Conservation Genetics

https://doi.org/10.1007/s10592-019-01155-7

RESEARCH ARTICLE

Isolation by elevation: mitochondrial divergence among sky island

populations of Sacramento Mountain salamander (Aneides hardii)
Megan J. Osborne1 · Samantha J. Cordova1 · Alexander C. Cameron1 · Thomas F. Turner1

Received: 1 October 2018 / Accepted: 6 February 2019

© Springer Nature B.V. 2019

Abstract
The Sacramento Mountain salamander (Aneides hardii) is a fully terrestrial plethodontid endemic to mountains in south-
central New Mexico. This species is of conservation concern but there is scant knowledge regarding the degree of genetic
divergence among populations. This information is vital for developing an appropriate species management strategy. To
address this issue, we generated sequence data from the mitochondrial cytochrome b gene and four nuclear loci to explore
population demographic history. Cytochrome b data revealed three deeply divergent (2.3–2.8%) mitochondrial lineages
corresponding to distinct mountains (Capitan, White, and Sacramento). Divergence dates suggest separation since the early
Pleistocene, and signatures of population expansion in the mid to late Pleistocene. Of 25 haplotypes identified, none were
shared among mountains, but genetic diversity differed among mitochondrial lineages and suggested different demographic
histories within lineages. A single microsatellite locus identified private alleles in each of the three lineages. Other nuclear
loci screened were invariant within and among populations, which corresponds to a lower mutation rate compared with
mtDNA. Together these results indicate that each mountain population of A. hardii represents a demographically and geneti-
cally distinct management unit.

Keywords  Demographic history · Fragmentation · Gene flow · Management unit · Mitochondrial DNA · Plethodontidae ·
Sky islands

Introduction more mesic environments (Warshall 1995; Holycross and

In arid regions of the southwestern United States and north-
ern Mexico, climatic oscillations during the late Pleistocene
induced major shifts in habitat availability. Warm inter-
glacial periods reduced forested habitat to high-elevation
refugia accompanied by extensive desertification at lower
elevations (Thompson and Anderson 2000; Smith and Far-
rell 2005) resulting in formation of isolated habitats known
as sky islands (Warshall 1995). Sky islands are vertically
arranged biotic communities distributed along a strong eco-
logical gradient, from lower and more xeric, to elevated and
Electronic supplementary material  The online version of this
article (https​://doi.org/10.1007/s1059​2-019-01155​-7) contains
supplementary material, which is available to authorized users.
* Megan J. Osborne
mosborne@unm.edu
1
Department of Biology and Museum of Southwestern
Biology, University of New Mexico, Albuquerque, NM,
USA
Douglas 2007). Intervening ‘seas’ of inhospitable habitat act
as major barriers to movement and impede colonization of
adjacent mountains. Sky island complexes represent unique
biological systems and can be centers of endemism and
harbor high species diversity (Warshall 1995; McCormak
2009). Species comprising these complexes are susceptible
to effects of global climate change (Parmesan 2006; Sorte
and Jetz 2010; Lyons et al. 2016) because they have small
geographic ranges, are demographically isolated and cannot
escape the effects of warming by migrating northward (Wil-
son et al. 2007; Moritz et al. 2008; Sekercioglu et al. 2008)
or by moving upslope.
Divergence of sky island populations has been docu-
mented in a variety of taxa, including plants (Boyd 2002),
arthropods (Masta 2000; Dyer and Jaenike 2005; Smith and
Farrell 2005), birds (McCormack et al. 2008) and reptiles
(Holycross and Douglas 2007). Consequences of isolation
among sky islands are particularly evident in terrestrial
ectotherms characterized by a narrow range of thermal tol-
erances and limited vagility (e.g. Polato et al. 2018). One

