Conserv Genet
DOI 10.1007/s10592-013-0526-3

RESEARCH ARTICLE

Inbreeding and strong population subdivision in an endangered
salamander
Andrew Storfer • Stephen G. Mech •
Matthew W. Reudink • Kristen Lew

Received: 1 May 2013 / Accepted: 22 August 2013
Ó Springer Science+Business Media Dordrecht 2013

Abstract Studies of genetic population structure and
genetic diversity are often critical components of endangered species conservation and management plans. Genetic
studies are thus particularly important for amphibians,
which are in global decline. We studied genetic variation
and population structure among 276 individuals from
approximately half of the known localities of the endangered Sonora tiger salamander, Ambystoma mavortium
stebbinsi, using ten microsatellite loci. Allelic diversity
was generally low (2.7 alleles per locus per population) and
overall observed heterozygosity (0.191) was significantly
lower than expected (0.332). Most populations showed
significant departures from Hardy–Weinberg equilibrium,
which are likely due to inbreeding. In addition, evidence of
recent bottlenecks was suggested by shifted allele frequency distributions in 5 of 16 populations, and ratios of
allele number to allele size range (M) values lower than
critical values in all populations. A high degree of genetic
subdivision (h = 0.133) was found over all populations,
and nearly all pairwise population combinations were
genetically subdivided. Thus, gene flow is limited even

A. Storfer (&)  S. G. Mech  M. W. Reudink  K. Lew
School of Biological Sciences, Washington State University,
Pullman, WA 99164-4236, USA
e-mail: astorfer@wsu.edu
Present Address:
S. G. Mech
Department of Biology, Albright College,
Reading, PA 19612-5234, USA
Present Address:
M. W. Reudink
Department of Biology, Thompson Rivers University,
Kamloops, BC V2C 0C8, Canada

over small distances, perhaps because high desert grassland
throughout the study area limits the efficacy of inter-pond
movement of salamanders. Further, population sizes and
gene flow of Sonora tiger salamanders are likely compromised by several contemporary ecological threats, including: frequent die-offs due to an infectious virus,
introductions of non-native species, and continuing cattle
grazing. Overall, these genetic data support the endangered
status of the Sonora tiger salamander and suggest the
subspecies exists in small, inbred populations.
Keywords Inbreeding  Genetic structure 
Microsatellites  Salamander  Ambystoma tigrinum

Introduction
Amphibians are declining globally, and studies of genetic
diversity and population genetic structure are crucial for
conservation and management planning (Blaustein and
Kiesecker 2002; Collins and Storfer 2003; Frankham 2003;
Stuart et al. 2004). Amphibian populations are increasingly
threatened by land use change, climate change, introduction of non-native, invasive species and emerging infectious diseases for which transport is enhanced by human
activities (Collins and Storfer 2003; Schloegel et al. 2009,
2010). Amphibian populations fluctuate widely in numbers
from year to year, and consequently, many years of census
data are necessary for accurately detecting population
trends (Pechmann et al. 1991; Alford and Richards 1999).
However, studies of genetic variation facilitate relatively
rapid inferences of population declines or even recent
bottlenecks (i.e., drastic reductions in population size; see
Cornuet and Luikart 1996; Luikart et al. 1998a, b; Garza
and Williamson 2001). Thus, genetic studies facilitate

123

Conserv Genet

management decisions over much shorter time scales than
census studies, which may be critical for saving rapidly
declining or endangered species.
Studies of genetic variation are a critical component of
management programs because genetic variation underlies
the potential for adaptation to future environmental change
(Frankel and Soulé 1981; Frankham 2003). Genetic analyses can be used to prioritize conservation of small populations that are prone to the potential negative impacts of
genetic drift or inbreeding (Lande 1995; Reed et al. 2003).
Small, genetically depauperate populations may consequently become targets of management programs with the
goal of augmenting genetic variation (Hedrick 1995;
Storfer 1999; Frankham 2003).
Genetic studies also determine the extent of gene flow and
consequent population genetic structure. Gene flow is often
seen as a creative force that maintains genetic variation
among populations (Slatkin 1987). However, anthropogenic
forces such as development can restrict gene flow among
populations and, in turn, reduce genetic diversity (Slatkin
1987; Frankham 2003). Data on gene flow among amphibian
populations can also help researchers determine if metapopulation structure exists (Marsh and Trenham 2001;
Storfer 2003; Smith and Green 2005). Identification of
source and sink populations can, in turn, lead to appropriate
management strategies, such as artificial colonization of
locally extirpated populations or augmentation of sink populations via translocations (Vitt et al. 2009). Alternatively, if
amphibian populations are not acting as metapopulations
because gene flow is low, translocations may not be warranted due to the possibility of disrupting local adaptation
(Storfer 1999).
The endangered Sonora tiger salamander, Ambystoma
mavortium stebbinsi (Lowe) is subject to many of the
factors that threaten amphibians globally, including introduction of exotic species, habitat alteration, and frequent
epizootics due to an iridoviral disease (Jancovich et al.
1997, 2005; Storfer et al. 2007). Previous molecular studies
support the distinctiveness of A. m. stebbinsi from other
tiger salamander subspecies (Jones et al. 1988; Storfer et al.
2004). The subspecies is only found within the San Rafael
Valley (SRV) of south-central Arizona (Fig. 1), where the
predominant land use is cattle grazing. Sonora tiger salamanders breed exclusively in human-made, earthen stock
ponds (‘‘cattle tanks’’) and have not been collected from
non human-modified habitats. However, it is thought that
prior to European settlement, A. m. stebbinsi occupied
natural cienegas (marshes) that historically covered parts of
the SRV (Jones et al. 1988, 1995). The southern boundary
of A. m. stebbinsi ends at the US-Mexico border because no
A. m. stebbinsi have been confirmed in Mexico despite
repeated surveys over the last 20 years (Jones et al. 1995,
unpubl.).

123

Fig. 1 Tiger salamander subspecies found in Arizona. Dark stippling
indicates the geographic range of A. m. nebulosum, light stippling
indicates the range of A. m. mavortium, and cross-hatching shows the
range of A. m. stebbinsi. Inset shows the localities sampled for
genetics within the range of A. m. stebbinsi

Our goal in this study were to estimate levels of genetic
variability and genetic population structure among populations of A. m. stebbinsi using microsatellite markers. Due
to the contemporary threats of exotic species introductions,
viral epizootics, and habitat alteration due to cattle grazing,
rapidly evolving microsatellite markers are a particularly
appropriate tool for assessing genetic structure in A. m.
stebbinsi. Therefore, we conducted analyses of genetic
diversity and population structure at 10 microsatellite loci
from 276 samples collected throughout the range of the
subspecies.

Materials and methods
Sampling and DNA isolation
Repeated surveys have identified a total of *40 SRV
ponds between the eastern side of the Patagonia Mountains
and the western and southern sides of the Huachuca

Conserv Genet

Mountains in southern Arizona that contain A. m. stebbinsi.
We sampled 16 of these ponds from 1997 to 1999, and
collected tissue samples from A. m. stebbinsi by walking
the perimeter of each stock tank with a 6 m seine net or the
center of the tank using a 300 deep-water seine (see Fig. 1;
Table 9 in Appendix for locations and sample sizes). When
we caught a salamander, we cut a piece of tissue (*1 cm2)
from the posterior end of the tail and clipped one toe (for
marking purposes). We suspended tissue samples in 95 %
ethanol in sterile 1.5 ml centrifuge tubes on ice, and stored
them at -20 °C upon return to the laboratory.
We extracted DNA from samples collected prior to 2002
using PuregeneÒ DNA isolation kits (Gentra, Inc., Minneapolis, MN) with a modified version of the standard
animal tissue isolation protocol. Our modifications followed three basic steps, all using 20 mg/ml Proteinase-K
solution: (1) tissues were homogenized in PuregeneÒ cell
lysis buffer and digested at 55 °C for 1 h with 10 ll of
Proteinase-K; (2) tissues were re-homogenized and digested at 55 °C overnight; (3) after incubation, samples were
homogenized a third time before proceeding with the
PuregeneÒ protocol. All extracted DNA was suspended in
PuregeneÒ DNA hydration solution and stored at -20 °C.
For samples collected during 2002, we extracted DNA with
a standard phenol–chloroform protocol following Sambrook et al. (1989).

