 188 samples were collected from

Hamersley Moir (HAM) and Moir Track (MT), the eastern source population (33°53’S,

119°55’E) and 49 samples from Twertup (TW), the western source population (33°58’S,

119°16’E). A total of 137 samples were collected from wild-born animals at the

translocation site during follow-up monitoring 2002 – 2013. All sampled individual had

a tissue sample taken from their ear, a microchip implanted and their sex recorded. Ear

tissue samples were stored in 20% DMSO2 saturated with NaCl at room temperature.

53

Table 3.1 Summary of Parantechinus apicalis samples used in this study. Eastern and

western sources represent the source populations and the locations where wild-born

animals were trapped: Hamersley Moir (HAM), Moir track (MT) and Twertup (TW).

Founding animals are individuals selected from the source populations to breed in the

captive colony at Perth Zoo. Captive represents animals born in captivity between 2000

and 2010. Release represents animals that were released to the Peniup Nature Reserve

between 2001 and 2010. This includes both captive-born and some founding animals. PG

represents wild-born animals caught at the translocation site.

Eastern

source

Western

source

Founding

animals

Year HAM MT TW East West Captive Release PG

2000 11 6 8

2001 2 2 1 41 41

2002 3 39 46 5

2003 2 4 36 43 14

2004 3 18 2 6 43

2005 45 5 10 3

2006 17 15 6 11

2007 22 7 5 3

2008 18 19 24

2009 5 2 20 34 7

2010 17 37 41 14

2011 2 1 21

2012 19 15

2013 4

54

Figure 3.1 Map of Parantechinus apicalis trapping sites within the Fitzgerald River

National Park and the location of the translocated population at Peniup Nature Reserve

(PG) (boxed in inset map of Western Australia). Sites HAM and MT represent the eastern

lineage and TW the western lineage.

3.3.2 DNA extraction and microsatellite genotyping

DNA was extracted using the ‘salting-out’ method (Sunnucks and Hales, 1996) with a

modification of a 56 °C incubation step and 10 mg/mL of Proteinase K being added to

300 µL TNES. Each individual was genotyped using the following 21 microsatellite loci

developed for P. apicalis and other dasyurids: pPa2D4, pPa2A12, pPa2B10, pPa7A1,

pPa7H9, pPa9D2, pPa1B10, pPa4B3, pPa8F10 (P. apicalis, Mills and Spencer, 2003) ;

pDG1A1, pDG1H3, pDG6D5 (Dasyurus geoffroii, Spencer et al. 2007) ; 3.1.2, 3.3.1,

3.3.2, 4.4.2, 4.4.10 (Dasyurus spp., Firestone, 1999) ; Sh3o, Sh6e (Sarcophilus laniarius,

Jones et al., 2003) ; Aa4A (Antechinus agilis, Banks et al., 2005) , Aa4J (A. agilis,

Kraaijieveld-Smit et al., 2002b) . PCRs (volume 10 µL) were performed using a QIAGEN

Multiplex PCR Kit and contained primer concentrations ranging from 0.04 to 1.5 µM and

10 – 20 ng of DNA (Table S2.1). Amplifications were performed using an Eppendorf

Mastercycler epgradientS Thermocycler with the following steps: 15 min at 95 °C, 35 to

40 cycles at 94 °C for of 30 s, the annealing temperature (46 °C to 58 °C) for 90 s, 72°C

for 60 s, and finally 60 °C for 30 mins (Table S2.1). PCR products were analysed in an

TW

HAM

MT

PG

55

ABI 3730 sequencer using a GeneScan-600 LIZ internal size standard and scored using

GENEMARKER version 1.90 (SoftGenetics).

3.3.3 Data analysis

Genotypic quality was assessed by calculating the allele-specific and locus-specific

genotypic error rates (Pompanon et al., 2005). We tested for the presence of null alleles

in the source population samples at each locus using MICROCHECKER (Van Oosterhout

et al., 2004). We analysed samples from each population by collection year when N ≥ 10

and as pooled samples (all collection years analysed together). Microsatellite variation

was quantified by calculating the allelic richness (AR) (the allele number per locus

estimate corrected for sample size) and gene diversity (H). Deviations from Hardy-

Weinberg Equilibrium were assessed by calculating the inbreeding coefficient (FIS) and

randomisation tests were performed to test the significance of the deviations. Positive FIS

values indicate a deficit of heterozygotes, while negative FIS values indicate an excess of

heterozygotes. Randomisation tests were also performed to test for genotypic

disequilibrium between each pair of loci. For these tests, the sequential Bonferroni

correction (Rice, 1989) was applied to control for type I statistical error. Genetic

differentiation between population samples were quantified using Weir & Cockerham’s

(1984) FST and were assessed for significance using randomisation tests. All above

genetic parameters and tests were calculated using FSTAT version 2.9.3.2 (Goudet,

2001). The number of rare alleles (Ar) with frequency less than 5% and the number of

unique alleles (Au) were calculated in GENALEX version 6.5 (Peakall and Smouse,

2012). Differences in H, Ar, Au, and AR between collection years and pooled sample

populations were tested using Wilcoxon’s signed-rank tests with loci as the pairing factor

using the R statistical package version 3.0.1 (R Core Team, 2014).

The effective population size (Ne) for each population sample and samples pooled cross

collection years were estimated using the single-sampled estimator of Ne as implemented

in the software package LDNE (Waples and Do, 2008). We assumed that all of our

population samples consisted of overlapping generations. We used a random mating

model and estimated linkage disequilibrium amongst alleles using only alleles with

frequencies > 5%, as this expected to give the best balance between precision and bias in

the Ne estimator (Waples and Do, 2010).

The occurrence of recent reductions in Ne was investigated by testing for an excess in

heterozygosity using the program BOTTLENECK (Piry et al., 1999). Both the stepwise

56

mutation model (SMM) and two-phase model (TPM) were used. These models were

chosen because they are considered to be the most appropriate for microsatellite data (Piry

et al., 1999). Variance for TPM was set to 12 and the proportion of SMM in TPM was

95% with 1000 iterations following approaches described by Luitkart and Cornuet (1998)

and Luikart et al. (1998).

To investigate the extent of genetic mixing between the eastern and western source

population lineages within the captive and translocated populations, we used a

Discriminant Analysis of Principal Components (DAPC) provided in the Adegenet

version 2.0.1 (Jombart, 2008, Jombart et al., 2010). DAPC grouped individuals to achieve

the largest between-group variance and the smallest within-group variance using linear

combinations of alleles (Jombart et al., 2010). To achieve this, it performs Principal

Component Analysis as a prior step to the Discriminant Analysis. We ran the find.cluster

command with the number of component (PCs) that allowed 90% of cumulative variance

to be retained (between 40-50 PCs) and selected two clusters based on the number of

source populations. Then we ran the dapc command but we used optim.a.score to select

the number of PCs to retain and the discriminant functions to save was the number of

group or collection year – 1.

A chi-square test was used to determine whether the observed genetic proportion from

each lineage in the DAPC analysis matched the expected genetic proportions from the

captive breeding pedigree record. The observed genetic proportions were calculated from

the proportion of individuals assigned to either the Eastern or Western cluster in the

DAPC cluster analysis. The expected genetic proportions of captive-born animals were

based on pedigree records. The expected genetic proportions of the translocated

population were calculated from genetic proportions of released animals up to two years

prior the year of interest. For example, to calculate the expected genetic proportion of the

collection year 2003, captive-bred individuals born between 2000 and 2001 were included

in the calculation. This is because the releases generally occurred in October whereas the

sampling occurred in January-May, prior to the mating season in March-April (i.e. the

offspring of these released animals would not be sampled until January-May the

following year).

Finally, pairwise relatedness estimates were calculated using the method of Lynch and

Ritland (Lynch and Ritland, 1999) implemented in GENALEX version 6.5 (Peakall and

Smouse, 2012). Differences in pairwise relatedness between population samples were

57

tested using Wilcoxon’s signed-rank tests implemented in the statistical package R

version 3.0.1 (R Core Team, 2014). Confidence limits for population mean values were

calculated using bootstrapping (1000 bootstraps) in R.

3.4 RESULTS

3.4.1 Effects of translocation on genetic variability

The allele-specific and locus-specific genotyping error rates were 0.016 and 0.026,

respectively. The average amplification success rate per locus was 0.946.

MICROCHECKER identified one locus (aPa1B10) as having null alleles in both source

populations. This locus was removed from further analysis.

Overall, estimates of genetic diversity of the captive and translocated population were

lower than the source populations (Figure 3.2a and 3.2b, Table 3.2). This pattern was

consistent over multiple years. Population samples from the translocated population in

years 2003 and 2006 showed the lowest levels relative to the source populations with 17

out of 18 comparisons for H and nine out of 18 comparisons for AR being significantly

lower than the source populations (Wilcoxon rank sum tests, P < 0.05).

58

(a)

(b)

Figure 3.2 Estimates of (a) allelic richness (AR) and (b) gene diversity (H) within the

source, captive and translocated populations. Standard error bars are given around the

means.

2.0

2.5

3.0

3.5

4.0

4.5

5.0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

AR

0.40

0.45

0.50

0.55

0.60

0.65

0.70

0.75

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

H

Year

East West

Captive population Translocated population

59

A total of 155 alleles across 20 loci were detected. Of these, 38 (24.5%) were unique to

the eastern source population and 17 (11.0%) were unique to the western source

population. The eastern source population possessed a significantly higher number of

unique alleles, on average, than any other population samples (Wilcoxon rank sum tests,

P < 0.01 in all cases, Table 3.2). However, there were no significant differences in rare

alleles between population samples (Friedman rank sum test, P = 0.858). The captive

population maintained a relatively similar total number of alleles to those found in the

wild-caught founders. A slight loss of allele number (8.8%) was observed from the wild-

born dibblers in the translocated population when compared to the captive population.

Wild-born individuals in the translocated population retained 13 (34.2%) and 6 (35.3%)

of the unique alleles from the eastern and western source populations respectively.

Estimates of Ne were much lower in the captive population (Ne = 24.5, range 21.0 to 28.5)

than the source populations (Eastern source population, 74.1, range 52.5 to 110.9;

Western source population, 54.1, range 36.7 to 91.4, Table 3.2). Ne of the translocated

population was comparable to the captive population with an overall estimate of 16.7

(range 14.5 to 19.1). Population bottlenecks were also detected more frequently in the

captive and translocated populations than the source populations (Table 3.2). All

bottlenecks in these populations were detected after they had become established.

60

Table 3.2 Estimates of genetic variation over 20 microsatellite loci within the source,

captive and translocated populations. N is an average sample size per locus. A is the total

number of alleles. Au is an average of unique alleles. Ar is an average number of rare

alleles (frequency < 5%). AR is allelic richness. H is gene diversity. FIS is inbreeding

coefficient. GD is genotypic disequilibrium. Ne is an effective population size. Standard

errors are given after mean values. Asterisks represent FIS values significantly different

to zero at P < 0.05.

Population N A Au Ar AR H FIS GD Ne Ne range Bottleneck

East

2005 40.2±2.3 113 0.3±0.2 1.2±0.3 4.2±0.3 0.64±0.05 0.11* 0 NA NA N

2006 16.3±0.3 103 0.1±0.1 0.9±0.4 4.3±0.4 0.63±0.06 -0.02 0 15.0 11.1−21.4 N

2007 20.4±0.3 108 0.1±0.1 1.3±0.4 4.3±0.4 0.64±0.05 0.02 0 42.9 26.9−90.3 N

2008 17.3±0.2 93 0 0.5±0.2 4.0±0.3 0.62±0.05 0.04 0 15.2 11.4−21.2 Y

2010 15.5±0.5 96 0 1.0±0.2 4.1±0.3 0.64±0.06 0.00 0 NA NA N

2012 19.0±0.1 95 0 0.6±0.2 3.9±0.3 0.64±0.05 -0.04 0 9.4 7.5−11.6 N

Overall 141.3±3.3 133 0.9±0.3 2.4±2.8 4.2±0.4 0.64±0.05 0.05* 1 74.1 52.5−110.9 N

West

2000 6.7±0.5 83 0.1±0.1 0.1±0.1 NA 0.64±0.06 0.13 0 NA NA -

2004 16.8±0.2 97 0.1±0.1 0.6±0.3 4.1±0.4 0.64±0.05 0.00 0 100.2 35.6−∞ N

2005 9.9±0.1 79 0 0 3.6±0.4 0.61±0.05 0.03 0 42.3 18.8−∞ N

Overall 42.9±0.9 112 0.4±0.2 1.5±1.6 4.0±0.4 0.63±0.05 0.03 0 54.1 36.7−91.4 N

Founders 24.7±0.3 114 0 1.6±2.2 4.3±0.4 0.64±0.05 0.03 0 69.7 40.0−204.1 N

Captive population

2001 41.0±0.0 96 0 0.6±0.3 3.9±0.4 0.60±0.05 0.01 14 5.6 4.0−7.0 N

2002 38.2±0.3 90 0 0.7±0.3 3.6±0.3 0.58±0.05 -0.08 7 5.3 3.8−6.9 N

2003 35.8±0.1 83 0 0.5±0.2 3.5±0.3 0.57±0.05 -0.06 7 6.0 3.9−8.0 N

2006 15.0±0.0 78 0 0.5±0.2 3.4±0.3 0.55±0.05 -0.07 0 2.1 1.9−2.5 Y

2008 18.9±0.1 84 0 0.2±0.1 3.7±0.4 0.58±0.06 -0.10 3 2.0 1.8−2.3 Y

2009 18.7±0.4 80 0 0.3±0.1 3.5±0.3 0.56±0.06 -0.03 1 4.2 3.0−6.1 Y

2010 36.9±0.1 97 0 0.8±0.3 3.8±0.3 0.60±0.05 -0.02 2 9.8 8.1−11.7 N

Overall 220.4±0.6 113 0 1.6±2.6 4.0±0.3 0.61±0.05 -0.01 50 24.5 21.0−28.5 N

Translocated

population

2003 13.7±0.1 67 0 0.4±0.1 3.0±0.2 0.51±0.05 0.00 0 4.0 2.6−7.8 N

2004 41.8±0.4 85 0 0.7±0.2 3.5±0.3 0.57±0.05 -0.05 0 10.8 8.7−13.3 N

2006 9.0±0.3 58 0.1±0.1 0.1±0.1 2.8±0.2 0.48±0.05 -0.08 0 16.7 6.5−1890.3 Y

2010 12.5±0.3 75 0 0.6±0.2 3.4±0.2 0.59±0.05 -0.03 0 5.4 3.0−8.8 N

2011 21.0±0.0 83 0 0.7±0.2 3.5±0.3 0.59±0.05 -0.10 0 4.2 3.1−5.7 N

2012 15.0±0.0 77 0 0.4±0.1 3.5±0.3 0.60±0.05 -0.03 0 8.2 5.6−11.8 Y

Overall 131.6±1 103 0.1±0.1 1.3±1.2 3.7±0.3 0.60±0.05 0.01 13 16.7 14.5− 19.1 N

61

3.4.2 Population structure of captive and translocated populations

There was low, but significant genetic differentiation between the source populations (FST

= 0.046). No significant temporal variation in FST was detected within these populations,

but there was significant temporal variation in both captive and translocated populations

(Table 3.4). Initially, pairwise FST values were lower between the captive and western

source population than between the captive and eastern source population, but this

changed with the opposite pattern evident in 2008 – 2010 (Table 3.4). A similar pattern

was observed in the translocated population, but mostly in collection year 2010 – 2012

(Table 3.4). Consistent with the pairwise FST values, there were two genetic clusters

detected the captive population by the DAPC analysis separating collection year 2000 –

2003 from 2006 – 2010 (Figure 3.3b). A similar pattern was observed in the translocated

population for collection year 2002 – 2007 and 2009 – 2013 (Figure 3.3c). This change

occurred after more wild-caught individuals from the eastern source population were

introduced to the captive population in 2007 and 2009 (Table 3.1 and Figure 3.3a). The

expected genetic proportion calculated from the pedigree showed that the proportion of

captive born dibblers with the eastern ancestry increased from 22% – 35% to 46% – 66%

after collection year 2006 (Table 3.3). This change was not detected in the translocated

population until 2009 (Figure 3.3c). The observed genetic proportions from DPAC

analysis were not significantly different from the expected based on the pedigree in most

years. However, significant deviations were detected in 2008 for the captive born dibblers

and in 2010 – 2012 for the wild born dibblers at the translocation site (Table 3.3). Both

population lineages were expected to become evenly mixed in the translocated population

by collection years 2011 and 2012, but the observed proportions from the eastern lineage

were significantly higher than expected (Table 3.3).

62

a) Source populations and founding individuals

b) Captive population

63

c) Translocated population

Figure 3.3 Scatterplot of the DAPC analysis showing the first two principal components.

Clusters in different colours represent different collection years, except the source

populations, which represented individuals pooled across the 2000-2013 collections. Dots

represent individuals. Insets show the histogram of discriminant analysis eigenvalues. a)

shows the source populations and founders introduced to the breeding program in

different years. b) shows the captive population. c) shows the translocated population.

64

Table 3.3 Observed and expected proportions of each source population in samples taken

from the captive and translocated populations. Observed proportions are calculated from

the proportion of individuals assigned to different source populations assuming two

admixed populations (K = 2) in the DAPC analysis. The expected proportions of the

captive and translocated population are based on average genetic values calculated from

the pedigree record. Significant P-values are denoted with a bold font.

n

Observed Expected

Sample East West East West Х2 P

Captive population

2001 41 0.44 0.56 0.35 0.65 1.6 0.207

2002 39 0.38 0.62 0.31 0.69 1.1 0.298

2003 36 0.25 0.75 0.27 0.73 0.1 0.744

2006 15 0.20 0.80 0.22 0.78 0.0 0.860

2008 19 0.74 0.26 0.46 0.54 5.9 0.016

2009 20 0.55 0.45 0.61 0.39 0.3 0.592

2010 37 0.54 0.46 0.66 0.34 2.2 0.134

Translocated population

2003 14 0.07 0.93 0.29 0.71 3.2 0.072

2004 43 0.19 0.81 0.30 0.70 2.6 0.108

2006 11 0.27 0.73 0.29 0.71 0.0 0.893

2010 14 1.00 0.00 0.38 0.62 22.8 <0.001

2011 21 1.00 0.00 0.46 0.54 25.1 <0.001

2012 15 1.00 0.00 0.54 0.46 12.7 <0.001

65

Multilocus FIS values of the captive- and Peniup-born dibblers were mostly negative,

indicating a heterozygosity excess, but they were not significantly different from zero

(Table 3.2). Significantly positive multilocus FIS values were observed in the eastern

source population only (Randomization tests, P < 0.002). The number of pairs of loci in

genotypic disequilibrium (GD) ranged from zero to 14. The highest number occurred in

the captive population, especially during the first few generations, but it declined over

time. The number of pairs of loci in GD in the source populations and translocated

population ranged from zero to one (Table 3.2).

66

Table 3.4 Pairwise FST estimates between source (East and West), captive and translocated populations of P. apicalis. FST estimates

significantly greater than zero (P < 0.05) after correction for multiple comparisons are highlighted in bold text.

East West Captive population Translocated population

Population Year 2006 2007 2008 2010 2012 2004 2005 2001 2002 2003 2006 2008 2009 2010 2003 2004 2006 2010 2011 2012

East 2005 0.009 0.005 0.018 0.012 0.024 0.026 0.047 0.042 0.058 0.070 0.064 0.047 0.037 0.034 0.116 0.080 0.100 0.041 0.058 0.047

2006 0.000 0.009 0.004 0.027 0.040 0.046 0.053 0.075 0.082 0.055 0.045 0.047 0.030 0.148 0.097 0.144 0.045 0.065 0.046

2007 -0.004 -0.004 0.013 0.038 0.062 0.049 0.073 0.087 0.063 0.043 0.042 0.031 0.142 0.103 0.137 0.036 0.058 0.046

2008 -0.005 0.025 0.047 0.072 0.055 0.070 0.083 0.074 0.049 0.039 0.036 0.147 0.102 0.149 0.042 0.065 0.057

2010 0.017 0.042 0.058 0.045 0.069 0.077 0.076 0.047 0.039 0.032 0.124 0.090 0.132 0.024 0.051 0.047

2012 0.052 0.053 0.058 0.083 0.091 0.107 0.073 0.077 0.055 0.138 0.102 0.148 0.067 0.080 0.080

West 2004 0.018 0.037 0.066 0.070 0.070 0.055 0.062 0.042 0.122 0.075 0.128 0.071 0.066 0.062

2005 0.039 0.064 0.065 0.078 0.068 0.072 0.050 0.145 0.081 0.138 0.072 0.092 0.068

Captive

population 2001 0.008 0.015 0.066 0.041 0.056 0.039 0.079 0.033 0.080 0.050 0.060 0.050

2002 0.005 0.081 0.060 0.059 0.052 0.099 0.030 0.065 0.064 0.065 0.066

2003 0.089 0.073 0.080 0.063 0.107 0.029 0.090 0.070 0.072 0.073

2006 0.044 0.068 0.050 0.163 0.090 0.183 0.064 0.086 0.057

2008 0.023 0.023 0.136 0.085 0.155 0.066 0.082 0.065

2009 0.028 0.156 0.101 0.154 0.055 0.075 0.067

2010 0.127 0.082 0.132 0.042 0.055 0.027

Translocated

population 2003 0.031 0.105 0.133 0.123 0.129

2004 0.062 0.080 0.073 0.082

2006 0.123 0.096 0.108

2010 0.033 0.021

2011 0.020

67

3.4.3 Genetic relatedness comparisons

Overall, pairwise relatedness values of wild-born dibblers in the translocation site were

consistently higher than wild-born dibblers in the source populations (Figure 3.4). Most

of the pairwise comparisons between these populations were significantly different

(Wilcoxon rank sum tests, P < 0.05 in 47 out of 54 comparisons). Pairwise relatedness of

the captive population was significantly different to the translocated population in most

collection years, except for 2003 and 2006 when only a few pairwise relatedness values

were significant (captive 2003 vs Peniup 2003, 2006; and captive 2006 vs Peniup 2003,

Wilcoxon rank sum tests, P < 0.05). On average, pairwise relatedness values for captive-

born dibblers were higher than the source populations but only significantly in some years

(e.g. 2006 and 2010, Wilcoxon rank sum tests, P < 0.001) (Figure 3.4).

Figure 3.4 Mean pairwise genetic relatedness of the source, captive and translocated

populations. Error bars are bootstrapped 95% confidence limits.

Means pairwise genetic relatedness of wild-born individuals used to establish the captive

population were lower between individuals from different sources than individuals from

the same source. Especially, between pairs of females sampled from different sources,

which had significantly lower pairwise relatedness values than between females from the

same population (V = 105.5, P < 0.001; Figure 3.5). In the source populations, pairwise

relatedness values between pairs of females were significantly higher than between pairs

of males or between female-male pairs (Wilcoxon rank sum tests, P < 0.01 in all

comparisons, Figure 3.6).

-0.05

0

0.05

0.1

0.15

0.2

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Pai

rwis

e re

late

dnes

s

Year

East West


68

Relatedness values between pairs of captive-born dibblers were significantly higher than

between pairs of the original founders of the captive population (Wilcoxon rank sum tests,

P < 0.01 in all cases), but were not significantly different from the source populations

except for comparisons between pairs of females (V = 1926022, P < 0.05) and female-

male pairs (V = 8309912, P < 0.001) from the eastern source population (Figure 3.6).

Dibblers born at the translocation site had significantly higher values of pairwise

relatedness than captive-born dibblers (Wilcoxon rank sum tests, P < 0.001 all cases

except captive pairs of females, Figure 3.6). They also had significantly higher pairwise

relatedness values than the source populations for pairs of males and female-male pairs

(Wilcoxon rank sum tests, P < 0.001), but not for pairs of females, which were

significantly lower than the western source population (V = 16722.5, P < 0.001) and not

significantly different from the eastern source population. It was also noteworthy that

pairwise relatedness values for pairs of females were significantly lower than for pairs of

males and female-male pairs at the translocation site as opposite to the source populations

(Wilcoxon rank sum tests, P < 0.001).

Figure 3.5 Mean pairwise genetic relatedness between female-female pairs (FF), male-

male pairs (MM) and female-male pairs (FM) of founding individuals selected from the

same or different source populations. Error bars are bootstrapped 95% confidence limits.

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

Same source Different source

Pai

rwis

e re

late

dnes

s

FF MM FM

69

Figure 3.6 Mean pairwise genetic relatedness between female-female pairs (FF), male-

male pairs (MM) and female-male pairs (FM) of the source populations (west and east),

the wild-caught founders of the captive population, and the captive and translocated

populations. Error bars are bootstrapped 95% confidence limits.

3.5 DISCUSSION

3.5.1 Genetic consequences of mixing subpopulations

Translocated populations often experience significant loss of genetic variation and

become genetically distinct from their source populations as a result of founder effects,

genetic bottlenecks and/or genetic drift (Biebach and Keller, 2009, Hundertmark and Van

Daele, 2010, Broders et al., 1999, Gautschi et al., 2002, Mock et al., 2004). In this study,

we show that a translocated population of dibblers has no significant loss of genetic

diversity after 10 generations. However, both the captive and translocated population

experienced reductions of Ne by 10 – 16 fold, which is similar to declines seen in other

translocated populations (Ottewell et al., 2014, Miller et al., 2011b, Ramstad et al., 2013,

Fitzsimmons et al., 1997, Jamieson, 2011). In addition, many unique, as well as rare

alleles, were lost from the translocated population. This is not unexpected given the

reduction in Ne and rare alleles being more prone to loss following founder events and

genetic bottlenecks than common alleles (Allendorf, 1986, Leberg, 1992, Nei et al.,

1975).

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

East West Founder Captive

population

Translocated

population

Pai

rwis

e re

late

dnes

s

FF MM FM

70

The maintenance of genetic diversity in the translocated population may be attributed to

several factors. First, insignificant loss of allelic richness could owe to the sufficient

number of founding animals that captured the majority of alleles from the source

populations and the efficiency of the captive breeding program in maintaining these

alleles (Table 3.2). A rapid population growth may also have shortened the duration of

genetic bottleneck, minimising its effect on gene diversity. This was supported by the

lack of genetic bottleneck signatures detected during the early phase of translocation and

a steady increase of the effective population size in the translocated population after the

establishment and the population crash in 2006 due to a lapse of in effective predator

control (Friend, pers. comm.). Indeed, a rapid population expansion was identified as the

main factor for the high retention of genetic diversity despite a significant reduction of

population size in the white-tailed deer (Odocoileus virginianus) following its

reintroduction to different parts of Mississippi (Deyoung et al., 2003) and the European

rabbit (Oryctolagus cuniculus) to Australia (Zenger et al., 2003). Second, admixture from

multiple source populations is likely to have bolstered genetic variation in the translocated

population counteracting subsequent losses that may have occurred (Huff et al., 2010,

Kennington et al., 2012, Ransler et al., 2011, Stockwell et al., 1996). For example, in

translocated populations of the brown anole, Anolis sagrei, a reduction in genetic

diversity following a founder event and an increase in genetic variation due to admixture

were suggested to occur simultaneously, resulting in the maintenance of haplotype

diversity in one population and higher haplotype diversity in another (Kolbe et al., 2007).

Third, multiple releases of captive-bred individuals may have replenished genetic

diversity lost due to post-released mortality and variance in reproductive success amongst

founders (Williams and Scribner, 2010, Jamieson, 2011). The size of the dibber

translocated population was reported to have crashed in 2006. Continuing releases of

captive animals to this population were likely to have offset the genetic impacts of the

population crash.

3.5.2 Consequences of admixture on population structure

The captive and translocated populations in this study were established using individuals

from two distinct genetic clusters within the Fitzgerald River NP. The pedigree record of

the captive population provided evidence of interbreeding between individuals from

different genetic clusters. Based on the NEWHYBRIDS analysis, both captive and

translocated populations were dominated by pure-breds from the western lineage.

However, the eastern lineage may be underestimated, especially through contributions

71

from hybrids, which were more frequently identified by NEWHYBRIDS as western

lineage pure-breds than eastern lineage pure-breds. Nonetheless, the number of eastern

pure-breds increased in 2007 and 2009 when more individuals from that lineage were

introduced to the captive colony. This showed that the initial genetic proportions of newly

established populations are influenced by the number and origins of the introduced

individuals. In the Mexican wolf (Canis lupus baileyi), a manipulation of founders with

three different ancestries was used to reduce levels of inbreeding within a reintroduced

population (Hedrick et al., 1997). However, careful manipulations of this type are

vulnerable to differential mortality and/or reproductive success among founders (Biebach

and Keller, 2012, Raisin et al., 2012).

3.5.3 Genetic mixing and relatedness

We found that interbreeding of founders from different genetic clusters reduced genetic

relatedness among their progeny. This is not surprising given that dibblers from different

genetic clusters were less likely to share alleles that are identical by descent as a result of

isolation by distance (Wright, 1943). A similar finding was reported in farmed pearl

oysters (Pinctada margaritifera). By pooling individuals from genetically divergent

populations, it lowered the levels of pairwise relatedness when compared to the wild

populations (Lemer and Planes, 2012). However, the reduction was short-lived due to

limited mate availability and continued interbreeding within the new population. A

change in the dispersal behaviour of males may have also occurred as genetic relatedness

values between pairs of males was much higher than in either of the source populations.

This finding demonstrates that the initial genetic similarity between founding individuals

is important for the captive breeding and translocation programs. For example, a number

of studies have found that background inbreeding of founders leads to higher levels of

genetic relatedness and inbreeding depression (Swinnerton et al., 2004, Mitchell et al.,

2011, Ellegren, 1999).

72

3.5.4 Conservation implications

This study shows that a large number of founders and rapid population growth can reduce

gene diversity loss and maintain allelic richness in the translocated populations. Selecting

individuals from multiple source populations maximised allelic diversity and lowered the

genetic similarity between admixed individuals. For future captive breeding and

translocation in dibblers, we recommend an initial population size of at least 30

individuals and that individuals selected as founders should come from multiple

locations/regions within a habitat of the source population to maximise allelic diversity

(including alleles unique to different source populations) and to reduce inbreeding

(Schwartz and McKelvey, 2009). Despite no significant loss of genetic diversity in this

study, the translocated population still experienced a significant reduction in effective

population size. Therefore, population monitoring is essential to assess if the translocated

population shows a decline in levels of genetic variation and/or effective population size

overtime, in which case, a top-up release of animals may be necessary to prevent further

losses.

74


75

CHAPTER FOUR

Admixture between genetically diverged island populations

bolsters genetic diversity within a newly established island

population of the dibbler (Parantechinus apicalis), but does

not prevent subsequent loss of genetic variation

4.1 ABSTRACT

Using individuals from multiple source populations is one way to bolster genetic variation

and avoid inbreeding in newly established populations. However, mixing isolated

populations, especially those on islands, can lead to outbreeding depression and mating

preferences may limited interbreeding between source population lineages. In this study,

we investigated the genetic consequences of mixing individuals from two island

populations of dibbler (Parantechinus apicalis) in an island translocation. Despite a high

level of genetic divergence between the source populations (FST = 0.46), and significant

differences in body size, individuals with different source population ancestries were able

to successfully interbreed in captivity and in the wild with no obvious effects on

reproductive fitness. Genetic diversity within the translocated population was

significantly higher than one of the source populations. Although equal numbers of

individuals from each source population were used to establish the captive breeding

population, the genetic contribution from one source population was higher than the other,

due to the higher mating success of larger males. Nevertheless, the genetic contributions

from both source populations were maintained over multiple generations and levels of

genetic diversity were significantly higher in the translocated population than one of the

source populations. Estimates of the effective population size were very low in all

populations (< 23.7). All island populations exhibited significant fluctuations in allele

frequencies and lost genetic variation between 2006 and 2012. Population viability

analysis suggests a supplementation program using 30 animals every ten years is required

to prevent the decline in population size and maintain at least 90% of genetic variation in

the translocated population.

