Genetic diversity and divergence in the fountain darter (Etheostoma fonticola): implications for conservation of an endangered species

Abstract

The endangered fountain darter Etheostoma fonticola is found only in the Comal and San Marcos rivers in the Guadalupe River basin in central Texas, USA. Comal River fountain darters were believed to be extirpated following a severe drought in the 1950s and were reintroduced in the early 1970s using 457 darters from the San Marcos River. In this study we used 23 microsatellite loci to describe and evaluate the genetic diversity, population structure and effective population size (N
e) of fountain darters. We also evaluated the genetic effect of the Comal River reintroduction and the influence of low-head dams (dams) on dispersal in both rivers. Bayesian analysis of individual genotypes and Analysis of Molecular Variation supported two distinct populations concordant with the two rivers. Estimates of N
e were much smaller (<10 %) than census size (N
c) in both rivers but did not indicate the populations are at risk of an immediate and rapid loss of genetic diversity. Coalescent-based estimates of the genetically effective number of founders (Nf) for the Comal River averaged about 49 darters and, together with the indices of genetic diversity and the bottleneck test (heterozygosity excess) results, were consistent with a founder event following the reintroduction in the Comal River. Finally, our results regarding the influence of dams on fountain darter dispersal were equivocal and did not support a conclusion. We recommend this issue be examined further as part of the fountain darter recovery program.

Full text

1 23
Conservation Genetics
ISSN 1566-0621
Conserv Genet
DOI 10.1007/s10592-016-0869-7
Genetic diversity and divergence in the
fountain darter (Etheostoma fonticola):
implications for conservation of an
endangered species
Jeffrey B.Olsen, Andrew P.Kinziger,
John K.Wenburg, Cara J.Lewis,
Catherine T.Phillips & Kenneth
G.Ostrand

1 23
Your article is protected by copyright and
all rights are held exclusively by Springer
Science+Business Media Dordrecht (outside
the USA). This e-offprint is for personal
use only and shall not be self-archived
in electronic repositories. If you wish to
self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.

RESEARCH ARTICLE
Genetic diversity and divergence in the fountain darter
(Etheostoma fonticola): implications for conservation
of an endangered species
Jeffrey B. Olsen
1
•Andrew P. Kinziger
2
•John K. Wenburg
1
•Cara J. Lewis
1
•
Catherine T. Phillips
3
•Kenneth G. Ostrand
4
Received: 25 February 2016 / Accepted: 12 July 2016
ÓSpringer Science+Business Media Dordrecht (outside the USA) 2016
Abstract The endangered fountain darter Etheostoma
fonticola is found only in the Comal and San Marcos rivers
in the Guadalupe River basin in central Texas, USA.
Comal River fountain darters were believed to be extir-
pated following a severe drought in the 1950s and were
reintroduced in the early 1970s using 457 darters from the
San Marcos River. In this study we used 23 microsatellite
loci to describe and evaluate the genetic diversity, popu-
lation structure and effective population size (N
e
)of
fountain darters. We also evaluated the genetic effect of the
Comal River reintroduction and the influence of low-head
dams (dams) on dispersal in both rivers. Bayesian analysis
of individual genotypes and Analysis of Molecular Varia-
tion supported two distinct populations concordant with the
two rivers. Estimates of N
e
were much smaller (\10 %)
than census size (N
c
) in both rivers but did not indicate the
populations are at risk of an immediate and rapid loss of
genetic diversity. Coalescent-based estimates of the
genetically effective number of founders (Nf) for the
Comal River averaged about 49 darters and, together with
the indices of genetic diversity and the bottleneck test
(heterozygosity excess) results, were consistent with a
founder event following the reintroduction in the Comal
River. Finally, our results regarding the influence of dams
on fountain darter dispersal were equivocal and did not
support a conclusion. We recommend this issue be exam-
ined further as part of the fountain darter recovery
program.
Keywords Fountain darter Genetic diversity 
Endangered species Reintroduction Effective founder
number
Introduction
The fountain darter Etheostoma fonticola is a small, spring-
endemic percid restricted to the upper San Marcos and
Comal rivers in the Guadalupe River basin in central
Texas, USA. The species is listed as endangered under the
U.S. Endangered Species Act (ESA 1973, as amended) and
captive refuge populations from each river have been
established at the San Marcos Aquatic Resources Center
(USFWS 1996). The fountain darter recovery plan calls for
conducting biological studies necessary for successful
monitoring, management and restoration (USFWS 1996).
Among the studies needed to inform the recovery effort is a
description and evaluation of genetic diversity in both the
San Marcos and Comal rivers. In this context there is a
need to develop a genetic baseline for on-going monitoring
of the wild populations and for management of the captive
refuge populations (USFWS 1996). There is also a need to
evaluate restoration actions and other factors that may
influence genetic diversity in fountain darters such as a past
Electronic supplementary material The online version of this
article (doi:10.1007/s10592-016-0869-7) contains supplementary
material, which is available to authorized users.
&Jeffrey B. Olsen
jeffrey_olsen@fws.gov
1
Conservation Genetics Laboratory, U.S. Fish and Wildlife
Service, MS331, Anchorage, AK 99503, USA
2
Department of Fisheries Biology, Humboldt State University,
One Harpst Street, Arcata, CA 95521, USA
3
Panama City Fish and Wildlife Conservation Office, U.S.
Fish and Wildlife Service, Panama City, FL 32405, USA
4
U.S. Fish and Wildlife Service, San Marcos Aquatic
Resources Center, San Marcos, TX 78666, USA
123
Conserv Genet
DOI 10.1007/s10592-016-0869-7
Author's personal copy