2007). As a result, we ensured that our analyses for this
study only included pure A. m. stebbinsi, by using two
analytical methods to exclude potential hybrids. First, we
examined all individuals for nine microsatellite alleles (all
those included herein except 13-1) unique to 47 A. m.
mavortium from seven sites analyzed previously (Storfer
et al. 2004), and any individual with such an allele was
removed from the analysis. Second, for all individuals
collected from four ponds known to have possible hybrid
populations (SCE, WHN, GYP, and HDI), we used a
maximum likelihood based assignment test in WHICHRUN
(Banks and Eichert 2000) to assign individuals to subspecies using their multilocus microsatellite genotype. ‘‘Pure’’
reference populations of each of the two subspecies were
used as potential parental populations, whereby WHICHRUN
then computed likelihood ratios of belonging in one
parental population versus another. Only those individuals
assigned to A. m. stebbinsi with likelihood ratios exceeding
105 (i.e., that an individual was at least 100,000 times more
likely to be A. m. stebbinsi than A. m. mavortium), were
used in the following analyses. In total, 39 individuals were
excluded due to unique alleles and another 9 were excluded
due to the assignment test.
Statistical analysis
Genetic diversity

Microsatellite characterization
We performed microsatellite genotyping via PCR reactions
of genomic DNA (PCR conditions in Mech et al. 2003) on
10 polymorphic microsatellite loci developed for A. m.
stebbinsi (ATS 4-11; ATS 4-20; ATS 4-25; ATS 5-6; ATS
5-7; ATS 5-8; ATS 10-7; ATS 12-3; ATS 13-1 and ATS
14-3); we used fluorescently labeled heavy strand PCR
primers and ran products on an ABI 377 automated
sequencer with GeneScan Version 3.1 software (Life
Technologies, Foster City, CA). To prevent sample crosscontamination, we used filter tips and negative controls in
all PCR reactions. We scored microsatellite alleles using
Genotyper Version 2.5 (Life Technologies). Alleles were
scored blind by S.G.M., and a known sample standard was
run on every gel to ensure scoring consistency. Alleles that
were much larger or much smaller than the next largest or
smallest allele were confirmed by re-amplification,
sequencing, and verification of the appropriate repeat
structure. All unique alleles (alleles found in only 1 individual) were confirmed by re-amplification and re-scoring.
We showed the possibility of limited hybridization
between A. m. stebbinsi and A.m. mavortium (Storfer et al.
2004), and hybridization between A. m. mavortium and the
California tiger salamander, A. californiense has been welldocumented (Riley et al. 2003; Fitzpatrick and Shaffer

We estimated allelic diversity, overall observed and
expected heterozygosity, and tested for linkage disequilibrium between all possible locus pairs using Fisher’s
exact tests in GENEPOP 3.4 (Raymond and Rousset 1995). In
GENEPOP, we also tested for deviations from Hardy–Weinberg frequencies, reflected as whether the inbreeding
coefficient (f, Weir and Cockerham 1984) is significantly
different from zero; exact p-values are calculated with the
Markov chain method (dememorization 10000; batches
100; iterations per batch 5000; Guo and Thomson 1992).
We performed a sequential Bonferroni correction to minimize type-I errors by adjusting p-values to correct for
multiple table-wide tests (Rice 1989).
For loci that deviated from Hardy–Weinberg expectations, we estimated the frequency of null alleles using
CERVUS 2.0 (Marshall et al. 1998). CERVUS uses the method
proposed by Summers and Amos (1997), which allows
inclusion of double null homozygotes, to compute the
frequency of null alleles (see Garner et al. 2004). To
determine whether null alleles were problematic for loci
that showed significant departures from Hardy–Weinberg
expectations, we calculated the frequency of expected
double null allele genotypes by squaring the expected null
frequency as calculated by CERVUS. We then compared our
observed frequency of double null allele genotypes

123

Conserv Genet

(individual samples that did not amplify) at each locus with
the expected frequency using a binomial test. Note that this
test should be conservative because some samples may not
amplify due to problems in the DNA template, thus
inflating the apparent number with null alleles.
Next, we tested for evidence of recent bottlenecks in
each population with two different methods. The first
method used the program BOTTLENECK (Piry et al. 1999),
which tests whether there are a significant number of loci
that exhibit a heterozygote excess. This excess is expected
to be transient in a population because recent, severe
reductions in effective population size will result in
reductions of allelic diversity faster than heterozygosity
(Cornuet and Luikart 1996). We used the Wilcoxon signrank test in BOTTLENECK to test for heterozygote excesses.
This test is more powerful than the sign test when average
sample size per site is 30 or less, and the other available
test—the standardized differences test—requires a minimum of 20 loci (Luikart et al. 1998b). We used a two-phase
mutation model (TPM), which is modified from the strict
stepwise mutation model (SMM), because it allows for (a
user-defined percentage of) multi-step mutations. Consequently, the TPM was shown to be best fit for most
microsatellite datasets (Di Rienzo et al. 1994; Piry et al.
1999). In BOTTLENECK, we chose 5 % multistep mutations
(Piry et al. 1999), which approximated the frequency of
multistep mutations in our dataset (Mech et al. 2003).
Because allele size range is reduced more slowly than
allelic diversity in cases of reduced effective population
size, the second method we used to test for bottlenecks was
calculating M (where M = k/r, where k = # of alleles per
locus and r = allelic size range for that locus across all
samples; Garza and Williamson 2001). To test significance
of M-values generated for each population, we used Critical_M.exe software (Garza 2001), which generates a
critical M value (that which 5 % of simulations were
below) based on number of individuals sampled and
number of loci using 10,000 replicates. Thus, observed
M values below the critical value suggest significant bottlenecks. In Critical_M.exe, we parameterized the general
SMM model (Garza and Williamson 2001) with the proportion of one-step mutations and average size of multistep mutations (D) that were obtained empirically for each
sampling site. Also needed to parameterize Critical_M.exe
are estimations of long term effective population sizes,
which we estimated based on a coalescent approach using
MIGRATE 1.7.6 (Beerli and Felsenstein 2001; Beerli 2002).
Specifically, a parameter (H—not to be confused with the
F-statistic analog by Weir and Cockerham 1984) equal to
4Nel (where Ne = effective population size l = mutation
rate), is generated by MIGRATE using a Monte Carlo
Markov Chain (MCMC) approach. We used a mutation

123

rate estimated for microsatellites at 10-5 per locus per
generation (Goldstein and Schlötterer 1999). Each population was analyzed separately due to the known low rates of
gene flow among tiger salamander populations (e.g., Spear
et al. 2005, 2006). We analyzed each locus separately using
a Brownian approximation of the SMM (note that, although
TPM is preferred for microsatellites, it is not an available
option in this program), with MCMC searches for 10 short
chains (500 gene trees each) and two long chains (5000
gene trees each). In each of the five replicate runs, the first
10,000 steps were discarded to allow ‘‘burn in’’ and stationarity of parameters was ensured by comparing standard
deviations of likelihood values among replicate runs.
Genetic population structure
We analyzed population genetic structure using analysis of
molecular variance (AMOVA; Excoffier et al. 1992) in
ARLEQUIN 2.0 (Excoffier and Lischer 2010). We also used
the Mantel test (Mantel 1967) in ARLEQUIN to test for a
significant isolation-by-distance relationship. We then used
GDA (Lewis and Zaykin 2001) to estimate Weir and
Cockerham’s (1984) analogs of F-statistics between all
possible population pairs. We did not use RST (Slatkin
1995), which was developed for microsatellites and follows
the SMM, because seven of our 10 loci violate assumptions
of the SMM (four of ten loci were imperfect, and three of
the remaining six loci had large jumps in allelic size distributions see Table 9 in Appendix). F-statistics were
bootstrapped with 10,000 replicates to determine confidence intervals and significance from zero. Because some
of our sample sizes were small (because we were sampling
an endangered species), we also jackknifed the h estimates
over populations using GDA to determine whether sampling error in small populations skewed results. We also
repeated calculations of overall F-statistics for populations
with at least 15 individuals sampled and again for populations with at least 20 individuals sampled. Loci that
showed significant departures from Hardy–Weinberg
equilibrium in multiple populations (due to the possibility
of null alleles) were excluded from the analyses and overall
results were calculated on the remaining loci, as well as the
full 10-locus dataset.

Results
Genetic diversity
A total of 276 individuals were genotyped for an average of
17.2 ± 2.2 (mean ± SE) individuals per population. An
average of 274.1 ± 0.8 individuals were genotyped at each

Conserv Genet
Table 1 Basic genetic
information. Measures of
genetic variation among 10
microsatellite loci in A. m.
stebbinsi among 16 sites in
SRV, AZ

Allelic diversity within each
sampling locality (A), number
of alleles per polymorphic locus
(Ap), average sample size for
each locality (N), and per locus
means across all localities