13
Vol.:(0123456789)

such species is the fully terrestrial Sacramento Mountain
salamander (Aneides hardii) that is endemic to three disjunct
mountain ranges in south-central New Mexico including the
Capitan, White, and Sacramento Mountains, located east of
the Madrean Sky Island archipelago. Yet, to our knowledge
only a single study has investigated patterns of genetic diver-
gence in this taxon. This study used allozymes and suggested
relatively low divergence among mountains top populations
of A. hardii (Pope and Highton 1980). Genetic isolation of
sky island populations raises management concerns that
mirror those observed in anthropogenically fragmented
populations including erosion of standing genetic diversity,
reduced adaptive potential, and elevated risk of inbreeding
depression (Frankham et al. 2009). However, in contrast to
recently fragmented populations, sky island populations
can be isolated over geological time scales and are often
highly divergent and independently evolving entities. Hence,
describing relationships among populations is imperative,
particularly as A. hardii is a species of conservation concern
in New Mexico and populations represent the basic units for
conservation and management (Waples and Gaggiotti 2006).
We sought to determine the degree of population struc-
ture in A. hardii both within and among sky islands. We
hypothesized that the main source of population fragmen-
tation in A. hardii are the low-elevation barriers between
sky island populations that have existed since the onset of
the most recent interglacial period (~ 11.5 kya). Divergence
dates older than this would suggest that gene flow was also
restricted during glacial periods; when dispersal corridors
presumably existed. We also described genetic diversity
within populations with respect to their demographic his-
tory. Finally, we attempted to identify if populations of A.
hardii fit criteria for recognizing evolutionary significant
units (ESUs; i.e., reciprocal monophyly in mitochondrial
gene tree and significant divergence among nuclear genes;
Mortiz 1994) and/or management units (i.e., divergent
mtDNA/nucDNA and demographic independence; Mortiz
1994; Palsboll et al. 2007).
Methods
Study Species—A. hardii is reliant on moist microhabitat
(e.g., coarse woody debris and talus rock deposits) associ-
ated with spruce-fir forests and alpine tundra found at eleva-
tions of 2400–3570 m (Petranka 1998). Similar to other ter-
restrial plethodontids and amphibians in general, the extent
to which A. hardii can disperse is constrained by desiccation
risk, so arid expanses of piñon-juniper woodland separat-
ing mountains are presumably highly effective barriers to
dispersal (Watling and Braga 2015).
Genetic Sampling—Samples (3–5 mm tail tips) were
collected from 212 individuals encompassing the range of
13
Conservation Genetics
A. hardii, with multiple sampling sites per mountain range
(Fig. 1). We isolated DNA using a standard proteinase-K
digestion and phenol/chloroform isolation method (Hillis
et al. 1996). We amplified a 741-bp segment of the mito-
chondrial cytochrome b gene and 602 base pairs of the
ND4 gene using primers developed specifically for this
study using A. hardii sequences available on GenBank
(AY728226.1). Polymerase Chain Reactions (PCRs) com-
prised template DNA, 1X Promega Flexi TAQ reaction
buffer, 2 mM MgCl2, 0.8 mM dNTPs, 0.5 µM of forward
(ND4: Anhard_nd4F 5′GGT​ ATG ​ G AA​ T TA​ T TC ​ G AG​
TAAC and Anhard_nd4R 5′CCT​GAR​ATT​AAC​TCT​GGT​
TTA; Cytochrome b: Anhard cytbF 5′AGT​ACA​CAT​TTG​
CCG​CGA​TG and Anhard_cytb1R 5′ACT​GGT​TGG​CCT​
CCA​ATT​CA) and reverse primer (ND4 or Cytochrome b),
and 0.5 U of TAQ polymerase. For cytochrome b and ND4,
PCR cycling conditions were: 90 °C initial denaturation for
2 minutes (min) followed by 30 cycles of 90 °C for 30 s (s),
60 °C for 30 s, and 72 °C for 40 s, plus a final elongation
step of 72 °C for 15 min. PCR products were purified using
the OMEGA CyclePure Kit. DNA was sequenced using the
Applied Biosystems BigDye Cycle Sequencing Kit (version
1.1) according to the manufacturer’s instructions. PCR prod-
ucts were sequenced, and raw DNA sequence reads were
edited and aligned using the software program Sequencher®
(version 5.4.6). Two tandem gene duplications within the
mitochondrial genome have been documented previously,
one occurring at the base of the Aneides clade and a second
specific to A. hardii (Mueller et al. 2004; Mueller and Boore
2005). Results from recent work indicate that the genomic
region where ND4 resides, and cytochrome b, are both
duplicated in A. hardii (Chong and Mueller 2017). Evalua-
tion of our ND4 sequence data identified widespread, intra-
individual polymorphism, which is a signature consistent
with the gene duplications identified by Chong and Mueller
(2017). For this reason, we only present results based on
cytochrome b, as amino acid translations of these sequences
indicated amplification of the functional copy of the gene.
In addition to mtDNA loci, we sequenced three nuclear
loci (nucDNA) for a subset of individuals from each local-
ity per mountain (n = 5 per sample locality) to maximize
the chance of detecting variation. Nuclear loci sequenced
included recombination activating 2 gene (RAG2; 607 bp),
Sushi Von Willebrand Factor Type A EGF and Pentraxin
Domain Containing 1 gene (SVEP1, 502 bp), and C-X-C
motif chemokine receptor 4 gene (CXCR4; 531 bp) origi-
nally described in Shen et al. (2013). Three other genes
were tested including dispatched RND transporter fam-
ily member 2 (DISP2, 1100 bp), dolichol kinase (DOLK,
672 bp) and FIC domain containing (FICD, 510 bp) genes
(Shen et al. 2013) but these did not amplify well. The
final suite of genes was selected because they had been
shown to have a high substitution rate when tested across
Conservation Genetics
Fig. 1  Digital elevation map depicting the arrangement of the sam-
pled Aneides hardii localities within the Capitan, White and Sacra-
mento Mountains located in Lincoln and Otero Co., New Mexico.
salamander species (Shen et al. 2013), amplified a product
in A. hardii, and captured variation between individuals
in other Aneides species. We attempted to increase primer
specificity by redesigning primer pairs for each locus using
available sequence data from other members of Aneides
including A. hardii via GenBank (Shen et al. 2013, 2015;
Reilly and Wake 2015). For all nucDNA loci, PCR was
conducted at 30 µl total volume and contained 3 µl tem-
plate DNA, 1X Promega Flexi TAQ reaction buffer, 2 mM
­MgCl2, 0.8 mM dNTPs, 0.2 µM of forward primer, 0.2 µM
of reverse primer and 0.225 U of TAQ polymerase. PCR
cycling conditions for RAG2 and CXCR4 consisted of ini-
tial denaturation at 94 °C for 4 min followed by 30 cycles
of 94 °C for 45 s, 40 s at 57.1 °C dropping 0.2 °C per
cycle, and 72 °C for 2 min, followed by a final elongation
step of 72 °C for 10 min. Cycling conditions were identical
for SVEP1 except that annealing began at 59.4 °C drop-
ping 0.1 °C per cycle. PCR products were purified using
the OMEGA CyclePure Kit. DNA was sequenced using
the Applied Biosystems BigDye Cycle Sequencing Kit
(version 1.1) according to the manufacturer’s instructions.
Genetic samples span the entirety of the geographic range as this spe-
cies is endemic to these three mountains located in south-central New
Mexico
Geneious vers 9.1.2 (Biomatters) was used to align and
edit raw DNA sequence reads.
Lastly, we used cross species amplification of micros-
atellites developed for other plethodontid salamanders.
We selected 84 microsatellite loci from the literature rep-
resenting eastern and western members of Plethodon and
Aneides (Connors and Cabe 2003; De Gross 2004; Spatola
et al. 2013; J.J. Apodaca unpublished). A single polymor-
phic microsatellite, Plel111 (DeGross 2004), amplified
consistently. All other primer pairs either did not amplify,
did not amplify consistently, were monomorphic, or did not
appear to amplify a repetitive sequence in A. hardii. We
assayed Plel111 across populations as a preliminary point
of comparison with previously collected allozyme data but
recognize that drawing conclusions from a single locus is
tenuous. Polymerase chain reactions for Plel111 contained
10 µl total volume contained the following: 3 µl template
DNA, 1X Promega FlexiTAQ reaction buffer, 2 mM MgCl2,
0.8 mM dNTPs, 0.5 µM of forward primer labeled (5′ GTA​
TCA​CCC​CAC​TCA​CTT​TGCTA) and reverse primer (5′
GTA​T GT ​ C CA​ C TG ​ C TC ​ GTC ​ T TT ​ C TT ​ ) , and 0.5 U of
13