reintroduction event and low-head dams that may impede
gene flow in both rivers.
One of the main threats to fountain darters is drought
(natural and anthropogenic) and essential habitat for the
fountain darter is strongly influenced by the amount of
spring water emerging from the Edwards Aquifer at the
headwaters of each river. The discharge is determined by
the amount of precipitation over aquifer recharge areas and
the amount of water extracted from the aquifer for human
use. Droughts in central Texas generally occur at least once
a decade and during a severe drought in the 1950s, Comal
Springs ceased flowing for six months (Schenck and
Whiteside 1976; Brune 2002). The fountain darter was
likely extirpated from the Comal River during this drought
(USFWS 1996) and physical separation prevented natural
recolonization from the San Marcos population. In
1974–1976 457 darters from the San Marcos River were
transplanted to the headsprings area of the Comal River
(Schenck and Whiteside 1976; USFWS 1996). The rein-
troduction was successful in the sense that fountain darters
are now found throughout the Comal River and the popu-
lation size is estimated at over 150,000 fish (Linam et al.
1993; Ed Oborny 2011 pers. comm.). However, to inform
restoration planning a genetic assessment of the reintro-
duction is needed to determine if the number of fountain
darters used was sufficient to retain genetic diversity rela-
tive to the source (San Marcos) population and to evaluate
the extent to which the two populations have diverged from
one another.
Habitat fragmentation may also threaten fountain dar-
ters. Dams and other flow-control structures can have
dramatic effects on lotic habitat by altering water chemistry
and flow (Baxter 1977), river geomorphology (Ligon et al.
1995), fish and macroinvertebrate communities (Lessard
and Hayes 2003; Santucci et al. 2005; Tiemann et al. 2004)
and can cause a reduction of gene flow among populations,
leading to reduced genetic diversity and increased genetic
differentiation (Wofford et al. 2005; McCraney et al.
2010). Low-head dams (small drop structures in rivers,
streams or channels constructed to impound water, meter
discharge, or maintain stream slope; Kern et al. 2015) and
weirs have been shown to influence gene flow and genetic
diversity in some darter species (e.g., Haponski et al. 2007;
Beneteau et al. 2009; Sterling et al. 2012). Low head dams
(hereafter dams) in the Comal and San Marcos rivers
impound water for recreational use and are a potential
source of habitat fragmentation for fountain darters. Five
dams in the Comal River and three dams in the San Marcos
River may partially restrict gene flow within each river,
resulting in genetic differentiation.
Here we use 23 microsatellite loci to describe and
evaluate the genetic diversity, population structure and
effective population size of fountain darters to inform
restoration planning and provide a baseline for genetic
monitoring using indices such as heterozygosity, allele
number, Fst and effective population size (Schwartz et al.
2007). We evaluate the genetic outcome of the reintro-
duction and the effect of dams on genetic diversity within
and between the Comal and San Marcos rivers by
addressing three questions: (1) have fountain darters in the
Comal River diverged from fountain darters in the San
Marcos River? (2) was the 457 fish reintroduction sufficient
to maintain genetic diversity relative to the source popu-
lation? (3) do dams impede gene flow? The results are
assessed with regard to on-going and future conservation
efforts including the selection and size of a refuge popu-
lation(s) for ex situ captive propagation.
Methods
Sample collection
Fountain darters were collected from seven locations on the
Comal River (n =147) and nine locations on the San
Marcos River (n =180, Table 1; Fig. 1). The sample
locations represent four sections in each river that are
separated by dams that could inhibit gene flow (Fig. 1).
Fish were anesthetized with tricane methane-sulfonate
(MS-222; Finquel, Argent Chemical Laboratories, Inc.,
Redmond, Washington) and preserved individually in vials
of 95 % ethanol. Fin tissue samples were collected for
genetic analysis, preserved in 95 % ethanol, and sent to the
Conservation Genetics Laboratory, Alaska Region, U.S.
Fish and Wildlife Service for analyses.
Laboratory analyses
Twenty-three microsatellite loci were used to estimate
genetic variation in the 327 fountain darter samples (Online
Resource 1, Online Resource 2). Total genomic DNA was
extracted from fin tissue (*25 mg) using proteinase K
with the DNeasy
TM
DNA isolation kit (Qiagen Inc.
Valencia, CA), quantified with fluorometry and diluted to a
standard concentration. An MJResearch DNA Engine
Ò
thermal cycler was used to perform polymerase chain
reactions (PCR) in 10 ll volumes; general conditions were:
2.5 mM MgCl
2
, 1X PCR buffer (20 mM Tris–HCl pH 8.0,
50 mM KCl), 200 lM of each dNTP, 0.40 lM fluores-
cently labeled forward primer, 0.40 lM unlabeled reverse
primer, 0.008 units Taq polymerase, and 1 ll of DNA
(30 ng/ll). Standard thermal cycling conditions were: ini-
tial denaturation cycle of 92 °C for 2 min, followed by
92 °C for 15 s, 52–60 °C for 15 s (locus-specific sequences
and annealing temperature are given in Online Resource 1),
72 °C for 30 s, (30 cycles) with a final single cycle of
Conserv Genet
123
Author's personal copy