Population

Locus

Overall

ATS
4-11

ATS
4-20

ATS
4-25

ATS
5-6

ATS
5-7

ATS
5-8

ATS
10-7

ATS
12-3

ATS
13-1

ATS
14-3

A

Ap

N

BOD

2

3

4

2

3

4

3

3

5

4

3.3

3.3

33.2

CAD

2

3

2

2

3

1

3

3

4

3

2.6

2.8

10.0

DAN

1

2

3

2

1

1

3

2

2

2

1.9

2.3

17.9

GYP

2

5

2

3

4

2

2

3

5

4

3.2

3.2

17.0

HDI

4

2

3

2

2

1

3

3

5

2

2.7

2.9

11.0

HUA

3

3

4

3

1

3

2

5

6

4

3.4

3.7

21.9

INZ
LES

3
3

2
4

3
2

3
1

3
2

2
2

5
5

4
2

3
2

5
2

3.3
2.5

3.3
2.7

10.0
10.0

MVT

3

6

2

2

1

3

2

1

4

2

2.6

3.0

14.0

OUT

3

3

3

2

2

1

2

4

5

5

3.0

3.2

17.0

PC1

2

2

1

4

4

2

1

2

3

3

2.4

2.8

17.0

PYL

3

2

2

2

1

2

2

2

3

4

2.3

2.4

12.6

SCE

3

4

3

5

4

3

4

3

4

4

3.7

3.7

39.6

TER

2

2

1

2

1

1

2

2

3

3

1.9

2.3

11.0

WHN

3

4

2

3

6

3

1

2

3

4

3.1

3.3

20.9

WTT

2

3

2

1

2

1

1

3

2

3

2.0

2.4

11.0

Mean

2.6

3.1

2.4

2.4

2.5

2.0

2.6

2.8

3.7

3.4

2.7

3.0

17.1

Overall

8

11

6

8

9

8

8

7

11

5

N

268

275

276

276

276

275

272

275

273

275

locus yielding an average of 8.1 ± 0.6 alleles per locus
overall (Table 1). However, within population allelic
diversity was low, with a mean of 2.7 alleles per locus.
Allele frequencies and sample sizes are given in Table 9 in
Appendix. Observed heterozygosity was 0.191, significantly lower than the expected heterozygosity of 0.332
(f = 0.433; p \ 0.0001; 1000 iterations; Table 2). Most
populations showed significant heterozygote deficiencies as
well (Table 2), suggesting the possibility of non-random
mating. Note that we removed a total of 48 individuals
from the analyses because they were potential hybrids
between A. m. stebbinsi and A. m. mavortium based on
previous work (Storfer et al. 2004). When these individuals
are included in the analysis, results do not differ appreciably, and in many cases f values increased. Thus,
although our inclusion criteria for ‘‘pure’’ A. m. stebbinsi
were stringent, they did not bias results appreciably.
We did not find any significant linkage disequilibrium
among loci. However, five loci (ATS 4-20, ATS 5-7, ATS
10-7, ATS 13-1 and ATS 14-3) showed significant departures from H–W equilibrium overall (Table 2); four of
these five loci showed significant heterozygote deficiency
in at least four sampling localities. Null allele frequencies
were estimated for each of these loci, and observed double
null frequencies were significantly lower than expected
proportions for all five loci (Table 3). This suggests that

departures from Hardy–Weinberg equilibrium (reflected as
significant heterozygote deficiencies) were not a result of
null alleles. Nonetheless, in our overall analyses of population structure below, we conducted two sets of analyses:
with these five loci excluded, as well as with all ten initial
loci.
Evidence of recent bottlenecks was at least partially
supported in separate analyses. Although BOTTLENECK
showed that no populations had a significant proportion of
loci that were in heterozygote excess, five populations
(CAD, DAN, INZ, LES and PC1) showed allelic distribution shifts, which suggests recent bottleneck events
(Luikart et al. 1998a). That is, bottlenecks cause transient
(fewer than a few dozen generations) distortions in allele
frequency distributions, whereby low frequency alleles
(\0.1) decline in occurrence relative to ‘‘intermediate’’
frequency alleles (0.1–0.2). In addition, seven populations
(BOD, GYP, HDI, HUA, INZ, OUT, and SCE) had a
significant proportion of loci in heterozygote deficiency
under the TPM. Migrate results generally showed H values
that were consistent across loci within populations, and
mean values are presented in Tables 4, 5. Calculations of
M statistic were also generally low, ranging from 0.272 to
0.741 (Table 5). All populations showed evidence of recent
bottlenecks because M values were lower than the corresponding critical M value (Mc) for each population.

123

123

0.331

f

0.469

f

0.384

0.27

0.429

0.274

0.679

0.7
0.864

0.6

0.484

0.769

0.416

–

0

0.6

0.484

-0.2

0.337

-0.2

0.337

0.446

0.353

0.505
-0.2

CAD

0.734

0.056

0.205

1

0.214

0.475
-0.055

1

0.108

1

0.381

–

0

–

0

0

0.056

1

0.298

1

0.514

0
–

DAN

0.265

0.229

0.31

0.076

0.508

0.647
0.279

-0.076

0.219

0.458

0.214

0

0.059

0.742

0.223

1

0.221

0.22

0.299

0.286

0.408

0.299
-0.185

GYP

0.308

0.227

0.324

-0.132

0.485

0.658
0.32

0.318

0.394

0.655

0.255

–

0

1

0.173

0

0.091

0.672

0.537

-0.111

0.247

0.398
0.091

HDI

0.516

0.15

0.307

0.728

0.656

0.671
0.667

0.753

0.362

1

0.304

-0.082

0.21

–

0

-0.214

0.376

0.485

0.174

0.661

0.138

0.172
-0.057

HUA

0.547

0.19

0.407

0.045

0.732

0.532
0.449

1

0.505

0.833

0.574

0

0.1

1

0.358

-0.029

0.195

0.449

0.532

1

0.189

0.358
0.455

INZ

0.486

0.23

0.436

0.25

0.526

0.505
0.617

1

0.505

0.857

0.668

0

0.1

1

0.526

0

0

-0.286

0.395

0.176

0.721

0.416
0.04

LES

0.128

0.35

0.399

0.458

0.389

0.701
0.398

–

0

0.447

0.254

-0.138

0.442

–

0

-0.444

0.452

0.286

0.495

0.065

0.762

0.5
0

MVT

0.523

0.165

0.34

0.834

0.69

0.617
0.531

0.033

0.608

1

0.214

–

0

1

0.114

1

0.114

0.034

0.426

0.871

0.444

0.169
-0.043

OUT

0.551

0.153

0.335

0.506

0.586

0.506
0.773

1

0.114

–

0

0.731

0.428

0.609

0.59

0.01

0.357

–

0

1

0.428

0.337
-0.231

PC1

0.484

0.157

0.299

0.681

0.705

0.468
0.795

-0.048

0.159

-0.091

0.212

0

0.077

–

0

-0.043

0.148

-0.091

0.212

1

0.517

0.495
0.231

PYL

0.394

0.209

0.343

0.686

0.474

0.419
0.704

0.114

0.648

-0.06

0.165

-0.03

0.097

0.936

0.383

0.211

0.19

-0.091

0.344

0.851

0.166

0.545
0.289

SCE

0.466

0.118

0.216

0.551

0.394

0.481
0.444

0

0.091

-0.053

0.173

–

0

–

0

-0.053

0.173

–

0

1

0.485

0.368
0.268

TER

0.365

0.259

0.404

0.161

0.735

0.585
0.841

0.022

0.438

–

0

0.481

0.181

0.561

0.535

-0.034

0.138

0.245

0.502

0.573

0.549

0.376
0.07

WHN

0.488

0.127

0.243

-0.127

0.325

0.416
0.574

-0.111

0.329

–

0

–

0

1

0.416

–

0

0.111

0.247

1

0.606

0.091
0

WTT

0.433

0.191

0.332

0.418

0.559

0.372

0.637

0.195

0.806

0.029

0.266

0.617

0.073

Mean

Included is expected heterozygosity (He), inbreeding coefficient, f (Weir and Cockerham 1984) for each locus within each populations and overall. Bolded numbers depict significant departures
from Hardy–Weinberg proportions after Bonferroni correction for multiple tests. A – means a population is fixed at a locus and thus, f, is undefined

0.17

Ho

0.792

0.317

He

f

Overall

0.42

He

ATS14-3

0.468
0.752

0.021

f

He
f

0.36

He

0.425

0.218

He

f

0.198

f

0.824

0.414

f

He

0.165

He

0.058

1

He

0.667

f

f

0.437

He

ATS13-1

ATS12-3

ATS10-7

ATS5-8

ATS5-7

ATS5-6

ATS4-25

0.394

He

ATS4-20

0.032
0

He
f

ATS4-11

BOD

Statistic

Locus

Table 2 Measures of genetic variation among 10 microsatellite loci in A. m. stebbinsi among 16 sites in SRV

Conserv Genet

Conserv Genet
Table 3 Five loci with significant heterozygote deficiencies
Locus

Null frequency

Expected null homozygotes

Observed null homozygotes

p value

ATS 4-20

0.5105

0.2607

0

\0.0001

ATS 5-7

0.7124

0.5075

0

\0.0001

ATS 10-7

0.4029

0.1623

0.0145

\0.0001

ATS 13-1

0.4604

0.212

0.0108

\0.0001

ATS 14-3

0.4232

0.1791

0

\0.0001

For each locus, the frequency of null alleles was estimated by CERVUS 2.0 (Marshall et al. 1998). Expected null homozygotes were calculated by
squaring the estimated null frequency. Observed null homozygotes were calculated by dividing the number of individuals with double null
genotypes (did not amplify) by total number of samples (276). p-values were computed by comparing observed versus expected proportions
using a binomial test

