Using genetic techniques to quantify reinvasion, survival and in-situ breeding rates during control/eradication operations.

Using genetic techniques to quantify reinvasion, survivaland in situ breeding rates during control operations

A. J . VEALE,* K. -A. EDGE,† P. MCMURTRIE,† R. M. FEWSTER,‡ M. N. CLOUT* and

D. M. GLEESON§¶*School of Biological Sciences, Tamaki Campus, University of Auckland, Private Bag 92019, Auckland, New Zealand, †Department

of Conservation Te Anau, PO Box 29, Te Anau 9640, New Zealand, ‡Department of Statistics, University of Auckland, Private

Bag 92019, Auckland, New Zealand, §Institute for Applied Ecology, University of Canberra, Bruce ACT 2601, Australia,

¶Ecological Genetics Laboratory, Landcare Research, Private Bag 92170, Auckland 1142, New Zealand

Abstract

Determining the origin of individuals caught during a control/eradication programme

enables conservation managers to assess the reinvasion rates of their target species and

evaluate the level of success of their control methods. We examine how genetic tech-

niques can focus management by distinguishing between hypotheses of ‘reinvasion’

and ‘survivor’, and defining kin groups for invasive stoats (Mustela erminea) on Secre-

tary Island, New Zealand. 205 stoats caught on the island were genotyped at 16 micro-

satellite loci, along with 40 stoats from the opposing mainland coast, and the age and

sex were determined for each individual. Using these data, we compare and combine a

variety of genetic techniques including genetic clustering, population assignment and

kinship-based techniques to assess the origin of each stoat. The population history

and individual movement could be described in fine detail, with results indicating that

both in-situ survival and breeding, and reinvasion are occurring. Immigration to the

island was found to be generally low, apart from in 1 year where around 8 stoats emi-

grated from the mainland. This increased immigration was probably linked to a stoat

population spike on the mainland in that year, caused by a masting event of southern

beech forest (Nothofagus sp.) and the subsequent rodent irruption. Our study provides

an example of some of the ways genetic analyses can feed directly into informing man-

agement practices for invasive species.

Keywords: assignment, connectivity, immigration, invasion, microsatellite, Mustela erminea

Received 26 October 2012; revision received 8 July 2013; accepted 9 July 2013

Introduction

Understanding the level of connectivity between

patches is vital to successfully manage populations. For

invasive species, the ultimate goal of management is

eradication (Myers et al. 2000; Simberloff 2003), and this

is now a common practice for many mammalian pest

species on islands (Towns & Broome 2003; Howald

et al. 2007). In order to eradicate a pest from an area,

the rate of dispersal and re-establishment from other

populations must either be zero or manageable to zero

(Bomford & O’Brien 1995; Broome et al. 2005). If this is

not possible, then management must refocus to

continuous or periodic control (Cromarty et al. 2002).

Unfortunately on ‘near-shore’ islands, neighbouring

source populations exist, and unaided reinvasion is

likely for some species; therefore, understanding the

level of connectivity among populations is vital to eval-

uate eradication programmes (Clout & Russell 2007;

Russell et al. 2009). Equally, understanding the levels of

invasion and harvesting success is important when

ongoing control is necessary, as this allows managers to

determine the likely ongoing residual pest densities and

evaluate their control strategies.

Population genetic analyses can identify population

structure patterns, which may in turn elucidate the

degree of population connectivity (Waples & GaggiottiCorrespondence: Andrew Veale, Fax: +6495291934;

E-mail: [email protected]

© 2013 John Wiley & Sons Ltd

Molecular Ecology (2013) doi: 10.1111/mec.12453

2006). These techniques can determine the origin of inva-

sive species populations, along with patterns of connec-

tivity within and between established populations

(Robertson & Gemmell 2004; Rollins et al. 2006; Schwartz

et al. 2007; Le Roux & Wieczorek 2009). There are three

broad categories of population genetic analyses used to

assess migration rates between patches. First, there are

indirect approaches, which use the partitioning of

genetic variation among populations to estimate migra-

tion levels (Slatkin & Barton 1989; Neigel 1997). Second,

a higher resolution set of methods are direct assignment

approaches, which select or exclude populations as pos-

sible sources for sampled individuals (Rannala & Moun-

tain 1997; Davies et al. 1999; Manel et al. 2005). For

invasive species during or after eradication attempts,

these methods have been successfully used to select

between hypotheses of ‘survival’ or ‘reinvasion’ for indi-

viduals detected post’eradication’ (e.g. Abdelkrim et al.

2005, 2007; Berry & Kirkwood 2010; Russell et al. 2010;

Veale et al. 2012a). Finally, there are kinship-based meth-

ods, which link individuals to a kin group rather than to

a population (e.g. Paetkau et al. 1995; Peery et al. 2008;

Broquet et al. 2009; Saenz-Agudelo et al. 2009; Palsbøll

et al. 2010). Kinship-based methods are well suited when

the degree of genetic divergence between populations is

low, such as for recently diverged populations, admixed

but subdivided populations, or in situations where cur-

rent or recent migration rates are high.

The stoat (Mustela erminea) is one of the principal

invasive species in New Zealand. They were introduced

to New Zealand in 1884 (Thomson 1922) to control

invasive rabbits (Oryctolagus cuniculus). Stoats subse-

quently have had well-documented catastrophic effects

on native New Zealand birds (King & Powell 2007).

They are one of the key nest predators in both alpine

grasslands and beech forests (Smith et al. 2008), and

they are a primary agent of decline for a large number

of threatened species of forest birds in New Zealand –

particularly in southern beech forest (Innes et al. 2010).

Due to the severe effects of stoat invasions – and the

associated extirpation of many endemic species on the

mainland, maintaining the stoat-free status of New Zea-

land’s offshore islands is extremely important (Parkes &

Murphy 2003; Bellingham et al. 2010). Unfortunately, of

all mammalian invasive species in New Zealand, stoats

pose one of the highest island invasion risks, having

invaded at least 90 islands (Veale et al. 2012b). Stoats

are competent swimmers (Veale 2013), and incursions

out to at least 3 km have been detected (Veale et al.

2012a). Stoats have been eradicated from numerous

islands around New Zealand, primarily through trap-

ping (Veale et al. 2012b). On these islands, trapping has

generally been maintained to prevent re-establishment,

although reincursions have occurred (King & Powell

2007). The reproductive biology of stoats makes them a

particularly insidious invasive species, in that all female

stoats become impregnated in spring (September –

October in New Zealand) including all infant female

stoats before they leave the nest (King & Moody 1982b).

