Genetic structure and diversity of the koala population in South Gippsland, Victoria: a remnant population of high conservation significance

Abstract

In the Australian state of Victoria, the history of koalas and their management has resulted in the homogenisation and reduction of genetic diversity in many contemporary populations. Decreased genetic diversity may reduce a species’ ability to adapt to future environmental pressures such as climate change or disease. The South Gippsland koala population is considered to be unique in Victoria, as it is believed to be a remnant population, not originating from managed populations that have low genetic variation. This study investigated genetic structure and diversity of koalas in South Gippsland, with comparison to other populations in Victoria (French Island/Cape Otway, FI and Raymond Island, RI), New South Wales and south east Queensland. Population analyses were undertaken using both microsatellite genotype and mitochondrial DNA sequence data. Non-invasive sampling of koala scats was used to source koala DNA, allowing 222 South Gippsland koalas to be genotyped. Using nuclear data the South Gippsland koala population was found to be significantly differentiated (Djost 95% CI SG–RI = 0.03–0.06 and SG–FI = 0.08–012) and more diverse (AR 95% CI SG = 4.7–5.6, RI = 3.1–3.3, FI = 3.0–3.3; p = 0.001) than other Victorian koala populations, supporting the premise that koalas in the South Gippsland region are part of a remnant population, not derived from translocated island stock. These results were also supported by mitochondrial data where eight haplotypes (Pc4, Pc17, Pc26, Pc27, and Pc56–Pc59) were identified in South Gippsland while a single haplotype (Pc27) was found in all island koalas tested. Compared to other Victorian koala populations, greater genetic diversity found in South Gippsland koalas, may provide this population with a greater chance of survival in the face of future environmental pressures. The South Gippsland koala population is, therefore, of high conservation significance, warranting the implementation of strategies to conserve this population and its diversity into the future.

Full text

Vol.:(0123456789)
1 3
Conservation Genetics
https://doi.org/10.1007/s10592-018-1049-8
RESEARCH ARTICLE
Genetic structure anddiversity ofthekoala population inSouth
Gippsland, Victoria: aremnant population ofhigh conservation
significance
FayeWedrowicz1,2 · JenniferMosse2· WendyWright2· FionaE.Hogan2
Received: 30 May 2017 / Accepted: 12 January 2018
© Springer Science+Business Media B.V., part of Springer Nature 2018
Abstract
In the Australian state of Victoria, the history of koalas and their management has resulted in the homogenisation and
reduction of genetic diversity in many contemporary populations. Decreased genetic diversity may reduce a species’ abil-
ity to adapt to future environmental pressures such as climate change or disease. The South Gippsland koala population is
considered to be unique in Victoria, as it is believed to be a remnant population, not originating from managed populations
that have low genetic variation. This study investigated genetic structure and diversity of koalas in South Gippsland, with
comparison to other populations in Victoria (French Island/Cape Otway, FI and Raymond Island, RI), New South Wales and
south east Queensland. Population analyses were undertaken using both microsatellite genotype and mitochondrial DNA
sequence data. Non-invasive sampling of koala scats was used to source koala DNA, allowing 222 South Gippsland koalas
to be genotyped. Using nuclear data the South Gippsland koala population was found to be significantly differentiated (Djost
95% CI SG–RI = 0.03–0.06 and SG–FI = 0.08–012) and more diverse (AR 95% CI SG = 4.7–5.6, RI = 3.1–3.3, FI = 3.0–3.3;
p = 0.001) than other Victorian koala populations, supporting the premise that koalas in the South Gippsland region are part
of a remnant population, not derived from translocated island stock. These results were also supported by mitochondrial data
where eight haplotypes (Pc4, Pc17, Pc26, Pc27, and Pc56–Pc59) were identified in South Gippsland while a single haplotype
(Pc27) was found in all island koalas tested. Compared to other Victorian koala populations, greater genetic diversity found
in South Gippsland koalas, may provide this population with a greater chance of survival in the face of future environmental
pressures. The South Gippsland koala population is, therefore, of high conservation significance, warranting the implemen-
tation of strategies to conserve this population and its diversity into the future.
Keywords Phascolarctos cinereus· Non-invasive DNA sampling· Scats· Population structure· Genetic diversity·
Microsatellite genotyping· Mitochondrial DNA sequencing
Introduction
The koala (Phascolarctos cinereus) is an arboreal Austral-
ian marsupial inhabiting eucalypt forests of Australia’s east
(Fig.1). A dietary specialist, koalas feed exclusively on the
foliage of certain eucalypt species (Martin and Handasyde
1999). Breeding occurs from October to May and females
generally bear one offspring each 1–3years (Handasyde
etal. 1990; Martin and Handasyde 1990). Mean home range
size varies from 0.5ha at Cape Otway in Victoria (Whis-
son etal. 2016) to 135ha in central Queensland (Ellis etal.
2002), likely driven by the density of preferred eucalypt spe-
cies in an area (Martin and Handasyde 1999).
In Australia, extensive habitat loss and hunting post Euro-
pean colonisation (~ 1788) decimated koala populations. By
the early 1900s, less than 1000 koalas remained on the Vic-
torian mainland, whilst introduced populations on French
and Phillip Islands flourished, eventually reaching unsus-
tainable densities and requiring intervention (Lewis 1954;
Menkhorst 2008). Between 1923 and 2006, over 12,000
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10592-018-1049-8) contains
supplementary material, which is available to authorized users.
* Faye Wedrowicz
faye.wedrowicz@gmail.com
1 Faculty ofScience, Monash University, Clayton, Australia
2 School ofApplied andBiomedical Sciences, Federation
University Australia, Churchill, VIC3842, Australia

Conservation Genetics
1 3
koalas were translocated from French and Phillip Islands to
the mainland, to curb population growth whilst simultane-
ously facilitating the re-establishment of koala populations
in Victoria (Lewis 1954; Menkhorst 2008). As only small
numbers of individuals were used to establish the island pop-
ulations during the late 1800s (French Island, n = 3 and Phil-
lip Island, n ~ 10–30), genetic diversity was reduced in island
populations relative to their ancestral population/s (Lee etal.
2011; Wedrowicz etal. 2017b). Although translocation of
individuals from French and Phillip Islands to the mainland
was successful in re-establishing koala populations through-
out Victoria (and in establishing koala populations in South
Australia), genetic diversity among and between contempo-
rary koala populations in Victoria and South Australia is low
(Houlden etal. 1996, 1999; Cristescu etal. 2009; Lee etal.
2011). Low genetic diversity can impact a species’ ability to
adapt to new environmental pressures such as climate change
or disease, even where population size is large (Bijlsma etal.
2000; Frankham 2005). This lack of variation is of genuine
concern for the future viability of southern koala populations
(in Victoria and South Australia), especially during the cur-
rent period of rapid environmental change.
Koala populations in Victoria and South Australia are
currently considered secure, mainly due to high koala densi-
ties of some populations (Department of the Environment
2015). The density of koalas in other southern populations
ranges from low to moderate though data are unavailable
for many (EaCRC 2011a). Conversely, widespread decline
of koala populations in the north of Australia (Queensland,
New South Wales and the Australian Capital Territory)
since the 1990s has resulted in northern koalas being listed
as vulnerable under the Environment Protection and Bio-
diversity Conservation Act 1999 (EPBC Act; Department
of the Environment 2015). Due to overabundance of some
koala populations in Victoria and South Australia, koalas
in these states are not listed as threatened under the EPBC
Act (EaCRC 2012; Department of the Environment 2015).
As most koala populations in Victoria and South Aus-
tralia were entirely founded by island stock they are likely to
lack genetic diversity. An exception is the koala population
in South Gippsland (Victoria), which is thought to be a rem-
nant population that has received very few translocations of
island stock (mainly in coastal South Gippsland; see Wedro-
wicz etal. 2017b). Koalas inhabiting this region may, there-
fore, retain greater levels of ancestral diversity; past studies
have indicated that koalas in South Gippsland have greater
genetic diversity compared to southern populations founded
by island stock (Houlden etal. 1999; Lee etal. 2011).
The koala population in South Gippsland has been shown
to be differentiated from and have significantly higher lev-
els of genetic diversity (with an average of six alleles per
locus) compared to koala populations from the French Island
and Mornington Peninsula (of French Island origin, with
an average of less than four alleles per locus) (Lee etal.
2011). Genetic differences between the South Gippsland
and Phillip Island populations have been demonstrated
Fig. 1 Spatial distribution of samples collected for this study. Shading
on the map of eastern Australia (left) shows koala distribution which
was adapted from Department of the Environment (2015). The map
on the right shows the South Gippsland region, with the Strzelecki
Ranges (STZ) bioregion indicated by the purple outline, the Gipps-
land Plain (GP) bioregion outlined in blue and the Wilsons Promon-
tory (WP) bioregion outlined in red. (Color figure online)