76

4.2 INTRODUCTION

Establishing new populations is an effective management tool for reducing the extinction

risk to endangered species restricted to a few remnant populations (Johnson et al., 2010,

Maguire and Lacy, 1990) or populations that are threatened by introduced predators or

disease (Ottewell et al., 2014, Huxtable et al., 2015). However, because translocations are

often based on small numbers of individuals, they are often prone to founder effects,

which can reduce genetic variation and lead to rapid genetic divergences from source

populations (Broders et al., 1999, Ramstad et al., 2013, Gautschi et al., 2002, Cardoso et

al., 2009). Survivorship differences among founders and their offspring may also reduce

the effective population size and subsequently exacerbate the loss of genetic diversity

(Jamieson, 2011, Biebach and Keller, 2012). Loss of genetic variation is of particular

concern because it reduces the evolutionary potential of the population and inbreeding in

small populations may lead to declines in fitness due to inbreeding depression, further

increasing extinction risks (Willi et al., 2006, Eldridge et al., 1999, Frankham, 1996). One

way to counterbalance the loss of genetic diversity when establishing new populations is

to use multiple source populations (Weeks et al., 2011). Indeed, several studies have

shown that new populations established using founders from multiple source populations

have higher genetic variation relative to one or more of the source populations (e.g.

Kennington et al., 2012, Huff et al., 2010, Ransler et al., 2011).

While there are clear advantages to using multiple sources populations when establishing

new populations, there are potential costs as well. Intrinsic (environment independent)

and extrinsic (environment dependent) incompatibilities between populations can reduce

fitness in hybrid and backcrossed offspring (Allendorf et al., 2001, Rhymer and

Simberloff, 1996, Lynch, 1991). In addition, differences in mating behaviour and mate

recognition may lead to pre-zygotic reproductive barriers between source populations

(Vines and Schluter, 2006, Rolán-Alvarez et al., 1999). There may also be survivorship

differences among founders from different source populations and their offspring due to

maladaptation to the release site (Brodie, 1992, Campbell and Waser, 2001) or post-

zygotic barriers (Arntzen et al., 2009, Álvarez and Garcia-Vazquez, 2011). All of these

factors can reduce the effective population size of a newly established population and

subsequently lead to loss of genetic variation and inbreeding (Frankham, 1995).

77

In this study, we investigate the genetic consequences of using multiple source

populations in an island translocation of the dibbler (Parantechinus apicalis), a small (40

– 100 g) primarily insectivore marsupial, which is endemic to the southwest of Australia

(Miller et al., 2003, Mills et al., 2004) and listed as Endangered under the Environment

Protection and Biodiversity Conservation Act (1999) and the 2014 IUCN Red List of

Threatened Species (Friend et al., 2008). Dibblers were once widely distributed in

Western Australia from Shark Bay on the central coast to Esperance on the southern

coastline and east to the Eyre Peninsula, South Australia (Friend, 2003, Baynes, 1990,

Baynes, 1987) (Figure 4.1). They now occur only in the Fitzgerald River National Park

and on two small islands off the coast from Jurien Bay (Morcombe, 1967, Fuller and

Burbidge, 1987). The islands, Boullanger and Whitlock Islands, have been separated from

the mainland for at least 500 years (Chalmers and Davies, 1984). They have a

polygynandrous mating system, with both males and females pairing with several mates

(Lambert and Mills, 2006). There is strong sexual dimorphism, with males being larger

than females (Mills and Bencini, 2000, Mills et al., 2004). Females breed once a year

during autumn (March to April). They produce as many as eight young per breeding

season. These young reach sexual maturity after 10 – 11 months (Woolley, 1995). In

captivity, P. apicalis live up to 3.5 years, but most don’t live past two years (Lambert and

Mills, 2006, Wolfe et al., 2004).

78

Figure 4.1 Past and present distribution of the Dibbler (Parantechinus apicalis) adapted

from Moro (2003). Past distribution is shown in the smaller map above. Present

distribution is shown in the larger map. Normal font represents natural populations and

italic font represents translocated populations.

To reduce the extinction risk to the species, a new back-up or ‘insurance’ population of

P. apicalis was established on Escape Island, which is situated three km offshore from

the town of Jurien Bay, close to Boullanger and Whitlock Islands (Figure 4.1). This

species was not known to occupy the island prior to the translocation (Lambert and Mills,

2006). In 1997, the captive-bred colony that was set up using four pairs of P. apicalis,

two pairs from Boullanger Island and two pairs from Whitlock Island in 1997 (Lambert

and Mills, 2006). An additional three males from Whitlock Island were introduced to the

captive breeding population in 1999. Pairing selection was based on the mean kinship

value. Males and females with the lowest mean kinship value were paired together. In

cases where a female showed aggression or no mating behaviours towards a selected

male, the next male on the list would be introduced until all females were allocated a

partner. Further details on husbandry and breeding of island P. apicalis are described in

Lambert and Mills (2006). A total of 33 adults and 55 sub-adults (51 females and 37

males) that were released on Escape Island in 1998 (N = 26), 1999 (N = 41), 2000 (N =

Boullanger Island

Whitlock Island

Escape Island

Peniup Nature Reserve Fitzgerald River National Park

Perth

Jurien

79

19) and 2001 (N = 2) (Moro, 2003). Eighty three of these animals were captive-bred and

five animals were wild-born (Lambert and Mills, 2006).

A previous study has shown there is substantial genetic differentiation between P.

apicalis populations on Boullanger and Whitlock Islands (Mills et al., 2004). In addition,

P. apicalis males on Boullanger Island are slightly heavier and larger than those on

Whitlock Island (Mills et al., 2004, Mills and Bencini, 2000). While the translocation

appears to have been successful (Moro, 2003), it is unclear how well the source

population lineages have introgressed and whether genetic variation within the population

has changed over time. The aim of this study is to investigate the genetic changes taking

place within the newly created island population between 1998 and 2012 (~ 15

generations). Our specific aims are to: i) determine the extent of mixing between the

source population lineages, ii) determine whether levels of genetic variation within the

translocated and source populations have changed over time, iii) investigate factors

influencing mating success in captivity, and iv) conduct a population viability analysis to

assess extinction probabilities and implications of various management options on levels

of genetic diversity within the newly established population.


4.3.1 Sampling and DNA extraction

All samples used in this study were obtained from the captive breeding program or were

collected during monitoring of P. apicalis populations on Escape (10.5 ha), Boullanger

(26 ha), and Whitlock (5.4 ha) Islands carried out by the Western Australian Department

of Parks and Wildlife. All wild-born and captive-bred dibblers had either an ear tissue or

hair sample taken and a microchip implanted. They were then measured, weighed and

their reproductive status recorded. All ear tissue and hair samples were stored in a 20%

DMSO2 solution saturated with NaCl at room temperature or 70% ethanol at –80 °C.

DNA from 65 ear notch samples was extracted using a salting-out method (Sunnucks and

Hales, 1996) with the following modification: DNA was incubated at 56 °C rather than

37 °C and 10 mg/mL instead of 0.1 mg/mL Proteinase K was added to 300 µL solution

of TNES. DNA from 14 hair samples (one from year 1997 and 13 from year 1999) was

extracted using BioBasic® ONE 4 ALL Genomic DNA Miniprep Kit, following the

protocol described in the instruction manual except DNA was eluted in 30 µL of Buffer

CE instead of 50 µL. DNA was extracted from a maximum of 30 samples per collection

80

year in each source population. In total, DNA was extracted from 76 samples from the

captive bred population (1997, N = 19; 1998, N = 11; 1999, N = 43; 2000, N = 3), 120

tissue samples from Boullanger Island (2006, N = 30; 2008, N = 30; 2010, N = 30; and

2012, N = 30), 53 from Whitlock Island (2006, N = 12; 2008, N = 17; 2010, N = 6; and

2012, N = 18), and 141 samples from Escape Island (2002, N = 15; 2003, N = 47; 2006,

N = 44; 2008, N = 7; and 2012, N = 24).

4.3.2 Microsatellite variation

Genotypes were determined at 14 microsatellite loci (3.1.2, 3.3.1, 3.3.2, 4.4.2, 4.4.10,

pPa2A12, pPa7A1, pPa1B1O, pPa2B1O, pPa4B3, pPa2D4, pDG1A1, Sh6e, and Aa4A)

that were characterized from P. apicalis and from closely related species (Table S4.1).

Polymerase chain reaction (PCR) was performed in a 10 µL reaction volume using the

QIAGEN Multiplex PCR Kit with primer concentrations ranging from 0.04 – 1.5 µM and

10 – 20 ng of DNA (Table S4.1). Amplifications were performed on an Eppendorf

Mastercycler epgradientS Thermocycler using the following parameters: 15 min at 95°C,

a total of 35 or 40 cycles of 30 s at 94°C, 90 s at the annealing temperature, 60 s at 72°C,

and concluding with 30 min at 60 °C (Table S4.1). PCR products were analysed on an

ABI 3730 sequencer using a GeneScan-600 LIZ internal size standard and scored using

GENEMARKER version 1.90 (SoftGenetics). Due to limited amount hair samples and

low quantity yield from DNA extraction, PCR per multiplex was carried out once per hair

sample. Genotyping success of hair samples was 54.6% and genotypes were screened for

unusual alleles. In addition, approximately 10% of tissue samples were re-amplified to

calculate genotyping error rates.

4.3.3 Data analysis

To assess genotype quality, we calculated the allele- and locus-specific genotypic error

rates (Pompanon et al., 2005). Individuals that failed in DNA extraction and with < 3

successful genotypes were excluded from further analysis (Captive, 1997 N = 2, 1999 N

= 1; Boullanger Island, 2006 N = 1; and Escape Island, 2002 N = 2, 2003 N = 4, 2006 N

= 10). We tested for the presence of null alleles in the source populations at each locus

using MICROCHECKER (Van Oosterhout et al., 2004). For these, and all subsequent

analyses, only population samples with at least 10 individuals were used. Estimates of

genetic variability within population samples were assessed by calculating the allelic

richness (an estimate of the number of alleles per locus corrected for sample size) and

gene diversity using the FSTAT software package (Goudet 2001). Deviations from

81

Hardy-Weinberg Equilibrium (HWE) were quantified using the inbreeding coefficient

(FIS) and tested for significance using randomization tests. Positive FIS values indicate

deficiency of heterozygotes, relative to random mating, whereas negative values indicate

a heterozygote excess. Genotypic disequilibrium between each pair of loci within each

population sample was assessed by testing the significance of association between

genotypes. Genetic differentiation between pairs of population samples was assessed

using Weir & Cockerham’s (1984) FST (θ). Estimates of FIS, pairwise FST values, tests for

differentiation among samples, deficits in heterozygotes, and genotypic disequilibrium

were calculated using the FSTAT software package (Goudet 2001). A sequential

Bonferroni correction (Rice, 1989) was applied to all tests to control for type I statistical

error. Differences in estimates of genetic variation and FIS values between population

samples were tested using Wilcoxon’s signed-rank tests with samples paired by locus

using the R version 3.0.1 statistical package (R Core Team, 2014).

We used the software package BOTTLENECK (Piry et al., 1999) to test for severe

reductions in effective population size (Ne). The tests were based on the principle that the

number of alleles decreases faster than the expected heterozygosity after a population

bottleneck. As a consequence, the expected heterozygosity should be higher than the

equilibrium heterozygosity predicted in a stable population from the observed number of

alleles (Maruyama & Fuerst 1985). Analyses were run using a two-phase model (TPM)

with 95% single-step mutation, 5% multiple-step mutations, and a variance of 12 among

multiple steps as recommended by Piry et al. (1999). A Wilcoxon signed rank test was

used to determine whether each site had an excess of heterozygosity (Piry et al., 1999).

Estimates of Ne were obtained from genotypic data by implementing the software package

LDNE (Waples and Do, 2008). The calculations were based on the assumption that

population samples consisted of overlapping generations. We used a random mating

model and estimated linkage disequilibrium amongst alleles using only alleles with

frequencies > 5% (Waples and Do, 2010). Estimates of the census population size were

based on the minimum number of animals known to be alive (Krebs, 1966) that were

recorded by trapping using Elliott traps during population monitoring on Escape Island

(2006, 2007, 2008, and 2012), Boullanger Island (2006, 2008, 2010, and 2012), and

Whitlock Island (2006, 2008, 2010 and 2012).

To investigate the extent of genetic mixing between source population lineages in the

translocated population on Escape Island, we carried out Bayesian clustering analysis

82

using the software package STRUCTURE 2.3.4 (Pritchard et al., 2000). The analyses

were performed with the source population locations set as prior information and the

number of genetic clusters (K) set to two because this was the number of source

populations used to establish the captive breeding and translocated populations. This was

confirmed by comparing the likelihood values for different values of K (1 – 10) and using

the ∆K method of Evanno et al. (2005b) to choose the most likely K (Table S4.2). In each

analysis, individuals were assigned a membership coefficient, which is the fraction of the

genome with membership to a particular cluster. Ten independent runs were performed

for each population sample using 100,000 iterations, with a burn-in period of 10,000

iterations. These parameter values yielded highly consistent results across independent

runs, indicating the number of iterations and burn-in period were sufficiently long (Table

S4.2).

Differences in body weight between captive-bred adult males and females and between

animals with different ancestries (N ≥ 4) were tested using Wilcoxon’s signed-rank tests.

A relationship between genetic proportions of Boullanger Island ancestry and body

weight was tested using Spearman's rank correlation. Differences in body size and weight

between wild-born adults on different islands were tested for each sex using Kruskal-

Wallis rank sum tests and then Wilcoxon’s signed-rank tests for Post-hoc analysis. The

differences between sexes were tested using Wilcoxon’s signed-rank tests.

We carried out a generalized linear model (GLM) to investigate factors influencing

mating success and reproductive outcomes (number of offspring and proportion of

offspring surviving). These factors included parental weight and age, the relatedness

between the mating pair (scored as 1 or 0 for the same or different source population,

respectively), and genetic background of offspring produced (Whitlock Island pure-bred

= 0, Whitlock Island pure-bred × F1 = 0.25, F1 = 0.5, Boullanger Island pure-bred × F1 =

0.75 and Boullanger Island pure-bred = 1). These analyses were carried out using R

package version 3.0.1 (R Core Team, 2014).

Population viability analyses of the island populations were conducted using the software

package VORTEX version 10.0 (Lacy and Pollak, 2014). Demographic parameters for

the simulation models were taken from previously published studies of P. apicalis from

the captive and Escape Island populations (Lambert and Mills, 2006, Moro, 2003, Mills

and Bencini, 2000, Woolley, 1991, Mills et al., 2004) and unpublished data from the

captive-bred colony at Perth Zoo (four years data, C. Lambert). We modelled genetic

83

diversity change using the allele frequency estimates for each population from the FSTAT

software package (Goudet 2001). We used all loci in the summary statistic. The maximum

capacity was set to the maximum Known To Be Alive (KTBA) estimate for each island

population (2006 – 2012 data, T. Friend). The models were run for 500 years using the

exponential population growth model with 1000 iterations. We also ran simulations of

multiple supplementation events (20%/20/30 animals every 5/10 years) to determine the

number of additional individuals needed to preserve > 90% of the initial gene diversity. I

also tested the robustness of the results by changing different parameters such as the

mortality rate of age 0 to 1 (from 29.4% to 54.7%, 54.7% is a combined mortality of 0 –

1 year old animals in captive-bred and wild-born dibblers on Escape Island). This model

predicted all populations would go extinct within 200 years. However, the mortality

estimate of wild-born Escape Island dibblers can also be affected by the recent

translocation event. Therefore, I only reported the results using the mortality estimate of

age 0 – 1 year from the captive population. Detailed descriptions of the demographic

parameters, management scenarios and assumptions made in the models are provided in

Table 4.1 and S4.4 (Appendices).

84

Table 4.1 Demographic and life history parameters used in population viability models

of Parantechinus apicalis translocated population at Escape Island and the sources of

data used. Full details and justification of the parameters used is provided in Table S4.4.

Parameter Dibbler

Breeding system Polygynous1A

Inbreeding Depression Recessive Lethals (8 Lethal

equivalents)2A

Adult males in breeding pool 84.3 (calculated from % male success in

breeding)

% males successful in breeding 61.91AU

Mean no. mates per male -

Age of first reproduction (Females) 1 (8-9 months3A)

Age of first reproduction (Males) 1 (10 months4A)

Max. age of reproduction 34A

No. litters/year 11A

Sex ratio at birth (in % males) 41.1%1A

Max no. progeny/litter 81A

% Adult females producing 90%5A

Mean no. of young/litter 6.2±0.2 (Whitlock6A), 7.4±0.1

(Boullanger6A), 7.0±1.1(Escape5A)

Mortality of females and males

0-1 years of age 54.7% 5A,1AU

>1 years of age 35%5A

Population carrying capacity (K) 47±157A

Dispersal between pops None, closed pop

Initial population size 885A

Years modelled 100

No. of iterations 1000

1A Lambert and Mills 2006 4A Mills and Bencini 2000 2A O’Grady et al. 2006 5A Moro 2003

1AU Lambert and Mills 2006 unpublished data 6A Mills et al. 2004 3A Woolley 1991 7A Assumption made – see Table S4.4 for rational

85

4.4 RESULTS

4.4.1 Genetic variation within populations

Across the genetic dataset, we had a successful amplification rate of 0.921 per locus. The

allele-specific and locus-specific genotypic error rates were 0.004 and 0.009 respectively.

Three loci were identified as having null alleles (3.3.2, pPa7A1, and Sh6e) within

population samples. Because there was no consistent pattern in the presence of null alleles

(i.e. the loci with null alleles varied among samples from the same population and

between populations), all loci were retained for further analysis. Overall estimates of

genetic diversity in the captive and translocated populations were higher than the source

populations (Table 4.2). Estimates of gene diversity and allelic richness were significantly

higher in the translocated population than the Whitlock Island source population in 18 out

of 20 pairwise tests (Wilcoxon rank sum tests, P < 0.05). However, there were no

significant differences in gene diversity and allelic richness between the translocated and

Boullanger Island populations in all comparisons. Significant differences in allelic

richness were also observed between collection years in the translocated population (χ2 =

10.8, P = 0.013) and in both allelic richness and gene diversity in the Whitlock Island

source population (allelic richness: χ2 = 7.4, P < 0.001; gene diversity: χ2 = 6.0, P =

0.014). In both populations, gene diversity and allelic richness were lower in the most

recent collections (Table 4.2). There was also a trend of decreasing genetic variation in

the Boullanger Island source population, but the differences between collection years

were not significant (allelic richness: χ2 = 6.8, P = 0.079 and gene diversity: χ2 = 2.7, P =

0.434).

Multilocus FIS values fluctuated between years both in the source and the translocated

populations, but were significantly different from zero in only a few cases (Table 4.2).

The Whitlock population had the highest overall multilocus FIS value and was

significantly different from zero (randomization tests, P < 0.013). The number of pairs in

loci in genotypic disequilibrium (GD) ranged from one to five, with the highest level

occurring in 1998 samples from the captive breeding population. The number of pairs of

loci in GD in the source population samples ranged from zero to one (Table 4.2).

86

Table 4.2 Genetic variation and results of bottleneck testing within the source, captive

and translocated populations of P. apicalis. N is the mean number of genotypes per locus,

FIS is the inbreeding coefficient, H is gene diversity, AR is allelic richness, GD is the

number of pairs of loci in genotypic disequilibrium. Standard errors are given after mean

values. FIS values significantly greater than zero (P < 0.05) are highlighted in bold text.

Sample N H AR FIS GD Bottleneck

Source population: Boullanger Island

2006 26.3±0.8 0.40 ± 0.05 2.4 ± 0.2 0.06 1 N

2008 29.6±0.2 0.35 ± 0.06 2.1 ± 0.2 –0.07 1 Y

2010 29.2±0.3 0.34 ± 0.06 2.1 ± 0.2 0.08 1 Y

2012 30.0±0.0 0.33 ± 0.07 2.0 ± 0.2 –0.01 1 Y

Overall 115.1±0.9 0.37 ± 0.06 2.2 ± 0.2 0.05 4 N

Source population: Whitlock Island

2006 9.4±0.6 0.18 ± 0.07 – 0.45 0 N

2008 16.3±0.3 0.15 ± 0.06 1.5 ± 0.2 0.01 0 N

2012 18.0±0.0 0.06 ± 0.03 1.2 ± 0.1 0.25 0 Y

Overall 48.6±0.8 0.13 ± 0.05 1.5 ± 0.2 0.28 0 N

Captive population

1997 15.9 ± 0.2 0.40 ± 0.06 2.4 ± 0.3 0.17 2 N

1998 6.8 ± 1.0 0.40 ± 0.08 – –0.04 0 Y

1999 42.0 ± 0.0 0.39 ± 0.05 2.2 ± 0.3 –0.06 5 N

Overall 67.7 ± 1.2 0.40 ± 0.05 2.3 ± 0.3 –0.02 9 N

Translocated population: Escape Island

2002 11.0 ± 0.4 0.41 ± 0.05 2.3 ± 0.2 –0.02 0 Y

2003 40.9 ± 0.7 0.41 ± 0.05 2.2 ± 0.2 0.05 3 N

2006 24.6 ± 1.5 0.44 ± 0.05 2.5 ± 0.3 0.26 0 N

2012 23.8 ± 0.2 0.38 ± 0.06 2.0 ± 0.2 0.04 1 Y

Overall 106.6 ± 2.3 0.42 ± 0.05 2.3 ± 0.2 0.11 6 N

87

4.4.2 Population bottlenecks and estimates of Ne

In general, all population samples were comprised of a high proportion of rare alleles (<

10% frequency) and had L-shaped allelic distributions indicating they were under

mutation-drift equilibrium. However, a mode shift consistent with a recent population

bottleneck was found in some collection year samples, especially in 2012 in all three

island populations (Wilcoxon sign test, P < 0.05) (Table 4.2).

Estimates of Ne were very low and, as expected, were much lower than census size

estimates (Nc) for all populations (Table 4.3). Overall Ne of the translocated population

was slightly lower than the Boullanger Island source population, but it was higher than

the Whitlock Island source population. While Nc of the source populations fluctuated

largely between years, Ne remained relatively stable. Ne of the translocated population

showed larger yearly fluctuations than the source populations. The Ne/Nc ratios ranged

from 0.03 to 0.06 for Whitlock Island, 0.10 to 0.31 for Boullanger Island, and 0.34 to 0.54

for the translocated population (Table 4.3).

88

Table 4.3 Estimates of the census (Nc) and effective population size (Ne) of P. apicalis in

the source and translocated populations. The overall estimate of Nc for each population is

based on the harmonic mean of the yearly estimates. NA is not available.

Ne

Sample Nc Estimate 95% CL Ne/Nc ratio

Source population: Boullanger Island

2008 63 9.5 2.9 – 26.8 0.15

2010 37 9.6 2.8 – 32.0 0.26

2012 68 6.8 2.2 – 21.5 0.10

Overall 52.1 16.2 5.8 – 36.6 0.31

Source population: Whitlock Island

2008 42 1.3 0.4 – 7.9 0.03

2012 29 1.3 0.1 – 223.0 0.04

Overall 34.3 2.2 0.5 – 11.3 0.06

Translocated population: Escape Island

2002 NA 11.5 2.7 – infinite NA

2003 NA 4.7 2.1 – 15.0 NA

2006 47 23.7 5.0 – infinite 0.50

2012 26 14.1 2.5 – infinite 0.54

Overall 33.5 11.3 3.3 – 26.1 0.34

4.4.3 Population structure and genetic mixing within the translocated population

Pairwise population FST indicated a substantial differentiation in allele frequencies

between the source populations (Table 4.4). There were also significant divergences

between the translocated and source populations and between population samples taken

on different years from the translocated population. Overall, FST values indicated the

captive breeding population and earlier collection years of the translocated population

were most similar to the Boullanger Island population. However, in the last two collection

years, 2006 and 2012, the translocated population had similar FST values when compared

to both source populations (Table 4.4).

89

Table 4.4 Pairwise FST estimates between source, captive and translocated populations of P. apicalis. FST estimates significantly greater than zero (P <

0.05) after correction for multiple comparisons are highlighted in bold text.

Boullanger Island Whitlock Island Captive breeding Translocated

Population Year 2008 2010 2012 2006 2008 2012 1997 1998 1999 2002 2003 2006 2012

Boullanger

Island

2006 0.060 0.056 0.102 0.325 0.373 0.471 0.083 0.037 0.098 0.074 0.084 0.161 0.168

2008 0.053 0.050 0.444 0.470 0.550 0.213 0.139 0.188 0.189 0.172 0.265 0.264

2010 0.021 0.392 0.415 0.502 0.192 0.134 0.200 0.192 0.173 0.268 0.260

2012 0.449 0.469 0.550 0.238 0.194 0.215 0.224 0.189 0.284 0.264

Whitlock

Island

2006 0.042 0.321 0.189 0.218 0.288 0.341 0.264 0.218 0.260

2008 0.172 0.239 0.286 0.319 0.407 0.292 0.263 0.295

2012 0.373 0.473 0.401 0.538 0.365 0.331 0.382

Captive

breeding

1997 -0.002 0.030 0.022 0.014 0.057 0.069

1998 -0.019 -0.016 -0.025 0.033 0.024

1999 0.001 0.004 0.073 0.073

Translocated 2002 -0.010 0.052 0.081

2003 0.046 0.049

2006 0.039

90

The clustering analyses also revealed clear genetic differences between the source

populations and changes in the genetic composition of the translocated population over

time. Individuals from Boullanger Island were predominantly assigned to one genetic

cluster and those from Whitlock Island were assigned to the other (Figure 4.3). Most

individuals from the captive breeding and translocated populations had membership to

both genetic clusters, indicating they had mixed ancestry (Figure 4.3). Overall,

individuals within the captive breeding and translocated population had a higher

proportion of membership to the genetic cluster associated with Boullanger Island, though

this pattern changed over time as the proportion of membership between the two genetic

clusters became more even in the translocated population (Figure 4.3).

The dominance of the genetic cluster associated with Boullanger Island in the captive

breeding and translocated populations was consistent with the pedigree record. Across a

total of 88 dibblers born in captivity, 85.2% of them had a strong Boullanger Island

ancestry (pure-bred Boullanger Island and 1st or 2nd generation backcross to Boullanger

Island). The remaining 14.8% of individuals were F1 hybrids or backcrosses to Whitlock

Island.

Figure 4.2 Summary of the clustering analysis on the source, captive breeding, and

translocated populations assuming two admixed populations (K = 2). Each individual is

represented by a bar showing the individual’s estimated membership to a particular

cluster (represent by different colours). Black lines separate samples collected over

different years from each of the populations.

Boullanger Whitlock 1997 1999 2002 2003 2006 2012

Island Island1998 2000 2008

1.0

0.8

0.6

0.4

0.2

0.0

Source populations Captive population Translocated population

91

4.4.4 Differences in body size between source populations and factors influencing

mating and reproductive success in captivity

There were significant differences between males and females with different genetic

backgrounds. Captive pure-bred males with a Boullanger Island background were

significantly heavier than females (males: 76.3 ± 3.5 g; females: 58.5 ± 1.0 g, Figure 4.3).

A significant difference was also found between females with Boullanger Island

background and those with hybrid ancestry, with Boullanger Island pure-breds being

heavier than hybrid females (W = 60, P = 0.003). In addition, there were strong positive

correlations between genetic proportion of Boullanger Island ancestry and body weight

of captive-born dibblers in both sexes (females: rho = 0.738, S = 403.0, P < 0.001; males:

rho = 0.797, S = 11.4, P = 0.032, Figure 4.3). On the islands, males were significantly

larger than females (Wilcoxon’s signed-rank tests, P < 0.05 in all cases, Figure 4.4).

Males on Boullanger Island were significantly larger and heavier than males on Whitlock

Island (body weight: W = 303.5, P = 0.001; head length: W = 274, P = 0.007; pes length:

W = 338, P < 0.001; pes short: W = 313.5, P < 0.001), while females only exhibited pes

size differences (pes length: W = 502, P = 0.005; pes short: W = 492.5, P = 0.008). Males

from Escape Island were heavier and had longer pes length than males on Whitlock Island

(body weight: W = 507, P = 0.005; pes length: W = 496.5, P < 0.001; pes short: W =

636.5, P < 0.001). However, they had shorter head length and pes length than males on

Boullanger Island (head length: W = 376.5, P < 0.001; pes length: W = 327.5, P = 0.003).

Body size and weight of females on Escape Island were not significantly different from

females on Boullanger and Whitlock Islands except for pes length (Wilcoxon’s signed-

rank tests, P > 0.05 in all cases except one, Figure 4.4). Females on Boullanger Island

had longer pes length than Escape Island females (W = 572.5, P = 0.021).

92

Figure 4.3 The relationship between Boullanger Island ancestral genetic proportions

calculated from the pedigree and individual body weight of adult male (shaded) and

female (unshaded) P. apicalis born in captivity.

Figure 4.4 Mean adult body size and weight (± standard error) of male (full symbols)

and female (open symbols) P. apicalis captured on Boullanger, Whitlock, and Escape

Islands between 2006 and 2013.

0

20

40

60

80

100

0 0.2 0.4 0.6 0.8 1

Body w

eight

(g)

Boullanger Island ancestry genetic proprotion

36

37

38

39

40

41

Hea

d l

eng

th (

mm

)

19.5

20.0

20.5

21.0

21.5

22.0

22.5

Boullanger

Island

Whitlock

Island

Escape

Island

Pes

len

gth

(m

m)

14.5

15.0

15.5

16.0

16.5

17.0

Boullanger

Island

Whitlock

Island

Escape

Island

Pes

sh

ort

(m

m)

40

45

50

55

60

65

Wei

gh

t (g

)

93

The generalized linear models revealed that mating success in captivity was associated

with male weight and male age with the model coefficients indicating that the younger,

larger males had higher mating success (Table 4.5). No other factor had an effect on

mating success and none of the factors examined, including whether or not the offspring

had mixed ancestry (admixed), were significantly associated with the number of offspring

or the proportion of offspring that survived (Table 4.5).

Table 4.5 Results of generalized linear models investigating factors influencing mating

and reproductive success of P. apicalis in captivity. The table shows coefficient values

(b), standard errors (SE), t-values and significant P-values. Significant factors are

highlighted in bold text.

Factor b SE t P

Mating success

Female weight 0.071 0.077 0.917 0.367

Female age –1.031 1.147 –0.899 0.376

Male weight 0.179 0.083 2.151 0.040

Male age –2.156 0.970 –2.221 0.034

Relatedness 0.105 2.652 0.039 0.969

Number of offspring

Female weight 0.006 0.015 0.378 0.719

Female age 0.012 0.129 0.093 0.929

Male weight 0.003 0.009 0.315 0.763

Male age –0.133 0.141 –0.946 0.381

Relatedness 0.038 0.710 0.053 0.959

Admixed offspring 0.167 0.311 0.538 0.610

Proportion of offspring surviving

Female weight 0.313 0.184 1.703 0.139

Female age –1.019 1.784 –0.571 0.589

Male weight 0.180 0.174 1.030 0.343

Male age 1.739 2.597 0.669 0.528

Relatedness –8.902 10.611 –0.839 0.434

Admixed offspring –0.046 4.033 –0.011 0.991

94

4.4.5 Population viability analysis

Due to the low genetic diversity and constricted population size of the island populations,

the PVA model constructed from island- and species-specific demographic rates (Table

4.1) predicted that genetic diversity will decline rapidly over time and that all island

populations will go extinct within 400 years (Figure 4.5). The translocated population is

expected to be extinct in 186 years, while the Boullanger and Whitlock Island populations

are predicted to persist for only 380 and 147 years. An augmentation strategy of 20%

supplementation of the recipient population every five and ten years was not sufficient to

preserve 90% of original genetic variation. There were no significant differences between

releasing 20 individuals at five-year interval and 30 individuals at 10-year interval as

shown in Figure 4.5. However, I recommended 30 individuals every 10 years because

this supplementation regime predicted to maintain genetic diversity closer to 90% of the

initial level after translocation. In addition, it is more cost sufficient to breed 30 animals

every 10 years than 20 animals at five-year interval. The effect of supplementation was

observed to boost the populations’ genetic variation and lasted up to four generations.