TAQ polymerase. PCR cycling conditions were: 90 °C ini-
tial denaturation for 2 min followed by 30 cycles of 90 °C for
30 s, 60 °C for 30 s, and 72 °C for 40 s, plus a final elonga-
tion step of 72 °C for 15 min. Fragment size analysis was
conducted on an ABI3130 automated capillary sequencer
by combining 1 µl of PCR product with 10 µl of formamide
and 0.4 µl of HD1000 size standard, which was denatured at
93 °C for 5 min. Genotype data were scored in GENEMAP-
PER vers 4. 0 (Applied Biosystems).
Statistical Analysis—Genetic Diversity
To determine the most appropriate model of DNA sequence
evolution for cytochrome b, we used jModeltest (Posada
2008) to evaluate alternative models using corrected Akaike
information criteria (Hurvich and Tsai 1989). Results from
jModeltest revealed the Kimura two parameter model
(Kimura 1980) as the most appropriate model of sequence
evolution which was implemented in subsequent analyses
when appropriate. We used MEGA (Kumar et al. 2016) to
calculate the mean, uncorrected percent sequence divergence
within and between mountain top populations. We calculated
standard measures of genetic diversity, including haplotype
diversity (h, Nei 1987), number of haplotypes and haplo-
type richness (­ HR) for mitochondrial DNA cytochrome b
sequences with the software program Contrib (Petit et al.
1998). Haplotype richness ­(HR) is calculated using rarefac-
tion to account for differences in sample size among col-
lections. All statistics are reported for each mountain top
population. We also estimated haplotype frequency by col-
lection locality. We used DNAsp vers. 5 (Librado and Rozas
2009) to calculate nucleotide diversity (π, Nei 1987), mean
nucleotide differences between sequences in the sample (k,
Tajima 1983), the number of segregating (polymorphic)
sites (S) and the number of mutations per site (θ, Nei 1987).
MICROSATELLITETOOLKIT (add-in for Microsoft Excel,
written by S. Park, available at http://anima​lgeno​mics.ucd.
ie/sdepa​rk/ms-toolk​it/) was used to check the data for scor-
ing errors and to estimate additional diversity statistics,
including observed heterozygosity (­ Ho), Nei’s unbiased gene
diversity ­(He). GENEPOP (Raymond and Rousett 1995) was
used to test for departures from Hardy–Weinberg equilib-
rium (HWE), using the procedure of Guo and Thompson
(1992). Average inbreeding coefficients (­ FIS) and allelic rich-
ness ­(AR) were obtained using FSTAT vers. 2.9.3.1 (Goudet
1995). The number of private alleles ­(PA, alleles unique to
one population) per mountain range was calculated using
GenAlEx vers 6.5 (Peakall and Smouse 2006).
Population structure
To visualize relationships among A. hardii cytochrome
b sequences, we created haplotype networks using the R
13
Conservation Genetics
package pegas (Paradis 2010; R Core Team 2013). Weir
and Cockerham’s (1984) analysis of molecular variance
(AMOVA) based on haplotype frequencies as implemented
in ARLEQUIN vers. 3.5 (Excoffier and Lischer 2010) was
used to examine the partitioning of genetic variance among
mountain tops (ϕCT), among sites within a mountain top
(ϕSC), and among sites (ϕST). We also calculated pairwise
ϕST among sampling localities. Significance was assessed
by 10,000 bootstrap replicates.
Lineage divergence estimates
We used the software program BEAST vers 2.5.1 (Bouckaert
et al. 2014) to estimate divergence times among three line-
ages of A. hardii: Sacramento, White and Capitan Moun-
tains. We used the rate of substitution for cytochrome b as
estimated by Mueller (2006) to implement the substitution
prior for generating divergence estimates among lineages.
We used a normal distribution for the clock rate with a rate
of 0.62 (SD = 0.16) substitutions per nucleotide site per
100 million years (Mueller 2006) and did not include an
outgroup as suggested by Drummond and Bouckaert (2015).
We employed the Kimura two parameter model of sequence
evolution under a strict clock model and a coalescent con-
stant population prior. Brown and Yang (2011) suggest that
a strict clock is most appropriate for shallow phylogenies
because of minimal rate variation among branches. We
conducted three independent runs of 50 million genera-
tions, sampling every 5000 generations with the first 25%
percent of trees discarded as burn-in. Effective sample size
(ESS) and convergence were checked for each run via Tracer
vers1.7.1 (Rambaut et al. 2018). For each run, no ESS val-
ues were < 200 and all runs appeared to have converged.
Tree files from multiple runs were then concatenated using
LogCombiner and the results were summarized using a
maximum clade credibility tree via the TreeAnnotator util-
ity in BEAST. We used FigTree vers 1.4.4 to visual trees and
display mean node heights as well as 95% highest posterior
densities (HPD). Divergence time estimates were calculated
to provide a framework for understanding the separation of
lineages. However, these estimates should be viewed with
an appropriate level of caution given that they are based on
a single gene tree, which may overestimate the age of line-
age separation.
We estimated a gene tree using MRBAYES vers 3.2.2
with the Kimura two parameter model and implemented in
the CIPRES Portal vers 3.3 cluster at the San Diego Super-
computer Center (Miller et  al. 2010). The analysis was
run for 1 × 107 generations with sampling of the Markov
chain every 100 generations. Of the resulting 100,000 trees,
10,000 were discarded as burn-in. Support for nodes was
determined by posterior probabilities obtained from the
Conservation Genetics

majority-rule consensus tree visualized using program Results

FigTree vers 1.4.2 (Rambaut 2014).
Demographic history
We examined A. hardii sequences for signals of recent
demographic expansion or population bottlenecks using Fu
and Li’s D­ * (Fu and Li 1993) and Tajima’s D (Tajima 1989a,
b) using DNAsp (Librado and Rozas 2009). These statistics
are affected by selection and demographic processes (Tajima
1989a, b; Fu and Li 1993). We also calculated Fu’s Fs, which
compares the probability of the observed number of hap-
lotypes versus the expected number of haplotypes under
neutral conditions. These statistics were calculated for each
mountain top population and significance was assessed using
10,000 coalescent simulations conducted in DNAsp (Lib-
rado and Rozas 2009). We used Arlequin 3.5 (Excoffier and
Lischer 2010) to conduct mismatch analyses (Harpending
and Rogers 2000; Rogers and Harpending 1992). To test for
demographic expansion against a null model of population
stability, we used the raggedness index (­ R2, Ramos-Onsins
and Rozas 2002). A significant R ­ 2 value suggests popula-
tion expansion. We used the sum of squared deviations
between the observed and expected mismatch distributions
to test for the signature of population stability (significant
values indicate stability). Significance of these test statistics
were assessed with 1000 bootstrap replicates in Arlequin
(Excoffier and Lischer 2010). Mismatch analysis also pro-
vides the value tau (τ), which is a moment estimator that
represents a unit of mutational time (Schenekar and Weiss
2011). This can be used to calculate the time since popu-
lation expansion occurred (t) using the equation t = 𝜏∕2u
where u is the cumulative probability of substitutions
(Schenekar and Weiss 2011). We used the excel spreadsheet
calculator (http://www.uni-graz.at/zooww​w/misma​tchca​lc/
index​.php) to estimate the time since expansion using the
sequence divergence rate of 1.24% per million years (Muel-
ler 2006). The excel calculator converts this estimate into the
cumulative number of substitutions per generation (assum-
ing a generation time of 3 years) and provides an estimate
of the time since expansion in years.
Genetic diversity
Cytochrome b data were obtained from 212 A. har-
dii individuals (Genbank Accession numbers:
MH908140–MH908164). The Sacramento Mountains
population had the highest h and ­HR (Table 1). The num-
ber of variable sites (S) was the same in the Capitan and
Sacramento Mountains populations, k and π were higher
in the Capitan population (Table  1) indicative of more
divergent haplotypes. All diversity metrics calculated from
cytochrome b were low in the White Mountains popula-
tion (Table 1). No haplotypes were shared among mountain
ranges (Table 2), and there were substantial haplotype fre-
quency differences among sites within mountains ranges.
West Capitan individuals (n = 6) shared a single unique hap-
lotype (haplotype 8). The southern-most sampled locality
(Timberon) had four haplotypes (haplotypes 7, 14, 15, 16)
that were not detected elsewhere in the Sacramento Moun-
tains (Table 2).
Nuclear sequence data revealed no variation among
mountains for all loci (RAG2 [Genbank Accession:
MH929435], CXCR4 [Genbank Accession: MH932715],
SVEP1 [Genbank Accession: MH929434]; results not
shown). The microsatellite locus conformed to Hardy–Wein-
berg expectations in all populations. Gene diversity and
heterozygosity were high for all populations (Table 1). The
Capitan population had the highest allelic richness and num-
ber of private alleles ­(PA = 15). Three of the private alleles
detected in the Capitan Mountains were unique to West
Capitan. Values of F ­ IS ranged from 0.015 (Sacramento) to
0.120 (White).
Population structure
Haplotype networks and the gene tree revealed groups of
haplotypes that were unique to each mountain range (Fig. 2).
There was strong statistical support for these groupings
(posterior probabilities = 1; Fig. 2b). P-distance was 2.8%