Delayed implantation then means that litters are not

born until the following spring. This effectively means

that all female stoats are always pregnant, and any

female stoat that arrives on an island is capable of

founding a new population.

In this study, microsatellite, sex and age data were

analysed to assess the rates of immigration to, and the

rates of survival on Secretary Island during a stoat-con-

trol/eradication operation. A pilot study of the stoat

incursion rate to 19 small islands in Fiordland predicted

that the incursion rate to Secretary Island would be

below 1 stoat per year based on the minimum swim-

ming distance (950 m) to the island (Elliott et al. 2010),

and this was felt to be a manageable reincursion rate.

Aims

This study was designed to better quantify the rates of

immigration and survival of stoats on the island during

the operation to date. The specific questions asked are:

(i) How successful has the trapping programme been at

eliminating extant individuals? (ii) How successful has

the trapping programme been at eliminating invading

individuals? (iii) What is the immigration rate to the

island? (iv) Is in situ breeding occurring on the island

and (v) What information and level of resolution can

each genetic analysis method provide for answering

these questions?

Methods

Study area

Secretary Island (8140 ha) is situated at the entrance to

Doubtful Sound in Fiordland, New Zealand (Fig. 1.). It

is the second largest island on the Fiordland coast and

is an order of magnitude larger than any other island

where stoat eradication has been attempted using a

trapping network. It is unknown when stoats first

invaded Secretary Island, but it is likely to have been

around 1900, at the time they invaded nearby Resolu-

tion Island (Hill & Hill 1987).

Control programme

Full details of the methods for the trapping programme

on Secretary Island are described in the Operational

Plan (Golding et al. 2005) and summarized by McMurt-

rie et al. (2011). In brief, 945 tunnels, each containing 2


2 A. J . VEALE ET AL.

Mark IV FennTM kill traps (later replaced by DOC 150TM

traps) were situated along trap lines established across

the island: yielding an average tunnel density of 1 tun-

nel per 8.6 ha. A coastal trap line comprising 180 double

set DOC 150TM traps was also established on the adjacent

mainland along Thompson Sound and Pendulo Reach,

and trapping was maintained on neighbouring Bauza

Island (where stoats had been eradicated in 2003). Stoat

trapping began on Secretary Island in July 2005; thereaf-

ter, traps were left set and baited and have been ser-

viced three times annually. In the initial trapping

session in July 2005, 95 stoats were caught on Secretary

Island, and in all years subsequent to this there have

been small but significant numbers of stoats caught with

higher summer peaks in 2010 and 2011 (See Supplemen-

tary Material for a full breakdown of stoat captures).

Sample collection

The location, date of trap check, tunnel type, type of

bait used, weight and sex of every stoat trapped across

the operation (including those caught on the mainland

in the coastal traps along Thompson Sound and Pendu-

lo Reach) were recorded, and each individual carcass

was bagged separately and frozen for future analyses.

Age & sex determination

Sex was determined from the presence/absence of a

baculum bone, unless the specimen was badly

degraded; in which case sex was determined genetically

using PCR with the primers P1-5EZ and ZFRev1 fol-

lowed by digestion with BSM1 as described by Statham

et al. (2007).

The age of each stoat was estimated using the cemen-

tum annuli of the canine teeth. This method is the most

accurate age determination method for stoats (King

1991). The teeth were sectioned and stained at Matson’s

Laboratory (Milltown, MT, USA). Cementum annuli

were then scored visually at 1009 and 4009 magnifica-

tions, and these were compared with the known age

stoat teeth obtained by King (1991).

New Zealand

SecretaryIsland

Thompson Sound

Pendulo Reach

0 5 km

Secretary Island

Bauza Island

N

Doubtful Sound/Patea

Fig. 1 Location of Secretary Island within

Fiordland National Park, New Zealand.


STOAT INVASION POPULATION GENETICS 3

DNA extraction and amplification

DNA extraction was attempted from all individual

stoats caught on Secretary Island, and from all main-

land samples that were not badly degraded. Tissue

samples were dissected in the laboratory, where around

50 mg of tissue from the tail or ear was removed. For

samples that were particularly degraded, clumps of

hair, unidentified tissue or bone were dissected. Bone

samples were digested using the protocol of Loreille

et al. (2007) before DNA extraction. Genomic DNA was

extracted using the X-tractor GeneTM CAS-1820 (Corb-

ett Robotics, Brisbane, Australia) following digest or

using standard phenol/chloroform/isoamyl alcohol

(IAE) (protocol 2 in Hoelzel & Green 1992).

Individual genetic profiling was carried out by PCR

amplification of 16 microsatellite markers as described

by Veale et al. (2012a). For degraded samples, PCR

amplification was replicated using a stepwise protocol

as recommended by Paetkau (2003). A perfect match

between two PCR amplifications was required in order

to accept a genotype in order to eliminate PCR errors

resulting in either allelic dropout or false alleles. Any

samples that differed by one locus were checked for

scoring or amplification errors at all loci (Paetkau 2003).

If these differences could not be explained by errors in

scoring/typing, samples were then subjected to a

further round of PCR and scoring (Poole et al. 2001;

Mowat & Paetkau 2002). MICROCHECKER (van Oo-

sterhout et al. 2006) was used to detect possible techni-

cal artefacts, such as null alleles, large allele dropout

and accidental scoring of stutter bands.

Analysis

For both the mainland population and the Secretary

Island population for each capture year, the allele fre-

quencies, observed and expected heterozygosities and

mean number of alleles per locus were calculated as

summaries of genetic diversity within the populations

in ARLEQUIN 3.5.1.3 (Excoffier et al. 2005). Allelic rich-

ness and private allele richness were also calculated for

each capture year using the rarefaction method imple-

mented in HP-RARE 1.0 (Kalinowski 2005). Our assign-

ment methods rely on assumptions of Hardy–Weinberg

proportions and not on linkage disequilibrium, so these

were checked in each population, using Fisher’s exact

tests. FIS, FIT and FST were estimated between all combi-

nations of Mainland and Secretary Island during each

year of the control. These calculations were made in R

(version 2.10.1), based on the methods of Weir and

Cockerham (1984).