72 °C for 10 min. One-half ll of PCR product was elec-
trophoresed and visualized with the Applied Biosystems
3730 Genetic Analyzer utilizing a polymer denaturing
capillary system. Microsatellite allele sizes were estimated
and scored by the computer program GeneMapper
Ò
ver-
sion 4.0. Applied Biosystems GeneScan
TM
-600 LIZ
Ò
size
standards, 20–600 bases, were loaded in all lanes as an
internal lane standard. Two independent researchers veri-
fied all scores manually, with discrepancies being resolved
by replicating the analysis for the samples in question and
repeating the double scoring process until scores matched.
Data for samples from at least one column (n =8) in each
96-well sample plate was automatically replicated to con-
firm that proper plate orientation was maintained
throughout genotyping efforts for the project.
Genetic diversity and divergence
We examined population structure to assess if fountain
darters in the Comal River have diverged from fountain
darters in the San Marcos River and if dams in both rivers
impede gene flow. We used complementary individual- and
population-level analyses. Individual-level analysis was
done using the software STRUCTURE v. 2.3.1 (Pritchard
et al. 2000). STRUCTURE is a Bayesian, model-based
algorithm that clusters genetically similar individuals and
estimates the most likely clustering scenario, assuming
Hardy–Weinberg and linkage equilibrium. We tested
K=2 to 16 clusters assuming admixture and correlated
allele frequencies (between the K clusters) and using a
burn-in of 20,000 replications followed by 50,000 Markov
Table 1 A summary of the
samples, including location,
date, collection designation, and
sample size (n) used for genetic
analysis of fountain darters (E.
fonticola) from the Comal (CR)
and San Marcos (SMR) rivers,
Guadalupe River Basin, Texas,
USA
River Location Label n Date Collection
Comal Houston St., Landa Lake CR1a 15 02/25/10 CR1
Comal Liberty Ave., Landa Lake CR1b 15 02/25/10 CR1
Comal Landa Lake, Landa Park CR1c 9 03/09/10 CR1
21 04/14/10
Comal Elizabeth Ave., Old Channel CR2 1 02/25/10 CR2
13 02/26/10
5 04/16/10
6 04/21/10
Comal Hinman Island Park, New Channel CR3a 9 03/31/10 CR3
12 04/09/10
9 04/14/10
Comal Above Hinman Weir, Old Channel CR3b 2 04/14/10 CR3
Comal Garden St. CR4 10 03/31/10 CR4
9 04/14/10
11 04/16/10
San Marcos Spring Lake, near hotel SMR1a 15 11/23/09 SMR1
15 02/05/10
San Marcos Spring Lake, near outflow dam SMR1b 15 11/23/09 SMR1
15 02/05/10
San Marcos Sewell Park SMR2a 15 03/19/10 SMR2
14 11/23/09
San Marcos City Park SMR2b 15 03/17/10 SMR2
15 04/09/10
San Marcos Rio Vista Park SMR2c 14 03/17/10 SMR2
San Marcos Cheatum St. SMR3a 15 03/17/10 SMR3
San Marcos I-35 SMR3b 15 03/19/10 SMR3
San Marcos Cypress Tree, Lower SMR4a 4 10/28/09 SMR4
San Marcos Todd Island SMR4b 3 10/28/09 SMR4
4 04/09/10
6 04/23/10
The location labels denote the river (CR, SMR), the river section (1, 2, 3, 4), and the location within the
river section (a, b, c)
Conserv Genet
123
Author's personal copy

chain Monte Carlo replications. Ten replications were
performed to confirm the consistency of the log-likelihood
probabilities and to estimate the variance. The most likely
cluster scenario was the one that produced the largest
penalized likelihood value (the mean of the likelihood
values from each replicate minus half the variance,
Pritchard et al. 2000) and the results were visualized using
the software Distruct v. 1.1 (Rosenberg 2004).
For the population-level analyses, we grouped the
samples into collections by river section (see collections,
Table 1) to both evaluate the dams as putative barriers to
gene flow and evaluate genetic divergence between the two
rivers. The analyses included pairwise tests of genetic
differentiation and tests of hierarchical structure. The
software GenePop v. 4.1.4 (Rousset 2008) was used to
conduct exact G-tests of allele frequency homogeneity to
Fig. 1 A map showing the sample locations for fountain darters (E.
fonticola) from the aComal (CR) and bSan Marcos (SMR) rivers,
Guadalupe River Basin, Texas, USA. The 16 sites (red circles) are
coded as in Table 1where the upper case letters (CR, SMR) denote
the river, the numbers (1, 2, 3, 4) denote the river section, and lower
case letters (a, b, c) denote the location within the river section. The
dams are coded (A =Comal River, B =San Marcos River) and
represented by straight black lines perpendicular to river flow
Conserv Genet
123
Author's personal copy

test for genetic differentiation between all pairs of collec-
tions within and between each river. Significant G-test
results for adjacent collections would be consistent with
restricted gene flow by the dams. The pvalue for each test
was estimated using a Markov chain algorithm with the
following parameter values: dememorisation =10,000,
batch length =500, batch number =5000. The software
FSTAT v. 2.9.3 (Goudet 2001) was used to calculate
pairwise estimates of Fst (Wright 1943), according to Weir
and Cockerham (1984), and quantify the relative level of
genetic differentiation between all pairs of collections.
Analysis of Molecular Variation (AMOVA) was used to
compute hierarchical F-statistics and quantify the relative
level of genetic variation among all collections (Fst),
among collections between rivers (Fct) and among col-
lections within rivers (Fsc) and to test for statistical sig-
nificance of each value. We performed AMOVA using the
software ARLEQUIN v. 3.5 (Excoffier et al. 2005).
Estimates of allele frequency, allelic richness (Ar) and
observed (Ho) and expected heterozygosity (He) were
computed for each locus and collection using FSTAT.
Randomization tests were used to test for conformity to
Hardy–Weinberg expectations (HWE) for each locus and
collection combination and to test for composite genotypic
disequilibrium among locus pairs over all collections.
These tests were performed using FSTAT and the threshold
for statistical significance (a=0.05) was corrected (a/
k) using the sequential Bonferroni method (Rice 1989) for
ksimultaneous tests where the value of kdecreased
sequentially by removing significant tests until no tests
were judged significant. Two initial values of kwere used
for the HWE test to evaluate each collection over all loci
(k=23) and each locus over all collections (k=8).
Effective population size and genetic evaluation
of the Comal River reintroduction
The analysis of effective population size (Ne) and genetic
evaluation of the Comal River reintroduction were each
performed after pooling collections for each river because
we found little evidence of population structure within
rivers. The contemporary N
e
of fountain darters in the
Comal and San Marcos rivers was estimated using the
linkage disequilibrium method implemented in the soft-
ware N
E
ESTIMATOR 2.01 (Do et al. 2014). The estimates
were derived with P
crit
=0.02 (the minimum frequency for
alleles to be included in the analysis) and confidence
intervals were generated using the jackknife method.
To evaluate if the reintroduction was sufficient to
maintain genetic diversity relative to the source population
we tested if genetic diversity in fountain darters from the
Comal River was lower than in fountain darters from the
San Marcos River. We computed estimates of He, Ar and
private allele richness (pAr) for each locus in each popu-
lation using the software HP-RARE v. 1.0 (Kalinowski
2005). A Wilcoxon paired-sample test implemented in the
software R v. 3.2.2 (http://www.r-project.org/) was used to
test the null hypothesis that, for each statistic, estimates
across loci from the Comal population were greater than or
equal to estimates from the San Marcos population.
The genetically effective number of individuals re-intro-
duced to the Comal River were estimated from the allele
frequency data using a coalescent-based maximum likelihood
approach in the software COALIT and NFCONE (Anderson
and Slatkin 2007). This method requires estimates of three
parameters: the intrinsic rate of population growth (R), the
number of generations since introduction (T), and the effec-
tive carrying capacity (Nk). We used multiple estimates for
each parameter to reflect a reasonable range of values based
on knowledge of fountain darter biology and habitat. For
example, we examined estimates of 35 and 70 for T because
fountain darters may reproduce one to two times per year
(USFWS 1996) and the reintroduction occurred 35 years
before the samples were collected. We examined estimates of
0.5, 1.0, 2.0 and 3.0 for R because fountain darters rapidly
expanded their range following the reintroduction at Landa
Lake and are now found throughout the Comal River
(Fig. 1a). We examined estimates of 5000, 50,000 and
100,000 for Nk because the effective carrying capacity likely
varies as a result of drought. Estimates of the effective number
of founders (Nf) were derived for all 24 combinations of R, T,
and Nk. The number of simulated coalescent trees computed
by COALIT was 100,000 and the number of sampling
replicates used by NFCONE was 10,000.
Finally, we evaluated the data for evidence of a genetic
bottleneck in the Comal River. The software Bottleneck v.
1.2.02 (Cornuet and Luikart 1996) was used to test if the
heterozygosity computed from observed allele frequencies
was greater (heterozygote excess) than the equilibrium
heterozygosity computed from allele frequencies expected at
drift-mutation equilibrium. This excess of heterozygosity may
be observed in a population following a bottleneck because
allelic diversity is reduced faster than heterozygosity during a
bottleneck. The analysis was run assuming a two-phase
mutation model with 30 % multi-step mutations and a variance
of 30 for the number of repeat units per multi-step mutation.
The simulation included 5000 iterations and the results were
evaluated using the sign test and standardized differences test.
Results
Genetic diversity and divergence
The analysis of individual genotypes using STRUCTURE
indicated that the most likely solution was K =2 clusters.
Conserv Genet
123
Author's personal copy