Conservation Genetics
1 3
using mitochondrial DNA haplotype data with four hap-
lotypes identified in South Gippsland and one at Phillip
Island (Houlden etal. 1999). Using nuclear DNA however,
the Phillip Island population was found to have similar lev-
els of genetic diversity to the South Gippsland population
(Houlden etal. 1996; Fowler etal. 1998).
Past population genetic studies of koalas in South Gipp-
sland have largely relied on spatially localised sample col-
lection or opportunistic sampling of deceased (e.g. road
kill) individuals or animals entering wildlife shelters due
to illness or trauma (Houlden etal. 1996, 1999; Lee etal.
2011). Wild individuals in the Strzelecki Ranges bioregion
(within South Gippsland, Fig.1) have not been systemati-
cally sampled in the past, so the full extent and distribution
of the genetic diversity within the regions koala population
is unknown.
While molecular techniques are now routinely used for
wildlife studies, sourcing DNA can be challenging. DNA
for genetic studies is often obtained invasively from blood
or biopsies sourced from captured or deceased animals.
Although providing high quality DNA, invasive sampling
has the potential to limit sample size, especially if the spe-
cies is elusive and/or rare. Hence, obtaining sufficient sam-
ples to avoid bias is not always possible. Collection of DNA
from non-invasive sources, such as scats, can be a more
appropriate sampling option for rare or elusive species in
difficult environments (Piggott and Taylor 2003). Sourc-
ing DNA from scats reduces the time, cost and expertise
associated with invasive sampling, and can thus facilitate
the collection of more samples, in a shorter period of time,
across a larger spatial area. As koalas spend most of their
time in the canopy of tall eucalyptus trees (often > 30ms),
animal capture for DNA sampling is difficult. Within the Str-
zelecki Ranges bioregion, obtaining DNA from wild koalas
is especially difficult due to the rugged terrain. Faecal pel-
lets (scats) found at the base of a tree provide an alternative,
accessible DNA source. Koala DNA isolated from host cells
coating the surface of koala scats has been shown to provide
DNA of quality sufficient to generate a unique identifying
genotype (Wedrowicz etal. 2013).
DNA isolated from koala scats was used to investigate
genetic variation within the South Gippsland koala popu-
lation, using microsatellite genotyping and mitochondrial
sequencing. The aims of this study were (1) to examine
genetic differentiation and diversity of the South Gippsland
koala population compared to other Victorian koala popula-
tions and more northern populations in New South Wales
and Queensland and (2) to consider fine-scale population
structure and the distribution of genetic variability in koalas
throughout the South Gippsland region.
Methods
South Gippsland study site
The South Gippsland study region covers an area of approxi-
mately 6,000 sq km and includes the Strzelecki Ranges and
Gippsland Plain bioregions1 (Fig.1). The Gippsland Plain
bioregion is dominated by agricultural land, although rela-
tively large forested areas exist in the east, consisting mainly
of plains grassy forest, lowland forest and heathy woodland
EVCs.2 Land use in the Strzelecki Ranges bioregion is more
diverse. The western half of the Strzelecki Ranges bioregion
mainly consists of agricultural land, with very few parks
and reserves. In the eastern half of the Strzelecki Ranges
bioregion, a large proportion of the landscape is under the
management of Grand Ridge Plantations Pty. Ltd. (HVP
Plantations) and is utilised for forestry. HVP’s estate con-
sists of plantation species, as well as native forest managed
for conservation purposes (EaCRC 2011b). Native habitat
containing preferred tree species, and therefore koalas, are
unevenly distributed across the region. Within the HVP
estate, koalas are found at densities of 0.25 koalas per hec-
tare in native forest containing locally preferred eucalypt
species including blue gum (Eucalyptus globulus), yellow
stringybark (E. muelleriana) and/or mountain grey gum (E.
cypellocarpa) (Allen 2015; Richard Appleton, HVP, pers.
comm.). Plantation species in the region include radiata pine
(Pinus radiata), blue gum, mountain ash (E. regnans) and
shining gum (E. nitens). Koalas are generally less common
within the plantations (R. Appleton, HVP, pers. comm.).
Sample collection
Koala scats were collected across South Gippsland between
March 2013 and December 2016 by researchers, forestry
staff, contractors and citizen scientists (Fig.1). As koala
density is greatest in the eastern part of the Strzelecki
bioregion, an intensive sampling strategy was designed; the
region was divided into nine areas, and searches for koalas
and koala scats were carried out at between five to ten sites
within each area. Areas considered to be good koala habitat
(according to koala habitat modelling undertaken by HVP)
were preferentially searched over less favourable habitat.
Additionally, scat samples were opportunistically obtained
from koalas at the Southern Ash Wildlife Shelter (SAWS),
1 Bioregions are areas of land defined by similarities in geologi-
cal and ecological characteristics; and are used by state government
agencies, and others, for biodiversity planning and land management
purposes.
2 Ecological vegetation classes (EVCs) are used for classifying veg-
etation types within bioregions.

Conservation Genetics
1 3
where sick and injured koalas from the South Gippsland
region are rehabilitated. Tissue samples, from deceased indi-
viduals, were also obtained from SAWS.
Seven reference populations were sampled for compari-
son; four from Victoria and three from New South Wales
and Queensland. All reference populations were sampled
by collecting scats. Victorian populations sampled included
French Island, two mainland locations where individuals
from French Island were released (Cape Otway and Mal-
lacoota) and Raymond Island, which was founded by koalas
translocated from Phillip Island. More northern populations
sampled were from New South Wales (south east and north
east) and south east Queensland (Fig.1). Prior information
regarding potential sub structuring within populations was
not known. A list of samples used for this study and their
locations is provided in online resource 1 (TableS1).
For all wild koalas, and some shelter koalas, four scats per
individual (where possible) were collected using a wooden
toothpick inserted into the side of each pellet. Scats were
stored by inserting the opposite end of the toothpick into
a foam block. Foam blocks supporting scats on toothpicks
were then encased in a plastic, open ended covering (rectan-
gular drain pipe sectioned into approximately 13cm lengths)
for protection. Each set of scats, stored in this way, consti-
tuted a sample from one individual. Details of each sample
were recorded on the plastic covering, including date, col-
lector details and spatial coordinates obtained from a hand
held Global Positioning System (GPS). For deceased shelter
animals, tissue samples were collected by SAWS by excis-
ing a small (~ 10 × 20mm) piece of ear tissue and storing in
methylated spirits. If possible, the location (nearest town)
from which the deceased individual koala had been retrieved
was georeferenced to obtain spatial data.
DNA isolation fromscats
Scats were stored at ambient temperature, on toothpicks until
surface washing (< 4 weeks). The surface of each scat was
individually washed in a vial, with 2mL of PBS buffer, by
rolling on a Ratek roller mixer (BTR5P) at full speed for
8min. DNA isolation was carried out either immediately
after surface washing or after storage of surface washes at
− 20°C (for a maximum of 10months). DNA was isolated
from two washes using the Qiagen QIAamp® DNA Stool
Mini Kit as previously described (Wedrowicz etal. 2013) or
the Axygen® AxyPrep™ MAG Soil, Stool, and Water DNA
Kit. Isolations using the Axygen® DNA Kit were carried out
following the manufacturer’s instructions with slight modi-
fication to the volumes of supernatant transferred after the
centrifugation step (400µL), SBW buffer added to the super-
natant (400µL) and binding enhancer (15µL). DNA was
isolated from two of the four washes to provide two separate
DNA isolates per individual. To minimise the risk of cross
contamination between samples, surface washing, DNA iso-
lation, PCR setup and electrophoresis of PCR products were
all carried out in separate work areas using equipment dedi-
cated to each work space and filter pipette tips were used.
DNA was isolated from tissue using the DNeasy® Blood &
Tissue Kit (Qiagen) following the manufacturer’s protocol.
Screening forDNA quantity andquality
DNA isolates from two washes from each sample were
screened for DNA quantity and quality as described in
Wedrowicz etal. (2017a). Total DNA was quantified using
the Qubit® dsDNA HS assay kit (Life Technologies) while
DNA quality was assessed by amplification of microsatel-
lite Pcv31 (Cristescu etal. 2009) and sexing markers using
standard PCR and electrophoresis. Primers, IMY1 and
IMY2 (Watson etal. 1998), used to amplify Y chromosome
DNA in male koalas sampled in Victoria and New South
Wales, did not produce amplification product for koalas sam-
pled from Queensland (Fig.1). A new primer set targeting
the Y chromosome was therefore designed from GenBank
sequence LC111530.1 (Katsura etal. 2016). Primers desig-
nated PCY-F (5′-TCT GGA GAA TCC CAA AAT GC-3′) and
PCY-R (5′-ATT CTT CCC TG T GTT TAG CG-3′) successfully
amplified a fragment of approximately 130bps in length for
male Queensland koalas. For each sample, the DNA iso-
late producing the brightest bands using gel electrophoresis
was chosen for microsatellite genotyping. DNA isolates that
failed both screening PCRs were not analysed using micros-
atellite genotyping but were retained for potential amplifica-
tion of mitochondrial DNA.
Microsatellite genotyping
Twelve microsatellite markers, K2.1, K10.1, Pcv2, Pcv6.1,
Pcv6.3, Pcv24.2, Pcv25.2, Pcv30, Pcv31 (Cristescu etal.
2009), Phc2, Phc4 and Phc13 (Houlden etal. 1996) were
used to genotype samples. Amplification and product sepa-
ration using capillary electrophoresis were conducted at the
Australian Genome Research Facility (AGRF), Melbourne,
Australia. Genotypes were replicated three or four times
according to total DNA concentration of the sample (Wed-
rowicz etal. 2013).
DNA binning and the production of consensus genotypes
were undertaken using R statistical software (R Core Team
2014). Raw microsatellite allele sizes were visualised and
binned using the MsatAllele package (Alberto 2009). Con-
GenR (Lonsinger and Waits 2015) was used to generate
consensus genotypes from replicate data, which were then
checked by eye. Genotypes with less than eight successfully
amplified and scored loci were removed. Allelematch (Galp-
ern etal. 2012) was used to identify identical or almost iden-
tical genotypes (pairs with less than three mismatched loci,