95

Figure 4.5 Predicted changes in the population size (N), gene diversity (H), and number

of alleles (A) of P. apicalis on Boullanger, Whitlock and Escape Islands with and without

different supplementation strategies (20 and 30 animals/five years and 30 animals/ten

years) for the next 500 years using based on population viability analysis with input

parameters as described in Table 4.1 and S4.3.

4.5 DISCUSSION

4.5.1 Phenotype and genetic differentiation between island populations

We found substantial genetic differentiation between P. apicalis populations on

Boullanger and Whitlock Islands, as well as significant differences in body weight and

size between males on the islands. These genetic divergences likely reflect the isolation

between the island populations ( < 500 years, Chalmers and Davies, 1984). The observed

differences in male body size could be a consequence of various factors. Directional

selection on male body mass in closely related species, the agile antechinus (Antechinus

agilis) and brown antechinus (A. stuartii) has been attributed to sexual selection (Holleley

et al., 2006, Kraaijeveld-Smit et al., 2003). Environmental variables such as island size

have also been found to positively correlate to body size in many species (Lomolino,

2005, McNab, 2002). Boullanger Island is much larger than Whitlock Island (26 ha vs

96

5.4 ha). Male dispersal distance could be correlated to the island size (Jenkins et al., 2007)

favouring larger males (Lomolino, 2005). Competition for food may also be important

(Kraaijeveld-Smit et al., 2003), though resource abundance (invertebrate prey and seabird

burrows used for shelter) is higher on Whitlock Island than Boullanger Island (Wolfe et

al., 2004), which is inconsistent with this explanation (Miller et al 2003).

It is noteworthy that not all individuals with equal genetic makeup from both ancestries

were intermediate, especially after few generations of recombination. In Escape Island P.

apicalis, only body weight and pes length were intermediate of the ancestry populations,

but not head length and pes short. Males on Escape Island had much smaller head length

than males on Boullanger Island and much longer pes short than males on Whitlock

Island. Such deviations could be driven by factors such as location adaptation and

epistatic interaction. Changes in phenotypes as a result of local adaptation has been

reported previously following hybridization between grey wolves (Canis lupus) and

coyotes (C. latrans), which resulted in larger skull size and body size that enhanced

hunting ability (Kays et al., 2010). In an intertidal snail (Bembicium vittatum)

translocation, variation in shell shape changed toward the phenotype that was most suited

to the local environment (Binks et al., 2007). Epistatic interaction has shown to vary

survival rates of individuals with different genetic background. For example, partial F1

hybrid viability and phenotypic effects between green sunfish (Lepomis cyanellus) and

longear sunfish (L. megalotis) were attributed to Dobzhansky-Muller incompatibilities at

several loci (López-Fernández and Bolnick, 2007). Maternal effect had also been

observed to result in offspring having body size similar to the maternal ancestry in a

reciprocal cross study between giant and normal size ninespine sticklebacks (Pungitius

pungitius, Ab Ghani et al., 2012).

Despite the genetic and morphological divergences between P. apicalis populations on

Boullanger and Whitlock Islands, individuals with mixed ancestry were successfully bred

in captivity, and there was a high proportion of individuals with mixed ancestry in the

captive and translocated populations across multiple generations. Analysis of the captive

breeding pedigree record also showed that the genetic similarity between mating pairs,

and whether or not the offspring produced had mixed ancestry, had no significant effect

on the number of offspring produced or the proportion of offspring that survived in

captivity. However, this could also reflect the lack of statistical power.

97

4.5.2 Genetic composition and influence of male body size on reproductive

success

Overall, the genetic composition of the captive population was more similar to the

Boullanger Island source population. This was despite the captive breeding population

being established with equal numbers of males and females from each of the source

populations and additional three males from Whitlock Island, though one of the females

from Whitlock Island died soon after being moved to captivity. Our analyses of mating

success suggest that the bias towards the Boullanger Island lineage partly reflects a mating

advantage of larger males. Larger, younger males were more likely to achieve a successful

mating than smaller, older males when paired with a female in captivity. These results are

consistent with a behavioral study of island P. apicalis, which showed that males with a

large body size or are younger had a reproductive advantage during courtship (Wolfe et

al., 2000). Previous studies have shown that body size is a major contributor to male

mating success in marsupials where male-male competition is important (Fisher and Lara,

1999, Clinchy et al., 2004, Miller et al., 2010a). In wild populations of A. agilis and A.

stuartii, large males were more successful in gaining mating access and at maintaining

intromission despite aggression from other males and resistance from females (Holleley

et al., 2006, Kraaijeveld-Smit et al., 2003).

Our analysis on mating success in captivity suggests that males with a Boullanger Island

ancestry should have a reproductive advantage over males with a Whitlock Island

ancestry due to their larger size. Therefore, we expected that the genetic composition of

the translocated population would become more similar to the Boullanger Island

population over time. Instead, the translocated population became evenly admixed by

both source population ancestries. This suggests factors other than mating success are

important or that factors other than body size determine male mating success in the wild

populations. Indeed, studies on other small marsupial species have shown that there are

other factors influencing mating success. For example, A. agilis male’s mating success is

influenced by mating order, the timing of mating and genetic compatibility (Kraaijeveld-

Smit et al., 2002b). A. agilis females, who can store sperm up to two weeks prior to

ovulation, do not refuse mating with a second, smaller male, which have a higher chance

of a successful fertilization due to mating order (Kraaijeveld-Smit et al., 2002b). In

another study, A. agilis females preferred genetically dissimilar males whereas males

mated readily with most females (Parrott et al., 2015). In absence of good quality or large

98

males, females may mate with smaller males or multiple males to ensure that enough

spermatozoa will be present at time of ovulation (Kraaijeveld-Smit et al., 2002a).

4.5.3 Consequences of genetic mixing on genetic diversity

Previous studies of translocations involving single source population had shown that

establishing new populations can lead to loss of genetic diversity, increased inbreeding,

and divergence from ancestral alleles frequencies (Cardoso et al., 2009, Jamieson, 2011,

Sigg et al., 2005). However, this study showed that the translocated population had higher

genetic variation than the source populations, which is consistent with other studies on

translocations involving multiple source populations (Kennington et al., 2012, Ransler et

al., 2011, Stockwell et al., 1996). The increased level of genetic variation in the

translocated population was not much greater than the most variable source population,

Boullanger Island. Similarly, Huff et al. (2010) found all reintroduced populations of

slimy sculpins (Cottus cognatus) exhibited higher levels of genetic diversity, but the

increases were only slightly higher than the single most genetically diverse source

population. The increase in genetic variation we found in this study is not unexpected

given that these island populations carried different subsets of alleles (Table S4.5) as a

result of lack of gene flow, genetic drift, and local selection (Eldridge et al., 1999).

All populations had low estimates of Ne (range 2.2 to 16.2), which is likely to reflect the

carrying capacity of the islands. Ne/Nc ratios (0.1 – 0.26) in the Boullanger island

population were consistent with those found in wild populations (Palstra and Ruzzante,

2008, Frankham, 1995). However, low Ne/Nc ratios (0.03 – 0.04) of on Whitlock Island

indicate that this population is more sensitive to genetic stochasticity than the Boullanger

Island population (Palstra and Ruzzante, 2008). Interestingly, Ne/Nc ratio (0.50 – 0.54) of

the Escape Island population was higher than both source populations. Frankham (1995)

identified fluctuating population size, variance in reproductive success, and unequal sex

ratio as the main factors affecting Ne/Nc ratios. In P. apicalis, facultative male die-off is

associated with nutrient inputs from seabirds (Wolfe et al., 2004). Seasonal variation of

nutrient inputs from seabirds may affect the abundance of invertebrates on these islands

and cause a fluctuation in population size contributing to the difference in Ne/Nc ratios

(Miller et al., 2003, Wolfe et al., 2004). However, if this is the case we would expect a

higher Ne/Nc ratios in the Whitlock and Escape Island populations. The pedigree record

also show high variance in male reproductive success between mating pairs in captivity,

suggesting that this could be the main contributing factor of the island populations.

99

However, higher Ne/Nc ratios of the Escape Island population and evenly mixed ancestry

proportions suggested that reproductive variance maybe lower in this population (see

Álvarez et al., 2015). The Ne/Nc ratios of island P. apicalis are comparable to other

marsupial species (the northern hairy-nosed wombat, Lasiorhinus krefftii, 0.18 and 0.59,

Taylor et. al. 1994; the eastern barred bandicoot, Perameles gunnii, 0.135, Sherwin and

Brown 1990; mountain pygmy possum, Burramys parvus, 0.62, Mitrovski et al. 2008).

The low estimates of Ne suggest that all populations are vulnerable to high levels of

genetic drift. Consistent with this expectation, both source and translocated populations

exhibit significant fluctuations in allele frequencies between collection years. In addition,

there was evidence of genetic bottlenecks, as well as significant declines in genetic

diversity and positive FIS values in the source and translocated populations over the

course of the study. Further loss of genetic diversity and elevated inbreeding is likely in

the years to come unless effective population sizes increase. Breeding designs such as

equalization of family size can increase Ne by reducing the variance of reproductive

success among individuals (but see Ryman and Laikre, 1991). This has been previously

reported in wild Atlantic salmon (Salmo salar, Saura et al. 2008, Perrier et al. 2014) and

cultured silver-lipped pearl oysters (Pinctada maxima, Lind et al. 2010) . However, this

approach also removes selection on fecundity and may negatively affect survivorship in

the wild (Lind et al., 2010, Trevarrow and Robison, 2009). Higher Ne estimates in the

translocated population could be because the initial Ne of the source populations was very

low. Benign captive environments could reduce population size fluctuations and lower

the reproductive success variance which may result in the increased Ne in the translocated

population. Alternatively, sourcing founders from multiple populations may dilute

inbreeding and increase individual reproductive success (Slate et al., 2000, Grueber et al.,

2010).

100

4.5.4 Management implications

Newly established populations are prone to loss of genetic variation, genetic drift and

inbreeding (Frankham, 1995). Our study shows that using multiple source populations

can increase genetic diversity within a newly established population of P. apicalis on an

offshore island. However, due to low effective population sizes and lack of gene flow

between populations, genetic diversity within the translocated population is predicted to

decline over time unless there is some types of intervention. A supplementation program

using 30 captive animals every ten years is predicted to prevent the decline in population

size and maintain at least 90% of genetic variation in all island populations. For future

captive breeding and translocation programs, we recommend targeting large males from

both island populations, pairing females with multiple males (≥ 3 males) in captivity and

that supplementation programs are used for all island populations to ensure their long-

term persistence.

102

Photo credit: Judy Dunlop

103

CHAPTER FIVE

Asymmetrical introgression between genetically distinct

populations of the burrowing bettong (Bettongia lesueur) in

a newly established translocated population

5.1 ABSTRACT

Translocations involving multiple source populations provide a way to maximise genetic

variation and reverse/avoid inbreeding depression. According to the IUCN, these

populations must be from the closest race or type. However, access to such populations

is not always possible, especially in rare Australian mammals that often only persist on

isolated offshore islands. The effects of mixing such populations and how readily they

interbreed remain largely unknown. Here, we investigate the genetic consequences of

mixing two isolated, island populations of boodies (Bettongia lesueur) used in a

translocation to mainland Australia. As expected, we found high levels of divergence

between the two source populations (FST = 0.42 and ϕST = 0.72 for nuclear and

mitochondrial DNA respectively) and higher levels of genetic variation in the

translocated population relative to one, but not both source populations. Despite clear

differences in body size, we found evidence of reciprocal interbreeding between the two

source population lineages. However, there was a bias towards crosses between males

from the smaller-sized Barrow Island source population and females from the larger-sized

Dryandra source population, which were originally from Dorre Island. The basis for the

asymmetrical mixing is unclear as it opposes the common trend of male-male competition

or female mate choice favouring larger dominant males. Our study shows that, given the

opportunity, boodies from highly diverged populations readily interbreed, without any

apparent loss to reproductive capacity or survivorship. However, further genetic

monitoring is required to assess whether intrinsic incompatibilities are causing the

asymmetrical introgression and whether evidence of fitness declines occur in subsequent

generations.

104

5.2 INTRODUCTION

Species used in translocation are often threatened or rare. Many have isolated populations

that are subjected to loss of genetic variation, high levels of inbreeding and elevated risks

of extinction (Wilcox, 2003, Larson et al., 2002, Sheean et al., 2012). While it has been

argued that any threatened populations with unique characteristics or distinct evolutionary

history should be conserved separately (Wayne et al., 1994), low levels of genetic

variation can pose a considerable extinction threat to populations (Bijlsma et al., 2000).

Furthermore, an isolated population may accumulate deleterious mutations and have low

evolutionary potential in changing environments (Moritz, 1999). Using individuals from

these populations in translocation programmes may reduce a probability of population

establishment and increase a risk of programme failure.

Mixing individuals from different source populations is one way to bolster genetic

variation and avoid inbreeding in these species (Ransler et al., 2011, Witzenberger and

Hochkirch, 2008, Hedrick, 1995). Hybridization between diverged populations reverses

deleterious effects of inbreeding by masking deleterious recessives (dominance) or

increasing heterozygosity at loci where heterozygotes have a selective advantage (over-

dominance) (Edmands and Timmerman, 2003). A well-known example is the genetic

restoration of the Florida panther (Puma concolor coryi). The introduction of Texas

panthers (P. concolor stanleyana) from a geographically nearby population increased

genetic diversity, reduced inbreeding, improved survival and fitness, and tripled the

number of panthers (Johnson et al., 2010, Pimm et al., 2006, Hostetler et al., 2010).

Further, long-isolated populations often carry different subsets of alleles as a result of

lack of gene flow, genetic drift, and local selection (Eldridge et al., 1999). Interbreeding

between individuals from these populations should increase evolutionary potential and

enable their offspring to survive in a wider range of environment (see Thompson et al.,

2010, Taylor et al., 2006). Consistent with this expectation, many translocations sourcing

from multiple source populations have shown an increase in genetic diversity over

multiple generations (Miller et al., 2012, Adams et al., 2011, Kennington et al., 2012,

Ransler et al., 2011, Binks et al., 2007).

While mixing differentiated populations can have favourable outcomes, it can also lead

to fitness reductions in the progeny (i.e. outbreeding depression) due to post-zygotic

isolation between source populations (Edmands, 2007, Tallmon et al., 2004, Allendorf et

al., 2001). Crossing between phenotypically different parents can produce offspring with

105

phenotypes that are unsuitable for the local environment. For example, a hybridisation of

two garter snakes populations (Thamnophis ordinoides) produced a mismatch in body

pattern and behaviour in hybrid snakes that had a higher mortality from predation in

comparison to purebred snakes (Brodie, 1992). Progeny may be unviable because of

abnormal structure and/or number of chromosomes (Fishman and Willis, 2001) or fitness

of progeny may be lower due to heterozygote disadvantage, harmful epistatic interaction

between alleles of the parents, or disrupting of co-adapted gene complexes (Charlesworth

and Willis, 2009). Common signs of intrinsic incompatibility include reduction in fertility

and viability of hybrid offspring such as sterility (Fishman and Willis, 2001), low survival

rate (Gharrett et al., 1999), slow growth rate (Huff et al., 2011), and decreased

reproductive success (Lancaster et al., 2007). In addition, pre-zygotic isolation such as

differences in morphology, behaviour, ecology, reproductive biology and gametic

compatibility, may prevent individuals from different source populations from

interbreeding (Alexandrino et al., 2005, Latch et al., 2006, Coyne and Orr, 2004). This

could reduce the effective population size or result in an uneven genetic contribution from

the source to the translocated population and induce genetic problems associated with a

small population size.

Predicting whether outbreeding depression will occur is difficult. Generally, the risk of

outbreeding depression becomes higher as the genetic distance between the parents

becomes greater, but the amount of divergence required for it to occur varies from species

to species (Edmands, 2002, Lynch, 1991). The different possible outcomes of mixing

diverged populations leave many conservation managers with a difficult decision when

choosing populations for use in translocations. This decision can affect an outcome of the

translocation and long-term persistence of the population. Allendorf et al. (2001) and

Edmands (2007) suggested that augmenting gene flow between fragmented populations

should only be carried out if the populations have lost substantial genetic variation and

the effects of inbreeding depression are apparent. However, such information is often not

available for populations of immediate conservation concern, and while awaiting for data

on the effects of inbreeding to be collected, this may put populations at risk of expirations.

Weeks et al. (2011) argued that by overestimating the risk of outcrossing breeding

depression, rational use of gene flow for genetic rescue is unnecessarily prevented. So

far, a meta-analysis of intentional outcrossing of inbred populations of vertebrates,

invertebrate and plants with a low outbreeding depression risk (evaluated using Frankham

et al.’s 2011 decision tree) has shown a positive outcome of genetic mixing (Frankham,

106

2015). However, there are only a few case studies available that employed outcrossing

for conservation purposes (Frankham, 2015). Inconsistent outcomes of hybridisation

between species, subspecies, and divergent populations and increasing use of

translocation mean that more studies are needed to allow better guidelines about when to

use multiple populations in translocation (see Close and Lowry, 1989).

The boodie or burrowing bettong (Bettongia lesueur) is a medium-sized, burrow-living

marsupial endemic to Western Australia. They are listed as Near Threatened in the 2014

IUCN Red List of Threatened Species (Richards et al., 2008a) and as Threatened by the

Environmental Protection and Biodiversity Conservation Act (1999). Prior to European

settlement, they were abundant throughout the middle and western half of Australia

(Short and Turner, 1993), but now only remain on Bernier, Dorre (hereafter referred to as

Shark Bay Islands), and Barrow Islands. Bettongia lesueur on the Shark Bay Islands are

considered to be a different subspecies to those on Barrow Island due to their significant

body size differences (Richards, 2012) and a long period of isolation from mainland

Australia (over 8000 years; Dortch and Morse, 1984). Several reintroductions to nearby

islands and some to mainland sites have been carried out using individuals from the Shark

Bay Islands or Barrow Island populations (Richards, 2012). However, only one

reintroduction to a mainland site at Lorna Glen has attempted to promote mixing between

populations. In 2010, 56 females and 53 males from Dryandra Field Breeding Facility

(originally established from 20 individuals collected from Dorre Island) and 27 females

and 40 males from Barrow Island were reintroduced to Lorna Glen in Western Australia

as part of Operation Rangeland Restoration conducted by the Department of Parks and

Wildlife (DPaW).

According to Frankham et al.’s (2011) framework, the likelihood of outbreeding

depression in this translocation is high and it is unknown whether animals from different

source populations would interbreed due to their differences. The aim of this project is to

investigate the genetic consequences of mixing two geographically isolated source

populations to create a newly established translocated population of Bettongia lesueur at

Lorna Glen. Our specific aims are to: i) assess the level of genetic divergence between

the two source populations, ii) determine whether levels of genetic variation were higher

in the translocated population relative to the source populations and whether these

patterns change over time, iii) examine the extent of mixing between the two source

populations and test for evidence of bidirectional introgression, and iv) investigate the

107

genetic bases of phenotypic variation in body size between source populations, fecundity,

and survival differences between generations.


5.3.1 Studied species

Bettongia lesueur is characterised by a short blunt head, with small rounded and erect

ears. They are yellowy grey with a light grey underside. The legs, feet, and tail are more

yellow in colour. Their fat tails are lightly haired and some have a distinctive white tip

(Burbidge and Short, 1995). They are omnivorous, nocturnal, and the only macropod that

shelters in burrows on a regular basis (Burbidge and Short, 1995). The average weight of

a Shark Bay boodie is 1.26 kg (Short and Turner, 1999) in comparison to Barrow Island

which is 0.68 kg (Short and Turner, unpublished data). The populations on the Shark Bay

Islands breed throughout the year except a period of anoestrous over summer (Short and

Turner, 1999). Breeding peaks over winter when the majority of rain falls. The population

at Dryandra Field Breeding Facility have a similar breeding cycle to those on the Shark

Bay Islands (Richards, 2012). On Barrow Island, breeding cycles are seasonally opposite

and peak in summer coinciding with cyclonic rain (Richards, 2012, Short and Turner,

1999). Females produce one young per litter and up to three young per year in captivity

(Tyndale Biscoe, 1968), but on the islands, the period of anoestrus over a dry season

means that two rather than three young are most likely to be produced per year (Short and

Turner, 1999). The young leave the pouch after 115 to 120 days and reach sexual maturity

after 280 days (Finlayson and Moseby, 2004, Tyndale Biscoe, 1968, Short and Turner,

1999). Boodies form a social group of one male to one to many females, but they tend to

forage independently at night rather than forming feeding aggregations (Sander et al.,

1997). There is no clear dominance hierarchy for reproduction, but the oldest female tends

to be the dominant individual while a young male tends to rank at the bottom of the group

(Short and Turner, 1999, Sander et al., 1997). There is no significant dimorphism between

sexes (Short and Turner, 1999). Males gain access to females by investing in knowledge

of the reproductive status and location of females within males’ day-range (Jarman,

1991). A male chases other males to defend females, but will often get supplanted,

especially those with several females in their social group (Sander et al., 1997). Based on

the social behaviour and space used, they exhibit a polygynous mating system (Sander et

al., 1997). Boodies can survive to at least three years of age (Short and Turner, 1999), but

108

animals up to 11 years had been reported in a translocated population (J. Short et al.

unpublished data).

5.3.2 Translocation history

The founders were released incrementally with an initial release of 20 boodies from

Dryandra in January, followed by an additional 67 boodies from Barrow Island in

February and 80 boodies from Dryandra in August. Twenty two of the boodies from

Dryandra that were released in August were subsequently recaptured and moved outside

the translocation site two months later. The last nine boodies from Dryandra were released

in October 2010.

5.3.3 Sampling and DNA extraction

All samples used in the study were collected during the establishment of the translocated

population at Lorna Glen (26°13’S, 121°33’E) in 2010 and during follow-up population

monitoring between 2010 and 2013. Six populations were sampled in total, two collected

from individuals translocated to Lorna Glen (representing each of the source populations:

Barrow Island 20°51’S 115°24’E, N = 67 and the Dryandra Field Breeding Facility

32°48′S 117°0′E, N = 109) and four collected from the population at the translocation site

once a year from 2010 to 2013 (2010 N = 11, 2011 N = 27, 2012 N = 48, and 2013 N =

24). All animals caught during trapping sessions had a tissue sample taken from their ear

and were measured, weighed, and their reproductive status noted. Individuals were

classified as adults if they had pouch young, one or more teats showed signs of prior

lactation, or had a fully developed pouch or testes. Females were recorded as producing

pouch young if they showed signs of lactation or trapped with pouch young. The ear tissue

biopsied from each animal was stored in 70% ethanol. DNA was extracted using a

‘salting-out’ method (Sunnucks and Hales, 1996) with a modification of 10 mg/mL

Proteinase K being added to 300 µL TNES and incubated at 56 °C. All samples were used

for mitochondrial DNA (mtDNA) and microsatellite analysis. Additional samples

obtained from Barrow Island (N = 49), Dorre Island (N = 5), and Bernier Island (N = 5)

populations were collected between 1999 and 2001 using cage traps (F. Donaldson,

unpublished data). The ear tissues were taken and stored in 70% ethanol. DNA was

extracted using Roche’s High Pure PCR (polymerase chain reaction) Template

Preparation Kit. These samples were used in the phylogenetic analysis only.

109

5.3.4 Mitochondrial DNA control region sequences

The D-loop region was amplified using primers L15999M and H16498M (Fumagalli et

al., 1997). PCR was performed using the following parameters: after an initial

denaturation at 94 °C for 2 min, 30 iterations of 94 °C for 35 s, 57 °C for 45 s, 72 °C for

1 min followed by a final extension of 10 min at 72 °C. Reactions were performed in 25

µL volumes which contained 20 ng of DNA, 2.5 μL of 10 × buffer (Fisher Biotech), 4 µL

of 25 mM MgCl2, 0.5 µL of 10 mM dNTP2, 0.5 µL of 10 µM of each primer, 0.25 µL of

0.5 U Tth Taq polymerase and 14.75 µL of dH2O. Products were sequenced in an ABI

3730 sequencer using a commercial service (Australian Genome Research Facility Ltd),

edited using SEQUENCER (Gene Codes Corporation, Ann Arbor, MI, USA), and aligned

with CLUSTAL W using default parameters (Thompson et al., 1997).

5.3.5 Microsatellites

We genotyped each individual at 18 microsatellite loci that were developed for other

macropod species, including 12 loci described in Donaldson and Vercoe (2008), three

loci from Petrogale assimilis (Spencer et al., 1995), one from P. xanthopus (Zenger et al.,

2002, Pope et al., 1996), and two from Potorous longies (Luikart et al., 1997) (Table

S5.1). PCR reactions (volume 10 µL) were performed using a QIAGEN Multiplex PCR

Kit with 10 ng of DNA and primer concentrations ranging from 0.04 to 0.4 µM (Table

S5.1). Amplifications were carried out using the following parameters: 15 min at 95 °C,

followed by 35 cycles of 30 s at 94 °C, 90 s at different annealing temperatures as

described in Table S5.1 and 60 s at 72 °C and concluding with 30 min at 60 °C. PCR

products were analysed in an ABI 3730 sequencer using a GeneScan-500 LIZ internal

size standard and scored using GENEMARKER version 1.90 (SoftGenetics).

5.3.6 Data Analysis

Mitochondrial DNA diversity was quantified by calculating the number of haplotypes,

gene diversity and nucleotide diversity using DNASP version 5 (Librado and Rozas,

2009). Pairwise ϕST values and tests for differentiation between samples taken from the

source populations and translocation site were calculated and tested using an analysis of

molecular variance (AMOVA) in ARLEQUIN version 3.0 (Excoffier et al., 2005).

Representatives of each haplotype were selected for Bayesian phylogenetic analysis. The

model used for the phylogenetic analysis was determined in MEGA 5.1 (Tamura et al.,

2011) with the phylogenetic analysis conducted using MRBAYES (Ronquist et al., 2012).

A Bayesian tree using HKY model was drawn in FIGTREE version 1.4 program

110

(Rambaut, 2012). D-loop sequences of woylies, Bettongia penicillata, from Faure Island

in Western Australia and Witchcliffe Rock Shelter in South Western Australia were used

as outgroups. The phylogenetic analysis also included D-loop sequences obtained from

B. lesueur collected on Barrow, Dorre, and Bernier Islands between 1999 and 2001.

We assessed the genotype quality by calculating the allele- and locus-specific genotypic

error rates (Pompanon et al., 2005). We also tested for the presence of null alleles in the

source population samples at each locus using MICROCHECKER (Van Oosterhout et

al., 2004). Estimates of the allelic richness (an estimate of the number of alleles per locus

corrected for sample size), gene diversity, inbreeding coefficient (FIS), pairwise genetic

divergence (FST), tests for differentiation among population samples and genotypic

disequilibrium were calculated using FSTAT version 2.9.3.2 (Goudet, 2001). The

significant deviation of FIS values from Hardy-Weinberg Equilibrium was determined by

randomization tests. Genetic divergences between pairs of population samples were

quantified using Weir & Cockerham’s (1984) FST (θ). Genotypic disequilibrium between

each pair of loci within each population sample was assessed by testing the significance

of association between genotypes. For these tests, a sequential Bonferroni correction

(Rice, 1989) was applied to control for type I statistical error. Differences in gene

diversity and allelic richness among population samples were statistically tested using

Wilcoxon’s signed-rank tests with samples paired by locus using the R version 3.0.1

statistical package (R Core Team, 2014).

Two methods were used to infer the extent of genetic mixing within the translocation site.

Firstly, we utilised a Bayesian clustering method in STRUCTURE 2.3.4 (Pritchard et al.,

2000). Analyses were performed assuming the presence of two genetic clusters (K = 2)

because this was the number of source populations. To confirm the number of genetic

clusters, we compared the likelihood values for different K values (1 – 10) and used the

ΔK method of Evanno et al. (2005a) to choose K (Table S5.2). In each analysis,

individuals were assigned a membership coefficient, which is the fraction of the genome

with membership to a particular cluster. Ten independent runs were performed using

100,000 iterations, with a burn-in period of 10,000 iterations. A chi-square test was used

to determine whether the average proportion of membership to each predefined cluster

matched the expected proportions based on the number of individuals translocated from

each source population. This number was adjusted to take into account individuals known

or assumed to have died within one month of release. After taking into account early

mortality, there were 59 founders from Barrow Island and 67 from Dryandra. Secondly,

111

we used NEWHYBRIDS version 1.1 (Anderson and Thompson, 2002) to assign

individuals born at the translocation site to one of the six generations. Genotype frequency

classes of each generation were specified as following: pure-bred Barrow Island

(1.000/0.000/0.000/0.000), pure-bred Dryandra (0.000/0.000/0.000/1.000), F1 hybrid

(0.000/0.500/0.500/0.000), F2 hybrid (0.250/0.250/0.250/0.250), backcross to pure-bred

Barrow Island (0.500/0.250/0.250/0.000), and backcross to pure-bred Dryandra

(0.000/0.250/0.250/0.500) (Anderson and Thompson, 2002). The z option was set as prior

information for individuals from the source populations. Uninformative priors (Jeffreys)

were given to both allele frequency and admixture distributions. Results presented are

based on the average of five replicates, which were run for 1,000,000 Markov chain

Monte Carlo (MCMC) sweeps following a burn-in period of 100,000. A posterior

probability value of 0.7 was used as a threshold to assign individuals to different

generation classes. Four samples fell below this threshold and were excluded from further

analysis. The results were cross-checked with CERVUS 3.0 results (Marshall et al.,

1998).

The direction of mixing between source populations was investigated by using CERVUS

3.0 (Marshall et al., 1998) to allocate parentage to F1 hybrid offspring born at the

translocation site. For these analyses, each yearly cohort of offspring were analysed

separately. Individuals born at the translocation site were allocated as offspring, while

individuals translocated to the translocation site (founders) and any adult offspring were

allocated as candidate parents. CERVUS 3.0 uses simulations to calculate critical values

of likelihood ratios, so that the confidence of parentage assignments can be determined.

Our simulations were based using an error rate of 1%, and assuming 99% of loci were

genotyped with 76.4 – 84.3% of total candidate mothers and 89.3 – 96.8% of total

candidate fathers sampled from the population. The percentage of candidate parents was

varied to account for missing samples. The analyses were run with the sex of candidate

parents known and confidence levels of 80% (relaxed) and 95% (strict). The CERVUS

results were cross-checked by comparing the mtDNA haplotypes of offspring with their

candidate mothers and by using eight known mother-offspring pairs.

We examined the direction of mixing further by comparing the proportions of mtDNA

haplotypes in F1 hybrids at the translocation site with the expected values based on the

haplotype frequencies and number of females translocated from each of the source

populations. The expected proportions were further adjusted to take into account females

who did not survive the first month in the translocation site and were, therefore, unlikely

112

to contribute to the gene pool. We also carried out these comparisons for pure-bred

Barrow Island and pure-bred Dryandra generations born at the translocation site to test

whether haplotype frequencies in the offspring matched the haplotype frequencies of the

female parents, on which the predicted values were based. Each generation was analysed

separately and deviations between observed and expected numbers of haplotypes were

tested using Chi-square Goodness of Fit.