Table 1  Genetic diversity Population n Cytochrome b Plel111

statistics for the mitochondrial
cytochrome b gene and the Nhaps HR h π k S He Ho AR FIS PA
microsatellite Plel111 
Capitan 36 8 7.000 0.740 0.003 2.284 10 0.983 0.921 41.000 0.064 15
White 68 4 2.800 0.632 0.002 1.162 3 0.958 0.851 29.420 0.120 7
Sacramento 108 13 8.600 0.796 0.002 1.474 9 0.953 0.939 26.310 0.015 3

The following metrics are presented: sample size (n), number of haplotypes (­Nhaps), haplotype richness
­(HR), haplotype diversity (h), nucleotide diversity (π), mean nucleotide differences between sequences in
the sample (k) and number of segregating sites (S). For the microsatellite gene diversity ­(He), observed
heterozygosity ­(Ho), allelic richness ­(AR), inbreeding co-efficient (­ FIS) and number of private alleles ­(PA)

13
 Conservation Genetics

Table 2  Sample size (n) and haplotype frequencies by collection locality and mountain range
Haplotype Capitan White Sacramento
West East Big Bear Ski Apache Observatory Rio Peñasco Rd Russia Canyon Timberon
n = 6 n = 30 n = 25 n = 43 n = 29 n = 26 n = 28 n = 25

1 – – – – – 0.577 – –
2 – – – – – 0.115 – –
3 – – – – 0.034 – 0.357 –
4 – 0.267 – – – – – –
5 – – 0.400 – – – – –
6 – – 0.600 0.442 – – – –
7 – – – – – – – 0.040
8 1.000 – – – – – – –
9 – 0.033 – – – – – –
10 – 0.033 – – – – – –
11 – – – 0.512 – – – –
13 – – – – 0.828 – 0.500 0.200
14 – – – – – – – 0.200
15 – – – – – – – 0.520
16 – – – – – – – 0.040
17 – – – 0.047 – – – –
18 – – – – 0.034 0.192 – –
19 – – – – – 0.115 0.036 –
20 – – – – 0.103 – – –
21 – – – – – – 0.071 –
22 – – – – – – 0.036 –
23 – 0.533 – – – – – –
24 – 0.067 – – – – – –
25 – 0.067 – – – – – –
26 – 0.033 – – – – – –

between Capitan and White, 2.5% between Capitan and Sac-
ramento, and finally, 2.3% between White and Sacramento
Mountain populations. AMOVA analysis indicated that a
significant proportion of variance could be attributed to dif-
ferences between mountain tops (ϕCT = 0.140, p = 0.02), as
well as to differences between sampling localities within
each mountain top (ϕSC = 0.370, p = 0.0001). Pairwise ϕST
values calculated between sampling localities were all highly
significant (i.e., p < 0.0001, Table 3).
Divergence time estimates and demographic history
~ 162,000–198,000 years ago for the Sacramento and White
Mountains populations to ~ 516,000 years ago for Capitan
Mountains. Fu and Li’s D* and Tajima’s D did not differ
significantly from zero for any of the mountain top popula-
tions (Table 4). Fu’s Fs was significantly negative in the
Sacramento Mountains population indicative of historical
population expansion (Table 4), however, we found little
evidence for population expansion via the 𝜏 and R ­ 2 statis-
tics (Table 5). Moreover, comparisons of observed mismatch
distributions (Supplementary Material) to those expected
under sudden demographic expansion were not significant.

Estimated times of lineage divergence between mountain top

populations date to the early Pleistocene, with the Capitan
lineage diverging first, followed by White and Sacramento
Mountains populations (Fig. 2b). It is important to note
that these estimates should be viewed with the caveat that
they are based on a single gene and are largely dependent
on the rate of sequence evolution used in their calculation.
The estimated time since population expansion ranged from
Discussion
Mitochondrial DNA sequence data support the presence
of three genetically distinct lineages restricted to each of
the mountain ranges occupied by A. hardii; the Capitan,
White, and Sacramento Mountains. Estimates of diver-
gence times among lineages suggest separation since the

13
Conservation Genetics

Fig. 2  a Haplotype network depicting the relationship among each
mountaintop lineage (Green-Capitan Mountain, Red-White Moun-
tains, Blue-Sacramento Mountains), where different shades cor-
respond to sampling localities within each mountain range. Nucleo-
tide substitutions are represented by dots. Numbers within the large
circles refer to the haplotype number and the size of each circle
reflects haplotype frequency. b Divergence time (in millions of years)
between mitochondrial lineages with 95% highest probability den-
sities shown below the branch nodes. Haplotype numbers present
within each mountain lineage are also included. (Color figure online)

Table 3  Pairwise ϕST among collection sites calculated based on haplotype frequencies

East Capitan West Capitan Big Bear Ski Apache Observatory Rio Peñasco Russia Canyon Timberon