A number of approaches were used to determine the

likely origin of each individual and estimate the

migration rates. First, Structure, version 2.3.4 (Pritchard

et al. 2000), was used to infer individual cluster assign-

ment, using the correlated allele frequencies and admix-

ture models as recommended by Falush et al. (2003).

The probability of each admixture model was tested for

clusters (K) ranging from 1 to 5. For each value of K,

models were tested using 5 independent MCMC simu-

lations each consisting of a burn-in period of 105, fol-

lowed by 106 analysis cycles. This was conducted both

with and without the use of a location prior (Secretary

Island or mainland). We assessed the most likely num-

ber of homogeneous clusters using the second-order

rate of change L” (K) or ΔK following Evanno et al.

2005; implemented in STRUCTURE HARVESTER (Earl & von

Holdt 2012). Results were then combined with the pro-

gram CLUMPP to average individual clustering outputs

between runs (Jakobsson & Rosenberg 2007) were and

visualized in DISTRUCT (Rosenberg 2004).

For genetic population assignment, we used GENE-

CLASS2 (Piry et al. 2004), implementing the Bayesian

assignment criterion of Rannala & Mountain (1997),

which was recommended as the best set of assignment

criterion by Cornuet et al. (1999). We attempted to

assign each individual to either the initial knockdown

population (2005) or to the mainland population. For

later years, we also evaluated population assignment

using the previous year’s Secretary Island captures as

one potential population and the mainland as the other.

For population assignment, individuals were high-

lighted as potential migrants that either had a higher

posterior probability of originating in the ‘mainland’

population or that had a low posterior probability

(<1%) of belonging to the ‘island’ population. We pres-

ent the assignment results graphically following the

methodology of Russell et al. (2010), using an updated

method of accommodating individuals with partially

missing data by plotting them at the same posterior

quantiles in the full-locus chart, as they would appear

at in a reduced-locus chart. For each reference popula-

tion, the charts show the 1% and 99% quantiles of the

posterior distribution of log-genotype probabilities.

These lines display the Bayesian posterior distribution

for the ‘fit’ to a population for an individual drawn

from this population, and should be regarded as a

simultaneous reflection of both the natural spread of

the population, and of our sampling uncertainty. We

refer to these plots as GeneCharts hereafter.

Relatedness was assessed using pairwise likelihood

ratios calculated in KINGROUP 2.0, a program that imple-

ments the methods developed by Queller & Goodnight

(1989) to calculate the genetic similarity between indi-

viduals (Konovalov et al. 2004). In KINGROUP, we

used the descending ratio (DR) method to partition

individuals into kin groups that had the highest overall



likelihood. In addition, we attempted to identify full-

sibling and parent–offspring relationships. Given the

very large number of comparisons, only likelihood

ratios with a probability of <0.01 were accepted as sig-

nificant. These analyses were combined with the age

data to assist in determining parent–offspring relation-

ships and to ascertain whether stoats assigned as

siblings were plausibly littermates.

Using the combined age and sex data, clustering,

population assignment and kinship outputs, the mini-

mum total number of individuals alive at any given

time on the island was estimated, along with the num-

ber of litters produced on the island, and the number of

immigrants that arrived each year. For defining

migrants, results were compared between Structure,

GENECHART and KINGROUP for each individual. The GENE-

CHART was generally given precedence over Structure

due to better resolution particularly in later years.

Potential migrants highlighted by Structure or in GENE-

CHART that had kin members present on the island

either that year or the previous year were defined as

‘residents’. The total minimum population on the island

during each year was estimated by adding the total

number of stoats caught during the year, with the total

survivors that were later caught, and the total number

of females and males required for the subsequent years’

litters to be produced that also were not caught. This

was estimated by examining the relatedness estimates

derived from KINGROUP, to determine if the father and

mother of each litter had been caught. The maximum

potential capture success attained during each year was

then estimated by dividing the estimated minimum

total stoat population with the total captures each year.

Results

From a total of 247 stoats caught on Secretary Island in

the period July 2005 – December 2011, 205 (83%) stoats

were genotyped for at least 13 loci and were used in

the analysis. Of these, 187 were genotyped at all 16 loci.

From the 77 stoats caught on the mainland, 40 were

genotyped. The age of 88% (218/247) of the stoats was

successfully estimated (those that could not either had

broken or degraded tooth samples, or teeth were not

collected).

The mainland stoat population was more diverse

than the Secretary Island population with respect to

allelic richness, the private allelic richness, the mean

number of alleles per locus and the average heterozy-

gosity (Table 1). In general, these measures of genetic

diversity remained relatively stable on Secretary Island

throughout the trapping period, other than in 2007

when there was a brief increase in average genetic

diversity.

We observed a small but statistically significant

differentiation between the initial Secretary Island pop-

ulation and the mainland (FST = 0.0326***). This differ-

entiation between the mainland and Secretary island

populations generally increased each year, other than in

2007 where there was no significant differentiation

(Table 2). The inbreeding coefficient (FIS) on Secretary

Island trended upwards over time (see Supporting

information). In general (other than for 2007), genetic

differentiation between adjacent capture years on Secre-

tary Island was low and nonsignificant, and increas-

ingly large and significant between more distant years.

All of these indices suggest that the Secretary Island

population is primarily replicated by recruits from the

island’s previous year’s population, with increasing dif-

ferentiation from the mainland population due to

genetic drift, and increasing inbreeding due to the low

numbers of progenitors each year.

Using the method of Evanno et al. (2005), the most

appropriate number of genetic clusters obtained from

the Structure analyses was two, and these corresponded

relatively well with a ‘mainland’ cluster and an ‘island’

cluster (Fig. 2), with stoats caught on Secretary Island

from 2005 to 2006 generally clustering separately from

the mainland stoats. Most stoats caught on Secretary

Island during 2007 appeared to be from the ‘mainland’

cluster, while most stoats caught in 2008–2009 appeared

to be from the ‘island’ cluster. In later years, a mix of

‘mainland’ and ‘admixture’ stoats is evident.