Conservation Genetics
1 3
potentially representing matching genotypes with errors) as
per Paetkau (2003). Identical genotypes, and genotypes with
a small number of mismatched loci that could not be refuted
as errors, were removed from the dataset.
Genetic statistics
The R package strataG (Archer etal. 2016) was used to
test for deviations from Hardy–Weinberg (HW) proportions
and to calculate the number of private alleles (AP) for each
population. Observed (HO) and expected (HE) heterozygo-
sity, allelic richness (AR) and the proportion of total sampled
alleles found in each population (A%) were calculated using
the diveRsity package (Keenan etal. 2013). The corPlot
function in diveRsity plots differentiation against locus poly-
morphism to investigate potential bias in FST type estimates
(Keenan 2014). It was found that FST was likely to be biased
for this data, in which case Djost (Jost 2008) is suggested as
a more suitable measure of genetic differentiation (Keenan
2014), although it is recommended that other measures of
differentiation be used in combination with FST (Meirmans
and Hedrick 2011). Both FST and Djost were therefore used
to estimate genetic differentiation. Genetic and geographic
distances were calculated and Mantel tests conducted in the
R package adegenet(Jombart 2008).
Population structure
Analyses for the detection of population structure were
carried out using the microsatellite genotype data and
the Bayesian clustering programs, STRU CTU RE 2.3.4
(Pritchard etal. 2000) and BAPS 6.0 (Corander etal. 2008).
Both STUC TUR E and BAPS group individuals into clusters
in such a way that deviations from HW proportions and link-
age disequilibrium (LD) are minimised, but differ in the way
that the number of populations (K) is inferred (Latch etal.
2006). Spatial Bayesian clustering methods in GENELAND
(Guillot etal. 2008) were used for the analysis of the fine
scale genetic data. Population structure was also analysed
using discriminant analysis of principal components (DAPC)
in adegenet, which is based on genetic distances rather than
minimisation of HW proportions and LD (Jombart 2008).
The STRU CTU RE software was run with admixture and
correlated allele frequencies using 3,000,000 Markov chain
Monte Carlo (MCMC) iterations preceded by a burn-in of
1,000,000 iterations for K from 1 to 20. The most likely
number of clusters inferred by STRU CTU RE was chosen
based on both Pritchard etal. (2010), where the most likely
K has the lowest posterior probability from high values that
have plateaued, and ∆K described by Evanno etal. (2005),
which is based on the rate of change between log probabili-
ties for consecutive values of K. BAPS was run ten times
each using the non-spatial model for maximum values of K
between 5 and 30 (5–15, 20, 25, 30). The most likely value
of K was chosen based on the 10 partitions with the lowest
marginal likelihoods (Corander etal. 2013).
Excluding samples without reliable spatial location data
(i.e. those sampled from the shelter), GENELAND was used
to test for fine scale genetic substructure in South Gippsland.
Population data were analysed in GENELAND using a spa-
tial model and correlated allele frequencies. The maximum
number of populations, K, was set to 20 with 1,000,000
iterations and an additional 50,000 burn-in iterations. The
thinning parameter was set to 1000 and 20 independent runs
were carried out. As recommended by Guillot (2012), the
most likely number of clusters inferred by GENELAND and/
or the best model amongst runs was chosen according to the
run with the highest average posterior probability. Conver-
gence was assessed by seeking evidence of non-convergence
as described in the GENELAND manual (Guillot 2012).
Landscape data
ArcGIS 10.0 (Esri 2010) was used to investigate differences
in habitat types between population clusters (inferred by
GENELAND) within South Gippsland. Ecological Vegeta-
tion Classification (EVC) data for public land were obtained
from DELWP (2017), while data for EVCs and dominant
tree species within plantation estate were provided by HVP.
Data from DELWP and HVP were merged to provide a sin-
gle layer containing vegetation information for both public
(DELWP) and privately (HVP) managed land across the
region (at a cell size of 25m × 25m). Cluster assignment
data for individuals, inferred by GENELAND, were then
overlayed onto the vegetation layer. Using the join function
in ArcGIS, individuals were then assigned to a habitat type,
based on their sampling location.
Mitochondrial sequence data
Three regions of the mitochondrial genome (mtDNA) were
targeted for sequencing, including approximately 700bp
of the mitochondrial control region (Fowler etal. 2000),
1001bp of the cytochrome B gene (cytB) and a 1559bp
stretch of DNA spanning genetic sequence of NADH dehy-
drogenase subunits five and six (ND5/6). Primers for the
amplification of cytB and ND5/6 were designed using koala
mitochondrial sequence (Genbank accession NC_00813:
Munemasa etal. 2006) and Primer-BLAST (Ye etal. 2012).
The control region PCR used primers KmtL1 and KmtH2
designed by Fowler etal. (2000); cytB DNA was ampli-
fied using primers cytB-F (5′-CCC ATC CAA CAT CTC TAC
CT-3′) and cytB-R (5′-ATG TGG TGG ATG CTA CTT GG-3′)
and the ND5/6 PCR used primers, ND-F (5′-CGC AAC AGG
AAA ATC AGC CC-3′) and ND-R (5′-TAG TTA GTG GTG
GCT TGG GG-3′).

Conservation Genetics
1 3
Mitochondrial PCRs were carried out using BIO-X-
ACT™ Short DNA Polymerase (Bioline) or MyTaq™
2× Red Mix (Bioline). Reactions carried out with BIO-X-
ACT™ Short DNA Polymerase consisted of 1× OptiBuffer,
0.25 X Hi-Spec additive, 2mM MgCl2, 0.5mM each dNTP,
0.25µM of each primer and 1 unit of BIO-X-ACT™ Short
DNA Polymerase made up to 20µL with water. PCRs using
MyTaq™ Red Mix were made up using 0.25µM of each
primer, 0.1µg/µL bovine serum albumin (BSA) and 1×
MyTaq™ Red Mix made up to a total volume of 40µL with
water.
PCR products were purified using the Wizard® SV Gel
and PCR Clean-Up System (Promega) and sequencing was
carried out at AGRF, Melbourne. Where a haplotype was
observed only once, PCR and sequencing were repeated
independently to confirm. Sequences were trimmed and
aligned using MEGA 6 (Tamura etal. 2013). Aligned
sequences were exported in FASTA format for use by the R
packages, apex (Jombart etal. 2017), pegas (Paradis 2010)
and ape (Paradis etal. 2004), in which sequences were con-
catenated and a haplotype network produced. Estimates of
nucleotide diversity (π), haplotype diversity (h) and pairwise
differences between populations were obtained using ARLE-
QUIN 3.5.2.2 (Excoffier and Lischer 2010) while haplotype
AMOVA and differentiation (ΦST ) were calculated using
GenAlEx 6.5 (Peakall and Smouse 2012).
Results
Sampling andmicrosatellite genotyping
Scats were collected from a total of 583 putative individ-
uals during this study. DNA quality was high with DNA
samples from 467 (80%) putative individuals passing the
quality screening. After removal of genotypes with less than
eight (out of 12) successfully amplified and scored loci, 429
(74%) samples provided reliable genotypic data. Matching
genotypes indicated that 67 individuals had been sampled
more than once. After removing duplicates, 362 individual
koalas had been sampled. The majority of the individuals
sampled (n = 222, 61%) were from the South Gippsland
(SG) region; 188 were obtained from scats and 34 from ear
tissue. Genotypes from scat samples were obtained from
both wild (n = 155) and shelter (n = 33) koalas in the SG
region. Genotypic data were also obtained from reference
populations in Victoria (total n = 93); Cape Otway (OTW,
n = 50), French Island (FI, n = 9), Mallacoota (MC, n = 3)
and Raymond Island (RI, n = 31) and from interstate popula-
tions (total n = 48); south east New South Wales (SENSW,
n = 12), north east New South Wales (NENSW, n = 24) and
south east Queensland (SEQLD, n = 12).
Using MICROCHECKER, there was no evidence for null
alleles within the microsatellite loci, except for Pcv2 in the
SG and OTW populations. Genotypes were in HW propor-
tions for all loci and populations, except for Pcv2 for the
SG, RI and OTW populations and Pcv25.2 for the FI group.
The Pcv2 locus was retained, as the spatial distribution of
Pcv2 genotypes in the SG and OTW populations identified
regions where homozygotes for offending alleles were clus-
tered, suggesting population structure.
Population structure
Using STRU CTU RE, the ∆K method described by Evanno
etal. (2005) indicated K = 2 as the most likely number of
population clusters, which divided Victorian samples from
more northern (SENSW, NENSW and SEQLD) koala popu-
lations with very high cluster membership; 96% of northern
individuals and 100% of southern individuals had a q value
higher than 0.90. The ∆K method of inferring the number of
population clusters can, however, suffer from falsely inflated
values at K = 2 (Campana etal. 2011), making it impor-
tant to also analyse other relevant values of K for biological
meaning. The ∆K plot showed an additional, less intense,
peak at K = 7 and the maximum log probability of the data
[LnP (D)] also indicated the most likely number of popu-
lations to be K = 7. The seven clusters inferred by STRU
CTU RE also indicated differentiation between Victorian
and more northern koala populations, however, Victorian
koalas were further divided into three main clusters, those
of French Island origin (FI, OTW and MC), Phillip Island
origin (RI) and South Gippsland (SG). Cluster assignment
for individuals of French Island and Phillip Island origin
were high, with 85% (52/61) and 74% (23/31), respectively,
of individuals having cluster assignment (q) values greater
than 0.8. One individual appearing to be of SG origin was
detected on RI, suggesting it had been translocated. Further
structure was also inferred in SG with four population clus-
ters being present, however, a greater amount of admixture
was found in SG evidenced by a lack of strong cluster mem-
bership (Fig.2).
The population structure determined by BAPS correlated
well with the STRU CTU RE results, however, the number
of clusters inferred by BAPS for the data set was double
that of STRU CTU RE, at K = 14 (Fig. 2). As with STRU
CTU RE, individuals sampled in Victoria were clustered into
six distinct groups, those of French Island origin (OTW/
FI/MC), those of Phillip island origin (RI) and SG which
was divided into four population sub clusters (Fig.2). BAPS
distinguished between the three interstate sample locations
where STRU CTU RE did not. Within the more northern ref-
erence populations, BAPS divided individuals into clusters
which matched their sampling location. The two sampling
sites in SENSW were clearly separated, divided according