East Capitan –
West Capitan 0. 528 –
Big Bear 0. 413 0. 634 –
Ski Apache 0.422 0. 610 0. 300 –
Observatory 0. 513 0. 756 0. 598 0. 578 –
Rio Peñasco 0. 357 0. 552 0. 436 0. 438 0. 532 –
Russia Canyon 0. 352 0. 543 0. 429 0. 432 0. 173 0. 365 –
Timberon 0. 335 0. 526 0. 413 0. 418 0. 417 0. 350 0. 272 –

All values were highly significant (p = 0.00001)

Table 4  Fu and Li’s D*, Population Fu and Li’s D* p Value Tajima’s D p Value Fu’s Fs p Value
Tajima’s D, and Fu’s Fs and
associated p-values Capitan − 0.294 0.371 − 0.161 0.494 − 0.499 0.450
White 0.861 0.746 1.668 0.944 1.603 0.816
Sacramento − 0.206 0.410 − 0.334 0.420 − 4.626 0.037

early Pleistocene. These data do not indicate secondary
contact between populations during recent glacial peri-
ods. Isolation is likely a result of high site fidelity and
the tendency for plethodontids to be sedentary (Marvin
2001). Landscape features could also serve as barriers to
impede secondary contact. For example, first and second
order streams have been documented to influence patterns
of gene flow among populations of Eastern Red-backed
Salamanders (Marsh et al. 2007).

13

Table 5  Population expansion statistics and associated p-values cal-
culated from cytochrome b among mountain ranges
Sudden demographic Spatial expansion
expansion
τ R2 SSD R2
Capitan 4.742 0.100 0.041 0.100
p Value 0.325 0.331 0.585
White 2.156 0.115 0.019 0.115
p Value 0.321 0.348 0.593
Sacramento 1.474 0.081 0.012 0.061
p Value 0.425 0.101 0.052
Statistics include: Tau (τ), raggedness statistic (­R2) where significant
values are indicative of population expansion, and sum of squared
deviations (SSD) between the observed mismatch distribution and the
expected distribution under a sudden demographic expansion model.
Significant results support population stability
Genetic diversity differed within major mitochondrial lin-
eages of A. hardii, with high diversity and deeper divergence
among haplotypes restricted to the Capitan Mountains, very
low diversity in the White Mountains, and high diversity
with low divergence among haplotypes in the Sacramento
Mountains lineage. These results suggest different demo-
graphic histories within each lineage. In contrast to mtDNA,
results from four nuclear loci investigated revealed all popu-
lations to be monomorphic with the exception of a single
microsatellite locus. The lack of nuclear variation may be a
result of low mutation rate rather than the presence of gene
flow. Pope and Highton (1980) also found that the majority
of allozyme loci (16 of 21 assessed) were monomorphic
across populations of A. hardii.
Broad scale fragmentation
Our results support the recognition of three demographically
and genetically distinct management units corresponding to
each of the mountains occupied by A. hardii. This is consist-
ent with the idea that sky islands provided isolated refugia
during the warmer, interglacial periods. The degree of diver-
gence and timing of isolation differs substantially among sky
island taxa; presumably a result of varied life-histories and
landscape features. For example, Mexican jays have high
site fidelity like A. hardii, and divergence predates extreme
glacial–interglacial cycles which have occurred over the past
700,000 years (McCormack et al. 2008). In contrast, diver-
gence times among New Mexico Ridge-nosed Rattlesnake
(Crotalus willardi obscurus) populations were much more
recent (i.e., Holocene) consistent with more recent gene
flow. In this case, the degree of genetic divergence observed
among sky islands warranted recognition of multiple ESUs
(Holycross and Douglas 2007). Sequence divergence in the
Striped Plateau Lizard (Sceloporus virgatus) also reflects the
13
Conservation Genetics
isolation of sky island populations where divergence was ten
times greater among sky islands compared to within island
divergence even though geographic distances between sites
within an island were often greater than distances separating
sky islands (Tennessen and Zamudio 2008).
Patterns of broad-scale fragmentation observed in A. har-
dii contrast those of other montane plethodontids. For exam-
ple, Shepard and Burbrink (2009) reported significant diver-
gence among montane populations of Plethodon fourchensis,
with lineage divergence dating to the middle Pleistocene
and likely caused by fragmentation of a widely-distributed
ancestor. Their results suggested that P. fourchensis popula-
tions expanded during interglacial periods coinciding with
the expansion of deciduous forests and contracted during
glacial periods. This finding contrasts with environmental
conditions of Madrean Sky Islands where range contrac-
tions, and hence fragmentation of montane species, are
associated with interglacial periods. During these periods
pine-oak forests contracted thereby isolating populations
on adjacent mountain tops (Smith and Farrell 2005; Masta
2000). We did not find evidence (e.g. shared haplotypes
between mountain range populations) that populations
of A. hardii came into contact during more recent glacial
periods when coniferous forests dominated the landscape
(Davis 1983). Thus, although movement may have extended
downslope during cooler periods, mountain-top popula-
tions remained isolated from one another; possibly a result
of other landscape features (rivers and streams) or limited
dispersal ability of A. hardii.
Divergence times estimated here based on mitochondrial
data, are older than those obtained from allozymes (Pope and
Highton 1980). The order of divergence also differs between
studies. Allozyme data suggests a more recent connection
between Capitan and White mountain potentially explained
by closer geographic distance between these populations.
In contrast, mitochondrial data indicates that the Capitan
lineage is the most divergent, while the Sacramento Moun-
tains lineage is most closely related to the White Mountains.
Greater sampling depth and more variable mitochondrial
markers may explain the discrepancy between studies. Addi-
tional data from microsatellites and nuclear gene sequences
is needed to further explore both contemporary population
dynamics and evolutionary history of A. hardii.
Inter and intrapopulation variation
Divergence time estimates suggest that the Capitan popula-
tion is the oldest of the three mitochondrial lineages of A.