Table 4 Results from BOTTLENECK
Population

Deficient/excess

p-value

Distribution shape

BOD

10/0

0.0005

Normal

CAD

4/5

0.5449

DAN

5/2

GYP

8/2

HDI
HUA

7/2
7/2

Table 5 Estimates of M, as an estimate of putative bottlenecks, for
each population (Garza and Williamson 2001)
Population

4Nel (H)

M

MC

Shifted

BOD

0.145

0.327

0.843

0.3437

Shifted

CAD

0.417

0.399

0.717

0.0020

Normal

DAN

0.092

0.697

0.802

Normal

GYP

0.074

0.402

0.798

Normal

HDI

0.148

0.307

0.725

0.177

0.272

0.657

0.0185
0.0049

INZ

7/3

0.0068

Shifted

HUA

LES

3/6

0.1797

Shifted

MVT

3/5

0.1250

Normal

INZ
LES

0.897
0.193

0.392
0.488

0.586
0.61

OUT

9/0

0.0009

Normal

MVT

0.519

0.456

0.682

Shifted
Normal

OUT

0.119

0.577

0.704

PC1

0.076

0.595

0.713

0.068

0.522

0.771

PC1
PYL

3/5
7/2

0.4720
0.8496

SCE

8/2

0.0068

Normal

PYL

TER

5/2

0.1875

Normal

SCE

0.157

0.443

0.816

0.099

0.741

0.762

WHN

5/4

0.3262

Normal

TER

WTT

4/3

0.2344

Normal

WHN

0.172

0.46

0.647

WTT

0.295

0.489

0.62

The TPM was used with 5 % mutations greater than one-step.
Included are the number of loci that are heterozygote deficient versus
those that show heterozygote excesses, the p-value based on the
Wilcoxon test, and whether the allelic frequency distribution shape
was normal or shifted

4Nel was calculated for each population using a coalescent approach
in MIGRATE (Beerli 2002). M was calculated in each population as the
ratio of number of alleles per locus (k) divided by allelic size range
for that locus (r). Mc was estimated using software from J. C. Garza
(M_crit.exe) and generates a critical value below which M values
indicate significant bottlenecks in their respective populations

Genetic population structure
The AMOVA showed significant subdivision at all levels:
ASC = 0.433, ACT = 0.133, and AST = 0.509 (p \ 0.0001
for all three statistics, Table 6), with the highest proportion
of variation (49.43 %) explained among individuals.
Results were nearly identical and did not differ significantly for all three F-statistics (f, F and h) whether all
populations were sampled, populations with greater than 15
individuals were sampled, or populations with a minimum
of 20 individuals were sampled (Table 7). Further, no
single population was responsible for the large overall h
value (jackknife h values averaged 0.118 ± 0.001). When

the five loci that showed significant departures from
Hardy–Weinberg proportions were excluded, both f and F
decreased significantly, but remained significantly higher
than zero (Table 7). Theta (h), however, did not change
significantly with exclusion of these five loci.
Pairwise h estimates were, with two exceptions, significantly greater than zero, indicating significant population
structure among all but two population pairs (Table 7). The
exceptions were INZ–OUT (6.45 km apart; h = 0.021;
p = 0.09) and TER–WTT (22.92 km apart; h = 0.04;
p = 0.052), although p-values for these two pairs were

123

Conserv Genet
Table 6 AMOVA of 10 microsatellite loci among 16 localities in SRV, Arizona
Source

d.f.

Sum of squares

Variance component

% Variation

p-value

A-Statistic

Among populations

15

164.007

0.253

13.31

\0.00001

ACT = 0.133

Among individuals within populations

260

611.823

0.707

37.27

\0.00001

ASC = 0.433

Within individuals

276

276

0.938

49.43

\0.00001

AST = 0.509

Total

551

1034.83

1.898

Results are from Arlequin Version 2.0 for all locations

Table 7 Resampling of F-statistics using all localities, localities with
a minimum of 15 individuals sampled and localities with a minimum
of 20 individuals sampled
Sampling
localities/loci
included

Fstatistic

Overall

Upper
95 %
CI

Lower
95 %
CI

p-value

All localities; all
10 loci

f

0.433

0.547

0.287

\0.0001

F

0.509

0.609

0.367

\0.0001

h

0.133

0.166

0.1

\0.0001

All localities; 5
loci

f

0.203

0.272

0.103

\0.0001

F

0.293

0.364

0.188

\0.0001

H

0.112

0.132

0.088

\0.0001

Localities w/C15
samples

f

0.454

0.577

0.307

\0.0001

F

0.523

0.625

0.391

\0.0001

h

0.125

0.174

0.077

\0.0001

f

0.429

0.57

0.269

\0.0001

F

0.506

0.628

0.367

\0.0001

h

0.135

0.19

0.078

\0.0001

Localities w/C20
samples

Also included were F-statistics for 5 loci, excluding loci ATS 4-20,
ATS 5-7, ATS 10-7, ATS 13-1 and ATS 14-3, which showed
departures from Hardy–Weinberg equilibrium over all populations

both below 0.1. There was no significant isolation-by-distance pattern revealed by the Mantel test (Matrix correlation = 0.230, Z = 168.1, P = 0.117, 15000 permutations).
Nei’s (1978) genetic distances ranged from 0.0097 (INZ–
OUT) to 0.2906 (LES–SCE). These same pairs had the
lowest and highest h values, respectively (Table 8).

Discussion
Genetic diversity
Overall, A. m. stebbinsi has low genetic diversity and
shows evidence of population bottlenecks among some
populations. The level of genetic diversity based on allelic
richness and heterozygosity is lower than that typical of
other amphibians (Scribner et al. 2001; Garner et al. 2003;
Hoffman and Blouin 2004; Spear and Storfer 2008; Murphy et al. 2010). However, similar levels of heterozygosity
have been documented in a few cases, including A. m.

123

melanostictum in Yellowstone (Spear et al. 2006), the
Pacific giant salamander, Dicamptodon tenebrosus, in a
fragmented landscape in British Columbia (Curtis and
Taylor 2004) and the hellbender, Cryptobranchus alleganiensus (Sabatino and Routman 2009; Crowhurst et al.
2011).
While no bottlenecks were reflected in heterozygote
excesses, A. m. stebbinsi showed evidence of population
bottlenecks in two ways. First, several populations (CAD,
DAN, INZ, LES and PC1) showed shifted allele frequency
distributions. Second, estimates of M ranged from 0.272 to
0.741, and all were lower than critical values estimated for
each particular population. In addition, nearly all of these
values were lower than 0.68, or the critical value proposed
by Garza and Williamson (2001) for concluding decline in
population numbers with at least seven loci.
Previous genetic work on A. m. stebbinsi suggests evidence of a historical bottleneck due to low allozyme variation (Jones et al. 1988), as well as low mitochondrial
DNA variation (Jones et al. 1995; Storfer et al. 2004). We
suggest there also have been recent bottlenecks because
microsatellites are generally reflective of more recent
population-level processes than allozymes or mtDNA data
(Goldstein and Schlötterer 1999; Jehle and Arntzen 2002).
There are several ecological factors that could explain
recent bottlenecks as opposed to natural fluctuations among
ponds containing A. m. stebbinsi. First, non-native species
such as fish and bullfrogs have been introduced sporadically throughout ponds in SRV, causing local extirpation in
some sites and population size reductions in others (Gerst
1997; Collins et al. 1988). During the study period, we
observed at least three predatory fish introductions
(including mosquito fish, Gambusia affinis and green sunfish, Lepomis cyanellus) that resulted in local extirpations
of tiger salamanders (Maret et al. 2006). Several other
studies have shown correlations between fish introductions
and amphibian declines (see Kats and Ferrer 2003). For
example, declines of A. macrodactylum were correlated
with trout stocked in lakes in the North Cascades of
Washington (Tyler et al. 1998). Bullfrogs, which are
excellent competitors and predators of other amphibian
species (Kats and Ferrer 2003), have also been introduced
into the SRV. Recent survey data show presence of