Conservation Genetics
1 3
to coastal or inland sampling regions. Individuals sampled
from Queensland were also separated into two groups,
broadly corresponding to individuals sampled in coastal or
more inland regions. Discriminant analysis of principal com-
ponents (DAPC) in adegenet by sampling location supported
population structuring provided by BAPS (Online resource
1, Fig. S1). Further population structuring in the SG koala
population was also indicated using DAPC, where five sub
clusters were inferred (Online resource 1, Fig. S2).
Fine scale genetic structure inSouth Gippsland
STRU CTU RE, BAPS and DAPC all gave an indication of
further population structure in SG, supported by FST values
ranging from 0.03 to 0.06 between the four BAPS assigned
populations in SG. Fine scale population structure in SG
was therefore investigated using GENELAND, where seven
population clusters were inferred. Six spatially well-defined
population clusters with more than five assigned individuals
were identified (Fig.3). Population structure in some regions
was not well defined, with individuals from multiple popula-
tion clusters present.
All six population clusters were significantly differenti-
ated using Djost, but five of the fifteen comparisons were not
significant using FST (Online resource 1, TableS2). Genetic
differentiation between SG population clusters ranged from
0.0002 (cluster 2 − cluster 3) to 0.09 (cluster 1 − cluster 5)
for Djost and between 0.003 (cluster 1 − cluster 3) to 0.12
(cluster 1 − cluster 5) for FST. Genotypes conformed to HW
proportions for all six populations at all loci except for locus
Pcv6.1 in population cluster 4, potentially indicative of fur-
ther fine scale structure. Genotypic AMOVA between Vic-
torian populations (SG, RI and FI) indicated 9% between
population variation, 4% variation between the six subpopu-
lations in SG and 0.4% variation between individuals within
subpopulations.
Using GIS, the spatial distribution of three (clusters
3, 4, and 5) of the seven population clusters inferred by
Fig. 2 a Inferred population structure using STRU CTU RE and
BAPS. Horizontal bar plots represent 362 individual koalas. Differ-
ent colours on the same horizontal bar represent the estimated pro-
portion of the individual’s ancestry assigned to a particular popula-
tion cluster. Solid black lines separate different sample locations
while dotted black lines separate samples from Cape Otway, French
Island and Mallacoota (all French Island descent). Sample areas are
labelled SEQLD South east Queensland (n = 12), NENSW North east
New South Wales (n = 24), SENSW South east New South Wales
(n = 12), OTW Cape Otway (n = 50), FI French Island (n = 9), MC
Mallacoota (n = 3), RI Raymond Island (n = 31) and SG South Gipp-
sland (n = 222). b Neighbour joining tree using genotypic data and
Provesti’s distance based on the main population clusters identified
using BAPS. Clusters with less than three individuals were excluded
while the four clusters identified in the South Gippsland population
were simplified to two by combining clusters with FST values less
than 0.04. (Color figure online)
▸

Conservation Genetics
1 3
GENELAND (Fig.3) appeared to correspond to the distri-
bution of differing habitat types (Online resource 1, Figs.
S3 and S4). Cluster 3 was located at a site that consisted
mainly of lowland forest (consisting of messmate and pep-
permint) and, to a lesser extent, damp forest (where mess-
mate, blue gum and mountain grey gum are the common tree
species). Koalas in cluster 4, were mainly found in native
forest within HVP estate, where blue gum is the dominant
species. On public land, koalas within cluster 4 were mostly
sampled within herb-rich foothill forest and damp forest
(consisting of messmate, blue gum and mountain grey gum).
Lastly, koalas in cluster 5 were concentrated where yellow
stringybark is the dominant tree species.
Fine‑scale isolation bydistance
Discrete population structure can arise due to the presence of
clinal structure (isolation by distance, IBD) and conversely,
the detection of clinal structure can result from the presence
of discrete structure (Meirmans 2012). Determining whether
population structure is discrete or clinal can therefore be
difficult (Ruiz-Gonzalez etal. 2015). A Mantel test between
all wild sampled individuals in SG indicated the presence of
IBD (p = 0.002). Tests for IBD within GENELAND inferred
populations were, however, not significant (p > 0.23) except
for population cluster 4 (p = 0.008). Detection limits may,
however, be affected by small sample sizes. It is therefore
unclear whether the population structure in SG is discrete
or clinal.
Concatenated mtDNA haplotypes
Concatenated DNA sequence data (control region, cytB and
ND5/6) were obtained for a subset of samples (n = 55) rep-
resenting populations from both Victorian (SG n = 15, OTW
n = 3 and RI n = 6) and more northern (QLD n = 9, NENSW
n = 9, SENSW n = 13) reference sites. After alignment and
trimming of mitochondrial DNA sequences, 641bp of the
Fig. 3 Population substructure in South Gippsland inferred using
GENELAND (n = 155). Each point represents a sampled individual
and colours are indicative of different population clusters, which are
overlayed with polygons delineating regions of population cluster-
ing for greater clarity. Spatially well-defined population clusters with
more than six individuals are numbered 1 through to 6. (Color figure
online)

Conservation Genetics
1 3
control region, 933bp of cytB and 1381bp of ND5/6 were
obtained. Sequence data were deposited in GenBank under
accession numbers KY979201–KY979210 (control region),
KY979211–KY979220 (cytB) and KY979221–KY979230
(ND5/6). Concatenated sequence consisting of 2955 DNA
base pairs identified 20 haplotypes across sampling areas
(Fig.4a). Apart from Hap16, which was found in both SG
and in French and Phillip Island derived populations, all
haplotypes were specific to a given region that was sampled.
Compared to the control region alone, inclusion of the cytB
(933bp) and ND5/6 (1381bp) regions provided an addi-
tional 26 variable sites which were able to differentiate indi-
viduals with identical control region haplotypes present in
separate regions (Online resource 1, TableS3). For example,
koalas sampled around 350km apart (from SG and coastal
SENSW) were found to have the same control region hap-
lotype, Pc17, but could be differentiated by a variable site
in the ND5/6 region. Sequence data for the cytB and ND5/6
mtDNA regions did not add greatly to the discrimination
of samples collected in the south, as different haplotypes
detected for cytB (three Victorian haplotypes) and ND5/6
(two Victorian haplotypes) were generally associated with
a specific control region haplotype.
Mitochondrial control region haplotypes
To obtain a large sample set for the investigation of hap-
lotype diversity in the South Gippsland region, mtDNA
was sequenced at the control region alone for an addi-
tional 150 randomly selected samples from Victoria (SG,
n = 110, OTW/FI, n = 20, RI, n = 20). Six previously unre-
ported mtDNA control region haplotypes were detected
in this study; four in the SG study area (Pc56, Pc57, Pc58
and Pc59) and one each in SENSW (Pc55) and NENSW
(Pc54) (detected in samples for which concatenated mtDNA
sequences were obtained). New control region haplotypes
Fig. 4 a Haplotype network based on 2955 bp of concatenated
mtDNA sequence (control region, cytB and ND5/6). Sampled popula-
tions were South Gippsland (SG, VIC), Raymond Island (RI, VIC),
Cape Otway/French Island (OTW/FI, VIC), south east New South
Wales (SENSW), north east New South Wales (NENSW) and south
east Queensland (SEQLD). Haplotypes are numbered 01–20 which
correspond to Hap01 to Hap20 in the text. The number of base pair
differences between haplotypes are shown in small white circles on
the lines adjoining haplotypes. Lines joining haplotypes differing by
one base pair are unlabelled. b mtDNA control region haplotype net-
work for Victorian samples

Conservation Genetics
1 3
were named using standardised labels as recommended by
Neaves etal. (2016). Relationships between control region
haplotypes detected in Victorian koalas sampled are illus-
trated by the haplotype network in Fig.4b. The control
region haplotype network was star shaped with the domi-
nant haplotype (Pc27) surrounded by five low frequency
variants (Pc4, Pc26, Pc57, Pc58, Pc59), suggestive of recent
expansion in evolutionary terms (Fig.4b). Two slightly more
divergent control region haplotypes, Pc17 and Pc56 were
also found in SG (Fig.4b). Control region haplotypes sup-
ported the population clusters identified by GENELAND,
with differences in the haplotype frequencies detected in
each population cluster (Table1).
Genetic differentiation betweenVictorian
populations
Private alleles are unique to a particular population and can
provide an indication of genetic distinctiveness. In SG, 38
alleles not found in other Victorian koala populations were
detected, indicating that koalas in SG are genetically distinct
from the island populations sampled. There was moderate
differentiation between the SG and RI (Djost 0.04 and FST
0.08) and SG and OTW/FI/MC (Djost 0.10 and FST 0.12)
populations.
Tests of differentiation (ΦST) using concatenated haplo-
typic data between the SG and island derived populations
were not significant (p = 0.08). Using control region haplo-
types alone, for which a greater amount of data were avail-
able, significant differentiation between the SG and island
derived populations was detected (ΦST = 0.07, p = 0.04),
indicating that the SG population is also differentiated from
populations of both French Island and Phillip Island origin
at the mtDNA control region.
Broad scale genetic differentiation
For both genotypic and haplotypic data, genetic differentia-
tion between populations increased with increasing distance
(Online resource 1, TableS4, Fig. S5). Using genotypic data,
pairwise population differentiation was highly correlated
to geographic distance (Djost: R2 = 62%, p = 0.0005, FST:
R2 = 44%, p = 0.007). Haplotypic data also indicated a pat-
tern of isolation by distance (Average pairwise differences:
R2 = 40%, p = 0.02) between populations (Online resource
1, Fig. S5).
When genotypic data were stratified as per the dendro-
gram in Fig.2a, AMOVA showed between population vari-
ation of 16%, between subpopulation variation of 10% and
within subpopulation variation of 3%. Variation in haplo-
typic data were more structured than the genotypic data,
Table 1 Genetic statistics for
the six population clusters
(with six or more individuals)
in South Gippsland inferred
by GENELAND (Fig.3) using
microsatellite genotypes (upper
section) and mitochondrial
control region sequence data
(lower section)
Haplotypes identified in only one population cluster are shown in bold
nmt number of sequences, nh number of haplotypes, npm number of polymorphic sites, pw average number
of pairwise differences, h ± sd% haplotype diversity, π ± sd% nucleotide diversity, N Number of individual
genotypes, A allelic diversity; the mean number of alleles per locus, A% the percentage of alleles, from all
populations, found in each specific population, AR allelic richness; the mean number of alleles per locus,
corrected for differences in sample size (based on a sample size of six), PA private alleles; alleles unique to
a single population, HO observed heterozygosity, HE expected heterozygosity
Cluster 1 2 3 4 5 6
N6 7 12 38 25 20
A3.2 4.0 4.1 5.2 4.7 3.8
A%55 67 69 81 77 65
AR2.9 3.3 3.3 3.3 3.0 3.0
PA130632
HO0.58 0.67 0.64 0.59 0.50 0.57
HE0.55 0.59 0.61 0.57 0.49 0.55
nmt 5 22 7 14 13 13
nh 232433
π ± sd% 0.06 ± 0.08 0.11 ± 0.10 0.09 ± 0.09 0.21 ± 0.10 0.12 ± 0.10 0.31 ± 0.21
Pc4 20% 27% – – – –
Pc17 – 5% – 14% 8% 23%
Pc27 80% 68% 57% 71% 85% 69%
Pc56–––––8%
Pc57 – – – 7% 8% –
Pc58 – – – 7% – –
Pc59 – – 43% –––