hardii, which is consistent with the presence of more diver-
gent haplotypes. Within the Capitan Mountains, the presence
of private alleles/haplotypes between East and West moun-
tain suggests a distinct break between these populations.
This is not surprising as West mountain is separated from
Conservation Genetics
the rest of the range by the low elevation Capitan Pass. Inter-
estingly, the habitat occupied by A. hardii differs between
the East and West mountain populations. Specifically, the
East mountain samples were collected in the vicinity of a
talus slope and most often from under rocks; habitat that was
absent at the West Capitan site. Differences in the quality
of microhabitat available within the West and East Capitan
sites may explain differences in local densities. Additionally,
the few samples that were obtained from the West Capitan
site were monomorphic, suggesting reduced diversity and
possibly low effective population size in this population.
Further sampling in this area is warranted to monitor the
status of the population and to further characterize diver-
sity. The East Capitan site contrasts the West Capitan site in
that the talus slope and surrounding forest harbored seven
divergent haplotypes. Additional surveys on East mountain
may also be warranted to further characterize diversity at
other locations, if additional A. hardii habitat and survey
sites can be identified. In contrast to our results, Pope and
Highton (1980) found the lowest heterozygosity in the Capi-
tan population with three of the five loci fixed for a single
variant. However, their sample size was also the lowest for
this collection and they only assessed variation in the East
Capitan population.
The White Mountain population exhibited low levels of
mitochondrial diversity and highest inbreeding coefficient.
Reduced mitochondrial diversity within the White Mountain
population is somewhat surprising given that the highest
local density of A. hardii was observed at the Ski Apache
site (n = 43). Tajima (1989b) found that the number of vari-
able sites is influenced more strongly by the current size of
the population, than is sequence diversity. Reduced genetic
diversity and relatively high local abundance in this popu-
lation may reflect a recent population bottleneck and sub-
sequent population recovery. High-severity wildfires like
the Little Bear Fire that impacted White Mountain in 2012,
affected salamander habitat and perhaps reduced salamander
abundances in the area. Allozyme data showed similar levels
of diversity between the White and Sacramento populations,
suggesting that the decline in diversity that was observed in
the White Mountain population has occurred very recently.
In contrast to the Capitan and White populations, the
Sacramento population exhibits a star-like arrangement
of haplotypes consistent with an expanding population of
more recent origin. Reduced genetic diversity in expand-
ing lineages has been documented in many taxa and is due
to the small number of founding organisms comprising
only a subset of the haplotype variation found throughout
the ancestral range (Templeton 1998). Moreover, rapidly
expanding founding populations exhibit high gene diversity,
but low nucleotide diversity, as was observed in the Sac-
ramento population of A. hardii (Grant and Bowen 1998).
Nucleotide diversity is affected by the size of the ancestral
population more severely than the number of variable sites
(Tajima 1989). Several other statistics (Fu’s Fs and ­R2) also
support historical expansion of the Sacramento population.
Higher haplotype diversity in the Sacramento population
is also consistent with the apparently higher density (and
presumably higher genetic effective population size) of A.
hardii in the Sacramento Mountains. Indications of haplo-
type frequency differences among sampling sites within the
Sacramento Mountains suggests that further examination
with microsatellites may reveal finer scale structure.
Management implications
Phylogenetic information can identify genetically distinct
units, which is critical for informed conservation decision-
making. Based on the mtDNA data presented here, we
recommend that A. hardii be recognized as three manage-
ment units corresponding to each of the inhabited mountain
ranges: Capitan, White, and Sacramento. This recommen-
dation is based on following lines of evidence including:
(1) substantial divergence and reciprocal monophyly of a
mitochondrial gene, which suggests isolation of lineages
for 1–2 million years with no evidence of gene flow; (2) a
disjunct distribution, which means that population dynam-
ics within each unit are completely independent of those in
other units. The degree of mitochondrial divergence between
populations of A. hardii is similar to that observed between
populations of other members of Aneides that have been des-
ignated as ESUs (Reilly et al. 2012). While the nuclear data
set failed to meet criteria for recognition of ESUs, allozyme
data suggest differentiation across mountains in the nuclear
genome (Pope and Highton 1980). Given the degree of geo-
graphic isolation among populations and the limited disper-
sal ability of A. hardii, the lack of nuclear divergence we
observed is likely a function of a low mutation rate rather
than the presence of gene flow among mountains. Purifying
selection can also result in reduced polymorphism in por-
tions of the genome (e.g. Palmer et al. 2010). Lack of recent
gene flow between mountain ranges is also supported by the
presence of microsatellite alleles unique to each mountain
(Capitan, White, and Sacramento). Investigating patterns
of differentiation with a different suite of nuclear markers
(e.g., microsatellites or SNPs) may be more reflective of
the degree of population connectivity as well as reveal fine-
scale population structuring within mountains. Nonetheless,
some discordance has been documented between nuclear and
mitochondrial markers on the location of breaks between
populations of other members of the Aneides clade. In
black salamanders (Aneides flavipunctatus), data suggested
that the mitochondrial boundary between two populations
remained stable despite nuclear gene flow from south to
north. This pattern would occur if there was male-biased
northward dispersal which would carry new nuclear alleles
13