These two clusters aggregated individuals by river (Fig. 2).
No evidence of genetic divergence was found using
STRUCTURE when the Comal and San Marcos river
samples were analyzed separately (the hierarchical method,
e.g. Va
¨ha
¨et al. 2007) to assess if fine-scale population
structure existed within each river. Nineteen of 28 paired
collections showed statistically significant differences in
allele frequencies when the G-test p-values were adjusted
for multiple tests (Table 2). Sixteen of the significant
results occurred when the paired collections were from
different rivers. Three of the Comal River collections
(CR1, CR2, CR4) showed significantly different allele
frequencies from one another but not from the CR3 col-
lection. Collections CR1 and CR2 were adjacent and sep-
arated by dam A1 (Fig. 1a). The San Marcos River
collections were not significantly different from one
another. The estimates of genetic variation among collec-
tions as measured by pairwise Fst ranged from -0.0008
(SMR3 9SMR4) to 0.0209 (CR1 9SMR3) and were
generally largest when the collections were from different
rivers (Table 2). The AMOVA estimate of Fst over all
collections was 0.0142 (Table 3). The estimates of genetic
variation between the two rivers (Fct) and among collec-
tions within rivers (Fsc) were 0.0102 and 0.0040,
respectively. The estimates of Fst, Fct and Fsc were all
significantly greater than zero.
The estimates of genetic diversity within collections as
measured by average heterozygosity (He) and allelic rich-
ness (Ar) were lowest in the Comal River collection CR4 at
0.624 and 6.99 and highest in the San Marcos River col-
lection SMR4 at 0.649 and 8.61 (Table 4). Tests of Hardy–
Weinberg equilibrium initially revealed a deficit of
heterozygote genotypes (p\0.05) at 13 of the 184 paired
locus x collection tests (Table 4). Six of these test were
judged statistically significant after correction for multiple
CR1 CR2 CR3 CR4 SMR2SMR1 SMR3 SMR4
Comal River San Marcos River
Fig. 2 A bar chart from the STRUCTURE analysis showing the most
likely cluster scenario (K =2) for fountain darters (E. fonticola) from
the Comal and San Marcos rivers, Guadalupe River Basin, Texas,
USA. Each individual is represented as a vertical line and the color
indicates the proportion of the individual genotype from the Comal
River cluster (green) and San Marcos River cluster (red). The black
vertical lines separate the sample locations (top labels) and the
horizontal brackets denote the collections as in Table 1
Table 2 A summary of the
pairwise estimates of Fst (below
diagonal) and the p-values from
the G-test of allele frequency
homogeneity (above diagonal)
for fountain darter (E. fonticola)
sample collections (Table 1)
from the Comal (CR) and San
Marcos (SMR) rivers,
Guadalupe River Basin, Texas,
USA
Aggr Comal River San Marcos River
CR1 CR2 CR3 CR4 SMR1 SMR2 SMR3 SMR4
CR1 NA 0.0007 0.1370 £0.0001 £0.0001 £0.0001 £0.0001 £0.0001
CR2 0.0067 NA 0.3718 0.0013 £0.0001 £0.0001 £0.0001 £0.0001
CR3 0.0029 0.0028 NA 0.0558 £0.0001 £0.0001 £0.0001 £0.0001
CR4 0.0129 0.0082 0.0068 NA £0.0001 £0.0001 £0.0001 £0.0001
SMR1 0.0114 0.0113 0.0099 0.0142 NA 0.2331 0.0062 0.4161
SMR2 0.0168 0.0145 0.0126 0.0133 0.0004 NA 0.1080 0.2361
SMR3 0.0209 0.0155 0.0119 0.0196 0.0031 0.0017 NA 0.2435
SMR4 0.0141 0.0139 0.0099 0.0134 0.0007 -0.0002 -0.0008 NA
A bold and underlined pvalue indicates a statistically significant difference in allele frequencies using the
sequential Bonferroni method (28 simultaneous tests, initial alpha =0.002)
Table 3 A summary of the AMOVA results for fountain darters (E.
fonticola) from the Comal and San Marcos rivers, Guadalupe River
Basin, Texas, USA
Among all collections Fst 0.0142*
Among collections within rivers Fsc 0.0040*
Among rivers Fct 0.0102*
The analysis was performed using the collections described in
Table 1and the estimates indicate the level of population structure
among all collections (Fst), among collections within the two rivers
(Fsc), and between the two rivers (Fct). An asterisk indicates the
value is significantly larger than zero (p\0.05)
Conserv Genet
123
Author's personal copy