To investigate the relationship between body size measurements and ancestry, we

compared the body weight, head length, and pes length of adults originating from the

source populations or born at the translocation site. The measurements were taken from

animals which were captured between 2010 and 2015 and subsequently genotyped. The

ancestries of offspring born at the translocation site were determined using

NEWHYBRIDS into the following generation classes: pure-bred Barrow Island, pure-

bred Dryandra, F1, F2, and backcrosses to each source population. The significance of

differences in body size measurements between pure-bred individuals born at the

translocation site and between the source populations were determined using unpaired t-

tests. We also used weighted regression to test whether variation in mean body size among

the different generations born at the translocation site conformed to a simple additive-

dominance genetic model following methods described by Kearsey and Pooni (1996).

The significance of additive, dominance and interaction effects were tested by adding

parameter coefficients for each effect to the base line model (overall mean value) and

comparing the new model to the previous model to determine whether there was a

significant improvement in fit to the data. For the model to test for additive effects, the

following parameter coefficients were used: Barrow Island pure-bred = 0, Barrow Island

pure-bred × F1 = 0.25, F1 = 0.5, F2 = 0.5, Dryandra pure-bred × F1 = 0.75, and Dryandra

pure-bred = 1. Parameter coefficients for the dominance effects were as follows: Barrow

Island pure-bred = 0, Barrow Island pure-bred × F1 = 0.5, F1 = 1, F2 = 0.5, Dryandra pure-

bred × F1 = 0.5 and Dryandra pure-bred = 0. Goodness of fit between observed and

expected generation means was tested using Chi-square. In situations where the Chi-

square indicates a simple additive or additive-dominance model does not adequately

explain variation in the trait, complicating factors such as maternal or epistatic effects

may be present (Kearsey and Pooni 1996).

To test for variation fecundity between different generation classes (BWI N = 15, BWI x

F1 N = 3, F1 N = 4, F2 N =2, and DRY N = 13), the number of offspring produced by

females between 2010 and 2015 were compared using generalized linear model fitted

113

with a quasipoisson error distribution. The initial model was constructed with number of

pouch young per genotyped female as the dependent variable and generation class, year

and their interaction as independent variables.

Capture histories between March 2010 and April 2015 of all 104 genotyped offspring

(BWI N = 37, F1 x BWI N = 7, F1 N = 20, F2 N = 5, F1 x DRY N = 4, and DRY N = 31)

were used to estimate survival rates using a Cormack-Jolly-Seber (CJS) model in Program

MARK (White and Burnham, 1999). The model was defined with two sexes and six

generations (strata) as previously determined by NEWHYBRIDS: pure-bred Barrow

Island, pure-bred Dryandra, F1, F2, and backcrosses to each source population. A full

model began with full variation of survival and capture probabilities of both sexes and all

offspring groups. Several reduced models were then tested by removing sex, groups

and/or capturing time differences. The models were compared using corrected AIC values

with the best fit model selected based on the lowest AIC value (Anderson et al., 1994).

The survival estimates were calculated by averaging survival estimates of various models

with a similar parameter structure. The average estimates were then weighted using

normalized AIC model weights (sensu Buckland et al., 1997, Burnham and Anderson,

2004). The annual survival rates were calculated by multiplying the survival estimates for

the intervals between trapping events. Unconditional standard errors were calculated

using the ‘Delta method’ and the 95% confidence intervals were calculated on the logit

scale before being back-transformed to probability (Cooch and White, 1997). The

survival estimates for March 2010, April 2010, and June 2010 intervals were excluded as

the β estimates because these intervals had very large standard errors, indicating that the

model was unable to estimate those survival rates adequately.

114

5.4 RESULTS

5.4.1 Mitochondrial DNA variation

Fifty three polymorphic sites were found in the 580-bp D-loop region, giving a total of

fifteen haplotypes. These haplotypes have been deposited in GenBank (Accession

numbers: KP257602-KP257616). Five distinctive clades were apparent corresponding to

each of the remnant populations. Three clades were found within the Barrow Island

population and one unique clade was detected in each of the Dorre and Bernier Island

populations (Figure 5.1). The AMOVA revealed 69.1% of the variation was between

island populations.

Figure 5.1 Bayesian posterior probabilities percentage tree based on 580-bp of D-loop

gene sequences from Bettongia lesueur. The tree is rooted with Bettongia penicillata from

Faure Island in Western Australia and Witchcliffe Rock Shelter in South Western

Australia (SW) as outgroups and B. lesueur haplotypes are grouped by the island on which

they occur. Representatives of each haplotype are selected from individuals born at the

translocation site between 2010 and 2013, individuals translocated to Lorna Glen from

Barrow Island and Dryandra Field Breeding Facility and individuals collected on Barrow

Island and the Shark Bay Islands between 1999 and 2001. Robustness is indicated by

Bayesian posterior probabilities percentage (≥ 50).

0.05

BarrowB

BarrowD

BernierC

BlesueurB

DorreA

BlesueurA

BernierB

BarrowC

BernierA

BpenicillataFaureIsl

DorreC

DorreD

BlesueurC

BarrowE

DorreB

BarrowA

BpenicillataSouthWes

M

B. penicillata (Faure Island)

B. penicillata (SW Australia)

A

D

E

F

B

C

G

H

I

J

K

L

N

O

0.05

Barrow Island

Bernier Island

Dorre Island

100

100100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

95 94

98

61

99

88

82

82 85

115

Each of the source populations used to establish the Lorna Glen translocated population

had both shared and unique haplotypes. Three haplotypes (A, B, and C) were found in

both the Dryandra and Barrow Island population samples. Individuals originating from

Barrow Island also carried unique haplotypes (D, E, G, and H), while individuals

originating from Dryandra had only one unique haplotype (I) in addition to shared

haplotypes (Figure 5.2). Haplotypes F, J, K, L, M, N, and O were not found in any of the

individuals used to establish the translocated population. Most of the individuals (90.1%)

from Dryandra had the haplotype I, while most of the individuals originating from Barrow

Island carried the haplotype A (57.5%, Figure 5.2).

Figure 5.2 Present distribution of Bettongia lesueur. Blue dots represent the locations of

remaining natural populations (Barrow, Bernier, and Dorre Islands). Red dots are two

translocated populations (Dryandra and Lorna Glen). Pie charts on the left represent

haplotype frequencies in Barrow Island and Dryandra individuals used to establish the

Lorna Glen translocated population. Pie charts on the right represent haplotype

frequencies in samples taken from the Lorna Glen population between 2010 and 2013.

Barrow

A

B

C

D

E

G

H

I

2010

2011

2012

2013

Dryandra

Lorna Glen

Bernier Island

Dorre Island

Barrow Island

Haplotype

116

Levels of mtDNA variation in the translocated population at Lorna Glen were comparable

to the levels found in the source populations (Table 5.1). Nevertheless, there were

significant differences in haplotype frequencies between the source populations and

between the source populations and samples taken from the translocation site (Table 5.2).

Pairwise ϕST values indicated that initially haplotype frequencies in the translocated site

were most similar to the Barrow Island source population. However, they became more

similar to the Dryandra source population after 2010 and remained stable thereafter

(Table 5.2).

117

Table 5.1 Sample sizes, estimates of genetic variation, inbreeding coefficient (FIS) and genotypic disequilibrium (GD) in the source (Barrow Island and

Dryandra) and translocated populations (Lorna Glen, LG) based on 12 microsatellite loci. N is the number of samples used in analysis. H is gene diversity.

Standard errors are given after mean values. FIS estimates significantly greater than 0 after correction for multiple comparisons are denoted with an

asterisk.

Microsatellites Mitochondrial DNA

Population N

Allelic

richness H FIS

Pairs of

loci in GD N

Number of

haplotypes H

Nucleotide

diversity

Barrow Island 65.8±0.1 3.8±0.5 0.56±0.07 0.02 0 62 7 0.58±0.01 0.008±0.0003

Dryandra 94.8±0.2 2.8±0.2 0.47±0.07 –0.03 2 91 4 0.19±0.01 0.004±0.0001

LG 2010 11.0±0.0 4.4±0.5 0.61±0.05 0.18* 0 11 2 0.22±0.06 0.001±0.0004

LG 2011 26.7±0.2 4.9±0.5 0.69±0.04 0.19* 16 26 4 0.64±0.01 0.012±0.0002

LG 2012 48.0±0.0 5.0±0.5 0.69±0.05 0.11* 24 49 4 0.50±0.01 0.010±0.0002

LG 2013 23.9±0.1 4.9±0.5 0.71±0.04 0.19* 3 24 4 0.47±0.02 0.009±0.0004

LG overall 112.6±0.2 5.1±0.5 0.70±0.04 0.17* 64 112 4 0.59±0.00 0.011±0.0001

118

Table 5.2 Pairwise FST and ϕST values between the source populations and samples

collected from the translocated population (LG) between 2010 and 2013 based on

microsatellite (below diagonal) and mtDNA data (above diagonal). Significant FST values

after correction for multiple comparisons and ϕST values are denoted with bold font.

Barrow

Island Dryandra LG2010 LG2011 LG2012 LG2013

Barrow Island – 0.72 -0.02 0.35 0.49 0.50

Dryandra 0.42 – 0.75 0.30 0.11 0.10

LG2010 0.01 0.38 – 0.26 0.43 0.44

LG2011 0.09 0.20 0.04 – 0.02 0.02

LG2012 0.18 0.10 0.11 0.02 – –0.03

LG2013 0.17 0.13 0.10 0.02 0.00 –

5.4.2 Microsatellite variation

Across the microsatellite data set, the amplification success rate was 0.973 per locus. The

allele- and locus-specific genotypic error rates were 0.030 and 0.039 respectively. Six

loci (Y151, Y170, Y76, Y112, T17-2 and Pa597) showed evidence of null alleles in the

source population samples using MICROCHECKER. The data was analysed with and

without null alleles. Both showed similar results but the presence of null alleles lowered

the posterior probability of some samples in NEWHYBRIDS. Therefore, only results

without null alleles were presented. Overall, estimates of genetic diversity were typically

higher in the translocated population at Lorna Glen than the source population samples

(Table 5.1). Pairwise tests revealed significantly higher allelic richness and gene diversity

in each of the samples taken from the Lorna Glen translocation site when compared to

the Dryandra source population samples (Wilcoxon rank sum tests, P < 0.05 in all cases).

However, there were no significant differences between samples from the translocation

site and the Barrow Island source population or between population samples taken from

the translocation site in different years.

All samples from the translocation site had significantly positive multilocus FIS values

(randomization tests, P < 0.008). Multilocus FIS values in the remaining population

samples (i.e. those from the source populations) were not significantly different from zero

(Table 5.1). The number of pairs of loci in genotypic disequilibrium (GD) ranged from

zero to 24, with the highest levels occurring in the 2011 and 2012 population samples

119

taken from the translocation site. The number of pairs of loci in GD in the source

population samples ranged from zero to two (Table 5.1).

5.4.3 Population structure and genetic mixing

Pairwise population ϕST and FST values based on mtDNA and microsatellite data indicated

there was substantial genetic differentiation between the source populations (Table 5.2).

There were also significant divergences among the different collection years sampled

from the translocation site, and between the translocation site and both source population

samples. Initially, ϕST and FST values indicated the translocated population was most

similar to the Barrow Island source population. However, it became more similar to the

Dryandra source population overtime. This pattern was more pronounced in the mtDNA

data.

The clustering analyses also revealed changes in the genetic composition of the

translocated population over time (Figure 5.3). Most offspring born at the translocation

site during 2010 had an ancestry matching the founders from the Barrow Island

population. There was also an individual with an ancestry matching the Dryandra source

population. Offspring with mixed ancestry started to appear at the translocation site from

March 2011. The lag in genetic mixing likely reflects the different release times of

founding animals (i.e. the majority of Barrow Island founders were released in February,

while most Dryandra founders were released mid-August). Genetic mixing between the

source populations was also clearly evident in the 2012 and 2013 population samples.

Overall, the proportion of membership to each predefined genetic cluster in the population

samples from the translocation site was not significantly different from the predicted

proportions based on the number of founders from each source population after taken

early mortality into account (Table 5.3).

120

Figure 5.3 Summary of the Bayesian clustering results for the Lorna Glen translocated

population assuming two admixed populations (K = 2). Each individual is represented by

a bar showing its estimated membership to a particular cluster (represent by different

colours). Black lines separate samples from different source populations (Barrow Island

and Dryandra) and collection years at the Lorna Glen translocation site.

Table 5.3 Observed and the initial expected proportions of membership to each

predefined cluster in samples taken from the Lorna Glen population. Observed

proportions are based on results from the STRUCTURE analysis of microsatellite data.

Expected proportions are based on the number of individuals translocated from each

source population after individuals with known or assumed mortality were removed. For

convenience, clusters have been labelled according to the source population they define

(BWI = Barrow Island and DRY = Dryandra).

Observed Expected

Sample N BWI DRY BWI DRY Х2 P

2010 11 0.895 0.105 0.468 0.532 – –

2011 27 0.609 0.391 0.468 0.532 2.1 0.143

2012 48 0.431 0.569 0.468 0.532 0.3 0.605

2013 24 0.450 0.550 0.468 0.532 0.0 0.858

Overall 113 0.537 0.463 0.468 0.532 2.1 0.143

121

Evidence of genetic mixing between source populations was also found with the

NEWHYBRIDS analysis. Across all of the offspring born at the translocation site that

had the posterior probability above the 0.7 threshold in NEWHYBRIDS (N = 109), 38.5%

were designated as pure-bred Barrow Island, 28.4% pure-bred Dryandra, 18.3% F1

hybrid, 4.6% F2 hybrid, 6.4% backcross to Barrow Island, and 3.7% backcross to

Dryandra. The classification in the NEWHYBRIDS analysis was consistent with

CERVUS for the first generation (i.e. pure-bred and F1 hybrid), but not for the second

generation (i.e. F2 hybrid and backcross) with many were miss-assigned as F1 hybrids or

pure-breds. The parentage analysis suggested that while introgression between source

populations was bi-directional, it was biased towards Dryandra females mating with

Barrow Island males. Of the 20 F1 hybrids identified in the NEWHYBRIDS analysis, 18

(90%) had a candidate mother from the Dryandra source population or was an adult

female born at the translocation site with a pure-bred Dryandra ancestry. By comparison,

all of the offspring identified as being pure-bred Barrow Island had candidate mothers

with Barrow Island ancestry and all of the offspring identified as pure-bred Dryandra had

candidate mothers with a Dryandra source population ancestry. This pattern was

supported by the mtDNA data which revealed higher than expected numbers of Haplotype

I, which was restricted to the Dryandra source population, in the F1 hybrid offspring

(Table 5.4). There were no significant differences between observed and expected

haplotype frequencies in the pure-bred Barrow Island and pure-bred Dryandra offspring

(Table 5.4).

Table 5.4 Observed and expected numbers of haplotypes in individuals born at the Lorna

Glen translocation site. Individuals were classified as pure Barrow Island, pure Dryandra

or F1 hybrid based on the results from the NEWHYBRIDS analysis. Expected numbers

for each haplotype are based on the haplotype frequencies and the number of females

translocated from each source population after females with known or assumed mortality

were removed. Haplotypes with low expected numbers were pooled prior to carrying out

the Chi-square test.

Haplotype

Sample A B C I others Х2 P

Pure Barrow

Island

24 (28.0) 6 (1.9) 11 (9.3) – 0 (1.9) 2.25 0.134

Pure Dryandra 1 (1.2) 0 (0.0) 0 (1.2) 30 (28.5) – 0.96 0.327

F1 hybrid 2 (6.1) 1 (0.4) 0 (2.4) 18 (11.7) 0 (0.4) 7.62 0.006

122

5.4.4 Effects of introgression on body size, reproductive fitness and survival

probability

Both females and males from the Dryandra source population were significantly heavier

and larger than those from the Barrow Island source population (body weight females: t

= 21.8, P = 0.038; body weight males: t = 21.3, P < 0.001; head length females: t = 19.2,

P < 0.001, head length males: t = 16.7, P < 0.001; pes length females: t = 28.3, P < 0.001,

pes length males: t = 35.2, P < 0.001). These differences were also apparent in pure-bred

offspring born at the translocation site (body weight females: t = 2.8, P = 0.038; body

weight males: t = 7.8, P < 0.001; head length females: t = 4.5, P < 0.001, head length

males: t = 6.1, P < 0.001; pes length females: t = 14.0, P < 0.001, pes length males: t =

19.3, P < 0.001), suggesting the body size differences between the source populations

have an underlying genetic basis. Significant differences between pure-bred Dryandra

and Dryandra founders were found in body weight (females: t = 3.4, P = 0.022; males: t

= 5.7, P < 0.001) but not pes and only head length of Dryandra females (t = 4.4, P <

0.001).

A genetic basis of variation in body size between source populations was supported by

the analysis of generation means performed on offspring born at the translocation site

(Figure 5.4). Significant additive genetic effects were detected for head length, pes length,

and body weight in males and for head length and pes length in females (Table 5.5). In

addition to additive effects, significant dominance effects were detected for head length

in females and pes length in males (Table 5.5). While a simple additive-dominance model

was able to account for variation in the observed means for traits, they did not for pes

length in females and head length in males (Table 5.5). This suggests that additional

genetic parameters are necessary to explain the observed variation.

123

Figure 5.4 Mean body size in the source populations (full symbols) and different

generations born at the Lorna Glen translocation site (open symbols). The fitted line is

the expected mean values for individuals born at the translocation site based on a purely

additive genetic model. Error bars are standard errors.

124

Table 5.5 Estimates of composite genetic effects underlying divergence in body size

between source populations used to establish the translocation site. The number of

asterisks indicates the significance of the improvement of fit when the parameter was

added. NS not significant; a marginally nonsignificant; P < 0.1; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P

< 0.001. The degrees of freedom for each Chi-square are equal to the number of

generation means minus the number of significant parameters.

Body size measurement

Sex Parameter Head length Pes length Weight

Females Intercept (mean) 70.0 ± 0.6∗∗∗ 92.8 ± 1.1∗∗∗ 1069.3 ± 40.9∗∗

Additive 4.2 ± 0.6∗ 8.7 ± 1.4∗ 189.4 ± 59.0 a

Dominance 6.2 ± 1.0∗

Χ2 1.17 NS 15.4∗∗∗ 5.1NS

Males Intercept (mean) 73.9 ± 1.0∗∗∗ 92.6 ± 0.4∗∗∗ 1031.3 ± 24.1∗∗∗

Additive 4.4 ± 1.4∗ 8.7 ± 0.4∗∗∗ 167.8 ± 33.3∗∗

Dominance 3.9 ± 0.7∗

Χ2 17.1∗∗ 1.9NS 9.1NS

We found no significant effect of generation class (χ2 = 3.12; P = 0.538), year (χ2 = 2.26;

P = 0.132) or a generation class-by-year interaction (χ2 = 1.52; P = 0.677) on the number

of pouch young produced by females. The most parsimonious model in MARK had

survival differences between males and females and encounter probability that varied

over time as the best fit model. On average, females had 6.3% higher survival than males.

A drop in the encounter probability was detected around April 2012 from an average of

76% to 50% (Figure S5.1). The model did not indicate any significant survival differences

between generations (Table 5.6)

125

Table 5.6: Annual survival estimates from Mark-Recapture and the 95% confidence interval of different generations and sex of Bettongia lesueur in the

translocation site at Lorna Glen for the years 2010-2015 (BWI = Barrow Island and DRY = Dryandra).

Year

Translocation site Sex

BWI F1 x BWI F1 F2 F1 x DRY DRY Male Female

2010 0.88 (0.84, 0.90) * * * * 0.88 (0.84, 0.90) 0.87 (0.81, 0.90) 0.89 (0.82, 0.93)

2011 0.88 (0.84, 0.90) 0.88 (0.84, 0.91) 0.88 (0.84, 0.91) * * 0.88 (0.84, 0.90) 0.87 (0.81, 0.90) 0.89 (0.82, 0.93)

2012 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.93 (0.91, 0.95) 0.95 (0.92, 0.97)

2013 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.93 (0.91, 0.95) 0.95 (0.92, 0.97)

2014 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.94 (0.92, 0.95) 0.93 (0.91, 0.95) 0.95 (0.92, 0.97)

2015 0.97 (0.96, 0.98) 0.97 (0.96, 0.98) 0.97 (0.96, 0.98) 0.97 (0.96, 0.98) 0.97 (0.96, 0.98) 0.97 (0.96, 0.98) 0.97 (0.95, 0.98) 0.97 (0.96, 0.98)

* No animal from this generation was caught this year.

126

5.5 DISCUSSION

5.5.1 Phenotypic and genetic differentiation between island populations

As expected there is substantial phenotypic and genetic differentiation between the Shark

Bay and Barrow Island populations of Bettongia lesueur. Each population formed a

unique clade or clades in the phylogenetic analysis of mtDNA data. They also formed

discrete genetic clusters following Bayesian cluster analysis of microsatellite data and

had large differences in allele and haplotype frequencies. There were haplotypes that were

found in both Barrow Island and Dryandra populations. Interestingly these haplotypes

were not found in Dorre Island samples. It is possible that these shared haplotypes were

inherited from the original 20 individuals from Dorre Island and the sample size from

Dorre Island in this study was too small (N = 5) to detect haplotypes at low frequency.

Alternatively, incomplete lineage sorting, or balancing selection may be possible

explanations for these haplotypes being found in both source populations (Hedrick,

2013). Neutral loci from microsatellites showed that individuals with shared haplotypes

belonged to Dryandra ancestry (N = 9) so it is less likely that these individuals were the

result of recent admixture. We were able to show that the body size differences between

the populations were maintained in pure-bred adults born at the translocation site and

raised in the same environment. Using information about ancestry derived from the

NEWHYBRIDS analysis, we showed there is an underlying genetic basis to the

differences in body size between populations with the detection of significant additive

and dominance genetic effects. These genetic divergences likely reflect the geographical

isolation of these islands to each other and the Australian mainland, which occurred

approximately 8,000 to 10,000 years ago (Dortch and Morse, 1984).

5.5.2 Genetic consequences of mixing geographically isolated island populations

Despite high levels of genetic differentiation between remnant populations, B. lesueur

translocated to a new site on the mainland were able to interbreed and produced viable

offspring, with no obvious fitness costs on fecundity and survivorship. Admixed

individuals were evident in the 2011 collection year after all of the founders were

released. After three years, more than half of the offspring born at the translocation site

were of hybrid or backcrossed origin. Significantly positive FIS values were evident in

each yearly collection taken from the translocated site, suggesting that while

interbreeding was taking place, mating between individuals with different ancestries was

127

non-random. By contrast, no deviations from random mating were observed in the source

population samples.

Non-random mating was also evident from the asymmetrical introgression between

source populations; crosses between smaller-sized Barrow Island males and larger-sized

Dryandra females were significantly more common than expected. The reason for the

asymmetry is unclear. Reproductive interference arising from male-male competition or

female mate choice is expected to favour larger dominant males rather than smaller males

in many species of the suborder Macropodiformes (Hynes et al., 2005, Sigg et al., 2005,

Miller et al., 2010a, Pope et al., 2012). It is also unlikely that the asymmetry was caused

by different breeding cycles because our trapping records indicated no obvious

differences in the frequency of pouch young carried by females originating from Barrow

Island and Dryandra. It is possible that the asymmetrical introgression is caused by cyto-

nuclear incompatibilities selected against hybrids resulting from crosses between

Dryandra males and Barrow Island females during gestation (see Arntzen et al., 2009,

Álvarez and Garcia-Vazquez, 2011). However, this is less likely because we observed bi-

directional rather than a unidirectional introgression.

Alternatively, there might be a physiological limitation making it too costly for smaller

Barrow Island females (pre- or post-parturition) to provide for large offspring that might

result from breeding with large Dryandra males. Freegard et al. (2008) who developed an

age estimation growth curve for B. lesueur, suggested that maternal weight may play role

in development variations of B. lesueur pouch young of the same age due to larger

females producing more milk, resulting in larger pouch young. Another study of B.

lesueur on Dorre and Bernier Island also found that females with weight range 1000-1600

g produced pouch young more frequently (from signs of lactation or trapped with pouch

young) than females weighed outside this range (Short and Turner, 1999). Smaller-size

Barrow Island females may not produce sufficient amount of milk to accommodate

larger-sized hybrid pouch young and could cause an early mortality. However, it is

difficult to measure this in the field because female boodies can reproduce throughout the

year. Any miscarriage or pouch young mortalities could be undetected. It is noteworthy

that there was a clear difference in body weight between pure-bred Dryandra and

Dryandra founders in both males and females, which indicated reduced condition and

these individuals were potentially stressed, which can impact their ability to compete for

mates and/or raise young.

128

As found in previous studies on recently established populations involving multiple

source populations, we found evidence of higher levels of genetic variation in the

translocated population than in one or more of the source populations (Huff et al., 2010,

Kennington et al., 2012, Ransler et al., 2011, Stockwell et al., 1996). This result is not

unexpected as long-isolated populations often carry different subsets of alleles as a result

of lack of gene flow, genetic drift, and local selection (Eldridge et al., 1999).

Alternatively, a large number of founders may help alleviate loss of genetic variation

which could explain the high diversity levels (Miller et al., 2009). The increased genetic

diversity was not much greater than the level found in the most variable source

population, Barrow Island. This result parallels Huff et al. (2010) who found all

reintroduced populations of slimy sculpins (Cottus cognatus) exhibited higher levels of

genetic diversity than any source populations, but the increases were only slightly higher

than the single most genetically diverse source population. While we found no evidence

of differences in genetic diversity among collection years, previous studies have shown

that translocated populations often lose genetic diversity as a result of the founder effect

(Miller et al., 2009), mating system (Adams et al., 2011, Sigg, 2006), or survivorship

differences of either founders or offspring (Biebach and Keller, 2012, Bolnick et al., 2008,

Arntzen et al., 2009).

Both microsatellites and mitochondrial DNA proportions reflected the founder

proportions from each source population after individuals with known or assumed

mortality were removed. However, significant deviations were observed when the

expected proportions did not take into account early mortality. These results are

consistent with the view that differential survival of the founders can substantially affect

the genetic composition of a translocated population (Biebach and Keller, 2012).

Although the genetic proportions of the translocated population were not significantly

different from expected in this study, these proportions could change in later generations

as the translocated population adapts to its new environment and local selection drive

changes in selected traits (e.g. Binks et al., 2007).

129

5.5.3 Considerations for conservation and management

Our results support the use of multiple source populations for enhancing genetic diversity,

which may be essential to provide newly established populations with adequate adaptive

variation to cope with new environments such as translocation sites. This adds further

support for the use of outcrossing as an effective management option if it is carried out

under appropriate circumstances. There are a number of guidelines to evaluate when

genetic mixing is appropriate (Frankham et al., 2011, Weeks et al., 2011, Hedrick and

Fredrickson, 2010). However, genetic monitoring is essential to evaluate outcomes of

admixture and to assess whether a newly established population maintains sufficient

genetic variation. This study shows that the extent of admixture can be influenced by

mortality and reproductive success. We were able to show that the genetic contributions

of each source population to the translocated population reflected the number of survived

founders from each source population. It is unclear why interbreeding between source

populations was asymmetric with a bias for crosses between females from the genetically

larger-sized Dryandra population and males from the genetically smaller-sized Barrow

Island population. Nevertheless, if this pattern continues, mtDNA diversity from Barrow

Island lineage will decline overtime because the small effective population size of Barrow

Island females. Supplementation of Barrow Island females may be necessary. However,

a longer monitoring is needed to determine the cause of asymmetry to ensure appropriate

actions are taken.

130

CHAPTER SIX

General Discussion

6.1 INTRODUCTION

The evolutionary potential of a newly established population is primarily determined by

the genetic composition of the founding individuals, particularly the amount of genetic

variation they carry. Selecting individuals for translocation is complicated by a number

of factors. First, the random selection of individuals from a source population may not

adequately capture the intended level of genetic diversity due to fine-scale population

structure that may occur within a landscape (Holderegger and Wagner, 2008). Genetic

heterogeneity across landscapes may be further complicated by behaviour (Croteau et al.,

2010, Double et al., 2005, Lampert et al., 2003, Hazlitt et al., 2006), for example,

differential dispersal patterns between sexes (e.g. Banks and Peakall, 2012, Peakall et al.,

2003). This means that sampling strategies for each sex may need to be designed

accordingly. Second, if animals are selected from multiple populations, there is a risk of

an uneven genetic contribution from each source population, as a result of differential

mortality, biased reproductive success, or release strategies i.e. the timing of release and

proportion of animals from each source population. In cases where source populations

have been isolated for thousands of years and are genetically differentiated, genetic

mixing between them may have deleterious effects (Allendorf et al. 2001; Frankham et

al. 2011; but see Frankham 2015; Weeks et al. 2011). These factors can influence the

maintenance of genetic variation within translocated populations, which affects long-term

persistence of the new population.

Using case studies of two endangered mammals, the dibbler and the burrowing bettong,

I have investigated the following i) how information about population structure and

dispersal patterns may assist in selecting individuals for captive breeding and

translocations, ii) the genetic outcomes of genetic mixing in translocations established

using multiple source populations, iii) the consequences of mixing source populations

that differ in body size and iv) factors that influenced the genetic contributions from each

source population in the translocations using multiple sources. The following is a

summary of my main findings, the management implications of my research, and overall

conclusions.

131

6.2 USING GENETIC STRUCTURE AND DISPERSAL PATTERN TO ASSIST

FOUNDER SELECTION

The Fitzgerald River National Park contains two distinct populations of dibblers that

reside on the western and central-eastern sides of the park (Chapter 2). A waterway and

isolation-by-distance are possible barriers to gene flow between these populations.

Waterways are a common barrier to gene flow in mammal species (Aars et al., 1998, Pfau

et al., 2001, Goossens et al., 2005, Eriksson et al., 2004, Quemere et al., 2010). For

example, the construction of dams and weirs that kept river flows high over the dry

summer months led to genetic divergences between different sides of the Murray River

in just 50 generations of the yellow-footed antechinus (Antechinus flavipes) (Lada et al.,

2008). However, as the river systems in the FRNP are dried up over summer (Water and

Rivers Commission, 2003), the main factor contributing to the observed genetic structure

is likely to be the large geographic distance between the regions and the limited dispersal

ability of P. apicalis.

Spatial autocorrelation analysis also revealed evidence of female philopatry and male-

biased dispersal in dibblers from the FRNP. This finding is consistent with many other

dasyurids (Peakall et al., 2003, Cockburn et al., 1985, Hazlitt et al., 2004, Soderquist and

Lill, 1995). Females had a positive genetic structure up to distances of 200 m, while males

showed no evidence of genetic heterogeneity, suggesting they move large distances.

Indeed, the trapping records show that males can move as far as 900 m. This distance is

comparable to other dasyurid species of a similar size such as agile antechinus (A. agilis),

brown antechinus (A. stuartii), and dusky antechinus (A. swainsonii) (Kraaijeveld-Smit

et al., 2007, Banks and Peakall, 2012, Fisher, 2005). Both landscape-scale and fine-scale

genetic structuring detected within the park demonstrated the importance of identifying

potential barriers to gene flow (i.e. landscape features and dispersal ability) and designing

sampling strategies accordingly to maximize genetic variation (e.g. sampling both

subpopulations) and avoid inbreeding (e.g. sampling females > 200 m apart or post

dispersal males).