Those individuals caught on Secretary Island from

2005 to 2006 that cluster more closely to the ‘mainland’

cluster could either be recent immigrants to Secretary

Island from the mainland or there could be insufficient

resolution in the data to assign them accurately to

either cluster. This latter hypothesis was supported by

comparing cluster assignment with and without the use

of a location prior in the analysis. The location prior for

these individuals was informative, in that it led to

a high degree of assignment probability for either

Table 1 Genetic properties of the sampled populations

Private

alleles

Allelic

richness

Private

allele

richness

Mean

heterozygosity/locus

Mainland 11 3.88 0.41 0.474

SI-2005 0 3.34 0.03 0.493

SI-2006 0 3.3 0.09 0.508

SI-2007 3 4.01 0.32 0.567

SI-2008 0 2.89 0.04 0.507

SI-2009 1 3.18 0.09 0.5

SI-2010 0 3.32 0.02 0.502

SI-2011 0 3.13 0.01 0.456



population, depending on which location was given for

each individual. Because the inclusion of a location

prior in the analysis was informative (thus creating

bias), only the ‘na€ıve’ clustering is shown, as this is

more appropriate for clustering the later individuals

caught whose origin is unknown, and it is a fairer

reflection of the uncertainty inherent in the data.

Using GeneCharts, it is possible to ascertain finer

details of the population structure across years (Fig. 3).

The initial Secretary Island population (captured 2005)

is clearly a subset of the genetic diversity of the main-

land – all Secretary Island stoats fall within the poster-

ior probability range of the mainland stoat population,

but not vice versa. Almost all initial captures were cor-

rectly assigned to their location of capture, indicating

adequate resolution in the data to determine resident

from migrant for future years. For the stoats caught in

2006, all but one assign to the initial island population

(Fig. 3a), suggesting they are descended from this pop-

ulation. In 2007 (the year after a major beech seed fall

on the mainland), the majority of stoats caught on Sec-

retary Island were either assigned to the mainland pop-

ulation, or have a low level of assignment to the

Secretary Island population – suggesting they are immi-

grants. Stoats captured on Secretary Island during both

2008 and 2009 largely assign to the island population

with a few individuals in each year (2 in each year) of

potential mainland origin using this method (Fig. 3b).

Stoats captured on Secretary Island during 2010 – while

assigning primarily to the original Secretary Island pop-

ulation – are becoming less distinguishable from main-

land stoats than the initial captures (Fig. 3c). All but 2

stoats caught in 2011 assign to the 2010 Secretary Island

population (Fig. 3d).

Using KINGROUP, we detected a number of litters and

parent–offspring relationships. In 2006, 10 l (born

October 2005) were distinguished, and these captures

showed spatial clustering (Fig. 4). All but one mother

was identified among the captures, with the one

remaining litter spatially clustered near the only adult

female caught that could not be genotyped. In 2007,

there were no litters detected indicating no breeding

during 2006. In 2008, there were two litters (born

October 2007). The father of one of these litters was

also caught and was also shown to be the father of

one of the litters born later in the year (caught 2009).

In 2009, three litters were detected. One of the moth-

ers that gave birth to a litter in 2007 (caught 2008)

also had a litter 2008 (caught 2009) and was subse-

quently trapped. In 2010, only 4/22 stoats genotyped

did not have a significant (P < 0.01) sibling-relation-

ship with another stoat caught on the island in that

year. In 2011, 10/15 stoats genotyped all clustered

together into a highly related group with a high

probability of sibling relationships between most indi-

viduals.

Combining the results from the multiple analyses, we

estimated the make-up and size of the population over

time (Table 3). The estimated proportion of the total

population of stoats caught each year has been high

throughout, although temporal extirpation has not been

achieved during the operation. At least 10% of the pop-

ulation present on the island was not trapped each year

during the first 3 years, and at least 24% of the more

Table 2 Pairwise FST between the Mainland stoat population and among each capture year on Secretary Island

Mainland SI-2005 SI-2006 SI-2007 SI-2008 SI-2009 SI-2010

SI-2005 0.0326***SI-2006 0.0414*** 0.006

SI-2007 0.015 0.0264*** 0.0204*SI-2008 0.0536*** 0.0188* 0.0317** 0.0400***SI-2009 0.0612*** 0.0483*** 0.0555*** 0.0245* -0.0003

SI-2010 0.0435*** 0.0436*** 0.0511*** 0.0094 0.0398*** 0.0119

SI-2011 0.0730*** 0.0751*** 0.0950*** 0.0525*** 0.0908*** 0.0391* 0.017

*p < 0.05; **p < 0.01; ***p < 0.001.

2005

2009

2010

2011

2008

2007

2006

Mainland Secretary IslandFig. 2 Graphical output from Structure

(Pritchard et al. 2000). Each vertical

estimated degree of cluster membership

for each individual (K = 2).



recent population has avoided trapping each year. The

probable number of immigrants has been low, other

than in 2007 when around eight stoats arrived on Secre-

tary Island from the mainland.

Discussion

By using a range of analyses on the data set gathered, it

has been possible to describe in fine detail the level of

success of the stoat trapping programme and the rates

of stoat invasion to Secretary Island. Because the age,

sex and genotype were determined for the majority of

stoats caught over the 6-year period, it is even possible

to provide details of the life history of many of the

stoats. For instance, the male stoat caught in trap 370

on Secretary Island in June 2008 was a 4-year-old stoat

born on Secretary Island in 2003. He avoided trapping

for 3 years and fathered two litters – one in October

2007 and one in October 2008. Descendants of his were

still present in 2011.

The Secretary Island stoat population has been

reduced significantly, and our study shows that both

(A) (B)

(C) (D)

–25 –20 –15 –10

–25

–20

–15

–10

Log10 genotype probability for population Mainland

Log1

0 ge

noty

pe p

roba

bilit

yfo

r pop

ulat

ion

SI-2

005

1% 99%

1%

99%

MainlandSI-2005SI-2006SI-2007

–22 –20 –18 –16 –14 –12 –10

–22

–20

–18

–16

–14

–12

–10

Log10 genotype probability for population MainlandLo

g10

geno

type

pro

babi

lity

for p

opul

atio

n S

I-200

5

1% 99%

1%

99%


–25 –20 –15 –10

–25

–20

–15

–10


Log1

0 ge

noty

pe p

roba

bilit

yfo

r pop

ulat

ion

SI-2

005

1% 99%

1%

99%


–20 –15 –10

–20

–15

–10


Log1

0 ge

noty

pe p

roba

bilit

yfo

r pop

ulat

ion

SI-2

010

1% 99%

1%

99%

MainlandSI-2010SI-2011

GeneChart population AssignmentMainland vs. Secretary Island 2005

for Secretary Island 2006 & 2007

GeneChart population assignmentMainland vs. Secretary Island 2005





for Secretary Island 2011

Fig. 3 GENECHARTS showing pairwise assignment of individuals between populations.



immigration from the mainland and in situ survival and

breeding have occurred. The initial trapping in 2005

killed all but approximately 10 females and 2 males,

and one of these two males was subsequently killed

November 2005 – February 2006. When the 10 litters

were born in October 2005, any females impregnated by

these two males were subsequently trapped before they

gave birth; therefore, there were no litters born in 2006.