tests, however these results were not isolated to a single
locus or collection.
Effective population size and genetic evaluation
of the Comal River reintroduction
The estimates of contemporary N
e
with 95 % confidence
intervals for the Comal and San Marcos populations were
899 (461—7668) and 9234 (1908—inf). The Wilcoxon
paired-sample test showed that the estimates of He, Ar and
pAr were all significantly larger across loci (p\0.05) in
the San Marcos population compared to the Comal popu-
lation (Table 5). Seventy-five private alleles were observed
in the San Marcos population and nine private alleles were
observed in the Comal population (Online Resource 3).
The mean frequency of private alleles was 0.010
(SD =0.006) for the San Marcos population and 0.007
(SD =0.005) for the Comal population.
The maximum likelihood estimates of the effective
number of founders (Nf) ranged from 37.21 to 77.42 for the
24 scenarios using combinations of intrinsic rate of popu-
lation growth (R), the number of generations since intro-
duction (T), and the effective carrying capacity (Nk,
Fig. 3). One scenario (R =0.5, T =70, Nk =5000)
resulted in a negative estimate of Nf and was not included
in the evaluation. The mean estimate over all scenarios
(49.35) suggests that the effective number of founders for
the Comal River was approximately 10.8 % of the number
of fish (457) transferred from the San Marcos River. The
Bottleneck test results showed 17 of 23 loci with an excess
of heterozygosity (relative to the expected heterozygosity
at drift-mutation equilibrium). An excess of heterozygosity
may be observed in a population following a bottleneck
because allelic diversity is lost much faster than
heterozygosity. The results of the sign test (p=0.062) and
standardized differences test (p=0.085) provided weak
evidence for a bottleneck.
Discussion
Genetic diversity and divergence
Two results stand out from our evaluation of genetic
diversity and divergence in the endangered fountain darter.
First, the individual-level Bayesian clustering analysis of
genotypes supported two populations represented by the
Comal and San Marcos rivers. The two populations are
separated by over 200 river kilometers along the mainstem
Guadalupe River and lower San Marcos River (Fig. 1).
This distance, the lack of suitable habitat between the
populations, and the high site fidelity in the species, is
thought to prevent fountain darter migration between the
two rivers. For example, Dammeyer et al. (2013) found that
fountain darters moved on average 10 ±17 m throughout
the year during a stable hydrologic regime, and movement
tended to be upstream and driven by habitat variables such
as low growing aquatic vegetation and stream flow. Many
darter species have limited dispersal ability with specific
habitat requirements that result in reproductively isolated
populations exhibiting substantial differentiation on a small
geographic scale (Haponski et al. 2009; Robinson et al.
2013; Fitzpatrick et al. 2014). The threatened Okaloosa
darter (Etheostoma okaloosae), for example, is restricted to
a 457 km
2
watershed in Florida and AMOVA results from
microsatellite loci revealed a high level of differentiation
(U
PR
=0.334) among collections from different drainages
(Austin et al. 2011). Similarly, the population-level
AMOVA results from this study indicated that genetic
divergence in fountain darters is largely due to differences
among collections from the two rivers. It is worth noting
that the level of differentiation between the Comal and San
Marcos rivers (Fct =0.0102) is, by comparison, low but
significant and shows that the reintroduced Comal popu-
lation has diverged at neutral loci from the San Marcos
source population.
Our population-level evaluation also revealed very little
evidence of genetic divergence among collections within
the two rivers. River barriers including dams and weirs
have been shown to influence dispersal of some darter
species (Haponski et al. 2007; Beneteau et al. 2009).
However, we only found evidence of genetic differentia-
tion at one dam (A1 in the Comal River) between two
adjacent fountain darter collections (CR1 9CR2). Indeed,
downstream movement is clearly evident in the Comal
River where fountain darters were reintroduced in the
Landa Lake area near the headwaters but now occupy the
entire length (5 km) of the river (Fig. 1). Collectively, our
results regarding the influence of dams on fountain darter
dispersal are equivocal and do not provide strong evidence
that movement across these barriers is inhibited. This
outcome is surprising but may indicate that periodic
opportunities arise, such as occasional flooding events, for
movement either across or around these dams. Another
possible explanation is that number of fountain darters
above and below the dams is large enough to minimize the
rate of genetic drift and thus limit the power to detect
genetic differentiation (Lowe and Allendorf 2010).
Effective population size and genetic evaluation
of the Comal River reintroduction
Two results stand out regarding the contemporary N
e
estimates. First, the estimates for both populations, while
larger than those reported for some other darter species
(e.g., Fluker et al. 2010; Fitzpatrick et al. 2014), are much
Conserv Genet
123
Author's personal copy