132

6.3 OUTCOMES OF TRANSLOCATIONS ESTABLISHED USING MULTIPLE

SOURCE POPULATIONS

6.3.1 Retention of genetic diversity

All three translocations that were examined in this thesis maintained at least 95% of their

original gene diversity. Genetic diversity was maintained over four generations in the

burrowing bettong translocation at Lorna Glen and at least ten generations in the dibbler

translocations on Escape Island and at Peniup Nature Reserve (Table 6.1). Only the

dibbler translocation to Escape Island and the burrowing bettong translocation to Lorna

Glen showed higher levels of genetic variation compared to both source populations

(Chapters 4 and 5). The genetic diversity in these populations was not significantly much

more than the level found in the most variable source population. The gain in genetic

diversity in the translocated populations is a consequence of the genetic divergences

between the source populations. Long isolated populations are less likely to share alleles

that are identical-by-decent (Eldridge et al., 1999). This was clearly evident by the levels

of genetic variation shared between source populations for the different translocations.

For example, in the Peniup translocation, established using two source populations within

the FRNP, only 3.2% of the total genetic variation was due to differences between the

source populations (Chapter 3). By contrast, 46% and 59% of the genetic variance resided

between the source populations in the Escape Island and Lorna Glen translocations

respectively (Chapters 4 and 5). Similar increases of genetic diversity after assisted gene

flow or translocations from multiple sources has been previously reported in mountain

pygmy possums (Burramys parvus, Weeks et al. 2015), Florida panthers (Felis concolor

coryi, Hedrick 1995), Rocky Mountain bighorn sheep (Ovis canadensis, Miller et al.

2012), and grey wolves (Canis lupus, Adams et al. 2011). The increase in genetic

diversity found in my studies contributes to mounting evidence of the benefits of

augmenting gene flow between populations. Further, the results showed that it is most

effective if the source populations exhibit some degree of genetic differentiation between

them.

133

Table 6.1 Descriptive statistics and genetic variation detected by microsatellite loci in the

source, captive and translocated populations in the dibbler and burrowing bettong

translocations. N is an average sample size per locus. H is gene diversity. AR is allelic

richness. FIS is inbreeding coefficient. Ne is an effective population size. Lost Ne% is a

calculation of the percentage decrease in Ne of the translocated population compared to

the source populations one and two. Standard errors are given after average values. Bold

font represents a significant P-value (P < 0.05).

Population N H AR FIS Ne Range Lost Ne%

Chapter 3 - Peniup dibbler translocation

Source one 172.6±4.3 0.67±0.04 4.4±0.4 0.05 85.2 62.3 – 122.0 –

Source two 43.0±1.1 0.66±0.05 4.2±0.4 0.03 59.3 39.5 – 104.6 –

Captive population 220.8±0.6 0.65±0.05 4.2±0.3 0.00 24.5 20.7 – 28.7 71.2 – 58.7

Translocated population 132.2±0.9 0.64±0.04 3.9±0.3 0.01 15.8 13.7 – 18.1 81.5 – 73.4

Chapter 4 - Escape Island dibbler translocation

Source one 115.1±0.9 0.37±0.06 2.2±0.2 0.05 16.2 5.8 – 36.6 –

Source two 48.6±0.8 0.13±0.05 1.5±0.2 0.28 2.2 0.5 – 11.3 –

Captive population 67.7±1.2 0.40±0.05 2.3±0.3 -0.02 1.8 1.3 – 2.5 88.9 – 18.2

Translocated population 106.6±2.3 0.42±0.05 2.3±0.2 0.11 11.3 3.3 – 26.1 30.2^

Chapter 5 - Burrowing bettong translocation

Source one 65.8±0.1 0.56±0.07 3.8±0.5 0.02 39.6 25.0 – 69.8 –

Source two 94.8±0.2 0.47±0.07 2.8±0.2 –0.03 40.7 23.4 – 79.2 –

Translocated population 112.6±0.2 0.70±0.04 5.1±0.5 0.17 2.7 2.4 – 3.0 93.2 – 93.4

^but gain of 513.6% when compared to source population two

6.3.2 Genetic similarity and inbreeding coefficient

Interbreeding between individuals from different source populations reduced genetic

relatedness among their offspring, especially between island populations as used for the

Escape Island and Lorna Glen translocations (Figure 6.1). However, the reduction was

short-lived as individuals born on translocation sites continued to interbreed (Figure 6.1b

and 6.1c). By contrast, inbreeding coefficient (FIS) values showed more complicated

patterns (Table 6.1). In the Peniup population of dibblers, FIS was lower and not

significantly different from the source populations (Table 6.1, Chapter 3). FIS values of

the captive population and wild-born dibblers on Escape Island were generally lower than

the source populations, though the overall FIS of wild-born Escape Island dibblers was

134

higher but it was driven by one collection year (Table 6.1, Chapter 4). FIS values in the

translocated burrowing bettong population were significantly and consistently higher than

the source populations, despite having significantly lower levels of genetic relatedness

(Table 6.1, Figure 6.1c). The consistently high FIS values could due to stronger mating

preferences isolating source population lineages and/or the significant reduction of

effective population size (Chapter 5). Augmenting gene flow between inbred populations,

for example African lions (Panthera leo, Trinkel et al. 2008) and Mexican wolves (C.

lupus baileyi, Hedrick & Fredrickson 2008), has been shown to lower the levels of

inbreeding and improve fitness of the inbred populations. Indeed, it has been

recommended to outcross highly inbred populations to reverse the deleterious effects of

inbreeding (Hedrick and Fredrickson, 2010, Frankham, 2015, Frankham et al., 2011). In

general, my results confirm that outcrossing lowers inbreeding levels, or at least reduces

genetic similarity among offspring (Chapters 3 and 4). However, there is a risk that

founder events and/or non-random reproductive success can also severely reduce the

effective population size and subsequently increase inbreeding levels within the newly

established population (Chapter 5).

135

a)

b)

c)

Figure 6.1 Comparison of relatedness between individuals within source, captive and

translocated populations of a) dibbler at the Peniup translocation, b) dibbler at the Escape

Island translocation and c) burrowing bettong at the Lorna Glen translocation. Error bars

are bootstrapped 95% confidence limits.

-0.05

0.00

0.05

0.10

0.15

0.20

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

r

Year

Source population one Source population two


-0.10

0.00

0.10

0.20

0.30

0.40

0.50

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

r

Year

Source population one Source population two


-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

Source

population

one

Source

population

two

Translocated

population

2010

Translocated

population

2011

Translocated

population

2012

Translocated

population

2013

r

136

6.3.3 Effective population size

All translocated populations experience founder effects, and in my study this was

manifested in the reduction of the effective population sizes (Ne) from 30.2% to 93.4% of

the source population estimates of Ne (Table 6.1). There were no correlations between

sample size and Ne in all chapters (Figure S6.1). To avoid the influence of sampling time

and population variation effects, I selected random individuals from the burrowing

bettong source populations and plotted sample size against Ne. There were no obvious

changes in the Barrow Island population, but a negative correlation was detected in the

Dryandra population (rho = – 0.89, P = 0.03, Figure S6.2). However, Ne estimates were

relatively similar across various sample size. Variance of these estimates were large. A

significant reduction of Ne was not surprising given that relatively small numbers of

individuals were selected from the source populations for the translocations. In addition,

mortality and/or varied reproductive success among founders further reduced effective

population size (Chapters 4 and 5). Previous studies on other mammal and bird

translocations reported approximately 3 – 11 fold reduction of Ne after the populations

became established (mountain bighorn sheep, Fitzsimmons et al. 1997; Saddleback,

Island robin, and takahe, Jamieson 2011; island tammar wallaby, Miller et al. 2011;

golden bandicoots, Ottewell et al. 2014; little spotted kiwi, Ramstad et al. 2013). Newly

established populations that experience the founder effect often lose significant amounts

of genetic variation and become genetically differentiated from the source through

genetic drift (e.g. Hedrick et al., 2001, Hundertmark and Van Daele, 2010, Broders et al.,

1999). However, none of the case studies in this thesis showed a significant reduction in

genetic diversity (Table 6.1). One possible explanation was suggested by Kolbe et al.

(2007) who proposed that founder effects and admixture effects can occur

simultaneously. Admixture effects may have offset significant loss of genetic variation

from founder effects. It is also noteworthy that the Escape Island population gained five

times larger effective population size than one of the source populations (Chapter 4). This

was likely to be the result of a small Ne in this source population and a benign captive

environment, which increases survivorship and lowers the reproductive success variance

in captivity, resulting in a comparatively large Ne.

137

6.4 CONSEQUENCES OF GENETIC MIXING ON OFFSPRING

Hybrids (F1) in all translocations were fertile and able to produce F2 and backcrosses

(Chapters 3, 4, and 5). This was a significant finding, particularly for the burrowing

bettong translocation to Lorna Glen, established from highly diverged source populations.

In cases where source populations exhibited morphological differences, my studies

showed that some traits of F1 or individuals that possessed both ancestries equally showed

intermediacy, while others did not. For example, male dibblers on Escape Island showed

intermediate body weight and pes length, but they also had smaller head length and longer

short pes length than both parental ancestries (Chapter 4). F1 of burrowing bettongs had

similar body size to the ancestry that was larger and heavier (Chapters 5). These non-

intermediate phenotypes could be driven by number of factors. Both additive and

maternal effects are likely to be contributing to morphological variation in the dibbler

translocation on Escape Island (Chapter 4) and burrowing bettongs at Lorna Glen

(Chapter 5). However, source population differences in body weight and size in the

dibbler were measured on different islands, so it is possible that the differences could be

due to environmental effects (Chapter 4). This finding is paralleled with a study of the

insular Rock Mountain bighorn sheep (Ovis Canadensis), which found an average birth

weight of F1 lambs to be similar to the level found in the purebreds with heavier birth

weight (Hogg et al., 2006). Similarly, in the nine-spine stickleback (Pungitius pungitius),

a hybrid’s phenotypic similarity to the larger-size ancestor was suggested to be a result

of additive and maternal effects (Ab Ghani et al., 2012). Larger mothers were observed

to provide a larger egg size, which resulted in a larger offspring (Ab Ghani et al., 2012).

Another interesting result was the lower mean body size and weight of F2 burrowing

bettong males compared to the F1 generation, which suggests possible negative epistatic

effects (Kearsey and Pooni, 1996). However, further monitoring of later generations is

required to test this possibility.

6.5 FACTORS INFLUENCING THE GENETIC CONTRIBUTIONS OF PARENTAL

LINEAGES

The genetic contribution from each source population was influenced by mortality, varied

reproductive success among founders, the proportion of founders from each source

population, and the timing of their release. Some mortality is expected in translocations

(Teixeira et al., 2007, Moseby et al., 2011). Early mortality has been found to affect as

much as half of the released animals (e.g. Capra ibex, Biebach & Keller 2012 and

138

Sphenodon guntheri, Nelson et al. 2002) with the level of genetic diversity in the

translocated population determined by admixture between the surviving individuals.

However, mortality attributed by fitness variance between the source populations could

result in unexpected genetic contribution to the newly established population;

particularly, if one of the source populations suffers from inbreeding depression (Keller

and Waller, 2002, Crnokrak and Roff, 1999). While no studies have investigated the

consequence of using inbred or low adaptive potential source populations in

translocations, there is mounting evidence that high inbreeding levels reduce

survivorship, reproductive output, and resistance to diseases, predation, and

environmental stress (see review Keller and Waller, 2002). In the burrowing bettong

translocation, after removing known and assumed mortality, we have 59 (35M 24F) and

67 (30M 37F) animals left respectively. The proportion of animals available for mating

from both source populations was similar. Moreover, the observed genetic contributions

estimated from microsatellite loci reflected the number of animals that survived and

became established after the translocation (Chapter 5). My studies also showed that whilst

reproductive success among of founding individuals can be influenced by many factors

(e.g. Holleley et al., 2006, Kraaijeveld-Smit et al., 2002b, Parrott et al., 2015), one of the

most important is male body size. This was most apparent in dibblers where males with

a large body size had a reproductive advantage during courtship in captivity (Chapter 4),

though there are likely to be other factors in addition to body size that govern reproductive

success of males in the wild (Parrott et al., 2015, Kraaijeveld-Smit et al., 2002b). It was

also apparent that changes in genetic contributions from source populations reflect the

number of founders from each source population, as well as the timing of release of these

individuals. The genetic composition of the captive and translocated populations in the

dibbler translocation at Peniup NR changed over time from having a genetic composition

dominated by the western lineage to one closer to the eastern lineage (Chapter 3). These

changes appear to be driven by different proportions and the timing of release of animals

from the source populations to the captive colony. Manipulations of the contributions

from different source populations are not uncommon in translocations and has been used

to reduce levels of inbreeding previously (Hedrick et al., 1997). However, as shown in

the dibbler translocation at Peniup NR, the genetic contributions from the minor lineages

are vulnerable to losses when they are too low (Chapter 3, Raisin et al., 2012).

139

6.6 IMPLICATIONS FOR CONSERVATION MANAGEMENT

6.6.1 Sampling strategies

Although 30 ˗ 50 individuals are recommended for translocations to capture 95%

heterozygosity or rare alleles with frequency less than 5% (Allendorf and Luikart, 2007,

Hedrick, 2000), localised sampling strategies are unlikely to adequately capture the full

range of genetic diversity, as shown in Chapter 2. Here, I was able to show that sampling

dibblers from only one side of the Fitzgerald River National Park would fail to capture

3.2% of the total genetic variation including many unique alleles. Natural populations are

often subdivided with different allele frequencies as a result of restricted gene flow due

to landscape features and dispersal ability/strategies (e.g. Chapter 2, Banks and Peakall,

2012, Lada et al., 2008). Therefore, population management and selection of individuals

to use in translocation should be planned accordingly. For example, mixing dibblers from

different subpopulations within the Fitzgerald River National Park in captive breeding

and translocation programs would maximise genetic variation and reduce genetic

similarity between founding individuals (Chapter 2). Individuals selected for these

programs, especially females and young males, should be collected at least 200 m apart

(Chapter 2).

In the past, there has been a reluctance to use multiple source populations for

translocations, but my study shows that, if done in the right circumstances, can have great

benefits. Newly established populations are likely to benefit from outcrossing if the

source populations are historically outbred, have high levels of inbreeding, and/or have

low genetic diversity (e.g. the Escape Island translocation, Chapter 3). If the source

populations (from the same species) have been isolated for longer than 500 years,

populations should be evaluated for the probability of outbreeding depression using the

Frankham et al. (2011) and Weeks et al. (2011) decision trees. Populations that have low

risk levels are likely to benefit from outcrossing (Chapters 3 and 4, Frankham, 2015). For

populations with a high probability of outbreeding depression (Chapter 5), but suffering

from inbreeding depression or showing high levels of inbreeding, intentional outcrossing

is also recommended, though population monitoring is essential for at least two

generations following such interventions.

140

6.6.2 Monitoring

Monitoring translocations after the release of individuals is crucially important for

assessing whether the translocation is successful and maintaining sufficient genetic

variation for both short- and long-term persistence (Schwartz et al., 2007). As shown in

this thesis, and in previous studies (Jamieson, 2011, Broders et al., 1999, Hundertmark

and Van Daele, 2010), translocated populations are prone to loss of genetic diversity and

fluctuations in allele frequencies due to founder effects and small Ne after populations

have become established. Continued genetic monitoring is essential for assessing whether

the translocation achieves the goal of maintaining 90% – 95% of the original genetic

diversity. Effective monitoring requires quantifying baseline levels of genetic diversity in

the source population(s) using a range of genetic parameters including gene diversity (H),

allelic richness (AR), inbreeding coefficient (FIS), and effective population size (Ne).

Changes in these genetic parameters over time can be an indicator of population

expansion or decline and, consequently is an indirect effectiveness measure of threatening

process management at the release site, which is fundamental to the success of mammal

translocations in Australia (Short, 2009, Moseby et al., 2011). If the translocated

population shows a decline in levels of genetic variation and/or Ne overtime, further

translocation of animals from the donor populations is necessary to prevent further loss

(e.g. Chapter 3).

If a translocation is founded from multiple isolated source populations, it is important to

monitor admixture between source population linages and the fitness outcomes of genetic

mixing. To monitor admixture, population structure needs to be monitored over multiple

generations (e.g. Chapter 3 – 5). Phenotypic measurements of different generations (i.e.

purebred, F1, F2, and backcross), which can be identified by molecular markers, can

provide an insight to the effects of genetic mixing on phenotypic variation (Chapters 4

and 5). To assess the outcome of genetic mixing on the fitness of offspring, monitoring

at least two to three generations of progeny is recommended because fitness declines may

not occur until the second (F2 or backcross) or later generations from disruption of co-

adapted gene complexes by recombination (Huff et al., 2011).

6.6.3 Population size and long-term persistence

It has been recommended that newly established populations should reach Ne of at least

1000 to maintain adequate adaptive potential or at least 100 to avoid inbreeding

depression (Willi et al., 2006, Frankham et al., 2014). However, this is unrealistic for

141

many mammal translocations in Australia, including those in this thesis, where small

islands and fenced enclosures are commonly used. For example, taking into account the

Ne/Nc ratio in island dibblers, a population size of approximately 3,000 individuals would

be required to maintain adaptive potential, which is well exceeded the carrying capacity

of the Jurien Bay islands (Chapter 4). Nevertheless, without interventions such as assisted

gene flow through translocations, the ability of the dibbler populations to cope with

environmental change will erode with time and this will reduce their long-term viability.

6.7 STUDY LIMITATIONS AND FUTURE RESEARCH

One common limitation shared in many translocation studies is that post-release

monitoring does not allow for the assessment of individual fitness within populations. To

measure fitness accurately in wild populations, extensive population monitoring is

required that includes measurements of longevity, reproductive success, size and

condition over several generations, and these data were not consistently available in my

study. Further, the different translocations I investigated, each had a different set of

challenges. For instance, although the burrowing bettong translocation had extensive data

from monitoring, the identification of offspring-mother pairs was very limited. This

information is helpful for paternity testing, which is needed to determine a males’

reproductive success. Due to the time constraints of my study, and the early stage of the

burrowing bettong translocation, it was not possible to investigate female reproductive

fitness because sample sizes of F1, F2, and backcrosses were too small. According to

Chapter 5, the mark-recapture analysis showed that there were also variations in the

probability of animals being captured, which subsequently affected the estimates of

survivorship of different offspring groups. These variations could due to the different

trapping protocols, a trial release of animals outside the fence, and/or unexpected floods,

which occurred at the Lorna Glen translocation site.

In the Escape Island translocation, the majority of the release animals were sub-adults so

a few analyses of captive dibblers were constrained by small adult sample sizes (Chapter

4). Furthermore, getting access to Escape Island was difficult (Friend, pers. comm.).

Consequently the Escape Island dibbler population was monitored irregularly and only

prior to the mating season (December to February) instead of during the post-mating

season (May to October) when monitoring was done for the Boullanger and Whitlock

Island source populations. As the result, data on litter size was not available for

comparison between the source and translocated populations (Chapter 4). In addition, it

142

was not possible to compare longevity of different offspring groups because both dibblers

and burrowing bettongs can live 3 to 5 years of age, and it is difficult to age an individual

after it reaches sexual maturity (Chapters 4 and 5).

For future studies, I recommend that conservation practitioners work in partnership with

population geneticists to formulate specific study questions and design appropriate

sampling methods prior to the translocation being carried out. Where possible, tissue

samples and other associated data should be collected from source populations for several

generations before translocation to enable baseline conditions to be assessed. Collected

tissues should be centrally archived to avoid the loss of important samples. Further, there

is still a lack of literature on long-term consequences of outbreeding depression. Most

translocation studies stop at the F1 generation, with only a few extending to F2 or F3

generations (Erickson and Fenster, 2006, Edmands, 1999, Fenster and Galloway, 2000,

Goldberg et al., 2005, Huff et al., 2011). Outbreeding depression consequences such as

lower growth rate (Huff et al., 2011), increased susceptibility to diseases (Goldberg et al.,

2005), and reduced survivorship (Edmands et al., 2005) can reduce Ne and further induce

founder effects, genetic drift and loss of allelic diversity. Beyond this, it is little known

whether fitness declines will continue due to further disruption of co-adapted gene

complexes or increase due to selection promoting beneficial gene combinations. A

hybridization study between divergent copepod populations over 15 generations by

Edmand et al. (2005) showed a rapid recovery from outbreeding depression in F2 hybrids.

The authors suggested that outbreeding depression is likely to be temporary and

incompatible gene interactions being rapidly purged by natural selection (Edmands et al.,

2005). In a 50 year study of a hybrid sunflower population showed shifting in morphology

occurred over 40 years and pollen viability reached its peak in the first 10-15 years and

was maintained thereafter until the end of the study (Carney et al., 2000). If recovery from

outbreeding depression is possible, it is crucial to maintain a large population size (>1000)

to allow natural selection to remove deleterious fitness effects over time (Weeks et al.,

2011). Deleterious effects of outbreeding depression can also be mitigated by different

management strategies (Weeks et al., 2011). For example, selective removal of crossed

individuals and introduce one of parental stocks to encourage backcrossing may avert the

immediate effects of outbreeding depression (Weeks et al., 2011). However, we also need

more studies to test the effectiveness of these management strategies.

143

6.8 CONCLUSION

Mammal declines across Australia have forced conservation managers to use

translocations in an effort to avoid further species extinctions. While normally considered

a last resort, translocations have now become commonplace, particularly in Western

Australia, yet we understand very little about the long-term genetic consequences of these

programs. This thesis has attempted to address this knowledge gap by collecting empirical

data on the genetic outcomes of translocations involving multiple source populations.

My study also showed that fine-scale population structure and species biology such as the

mating system should be taken into consideration when planning captive breeding and

translocation programs. These factors can greatly influence the levels of genetic variation

within newly established populations. While establishing a new population with animals

from multiple source populations can bolster genetic variation, reduce relatedness, and

reduce inbreeding (except Chapter 5), the translocated populations examined here still

experienced a significant reduction in effective population size.

The genetic composition of translocated populations was influenced by mortality,

variation in reproductive success, the timing of release and the origins of founding

individuals. Further, these factors ultimately impact upon the levels of genetic variation

within the translocated population. To avoid a loss of genetic diversity, a new population

should expand quickly to thousands of individuals. If this cannot be achieved within

several generations, habitat connectivity or animal supplementations (< 20% of the

translocated population size) are recommended to prevent further loss of genetic diversity

and to avoid upsetting local adaptation (Hedrick, 1995). Lastly, we need more long-term

studies particularly in translocations sourcing from diverged populations to understand

the consequences of admixture and to inform conservation agencies on the most effective

strategies for successful animal translocations. Examples of such studies include Florida

panthers (Hedrick, 1995, Hostetler et al., 2010), genetic rescue of the mountain pygmy

possums (Weeks et al., 2015), and the burrowing bettong at Lorna Glen from this thesis

(Chapter 5).

144

REFERENCES

AARS, J., IMS, R. A., LIU, H. P., MULVEY, M. & SMITH, M. H. 1998. Bank voles in

linear habitats show restricted gene flow as revealed by mitochondrial DNA

(mtDNA). Molecular Ecology, 7, 1383-1389.

AB GHANI, N. I., HERCZEG, G. & MERILÄ, J. 2012. Body size divergence in nine-

spined sticklebacks: disentangling additive genetic and maternal effects.

Biological Journal of the Linnean Society, 107, 521-528.

ABBOTT, I. 2000. Improving the conservation of threatened and rare mammal species

through translocation to islands: case study Western Australia. Biological

Conservation, 93, 195-201.

ADAMS, J. R., VUCETICH, L. M., HEDRICK, P. W., PETERSON, R. O. &

VUCETICH, J. A. 2011. Genomic sweep and potential genetic rescue during

limiting environmental conditions in an isolated wolf population. Proceedings -

Royal Society. Biological sciences, 278, 3336-3344.

ALEXANDRINO, J., BAIRD, S. J. E., LAWSON, L., MACEY, J. R., MORITZ, C. &

WAKE, D. B. 2005. Strong selection against hybrids at a hybrid zone in the

Ensatina ring species complex and its evolutionary implications. Evolution, 59,

1334-1347.

ALLENDORF, F. W. 1986. Genetic drift and the loss of alleles versus heterozygosity.

Zoo Biology, 5, 181-190.

ALLENDORF, F. W., ENGLAND, P. R., LUIKART, G., RITCHIE, P. A. & RYMAN,

N. 2008. Genetic effects of harvest on wild animal populations. Trends in Ecology

and Evolution, 23, 327-337.

ALLENDORF, F. W., LEARY, R. F., SPRUELL, P. & WENBURG, J. K. 2001. The

problems with hybrids: setting conservation guidelines. Trends in Ecology and

Evolution, 16, 613-622.

ALLENDORF, F. W. & LUIKART, G. 2007. Conservation and the genetics of

populations, Carlton, Blackwell Publishing.

ALLENDORF, F. W., LUIKART, G. H. & AITKEN, S. N. 2012. Conservation and the

genetics of populations, Hoboken, Wiley.

ÁLVAREZ, D. & GARCIA-VAZQUEZ, E. 2011. Maintenance of asymmetric

hybridization between Atlantic salmon (Salmo salar) and brown trout (Salmo

trutta) via postzygotic barriers and paternal effects. Canadian Journal of

Fisheries & Aquatic Sciences, 68, 593-602.

ÁLVAREZ, D., LOURENÇO, A., ORO, D. & VELO-ANTÓN, G. 2015. Assessment of

census (N) and effective population size (Ne) reveals consistency of Ne single-

sample estimators and a high Ne/N ratio in an urban and isolated population of

fire salamanders. Conservation Genetics Resources, 7, 705-712.

ANDERSEN, A. N., WOINARSKI, J. C. Z. & PARR, C. L. 2012. Savanna burning for

biodiversity: fire management for faunal conservation in Australian tropical

savannas. Austral Ecology, 37, 658-667.

ANDERSON, D. R., BURNHAM, K. P. & WHITE, G. C. 1994. AIC model selection in

overdispersed capture-recapture data. Ecology, 75, 1780-1793.

ANDERSON, E. & THOMPSON, E. 2002. A model-based method for identifying

species hybrids using multilocus genetic data. Genetics, 160, 1217-1229.

ANTAO, T., LOPES, A., LOPES, R. J., BEJA-PEREIRA, A. & LUIKART, G. 2008.

LOSITAN: A workbench to detect molecular adaptation based on a FST-outlier

method. BMC Bioinformatics, 9, 323-323.

ARNTZEN, J. W., JEHLE, R., BARDAKCI, F., BURKE, T. & WALLIS, G. P. 2009.

Asymmetric viability of reciprocal-cross hybrids between crested and marbled

newts (Triturus cristatus and T. marmoratus). Evolution, 63, 1191-1202.

145

BANKS, S. C., FINLAYSON, G. R., LAWSON, S. J., LINDENMAYER, D. B.,

PAETKAU, D., WARD, S. J. & TAYLOR, A. C. 2005. The effects of habitat

fragmentation due to forestry plantation establishment on the demography and

genetic variation of a marsupial carnivore, Antechinus agilis. Biological


BANKS, S. C. & PEAKALL, R. 2012. Genetic spatial autocorrelation can readily detect

sex-biased dispersal. Molecular Ecology, 21, 2092-2105.

BAYNES, A. 1987. The original mammal fauna of the Nullarbor and southern peripheral

regions: Evidence from skeletal remains in superficial cave deposits. In:

MCKENZIE, N. L. & ROBINSON, A. C. (eds.) A biological survey of the

Nullarbor region South and Western Australia in 1984. Adelaide: S.A.

Department of Environment and Planning, Adelaide and W.A. Department of

Conservation and Land Management, Perth.

BAYNES, A. 1990. The mammal of Shark Bay, Western Australia. In: BERRY, P. F.,

BRADSHAW, S. D. & WILSON, B. R. (eds.) Research in Shark Bay. Report of

the France-Australe Expedition Committee. Perth: Western Australian Museum.

BEAUMONT, M. A. & NICHOLS, R. A. 1996. Evaluating loci for use in the genetic

analysis of population structure. Proceedings: Biological Sciences, 263, 1619-

1626.

BECK, B. B., RAPAPORT, L. G., STANLEY PRICE, M. R. & WILSON, A. C. 1994.

Reintroduction of captive-born animals. In: OLNEY, P. J. S., MACE, G. M. &

FEISTNER, A. T. C. (eds.) Creative Conservation: Interactive management of

wild and captive animals. London: Chapman & Hall.

BENCINI, R., MCCULLOCH, C., MILLS, H. R. & START, A. N. 2001. Habitat and

diet of the dibbler (Parantechinus apicalis) on two islands in Jurien Bay, Western

Australia. Wildlife Research, 28, 465-468.

BIEBACH, I. & KELLER, L. F. 2009. A strong genetic footprint of the re-introduction

history of Alpine ibex (Capra ibex ibex). Molecular Ecology, 18, 5046-5058.

BIEBACH, I. & KELLER, L. F. 2012. Genetic variation depends more on admixture than

number of founders in reintroduced Alpine ibex populations. Biological


BIJLSMA, R., BUNDGAARD, J. & BOEREMA, A. C. 2000. Does inbreeding affect the

extinction risk of small populations? Predictions from Drosophila. Journal of

Evolutionary Biology, 13, 502-514.

BILSKI, D. R., PIE, M. R. & PASSOS, F. C. 2013. Variable inbreeding effects across

life-history stages in a captive carnivorous mammal population. Animal


BINKS, R. M., KENNINGTON, J. W. & JOHNSON, M. S. 2007. Rapid evolutionary

responses in a translocated population of intertidal snail (Bembicium vittatum)

utilise variation from different source populations. Conservation Genetics, 8,

1421-1429.

BOLNICK, D. I., TURELLI, M., LÓPEZ-FERNÁNDEZ, H., WAINWRIGHT, P. C. &

NEAR, T. J. 2008. Accelerated mitochondrial evolution and "Darwin's

Corollary": Asymmetric viability of reciprocal F1 hybrids in Centrarchid fishes.

Genetics, 178, 1037-1048.

BOS, D. G. & CARTHEW, S. M. 2008. Patterns of movement in the small dasyurid

(Ningaui yvonneae). Australian Journal of Zoology, 55, 299-307.

BRODERS, H. G., MAHONEY, S. P., MONTEVECCHI, W. A. & DAVIDSON, W. S.

1999. Population genetic structure and the effect of founder events on the genetic

variability of moose, Alces alces, in Canada. Molecular Ecology, 8, 1309-1315.

BRODIE, E. D. 1992. Correlational selection for color pattern and antipredator behavior

in the garter snake Thamnophis ordinoides. Evolution, 46, 1284-1298.

146

BRYANT, E. H. & REED, D. H. 1999. Fitness decline under relaxed selection in captive

populations. Conservation Biology, 13, 665-669.

BUCKLAND, S. T., BURNHAM, K. P. & AUGUSTIN, N. H. 1997. Model selection:

An integral part of inference. Biometrics, 53, 603-618.

BURBIDGE, A. A., JOHNSON, K. A., FULLER, P. J. & SOUTHGATE, R. I. 1988.

Aboriginal knowledge of the mammals of the central deserts of Australia. Wildlife

Research, 15, 9-39.

BURBIDGE, A. A. & MCKENZIE, N. L. 1989. Patterns in the modern decline of western

Australia's vertebrate fauna: causes and conservation implications. Biological


BURBIDGE, A. A., MCKENZIE, N. L., BRENNAN, K. E. C., WOINARSKI, J. C. Z.,

DICKMAN, C. R., BAYNES, A., GORDON, G., MENKHORST, P. W. &

ROBINSON, A. C. 2009. Conservation status and biogeography of Australia’s

terrestrial mammals. Australian Journal of Zoology, 56, 411-422.

BURBIDGE, A. A. & SHORT, J. C. 1995. Burrowing bettong. In: STRAHAN, R. (ed.)

The Mammals of Australia. Chatswood, NSW: Australian Museum and Reed

Books.

