Genetic consequences of genetic mixing in mammal ... consequences of genetic mixing in mammal...
-
Upload
truongdiep -
Category
Documents
-
view
216 -
download
0
Transcript of 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.
v
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 Materials and methods 52
3.3.1 Study site and sample collection 52
3.3.2 DNA extraction and microsatellite genotyping 54
3.3.3 Data analysis 55
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 Materials and methods 79
4.3.1 Sampling and DNA extraction 79
4.3.2 Microsatellite variation 80
4.3.3 Data analysis 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 Materials and methods 107
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
ix
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%
Thavornkanlapachai, R., W.J. Kennington, K. Ottewell, T. Friend, and H.R.
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
Analysis Writing Total
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
Analysis Writing Total
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
Analysis Writing Total
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
xvi
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:
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.
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:
Thavornkanlapachai, R., W.J. Kennington, K. Ottewell, T. Friend, and H.R.
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.
25
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.
47
48
Photo credit: Perth Zoo
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 MATERIALS AND METHODS
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
Captive population Translocated population
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.
73
74
Photo credit: Perth Zoo
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 MATERIALS AND METHODS
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.
101
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 MATERIALS AND METHODS
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
Captive population Translocated population
-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
Captive population Translocated population
-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
Conservation, 122, 581-597.
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
Conservation, 147, 197-203.
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
Conservation, 16, 633-640.
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
Conservation, 50, 143-198.
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.
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.
Molecular Ecology, 17, 325-333.
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.
Molecular Ecology, 6, 1199-1201.
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.
Molecular Ecology, 14, 441-456.
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
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.
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
Conservation, 141, 355-363.
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
Conservation, 111, 305-315.
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.
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.
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
Conservation, 124, 2915-2922.
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
Conservation, 96, 185-196.
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
LnP(K) Ln'(K) |Ln''(K)| ∆K
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
found in the studbook
Y 7 0
1997 BW3 BW2 This information was
found in the studbook
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)
25g X X Male 2 years + N 0 0
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)
11g X X Male 3 years + N 0 0
1999 BP14
(63g)
BW1
(65g)
2g X X Male 3 years + N 0 0
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)
9g X X Male 3 years + N 0 0
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)
8g X X Male 3 years + N 0 0
1999 BP60
(52g)
WW6
(54g)
2g X X Male 3 years + N 0 0
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
found in the studbook
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
= 97; Escape Island = 47
SD in K due to EV Whitlock Island = 12; Boullanger Island
= 27; Escape Island = 15
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
LnP(K) Ln'(K) |Ln''(K)| ∆K
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