Table 4 A summary of genetic diversity estimates for 23 microsatellite loci in eight fountain darter (E. fonticola) collections from the Comal
(CR) and San Marcos (SMR) rivers, Guadalupe River Basin, Texas, USA
Locus Stat Comal River San Marcos River
CR1 (60) CR2 (25) CR3 (32) CR4 (30) SMR1 (60) SMR2 (73) SMR3 (30) SMR4 (17)
Eche001 He 0.546 0.381 0.607 0.346 0.424 0.313 0.505 0.524
Ho 0.533 0.400 0.531 0.333 0.433 0.288 0.500 0.471
Ar 3.26 2.68 3.78 3.63 3.61 3.49 3.97 4.00
Eche002 He 0.307 0.256 0.177 0.129 0.187 0.143 0.128 0.404
Ho0.233 0.240 0.188 0.067 0.167 0.123 0.133 0.294
Ar 3.18 2.96 3.02 2.49 2.73 2.64 2.49 4.00
EosC112 He 0.000 0.000 0.031 0.000 0.017 0.027 0.033 0.000
Ho 0.000 0.000 0.031 0.000 0.017 0.027 0.033 0.000
Ar 1.00 1.00 1.53 1.00 1.28 1.47 1.57 1.00
EosC2 He 0.342 0.528 0.478 0.582 0.405 0.523 0.452 0.557
Ho 0.350 0.600 0.469 0.533 0.400 0.479 0.467 0.588
Ar 2.64 3.00 2.99 3.00 2.96 2.95 2.00 3.00
EosC3 He 0.325 0.431 0.441 0.357 0.382 0.407 0.506 0.467
Ho 0.317 0.440 0.438 0.367 0.400 0.425 0.600 0.588
Ar 2.82 2.90 2.78 2.82 2.28 2.66 2.57 2.00
EosC6 He 0.835 0.762 0.829 0.783 0.887 0.881 0.851 0.875
Ho 0.817 0.720 0.875 0.667 0.917 0.890 0.833 0.882
Ar 7.05 7.26 6.96 7.02 10.22 9.90 9.33 9.00
Esc132b He 0.937 0.928 0.957 0.940 0.956 0.957 0.959 0.949
Ho 0.950 1.000 0.969 0.933 0.967 0.932 0.967 0.941
Ar 16.02 16.70 18.79 16.59 18.35 19.15 19.81 18.00
Esc26b He 0.872 0.885 0.903 0.881 0.913 0.925 0.925 0.932
Ho 0.883 0.840 0.938 0.933 0.817 0.890 0.833 0.882
Ar 10.03 11.81 11.67 9.33 13.13 13.27 13.82 14.00
Esc57 He 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.059
Ho 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.059
Ar 1.28 1.00 1.00 1.00 1.00 1.00 1.00 2.00
EfoA12 He 0.466 0.548 0.507 0.445 0.415 0.372 0.474 0.412
Ho 0.450 0.440 0.438 0.400 0.367 0.411 0.533 0.412
Ar 4.37 5.47 4.68 3.94 4.93 4.80 5.43 4.94
EfoD4 He 0.897 0.892 0.910 0.918 0.915 0.912 0.914 0.930
Ho 0.948 1.000 0.969 0.967 0.898 0.904 0.900 0.941
Ar 11.43 11.68 12.13 12.50 13.19 12.19 13.07 13.70
EfoE3 He 0.917 0.930 0.923 0.934 0.930 0.931 0.928 0.947
Ho 0.850 0.800 0.903 0.867 0.900 0.918 0.933 0.824
Ar 13.04 13.66 13.46 13.32 14.68 14.08 14.04 15.58
EfoE12 He 0.711 0.683 0.737 0.751 0.748 0.685 0.645 0.735
Ho 0.700 0.640 0.688 0.767 0.733 0.671 0.567 0.706
Ar 5.64 5.59 5.74 4.98 5.22 5.30 5.95 5.88
EfoE109 He 0.096 0.273 0.119 0.259 0.224 0.249 0.263 0.231
Ho 0.100 0.320 0.125 0.300 0.250 0.282 0.300 0.250
Ar 1.85 2.00 1.94 2.00 2.45 2.53 2.53 3.00
EfoE116 He 0.269 0.255 0.228 0.216 0.228 0.287 0.343 0.268
Ho 0.283 0.200 0.250 0.200 0.250 0.329 0.367 0.177
Ar 3.12 2.87 2.74 2.77 2.94 3.02 2.96 2.94
EfoE125 He 0.806 0.838 0.839 0.793 0.855 0.852 0.890 0.869
Conserv Genet
123
Author's personal copy

smaller than the estimates of census size (N
c
). Current
census size estimates suggest each river supports at least
150,000 fish and thus the N
e
/N
c
ratios are 0.06 for the San
Marcos River (N
e
=9,234) and 0.006 for the Comal River
(N
e
=899) populations. While these ratios are below the
average (0.11) reported by Frankham (1995) from a survey
of 102 species, they are not unusual (e.g., Hauser et al.
2002). Frankham (1995) found that fluctuations in census
population size is the most important factor influencing N
e
across species and, given that the fountain darter habitat is
exposed to periodic drought, it is likely that this species
exhibits large shifts in N
c
as habitat is lost and restored.
The second notable result is that the effective size of the
San Marcos population is about 10.3 times larger than the
Comal population. The restored Comal population was
founded from a small effective number of fish from the San
Marcos population (see below). Such recent bottlenecks
can downwardly bias estimates of contemporary N
e
using
the linkage disequilibrium method (Waples 2005). Other
factors could also influence the difference in N
e
between
the two populations including differences in carrying
capacity of the two rivers and differences in the influence
of drought on available habitat. Further research should
examine these and other factors that may influence N
e
in
fountain darters.
Table 4 continued
Locus Stat Comal River San Marcos River
CR1 (60) CR2 (25) CR3 (32) CR4 (30) SMR1 (60) SMR2 (73) SMR3 (30) SMR4 (17)
Ho 0.850 0.760 1.000 0.867 0.900 0.822 0.833 0.765
Ar 7.17 6.63 8.25 6.83 9.00 8.91 10.68 9.82
EfoE134 He 0.902 0.892 0.866 0.910 0.946 0.938 0.953 0.952
Ho0.717 0.920 0.781 0.900 0.850 0.889 0.867 0.813
Ar 12.25 11.87 11.34 12.85 16.13 15.50 16.17 14.00
EfoE150 He 0.893 0.920 0.903 0.876 0.916 0.922 0.929 0.938
Ho 0.917 0.880 0.936 0.900 0.783 0.904 0.900 0.941
Ar 11.81 12.99 12.82 9.74 15.04 15.07 14.76 16.46
EfoE159 He 0.793 0.779 0.807 0.812 0.787 0.800 0.805 0.827
Ho0.617 0.640 0.813 0.733 0.583 0.667 0.767 0.813
Ar 4.99 5.75 5.50 5.53 5.46 6.27 5.94 7.00
EfoE185 He 0.924 0.894 0.912 0.918 0.940 0.939 0.916 0.947
Ho 0.933 1.000 0.903 0.833 0.950 0.959 0.967 0.882
Ar 13.92 12.56 12.95 11.96 16.16 15.17 14.89 16.47
EfoE189 He 0.770 0.768 0.786 0.726 0.821 0.809 0.809 0.792
Ho 0.800 0.760 0.656 0.733 0.783 0.819 0.767 0.765
Ar 7.09 7.07 6.88 7.20 9.08 8.59 7.95 8.82
EfoE192 He 0.902 0.864 0.874 0.911 0.939 0.931 0.928 0.921
Ho 0.883 0.840 0.719 0.800 0.883 0.904 0.833 1.000
Ar 11.77 10.57 11.49 12.07 14.65 14.25 13.70 13.65
EfoE193 He 0.901 0.900 0.878 0.856 0.894 0.890 0.887 0.884
Ho 0.917 0.880 0.906 0.800 0.950 0.863 0.833 0.941
Ar 10.10 11.20 10.01 9.01 10.94 10.95 11.14 9.88
Avg He 0.627 0.635 0.640 0.624 0.640 0.639 0.654 0.670
Ho 0.612 0.623 0.631 0.604 0.617 0.626 0.641 0.649
Ar 7.17 7.31 7.45 6.99 8.45 8.35 8.46 8.61
The estimates include expected heterozygosity (He), observed heterozygosity (Ho), and allelic richness (Ar). Estimates of Ho lower than
expected based on Hardy–Weinberg expectations are underlined and in bold (p\0.05). The sample size of each collection is in parentheses
Table 5 A summary of estimates of mean heterozygosity (He) mean
allelic richness (Ar) and mean private allele richness (pAr) across loci
for fountain darter (E. fonticola) from the Comal and San Marcos
rivers, Guadalupe River Basin, Texas, USA
Population HeAr pAr
San Marcos 0.645 13.47 3.05
Comal 0.634 10.90 0.48
pvalue 0.022 \0.001 \0.001
The pvalues are from the Wilcoxon paired-sample test of the null
hypothesis Comal CSan Marcos for each statistic across all loci
Conserv Genet
123
Author's personal copy