BURNHAM, K. P. & ANDERSON, D. R. 2004. Multimodel inference: understanding

AIC and BIC in model selection. Sociological Methods & Research, 33, 261-304.

CAMPBELL, D. R. & WASER, N. M. 2001. Genotype-by-environment interaction and

the fitness of plant hybrids in the wild. Evolution, 55, 669-676.

CARDOSO, M. J., ELDRIDGE, M. D. B., OAKWOOD, M., RANKMORE, B.,

SHERWIN, W. B. & FIRESTONE, K. B. 2009. Effects of founder events on the

genetic variation of translocated island populations: implications for conservation

management of the northern quoll. Conservation Genetics, 10, 1719-1733.

CARNEY, S. E., GARDNER, K. A. & RIESEBERG, L. H. 2000. Evolutionary changes

over the fitty-year history of a hybrid population of sunflowers (Helianthus).

Evolution, 54, 462-474.

CHALMERS, C. E. & DAVIES, S. M. 1984. Coastal management plan-Jurien Bay area.

Bulletin 176. In: DEPARTMENT OF CONSERVATION AND

ENVIRONMENT (ed.). Western Australia: Department of Conservation and

Environment,.

CHAPMAN, A. D. 2009. Numbers of living species in Australia and the world. In:

DEPARTMENT OF THE ENVIRONMENT WATER HERITAGE AND THE

ARTS (ed.) Second ed. Toowoomba, Australia: Australian Biological Resources

Study (ABRS).

CHAPMAN, A. D. & NEWBEY, K. R. 1995. A biological survey of the Fitzgerald area,

Western Australia. CALMScience, Supplement 3, 1-258.

CHARLESWORTH, D. & WILLIS, J. H. 2009. The genetics of inbreeding depression.

Nature Reviews Genetics, 10, 783-796.

CLINCHY, M., TAYLOR, A. C., ZANETTE, L. Y., KREBS, C. J. & JARMAN, P. J.

2004. Body size, age and paternity in common brushtail possums (Trichosurus

vulpecula). Molecular Ecology, 13, 195-202.

CLOSE, R. L. & LOWRY, P. S. 1989. Hybrids in marsupial research. Australian Journal

of Zoology, 37, 259-267.

COCKBURN, A., SCOTT, M. P. & SCOTTS, D. J. 1985. Inbreeding avoidance and

male-biased natal dispersal in Antechinus spp. (Marsupialia: Dasyuridae). Animal

Behaviour, 33, 908-915.

COLTMAN, D. W., PILKINGTON, J. G., SMITH, J. A. & PEMBERTON, J. M. 1999.

Parasite-mediated selection against inbred Soay sheep in a free-living, island

population. Evolution, 53, 1259-1267.

147

COOCH, W. & WHITE, G. C. 1997. Program Mark: A gentle introduction. 13 ed.:

http://www.phidot.org/software/mark/docs/book/.

COYNE, J. A. & ORR, H. A. 2004. Speciation, Sunderland, MA., Sinauer Association.

CRNOKRAK, P. & ROFF, D. A. 1999. Inbreeding depression in the wild. Heredity, 83,

260-270.

CROTEAU, E. K., HEIST, E. J. & NIELSEN, C. K. 2010. Fine-scale population structure

and sex-biased dispersal in bobcats (Lynx rufus) from southern Illinois. Canadian

Journal of Zoology, 88, 536-545.

DE BARBA, M., WAITS, L. P., GARTON, E. O., GENOVESI, P., RANDI, E.,

MUSTONI, A. & GROFF, C. 2010. The power of genetic monitoring for studying

demography, ecology and genetics of a reintroduced brown bear population.

Molecular Ecology, 19, 3938-3951.

DEC no date. Operation Rangelands Restoration.

http://www.dec.wa.gov.au/content/view/3055/2147/.

DEYOUNG, R. W., DEMARAIS, S., HONEYCUTT, R. L., ROONEY, A. P.,

GONZALES, R. A. & GEE, K. L. 2003. Genetic consequences of white-tailed

deer (Odocoileus virginianus) restoration in Mississippi. Molecular Ecology, 12,

3237-3252.

DIECKMANN, U., O'HARA, B. & WEISSER, W. 1999. The evolutionary ecology of

dispersal. Trends in Ecology and Evolution, 14, 88-90.

DONALDSON, F. R. & VERCOE, P. E. 2008. Cross-family amplification:

microsatellites isolated from Macropodidae are polymorphic in Potoroidae.

Molecular Ecology Resources, 8, 452-454.

DORTCH, C. E. & MORSE, K. 1984. Prehistoric stone artefacts on some offshore islands

in Western Australia. Australian Archaeology, 19, 31-47.

DOUBLE, M. C., PEAKALL, R., BECK, N. R. & COCKBURN, A. 2005. Dispersal,

philopatry, and infidelity: dissecting local genetic structure in superb fairy-wrens

(Malurus cyaneus). Evolution, 59, 625-635.

EARNHARDT, J. M. 1999. Reintroduction programmes: genetic trade-offs for

populations. Animal Conservation, 2, 279-286.

EDMANDS, S. 1999. Heterosis and outbreeding depression in interpopulation crosses

spanning a wide range of divergence. Evolution, 53, 1757-1768.

EDMANDS, S. 2002. Does parental divergence predict reproductive compatibility?

Trends in Ecology and Evolution, 17, 520-527.

EDMANDS, S. 2007. Between a rock and a hard place: evaluating the relative risks of

inbreeding and outbreeding for conservation and management. Molecular

Ecology, 16, 463-475.

EDMANDS, S., FEAMAN, H. V., HARRISON, J. S. & TIMMERMAN, C. C. 2005.

Genetic consequences of many generations of hybridization between divergent

copepod populations. Journal of Heredity, 96, 114-123.

EDMANDS, S. & TIMMERMAN, C. C. 2003. Modeling factors affecting the severity

of outbreeding depression. Conservation Biology, 17, 883-892.

ELDRIDGE, M. D. B., KING, J. M., LOUPIS, A. K., SPENCER, P. B. S., TAYLOR, A.

C., POPE, L. C. & HALL, G. P. 1999. Unprecedented low levels of genetic

variation and inbreeding depression in an island population of the black-footed

rock-wallaby. Conservation Biology, 13, 531-541.

ELLEGREN, H. 1999. Inbreeding and relatedness in Scandinavian grey wolves Canis

Lupus. Hereditas, 130, 239-244.

ERICKSON, D. L. & FENSTER, C. B. 2006. Intraspecific hybridization and the recovery

of fitness in the native legume Chamaecrista fasciculata. Evolution, 60, 225-233.

http://www.phidot.org/software/mark/docs/book/

http://www.dec.wa.gov.au/content/view/3055/2147/

148

ERIKSSON, J., HOHMANN, G., BOESCH, C. & VIGILANT, L. 2004. Rivers influence

the population genetic structure of bonobos (Pan paniscus). Molecular Ecology,

13, 3425-3435.

EVANNO, G., REGNAUT, S. & GOUDET, J. 2005a. Detecting the number of clusters

of individuals using the software STRUCTURE: a simulation study. Molecular

ecology, 14, 2611-2620.

EVANNO, G., REGNAUT, S. & GOUDET, J. 2005b. Detecting the number of clusters

of individuals using the software STRUCTURE: A simulation study. Molecular

Ecology, 14, 2611-2620.

EXCOFFIER, L., LAVAL, G. & SCHNEIDER, S. 2005. Arlequin (version 3.0): an

integrated software package for population genetics data analysis. Evolutionary

Bioinformatics, 1, 47-50.

FENSTER, C. B. & GALLOWAY, L. F. 2000. Inbreeding and outbreeding depression in

natural populations of Chamaecrista fasciculata (Fabaceae). Conservation

Biology, 14, 1406-1412.

FINLAYSON, G. R. & MOSEBY, K. E. 2004. Managing confined populations: the

influence of density on the home range and habitat use of reintroduced burrowing

bettongs (Bettongia lesueur). Wildlife Research, 31, 457-463.

FINLAYSON, H. H. 1961. On central Australian mammals. Part IV - The distribution

and status of central Australian species. Records of the South Australian Museum,

14, 141-191.

FIRESTONE, K. B. 1999. Isolation and characterization of microsatellites from

carnivorous marsupials (Dasyuridae: Marsupialia). Molecular Ecology, 8, 1084-

1086.

FISCHER, J. & LINDENMAYER, D. B. 2000. An assessment of the published results of

animal relocations. Biological Conservation, 96, 1-11.

FISHER, D. O. 2005. Population density and presence of the mother are related to natal

dispersal in male and female Antechinus stuartii. Australian Journal of Zoology,

53, 103-110.

FISHER, D. O. & LARA, M. C. 1999. Effects of body size and home range on access to

mates and paternity in male bridled nailtail wallabies. Animal Behaviour, 58, 121-

130.

FISHMAN, L. & WILLIS, J. H. 2001. Evidence for Dobzhansky-Muller incompatibilites

contributing to the sterility of hybrids between Mimulus guttatus and M. nasutus.

Evolution, 55, 1932-1942.

FITZSIMMONS, N. N., BUSKIRK, S. W. & SMITH, M. H. 1997. Genetic changes in

reintroduced rocky mountain bighorn sheep populations. The Journal of Wildlife

Management, 61, 863-872.

FORBES, S. H. & HOGG, J. T. 1999. Assessing population structure at high levels of

differentiation: microsatellite comparisons of bighorn sheep and large carnivores.

Animal Conservation, 2, 223-233.

FRANKHAM, R. 1995. Effective population size/adult population size ratios in wildlife:

a review. Genetics Research, 66, 95-107.

FRANKHAM, R. 1996. Relationship of genetic variation to population size in wildlife.

Conservation Biology, 10, 1500-1508.

FRANKHAM, R. 1998. Inbreeding and extinction: island populations. Conservation

Biology, 12, 665-675.

FRANKHAM, R. 2008. Genetic adaptation to captivity in species conservation programs.


FRANKHAM, R. 2015. Genetic rescue of small inbred populations: Meta-analysis

reveals large and consistent benefits of gene flow. Molecular Ecology, 24, 2610-

2618.

149

FRANKHAM, R., BALLOU, J. D., ELDRIDGE, M. D. B., LACY, R. C., RALLS, K.,

DUDASH, M. R. & FENSTER, C. B. 2011. Predicting the probability of

outbreeding depression. Conservation Biology, 25, 465-475.

FRANKHAM, R., BRADSHAW, C. J. A. & BROOK, B. W. 2014. Genetics in

conservation management: revised recommendations for the 50/500 rules, Red

List criteria and population viability analyses. Biological Conservation, 170, 56-

63.

FREEGARD, C. J., CALVER, M. C., RICHARDS, J. D. & BRADLEY, S. 2008. Age

estimation of pouch young and growth of the burrowing bettong Bettongia lesueur

(Marsupialia: Potoroidae) on Heirisson Prong, Shark Bay, Western Australia.

Australian Mammalogy, 29, 157-167.

FRIEND, T. 2003. Dibbler (Parantechinus apicalis) Recovery Plan July 2003-June 2013.

Wildlife Management Program No. 38. Perth: Department of Conservation and

Land Management.

FRIEND, T., BURBIDGE, A. & MORRIS, K. 2008. Parantechinus apicalis. The IUCN

Red List of Threatened Species. Version 2014.3 [Online]. www.iucnredlist.org.

[Accessed 31/12/2014.

FRIEND, T. & COLLINS, P. Male die-off in the dibbler: no evidence in mainland

populations (Abstract). Australian Mammal Society: 51st Scientific Meeting

Program and Abstracts, 2005 Albany, Western Australia.

FULLER, P. J. & BURBIDGE, A. A. 1987. Discovery of the Dibbler, Parantechinus

apicalis, on islands at Jurien Bay. The Western Australian Naturalist, 16, 177-

181.

FUMAGALLI, L., POPE, L. C., TABERLET, P. & MORITZ, C. 1997. Versatile primers

for the amplification of the mitochondrial DNA control region in marsupials.


GARAVANTA, C. A. M., WOOLLER, R. D. & RICHARDSON, K. C. 2000. Movement

patterns of honey possums, Tarsipes rostratus, in the Fitzgerald River National

Park, Western Australia. Wildlife Research, 27, 179-183.

GAUTSCHI, B., WIDMER, A., JOSHI, J. & KOELLA, J. 2002. Increased frequency of

scale anomalies and loss of genetic variation in serially bottlenecked populations

of the dice snake, Natrix tessellata. Conservation Genetics, 3, 235-245.

GHARRETT, A. J., SMOKER, W. W., REISENBICHLER, R. R. & TAYLOR, S. G.

1999. Outbreeding depression in hybrids between odd- and even-broodyear pink

salmon. Aquaculture, 173, 117-129.

GOLDBERG, T. L., GRANT, E. C., INENDINO, K. R., KASSLER, T. W., CLAUSSEN,

J. E. & PHILIPP, D. P. 2005. Increased infectious disease susceptibility resulting

from outbreeding depression. Conservation Biology, 19, 455-462.

GOOSSENS, B., CHIKHI, L., JALIL, M. F., ANCRENAZ, M., LACKMAN-

ANCRENAZ, I., MOHAMED, M., ANDAU, P. & BRUFORD, M. W. 2005.

Patterns of genetic diversity and migration in increasingly fragmented and

declining orang-utan (Pongo pygmaeus) populations from Sabah, Malaysia.


GOTO, S., IIJIMA, H., OGAWA, H. & OHYA, K. 2011. Outbreeding depression caused

by intraspecific hybridization between local and nonlocal genotypes in Abies

sachalinensis. Restoration Ecology, 19, 243-250.

GOUDET, J. 2001. FSTAT, a program to estimate and test gene diversities and fixation

indices Version 2.9.3. http://www2.unil.ch/popgen/softwares/fstat.htm.

GREENWOOD, P. J. 1980. Mating systems, philopatry and dispersal in birds and

mammals. Animal Behaviour, 28, 1140-1162.

GREGORY, A. J., KALER, R. S., PREBYL, T. J., SANDERCOCK, B. K. & WISELY,

S. M. 2012. Influence of translocation strategy and mating system on the genetic

http://www.iucnredlist.org/

http://www2.unil.ch/popgen/softwares/fstat.htm

150

structure of a newly established population of island ptarmigan. Conservation

Genetics, 13, 465-474.

GRUEBER, C. E., LAWS, R. J., NAKAGAWA, S. & JAMIESON, I. G. 2010.

Inbreeding depression accumulation across life-history stages of the endangered

Takahe. Conservation Biology, 24, 1617-1625.

GUILLOT, G., MORTIER, F. & ESTOUP, A. 2005. Geneland: A computer package for

landscape genetics. Molecular Ecology Notes, 5, 712-715.

HARTL, D. L. & CLARK, A. G. 1997. Principles of population genetics, Massachusetts,

Sinauer associates, Inc.

HATHAWAY, L., ANDERSSON, S. & PRENTICE, H. C. 2009. Experimental crosses

within European Silene latifolia (Caryophyllaceae): intraspecific differentiation,

distance effects, and sex ratio. Botany, 87, 231-240.

HAZLITT, S. L., ELDRIDGE, M. D. B. & GOLDIZEN, A. W. 2004. Fine-scale spatial

genetic correlation analyses reveal strong female philopatry within a brush-tailed

rock-wallaby colony in southeast Queensland. Molecular Ecology, 13, 3621-

3632.

HAZLITT, S. L., GOLDIZEN, A. W. & ELDRIDGE, M. D. B. 2006. Significant patterns

of population genetic structure and limited gene flow in a threatened macropodid

marsupial despite continuous habitat in southeast Queensland, Australia.

Conservation Genetics, 7, 675-689.

HEDRICK, P. W. 1995. Gene flow and genetic restoration: the Florida panther as a case

study. Conservation Biology, 9, 996-1007.

HEDRICK, P. W. 2000. Genetics of populations, London, Jones and Bartlett publishers,

Inc.

HEDRICK, P. W. 2013. Adaptive introgression in animals: examples and comparison to

new mutation and standing variation as sources of adaptive variation. Molecular

Ecology, 22, 4606-4618.

HEDRICK, P. W. & FREDRICKSON, R. 2010. Genetic rescue guidelines with examples

from Mexican wolves and Florida panthers. Conservation Genetics, 11, 615-626.

HEDRICK, P. W. & FREDRICKSON, R. J. 2008. Captive breeding and the

reintroduction of Mexican and red wolves. Molecular Ecology, 17, 344-350.

HEDRICK, P. W., GUTIERREZ-ESPELETA, G. A. & LEE, R. N. 2001. Founder effect

in an island population of bighorn sheep. Molecular Ecology, 10, 851-857.

HEDRICK, P. W. & KALINOWSKI, S. T. 2000. Inbreeding depression in conservation

biology. Annual Review of Ecology and Systematics, 31, 139-162.

HEDRICK, P. W., MILLER, P. S., GEFFEN, E. & WAYNE, R. 1997. Genetic evaluation

of the three captive mexican wolf lineages. Zoo Biology, 16, 47-69.

HEMMINGS, N. L., SLATE, J. & BIRKHEAD, T. R. 2012. Inbreeding causes early

death in a passerine bird. Nature Communications, 3, 1-4.

HENDRY, A. P., NOSIL, P. & RIESEBERG, L. H. 2007. The speed of ecological

speciation. Functional Ecology, 21, 455-464.

HOGG, J. T., FORBES, S. H., STEELE, B. M. & LUIKART, G. 2006. Genetic rescue of

an insular population of large mammals. Proceedings of the Royal Society B:

Biological sciences, 273, 1491-1499.

HOLDEREGGER, R. & WAGNER, H. H. 2008. Landscape genetics. BioScience, 58,

199-207.

HOLLELEY, C. E., DICKMAN, C. R., CROWTHER, M. S. & OLDROYD, B. P. 2006.

Size breeds success: Multiple paternity, multivariate selection and male

semelparity in a small marsupial, Antechinus stuartii. Molecular Ecology, 15,

3439-3448.

HOSTETLER, J. A., ONORATO, D. P., NICHOLS, J. D., JOHNSON, W. E., ROELKE,

M. E., O’BRIEN, S. J., JANSEN, D. & OLI, M. K. 2010. Genetic introgression

151

and the survival of Florida panther kittens. Biological Conservation, 143, 2789-

2796.

HUFF, D. D., MILLER, L. M., CHIZINSKI, C. J. & VONDRACEK, B. 2011. Mixed-

source reintroductions lead to outbreeding depression in second-generation

descendents of a native North American fish. Molecular Ecology, 20, 4246-4258.

HUFF, D. D., MILLER, L. M. & VONDRACEK, B. 2010. Patterns of ancestry and

genetic diversity in reintroduced populations of the slimy sculpin: implications for

conservation. Conservation Genetics, 11, 2379-2391.

HUFFORD, K. M., KRAUSS, S. L. & VENEKLAAS, E. J. 2012. Inbreeding and

outbreeding depression in Stylidium hispidum: implications for mixing seed

sources for ecological restoration. Ecology and Evolution, 2, 2262-2273.

HUNDERTMARK, K. J. & VAN DAELE, L. J. 2010. Founder effect and bottleneck

signatures in an introduced, insular population of elk. Conservation Genetics, 11,

139-147.

HUXTABLE, S. J., LEE, D. V., WISE, P. & THE SAVE THE TASMANIAN DEVIL

PROGRAM 2015. Metapopulation management of an extreme disease scenario.

In: ARMSTRONG, D., HAYWARD, M. W., MORO, D. & SEDDON, P. J. (eds.)

Advances in reintroduction biology. Clayton, South Victoria: CSIRO publishing.

HYNES, E. F., RUDD, C. D., TEMPLE-SMITH, P. D., SOFRONIDIS, G., PARIS, D.,

SHAW, G. & RENFREE, M. B. 2005. Mating sequence, dominance and paternity

success in captive male tammar wallabies. Reproduction, 130, 123-130.

IUCN. 1987. The IUCN position statement on translocation of living organisms:

Introductions, reintroductions and re-stocking. Prepared by the Species Survival

Commission in collaboration with the Commission on Ecology, and the

Commission on Environmental Policy, Law and Administration [Online].

IUCN 2015. The IUCN Redlist of Threatened Species. Version 2015.1. 01/06/2015 ed.:

http://www.iucnredlist.org.

IUCN/SSC 2013. Guidelines for reintroductions and other conservation translocations.

Gland, Switzerland: IUCN Species Survival Commission, viiii + 57 pp.

JAKOBSSON, M. & ROSENBERG, N. A. 2007. CLUMPP: a cluster matching and

permutation program for dealing with label switching and multimodality in

analysis of population structure. Bioinformatics, 23, 1801-1806.

JAMIESON, I. G. 2011. Founder effects, inbreeding, and loss of genetic diversity in four

avian reintroduction programs. Conservation Biology, 25, 115-123.

JARMAN, P. J. 1991. Social behavior and organization in the Macropodoidea. In:

SLATER, P. J. B., ROSENBLATT, J. S., BEER, C. & MILINSKI, M. (eds.)

Advances in the Study of Behavior. Academic Press.

JENKINS, D. G., BRESCACIN, C. R., DUXBURY, C. V., ELLIOTT, J. A., EVANS, J.

A., GRABLOW, K. R., HILLEGASS, M., LYON, B. N., METZGER, G. A.,

OLANDESE, M. L., PEPE, D., SILVERS, G. A., SURESCH, H. N.,

THOMPSON, T. N., TREXLER, C. M., WILLIAMS, G. E., WILLIAMS, N. C.

& WILLIAMS, S. E. 2007. Does size matter for dispersal distance? Global

Ecology and Biogeography, 16, 415-425.

JOHNSON, W. E., ONORATO, D. P., ROELKE, M. E., LAND, E. D., CUNNINGHAM,

M., BELDEN, R. C., MCBRIDE, R., JANSEN, D., LOTZ, M., SHINDLE, D.,

HOWARD, J., WILDT, D. E., PENFOLD, L. M., HOSTETLER, J. A., OLI, M.

K. & O'BRIEN, S. J. 2010. Genetic restoration of the Florida panther. Science,

329, 1641-1645.

JOMBART, T. 2008. adegenet: a R package for the multivariate analysis of genetic

markers. Bioinformatics, 24, 1403-1405.

http://www.iucnredlist.org/

152

JOMBART, T., DEVILLARD, S. & BALLOUX, F. 2010. Discriminant analysis of

principal components: a new method for the analysis of genetically structured

populations. BMC Genetics, 11, 94-94.

JONES, M. E., PAETKAU, D., GEFFEN, E. & MORITZ, C. 2003. Microsatellites for

the Tasmanian devil (Sarcophilus laniarius). Molecular Ecology Notes, 3, 277-

279.

JULE, K. R., LEAVER, L. A. & LEA, S. E. G. 2008. The effects of captive experience

on reintroduction survival in carnivores: a review and analysis. Biological


KAYS, R., CURTIS, A. & KIRCHMAN, J. J. 2010. Rapid adaptive evolution of

northeastern coyotes via hybridization with wolves. Biology Letters, 6, 89-93.

KEARSEY, M. J. & POONI, H. S. 1996. The genetical analysis of quantitative traits,

London, Chapman and Hall.

KELLER, L. F. & WALLER, D. M. 2002. Inbreeding effects in wild populations. Trends

in Ecology and Evolution, 17, 230-241.

KENNINGTON, J. W., HEVROY, T. H. & JOHNSON, M. S. 2012. Long-term genetic

monitoring reveals contrasting changes in the genetic composition of newly

established populations of the intertidal snail Bembicium vittatum. Molecular

Ecology, 21, 3489-3500.

KIMURA, M. 1962. On the probability of fixation of mutant genes in a population.

Genetics, 47, 713-719.

KOENIG, W. D., VAN VUREN, D. & HOOGE, P. N. 1996. Detectability, philopatry,

and the distribution of dispersal distances in vertebrates. Trends in Ecology and

Evolution, 11, 514-517.

KOLBE, J. J., GLOR, R. E., SCHETTINO, L. R., LARA, A. C., LARSON, A. & LOSOS,

J. B. 2007. Multiple sources, admixture, and genetic variation in introduced Anolis

lizard populations. Conservation Biology, 21, 1612-1625.

KÖRTNER, G., PAVEY, C. R. & GEISER, F. 2007. Spatial ecology of the mulgara in

arid Australia: impact of fire history on home range size and burrow use. Journal

of Zoology, 273, 350-357.

KRAAIJEVELD-SMIT, F. J. L., LINDENMAYER, D. B., TAYLOR, A. C.,

MACGREGOR, C. & WERTHEIM, B. 2007. Comparative genetic structure

reflects underlying life histories of three sympatric small mammal species in

continuous forest of south-eastern Australia. Oikos, 116, 1819-1830.

KRAAIJEVELD-SMIT, F. J. L., WARD, S. J. & TEMPLE-SMITH, P. D. 2002a.

Multiple paternity in a field population of a small carnivorous marsupial, the agile

antechinus, Antechinus agilis. Behavioral Ecology and Sociobiology, 52, 84-91.

KRAAIJEVELD-SMIT, F. J. L., WARD, S. J. & TEMPLE-SMITH, P. D. 2003. Paternity

success and the direction of sexual selection in a field population of a semelparous

marsupial, Antechinus agilis. Molecular Ecology, 12, 475-484.

KRAAIJEVELD-SMIT, F. J. L., WARD, S. J., TEMPLE-SMITH, P. D. & PAETKAU,

D. 2002b. Factors influencing paternity success in Antechinus agilis: last-male

sperm precedence, timing of mating and genetic compatibility. Journal of

Evolutionary Biology, 15, 100-107.

KREBS, C. J. 1966. Demographic changes in fluctuating populations of Microtus

californicus. Ecological Monographs, 36, 239-273.

LACY, R. C. & POLLAK, J. P. 2014. Vortex: A stochastic simulation of the extinction

process. Version 10.0. Brookfield, Illinois USA: Chicago Zoological Society.

LADA, H., MAC NALLY, R. & TAYLOR, A. C. 2008. Distinguishing past from present

gene flow along and across a river: the case of the carnivorous marsupial

(Antechinus flavipes) on southern Australian floodplains. Conservation Genetics,

9, 569-580.

153

LAIKRE, L. & RYMAN, N. 1991. Inbreeding depression in a captive wolf (Canis lupus)

population. Conservation Biology, 5, 33-40.

LAMBERT, C. & MILLS, H. 2006. Husbandry and breeding of the Dibbler

Parantechinus apicalis at Perth Zoo. International Zoo Yearbook, 40, 290-301.

LAMPERT, K., RAND, A., MUELLER, U. & RYAN, M. 2003. Fine-scale genetic

pattern and evidence for sex-biased dispersal in the tungara frog, Physalaemus

pustulosus. Molecular Ecology, 12, 3325-3334.

LANCASTER, M. L., BRADSHAW, C. J. A., GOLDSWORTHY, S. D. & SUNNUCKS,

P. 2007. Lower reproductive success in hybrid fur seal males indicates fitness

costs to hybridization. Molecular Ecology, 16, 3187-3197.

LARSON, S., JAMESON, R., BODKIN, J., STAEDLER, M. & BENTZEN, P. 2002.

Microsatellite DNA and mitochondrial DNA variation in remnant and

translocated sea otter (Enhydra lutris) populations. Journal of Mammalogy, 83,

893-906.

LATCH, E. K., HARVESON, L. A., KING, J. S., HOBSON, M. D. & RHODES, O. E.

2006. Assessing hybridization in wildlife populations using molecular markers: A

case study in wild turkeys. The Journal of Wildlife Management, 70, 485-492.

LAWSON HANDLEY, L. J. & PERRIN, N. 2007. Advances in our understanding of

mammalian sex-biased dispersal. Molecular Ecology, 16, 1559-1578.

LEBERG, P. L. 1992. Effects of population bottlenecks on genetic diversity as measured

by allozyme electrophoresis. Evolution, 46, 477-494.

LEMER, S. & PLANES, S. 2012. Translocation of wild populations: conservation

implications for the genetic diversity of the black-lipped pearl oyster Pinctada

margaritifera. Molecular Ecology, 21, 2949-2962.

LENORMAND, T. 2002. Gene flow and the limits to natural selection. Trends in Ecology

and Evolution, 17, 183-189.

LEVY, E., KENNINGTON, J. W., TOMKINS, J. L. & LEBAS, N. R. 2010. Land

clearing reduces gene flow in the granite outcrop-dwelling lizard, Ctenophorus

ornatus. Molecular Ecology, 19, 4192-4203.

LEVY, E., TOMKINS, J. L., LEBAS, N. R. & KENNINGTON, W. J. 2013. Contrasting

effects of landscape features on genetic structure in different geographic regions

in the ornate dragon lizard, Ctenophorus ornatus. Molecular Ecology, 22, 3904-

3915.

LIBRADO, P. & ROZAS, J. 2009. DnaSP v5: a software for comprehensive analysis of

DNA polymorphism data. Bioinformatics, 25, 1451-1452.

LIND, C. E., EVANS, B. S., TAYLOR, J. J. U. & JERRY, D. R. 2010. The consequences

of differential family survival rates and equalizing maternal contributions on the

effective population size (Ne) of cultured silver-lipped pearl oysters, Pinctada

maxima. Aquaculture Research, 41, 1229-1242.

LOMOLINO, M. V. 2005. Body size evolution in insular vertebrates: generality of the

island rule. Journal of Biogeography, 32, 1683-1699.

LÓPEZ-FERNÁNDEZ, H. & BOLNICK, D. I. 2007. What causes partial F1 hybrid

viability? Incomplete penetrance versus genetic variation. PLoS ONE, 2, e1294.

LUIKART, G., ALLENDORF, F. W., CORNUET, J. M. & SHERWIN, W. B. 1998.

Distortion of allele frequency distributions provides a test for recent population

bottlenecks. Journal of Heredity, 89, 238-247.

LUIKART, G. & CORNUET, J.-M. 1998. Empirical evaluation of a test for identifying

recently bottlenecked populations from allele frequency data. Conservation

Biology, 12, 228-237.

LUIKART, G., PAINTER, J., CROZIER, R. H., WESTERMAN, M. & SHERWIN, W.

B. 1997. Characterization of microsatellite loci in the endangered long-footed

potoroo Potorous longipes. Molecular Ecology, 6, 497-498.

154

LYNCH, M. 1991. The genetic interpretation of inbreeding depression and outbreeding

depression. Evolution, 45, 622-629.

LYNCH, M. & RITLAND, K. 1999. Estimation of pairwise relatedness with molecular

markers. Genetics, 152, 1753-1766.

MADSEN, T., STILLE, B. & SHINE, R. 1996. Inbreeding depression in an isolated

population of adders Vipera berus. Biological Conservation, 75, 113-118.

MAGUIRE, L. A. & LACY, R. C. 1990. Allocating scarce resources for conservation of

endangered subspecies: partitioning zoo space for tigers. Conservation Biology,

4, 157-166.

MANSERGH, I., HEINZE, D., WEEKS, A. & PERRIN, L. 2013. Gene-pool mixing:

Lessons from the high country (online). Wildlife Australia, 50, 32-35.

MARSHALL, T. C., SLATE, J., KRUUK, L. E. B. & PEMBERTON, J. M. 1998.

Statistical confidence for likelihood-based paternity inference in natural

populations. Molecular Ecology, 7, 639-655.

MARUYAMA, T. & FUERST, P. A. 1985. Population bottlenecks and nonequilibrium

models in population genetics. II. Number of alleles in a small population that was

formed by a recent bottleneck. Genetics, 111, 675-689.

MASTERS, P. 2003. Movement patterns and spatial organisation of the mulgara,

Dasycercus cristicauda (Marsupialia : Dasyuridae), in central Australia. Wildlife

Research, 30, 339-344.

MCNAB, B. K. 2002. Minimizing energy expenditure facilitates vertebrate persistence

on oceanic islands. Ecology Letters, 5, 693-704.