Conserv Genet
Table 8 Genetic distance
BOD

CAD

DAN

GYP

HDI

HUA

INZ

LES

MVT

OUT

PC1

PYL

SCE

TER

WHN

WTT

BOD

*

0.1357

0.0651

0.0480

0.0623

0.0806

0.0521

0.2496

0.1969

0.0572

0.1332

0.1585

0.1906

0.1185

0.1489

0.1326

CAD

0.0848

*

0.1844

0.0403

0.0819

0.0728

0.0425

0.1394

0.1031

0.0676

0.0799

0.0739

0.1623

0.1181

0.0965

0.1242

DAN

0.0239

0.0837

*

0.0722

0.1336

0.1087

0.1487

0.3071

0.2619

0.0845

0.1543

0.1884

0.2755

0.0733

0.2221

0.0919

GYP

0.0229

0.0173

0.0258

*

0.0309

0.0251

0.0403

0.1941

0.1288

0.0241

0.0568

0.0612

0.1685

0.0318

0.1104

0.0527

HDI

0.0315

0.0510

0.0503

0.0146

*

0.0724

0.0409

0.1945

0.1415

0.0617

0.1357

0.1232

0.1625

0.1404

0.1174

0.1577

HUA

0.0413

0.0368

0.0462

0.0125

0.0361

*

0.0536

0.2381

0.1785

0.0495

0.0991

0.0878

0.1509

0.0689

0.1420

0.1273

INZ

0.0271

0.0323

0.0609

0.0183

0.0223

0.0249

*

0.1793

0.1533

0.0207

0.1183

0.1221

0.1265

0.1474

0.0595

0.1606

LES

0.1947

0.1319

0.1893

0.1395

0.1543

0.1851

0.1733

*

0.1473

0.2113

0.2103

0.1928

0.3085

0.2573

0.1983

0.2239

MVT

0.1359

0.0836

0.1525

0.0807

0.0983

0.1218

0.1302

0.1303

*

0.1722

0.1516

0.1178

0.2588

0.2008

0.1301

0.2035

OUT

0.0278

0.0406

0.0329

0.0117

0.0335

0.0258

0.0097

0.1741

0.1274

*

0.0988

0.1008

0.1413

0.0813

0.0802

0.0726

PC1

0.0741

0.0486

0.0669

0.0289

0.0806

0.0539

0.0772

0.1702

0.1067

0.0574

*

0.1136

0.2370

0.0789

0.1688

0.0596

PYL

0.0990

0.0480

0.0809

0.0343

0.0760

0.0527

0.0814

0.1480

0.0773

0.0631

0.0615

*

0.2019

0.1058

0.1506

0.1184

SCE

0.1212

0.1115

0.1675

0.1242

0.1013

0.0880

0.0800

0.2906

0.2132

0.0864

0.1694

0.1433

*

0.2537

0.1097

0.2651

TER

0.0536

0.0516

0.0215

0.0113

0.0597

0.0293

0.0699

0.1660

0.1159

0.0342

0.0328

0.0401

0.1573

*

0.2063

0.0401

WHN

0.0936

0.0778

0.1310

0.0694

0.0810

0.0930

0.0435

0.1919

0.1060

0.0514

0.1259

0.1125

0.0710

0.1306

*

0.1819

WTT

0.0610

0.0627

0.0284

0.0217

0.0753

0.0619

0.0873

0.1464

0.1276

0.0327

0.0260

0.0498

0.1733

0.0124

0.1159

*

Pairwise Fst values (above diagonal, from Arlequin 2.0) and Nei’s (1978) genetic distance (below diagonal, from GDA). Bolded Fst values are not significantly
different from zero at the a = 0.05 level

bullfrogs in several SRV ponds (including BOD, INZ,
OUT and PC1; Storfer unpubl.). Note that two of these
ponds (INZ, PC1) showed shifted allelic distributions.
The second ecological reason that may explain recent
bottlenecks is disease. Salamanders in SRV ponds regularly experience epizootics due to an iridovirus (A. tigrinum
virus or ATV), and evidence suggests this virus has been
recently introduced to SRV (Jancovich et al. 1997, 2005;
Storfer et al. 2007). Although a chytrid fungus is apparently the pathogen most closely linked to amphibian
declines (Stuart et al. 2004; Murray et al. 2009; McCallum
2012), iridoviruses are responsible for the majority of
amphibian epizootics in North America (Green et al. 2002;
Miller et al. 2011). ATV is a virus that is highly infectious,
and frequently lethal, even at low doses (Jancovich et al.
1997, 2005; Brunner et al. 2004, 2007; Storfer et al. 2007).
Die-offs have occurred in at least seven populations studied
herein (BOD, HDI, HUA, INZ, OUT, PC1 and SCE).
During an epizootic event, larval mortality rates can
approach 100 % in a given year, thus potentially causing
population-level bottlenecks (Brunner et al. 2007; Schock
et al. 2009; Greer et al. 2009). Other populations show
evidence of repeated fluctuations in population size, which
is also a source of bottlenecks. That is, repeated temporal
surveys have shown that animals can often return, and
mark-recapture data suggest that these animals are survivors, rather than re-colonists from neighboring ponds
(Brunner et al. 2004). Coupling the low rates of gene flow
observed in this study, die-offs resulting in few survivors, and unlikely recolonization, it is expected that

ATV-induced epizootics contribute to the observed
declines in genetic diversity.
The third ecological reason for potential bottlenecks is
that the SRV is a human-dominated landscape that has
been drastically modified for cattle grazing over the past
150 years. Extensive grazing throughout the valley has
lowered water tables, removing most cienegas that were the
likely historical habitat for A.m. stebbinsi (Collins et al.
1988). The fact that A. m. stebbinsi now relies exclusively
on human-made habitats for breeding and that human
disturbance has generally made SRV drier apparently
limits gene flow among ponds, thus limiting genetic variability. Other work shows lower genetic variability in
disturbed versus undisturbed sites, suggesting gene flow is
limited by disturbance. For example, in Britain the common frog, Rana temporaria, had higher inbreeding levels
disturbed, urban sites as opposed to in rural areas (Hitchings and Beebee 1997). In addition, giant salamanders
showed a significant positive correlation of genetic variation with forest stand age after clear-cutting (Curtis and
Taylor 2004).
Taken together, the above arguments suggest that
within-population departures from Hardy–Weinberg equilibrium are a result of inbreeding resulting from population
bottlenecks and not due to null alleles. Our data showed a
significantly lower proportion of null homozygotes than
expected, several populations showed evidence of shifted
allele frequency distributions, and all populations had
M values below the critical value. Evidence of inbreeding
depression is reflected in lab-rearing experiments that show

123

Conserv Genet

lower embryonic survival, lower larval survival, and lower
disease resistance in A. m. stebbinsi relative to A. m. nebulosum (Storfer unpubl.), which has higher genetic variation and lower levels of inbreeding (Jones et al. 1988,
1995).
Genetic population structure
A. t. stebbinsi shows a high degree of population genetic
structure, with nearly all population pairs significantly
subdivided from one another, despite some inter-pond
distances that are less than one kilometer. In addition, there
was no significant isolation-by-distance pattern, probably
because nearly all populations were significantly subdivided from each other. Levels of gene flow documented
herein are generally lower than microsatellite studies of
pond breeding frogs (Newman and Squire 2001; Garner
et al. 2003; Emel and Storfer 2012) but similar to that
reported for the hellbender (Crowhurst et al. 2011). Sonora
tiger salamanders also show lower gene flow than their
sister taxon, A. californiense, which has a high degree of
genetic connectivity and dispersal on a local scale (Trenham et al. 2001).
Several factors could limit gene flow among ponds in
SRV. First, SRV ponds are generally separated by open,
dry grassland (Jones et al. 1988), and movement across
grassy fields in SRV could increase risk of both desiccation
and predation. Drying of these habitats has been exacerbated by cattle grazing, which has lowered the water table
in SRV (Gerst 1997). Several studies have shown ambystomatids tend to avoid open habitats (deMaynadier and
Hunter 1999; Trenham 2001; Rothermel and Semlitsch
2002; Spear et al. 2006). A radiotelemetry study of the
eastern tiger salamander, A. tigrinum, showed avoidance of
grassy fields (Madison and Farrand 1998), and genetic
subdivision was found among populations of A. tigrinum
separated by fields in Nebraska (Routman 1993). Another
study showed that juvenile amphibians (including A.
maculatum) moved less and were recaptured less often in
fields as compared to forested areas (Rothermel and Semlitsch 2002).
Second, because habitats in SRV are managed and relatively permanent, presence of perennibranchiate (sexually
mature, gilled) adults is common in A. m. stebbinsi (Jones
et al. 1988). These individuals are unable to disperse among
ponds, thus limiting gene flow and potentially genetic

123

variability among ponds. Support for this prediction comes
from a study of among populations of A. m. melanostictum in
northern Yellowstone, where presence of aquatic adults
appears to limit gene flow (Spear et al. 2005, 2006). Another
study showed that perennibranchiate tiger salamander populations had lower genetic variability than metamorphic
populations (Shaffer and Breden 1989).
Small population sizes could also contribute to differentiation among SRV populations via genetic drift. Low
effective population sizes that are likely caused by a
combination of disease, habitat disturbance and introduction of exotic species may mean that the magnitude of
genetic drift is sufficiently large to mask gene flow. Thus,
among populations separated by short distances that are
apparently subdivided genetically, gene flow may be
swamped by genetic drift.
In summary, the Sonora tiger salamander shows evidence of a high degree of inbreeding and strong genetic
population structure. The high level of inbreeding and low
genetic variability are of particular concern if there is no
longer sufficient genetic variability for the subspecies to
survive, for example, introduction of new viral strains
(Jancovich et al. 2005; Picco et al. 2010). The subspecies
appears to be particularly prone to drastic reductions in
population size and, consequently, local extinction. In
addition, the high degree of genetic structuring among
populations suggests low probability of re-colonization
following local extinction events. Therefore, unlike the
endangered California tiger salamander (Trenham et al.
2001; Smith and Green 2005), Sonora tiger salamanders
apparently do not operate as a metapopulation, and continued management may include artificial re-colonizations.
Based on the genetic data presented herein, the endangered
status of the species is supported.
Acknowledgments We thank R. Ziemba, J. Snyder, M. Myers and
C. Steele for help collecting tissue samples. Animals and tissue
samples were collected under Arizona State Department of Fish and
Game Collecting Permit #SP744166 and US Fish and Wildlife Permit
#TE043941-1. This research followed the guidelines of Institutional
Animal Care and Use Committees at Arizona State University (#99670R) and Washington State University (ASAF #3203). This work
was supported by US Army Contract #DABT63-99-P-0087 and NSF
#0548415 to A.S.