Conservation Genetics
1 3
with 61% of haplotypic variation found between SEQLD
and more southern sample regions (NENSW, SENSW and
VIC), alongside 20% between subpopulation variation and
19% within subpopulation variation.
Genetic diversity
The SG koala population was found to have greater genetic
diversity than populations originating from French or Phillip
Islands (OTW or RI, respectively). The SG koala population
had a mean of 7.2 alleles per locus while the OTW and RI
populations both had an average of 3.3 alleles per locus.
Allelic richness (mean alleles per locus corrected for dif-
ferences in sample size) was also found to be significantly
greater in the SG population (AR = 5.1) compared to either
of the island populations (OTW AR = 3.2; RI AR = 3.3,
F = 9.2, p = 0.001) with an average of two extra alleles per
locus (Table2).
Using concatenated haplotypes, nucleotide diversity was
higher in the SG (0.14 ± 0.007) population compared to the
OTW/FI (0.00) and RI (0.00) populations which comprised
a single haplotype (Table2; Fig.4a). Overall, eight Victo-
rian control region haplotypes were identified (Fig.4b). All
eight were present in the SG koala population, while only
one, the most common SG haplotype (Pc27), was identi-
fied in the island populations (OTW/FI and RI, χ2 = 14.5,
p = 0.04). These data indicate that the SG koala population
is distinct and has significantly greater genetic diversity than
other Victorian koala populations sampled.
The SG and SENSW koala populations had compara-
ble allelic richness with respective averages of 4.2 and 4.1
alleles per locus (Table2) while nucleotide diversity (con-
catenated mtDNA data) was slightly lower in the SENSW
(0.11 ± 0.005) population compared to the SG population
(0.14 ± 0.007). Compared to the SG population, higher
nuclear genetic diversity was found in the NENSW and
SEQLD populations, with allelic richness of 6.1 (T = − 2.3,
p = 0.04) and 5.3 (T = − 3.4, p = 0.006) respectively
(Table2). Similarly, nucleotide diversity using concatenated
haplotypes was highest in the SEQLD (0.21 ± 0.015) and
NENSW (0.20 ± 0.014) koala populations.
Discussion
The koala is an iconic species, endemic to Australia and
is the last surviving member of the Phascolarctidae family.
Low genetic variation is of genuine concern for the future
viability of koala populations in Victoria and South Aus-
tralia, as a lack of genetic diversity can affect population
fitness, hindering the ability to adapt to future environmen-
tal change. In this study, we have shown that koalas in the
Table 2 Summary of
genetic statistics for sampled
populations using microsatellite
genotypes (upper section) and
concatenated mtDNA sequence
(lower section)
In the upper section, comparisons made between Victorian populations only are indicated by ‘VIC’ in
superscript following the parameter label
nmt number of sequences, nh number of haplotypes, npm number of polymorphic sites, pw average num-
ber of pairwise differences, h ± sd% haplotype diversity, π ± sd% nucleotide diversity, N number of indi-
vidual genotypes, A allelic diversity; the mean number of alleles per locus, A% the percentage of alleles,
from all populations, found in each specific population, AR allelic richness; the mean number of alleles per
locus, corrected for differences in sample size, PA private alleles; alleles unique to a single population, HO
observed heterozygosity, HE expected heterozygosity
VIC (SG) VIC (OTW) VIC (RI) NSW (SE) NSW (NE) QLD (SE)
N222 50 31 12 24 12
A7.2 3.3 3.3 4.6 7.9 6.5
A%63 32 32 43 69 56
A%
VIC 99 53 53 – – –
AR4.17 2.83 3.03 4.07 6.08 5.28
AR
VIC 5.11 3.15 3.27 – – –
PA11 0 0 4 12 10
PA
VIC 38 0 1 – – –
HO0.59 0.44 0.50 0.65 0.68 0.57
HE0.60 0.45 0.52 0.61 0.73 0.68
nmt 15 3 6 13 9 9
Nh 5 1 1 3 5 7
npm 11 0 0 12 16 15
Pw 4.1 0 0 3.4 5.9 6.1
h ± sd% 73 ± 6 – – 56 ± 14 74 ± 5 84 ± 2
π ± sd% 0.14 ± 0.007 – – 0.11 ± 0.005 0.20 ± 0.014 0.21 ± 0.015

Conservation Genetics
1 3
South Gippsland region are differentiated from all other
koala populations sampled. They also have a significantly
greater level of genetic diversity compared to other Victo-
rian koala populations, retaining a greater proportion of the
ancestral diversity that was lost post European settlement.
The importance of conserving the koala gene pool in the
South Gippsland region therefore cannot be overstated, as
they carry additional genetic diversity that is not present in
populations established by island animals.
Conservation genetics can provide information that may
alert managers to issues affecting a population’s genetic
health such as population isolation, limited gene flow and
inbreeding. Obtaining sufficient, truly representative, sam-
ple sizes for genetic studies can, however, be difficult, slow
and expensive. This study demonstrates the power of non-
invasive sampling, using DNA from koala scats to obtain
genetic data with the ability to inform and monitor conserva-
tion strategies. Probably one of the most concerning future
environmental challenges for the koala, will be the effects
of climate change (Ellis etal. 2010) which may alter koala
habitat distribution and suitability (Adams-Hosking etal.
2012; González-Orozco etal. 2016) and modify leaf chemis-
tries, potentially rendering currently preferred koala dietary
species unsuitable (Moore and Foley 2000; DeGabriel etal.
2010). Molecular technologies, such as those described here,
provide a tool for longitudinal genetic monitoring of popula-
tion responses to environmental change. The rapid collection
of contemporary, empirical data, will expedite the acquisi-
tion of knowledge, allowing evidence based conservation
strategies to be implemented and monitored over time.
South Gippsland koalas are genetically distinct
fromother Victorian populations
Genetic structure in populations occurs due to deviations
from random mating which may result from differing lev-
els of population isolation or fragmentation (Frankham
etal. 2012). Population structure can help to reveal popula-
tion ancestries, while the extent of differentiation between
populations can provide a measure of how different two
populations are; something that is particularly important
to know when assessing risks associated with moving ani-
mals between populations for conservation purposes such
as genetic rescue (Frankham etal. 2011; Frankham 2016).
As inferred by the STRU CTU RE and BAPS plots (Fig.2),
the South Gippsland, Cape Otway (French Island origin) and
Raymond Island (Phillip Island origin) populations were all
moderately differentiated from one another (Djost 0.04–0.10,
FST 0.08–0.12), indicating that the South Gippsland koala
population is distinct from koalas originating from both
French and Phillip Islands. Previously reported levels of
genetic differentiation between French Island derived pop-
ulations and the South Gippsland population have ranged
from weak (FST = 0.05; Houlden etal. 1996) to moderate
(FST = 0.11; Seymour etal. 2001) to strong (FST = 0.25; Lee
etal. 2011).
In this wide-scale study, population substructure was evi-
dent within the South Gippsland region (Fig.3; Table1).
Variable genetic differentiation estimates between studies
may therefore be attributable to differing sampling locations,
regimes and sizes. Discrepancies may also be due to the loss
of particular subsets of the South Gippsland koala popula-
tion due to events occurring between studies. For example, a
high proportion of the samples used for the Lee etal. (2011)
study were obtained from individuals who had succumbed to
the 2009 bushfires; although searches were undertaken, few
samples were obtained from the area affected by the 2009
bushfires for this study. It is therefore possible that different
subsets of the diversity present in South Gippsland koalas
have been sampled by different studies.
South Gippsland koalas have greater genetic
diversity thanother Victorian populations
Genetic variation is important as it provides populations
with the capacity to adapt and survive environmental
changes, while decreased variation is found to negatively
affect survival, growth and reproduction rates (Reed and
Frankham 2003; Frankham etal. 2012). Some koala popu-
lations founded by island stock are currently at a high den-
sity, presently appearing unaffected by their low diversity.
These low diversity populations have, however, only existed
for a relatively short time. Stochastic factors play a signifi-
cant role in determining the outcome of low diversity for
a population (Reed 2010). Currently overabundant koala
populations may not yet have been subjected to pressures
severe enough to cause widespread population decline and
extirpation. Monitoring these populations for early signs of
population decline may assist in their conservation should
the negative effects of low genetic diversity become apparent
in the future. Indeed, examples exist where koalas were once
overabundant but have declined to extremely low densities;
these include Wilsons Promontory, Phillip Island and the
Grampians National Park (Wedrowicz etal. 2017b).
Both genotypic and haplotypic data revealed a signifi-
cantly greater level of genetic diversity in South Gippsland.
These results indicate that the relatively small numbers of
island koalas translocated to the area did not result in the
swamping of local genetic diversity and translocated koalas
may not have successfully integrated with resident popula-
tions at any level. It may be that the South Gippsland pop-
ulation had recovered to sufficient size by the time these
translocations occurred, such that low levels of integration
would have had little effect on levels of differentiation and
diversity.