Populations established as the result of a reintroduction
often exhibit reduced genetic variability (Leberg 1990).
Our results comparing estimates of heterozygosity, allelic
richness and private alleles indicate that the reintroduced
Comal population has lost genetic diversity relative to the
San Marcos source population. Assuming the genetic
diversity in the 457 fountain darters reintroduced into the
Comal River was representative of the San Marcos popu-
lation, our evaluation suggests that much of the observed
loss of diversity and population divergence occurred soon
after the reintroduction (founder effect) although the pop-
ulations continue to diverge as a result of genetic drift.
Rapid divergence at neutral loci is consistent with a
founder event and has been documented in other small
freshwater fishes following colonization (e.g., Kinziger
et al. 2011). In the present study, the mean estimate of the
effective number of founders (Nf=49.35) was small
compared to the number of fish (N
c
=457) used for the
reintroduction. The Nf/N
c
ratio (0.108) was close to the
average (0.11) across taxon reported by Frankham (1995).
We used the equation:
Ht=H0¼11=2Ne
ðÞ½
t;
(Frankham et al. 2002) to estimate the duration of the
bottleneck in generations (t) assuming estimates of
heterozygosity at time t (H
t
) from the Comal population
(0.634), initial heterozygosity (H
0
) from the San Marcos
population (0.645) and N
e
from the effective founder
number (49.35). The result suggests that the observed loss
of heterozygosity in the Comal population relative to the
San Marcos population could have occurred in as few as
t=1.8 generations. In addition, the signature of a founder
event (heterozygosity excess) was evident in 17 of the 23
loci although the bottleneck tests provided weak statistical
support (0.05 \P\0.10). These tests are sensitive to the
duration of the bottleneck and it is likely that, given the
current N
c
, the Comal River reintroduction was followed
by a period of rapid population growth that has dampened
the bottleneck signal (Hundertmark and Van Daele 2010;
Peery et al. 2012).
The Comal population exhibited nine alleles not found
in the San Marcos population (private alleles). It is possible
that these private alleles originated from fountain darters
that survived the drought in the early 1950s and that
remained undetected despite the extensive survey effort
prior to the reintroduction in 1974–76 (Schenck and
Whiteside 1976; USFWS 1996). However, we believe the
results suggest that this possibility is unlikely. For exam-
ple, the mean private allele richness per locus and mean
frequency of private alleles were both very low in the
Comal population (0.48, 0.007) compared to the San
Marcos population (3.05, 0.010). Remnant Comal River
fountain darters would likely have possessed a higher
number of private alleles than observed in the present
Comal River sample. In addition, the relatively low level of
genetic differentiation (Fst =0.0142) between the two
populations is consistent with recent divergence from
genetic drift following the reintroduction and rapid bot-
tleneck as opposed to a long-standing (pre-reintroduction)
separation that included a long-term ([20 generations)
Fig. 3 A scatter plot showing
the maximum likelihood
estimates of the effective
number of founders for fountain
darters (E. fonticola) in the
Comal River, Guadalupe River
Basin, Texas, USA. The
estimates were derived for 24
scenarios using three values for
the effective carrying capacity
(NK =5,000, 50,000, 100,000),
four values for the intrinsic rate
of population growth (R =0.5,
1.0, 2.0, 3.0), and two values for
generations since founding
(T =35, 70). Each symbol
represents one of the three
values of NK (91000) and one
of the two values of T
Conserv Genet
123
Author's personal copy