Appendix
See Table 9.

ATS5-6

ATS4-25

–

0.721

0.029

0.221

233

235

–

209

201

–

–

199

–

0.971

197

207

–

195

203

0.029

–

193

34

–

–

229

231

(N)

–

0.029

–

–

–

–

0.800

–

–

0.200

10

0.200

–

0.800

–

–

10

–

225

374

223

–

–

363

–

34

–

361

–

–

–

0.050

–

0.150

0.800

–

10

–

–

0.400

–

0.600

–

–

–

10

CAD

(N)

–

–

345

–

339

343

0.029

–

335

337

0.221

–

223

333

–

195

0.750

0.016

193

–

–

191

331

0.984

189

329

–

187

34

–

183

(N)

–

ATS4-20

31

(N)

175

ATS4-11

BOD

Allele

Locus

–

–

–

–

0.972

–

–

0.028

18

–

0.056

0.833

0.111

–

–

18

–

–

–

–

–

–

–

–

0.500

0.500

–

18

–

–

–

–

1.000

–

–

–

18

DAN

–

–

0.059

–

0.882

0.059

–

–

17

0.176

–

0.824

–

–

–

17

–

0.029

0.029

–

–

–

0.059

–

0.118

0.765

–

17

–

–

0.176

–

0.824

–

–

–

17

GYP

–

–

–

–

0.955

–

–

0.045

11

–

0.091

0.636

0.273

–

–

11

–

–

–

–

–

–

0.136

–

–

0.864

–

11

–

0.045

0.136

–

0.773

–

–

0.045

11

HDI

Table 9 Sample sizes (N) and allele frequencies for each population

–

–

–

–

0.773

0.182

–

0.045

22

0.023

–

0.909

–

0.023

0.045

22

–

–

–

–

–

–

0.024

–

0.048

0.929

–

21

0.023

–

0.068

–

0.909

–

–

–

22

HUA

–

–

–

–

0.900

0.050

–

0.050

10

0.250

–

0.650

–

–

0.100

10

–

–

–

–

0.100

–

–

–

–

0.900

–

10

–

0.100

0.100

–

0.800

–

–

–

10

INE

–

–

–

–

1.000

–

–

–

10

0.250

–

0.750

–

–

–

10

–

–

–

0.050

0.300

0.400

–

–

–

0.250

–

10

–

–

0.200

0.050

0.750

–

–

–

10

LES

–

–

–

–

0.679

–

–

0.321

14

0.393

–

0.607

–

–

–

14

0.071

–

–

–

–

0.214

0.071

–

0.143

0.429

0.071

14

–

–

0.321

0.036

0.643

–

–

–

14

MVT

–

–

–

0.059

0.941

–

–

–

17

0.206

–

0.735

–

–

0.059

17

–

–

–

–

–

–

–

0.029

0.265

0.706

–

17

–

–

0.059

0.029

0.912

–

–

–

17

OUT

0.029

0.029

–

0.147

0.794

–

–

–

17

–

–

1.000

–

–

–

17

–

–

–

–

–

–

–

–

0.294

0.706

–

17

–

–

0.206

–

0.794

–

–

–

17

PC1

–

–

–

–

0.923

0.077

–

–

13

0.115

–

0.885

–

–

–

13

–

–

–

–

–

–

–

–

–

0.462

0.538

13

–

–

0.308

–

0.654

–

0.038

–

13

PYL

–

–

–

0.025

0.900

0.037

0.025

0.013

40

0.200

0.013

0.787

–

–

–

40

–

–

–

–

–

–

0.013

–

0.050

0.912

0.025

40

–

–

0.389

0.056

0.556

–

–

–

36

SCE

–

–

–

–

0.909

0.091

–

–

11

–

–

1.000

–

–

–

11

–

–

–

–

–

–

–

–

0.364

0.636

–

11

–

–

–

–

0.773

0.227

–

–

11

TER

–

–

–

–

0.929

0.024

–

0.048

21

0.571

–

0.429

–

–

–

21

–

–

–

0.095

–

–

0.071

–

0.190

0.643

–

21

–

–

0.175

0.050

0.775

–

–

–

20

WHN

–

–

–

–

1.000

–

–

–

11

0.136

–

0.864

–

–

–

11

–

–

–

–

–

–

–

–

0.455

0.455

0.091

11

–

–

0.045

–

0.955

–

–

–

11

WTT

Conserv Genet

123

123

ATS12-3

ATS10-7

ATS5-8

–

0.191

0.029

0.779

–

–

213

221

223

225

227

–

314

–

0.067

304

193

–

302

191

0.733

300

34

0.200

298

(N)

–

296

–

0.030

370

372

–

0.197

368

294

0.742

366

–

0.030

364

30

–

362

(N)

–

292

–

–

260

360

0.015

258

356

–

256

33

–

254

(N)

–

–

246

0.074

0.912

244

248

–

250

34

(N)

242

ATS5-7

BOD

Allele

Locus

Table 9 continued

–

0.200

0.700

0.100

–

–

–

10

–

–

–

0.750

0.050

0.200

–

–

10

–

–

–

1.000

–

–

–

–

10

–

–

0.200

0.100

–

–

–

0.700

–

10

CAD

–

0.056

0.944

–

–

–

–

18

–

–

–

0.778

0.111

0.111

–

–

18

–

–

–

1.000

–

–

–

–

18

–

–

–

–

–

–

–

1.000

–

18

DAN

–

–

0.882

0.088

0.029

–

–

17

–

–

–

0.882

0.118

–

–

–

17

–

–

–

0.971

0.029

–

–

–

17

–

–

0.029

0.029

–

–

–

0.882

0.059

17

GYP

–

–

0.773

0.091

0.136

–

–

11

–

0.091

–

0.864

0.045

–

–

–

11

–

–

–

1.000

–

–

–

–

11

–

0.091

–

–

–

–

–

0.909

–

11

HDI

–

0.091

0.795

0.068

–

0.023

0.023

22

–

–

–

0.818

–

0.182

–

–

22

–

–

–

0.886

–

0.091

0.023

–

22

–

–

–

–

–

–

–

1.000

–

22

HUA

0.100

0.100

0.700

0.100

–

–

–

10

0.100

0.100

0.100

0.650

0.050

–

–

–

10

–

–

0.050

0.950

–

–

–

–

10

–

–

0.100

0.100

–

–

–

0.800

–

10

INE

–

0.400

0.600

–

–

–

–

10

–

–

–

0.550

0.050

0.100

0.200

0.100

10

–

–

0.050

0.950

–

–

–

–

10

–

–

–

–

0.500

–

–

0.500

–

10

LES

–

–

1.000

–

–

–

–

14

–

–

–

0.857

0.143

–

–

–

14

–

0.036

0.250

0.714

–

–

–

–

14

–

–

–

–

–

–

–

1.000

–

14

MVT

–

–

0.588

0.206

0.088

0.118

–

17

–

–

0.118

0.882

–

–

–

–

17

–

–

–

1.000

–

–

–

–

17

–

–

–

–

–

–

–

0.941

0.059

17

OUT

–

0.059

0.941

–

–

–

–

17

–

–

–

1.000

–

–

–

–

17

–

–

0.294

0.706

–

–

–

–

17

–

–

–

–

–

0.059

0.088

0.588

0.265

17

PC1

–

–

0.917

–

0.083

–

–

12

–

–

–

0.885

–

0.115

–

–

13

–

–

–

0.962

–

–

0.038

–

13

–

–

–

–

–

–

–

1.000

–

13

PYL

–

–

0.388

0.412

0.200

–

–

40

–

–

–

0.912

0.063

0.013

0.013

–

40

–

–

–

0.950

–

–

0.037

0.013

40

–

–

0.063

0.037

–

–

–

0.775

0.125

40

SCE

–

–

0.955

0.045

–

–

–

11

–

–

–

0.909

0.091

–

–

–

11

–

–

–

1.000

–

–

–

–

11

–

–

–

–

–

–

–

1.000

–

11

TER

–

–

0.690

0.310

–

–

–

21

–

–

–

1.000

–

–

–

–

21

–

–

–

0.905

0.048

–

0.048

–

21

0.024

0.143

0.048

0.024

0.095

–

–

0.667

–

21

WHN

–

–

0.818

0.091

0.091

–

–

11

–

–

–

1.000

–

–

–

–

11

–

–

–

1.000

–

–

–

–

11

–

–

–

–

–

–

–

0.727

0.273

11

WTT

Conserv Genet

0.045

0.818
0.095
0.773
0.125
0.462
0.588
0.500
0.250
0.500
0.400
0.500
0.636
0.676
0.882