MCNEELY, J. A., MILLER, K. R., REID, W. V., MITTERMEIER, R. A. & WERNER,

T. B. 1990. Conserving the world's biological diversity, Washington DC, USA,

International Union for the Conservation of Nature and Natural Resources, World

Resources Institute, Conservation International, and WWF - US and World Bank.

METCALF, J. L., SIEGUE, M. R. & MARTIN, A. P. 2008. Hybridization dynamics

between Colorado's native cutthroat trout and introduced rainbow trout. Journal

of heredity, 99, 149-156.

MILLER, E., ELDRIDGE, M., MORRIS, K., ZENGER, K. & HERBERT, C. 2011a.

Genetic consequences of isolation: island tammar wallaby (Macropus eugenii)

populations and the conservation of threatened species. Conservation Genetics,

12, 1619-1631.

MILLER, E. J., ELDRIDGE, M. D. B., COOPER, D. W. & HERBERT, C. A. 2010a.

Dominance, body size and internal relatedness influence male reproductive

success in eastern grey kangaroos (Macropus giganteus). Reproduction, Fertility

and Development, 22, 539-549.

MILLER, H. C. & LAMBERT, D. M. 2004. Genetic drift outweighs balancing selection

in shaping post-bottleneck major histocompatibility complex variation in New

Zealand robins (Petroicidae). Molecular Ecology, 13, 3709-3721.

MILLER, J. M., POISSANT, J., HOGG, J. T. & COLTMAN, D. W. 2012. Genomic

consequences of genetic rescue in an insular population of bighorn sheep (Ovis

canadensis). Molecular Ecology, 21, 1583-1596.

MILLER, K. A., NELSON, N. J., SMITH, H. G. & MOORE, J. A. 2009. How do

reproductive skew and founder group size affect genetic diversity in reintroduced

populations? Molecular Ecology, 18, 3792-3802.

MILLER, K. A., TOWNS, D. R., ALLENDORF, F. W., RITCHIE, P. A. & NELSON,

N. J. 2011b. Genetic structure and individual performance following a recent

founding event in a small lizard. Conservation Genetics, 12, 461-473.

MILLER, S., BENCINI, R., MILLS, H. & MORO, D. 2003. Food availability for the

dibbler (Parantechinus apicalis) on Boullanger and Whitlock Islands, Western

Australia. Wildlife Research, 30, 649-649.

155

MILLER, W., WRIGHT, S. J., ZHANG, Y., SCHUSTER, S. C. & HAYES, V. M. 2010b.

Optimization methods for selecting founder individuals for captive breeding or

reintroduction of endangered species. Pacific Symposium on Biocomputing 2010,

43-53.

MILLS, H. & BENCINI, R. 2000. New evidence for facultative male die-off in island

populations of dibblers, Parantechinus apicalis. Australian Journal of Zoology,

48, 501-510.

MILLS, H. R., BRADSHAW, F. J., LAMBERT, C., BRADSHAW, S. D. & BENCINI,

R. 2012. Reproduction in the marsupial dibbler, Parantechinus apicalis;

differences between island and mainland populations. General and Comparative

Endocrinology, 178, 347-354.

MILLS, H. R., MORO, D. & SPENCER, P. B. S. 2004. Conservation significance of

island versus mainland populations: a case study of dibblers (Parantechinus

apicalis) in Western Australia. Animal Conservation, 7, 387-395.

MILLS, H. R. & SPENCER, P. B. S. 2003. Polymorphic microsatellites identified in an

endangered dasyurid marsupial, the dibbler (Parantechinus apicalis). Molecular

Ecology Notes, 3, 218-220.

MITCHELL, A. A., LAU, J., CHEMNICK, L. G., THOMPSON, E. A., ALBERTS, A.

C., RYDER, O. A. & GERBER, G. P. 2011. Using microsatellite diversity in wild

Anegada iguanas (Cyclura pinguis) to establish relatedness in a captive breeding

group of this critically endangered species. Conservation Genetics, 12, 771-781.

MITROVSKI, P., HOFFMANN, A. A., HEINZE, D. A. & WEEKS, A. R. 2008. Rapid

loss of genetic variation in an endangered possum. Biology Letters, 4, 134-138.

MOCK, K. E., LATCH, E. K. & RHODES, O. E. 2004. Assessing losses of genetic

diversity due to translocation: long-term case histories in Merriam's turkey

(Meleagris gallopavo merriami). Conservation Genetics, 5, 631-645.

MOORE, S., CAVANA, M., GILLEN, K., HART, C., HOPPER, S., ORR, K. &

SCHMIDT, W. 1991. Fitzgerald River National Park Management Plan 1991-

2001. In: DEPARTMENT OF CONSERVATION AND LAND

MANAGEMENT (ed.). Perth: Department of Conservation and Land

Management.

MOORHOUSE, T. P., GELLING, M. & MACDONALD, D. W. 2009. Effects of habitat

quality upon reintroduction success in water voles: evidence from a replicated

experiment. Biological Conservation, 142, 53-60.

MORCOMBE, M. K. 1967. The rediscovery after 83 years of the Dibbler Antechinus

apicalis (Marsupialia, Dasyuridae). The Western Australian Naturalist, 10, 103-

111.

MORITZ, C. 1999. Conservation units and translocations: strategies for conserving

evolutionary processes. Hereditas, 130, 217-228.

MORO, D. 2003. Translocation of captive-bred dibblers Parantechinus apicalis

(Marsupialia: Dasyuridae) to Escape Island, Western Australia. Biological


MOSEBY, K. E., READ, J. L., PATON, D. C., COPLEY, P., HILL, B. M. & CRISP, H.

A. 2011. Predation determines the outcome of 10 reintroduction attempts in arid

South Australia. Biological Conservation, 144, 2863-2872.

NEAVES, L. E., ZENGER, K. R., PRINCE, R. I. T., ELDRIDGE, M. D. B. & COOPER,

D. W. 2009. Landscape discontinuities influence gene flow and genetic structure

in a large, vagile Australian mammal, Macropus fuliginosus. Molecular Ecology,

18, 3363-3378.

NEI, M., MARUYAMA, T. & CHAKRABORTY, R. 1975. The bottleneck effect and

genetic variability in populations. Evolution, 29, 1-10.

156

NELSON, N. J., KEALL, S. N., BROWN, D. & DAUGHERTY, C. H. 2002. Establishing

a new wild population of tuatara (Sphenodon guntheri). Conservation Biology, 16,

887-894.

NEVEU, H., HAFEN, T., ZIMMERMANN, E. & RUMPLER, Y. 1998. Comparison of

the genetic diversity of wild and captive groups of Microcebus murinus using the

random amplified polymorphic DNA method. Folia Primatologica, 69, 127-135.

NEWMAN, D. & PILSON, D. 1997. Increased probability of extinction due to decreased

genetic effective population size: experimental populations of Clarkia pulchella.

Evolution, 51, 354-362.

NIELSEN, J. F., ENGLISH, S., GOODALL-COPESTAKE, W. P., WANG, J.,

WALLING, C. A., BATEMAN, A. W., FLOWER, T. P., SUTCLIFFE, R. L.,

SAMSON, J., THAVARAJAH, N. K., KRUUK, L. E. B., CLUTTON-BROCK,

T. H. & PEMBERTON, J. M. 2012. Inbreeding and inbreeding depression of early

life traits in a cooperative mammal. Molecular Ecology, 21, 2788-2804.

O’GRADY, J. J., BROOK, B. W., REED, D. H., BALLOU, J. D., TONKYN, D. W. &

FRANKHAM, R. 2006. Realistic levels of inbreeding depression strongly affect

extinction risk in wild populations. Biological Conservation, 133, 42-51.

OTTEWELL, K., DUNLOP, J., THOMAS, N., MORRIS, K., COATES, D. & BYRNE,

M. 2014. Evaluating success of translocations in maintaining genetic diversity in

a threatened mammal. Biological Conservation, 171, 209-219.

PALSTRA, F. P. & RUZZANTE, D. E. 2008. Genetic estimates of contemporary

effective population size: what can they tell us about the importance of genetic

stochasticity for wild population persistence? Molecular Ecology, 17, 3428-3447.

PARROTT, M. L., WARD, S. J., TEMPLE-SMITH, P. D. & SELWOOD, L. 2015.

Females choose mates based on genetic relatedness in a small Dasyurid marsupial,

the Agile Antechinus (Antechinus agilis). PLoS ONE, 10, 1-16.

PAVEY, C. R., GOODSHIP, N. & GEISER, F. 2003. Home range and spatial

organisation of rock-dwelling carnivorous marsupial, Pseudantechinus

macdonnellensis. Wildlife Research, 30, 135-142.

PEAKALL, R., RUIBAL, M. & LINDENMAYER, D. B. 2003. Spatial autocorrelation

analysis offers new insights into gene flow in the Australian bush rat, Rattus

fuscipes. Evolution, 57, 1182-1195.

PEAKALL, R. & SMOUSE, P. E. 2012. GenAlEx 6.5: genetic analysis in Excel.

Population genetic software for teaching and research—an update.

Bioinformatics, 28, 2537-2539.

PERRIER, C., NORMANDEAU, É., DIONNE, M., RICHARD, A. & BERNATCHEZ,

L. 2014. Alternative reproductive tactics increase effective population size and

decrease inbreeding in wild Atlantic salmon. Evolutionary Applications, 7, 1094-

1106.

PERTH ZOO 2013. Dibbler. http://perthzoo.wa.gov.au/conservation/native-species-

breeding-programs/dibbler/.

PFAU, R. S., VAN DEN BUSSCHE, R. A. & MCBEE, K. 2001. Population genetics of

the hispid cotton rat (Sigmodon hispidus): patterns of genetic diversity at the major

histocompatibility complex. Molecular Ecology, 10, 1939-1945.

PICKUP, M., FIELD, D. L., ROWELL, D. M. & YOUNG, A. G. 2013. Source population

characteristics affect heterosis following genetic rescue of fragmented plant

populations. Proceedings - Royal Society. Biological sciences, 280, 2012-2058.

PIERTNEY, S. B., MACCOLL, A. D. C., BACON, P. J. & DALLAS, J. F. 1998. Local

genetic structure in red grouse (Lagopus lagopus scoticus): evidence from

microsatellite DNA markers. Molecular Ecology, 7, 1645-1654.

PIMM, S. L., DOLLAR, L. & BASS, O. L. 2006. The genetic rescue of the Florida

panther. Animal Conservation, 9, 115-122.

http://perthzoo.wa.gov.au/conservation/native-species-breeding-programs/dibbler/

http://perthzoo.wa.gov.au/conservation/native-species-breeding-programs/dibbler/

157

PIRY, S., LUIKART, G. & CORNUET, J. M. 1999. BOTTLENECK: a computer

program for detecting recent reductions in the effective size using allele frequency

data. Journal of Heredity, 90, 502-503.

POMPANON, F., BONIN, A., BELLEMAIN, E. & TABERLET, P. 2005. Genotyping

errors: Causes, consequences and solutions. Nature Reviews Genetics, 6, 847-859.

POPE, L. C., ESTOUP, A. & MORITZ, C. 2000. Phylogeography and population

structure of an ecotonal marsupial, Bettongia tropica, determined using mtDNA

and microsatellites. Molecular Ecology, 9, 2041-2053.

POPE, L. C., SHARP, A. & MORITZ, C. 1996. Population structure of the yellow-footed

rock-wallaby Petrogale xanthopus (Gray, 1854) inferred from mtDNA sequences

and microsatellite loci. Molecular Ecology, 5, 629-640.

POPE, L. C., VERNES, K., GOLDIZEN, A. W. & JOHNSON, C. N. 2012. Mating

system and local dispersal patterns of an endangered potoroid, the northern

bettong (Bettongia tropica). Australian Journal of Zoology, 60, 278-287.

POTTER, S., ELDRIDGE, M. D. B., COOPER, S. J. B., PAPLINSKA, J. Z. &

TAGGART, D. A. 2012. Habitat connectivity, more than species' biology,

influences genetic differentiation in a habitat specialist, the short-eared rock-

wallaby (Petrogale brachyotis). Conservation Genetics, 13, 937-952.

PRITCHARD, J., STEPHENS, M. & DONNELLY, P. 2000. Inference of population

structure using multilocus genotype data. Genetics, 155, 945 - 959.

QUEMERE, E., CROUAU-ROY, B., RABARIVOLA, C., LOUIS JR, E. E. & CHIKHI,

L. 2010. Landscape genetics of an endangered lemur (Propithecus tattersalli)

within its entire fragmented range. Molecular Ecology, 19, 1606-1621.

R CORE TEAM 2014. R: A language and environment for statistical computing. Vienna,

Austria: R Foundation for Statistical Computing.

RAISIN, C., FRANTZ, A. C., KUNDU, S., GREENWOOD, A. G., JONES, C. G.,

ZUEL, N. & GROOMBRIDGE, J. J. 2012. Genetic consequences of intensive

conservation management for the Mauritius parakeet. Conservation Genetics, 13,

707-715.

RAMBAUT, A. 2012. FigTree Version 1.4.0 Tree Figure Drawing Tool. Institute of

Evolutionary Biology, University of Edinburgh: http://tree.bio.ed.ac.uk/.

RAMSTAD, K. M., COLBOURNE, R. M., ROBERTSON, H. A., ALLENDORF, F. W.

& DAUGHERTY, C. H. 2013. Genetic consequences of a century of protection:

serial founder events and survival of the little spotted kiwi (Apteryx owenii).

Proceedings of the Royal Society of London B: Biological Sciences, 280, 1-10.

RANSLER, F. A., QUINN, T. W. & OYLER-MCCANCE, S. J. 2011. Genetic

consequences of trumpeter swan (Cygnus buccinator) reintroductions.

Conservation Genetics, 12, 257-268.

RHYMER, J. M. & SIMBERLOFF, D. 1996. Extinction by hybridization and

introgression. Annual Review of Ecology and Systematics, 27, 83-109.

RICE, W. R. 1989. Analyzing tables of statistical tests. Evolution, 43, 223-225.

RICHARDS, C. M. 2000. Inbreeding depression and genetic rescue in a plant

metapopulation. The American Naturalist, 155, 383-394.

RICHARDS, J., MORRIS, K. & BURBIDGE, A. 2008a. Bettongia lesueur. The IUCN

Red List of Threatened Species [Online].

http://www.iucnredlist.org/details/2784/0. [Accessed 09/12/2014.

RICHARDS, J., MORRIS, K. & BURBIDGE, A. 2008b. Bettongia lesueur. The IUCN

Red List of Threatened Species 2008.

http://dx.doi.org/10.2305/IUCN.UK.2008.RLTS.T2784A9480530.en.

RICHARDS, J. D. 2007. Western Barred Bandicoot, Burrowing Bettong and Banded

Hare-Wallaby Recovery Plan 2007-2011. Wildlife Management Program No. 49.

Perth: Department of Environment and Conservation.

http://tree.bio.ed.ac.uk/

http://www.iucnredlist.org/details/2784/0

http://dx.doi.org/10.2305/IUCN.UK.2008.RLTS.T2784A9480530.en

158

RICHARDS, J. D. 2012. Western Barred Bandicoot Perameles bougainville, Burrowing

Bettong Bettongia lesueur and Banded Hare-Wallaby Lagostrophus fasciatus

National Recovery Plan. Wildlife Management Program No. 49. Perth:

Department of Environment and Conservation.

ROBERT, A. 2009. Captive breeding genetics and reintroduction success. Biological


ROBERTS, D. G., GRAY, C. A., WEST, R. J. & AYRE, D. J. 2010. Marine genetic

swamping: hybrids replace an obligately estuarine fish. Molecular Ecology, 19,

508-520.

ROLÁN-ALVAREZ, E., ERLANDSSON, J., JOHANNESSON, K. & CRUZ, R. 1999.

Mechanisms of incomplete prezygotic reproductive isolation in an intertidal snail:

testing behavioural models in wild populations. Journal of Evolutionary Biology,

12, 879-890.

RONQUIST, F., TESLENKO, M., VAN DER MARK, P., AYRES, D. L., DARLING,

A., HOHNA, S., LARGET, B., LIU, L., SUCHARD, M. A. & HUELSENBECK,

J. P. 2012. MrBayes 3.2: Efficient bayesian phylogenetic inference and model

choice across a large model space. Systematic Biology, 61, 539-542.

ROSENBERG, N. A. 2004. Distruct: a program for the graphical display of population

structure. Molecular Ecology Notes, 4, 137-138.

RYMAN, N. & LAIKRE, L. 1991. Effects of supportive breeding on the genetically

effective population size. Conservation Biology, 5, 325-329.

SANDER, U., SHORT, J. & TURNER, B. 1997. Social organisation and warren use of

the burrowing bettong, Bettongia lesueur (Macropodoidea: Potoroidae). Wildlife

Research, 24, 143-157.

SAURA, M., CABALLERO, A., CABALLERO, P. & MORAN, P. 2008. Impact of

precocious male parr on the effective size of a wild population of Atlantic salmon.

Freshwater Biology, 53, 2375-2384.

SCHWARTZ, M. K., LUIKART, G. & WAPLES, R. S. 2007. Genetic monitoring as a

promising tool for conservation and management. Trends in Ecology and

Evolution, 22, 25-33.

SCHWARTZ, M. K. & MCKELVEY, K. S. 2009. Why sampling scheme matters: the

effect of sampling scheme on landscape genetic results. Conservation Genetics,

10, 441-452.

SCRIBNER, K. T. & CHESSER, R. K. 1993. Environmental and demographic correlates

of spatial and seasonal genetic structure in the Eastern cottontail (Sylvilagus

floridanus). Journal of Mammalogy, 74, 1026-1044.

SHEEAN, V. A., MANNING, A. D. & LINDENMAYER, D. B. 2012. An assessment of

scientific approaches towards species relocations in Australia. Austral Ecology,

37, 204-215.

SHERWIN, W. S. & BROWN, P. R. 1990. Problems in the estimation of the effective

size of a population of the Eastern Barred bandicoot (Perameles gunnii) at

Hamilton, Victoria. In: SEEBECK, J. H., BROWN, P. R., WALLIS, R. L. &

KEMPER, C. M. (eds.) Bandicoots and Bilbies. Chipping Norton, NSW: Surrey

Beatty & Sons.

SHORT, J. 2009. The characteristics and success of vertebrate translocations within

Australia: a progress report to Department of Agriculture, Fisheries and Forestry.

Western Australia: Wildlife Research and Management.

SHORT, J. & TURNER, B. 1993. The distribution and abundance of the burrowing

bettong (Marsupialia: Macropodoidea). Wildlife Research, 20, 525-534.

SHORT, J. & TURNER, B. 1999. Ecology of burrowing bettongs, Bettongia lesueur

(Marsupialia: Potoroidae), on Dorre and Bernier Islands, Western Australia.

Wildlife Research, 26, 651-669.

159

SHORT, J. & TURNER, B. 2000. Reintroduction of the burrowing bettong Bettongia

lesueur (Marsupialia: Potoroidae) to mainland Australia. Biological


SIGG, D. P. 2006. Reduced genetic diversity and significant genetic differentiation after

translocation: Comparison of the remnant and translocated populations of bridled

nailtail wallabies (Onychogalea fraenata). Conservation Genetics, 7, 577-589.

SIGG, D. P., GOLDIZEN, A. W. & POPLE, A. R. 2005. The important of maitng system

in translocation programs: reproductive success of released male bridled nailtail

wallabies. Biological Conservation, 123, 289-300.

SLATE, J., KRUUK, L. E. B., MARSHALL, T. C., PEMBERTON, J. M. & CLUTTON-

BROCK, T. H. 2000. Inbreeding depression influences lifetime breeding success

in a wild population of red deer (Cervus elaphus). Proceedings of the Royal

Society of London. Series B: Biological Sciences, 267, 1657-1662.

SMITH, A. P. & QUIN, D. G. 1996. Patterns and causes of extinction and decline in

Australian conilurine rodents. Biological Conservation, 77, 243-267.

SMITH, S. & HUGHES, J. 2008. Microsatellite and mitochondrial DNA variation defines

island genetic reservoirs for reintroductions of an endangered Australian

marsupial, Perameles bougainville. Conservation Genetics, 9, 547-557.

SODERQUIST, T. & LILL, A. 1995. Natal dispersal and philopatry in the carnivorous

marsupial Phascogale tapoatafa (Dasyuridae). Ethology, 99, 297-312.

SPENCER, P. B. S., CARDOSO, M., HOW, R. A., WILLIAMS, J., BUNCE, M. &

SCHMITT, L. H. 2007. Cross-species amplification at microsatellite loci in

Australian quolls including the description of five new markers from the Chuditch

(Dasyurus geoffroii). Molecular Ecology Notes, 7, 1100-1103.

SPENCER, P. B. S., ODORICO, D. M., JONES, S. J., MARSH, H. D. & MILLER, D. J.

1995. Highly variable microsatellites in isolated colonies of the rock-wallaby

(Petrogale assimilis). Molecular Ecology, 4, 523-525.

START, A. & BACZOCHA, N. 1997. Status and Ecology of the Dibbler (Parantechinus

apicalis) in Western Australia: Project Number 496: 1996 Annual Report.

Department of Conservation and Land Management.

STIEBENS, V. A., MERINO, S. E., RODER, C., CHAIN, F. J. J., LEE, P. L. M. &

EIZAGUIRRE, C. 2013. Living on the edge: How philopatry maintains adaptive

potential. Proceedings - Royal Society. Biological sciences, 280, 1-9.

STOCKWELL, C. A., MULVEY, M. & VINYARD, G. L. 1996. Translocations and the

preservation of allelic diversity. Conservation Biology, 10, 1133-1141.

STRAHAN, R. 1983. The Australian Museum complete book of Australian mammals,

London-Sydney-Melbourne, Angus and Robertson Publishers.

SUNNUCKS, P. & HALES, D. F. 1996. Numerous transposed sequences of

mitochondrial cytochrome oxidase I-II in aphids of the genus Sitobion

(Hemiptera: Aphididae). Molecular Biology and Evolution, 13, 510-524.

SWINNERTON, K. J., GROOMBRIDGE, J. J., JONES, C. G., BURN, R. W. &

MUNGROO, Y. 2004. Inbreeding depression and founder diversity among

captive and free-living populations of the endangered pink pigeon Columba

mayeri. Animal Conservation, 7, 353-364.

TALLMON, D. A., LUIKART, G. & WAPLES, R. S. 2004. The alluring simplicity and

complex reality of genetic rescue. Trends in Ecology and Evolution, 19, 489-496.

TAMURA, K., PETERSON, D., PETERSON, N., STECHER, G., NEI, M. & KUMAR,

S. 2011. MEGA5: Molecular evolutionary genetics analysis using maximum

likelihood, evolutionary distance, and maximum parsimony methods. Molecular

Biology and Evolution, 28, 2731-2739.

160

TAYLOR, A. C., SHERWIN, W. B. & WAYNE, R. K. 1994. Genetic variation of

microsatellite loci in a bottlenecked species: the northern hairy-nosed wombat

Lasiorhinus krefftii. Molecular Ecology, 3, 277-290.

TAYLOR, E. B., BOUGHMAN, J. W., GROENENBOOM, M., SNIATYNSKI, M.,

SCHLUTER, D. & GOW, J. L. 2006. Speciation in reverse: morphological and

genetic evidence of the collapse of a three-spined stickleback (Gasterosteus

aculeatus) species pair. Molecular Ecology, 15, 343-355.

TEIXEIRA, C. P., DE AZEVEDO, C. S., MENDL, M., CIPRESTE, C. F. & YOUNG,

R. J. 2007. Revisiting translocation and reintroduction programmes: the

importance of considering stress. Animal Behaviour, 73, 1-13.

THOMPSON, J. D., GIBSON, T. J., PLEWNIAK, F., JEANMOUGIN, F. & HIGGINS,

D. G. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple

sequence alignment aided by quality analysis tools. Nucleic acids research, 25,

4876-4882.

THOMPSON, S. L., LAMOTHE, M., MEIRMANS, P. G., PERINET, P. & ISABEL, N.

2010. Repeated unidirectional introgression towards Populus balsamifera in

contact zones of exotic and native poplars. Molecular Ecology, 19, 132-145.

TREVARROW, B. & ROBISON, B. 2009. Genetic backgrounds, standard lines, and their

hunsbandry. In: DETRICH, H. W., MONTE, W. & ZON, L. I. (eds.) Essential

zebrafish methods: genetics and genomics. Oxford, UK: Elsevier Inc.

TRINKEL, M., FERGUSON, N., REID, A., REID, C., SOMERS, M., TURELLI, L.,

GRAF, J., SZYKMAN, M., COOPER, D., HAVERMAN, P., KASTBERGER,

G., PACKER, C. & SLOTOW, R. 2008. Translocating lions into an inbred lion

population in the Hluhluwe-iMfolozi Park, South Africa. Animal Conservation,

11, 138-143.

TYNDALE BISCOE, C. H. 1968. Reproduction and post-natal development in the

marsupial Bettongia lesueur (Quoy & Gaimard). Australian Journal of Zoology,

16, 577-602.

VAN OOSTERHOUT, C., HUTCHINSON, W., DEREK, P., WILLS, M. & SHIPLEY,

P. 2004. MICRO-CHECKER: Software for identifying and correcting genotyping

errors in microsatellite data. Molecular Ecology Notes, 4, 535-538.

VINES, T. H. & SCHLUTER, D. 2006. Strong assortative mating between allopatric

sticklebacks as a by-product of adaptation to different environments. Proceedings

- Royal Society. Biological sciences, 273, 911-916.

WAPLES, R. S. & DO, C. 2008. LDNe: A program for estimating effective population

size from data on linkage disequilibrium. Molecular Ecology Resources, 8, 753-

756.

WAPLES, R. S. & DO, C. 2010. Linkage disequilibrium estimates of contemporary Ne

using highly variable genetic markers: a largely untapped resource for applied

conservation and evolution. Evolutionary Applications, 3, 244-262.

WATER AND RIVERS COMMISSION 2003. State of the waterways in the Fitzgerald

River catchment.

WAYNE, R. K., BRUFORD, M. W., GIRMAN, B. D., REBHOLZ, W. E. R.,

SUNNUCKS, P. & TAYLOR, A. C. 1994. Molecular genetics of endangered

species. In: OLNEY, P. J. S., MACE, G. M. & FEISTNER, A. T. C. (eds.)

Creative Conservation: Interactive management of wild and captive animals.

London, UK: Chapman & Hall.

WEEKS, A. R., MORO, D., THAVORNKANLAPACHAI, R., TAYLOR, H. R.,

WHITE, N. E., WEISER, E. L. & HEINZE, D. 2015. Conserving and enhancing

genetic diversity in translocation programs. In: ARMSTRONG, D., HAYWARD,

M. W., MORO, D. & SEDDON, P. J. (eds.) Advances in reintroduction biology

of Australian and New Zealand fauna. South Australia: CSIRO Publishing.

161

WEEKS, A. R., SGRO, C. M., YOUNG, A. G., FRANKHAM, R., MITCHELL, N. J.,

MILLER, K. A., BYRNE, M., COATES, D. J., ELDRIDGE, M. D. B.,

SUNNUCKS, P., BREED, M. F., JAMES, E. A. & HOFFMANN, A. A. 2011.

Assessing the benefits and risks of translocations in changing environments: a

genetic perspective. Evolutionary Applications, 4, 709-725.

WEIR, B. S. & COCKERHAM, C. C. 1984. Estimating F-Statistics for the analysis of

population structure. Evolution, 38, 1358-1370.

WHITE, G. C. & BURNHAM, K. P. 1999. Program MARK: survival estimation from

populations of marked animals. Bird Study, 46, 120-139.

WILCOX, K. 2003. The consequences of translocation on the genetic variability of the

endangered dibbler (Parantechinus apicalis) on Escape Island, Western

Australia. Bachelor of Science in Animal Science Dissertation, The University of

Western Australia.

WILLI, Y., KLEUNEN, M. V., DIETRICH, S. & FISCHER, M. 2007. Genetic rescue

persists beyond first-generation outbreeding in small populations of a rare plant.

274, 2357-2364.

WILLI, Y., VAN BUSKIRK, J. & HOFFMANN, A. A. 2006. Limits to the adaptive

potential of small populations. Annual Review of Ecology, Evolution, and

Systematics, 37, 433-458.

WILLIAMS, B. W. & SCRIBNER, K. T. 2010. Effects of multiple founder populations

on spatial genetic structure of reintroduced American martens. Molecular

Ecology, 19, 227-240.

WILLIAMS, S. E. & HOFFMAN, E. A. 2009. Minimizing genetic adaptation in captive

breeding programs: a review. Biological Conservation, 142, 2388-2400.

WITZENBERGER, K. A. & HOCHKIRCH, A. 2008. Genetic consequences of animal

translocations: a case study using the field cricket, Gryllus campestris L.

Biological Conservation, 141, 3059-3068.

WOINARSKI, J. C. Z., BURBIDGE, A. A. & HARRISON, P. L. 2014. Dibbler. The

action plan for Australian mammals 2012. Melbourne: CSIRO Publishing.

WOINARSKI, J. C. Z., MILNE, D. J. & WANGANEEN, G. 2001. Changes in mammal

populations in relatively intact landscapes of Kakadu National Park, Northern

Territory, Australia. Austral Ecology, 26, 360-370.

WOLF, C. M., GARLAND JR, T. & GRIFFITH, B. 1998. Predictors of avian and

mammalian translocation success: reanalysis with phylogenetically independent

contrasts. Biological Conservation, 86, 243-255.

WOLF, C. M., GRIFFITH, B., REED, C. & TEMPLE, S. A. 1996. Avian and mammalian

translocations: update and reanalysis of 1987 survey data. Conservation Biology,

10, 1142-1154.

WOLFE, K., MILLS, H., GARKAKLIS, M. J. & BENCINI, R. 2004. Post-mating

survival in a small marsupial is associated with nutrient inputs from seabirds.

Ecology, 85, 1740-1746.

WOLFE, K. M., ROBERTSON, H. & BENCINI, R. 2000. The mating behaviour of the

dibbler, Parantechinus apicalis, in captivity. Australian Journal of Zoology, 48,

541-550.

WOOLLEY, P. A. 1991. Reproductive pattern of captive Boullanger Island dibbers,

Parantechinus apicalis (Marsupialia : Dasyuridae). Wildlife Research, 18, 157-

163.

WOOLLEY, P. A. 1995. Southern dibbler. In: STRAHAN, R. (ed.) The Mammals of

Australia. Chatswood, NSW: Australian Museum and Reed Books.

WOOLLEY, P. A. 2008. Dibbler. In: VAN DYCK, S. & STRAHAN, R. (eds.) The

mammals of Australia. Third ed. Sydney: New Holland.

162

WRIGHT, S. 1922. Coefficients of inbreeding and relationship. The American Naturalist,

56, 330-338.

WRIGHT, S. 1939. Statistical genetics in relation to evolution (Actualites scientifiques

et industrielles; no. 802), Paris, Hermann & Cie.

WRIGHT, S. 1943. Isolation by distance. Genetics, 28, 114-138.

ZENGER, K. R. & COOPER, D. W. 2001. Characterization of 14 macropod

microsatellite genetic markers. Animal Genetics, 32, 166-167.

ZENGER, K. R., MCKENZIE, L. M. & COOPER, D. W. 2002. The first comprehensive

genetic linkage map of a marsupial: the tammar wallaby (Macropus eugenii).

Genetics, 162, 321-330.

ZENGER, K. R., RICHARDSON, B. J. & VACHOT-GRIFFIN, A. M. 2003. A rapid

population expansion retains genetic diversity within European rabbits in

Australia. Molecular Ecology, 12, 789-794.