into a region but not new mitochondrial haplotypes (Reilly
et al. 2012). Although the ecology of A. hardii remains
understudied, female philopatry is well documented in both
western and eastern members of Plethodontidae (e.g., Ensat-
ina sp., Staub et al. 1995; Plethodon cinereus; Liebgold et al.
2011). It may also be prudent to conduct a geometric mor-
phometric analysis of museum specimens collected from
across the range of A. hardii to determine if there are any
morphological differences, especially in cranial morphol-
ogy, which can closely track differences in diet (e.g., Maerz
et al. 2006). However, it has also been shown that in pletho-
dontid salamanders, there can be extreme morphological
stasis, such that the same morphology has been maintained
over long periods of time encompassing dramatic climatic
change (e.g., Wake 1983). Larson (1984, 1989) also demon-
strated that in numerous salamanders, speciation had been
decoupled from morphological evolution, specifically there
have been more speciation events than morphological inno-
vations (Highton 1990). Hence, absence of morphological
divergence would not negate the recommendation to treat A.
hardii as three distinct ESUs.
Acknowledgements  We gratefully acknowledge the Share with Wild-
life program at New Mexico Department of Game and Fish (NMDGF)
and State Wildlife Grant T-32-4, 8 for providing funding to complete
this project. We also sincerely thank L. Cordova and J. Williams (and
their interns) from the U.S. Forest Service for field assistance. A.
Sanchez, T. Pilger, D. Camak, T. Giermakowski, and the Museum of
Southwestern Biology Division of Herpetology (UNM) and V. Seam-
ster and L. Pierce (NMDGF) are gratefully acknowledged. The Chir-
icahua Desert Museum provided financial support. Salamanders were
collected under NMGF permit #3598 and UNM’s IACUC protocol
number #13-1000983-MC.
References
Bouckaert R, Heled J, Kühnert D, Vaughan T, Wu C-H, Xie D, Suchard
MA, Rambaut A, Drummond AJ (2014) BEAST 2: a software
platform for bayesian evolutionary analysis. PLOS Comput Biol
10:e1003537
Boyd A (2002) Morphological analysis of Sky Island populations of
Macromeria viridiflora (Boraginaceae). Syst Bot 27:116–126
Brown RP, Yang Z (2011) Rate variation and estimation of divergence
times using strict and relaxed clocks. BMC Evol Biol 11(1):271
Chong RA, Mueller RL (2017) Polymorphic duplicate genes and per-
sistent non-coding sequences reveal heterogeneous patterns of
mitochondrial DNA loss in salamanders. BMC Genom 18:992
Connors LM, Cabe PR (2003) Isolation of dinucleotide microsatellite
loci from red-backed salamander (Plethodon cinereus). Mol Ecol
Notes 3:131–133
Davis MB (1983) Quaternary history of deciduous forests of eastern
North America and Europe. Ann Mo Bot Gard 70:550–563
DeGross DJ (2004) Gene flow and the relationship of Plethodon stormi
and P. elongatus assessed with 11 novel microsatellite loci. M.Sc.
thesis, Oregon State University
Drummond AJ, Bouckaert RR (2015) Bayesian evolutionary analysis
with BEAST. Cambridge University Press, Cambridge
13
Conservation Genetics
Dyer KA, Jaenike J (2005) Evolutionary dynamics of a spatially
structured host-parasite association: Drosophila innubila and
male-killing Wolbachia. Evol 59:1518–1528
Excoffier L, Lischer HE (2010) Arlequin Suite Ver 3.5: a new series
of programs to perform population genetics analyses under
Linux and Windows. Mol Ecol Resour 10:564–567
Frankham R, Ballou JD, Briscoe DA (2009) Introduction to con-
servation genetics, 2nd  edn. Cambridge University Press,
Cambridge
Fu YX, Li WH (1993) Statistical tests of neutrality of mutations.
Genetics 133(3):693–709
Grant WAS, Bowen BW (1998) Shallow population histories in deep
evolutionary lineages of marine fishes: insights from sardines and
anchovies and lessons for conservation. J Hered 89:415–426
Guo SW, Thompson EA (1992) Performing the exact test of
Hardy-Weinberg proportion for multiple alleles. Biometrics
48(2):361–372
Harpending H, Rogers A (2000) Genetic perspectives on human origins
and differentiation. Annu Rev Genom Hum Genet 1:361–385
Highton R (1990) Taxonomic treatment of genetically differentiated
populations. Herpetologica 46:114–121
Hillis DM, Mable BK, Larson A, Davis SK, EA Zimmer (eds) (1996)
Nucleic acids IV: sequencing and cloning. In: Molecular systemat-
ics. Sinaeur Associates, Sunderland
Holycross AT, Douglas ME (2007) Geographic isolation, genetic
divergence, and ecological non-exchangeability define ESUs in
a threatened sky-island rattlesnake. Biol Conserv 134:142–154
Hurvich CM, Tsai CL (1989) Regression and time series model selec-
tion in small samples. Biometrika 76(2):297–307
Kimura M (1980) A simple method for estimating evolutionary rates
of base substitutions through comparative studies of nucleotide
sequences. J Mol Evol 16(2):111–120
Kumar S, Stecher G, Tamura K (2016) MEGA7: molecular evolution-
ary genetics analysis version 7.0 for bigger datasets. Mol Biol
Evol 33:1870–1874
La Sorte FA, Jetz W (2010) Projected range contractions of mon-
tane biodiversity under global warming. Proc R Soc Lond B
277:3401–3410
Larson A (1984) Neontological inferences of evolutionary pattern
and process in the salamander family Plethodontidae. Evol Biol
17:119–217
Larson A (1989) The relationship between speciation and morpho-
logical evolution. In: Otte D, Endler JA (eds) Speciation and its
consequences. Sinauer Associates, Sunderland, pp 579–598
Librado P, Rozas J (2009) DnaSP v5: a software for comprehen-
sive analysis of DNA polymorphism data. Bioinformatics
25:1451–1452
Liebgold EB, Brodie ED, Cabe PR (2011) Female philopatry and male-
biased dispersal in a direct-developing salamander, Plethodon
cinereus. Mol Ecol 20:249–257
Lyons MP, Shepard DB, Kozak KH (2016) Determinants of range lim-
its in montane woodland salamanders (Genus Plethodon). Copeia
104:101–110
Maerz JC, Myers EM, Adams DC (2006) Trophic polymorphism in a
terrestrial salamander. Evol Ecol Res 8:23
Marsh DM, Page RB, Hanlon TJ, Bareke H, Corritone R, Jetter N,
Beckman NG, Gardner K, Seifert DE, Cabe PR (2007) Ecologi-
cal and genetic evidence that low-order streams inhibit dispersal
by red-backed salamanders (Plethodon cinereus). Can J Zool
85:319–327
Marvin GA (2001) Age, growth, and long-term site fidelity in the
terrestrial plethodontid salamander Plethodon kentucki. Copeia
2001:108–117
Masta SE (2000) Phylogeography of the jumping spider Habronat-
tus pugillis (Araneae: Salticidae): recent vicariance of sky island
populations? Evolution 54:1699–1711
Conservation Genetics
McCormack JE, Peterson AT, Bonaccorso E, Smith TB (2008) Spe-
ciation in the highlands of Mexico: genetic and phenotypic
divergence in the Mexican jay (Aphelocoma ultramarina). Mol
Ecol 17:2505–2521
McCormack JE, Huang H, Knowles LL, Gillespie R, Clague D
(2009) Sky islands. Encycl Isl 4:841–843
Miller MA, Pfeiffer W, Schwartz T (2010) Creating the CIPRES Sci-
ence Gateway for inference of large phylogenetic trees. In: Gate-
way Computing Environments Workshop (GCE), Ieee, pp 1–8
Moritz C (1994) Defining ‘evolutionarily significant units’ for con-
servation. Trends Ecol Evol 9:373–375