Conservation Genetics
1 3
Greater genetic diversity in the South Gippsland koala
population could confer increased evolutionary potential
relative to island derived populations in Victoria. How-
ever, although neutral genetic markers are commonly used
to estimate evolutionary potential, the relationship may be
weak (Reed and Frankham 2001); further work to directly
estimate evolutionary potential using adaptive loci, in both
South Gippsland and island derived populations, should be
used to evaluate the risk of future declines due to low evo-
lutionary potential.
Greater diversity in South Gippsland, compared to both
Cape Otway (French Island origin) and Raymond Island
(Phillip Island origin) populations, provides strong sup-
port that South Gippsland koalas are derived from rem-
nant populations having survived in the region at a time
when most other Victorian populations are thought to have
become extirpated or reduced to extremely low numbers
(Lewis 1934, 1954). This reinforces and extends studies
conducted by Houlden etal. (1999) and Lee etal. (2011),
which demonstrated genetic differences and greater diversity
in the South Gippsland koala population compared to island
derived populations (both French and Phillip Islands using
mtDNA and French Island alone using nuclear DNA).
Population substructure ispresent intheSouth
Gippsland koala population
Subtle population substructure within the South Gippsland
koala population was detected using both genotype and
haplotype data, although it is unclear whether the observed
structure is discrete or clinal. In either case, the presence of
genetic structure across the region indicates that gene flow is
restricted. Further work is needed to investigate the reasons
for population substructure in South Gippsland.
Predominant eucalypt species vary across discrete koala
habitats within South Gippsland. Three population clusters
identified by GENELAND correspond with differences
between dominant tree species within the region occupied
by each inferred cluster. Previous studies have shown high
levels of site fidelity in koalas, demonstrating a strong ten-
dency for philopatry (Mitchell 1990; Whisson etal. 2016).
This may indicate a preference for individuals to remain in
areas containing habitat similar to their natal area (Stamps
and Swaisgood 2007), suggesting that the population sub-
structure observed may reflect, in part, recent patterns of
koala dispersal.
Another possibility is that the koala population in South
Gippsland was continuous pre-European settlement but, as
the forests were cleared for agriculture and the koala popula-
tion dwindled, small numbers of individuals survived within
isolated patches of habitat. When mass farm failures and
abandonment occurred in early 1900s, leading to reaffor-
estation and conversion of much of the land to plantation
(Legg 1986; Wedrowicz etal. 2017b), re-expansion of koala
populations across the landscape may have resulted in the
fine scale pattern of genetic structure observed here, where
each cluster represents a koala colony isolated during the
period of severe forest fragmentation.
Past and continuing levels of habitat fragmentation are
also likely to have influenced patterns of genetic structure in
South Gippsland. Further analyses using landscape genetic
approaches (Storfer etal. 2006) would be useful to identify
potential barriers to koala gene flow and gain insights into
how koalas utilise differing landscapes within South Gipps-
land. Landscape genetic methods may also provide a greater
understanding of the nature of the genetic structure detected
in South Gippsland (Ruiz-Gonzalez etal. 2015).
Southern koala populations appear lessdiverse
thannorthern populations
Isolation by distance occurs where gene flow between popu-
lations is sufficiently limited so as to result in the differen-
tiation of neighbouring populations (Frankham etal. 2012).
Both genotypic and haplotypic data showed a strong pattern
of isolation by distance indicating that, historically, koalas
(and their habitat) are likely to have been either continuously
distributed (with limited dispersal) along Australia’s east or
consisting of a series of subpopulations for which low lev-
els of migration could occur between adjacent populations
(stepping stone model; Frankham etal. 2012). Due to local
extinctions and habitat degradation, few koala populations
are likely to remain connected by the low levels of gene flow
that historically occurred across their range. This may have
a negative effect on the conservation of genetic diversity for
the species (Weeks etal. 2016).
Reconnecting nearby patches of koala habitat via cor-
ridors or stepping stones would be one strategy that could
increase gene flow towards historic levels, thereby minimis-
ing further losses of genetic diversity. Reconnecting habitat
will also be important because one response of wild popula-
tions to climatic changes may be to shift to their distribution
to more suitable habitat (Nuñez etal. 2013; McGuire etal.
2016), something that may not possible where habitats are
separated by large distances.
Compared to koala populations in Victoria and South
Australia, genetic diversity tends to be higher in the more
northern populations (Houlden etal. 1996, 1999). Koala
populations are likely to have undergone substantial losses
of genetic diversity Australia wide due to dramatic declines
post European settlement. A number of European species,
however, exhibit decreasing diversity in the direction of post
glacial population expansion (Hewitt 1999). During the cold,
dry conditions of the glacial periods, potential koala habitat
and therefore koala populations may have been restricted to
refugia in Queensland and/or north east New South Wales

Conservation Genetics
1 3
(Adams-Hosking etal. 2011). More favourable climatic con-
ditions in the preceding interglacial period may have allowed
population expansion, with each subsequent founding event
resulting in reduced genetic diversity in populations as they
expanded southwards (Hewitt 1999). Lower genetic diver-
sity in southern koala populations, such as South Gippsland
and south east New South Wales, relative to more northern
koala populations may thus be due, in part, to the koala’s
evolutionary history.
Genetic diversity present intheSouth Gippsland
koala population must be conserved
Genetic diversity provides populations the ability to tolerate
environmental changes, with the risk of extinction expected
to be higher where genetic diversity is low (Frankham 2005;
Frankham etal. 2012). Climates vary across the koala’s
range, as do genetic and morphological (Briscoe etal. 2015)
characteristics of koalas. How koalas in any one region will
respond to climatic and habitat changes is thus difficult to
know. Conserving diversity across the entire range of the
koala is therefore important.
Conclusions
The South Gippsland koala population is a remnant Victo-
rian population, not derived from the koala translocation
program. It has the highest known level of genetic diversity
of all koala populations in Victoria and South Australia.
Consequently, conservation of the South Gippsland koala
population and its genetic diversity into the future is of high
importance. The South Gippsland koala population requires
a different management approach compared to other Victo-
rian koala populations (where the focus is on the manage-
ment of overpopulation), with an emphasis on conservation
of this population and its genetic diversity. Due to high
population densities in several southern koala populations,
koalas in Victoria and South Australia were excluded from
the 2012 EPBC listing of the koala as Vulnerable (EaCRC
2011a, c, 2012); the South Gippsland koala population ought
to be an exception to that exclusion.
Koala management in Victoria is currently concentrated
on preventing the devastating effects of overpopulation
(Menkhorst 2008; DELWP 2015). As discussed, however,
the lack of genetic diversity in high density populations may
increase their chance of future declines. Other remnant koala
populations (outside of South Gippsland) have not been
identified in Victoria to date. Population remnants may be
at low density, so an ability to carry out analyses from scat
samples will greatly facilitate further investigation. Further
widespread genetic surveys, in Gippsland and across Vic-
toria, may highlight additional populations of conservation
priority and inform strategies to minimise further losses of
genetic diversity in southern koala populations.
Acknowledgements Grand Ridge Plantations Pty. Ltd. (HVP Plan-
tations) and the Holsworth Wildlife Research Endowment—Equity
Trustees Charitable Foundation are thanked for funding this study. For
support and the collection of scat samples we thank Richard Appleton
(HVP), Russell Cluning and staff (Hazelwood Forestry), Chris Allen
and the SAT team (NSW Office of Environment and Heritage), Dr.
Mike Weston, Dr. Raylene Cooke and students from the School of Life
and Environmental Sciences (Deakin University), Jim Whelan (Parks
Victoria), Colleen Wood (Southern Ash Wildlife Shelter), Cheyne
Flanagan (Port Macquarie Koala Hospital), Olivia Woosnam (OWAD
Environment) and colleagues, Susie Zent (Friends of the Gippsland
Bush), Nicole Walsh (Friends of the Strzelecki Koala) and the many
members of the public who kindly took the time to collect koala scats
for this project. Elise Wedrowicz, Shane Wedrowicz and Greg Somer-
ville are thanked for volunteering their time to assist with fieldwork.
This research was also supported in part by the Monash eResearch Cen-
tre and eSolutions Research Support Services through the use of the
MonARCH HPC Cluster. We also thank two anonymous reviewers for
their time and help improving this manuscript. A research permit was
obtained under provisions of the Wildlife Act 1975 and National Parks
Acts 1975 from the Department of Sustainability and Environment
(Permit No. 10004020). Research was also approved by the Monash
University Biological Sciences Animal Ethics Committee (AEC No.
GIPP/2011/03).
References
Adams-Hosking C, Moss P, Rhodes J, Grantham H, McAlpine C
(2011) Modelling the potential range of the koala at the last
glacial maximum: future conservation implications. Aust Zool
35:983–990. https://doi.org/10.7882/AZ.2011.052
Adams-Hosking C, McAlpine C, Rhodes JR, Grantham HS, Moss
PT (2012) Modelling changes in the distribution of the criti-
cal food resources of a specialist folivore in response to
climate change. Divers Distrib 18:847–860. https://doi.
org/10.1111/j.1472-4642.2012.00881.x
Alberto F (2009) MsatAllele_1. 0: an R package to visualize the bin-
ning of microsatellite alleles. J Hered 100:394–397. https://doi.
org/10.1093/jhered/esn110
Allen C (2015) Koala distribution, abundance, habitat and Chlamydia
prevalence studies in the Strzelecki Ranges. NSW Office of Envi-
ronment and Heritage, NSW
Archer FI, Adams PE, Schneiders BB (2016) strataG: an R package
for manipulating, summarizing and analysing population genetic
data. Mol Ecol Resour. https://doi.org/10.1111/1755-0998.12559
Bijlsma R, Bundgaard J, Boerema AC (2000) Does inbreed-
ing affect the extinction risk of small populations?: predic-
tions from Drosophila. J Evol Biol 13:502–514. https://doi.
org/10.1046/j.1420-9101.2000.00177.x
Briscoe NJ, Krockenberger A, Handasyde KA, Kearney MR (2015)
Bergmann meets Scholander: geographical variation in body
size and insulation in the koala is related to climate. J Biogeogr
42:791–802. https://doi.org/10.1111/jbi.12445
Campana MG, Hunt HV, Jones H, White J (2011) CorrSieve: software
for summarizing and evaluating structure output. Mol Ecol Resour
11:349–352. https://doi.org/10.1111/j.1755-0998.2010.02917.x
Corander J, Marttinen P, Sirén J, Tang J (2008) Enhanced Bayes-
ian modelling in BAPS software for learning genetic struc-
tures of populations. BMC Bioinform 9:1. https://doi.
org/10.1186/1471-2105-9-539