163

APPENDICES

Table S2.1 Characteristics of the 21 microsatellite loci that were selected for use in characterizing the genetic variability of the mainland

Parantechinus apicalis.

Locus Species Nature of

repeat

Size (bp) Multiplex Cycle DNA

(10ng/µL

)

PCR

Anneal

Temp

(C°)

Primer sequence (5' - 3') References

pPa2D4 Parantechinus

apicalis

NA 193-203 2 x35 1 56 CAATCTGTCAATAACCTTCCCCC Mills and

Spencer 2003 TGGAGGACCTCCAGAAAGTTAGC

pPa2A12 P. apicalis (GT)21 103-133 2 x35 1 56 ATCCTGGAGAAGAGAAGACCTGC Mills and

Spencer 2003 GTGGCTTATTCCATGCTTGTAGG

pPa2B10 P. apicalis (GT)23 172-182 4 x40 1 58 GAGAAAAAATATGCACAAGCACC Mills and

Spencer 2003 AAGGAGAAAAAGTTAATACCATCCC

pPa7A1 P. apicalis (GAA)85 298-315 2 x35 1 56 CTCCACCTCTCTAGACATGACCC Mills and

Spencer 2003 TTTACTTGCTTTGTACTAGAGGCC

pPa7H9 P. apicalis NA 159-173 2 x35 1 56 AAATAACAACAATAGTTCATTATGT Mills and

Spencer 2003 ATTATTTGCTTACTTTGAAGATATA

pPa9D2 P. apicalis (GT)4AT(GT)

11GC(GT)3

94-103 4 x40 1 58 TGGAAAGCAATATGGTAGAAGTGTG Mills and

Spencer 2003 TTCAAGGGTTCAAAACAACATTCTT

pPa1B10 P. apicalis (GAAA)46 197-236 2 x35 1 56 AAGGAGGGATGGAGGAGGAA Mills and

Spencer 2003 CAGTGTTCGAATGACATTGGCTAC

pPa4B3 P. apicalis (GT)15 125-135 4 x40 1 58 GAAGGACAACATTCCCGATTGT Mills and

Spencer 2003 CCTACCCTAATTGCAAATCCTTTC

pPa8F10 P. apicalis (AC)19 91-103 4 x40 1 58 CAATCTAGGAATCACAGAACTCCC Mills and

Spencer 2003 TTTGCATCTACCTAATTGCGTGT

pDG1A1 Dasyurus

geoffroii

(AG)20 208-224 3 x35 1 54 ATTTGCTTCTTGCTCCCTACAGC Spencer et al.,

2007 TTTCACTCCTTCTGAGTTTATCACC

pDG1H3 D. geoffroii (TG)17 190-204 4 x40 1 58 GTGGATTGACACAATCAGAGTGG Spencer et al.,

2007 GCAATTCCATCTTTATTGCATGC

164

pDG6D5 D. geoffroii (AC)22 101-135 3 x35 1 54 CCTCCAGACAAATGCAACC Spencer et al.,

2007 TCTCTGAATTTACTGATAGTATCTTTGG

3.1.2 Dasyurus spp. (CA)18 177-185 3 x35 1 54 AGGAAACTTCACAAGTGTCGA Firestone, 1999

ATTAATGACTCATCTGTTGTTGG

3.3.1 Dasyurus spp. (CA)20 130-152 3 x35 1 54 CAGCCCTTGAGTCTTGAGATT Firestone, 1999

CATACCACCCCAGGAGTTTC

3.3.2 Dasyurus spp. (CA)21 156-172 1 x35 2 46 AATAGCAGAGACTCGATCC Firestone, 1999

AGCCTTTATTACCTGGGAAG

4.4.2 Dasyurus spp. (CA)19 125-133 2 x35 1 56 GAAATCCAAGCTCATTTTAG Firestone, 1999

AATCAACTCTGGAATGCATC

4.4.10 Dasyurus spp. (CA)29 219-242 3 x35 1 54 AATGCTAGATTTCACTCCC Firestone, 1999

CCTCACATTTCTGGAACTG

Sh3o Sarcophilus

laniarius

(CA)22 195-197 1 x35 2 46 CTCAATGCCAAAGGTATCTTC Jones et al., 2003

CATAGTTCCAAATCACTCTCCAG

Sh6e S. laniarius (CA)6

(A)2(CA)18

166-181 3 x35 1 54 GATTCTAGAAGGGATAGCAAGC Jones et al., 2003

GACACTCCATAGAAATGCACTG

Aa4A Antechinus

agilis

NA 162-174 1 x35 2 46 TTTGATCCTCAGAGACTTGAT Banks et al.,

2005 CCAAATCTACGTAAAATATCC

Aa4J A. agilis NA 164-182 1 x35 2 46 TCTTCAGTCTCTCAATGAGTT Kraaijeveld-Smit

et al., 2002 AGAACACTCTAACAACATCCT

165

Table S2.2 Inferring the value of K, the number of Parantechinus apicalis populations

within the Fitzgerald River National Park, using STRUCTURE. K is the number of

genetic clusters. LnP(K) is the posterior probability of the data for a given K. Ln`(K) is

the model choice criterion. The most likely K is identified following Evanno et al. (2005b)

by selecting the highest mean LnP(K) with the smallest standard deviation. The strength

of signal is indicated by ∆K.

K Replications Mean

LnP(K)

Stdev

LnP(K) Ln'(K) |Ln''(K)| ∆K

1 10 -8880.0 0.2 — — —

2 10 -8716.6 1.3 163.4 139.5 106.5

3 10 -8692.7 23.7 23.9 26.1 1.1

4 10 -8695.0 18.0 -2.3 9.8 0.5

5 10 -8707.1 57.3 -12.1 78.2 1.4

6 10 -8641.0 41.7 66.1 31.2 0.7

7 10 -8606.1 25.0 34.9 111.6 4.5

8 10 -8682.7 58.3 -76.6 80.8 1.4

9 10 -8840.1 77.1 -157.4 1.5 0.0

10 10 -8999.0 206.7 -158.9 — —

166

Table S2.3 Inferring the value of K, the number of Parantechinus apicalis populations

within the Fitzgerald River National Park, using GENELAND. K is the number of genetic

clusters. The most likely K is identified by the most consistent numbers of K.

Run Replications

Average log

posterior probability K

1 10 -8365.0 2

2 10 -8368.1 2

3 10 -8373.1 2

4 10 -8372.3 2

5 10 -8371.9 2

6 10 -8378.7 2

7 10 -8370.8 2

8 10 -8260.3 3

9 10 -8362.7 2

10 10 -8202.6 3

167

Table S2.4 Outcomes of spatial autocorrelation of Parantechinus apicalis genetic cluster

one containing a broad scale data set with a maximum distance of 19 kilometres. The

correlation r is shown for each distance class along with the number of pairwise

comparisons n for the calculation of r, upper U and lower L bounds form the 95%

confidence interval and the upper Ur error bar and lower bounds Lr error bar about r as

determined by bootstrap resampling, the probability of P of a one-tailed test for positive

autocorrelation and the estimated x-intercept. Significant P-values are denoted with

asterisks.

Distance class

(Km) 0-1 1-6 6-7 7-8 8-12 12-14 14-19

n 8212 433 1269 1011 339 836 146

r 0.002 -0.003 -0.005 -0.002 0.000 -0.001 -0.025

U 0.001 0.012 0.005 0.006 0.010 0.006 0.024

L -0.001 -0.015 -0.006 -0.007 -0.010 -0.006 -0.026

P 0.009 0.673 0.950 0.774 0.522 0.683 0.974

Ur 0.005 0.011 0.003 0.007 0.016 0.008 -0.005

Lr -0.002 -0.016 -0.013 -0.012 -0.016 -0.011 -0.045

Intercept 2.976

168

Table S2.5 A fine-scale spatial autocorrelation of Parantechinus apicalis genetic cluster

one containing combined dataset, females and males for distance classes from 100 to 600

meters. The correlation r is shown for each distance class along with the number of

pairwise comparisons n for the calculation of r, upper U and lower L bounds form the

95% confidence interval and the upper Ur error bar and lower bounds Lr error bar about

r as determined by bootstrap resampling, the probability of P of a one-tailed test for

positive autocorrelation and the estimated x-intercept. Significant P-values are denoted

with asterisks.

Distance class (m) 100 200 300 400 500 600

Combined n 284 745 878 853 690 406

r 0.024 0.009 0.004 -0.007 0.000 -0.008

U 0.022 0.011 0.009 0.009 0.011 0.015

L -0.017 -0.011 -0.010 -0.010 -0.011 -0.014

P 0.018 0.059 0.183 0.923 0.490 0.873

Ur 0.048 0.023 0.015 0.005 0.012 0.008

Lr 0.000 -0.003 -0.006 -0.019 -0.011 -0.025

Intercept 0.337

Males n 148 428 528 444 377 149

r -0.002 -0.001 0.005 -0.009 0.004 -0.006

U 0.027 0.015 0.012 0.013 0.015 0.023

L -0.024 -0.013 -0.012 -0.014 -0.015 -0.024

P 0.528 0.557 0.224 0.906 0.261 0.685

Ur 0.028 0.016 0.019 0.008 0.022 0.018

Lr -0.030 -0.016 -0.009 -0.025 -0.012 -0.029

Intercept 0.335

Females n 136 317 350 409 313 257

r 0.051 0.023 0.003 -0.005 -0.005 -0.010

U 0.031 0.019 0.015 0.015 0.016 0.018

L -0.024 -0.018 -0.016 -0.014 -0.017 -0.017

P 0.002 0.007 0.321 0.781 0.738 0.851

Ur 0.092 0.043 0.023 0.012 0.012 0.012

Lr 0.017 0.003 -0.014 -0.022 -0.021 -0.031

Intercept 0.340

169

Table S4.1 Characteristics of the 14 microsatellite loci used in characterizing the genetic variability of the mainland Parantechinus apicalis.

Locus Species Nature of

repeat

Size (bp) Multiplex Cycle DNA

(10ng/µL)

PCR

Anneal

Temp

(C°)

Primer sequence (5' - 3') References

pPa2D4 P. apicalis NA 193-197 2 x35 1 56 CAATCTGTCAATAACCTTCCCCC Mills and Spencer

2003 TGGAGGACCTCCAGAAAGTTAGC

pPa2A12 P. apicalis (GT)21 129-131 2 x35 1 56 ATCCTGGAGAAGAGAAGACCTGC Mills and Spencer

2003 GTGGCTTATTCCATGCTTGTAGG

pPa2B10 P. apicalis (GT)23 176-186 3 x40 2 58 GAGAAAAAATATGCACAAGCACC Mills and Spencer

2003 AAGGAGAAAAAGTTAATACCATCCC

pPa7A1 P. apicalis (GAA)85 298-315 2 x35 1 56 CTCCACCTCTCTAGACATGACCC Mills and Spencer

2003 TTTACTTGCTTTGTACTAGAGGCC

pPa7H9 P. apicalis NA 166 2 x35 1 56 AAATAACAACAATAGTTCATTATGT Mills and Spencer

2003 ATTATTTGCTTACTTTGAAGATATA

pPa1B10 P. apicalis (GAAA)46 220-310 2 x35 1 56 AAGGAGGGATGGAGGAGGAA Mills and Spencer

2003 CAGTGTTCGAATGACATTGGCTAC

pDG1H3 Dasyurus

geoffroii

(TG)17 192-193 3 x40 2 58 GTGGATTGACACAATCAGAGTGG Spencer et al.,

2007 GCAATTCCATCTTTATTGCATGC

3.3.2 Dasyurus spp. (CA)21 158-191 1 x35 2 46 AATAGCAGAGACTCGATCC Firestone, 1999

AGCCTTTATTACCTGGGAAG

4.4.2 Dasyurus spp. (CA)19 127-129 2 x35 1 56 GAAATCCAAGCTCATTTTAG Firestone, 1999

AATCAACTCTGGAATGCATC

4.4.10 Dasyurus spp. (CA)29 221-231 3 x40 2 58 AATGCTAGATTTCACTCCC Firestone, 1999

CCTCACATTTCTGGAACTG

Sh3o Sarcophilus

laniarius

(CA)22 194-196 1 x35 2 46 CTCAATGCCAAAGGTATCTTC Jones et al., 2003

CATAGTTCCAAATCACTCTCCAG

170

Sh6e Sarcophilus

laniarius

(CA)6

(A)2(CA)18

175-182 3 x40 2 58 GATTCTAGAAGGGATAGCAAGC Jones et al., 2003

GACACTCCATAGAAATGCACTG

Aa4A Antechinus

agilis

NA 165-167 1 x35 2 46 TTTGATCCTCAGAGACTTGAT Banks et al., 2005

CCAAATCTACGTAAAATATCC

Aa4J Antechinus

agilis

NA 166-178 1 x35 2 46 TCTTCAGTCTCTCAATGAGTT Kraaijeveld-Smit

et al., 2002 AGAACACTCTAACAACATCCT

171

Table S4.2 Inferring the value of K, the number of populations of the island

Parantechinus apicalis, using STRUCTURE. K is the number of genetic clusters.

LnP(K) is the posterior probability of the data for a given K. Ln`(K) is the model choice

criterion. The most likely K is identified following Evanno et al. (2005b) by selecting

the highest mean LnP(K) with the smallest standard deviation. The strength of signal is

indicated by ∆K.

K Replications Mean

LnP(K)

Stdev


1 10 -3047.8 0.1 — — —

2 10 -2284.9 2.7 762.9 729.3 275.0

3 10 -2251.3 11.7 33.6 45.0 3.8

4 10 -2262.8 57.4 -11.5 181.0 3.2

5 10 -2455.3 198.7 -192.5 362.5 1.8

6 10 -2285.3 118.9 170.0 155.4 1.3

7 10 -2270.7 104.6 14.6 130.5 1.2

8 10 -2386.5 170.5 -115.9 54.4 0.3

9 10 -2448.0 293.6 -61.5 178.7 0.6

10 10 -2330.8 146.5 117.2 — —

172

Table S4.3 Breeding pair record for the Dibbler, Parantechinus apicalis, translocation to Escape Island at Perth Zoo (modified from Lambert and Mills

(2006), unpublished data). The animals used in the breeding program were wild-born dibblers from Boullanger Island (BW), wild-born dibblers from

Whitlock Island (WW), captive-born Boullanger Island purebred (BP), hybrid (H), and Boullanger Island backcross (BBC).

Year

IDENTIFICATIO

N

WEIGHT

AGE BEHAVIOUR

OTHER INFO

Mate

success

fully

(Y/N)

Number

of

Offspring

survive

Number

of

Offspring

die

Female Male Male

Heavier

by:

Female

heavier

by:

Same

Age

Male

Older

Female

Older

Female

Aggression

No mating

behaviour

1997 BW4 BW1 This information was

found in the studbook

Y 6 0

1997 WW8 BW2 This information was


Y 7 0

1997 BW3 BW2 This information was


Y 6 2

1998 H16

(49g)

BP24

(73g)

24g X Y 0 5

1998 H16

(49g)

WW7

(58g)

9g X Male 2 years +

1998 H17

(41g)

BP24

(67g)

26g X N 0 0

1998 H17

(41g)

WW7

(53g)

12g X Male 2 years +

1998 BP9

(58g)

BP21

(91g)

33g X Y 7 0

1998 BP9

(58g)

H19

(54g)

4g X

173

1998 BP10

(60g)

BP21

(83g)

23g X Y 0 8

1998 BP10

(60g)

H19

(55g)

5g X

1998 BP22

(56g)

H15

(62g)

6g X Y 8 1

1998 BP23

(57g)

H15

(57g)

Even Even X Y 5 3

1998 BP25

(58g)

BW1

(70g)

12g X Male 2 years + Y 4 1

1998 BP13

(57g)

BP24

(78g)

21g X Y 0 6

1998 BP29

(49g)

BW1

(70g)

21g X X Male 2 years + N 0 0

1998 BW3

(53g)

BW2

(78g)


Female 2 years +

1999 BP9

(66g)

BP21

(83g)

17g X Y 7 0

1999 BP11

(60g)

BBC37

(72g)

12g X Y 7 0

1999 BP14

(63g)

BP39

(77g)

14g X Y 6 2

1999 H18

(48g)

WW80

(54g)

6g X Y 0 8

1999 BBC34

(54g)

WW80

(51g)

3g X Y 6 0

1999 BP40

(63g)

BP57

(70g)

7g X Y 8 0

1999 BP41

(59g)

H15

(70g)

11g X Y 8 0

1999 BP60

(52g)

BW1

(70g)

10g X Male 3 years + N 0 0

No young resulted

174

1999 BP11

(64g)

WW7

(53g)


1999 BP14

(63g)

BW1

(65g)


1999 H18

(48g)

WW78

(52g)

4g X X N 0 0

1999 BP22

(70g)

WW78

(49g)

21g X X N 0 0

1999 BP22

(61g)

WW79

(51g)

10g X X N 0 0

1999 BP22

(61g)

WW7

(52g)


1999 BP22

(61g)

H15

(57g)

4g X X N 0 0

1999 BBC34

(55g)

WW79

(50g)

5g X X N 0 0

1999 BP40

(52g)

WW80

(45g)

7g X X N 0 0

1999 BP40

(60g)

WW78

(52g)

8g X X N 0 0

1999 BP40

(60g)

WW79

(51g)

9g X X N 0 0

1999 BP40

(60g)

WW7

(52g)


1999 BP60

(52g)

WW6

(54g)


2000 BP11 H15 This information was

found in the studbook.

One of the baby got

transfer to Adelaide

zoo

Y 2 0

2000 BP9 BP21 This information was


Y 1 2

175

Table S4.4 Description for choice of demographic and genetic parameters used in

constructing population viability analysis for Parantechinus apicalis island populations

on Boullanger, Whitlock and Escape Islands using the software VORTEX version 10.0

(Lacy and Pollak, 2014).

Vortex Parameter Value

Scenario settings

No. iterations 1000

No. “years” 500

Duration of each “year” in days 365 (reproduce once a year)

Extinction definition Only 1 sex remains

Number of populations Whitlock Island = 1; Boullanger

Island = 1; Escape Island = 1

Species description

Inbreeding depression

- Lethal equivalents 8

- % due to recessive lethal 50

EV concordance of reproduction &

survival

- EV correlation among populations 0.5 (default)

- Number of types of catastrophes 0

Reproductive system

Reproductive system Polygynous

Age of first offspring for

Females/Males

1 (8-9 months for females; 10 months

for males)

Maximum age of reproduction 3 (3 years)

Maximum number of broods/year 1

Maximum number of progeny per

brood

8

Sex ratio at birth 41.1

Reproductive rates

% Adult females breeding 90

- EV in % breeding 0

Distribution of broods per year

- 0 broods 0

- 1 broods 100

Number of offspring per female brood

- Mean Whitlock Island = 6.2; Boullanger Island

= 7.4; Escape Island = 7.0

- Standard Deviation Whitlock Island = 0.2; Boullanger Island

= 0.1; Escape Island = 1.1

Mortality rates

Mortality Age 0 to 1 (± SD) 29.4 ± 0

Annual mortality after Age 1 (± SD) 35 ± 0

176

Mate monopolisation

% males in breeding pool 84.33

Initial population size

Initial population size Whitlock Island = 42; Boullanger Island

= 97; Escape Island = 88

Stable age distribution Default values

Carrying capacity

Carrying capacity (K) Whitlock Island = 42; Boullanger Island


SD in K due to EV Whitlock Island = 12; Boullanger Island


Supplementation

First year of supplement

Last year of supplement

Interval between supplements

Optional criteria for supplements

Supplement from after age 1

1

500

5/10 years

1 (Default)

One migrate per year, 20% of K, 20 and

30 animals

Genetic management

Number of neutral loci to be modelled 14

Read starting allele frequencies from

file

see text below

Designation of “years” ˗ although P. apicalis reaches reproductive age at 8 ˗ 9 months

for females and 10 - 11 months for males, they are seasonal breeders and only breed

once a year. Therefore, we constructed the population viability model using calendar

years.

Inbreeding depression ˗ it is unknown how much inbreeding depression affects P.

apicalis survival. We used the value of eight lethal equivalents per generation with 50%

due to lethal recessives to represent the combined mean effect of inbreeding on first

year survival and survival to sexual maturity (O’Grady et al., 2006).

Environmental variation ˗ we accepted the default values of environmental variation

provided by VORTEX and did not incorporate catastrophic events into the models as

we were primarily interested in genetic rather than demographic patterns.

Percentage breeding success ˗ in captivity 1998, 12 mating observed out of 14 pairing

attempts which indicated 85.7% male mating success. In 1999, eight mating observed

out of 21 pairing attempts which indicated 38.1% male mating success. An average

male mating success rate across two years is 61.9%. This value is similar to the

177

paternity assignment based on exclusion results of Antechinus agilis (69% and 78%,

Kraaijeveld-Smit et al., 2003) and A. stuartii (56%, Holleley et al., 2006). Percentage

females producing is taken from Moro’s study (2003) from collection year 2001 when

the population was more likely to have stabilized and after the last set of founders were

released. This value is similar to A. stuartii (92%, Holleley et al., 2006).

Sex ratio at birth ˗ we averaged sex ratios from captive-bred dibblers born between year

1997 to 1999 (Lambert and Mills, 2006). We excluded year 2000 because the sample

size was small (1M: 3F).

Mortality of females and males ˗ morality is estimated from morality rates of captive

dibblers for 0 ˗ 12 months (29.4%) (Lambert and Mills, 2006, unpublished data) and

estimated a mortality rate for > 12 months old (35%) from the survival rate of dibblers

on Escape Island (Moro, 2003).

Population carrying capacity ˗ there is no information on the population carrying

capacity of the island populations. Based on monitoring trips to all islands between

2006 and 2012, we used the maximum estimates of Known to be alive (KTBA) as carry

capacity values and standard deviations. The maximum KTBA on Boullanger Island is

97 with a mean of 52 and standard deviation of 27 individuals. The maximum KTBA on

Whitlock Island is 42 with a mean of 28 and a standard deviation of 12 individuals. The

maximum KTBA on Escape Island is 47 with a mean of 27 and a standard deviation of

15 individuals.

Initial population size ˗ we used carrying capacities as the initial population size for

Boullanger and Whitlock Island. For the population on Escape Island, we used the total

number of animals released on Escape Island during the translocation.

Genetic input ˗ we seeded the demographic models using the allele frequencies of the

overall estimate of each island population.

178

Table S4.5 Allele frequencies of the source, captive and translocated populations of the

island Parantechinus apicalis.

Locus/Allele

Boullanger

Island

Whitlock

Island

Captive

population

Translocated

population

3.3.2 115 48 68 102

160 0.000 0.000 0.000 0.029

162 0.017 0.042 0.007 0.020

164 0.983 0.927 0.890 0.725

166 0.000 0.031 0.103 0.225

Aa4A 112 50 69 104

166 0.397 0.990 0.710 0.663

168 0.603 0.000 0.290 0.313

170 0.000 0.010 0.000 0.010

172 0.000 0.000 0.000 0.014

pPa2A12 120 51 73 119

129 0.029 1.000 0.507 0.462

131 0.971 0.000 0.493 0.538

4.4.2 117 48 61 108

127 0.748 1.000 0.361 0.352

129 0.252 0.000 0.639 0.648

pPa1B1O 113 52 73 103

220 0.593 1.000 1.000 0.995

224 0.407 0.000 0.000 0.005

pPa2D4 112 45 72 93

193 0.004 0.000 0.000 0.005

195 0.817 1.000 0.924 0.995

197 0.179 0.000 0.076 0.000

pPa7A1 115 49 72 92

298 0.009 0.041 0.056 0.033

301 0.000 0.643 0.167 0.402

304 0.061 0.316 0.063 0.022

307 0.300 0.000 0.139 0.103

310 0.265 0.000 0.111 0.087

313 0.357 0.000 0.431 0.353

315 0.009 0.000 0.035 0.000

3.1.2 111 48 65 102

177 0.000 0.000 0.000 0.005

179 0.000 0.010 0.000 0.034

181 0.027 0.042 0.000 0.059

183 0.135 0.188 0.485 0.490

179

185 0.640 0.760 0.392 0.338

187 0.005 0.000 0.000 0.000

189 0.194 0.000 0.123 0.074

4.4.10 119 50 67 107

221 0.282 0.150 0.313 0.439

230 0.017 0.020 0.000 0.014

232 0.702 0.830 0.687 0.547

3.3.1 120 50 67 117

138 0.004 0.000 0.007 0.000

144 0.004 0.000 0.000 0.000

146 0.596 1.000 0.515 0.585

148 0.383 0.000 0.478 0.415

152 0.013 0.000 0.000 0.000

1A1 112 46 64 106

208 0.402 1.000 0.711 0.675

212 0.598 0.000 0.289 0.325

Sh6e 120 53 73 116

173 0.000 0.019 0.021 0.004

175 0.038 0.585 0.253 0.418

177 0.963 0.396 0.726 0.578

4B3 112 49 61 119

125 0.388 1.000 0.336 0.391

127 0.232 0.000 0.000 0.008

129 0.379 0.000 0.664 0.601

2B10 114 41 63 104

176 0.018 0.024 0.000 0.077

180 0.978 0.976 0.865 0.846

184 0.004 0.000 0.135 0.077

180

Figure S5.1: Encounter probability of Bettongia lesueur born at the Lorna Glen

translocation site during population monitoring between July 2010 and April 2015.

Error bars represent standard errors.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Jul-

10

Oct

-10

Jan-1

1

Apr-

11

Jul-

11

Oct

-11

Jan-1

2

Apr-

12

Jul-

12

Oct

-12

Jan

-13

Apr-

13

Jul-

13

Oct

-13

Jan-1

4

Apr-

14

Jul-

14

Oct

-14

Jan-1

5

Apr-

15

Enco

unte

r pro

bab

ilit

y

181

Table S5.1 Characteristics of the 18 microsatellite loci that were optimized for use in characterizing genetic variability within the source and

translocated populations of the burrowing bettong, Bettongia lesueur.

Locus Multiplex

Primer

Conc.

(µM)

(F+R)

Anneal

Temperature

(C°)

Size Range

(bp)

Fluorescent

dye Primer sequence (5'-3') Reference Min Max

Bt64 1 0.12 57 175 214 PET AATAGGAATCCATATGCTGATGC Pope et al. 2000

ATAGCCAACTGGGTAATTTAGTG

Bt76 1 0.09 57 204 233 NED CGATGGTAGGCAACAACGAATAG Pope et al. 2000

ATAACCAGTTCTTCATAAAATCC

PI3 1 0.5 57 135 163 FAM GCTGGGAGGTTTGTTGATTTAC Luikart et al. 1997

GAGTCAAGAAATCAAACTGCCC

Y105 1 0.08 57 229 235 VIC GGTAATGAGTCAGTGTGATGAGG Zenger et al. 2002

GGTAGGAGGAAAGGGAGAAAAG

Bt80 2 0.2 54 185 204 FAM CACTTTTACCAGGCTACCTAACC Pope et al. 2000

CCCCTGATGAGATTATACTAAAC

T31-1 2 0.6 54 221 237 VIC TCAGGATTTTATTCTTCCATCTTTC Zenger and Cooper 2001

TTGGGGAGAAGATTTTTGAGAG

Y148 2 0.18 54 167 183 NED CTGTAGAATGTAACTTCCAGAA Pope et al. 1996

CTTTGGATTGAGAGACTAGGAT

Y170 3 0.2 62 145 177 NED GGACTCAAACCCAACACTAGC Pope et al. 1996

TGCATGCCTTTGTCATACACG

Y151 3 0.15 57 194 233 VIC ATATTACCTGCAAACTGGAAC Pope et al. 1996

AGCCATTGCAGTAACTCCAAC

T17-2 3 0.2 57 96 110 FAM AGCTCAGAGTTCCAACCCAATC Zenger and Cooper 2001

GAAACTTCTCCCAAGTGTTTCTGG

Pa593 3 0.4 60 124 139 PET TAGGGCACGTCATAAGGATGCAG Spencer et al. 1995

GCTTTCCAGTTTCTGACTTTTCATG

182

PI26 4 0.3 54 164 184 PET TTGTGAAATAGTGCATTTTCTGC Luikart et al. 1997

GGCTTCTGAGCAGTCAGTTCTG

PI22 4 0.2 54 124 156 FAM GGTGCTCATTTCATTTAGAGTTTG Luikart et al. 1997

ATTCTTCCTTTCCAAACTCAAGG

Pa597 4 0.5 54 94 128 VIC ACATACTCTATGCAACATTGGCTT Spencer et al. 1995

CTAGTAGAAAGGAAAAGAATTCAGA

Y76 4 0.8 54 163 181 NED AGAGTAGTAATTTCAGTCCTTTG Pope et al. 1996

CTGAACCTTATTCTCCCACAT

Y175 5 0.1 57 163 289 NED TGGGACATTTCCTGACCTAC Zenger et al. 2002

CCTCTTTAGGCTTCTTGACCTAC

Y112 5 0.3 57 177 225 VIC CATGTACTGCTGAGAATAGGCAC Zenger et al. 2002

CCTGGAGAAGTCTATCTCCCAAC

Pa385 5 0.18 57 170 170 FAM GCTCTACCAGGCTGATTGGGA Spencer et al. 1995

TGAGTATCTCTTTTGCTGCTTGAA

183

Table S5.2 Inferring the value of K, the number of populations in the Bettongia lesueur

translocation at Lorna Glen, using STRUCTURE. K is the number of genetic clusters.

LnP(K) is the posterior probability of the data for a given K. Ln`(K) is the model choice

criterion. The most likely K is identified following Evanno et al. (2005b) by selecting the

highest mean LnP(K) with the smallest standard deviation. The strength of signal is

indicated by ∆K.

K Replications Mean

LnP(K)

Stdev


1 10 -5546.3 0.3 — — —

2 10 -3602.9 2.0 1943.5 1909.6 965.3

3 10 -3569.0 16.3 33.9 17.3 1.1

4 10 -3552.3 62.1 16.6 21.5 0.3

5 10 -3557.1 98.0 -4.8 104.9 1.1

6 10 -3666.9 341.8 -109.7 50.2 0.1

7 10 -3826.8 506.5 -160.0 203.0 0.4

8 10 -3783.8 566.0 43.0 77.0 0.1

9 10 -3663.8 110.6 120.0 188.9 1.7

10 10 -3732.7 114.0 -68.9 — —

184

a) Chapter two

b) Chapter three

c) Chapter four

d) Chapter five

Figure S6.1: Relationship between effective population size (Ne) and sample size in a)

Chapter two, b) Chapter three, c) Chapter four, and d) Chapter five.

0

100

200

300

400

500

0 5 10 15 20

Ne

Sample size

Hamersley Moir

Twertnup

0

20

40

60

80

100

0 10 20 30 40 50 60

Ne

Sample size

East

West

Captive population

Translocated population

Source: rho = -0.145, P = 0.78

Captive: rho = 0.058, P = 0.91

Translocated: rho = -0.087, P = 0.87

0

5

10

15

20

25

0 10 20 30 40 50

Ne

Sample size

Boullanger Island

Whitlock Island

Captive population

Escape Island

0

10

20

30

40

50

0 20 40 60 80 100

Ne

Sample size

Barrow Island

Dryandra

Lorna Glen

Lorna Glen: rho = -0.4, P = 0.75

Source: rho = -0.145, P = 0.78

Escape Island: rho = -0.2, P = 0.92

185

Figure S6.2: Relationship between effective population size (Ne) and sample size of the

burrowing bettong source populations using a random resampling method.

0

10

20

30

40

50

60

0 20 40 60 80 100

Ne

Sample size

Dryandra

Barrow Island

Dryandra: rho = -0.89, P = 0.03

Barrow: rho = 0.54, P = 0.3

genetic consequences of genetic mixing in mammal ... consequences of genetic mixing in mammal...

Documents