There were survivors both male and female from the

litters born in 2005 that avoided trapping and bred in

2006 leading to litters in October 2007. There continued

to be low numbers of stoats surviving each subsequent

year, and primarily through this breeding, the popula-

tion has been maintained.

While having stoat captures of up to 31 stoats in

recent years may appear serious, these numbers do

need to be put in perspective. There were 75 female

stoats on Secretary Island in the winter of 2005 all of

which will have been pregnant (King and Moody 1982).

The average litter size for stoats is around 8 (King &

Powell 2007), though on Secretary Island in 2005,

approximately four kits survived to independence per

mother on average. There would therefore be a summer

peak population of around 416. The current level of

trapping therefore maintains the peak stoat population

at around 7% of the likely peak stoat numbers on the

island without trapping. Most of these stoats are

rapidly trapped rather than dying of natural mortality

related to resource limitation due to prey population

decreases. No unfenced area on the New Zealand main-

land can maintain stoat densities at these low levels,

and in all environmentally comparable mainland sites,

there is a suite of other invasive predatory mammals

present.

The fact there have been stoats that have survived

each trapping round was not unexpected, because stoat

sign including footprints, positive indication by preda-

tor detection dogs and footage of stoats on trail cameras

1

11

1

1

1

1

1

2

2

33

3

345

5

66

5

7

7

8

9

9

Mother

Ungenotyped

Immigrant

Trap

Legend

Fig. 4 Map of Secretary Island showing

stoat captures during 2006. Colours and

number represent each litter born Octo-

ber 2005. Several traps caught more than

one stoat, and these were either litter-

mates or were ungenotyped.



used for the red deer eradication operation has been

seen on the island in areas where stoats were not

caught in the given trapping period. These stoats may

have avoided traps (King et al. 2009), or some may have

not actually encountered traps. King (1980) showed that

trapping success for female stoats decreased when trap

spacing increased above 400 m – with some female

stoats occupying ranges between trap lines. Trap line

spacing on Secretary Island is significantly lower than

this due to the large size of the island, the difficult

topography and financial constraints. In 2008, a number

of stoats were caught that were descended from the ori-

ginal population, indicating that some stoats born in

2005 had avoided trapping throughout 2006 and 2007

and had litters in 2007 – with these offspring becoming

detectable in 2008 and 2009. A similar situation is

known to have occurred on Maud Island, where female

stoats invaded the island, avoided capture and were

only detectable through their litters (Crouchley 1994).

Some level of reinvasion was always predicted and

factored into the planning for this operation (Golding

et al. 2005). Invasion rates detected in this study were

relatively low (0–2) for most years with the notable

exception of 2007 when approximately eight stoats

arrived on Secretary Island. During the summer of

2006–2007, there was a plague of stoats on the mainland

(DOC Southland unpublished data). This stoat plague

was related to the mast event (heavy seed fall) pro-

duced by southern beech (Nothofagus spp.) forests,

which occurs every few years. Beech forest masting

leads to irruptions of rodents (Mus musculus and Rattus

spp.), which subsequently produces sudden massive

increases in stoat populations (King 1983, 2002). There

was a second lesser stoat plague in 2010 (DOC South-

land unpublished data). For this year, it is more diffi-

cult to assess how many stoats caught on Secretary

Island are immigrants, as by this stage, the litters being

born on the island appear to be largely admixed (main-

land/island). Four of the stoats caught on Secretary

Island during this year had no statistically significant

sibling relationships and could potentially have origi-

nated from the mainland. If this proportion of immi-

grants reflects the population, and then approximately

six stoats swam from the mainland this year.

The fact that we detected private alleles on Secretary

Island in 2007 and 2009, which were not detected on

the mainland coastal trapping nor the initial captures

on Secretary Island, is perhaps surprising. We verified

these alleles with multiple amplifications, and they

appear to be real. All these alleles have been found else-

where in Fiordland, about 40 km south on the mainland

opposite Resolution Island (A. J. Veale unpublished

data). It is probable that they therefore came from the

mainland and remained undetected in the mainland

coastal sampling used in this study due to their low

frequency.

In some situations, it could be useful to combine

mitochondrial data with the microsatellite data used in

this study, particularly to evaluate the number of

females contributing to a population. This analysis was

not performed because the mitochondrial diversity of

stoats in New Zealand is relatively low, with only two

haplotypes recorded in Fiordland (Holland 2004). This

diversity is too low for the expected benefits to justify

the cost of sequencing all of the individuals caught.

Genetic methods

Multiple methods, including population level statistics,

clustering, population assignment and kin-group

assignment were used to describe connectivity in this

study, and each method provided different and comple-

mentary information. Structure and GENECHARTS pro-

vided consistent answers, though with differing

resolution levels. These two methods answer slightly

different questions, with Structure looking for cryptic

genetic structure not necessarily associated with sam-

pling locations, while GENECHARTS assesses the fit of

individuals to specified populations. Structure was

unable to provide high-resolution assignment for some

individuals due to the low differentiation between pop-

ulations, with genetic drift and ongoing gene flow mak-

ing this particularly difficult for individuals caught in

the later periods of trapping. Overall, GENECHARTS

proved superior at differentiating these populations –

particularly for later years given the potential for com-

paring the genetic assignment of adjacent years. This

method also provides useful information, such as the

degree of separation between populations and variation

within them. This graphical method of assignment has

Table 3 Summary of capture success, number of immigrants

to the island and numbers of litters born on the island during

the operation to 2011

Total

caught

Min. Total

population

Max.%

Caught

No.

immigrants Litters

SI-2005 104 116* 90 0 10

SI-2006 50 56 89 0–1 0

SI-2007 12 19 90 7–8 2

SI-2008 10 17 76 0–2 3

SI-2009 13 25 76 0–3 ~5–6SI-2010 31 34 91 0–6 ~2SI-2011 27 ? ? 2 ?

*Total minimum population in 2005 is the population at the

start of trapping in July (winter), not during the summer peak

in December – January.