Conservation Genetics
1 3
Corander J, Marttinen P, Sirén J, Tang J (2013) BAPS: Bayesian analy-
sis of population structure. Manual Version 60
Cristescu R, Cahill V, Sherwin WB, Handasyde K, Carlyon K, Whisson
D, Herbert CA, Carlsson BLJ, Wilton AN, Cooper DW (2009)
Inbreeding and testicular abnormalities in a bottlenecked popula-
tion of koalas (Phascolarctos cinereus). Wildl Res 36:299–308.
https://doi.org/10.1071/WR08010
DeGabriel JL, Moore BD, Marsh KJ, Foley WJ (2010) The effect of
plant secondary metabolites on the interplay between the internal
and external environments of marsupial folivores. Chemoecology
20:97–108. https://doi.org/10.1007/s00049-009-0037-3
DELWP (2015) Cape Otway Koala management actions. http://www.
wildlife.vic.gov.au/our-wildlife/koalas/koalas-at-cape-otway.
Accessed 27 Mar 2017
DELWP (2017) Native vegetation: modelled 2005 ecological vegeta-
tion classes (with bioregional conservation status) Department of
Environment, Land, Water & Planning. http://www.data.vic.gov.
au. Accessed 1 May 2015
Department of the Environment (2015) Phascolarctos cinereus (com-
bined populations of QLD, NSW and the ACT) in species profile
and threats database. Commonwealth of Australia. http://www.
environment.gov.au/sprat. Accessed 23 Dec 2015
EaCRC (2011a) The koala - saving our national icon. The Senate
Printing Unit, Parliament House. http://www.aph.gov.au/Parlia-
mentary_Business/Committees/Senate/Environment_and_Com-
munications/Completed_inquiries/2010-13/koalas/report/index.
Accessed 1 Apr 2013
EaCRC (2011b) Official committe hansard: status, health and sus-
tainability of Australia’s koala population. Commonwealth of
Australia. http://www.aph.gov.au/Parliamentary_Business/Com-
mittees/Senate/Environment_and_Communications/Completed_
inquiries/2010-13/koalas/hearings/index. Accessed 4 Mar 2016
EaCRC (2012) Koala protected under national environment law. The
Hon. Tony Burke MP, Minister for Sustainability, Environment,
Water, Population and Communities. http://parlinfo.aph.gov.au/.
Accessed 02 May 2012
Ellis WAH, Melzer A, Carrick FN, Hasegawa M (2002) Tree use, diet
and home range of the koala (Phascolarctos cinereus) at Blair
Athol, central Queensland. Wildl Res 29:303–311. https://doi.
org/10.1071/WR00111
Ellis W, Melzer A, Clifton I, Carrick F (2010) Climate change and
the koala Phascolarctos cinereus: water and energy. Aust Zool
35:369–377. https://doi.org/10.7882/AZ.2010.025
ESRI (2010) ArcGIS 10.0. Environmental Systems Research Institute,
Redlands
Evanno G, Regnaut S, Goudet J (2005) Detecting the num-
ber of clusters of individuals using the software STRU CTU
RE: a simulation study. Mol Ecol 14:2611–2620. https://doi.
org/10.1111/j.1365-294X.2005.02553.x
Excoffier L, Lischer HE (2010) Arlequin suite ver 3.5: a new series
of programs to perform population genetics analyses under
Linux and Windows. Mol Ecol Resour 10:564–567. https://doi.
org/10.1111/j.1755-0998.2010.02847.x
Fowler EV, Hoeben P, Timms P (1998) Randomly amplified polymor-
phic DNA variation in populations of eastern Australian koalas,
Phascolarctos cinereus. Biochem Genet 36:381–393. https://doi.
org/10.1023/A:1018701630713
Fowler EV, Houlden BA, Hoeben P, Timms P (2000) Genetic diver-
sity and gene flow among southeastern Queensland koalas
(Phascolarctos cinereus). Mol Ecol 9:155–164. https://doi.
org/10.1046/j.1365-294x.2000.00844.x
Frankham R (2005) Genetics and extinction. Biol Conserv 126:131–
140. https://doi.org/10.1016/j.biocon.2005.05.002
Frankham R (2016) Genetic rescue benefits persist to at least the F3
generation, based on a meta-analysis. Biol Conserv 195:33–36.
https://doi.org/10.1016/j.biocon.2015.12.038
Frankham R, Ballou JD, Eldridge MD, Lacy RC, Ralls K, Dudash
MR, Fenster CB (2011) Predicting the probability of out-
breeding depression. Conserv Biol 25:465–475. https://doi.
org/10.1111/j.1523-1739.2011.01662.x
Frankham R, Ballou JD, Briscoe DA (2012) Introduction to Conserva-
tion Genetics, 2ndedn. Cambridge University Press, New York
Galpern P, Manseau M, Hettinga P, Smith K, Wilson P (2012)
Allelematch: an R package for identifying unique multi-
locus genotypes where genotyping error and missing data
may be present. Mol Ecol Resour 12:771–778. https://doi.
org/10.1111/j.1755-0998.2012.03137.x
González-Orozco CE, Pollock LJ, Thornhill AH, Mishler BD, Knerr
N, Laffan SW, Miller JT, Rosauer DF, Faith DP, Nipperess DA
(2016) Phylogenetic approaches reveal biodiversity threats under
climate change. Nat Clim Change 6:1110–1114. https://doi.
org/10.1038/nclimate3126
Guillot G (2012) Geneland manual: population genetic and morpho-
metric data analysis using R and the Geneland program. http://
www.imm.dtu.dk/~gigu/Geneland/Geneland-Doc.pdf. Accessed
8 Sept 2014
Guillot G, Santos F, Estoup A (2008) Analysing georeferenced popula-
tion genetics data with Geneland: a new algorithm to deal with
null alleles and a friendly graphical user interface. Bioinformat-
ics 24:1406–1407. https://doi.org/10.1093/bioinformatics/btn136
Handasyde K, McDonald I, Than K, Michaelides J, Martin R (1990)
Reproductive hormones and reproduction in the koala. In: Lee
AK, Handasyde KA, Sanson GD (eds) Biology of the koala. Sur-
rey Beatty and Sons, Chipping Norton, pp203–210
Hewitt GM (1999) Post-glacial re-colonization of European biota. Biol
J Linn Soc 68:87–112. https://doi.org/10.1111/j.1095-8312.1999.
tb01160.x
Houlden BA, England PR, Taylor AC, Greville WD, Sherwin
WB (1996) Low genetic variability of the koala Phascolarc-
tos cinereus in south-eastern Australia following a severe
population bottleneck. Mol Ecol 5:269–281. https://doi.
org/10.1046/j.1365-294X.1996.00089.x
Houlden BA, Costello BH, Sharkey D, Fowler EV, Melzer A, Ellis
W, Carrick F, Baverstock PR, Elphinstone MS (1999) Phylogeo-
graphic differentiation in the mitochondrial control region in the
koala, Phascolarctos cinereus (Goldfuss 1817). Mol Ecol 8:999–
1011. https://doi.org/10.1046/j.1365-294x.1999.00656.x
Jombart T (2008) Adegenet: a R package for the multivariate analysis
of genetic markers. Bioinformatics 24:1403–1405. https://doi.
org/10.1093/bioinformatics/btn129
Jombart T, Archer F, Schliep K, Kamvar Z, Harris R, Paradis E, Goudet
J, Lapp H (2017) Apex: phylogenetics with multiple genes. Mol
Ecol Resour 17:19–26. https://doi.org/10.1111/1755-0998.12567
Jost L (2008) GST and its relatives do not measure dif-
ferentiation. Mol Ecol 17:4015–4026. https://doi.
org/10.1111/j.1365-294X.2008.03887.x
Katsura Y, Kondo H, Ryan J, Harley V, Satta Y (2016) Two step sub
functionalisation of SRY during evolution of mammalian sex
determination. (unpublished)
Keenan K (2014) The diversity package. https://cran.r-project.org/web/
packages/diveRsity/vignettes/diveRsity.pdf. Accessed July 2016
Keenan K, McGinnity P, Cross TF, Crozier WW, Prodöhl PA (2013)
Diversity: an R package for the estimation and exploration of pop-
ulation genetics parameters and their associated errors. Methods
Ecol Evol 4:782–788. https://doi.org/10.1111/2041-210X.12067
Latch EK, Dharmarajan G, Glaubitz JC, Rhodes OE (2006) Relative
performance of Bayesian clustering software for inferringpopu-
lation substructure and individual assignment at low levels of
population differentiation. Conserv Genet 7:295–302. https://doi.
org/10.1007/s10592-005-9098-1
Lee T, Zenger KR, Close RL, Phalen DN (2011) Genetic analysis
reveals a distinct and highly diverse koala (Phascolarctos cinereus)