–

0.750
129

0.450

0.286

–
–

0.136
0.700

–
–

0.154
–

–
0.176

0.059
–

–
–

–
0.050

0.100
–

0.295
–

–
–

0.059
0.118

–

0.118
127

–

–
121

–

0.136

–
0.310
–
0.013
0.115
0.206
0.206
–
–
0.350
0.091
–
0.059
–
0.044
105

0.250

11
21

0.310
0.091

11
40

0.163
0.269

13
17

0.206
0.059

17
14

0.750
0.500

10
10

0.100
0.114

22
11

0.364
0.206

17
17

–
0.300
0.088

10
34
(N)

0.029
371

103

ATS14-3

0.727

–
–
–
–
–
–
0.059
–
–
–
0.045
–
0.059
–

–

0.191
369

0.300

0.286

–
–

0.682
0.200

–
–

0.700
0.647

–
–

0.382
0.429

–
–

0.400
0.250

–
–

0.409
0.136

0.045
–

0.441
0.361

–
–
367

0.450

–
341

0.050

–

–
0.143

–
–

–
0.050

–
–

0.050
0.059

–
–

0.029
0.321

–
–

0.600
0.100

–
0.045

0.023
0.227

–
–

0.059
–

–

0.044
339

–

–

–

–

–
–
–
–
–
–
–
–
–
–
–
–
0.029
–

–

337

–

–

–
–

0.045
–

–
–

–
–

–
0.029

–
–

–
–

–
–

–
–

–
–

–
–

–
–

–
–

–

–

–

333

0.029

335

0.273

–
–
–
–
–
–
–
0.071
–
–
0.068
0.045
–
–

–

331

–

0.571
0.273
0.738
0.250
0.294
0.500
0.179
–
0.650
0.409
0.545
0.412
0.639
0.706
329

0.200

11
21

–
–

11
40

0.013
–

10
17

–
–

17
14

–
–

10
10

–
–

22
11

–
–

17
18

–
–
–

10
34
(N)

327

ATS13-1

Table 9 continued

BOD
Allele
Locus

CAD

DAN

GYP

HDI

HUA

INE

LES

MVT

OUT

PC1

PYL

SCE

TER

WHN

WTT

Conserv Genet

References
Alford RA, Richards SJ (1999) Global amphibian declines: a problem
in applied-ecology. Ann Rev Ecol Syst 30:133–165
Banks MA, Eichert W (2000) WHICHRUN (version 3.2) a computer
program for population assignment of individuals based on
multilocus genotype data. J Hered 91:87–89
Beerli P (2002) MIGRATE-N Documentation and program, part of
LAMARC software package. http://popgen.sc.fsu.edu/Migrate/
Migrate-n.html. Accessed 29 March 2013
Beerli P, Felsenstein J (2001) Maximum likelihood estimation of a
migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proc Natl Acad Sci USA
98:4563–4568
Blaustein AR, Kiesecker JM (2002) Complexity in conservation: lessons
from the global decline of amphibians. Ecol Lett 5:597–608
Brunner JL, Schock DM, Davidson EW, Collins JP (2004) Intraspecific reservoirs: complex life history and the persistence of a
lethal ranavirus. Ecology 85:560–566
Brunner JL, Schock DM, Collins JP (2007) Transmission dynamics of
the amphibian ranavirus, Ambystoma tigrinum virus. Dis Aquat
Org 77:87–95
Collins JP (1981) Distribution, habitats, and life history in the tiger
salamander, Ambystoma tigrinum, in east-central and southeast
Arizona. Copeia 1981:666–675
Collins JP, Storfer A (2003) Global amphibian declines: sorting the
hypotheses. Divers Dist 9:89–98
Collins JP, Jones TR, Berna HA (1988) Conserving genetically
distinctive populations: the case of the Huachuca tiger salamander (Ambystoma tigrinum stebbinsi). In: Szaro RC, Severson KC,
Patton DR (eds) Management of amphibians, reptiles and small
mammals in North America. USDA Forest Service GTR-RM166, Rocky Mountain Forest and Range Experiment Station, Fort
Collins, CO, pp 45–53
Collins JP, Brunner JL, Miera V, Parris MJ, Schock DM, Storfer A
(2003) Ecology and evolution of infectious disease. In: Semlitsch RD (ed) Amphibian conservation Smithsonian Institution,
Washington, DC, p 137–151
Cornuet JM, Luikart G (1996) Description and power analysis of two
tests for detecting recent population bottlenecks from allele
frequency data. Genetics 144:2001–2014
Crowhurst RS, Faries KM, Collantes J, Briggler JT, Koppelman JB,
Eggert LS (2011) Genetic relationships of hellbenders in the
Ozark highlands of Missouri and conservation implications for
the Ozark subspecies (Cryptobranchus alleganiensis bishopi).
Conserv Genet 12:637–646
Curtis JMR, Taylor EB (2004) The genetic structure of coastal giant
salamanders (Dicamptodon tenebrosus) in a managed forest.
Biol Conserv 115:45–54
DeMaynadier PG, Hunter ML (1999) Forest canopy closure and
juvenile emigration by pond-breeding amphibians in Maine.
J Wildl Manag 63:441–450
Di Rienzo A, Peterson AC, Garza JC et al (1994) Mutational
processes of simple-sequence repeat loci in human populations.
Proc Natl Acad Sci USA 91:3166–3170
Emel SL, Storfer A (2012) A decade of amphibian population genetic
studies: synthesis and recommendations. Conserv Genet 13:
1685–1689
Excoffier L, Lischer HEL (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
Excoffier L, Smouse PE, Quattro JM (1992) Analysis of molecular
variance inferred from metric distances among DNA haplotypes:
application to human mitochondrial DNA restriction sites.
Genetics 131:479–491

123

Conserv Genet
Fitzpatrick BM, Shaffer HB (2007) Introduction history and habitat
variation explain the landscape genetics of hybrid tiger salamanders. Ecol Appl 17:598-608
Frankel OH, Soulé ME (1981) Conservation and evolution. Cambridge University Press, Cambridge
Frankham R (2003) Genetics and conservation biology. Comp Rend
Biol 326:S22–S29
Garner TWJ, Angelone S, Pearman PB (2003) Genetic depletion in
Swiss populations of Rana latastei: conservation implications.
Biol Conserv 114:371–376
Garner TWJ, Pearman PB, Angelone S (2004) Genetic diversity
across a vertebrate species’ range: a test of the central-peripheral
hypothesis. Mol Ecol 13:1047–1053
Garza JC (2001) Critical_M.exe software. www.pfeg.noaa.gov/tib/
staff/carlos_garcia/carlos%20software.html. Accessed Aug 2007
Garza JC, Williamson EG (2001) Detection of reduction in population
size using data from microsatellite loci. Mol Ecol 10:305–318
Gerst JL (1997) Endangered and threatened wildlife and plants;
determination of endangered status for three wetland species
found in southern Arizona and northern Sonora, Mexico. Fed
Regist 62:665–689
Goldstein DB, Schlötterer C (1999) Microsatellites: evolution and
applications. Oxford University Press, Oxford
Green DE, Converse KA, Schrader AK (2002) Epizootiology of sixtyfour amphibian morbidity and mortality events in the USA,
1996–2001. Ann N Y Acad Sci 969:323–339
Greer AL, Brunner JL, Collins JP (2009) Spatial and temporal patterns
of Ambystoma tigrinum virus prevalence in tiger salamanders
Ambystoma tigrinum nebulosum. Dis Aquat Org 95:1–6
Guo SW, Thomson EA (1992) Performing the exact test of Hardy–
Weinberg proportions for multiple alleles. Biometrics 48:
361–372
Hedrick PW (1995) Gene flow and genetic restoration: the Florida
panther as a case study. Conserv Biol 9:995–1007
Hitchings SP, Beebee TJC (1997) Genetic substructuring as a result of
barriers to gene flow in urban Rana temporaria (common frog)
populations: implications for biodiversity conservation. Heredity
79:117–127
Hoffman EA, Blouin MS (2004) Historical data refute recent range
contraction as cause of low genetic diversity in isolated frog
populations. Mol Ecol 13:271–276
Jancovich JK, Davidson EW, Morado JF, Jacobs BL, Collins JP
(1997) Isolation of a lethal virus from the endangered tiger
salamander, Ambystoma tigrinum stebbinsi Lowe. Dis Aquat Org
31:161–167
Jancovich JK, Mao JH, Chinchar VG et al (2003) Genomic sequence
of a ranavirus (family Iridoviridae) associated with salamander
mortalities in North America. Virology 316:90–103
Jancovich JK, Davidson EW, Parameswaran N, Mao J, Chinchar VG,
Collins JP, Jacobs BL, Storfer A (2005) Evidence for emergence
of an amphibian disease due to human-enhanced spread. Mol
Ecol 14:213–224
Jehle R, Arntzen JW (2002) Review: microsatellite markers in
amphibian conservation genetics. Herpetol J 12:1–9
Jones TR, Collins JP, Kocher TD, Mitton JB (1988) Systematic status
and distribution of Ambystoma tigrinum stebbinsi Lowe
(Amphibia: Caudata). Copeia 1988:621–635
Jones TR, Routman EJ, Begun DJ, Collins JP (1995) Ancestry of an
isolated subspecies of salamander, Ambystoma tigrinum stebbinsi Lowe: the evolutionary significance of hybridization. Mol
Phylog Evol 4:194–202
Kats LB, Ferrer RP (2003) Alien predators and amphibian declines.
Divers Dist 9:99–110
Lande R (1995) Mutation and conservation. Conserv Biol 9:782–791