? = unknown.



significant benefits over using the tabular outputs from

Gene-Class2, in that by looking at the quantiles of the

posterior distribution for each population, individuals

that do not assign well to either population are high-

lighted and we can see the degree of separation of the

populations. In this study, individuals that did not

assign well to either population are likely to be from

the mainland, because the sampling on the island was

very thorough, and the sampling on the larger and

genetically more diverse mainland was less extensive.

There are no likely modes by which stoats could arrive

on Secretary Island other than by swimming from the

adjacent mainland (Veale et al. 2012b), so having stoats

immigrate from an unsampled (distant) population is

unlikely; it is more likely that they originate from an

under-sampled population (the mainland).

Using 16 loci in a relatively isolated population to

determine precise familial relationships between indi-

viduals is difficult; however, this was aided by the fact

that the precise age and sex of most stoats caught was

known, and the breeding cycle of stoats is discrete and

annual. Looking at patterns of relatedness in this situa-

tion supplements the other techniques and allows us to

estimate the number of litters born each year. Given

that some immigration is occurring, the distinct genetic

signal of the original island population is rapidly mud-

died by ongoing geneflow. Using pairwise individual

relatedness added greatly to the resolution of this

otherwise indistinguishable data and should be used

more often when attempting to describe the origin of

individuals.

Understandably, the genetic differentiation between

this population and the mainland was low, with an FST

of 0.03, but despite this genetic similarity, recent immi-

grants were detectable. While attempting to directly

detect recent immigrants between natural populations

using genetic methods is regularly made, for example:

(Cegelski et al. 2003; Berry et al. 2004; Lancaster et al.

2011; Schregel et al. 2012), these populations usually

have more ancient heritage and accordingly, more sig-

nificant population structure. It is useful to know that

genetic assignment can be achieved for recently

diverged populations with ongoing connectivity. This

study therefore underscores the power and usefulness

of these techniques to assess animal movement, even in

more challenging scenarios.

Management implications

The management implications of population genetic

studies of invasive species on islands are usually that

either of the eradication procedures need to be

improved, such as on Poirier Island in the St Anne

Archipelago where a ship rat (Rattus rattus) eradication

was shown to have failed (Abdelkrim et al. 2007), or

they indicate that the reinvasion ability of the target

species has been underestimated; therefore, increased

vigilance is required to maintain the pest-free status of

the island, such as the reinvasion of ship rats on Pearl

Island (Russell et al. 2010).

Our results have helped to inform management of

this island in several more specific ways resulting in a

number of adjustments to the programme. First, extra

traps have been installed on Secretary Island, with traps

placed in the largest gaps between the original trap

lines. This was undertaken partially as a result of the

genetic analyses indicating that some individuals had

avoided trapping for considerable periods, and the

result that the population is primarily replicated from

in situ breeding. These new traps have already caught

several stoats. Without the information obtained from

our study, it may have been assumed that recent cap-

tures originated more from the mainland, and therefore,

extra effort could have been focused on increasing

mainland trapping.

Second, the locations of litters were identified in this

study using kinship analyses; therefore, extra trapping

and search efforts were concentrated in these areas.

Detecting mustelids at low densities is difficult (Cho-

quenot et al. 2001; King et al. 2009); consequently, any

information that can help focus the limited resources

available is vital to the success of the trapping opera-

tion. The Department of Conservation has attempted

multiple techniques (detection dogs, lures, various trap

types) striving to catch these trap-avoiding animals.

Finally, the fact that reinvasion rates calculated in this

study appear to show a link with beech mast events on

the mainland has led to the suggestion of pulsed stoat

and rodent control on the opposing mainland to

decrease these reinvasion spikes. This increase in rein-

vasion during beech mast events was not evident in the

trapping data alone, as variation between years in the

number of litters born on Secretary Island dispropor-

tionately affects the overall population size due to the

large potential size of litters.

Conclusion

While using genetic techniques to describe biological

invasions is becoming common practice (Rollins et al.

2006; Schwartz et al. 2007; Le Roux & Wieczorek 2009),

many of these studies lack focused management-ori-

ented objectives (Fitzpatrick et al. 2012). In our study,

the origin of almost all stoat individuals caught over a

6-year control period on Secretary Island was assessed

through genetic population assignment and kinship

analysis, and through this, it was possible to provide

fine-scale details of the sources of recruitment for this



island and the level of success of the control pro-

gramme. This information has enabled the stoat-control

programme to focus their control efforts in the most

cost-effective manner in order to minimize residual pest

densities.

Acknowledgements

We wish to thank the editor and three unnamed reviewers for

their constructive comments on the manuscript. We also thank

Dr. Robyn Howitt, Julia Allwood, Diana Prada, Dr. Craig Mil-

lar, Selina Patel, Dr. Shane Lavery and Annette Evans for their

assistance and time in the laboratory, and the staff at Matson’s

laboratory for their expertise in helping age the stoat teeth.

Thanks to Dr. Elaine Murphy and members of the Island Erad-

ication Advisory Group (DOC) for their advice. We thank New

Zealand’s Ministry of Business, Innovation and Employment

(Science and Innovation Group) for funding to the Invasive

Mammal Impacts Program (contract C09X0909) and core fund-

ing to Crown Research Institutes. Most importantly, we thank

the staff and volunteers who undertook the trapping (in the

rain) on Secretary Island and collected and carried back every

stoat carcass for analysis.

References

Abdelkrim J, Pascal M, Calmet C, Samadi S (2005) Importance

of assessing population genetic structure before eradication

of invasive species: examples from insular Norway rat popu-

lations. Conservation Biology, 19, 1509–1518.Abdelkrim J, Pascal M, Samadi S (2007) Establishing causes of

eradication failure based on genetics: case study of ship rat

eradication in Ste. Anne archipelago. Conservation Biology, 21,

719–730.Bellingham PJ, Towns DR, Cameron EK et al. (2010) New

Zealand island restoration: seabirds, predators, and the

importance of history. New Zealand Journal of Ecology, 34,

115–136.Berry O, Kirkwood R (2010) Measuring recruitment in an inva-

sive species to determine eradication potential. Journal of

Wildlife Management, 74, 1661–1670.

Berry O, Tocher MD, Sarre SD (2004) Can assignment tests

measure dispersal? Molecular Ecology, 13, 551–561.

Bomford M, O’Brien P (1995) Eradication or control for verte-

brate pests. Wildlife Society Bulletin, 23, 249–255.

Broome K, Cromarty P, Cox A (2005) Rat eradications – How

to get it right without a recipe.