Conservation Genetics
1 3
population in South Gippsland, Victoria, Australia. Aust Mammal
34:68–74. https://doi.org/10.1071/AM10035
Legg S (1986) Farm abandonment in South Gippsland’s Strzelecki
Ranges, 1870–1925: challenge or tragedy. Gippsland Herit J 1:14–22
Lewis F (1934) The koala in Victoria. Vic Nat 51:73–76
Lewis F (1954) The rehabilitation of the koala in Victoria. Vic Nat
70:197–200
Lonsinger RC, Waits LP (2015) ConGenR: rapid determination of con-
sensus genotypes and estimates of genotyping errors from repli-
cated genetic samples. Conserv Genet Resour 7:841–843. https://
doi.org/10.1007/s12686-015-0506-7
Martin R, Handasyde K (1990) Population dynamics of the koala (Phas-
colarctos cinereus) in southeastern Australia. In: Lee AK, Han-
dasyde KA, Sanson GD (eds) Biology of the koala. Surrey Beatty
and Sons, Chipping Norton, 75–84
Martin R, Handasyde K (1999) The koala: natural history, conservation
and management. University of New South Wales, Sydney
McGuire JL, Lawler JJ, McRae BH, Nuñez TA, Theobald DM (2016)
Achieving climate connectivity in a fragmented landscape. Proc
Natl Acad Sci USA 113:7195–7200. https://doi.org/10.1073/
pnas.1602817113
Meirmans PG (2012) The trouble with isolation by distance. Mol Ecol
21:2839–2846. https://doi.org/10.1111/j.1365-294X.2012.05578.x
Meirmans PG, Hedrick PW (2011) Assessing population structure:
FST and related measures. Mol Ecol Resour 11:5–18. https://doi.
org/10.1111/j.1755-0998.2010.02927.x
Menkhorst P (2008) Hunted, marooned, re-introduced, contracepted: A
history of Koala management in Victoria. In: Lunney D, Munn A,
Meikle W (eds) Too close for comfort: contentious issues in human-
wildlife encounters. Royal Zoological Society of New South Wales,
Mosman, pp73–92
Mitchell P (1990) The home ranges and social activity of koalas-a quan-
titative analysis. In: Biology of the koala. Surrey Beatty & Sons Pty
Ltd, Chipping Norton, pp171–187
Moore BD, Foley WJ (2000) A review of feeding and diet selection in
koalas (Phascolarctos cinereus). Aust J Zool 48:317–333. https://
doi.org/10.1071/ZO99034
Munemasa M, Nikaido M, Donnellan S, Austin CC, Okada N, Hasegawa
M (2006) Phylogenetic analysis of diprotodontian marsupials based
on complete mitochondrial genomes. Genes Genet Syst 81:181–191.
https://doi.org/10.1266/ggs.81.181
Neaves LE, Frankham GJ, Dennison S, FitzGibbon S, Flannagan C, Gil-
lett A, Hynes E, Handasyde K, Helgen KM, Tsangaras K (2016)
Phylogeography of the koala (Phascolarctos cinereus) and harmo-
nising data to inform conservation. PLoS ONE 11:e0162207. https://
doi.org/10.1371/journal.pone.0162207
Nuñez TA, Lawler JJ, McRae BH, Pierce DJ, Krosby MB, Kavanagh
DM, Singleton PH, Tewksbury JJ (2013) Connectivity planning
to address climate change. Conserv Biol 27:407–416. https://doi.
org/10.1111/cobi.12014
Paetkau D (2003) An empirical exploration of data quality in DNA-
based population inventories. Mol Ecol 12:1375–1387. https://doi.
org/10.1046/j.1365-294X.2003.01820.x
Paradis E (2010) pegas: an R package for population genetics with an
integrated–modular approach. Bioinformatics 26:419–420. https://
doi.org/10.1093/bioinformatics/btp696
Paradis E, Claude J, Strimmer K (2004) APE: analyses of phylogenetics
and evolution in R language. Bioinformatics 20:289–290. https://
doi.org/10.1093/bioinformatics/btg412
Peakall R, Smouse PE (2012) GenAlEx 6.5: genetic analysis in Excel.
Population genetic software for teaching and research—an
update. Bioinformatics 28:2537–2539. https://doi.org/10.1093/
bioinformatics/bts460
Piggott MP, Taylor AC (2003) Remote collection of animal DNA and
its applications in conservation management and understanding the
population biology of rare and cryptic species. Wildl Res 30:1–13.
https://doi.org/10.1071/WR02077
Pritchard JK, Stephens M, Donnelly P (2000) Inference of population
structure using multilocus genotype data. Genetics 155:945–959
Pritchard J, Wen X, Falush D (2010) Documentation for structure soft-
ware: version 2.3. University of Chicago, Chicago
R Core Team (2014) R: a language and environment for statistical com-
puting. R Foundation for Statistical Computing, Vienna
Reed DH (2010) Albatrosses, eagles and newts, Oh My!: exceptions
to the prevailing paradigm concerning genetic diversity and
population viability? Anim Conserv 13:448–457. https://doi.
org/10.1111/j.1469-1795.2010.00353.x
Reed DH, Frankham R (2001) How closely correlated are molecular and
quantitative measures of genetic variation? A meta-analysis. Evo-
lution 55:1095–1103. https://doi.org/10.1111/j.0014-3820.2001.
tb00629.x
Reed DH, Frankham R (2003) Correlation between fitness and
genetic diversity. Conserv Biol 17:230–237. https://doi.
org/10.1046/j.1523-1739.2003.01236.x
Ruiz-Gonzalez A, Cushman SA, Madeira MJ, Randi E, Gómez-Moliner
BJ (2015) Isolation by distance, resistance and/or clusters? Lessons
learned from a forest-dwelling carnivore inhabiting a heterogene-
ous landscape. Mol Ecol 24:5110–5129. https://doi.org/10.1111/
mec.13392
Seymour AM, Montgomery ME, Costello BH, Ihle S, Johnsson G, St.
John B, Taggart D, Houlden BA (2001) High effective inbreeding
coefficients correlate with morphological abnormalities in popula-
tions of South Australian koalas (Phascolarctos cinereus). Anim
Conserv 4:211–219. https://doi.org/10.1017/s1367943001001251
Stamps JA, Swaisgood RR (2007) Someplace like home: experience,
habitat selection and conservation biology. Appl Anim Behav Sci
102:392–409. https://doi.org/10.1016/j.applanim.2006.05.038
Storfer A, Murphy MA, Evans JS, Goldberg CS, Robinson S, Spear SF,
Dezzani R, Delmelle E, Vierling L, Waits LP (2006) Putting the
‘landscape’ in landscape genetics. Heredity 98:128–142. https://doi.
org/10.1038/sj.hdy.6800917
Tamura K, Stecher G, Peterson D, Filipski A, Kumar S (2013) MEGA6:
molecular evolutionary genetics analysis version 6.0. Mol Biol Evol
30:2725–2729. https://doi.org/10.1093/molbev/mst197
Watson C, Margan S, Johnston P (1998) Sex-chromosome elimination in
the bandicoot Isoodon macrourus using Y-linked markers. Cytogenet
Genome Res 81:54–59. https://doi.org/10.1159/000015008
Wedrowicz F, Karsa M, Mosse J, Hogan FE (2013) Reliable genotyp-
ing of the koala (Phascolarctos cinereus) using DNA isolated from
a single faecal pellet. Mol Ecol Resour 13:634–641. https://doi.
org/10.1111/1755-0998.12101
Wedrowicz F, Mosse J, Wright W, Hogan FE (2017a) Validating the
use of non-invasively sourced DNA for population genetic studies
using pedigree data. Web Ecol 17:9–18. https://doi.org/10.5194/
we-17-9-2017
Wedrowicz F, Wright W, Schlagloth R, Santamaria F, Cahir F (2017b)
Landscape, koalas and people: a historical account of koala popula-
tions and their environment in South Gippsland. Aust Zool 38:518–
536. https://doi.org/10.7882/AZ.2017.007
Weeks AR, Stoklosa J, Hoffmann AA (2016) Conservation of genetic
uniqueness of populations may increase extinction likelihood of
endangered species: the case of Australian mammals. Front Zool
13, 31. https://doi.org/10.1186/s12983-016-0163-z
Whisson DA, Dixon V, Taylor ML, Melzer A (2016) Failure to respond
to food resource decline has catastrophic consequences for koalas
in a high-density population in southern Australia. PLoS ONE
11:e0144348. https://doi.org/10.1371/journal.pone.0144348
Ye J, Coulouris G, Zaretskaya I, Cutcutache I, Rozen S, Madden TL
(2012) Primer-BLAST: a tool to design target-specific primers for
polymerase chain reaction. BMC Bioinform 13:1–11. https://doi.
org/10.1186/1471-2105-13-134