123

Lewis PO, Zaykin D (2001). Genetic data analysis: computer program
for the analysis of allelic data. Version 1.1. http://lewis.eeb.
uconn.edu/lewishome/software.html. Accessed Aug 2007
Luikart G, Allendorf FW, Cornuet J-M, Sherwin WB (1998a)
Distortion of allele frequency distributions provides a test for
recent population bottlenecks. J Hered 89:238–247
Luikart G, Sherwin WB, Steele BM, Allendorf FW (1998b)
Usefulness of molecular markers for detecting bottlenecks via
monitoring genetic change. Mol Ecol 7:963–974
Madison DM, Farrand L (1998) Habitat use during breeding and
emigration in radio implanted tiger salamanders, Ambystoma
tigrinum. Copeia 1998:402–410
Mantel N (1967) The detection of disease clustering and a generalized
regression approach. Cancer Res 27:209–220
Maret TJ, Snyder JD, Collins JP (2006) Altered drying regime
controls distribution of endangered salamanders and introduced
predators. Biol Conserv 127:129–138
Marsh DM, Trenham PC (2001) Metapopulation dynamics and
amphibian conservation. Conserv Biol 15:40–49
Marshall TC, Slate J, Kruuk L (1998) Statistical confidence for
likelihood-based paternity inference in natural populations. Mol
Ecol 7:639–655
McCallum H (2012) Disease and the dynamics of extinction. Phil
Trans R Soc B 367:2828–2839
Mech SG, Storfer A, Ernst JA, Reudink MW, Maloney SC (2003)
Polymorphic microsatellite loci for tiger salamanders, Ambystoma tigrinum. Mol Ecol Notes 3:79–81
Miller D, Gray MJ, Storfer A (2011) Ecopathology of ranaviruses
infecting amphibians. Viruses 3:2351–2373. doi:10.3390/v3112351
Murphy MA, Dezzani R, Pilliod DS, Storfer A (2010) Landscape
genetics of high mountain frog metapopulations: an application
of gravity models. Mol Ecol 19:3634–3649
Murray KA, Skerratt LF, Speare R et al (2009) Impact and dynamics
of disease in species threatened by the chytrid fungus (Batrachochytrium dendrobatidis). Conserv Biol 5:1242–1252
Nei M (1978) Estimation of average heterozygosity and genetic
distance from a small number of individuals. Genetics 89:
583–590
Newman RA, Squire T (2001) Microsatellite variation and fine-scale
population structure in the wood frog (Rana sylvatica). Mol Ecol
10:1087–1100
Pechmann JHK, Scott DE, Semlitsch RD et al (1991) Declining
amphibian populations: the problem with separating human
impacts from natural fluctuations. Science 253:892–895
Picco AM, Karam AP, Collins JP (2010) Pathogen host switching in
commercial trade with management recommendations. Ecohealth 7:252–256
Piry S, Luikart G, Cournet J-M (1999) BOTTLENECK: a computer
program for detecting recent reductions in the effective population size using allele frequency data. J Hered 90:502–503
Raymond M, Rousset F (1995) GENEPOP (version 1.2): population
genetics software for exact tests and ecumenicism. J Hered
86:248–249
Reed DH, Lowe EH, Briscoe DA, Frankham R (2003) Inbreeding and
extinction: effects of rate of inbreeding. Conserv Genet 4:
405–410
Rice WR (1989) Analyzing tables of statistical tests. Evolution
43:223–225
Riley SPD, Shaffer HB, Voss S, Fitzpatrick BM (2003) Hybridization
between a rare, native tiger salamander (Ambystoma californiense) and its introduced congener. Ecol Appl 13:1263–1275
Rothermel BB, Semlitsch RD (2002) An experimental investigation
of landscape resistance of forest versus old-field habitats to
emigrating juvenile amphibians. Conserv Biol 16:1324–1332

Conserv Genet
Routman E (1993) Population structure and genetic diversity of
metamorphic and paedomorphic populations of the tiger salamander, Ambystoma tigrinum. J Evol Biol 6:329–357
Sabatino SJ, Routman E (2009) Phylogeography and conservation
genetics of the hellbender salamander (Cryptobranchus alleganiensis). Conserv Genet 10:1235–1246
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning, 2nd
edn. Cold Spring Harbor, Plainview
Schloegel LM, Picco AM, Kilpatrick AM et al (2009) Magnitude of
the US trade in amphibians and presence of Batrachochytrium
dendrobatidis and ranavirus infection in imported North American bullfrogs (Rana catesbeiana). Biol Conserv 142:1420–1426
Schloegel LM, Dazack P, Cunningham AA et al (2010) Two
amphibian diseases, chytridiomycosis and ranaviral disease, are
now globally notifiable to the World Organization for Animal
Health (OIE): an assessment. Dis Aquat Org 92:101–108
Schock DM, Bollinger TK, Collins JP (2009) Mortality rates differ
among amphibian populations exposed to three strains of a lethal
ranavirus. Ecohealth 6:438–448
Scribner KT, Arntzen JW, Cruddace N, Oldham RS, Burke T (2001)
Environmental correlates of toad abundance and population
genetic diversity. Biol Conserv 98:201–210
Shaffer HB, Breden F (1989) The relationship between allozyme
variation and life history: non-transforming salamanders are less
variable. Copeia 1989:1016–1023
Slatkin M (1987) Gene flow and the geographic structure of natural
populations. Science 236:787–792
Slatkin M (1995) A measure of population subdivision based on
microsatellite allele frequencies. Genetics 139:457–462
Smith MA, Green DM (2005) Are all amphibian populations
metapopulations? Dispersal and the metapopulation paradigm
in amphibian ecology. Ecography 28:110–128
Spear SF, Peterson CR, Matocq M, Storfer A (2005) Landscape
genetics of the blotched tiger salamander, Ambystoma tigrinum
melanostictum. Mol Ecol 14:2553–2564
Spear SF, Peterson CR, Matocq M, Storfer A (2006) Molecular
evidence for recent population size reductions of tiger

salamanders (Ambystoma tigrinum) in Yellowstone National
Park. Conserv Genet 7:605–611
Spear SF, Storfer A (2008) Landscape genetic structure of tailed frogs
in protected versus managed forests. Mol Ecol 17:4642–4656
Storfer A (1999) Gene flow and endangered species translocations: a
topic revisited. Biol Conserv 87:173–180
Storfer A (2003) Amphibian declines: future directions. Divers
Distrib 9:151–163
Storfer A, Mech SG, Reudink MW, Ziemba RE, Warren JL, Collins
JP (2004) Introgression by non-native species in the endangered
tiger salamander, Ambystoma tigrinum stebbinsi. Copeia 2004:
783–796
Storfer A, Alfaro MR, Ridenhour BJ, Jancovich JK, Mech SG, Parris
MJ, Collins JP (2007) Phylogenetic concordance analysis shows
an emerging pathogen is novel and endemic. Ecol Lett 10:
1075–1083
Stuart SN, Chanson JS, Cox NA, Young BE, Rodrigues ASL,
Fischman DL, Waller RW (2004) Status and trends of amphibian
declines and extinctions worldwide. Science 306:1783–1786
Summers K, Amos W (1997) Behavioral, ecological and molecular
genetic analyses of dart-poison frog, Dendrobates ventrimaculatus. Behav Ecol 8:260–267
Trenham PC (2001) Terrestrial habitat use by adult California tiger
salamanders. J Herpetol 35:343–346
Trenham PC, Koenig WD, Shaffer HB (2001) Spatially autocorrelated demography and interpond dispersal in the salamander
Ambystoma californiense. Ecology 82:3519–3530
Tyler TJ, Liss WJ, Ganio LM, Larson GL, Hoffman RL, Deimling E,
Lomnicky G (1998) Interaction between introduced trout and
larval salamanders (Ambystoma macrodactylum) in high elevation lakes. Conserv Biol 12:94–105
Vitt P, Havene K, Hoegh-Guldberg O (2009) Assisted migration: part
of an integrated conservation strategy. Trends Ecol Evol 9:
473–474
Weir BS, Cockerham CC (1984) Estimating F-statistics for the
analysis of population structure. Evolution 38:1358–1370

123