Broquet T, Yearsley J, Hirzel AH, Goudet J, Perrin N (2009)

Inferring recent migration rates from individual genotypes.

Molecular Ecology, 18, 1048–1060.Cegelski CC, Waits LP, Anderson NJ (2003) Assessing popula-

tion structure and gene flow in Montana wolverines (Gulo

gulo) using assignment-based approaches. Molecular Ecology,

12, 2907–2918.Choquenot D, Ruscoe WC, Murphy E (2001) Colonisation of

new areas by stoats: time to establishment and require-

ments for detection. New Zealand Journal of Ecology, 25,

83–88.

Clout MN, Russell JC (2007) The invasion ecology of mammals:

a global perspective. Wildlife Research, 35, 180–184.

Cornuet J-M, Piry S, Luikart G, Estoup A, Solignac M (1999)

New methods employing multilocus genotypes to select or

exclude populations as origins of individuals. Genetics, 153,

1989–2000.Cromarty PL, Broome KG, Cox A et al. (2002) Eradication plan-

ning for invasive alien animal species on islands – the

approach developed by the New Zealand Department of

Conservation. In: Turning the tide: the eradication of invasive

species (eds Veitch CR, Clout MN), pp. 85–91. IUCN, Gland,

Switzerland.

Crouchley D (1994) Stoat control on Maud Island 1982–1993.

Ecological Management, 2, 39–45.Davies N, Villablanca FX, Roderick GK (1999) Determining the

source of individuals: multilocus genotyping in nonequilibri-

um population genetics. Trends in Ecology & Evolution, 14,

17–21.Earl DA, von Holdt BM (2012) STRUCTURE HARVESTER: a

website and program for visualizing STRUCTURE output

and implementing the Evanno method. Conservation Genetics

Resources, 4, 359–361.Elliott G, Willans M, Edmonds H, Crouchley D (2010) Stoat

invasion, eradication and re-invasion of islands in Fiordland.

New Zealand Journal of Zoology, 37, 1–12.

Evanno G, Regnaut S, Goudet J (2005) 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 ver 3.0: an

integrated software package for population genetic analysis.

Evolutionary Bioinformatics Online, 1, 47–50.Falush D, Stephens M, Pritchard JK (2003) Inference of popula-

tion structure using multilocus genotype data: linked loci

and correlated allele frequencies. Genetics, 164, 1567–1587.

Fitzpatrick BM, Fordyce JA, Niemiller ML, Reynolds RG (2012)

What can DNA tell us about biological invasions? Biological

Invasions, 14, 245–253.Golding C, McMurtrie P, Edge K-A, Willans M (2005) Secretary

Island Operational Plan Part A – Stoat Eradication. New Zea-

land Department of Conservation, Te Anau.

Hill S, Hill J (1987) Richard Henry of Resolution Island. John

McIndoe Ltd, Dunedin, New Zealand.

Hoelzel AR, Green A (1992) Analysis of population level varia-

tion by sequencing PCR-amplified DNA. In: Molecular genetic

analysis of populations: a practical approach (ed. Hoelzel AR),

pp. 159–187. Oxford University Press, Oxford.

Howald G, Donlan CJ, Galvan JP et al. (2007) Invasive rodent

eradication on islands. Conservation Biology, 21, 1258–1268.

Innes J, Kelly D, Overton JM, Gillies C (2010) Predation and

other factors currently limiting New Zealand forest birds.

New Zealand Journal of Ecology, 34, 86–114.Jakobsson M, Rosenberg NA (2007) CLUMPP: a cluster match-

ing and permutation program for dealing with label switch-

ing and multimodality in analysis of population structure.

Bioinformatics, 23, 1801–1806.Kalinowski ST (2005) HP-Rare: a computer program for per-

forming rarefaction on measures of allelic diversity. Molecu-

lar Ecology Notes, 5, 187–189.

King CM (1980) Field experiments on the trapping of stoats

(Mustela erminea). New Zealand Journal of Zoology, 7, 261–

266.

King CM (1983) The relationships between beech (Nothofagus

sp.) seedfall and populations of mice (Mus musculus), and



the demographic and dietary responses of stoats (Mustela

erminea), in three New Zealand forests. Journal of Animal Ecol-

ogy, 52, 141–166.

King CM (1991) A review of age-determination methods for

the stoat Mustela erminea. Mammal Review, 21, 31–49.

King CM (2002) Cohort variation in the life-history parameters

of stoats Mustela erminea in relation to fluctuating food

resources: a challenge to boreal ecologists. Acta Theriologica,

47, 225–244.

King CM, Moody JE (1982) The biology of the stoat (Mustela

erminea) in the national parks of New Zealand. 4. Reproduc-

tion. New Zealand Journal of Zoology, 9, 103–118.King CM, Powell RA (2007) The natural history of weasels and

stoats: ecology, behaviour and management. Oxford University

Press, New York.

King CM, McDonald RM, Martin RD, Dennis T (2009) Why is

eradication of invasive mustelids so difficult? Biological Con-

servation, 142, 806–816.Konovalov DA, Manning C, Henshaw MT (2004) KINGROUP:

a program for pedigree relationship reconstruction and kin

group assignments using genetic markers. Molecular Ecology

Notes, 4, 779–782.Lancaster ML, Taylor AC, Cooper SJB, Carthew SM (2011) Lim-

ited ecological connectivity of an arboreal marsupial across a

forest/plantation landscape despite apparent resilience to

fragmentation. Molecular Ecology, 20, 2258–2271.Le Roux J, Wieczorek AM (2009) Molecular systematics and

population genetics of biological invasions: towards a better

understanding of invasive species management. Annals of

Applied Biology, 154, 1–17.

Loreille OM, Diegoli TM, Irwin JA, Coble MD, Parsons TJ

(2007) High efficiency DNA extraction from bone by total

demineralization. Forensic Science International-Genetics, 1,

191–195.

Manel S, Gaggiotti OE, Waples RS (2005) Assignment methods:

matching biological questions with appropriate techniques.

Trends in Ecology & Evolution, 20, 136–142.McMurtrie P, Edge K-A, Crouchley D et al. (2011) Eradication

of stoats (Mustela erminea) from Secretary Island, New

Zealand. In: Island invasives: eradication and management (eds

Veitch CR, Clout MN & Towns DR), pp. 455–460. IUCN,

Gland, Switzerland.