Moritz C, Patton JL, Conroy CJ, Parra JL, White GC, Beissinger
SR (2008) Impact of a century of climate change on small-
mammal communities in Yosemite National Park, USA. Science
322:261–264
Mueller RL (2006) Evolutionary rates, divergence dates, and the
performance of mitochondrial genes in Bayesian phylogenetic
analysis. Syst Biol 55(2):289–300
Mueller RL, Boore JL (2005) Molecular mechanisms of extensive
mitochondrial gene rearrangement in plethodontid salamanders.
Mol Biol Evol 22:2104–2112
Mueller RL, Macey JR, Jaekel M, Wake DB, Boore JL (2004)
Morphological homoplasy, life history evolution, and histori-
cal biogeography of plethodontid salamanders inferred from
complete mitochondrial genomes. Proc Natl Acad Sci USA
101:13820–13825
Nei M (1987) Molecular evolutionary genetics. Columbia University
Press, New York
Palmer CA, Watts RA, Hastings AP, Houck LD, Arnold SJ (2010)
Rapid evolution of plethodontid modulating factor, a hypervari-
able salamander courtship pheromone, is driven by positive selec-
tion. J Mol Evol 70(5):427–440
Palsbøll PJ, Berube M, Allendorf FW (2007) Identification of man-
agement units using population genetic data. Trends Ecol Evol
22:11–16
Paradis E (2010) pegas: An R package for population genetics with an
integrated–modular approach. Bioinformatics 26:419–420
Parmesan C (2006) Ecological and evolutionary responses to recent
climate change. Annu Rev Ecol Evol Syst 37:637–669
Peakall ROD, Smouse PE (2006) GENALEX 6: genetic analysis in
Excel. Population genetic software for teaching and research. Mol
Ecol notes 6(1):288–295
Petit RJ, Mousadik AE, Pons O (1998) Identifying populations for
conservation on the basis of genetic markers. Conserv Biol
12:844–855
Petranka JW (1998) Salamanders of the United States and Canada.
Smithsonian Institution Press, Washington, DC
Polato NR, Gill BA, Shah AA, Gray MM, Casner KL, Barthelet
A, … Zamudio KR (2018) Narrow thermal tolerance and low
dispersal drive higher speciation in tropical mountains. PNAS
115:12471–12476
Pope MH, Highton R (1980) Geographic genetic variation in the
Sacramento Mountain salamander, Aneides hardii. J Herpetol
14:343–346
Posada D (2008) jModelTest: phylogenetic model averaging. Mol Biol
Evol 25:1253–1256
R Core Team (2013) R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Aus-
tria. http://www.R-proje​ct.org/
Rambaut A (2014) FigTree 1.4. 2 software. Institute of Evolutionary
Biology, Univ. Edinburgh
Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA (2018) Pos-
terior summarisation in Bayesian phylogenetics using Tracer 1.7.
Syst Biol 10:901–904
Ramos-Onsins SE, Rozas J (2002) Statistical properties of new neutral-
ity tests against population growth. Mol Biol Evol 19:2092–2100
Raymond M, Rousset F (1995) GENEPOP (version 1.2): population
genetics software for exact tests and ecumenicism. J Heredity
86:248–249
Reilly SB, Wake DB (2015) Cryptic diversity and biogeographical
patterns within the black salamander (Aneides flavipunctatus)
complex. J Biogeogr 42:280–291
Reilly SB, Marks SB, Jennings WB (2012) Defining evolutionary
boundaries across parapatric ecomorphs of Black Salamanders
(Aneides flavipunctatus) with conservation implications. Mol Ecol
21:5745–5761
Rogers AR, Harpending H (1992) Population growth makes waves in
the distribution of pairwise genetic differences. Mol Biol Evol
9:552–569
Schenekar T, Weiss S (2011) High rate of calculation errors in mis-
match distribution analysis results in numerous false inferences
of biological importance. Heredity 107:511–512
Sekercioglu CH, Schneider SH, Fay JP, Loarie SR (2008) Climate
change, elevational range shifts, and bird extinctions. Conserv
Biol 22:140–150
Shen XX, Liang D, Feng YJ, Chen MY, Zhang P (2013) A versatile and
highly efficient toolkit including 102 nuclear markers for verte-
brate phylogenomics, tested by resolving the higher-level relation-
ships of the Caudata. Mol Biol Evol 30:2235–2248
Shen XX, Liang D, Chen MY, Mao RL, Wake DB, Zhang P (2015)
Enlarged multilocus data set provides surprisingly younger time
of origin for the Plethodontidae, the largest family of salamanders.
Syst Biol 65:66–81
Shepard DB, Burbrink FT (2009) Phylogeographic and demographic
effects of Pleistocene climatic fluctuations in a montane salaman-
der, Plethodon fourchensis. Mol Ecol 18:2243–2262
Smith CI, Farrell BD (2005) Phylogeography of the longhorn cactus
beetle Moneilema appressum LeConte (Coleoptera: Cerambyci-
dae): was the differentiation of the Madrean sky islands driven by
Pleistocene climate changes? Mol Ecol 14:3049–3065
Spatola BN, Peterman WE, Stephens NT, Connette GM, Shepard DB,
Kozak KH et al (2013) Development of microsatellite loci for
the western slimy salamander (Plethodon albagula) using 454
sequencing. Conserv Genet Resour 5:267–270
Staub NL, Brown CW, Wake DB (1995) Patterns of growth and move-
ments in a population of Ensatina eschscholtzii platensis (Cau-
data: Plethodontidae) in the Sierra Nevada, California. J Herpetol
29:593–599
Tajima F (1983) Evolutionary relationship of DNA sequences in finite
populations. Genetics 105:437–460
Tajima F (1989a) Statistical method for testing the neutral mutation
hypothesis by DNA polymorphism. Genetics 123:585–595
Tajima F (1989b) The effect of change in population size on DNA
polymorphism. Genetics 123:597–601
Templeton AR (1998) Nested clade analyses of phylogeographic data:
testing hypotheses about gene flow and population history. Mol
Ecol 7:381–397
Tennessen JA, Zamudio KR (2008) Genetic differentiation among
mountain island populations of the striped plateau lizard, Scelopo-
rus virgatus (Squamata: Phrynosomatidae). Copeia 2008:558–564
Thompson RS, Anderson KH (2000) Biomes of western North Amer-
ica at 18,000, 6000 and 0 14C yr BP reconstructed from pollen
and packrat midden data. J Biogeogr 27:555–584
Wake DB, Roth G, Wake MH (1983) On the problem of stasis in organ-
ismal evolution. J Theor Biol 101:211–224
Waples RS, Gaggiotti O (2006) INVITED REVIEW: what is a popula-
tion? An empirical evaluation of some genetic methods for identi-
fying the number of gene pools and their degree of connectivity.
Mol Ecol 15:1419–1439
Warshall P (1995) The Madrean sky island archipelago: a plan-
etary overview. Biodiversity and management of the Madrean
Archipelago: the sky islands of southwestern United States and
13
 Conservation Genetics

Northwestern Mexico. In: DeBano LF (ed) Biodiversity and the Wilson RJ, Gutierrez D, Gutierrez J, Monserrat VJ (2007) An eleva-
management of the Madrean Archipelago: The sky islands of tional shift in butterfly species richness and composition accom-
southwestern United States and Northwestern Mexico. DIANE panying recent climate change. Glob Change Biol 13:1873–1887
Publishing, Darby, pp 6–18
Watling JI, Braga L (2015) Desiccation resistance explains amphibian Publisher’s Note Springer Nature remains neutral with regard to
distributions in a fragmented tropical forest landscape. Landsc jurisdictional claims in published maps and institutional affiliations.
Ecol 30:1449–1459
Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis
of population structure. Evolution 38:1358–1370

13