Mowat G, Paetkau D (2002) Estimating marten Martes ameri-

cana population size using hair capture and genetic tagging.

Wildlife Biology, 8, 201–209.Myers JH, Simberloff D, Kuris AM, Carey JR (2000) Eradication

revisited: dealing with exotic species. Trends in Ecology &

Evolution, 15, 316–320.

Neigel JE (1997) A comparison of alternative strategies for esti-

mating gene flow from genetic markers. Annual Review of

Ecology and Systematics, 28, 105–128.van Oosterhout C, Weetman D, Hutchinson WF (2006)

Estimation and adjustment of microsatellite null alleles in

nonequilibrium populations. Molecular Ecology Notes, 6,

255–256.Paetkau D (2003) An empirical exploration of data quality in

DNA-based population inventories. Molecular Ecology, 12,

1375–1387.

Paetkau D, Calvert W, Stirling I, Strobeck C (1995) Microsatel-

lite analysis of population structure in canadian polar bears.

Molecular Ecology 4, 347–354.

Palsbøll PJ, Peery MZ, Berube M (2010) Detecting populations

in the ‘ambiguous’ zone: kinship-based estimation of popula-

tion structure at low genetic divergence. Molecular Ecology

Resources 10, 797–805.Parkes J, Murphy E (2003) Management of introduced mam-

mals in New Zealand. New Zealand Journal of Zoology, 30,

335–359.

Peery MZ, Beissinger SR, House RF et al. (2008) Characterizing

source-sink dynamics with genetic parentage assignments.

Ecology 89, 2746–2759.Piry S, Alapetite A, Cornuet JM et al. (2004) GENECLASS2: a

software for genetic assignment and first-generation migrant

detection. Journal of Heredity, 95, 536–539.

Poole KG, Mowat G, Fear DA (2001) DNA-based population

estimate for grizzly bears Ursus arctos in northeastern British

Columbia, Canada. Wildlife Biology, 7, 105–115.Pritchard JK, Stephens M, Donnelly P (2000) Inference of popu-

lation structure using multilocus genotype data. Genetics,

155, 945–959.

Queller DC, Goodnight KF (1989) Estimating relatedness using

genetic markers. Evolution, 43, 258–275.

Rannala B, Mountain J (1997) Detecting immigration by using

multilocus genotypes. Proceedings of the National Academy of

Sciences of the United States of America, 94, 9197–9201.Robertson BC, Gemmell NJ (2004) Defining eradication units to

control pests. Journal of Applied Ecology, 41, 1042–1048.Rollins LA, Woolnough AP, Sherwin WB (2006) Population

genetic tools for pest management: a review. Wildlife

Research, 33, 251–261.Rosenberg NA (2004) DISTRUCT: a program for the graphical

display of population structure. Molecular Ecology Notes, 4,

137–138.

Russell JC, Mackay JWB, Abdelkrim J (2009) Insular pest con-

trol within a metapopulation context. Biological Conservation,

142, 1404–1410.Russell JC, Miller SD, Harper GA et al. (2010) Survivors or

reinvaders? Using genetic assignment to identify invasive

pests following eradication. Biological Invasions, 12, 1747–

1757.

Saenz-Agudelo P, Jones GP, Thorrold SR, Planes S (2009) Esti-

mating connectivity in marine populations: an empirical eval-

uation of assignment tests and parentage analysis under

different gene flow scenarios.Molecular Ecology, 18, 1765–1776.Schregel J, Kopatz A, Hagen SB et al. (2012) Limited gene flow

among brown bear populations in far Northern Europe?

Genetic analysis of the east-west border population in the

Pasvik Valley. Molecular Ecology, 21, 3474–3488.Schwartz MK, Luikart G, Waples RS (2007) Genetic monitoring

as a promising tool for conservation and management.

Trends in Ecology & Evolution, 22, 25–33.

Simberloff D (2003) Eradication – preventing invasions at the

outset. Weed Science, 51, 247–253.

Slatkin M, Barton NH (1989) A comparison of three indirect

methods for estimating average levels of gene flow. Evolu-

tion, 43, 1349–1368.Smith DHV, Wilson DJ, Moller H, Murphy EC (2008) Using

artificial nests to explore predation by introduced predators

inhabiting alpine areas in New Zealand. New Zealand Journal

of Zoology, 35, 119–128.Statham MJ, Turner PD, O’Reilly C (2007) Molecular sex identifi-

cation of five mustelid species. Zoological Studies, 46, 600–608.



Thomson GM (1922) The Naturalisation of Animals and Plants in

New Zealand. Cambridge University Press, Cambridge.

Towns DR, Broome KG (2003) From small Maria to massive

Campbell: forty years of rat eradications from New Zealand

islands. New Zealand Journal of Zoology, 30, 377–398.

Veale AJ (2013) Observations of stoats (Mustela erminea) swim-

ming. New Zealand Journal of Zoology, 40, 166–169.

Veale AJ, Clout MN, Gleeson DM (2012a) Genetic population

assignment reveals a long-distance incursion to an island by a

stoat (Mustela erminea). Biological Invasions, 14, 735–742.Veale AJ, Hannaford OD, Russell JC, Clout MN (2012b) Model-

ling the distribution of stoats on New Zealand offshore

islands. New Zealand Journal of Ecology, 36, 38–47.

Waples RS, Gaggiotti O (2006) What is a population? An

empirical evaluation of some genetic methods for identifying

the number of gene pools and their degree of connectivity.

Molecular Ecology, 15, 1419–1439.

Weir BS, Cockerham CC (1984) Estimating F-statistics for the

analysis of population structure. Evolution, 38, 1358–1370.

A.J.V. Designed, performed, analysed and wrote the arti-

cle. K.A.E. Designed, performed and edited article. P.M.

Designed, performed and edited article. R.F. Analysed

and edited article. M.N.C. Designed and edited article.

D.M.G. Designed and edited article.

Data accessibility

Sample data (age, sex, location, time of capture), micro-

satellite data and GeneChart R script available from

Dryad, doi:10.5061/dryad.r634p.

Supporting information

Additional supporting information may be found in the online ver-

sion of this article.

Fig. S1 Stoat capture tally on Secretary Island from 2005 until

December 2011.

Table S1 Inbreeding Coefficient (FIS) between the Mainland

stoat population and each capture year on Secretary Island.



Using genetic techniques to quantify reinvasion, survival and in-situ breeding rates during control/eradication operations.

Documents