bottleneck in the Comal population following the drought
in the 1950s. We believe, given the small number of private
alleles in the Comal population sample, it is more likely
that these alleles were simply not sampled from the San
Marcos population or that they arose naturally through
mutation given the relatively high mutation rate of
microsatellites (Jarne and Lagoda 1996) and short gener-
ation time (*1 year) of fountain darters.
Conservation implications
The results from this study have implications with respect
to conservation of the federally endangered fountain darter.
First, the genetic data revealed two populations, the Comal
and San Marcos populations, with no evidence of popula-
tion structure within the San Marcos River and only weak
evidence of population structure within the Comal River.
Presently, fountain darters in both rivers are treated sepa-
rately for conservation planning including ex situ captive
propagation of two refuge populations at the San Marcos
Aquatic Resources Center (USFWS 1996). Our results
support this approach and captive propagation of the San
Marcos population but indicate the need to re-evaluate
captive propagation of the Comal population. That is, our
results are consistent with the Comal population being
descendent from the San Marcos population and thus it
may only be necessary to maintain the source population
(San Marcos) in captivity. It is important to note, however,
that our genetic results are from neutral loci and we make
no inference regarding local adaptation by reintroduced
fountain darters in the Comal River. Evidence for rapid
adaptation has been shown in other freshwater fish species.
Stockwell and Vinyard (2000) found divergence of genetic-
based life history characteristics in the western mosquito-
fish Gambusia affinis, after approximately 60 years. Fur-
ther study is needed to assess if the Comal and San Marcos
river fountain darters differ at adaptively important traits
and genes.
The rate at which genetic diversity is lost is related to N
e
not N
c
and thus N
e
is an important parameter to monitor in
populations of conservation concern (Schwartz et al. 2007).
The estimates of N
e
here do not suggest these populations
are at risk of an immediate and rapid loss of genetic
diversity (Frankham et al. 2002). However, the fact that the
estimates of N
e
are much smaller than N
c
for both popu-
lations warrants attention and highlights the importance of
including N
e
, as well as He and Ar, as part of a genetic
monitoring program for fountain darters in order to fully
evaluate the genetic consequences of reductions in census
size following events such as droughts. Similarly, the
estimates of the effective number of founders used in the
Comal River reintroduction (range 37.21–77.42,
mean =49.35) were well below the number of fish (457)
used for the reintroduction. These results suggest that
future reintroductions into either river will require a larger
sample (457) from the captive refuge populations in
order to avoid a founder event that results in a rapid loss of
genetic diversity relative to the source population. Finally,
while this study did not support a conclusion regarding the
influence of dams on fountain darter gene flow, the issue is
a critical one for conservation planning relative to water
use in both rivers (USFWS 1996). Dammeyer et al. (2013)
showed that fountain darters tend to move upstream and
this behavior may be disrupted by the dams. For this reason
we suggest that the influence of dams on fountain darter
movement continue to be explored using a combination of
genetic and ecological studies targeting fish adjacent to
dams.
Acknowledgments This project was partially funded by the Edwards
Aquifer Recovery Implementation Program. Additional thanks to
anonymous for valuable suggestions regarding the composition of this
manuscript. The findings and conclusions in this article are those of
the author(s) and do not necessarily represent the views of the U.S.
Fish and Wildlife Service.
References
Anderson EC, Slatkin M (2007) Estimation of the number of
individuals founding colonized populations. Evolution
61(4):972–983
Austin J, Jelks H, Tate B, Johnson A, Jordan F (2011) Population
genetic structure and conservation genetics of threatened
Okaloosa darters (Etheostoma okaloosae). Conserv Genet
12:981–989
Baxter RM (1977) Environmental effects of dams and impoundments.
Ann R Ecol Syst 8:255–283
Beneteau C, Mandrak N, Heath D (2009) The effects of river barriers
and range expansion of the population genetic structure and
stability in Greenside Darter (Etheostoma blennioides) popula-
tions. Conserv Genet 10:477–487
Brune GM (2002) Springs of Texas. Texas A&M University Press,
College Station
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
Dammeyer NT, Phillips CT, Bonner TH (2013) Site fidelity and
movement of Etheostoma fonticola with implications to endan-
gered species management. Trans Am Fish Soc 142:1049–1057
Do C, Waples RS, Peel D, Macbeth GM, Tillett BJ, Ovenden JR
(2014) NEESTIMATOR v2: re-implementation of software for
the estimation of contemporary effective population size (Ne)
from genetic data. Mol Ecol Res 14:209–214
[ESA] Endangered Species Act of 1973, Pub. L. No. 93- 205, 87 Stat.
884 (Dec. 28, 1973)
Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: An
integrated software package for population genetics data anal-
ysis. Evol Bioinform Online 1:47–50
Fitzpatrick SW, Crockett H, Funk WC (2014) Water availability
strongly impacts population genetic patterns of an imperiled
Great Plains endemic fish. Conserv Genet 15:771–788
Fluker BL, Kuhajda BR, Lang NJ, Harris PM (2010) Low genetic
diversity and small long-term population sizes in the spring
Conserv Genet
123
Author's personal copy

endemic watercress darter, Etheostoma nuchale. Conserv Genet
11:2267–2279
Frankham R (1995) Effective population size/adult population ratios
in wildlife: a review. Genet Res 66:95–107
Frankham R, Ballou JD, Briscoe DA (2002) Introduction to conser-
vation genetics. Cambridge University Press, Cambridge
Goudet J (2001) FSTAT, a program to estimate and test gene
diversities and fixation indices (version 2.9.3). http://www.unil.
ch/izea/softwares/fstat.html
Haponski AE, Marth TA, Stepien CA (2007) Genetic divergence
across a low-head dam: a preliminary analysis using logperch
and greenside darters. J Great Lakes Res 33(SI2):117–126
Haponski AE, Bollin TL, Jedlicka MA, Stepien CA (2009) Landscape
genetic patterns of the rainbow darter Etheostoma caeruleum:a
catchment analysis of mitochondrial DNA sequences and nuclear
microsatellites. J Fish Biol 75:2244–2268
Hauser L, Adcock GJ, Smith PJ, Bernal Ramı
´rez JH, Carvalho GR
(2002) Loss of microsatellite diversity and low effective
population size in an overexploited population of New Zealand
snapper (Pagrus auratus). Proc Natl Acad Sci USA
99:11742–11747
Hundertmark K, Van Daele L (2010) Founder effect and bottleneck
signatures in an introduced, insular population of elk. Conserv
Genet 11:139–147
Jarne P, Lagoda JL (1996) Microsatellites, from molecules to
populations and back. Trends Ecol Evol 11:424–429
Kalinowski ST (2005) HP-RARE 1.0: a computer program for
performing rarefaction on measures of allelic richness. Mol Ecol
Notes 5:187–189
Kern EW, Hotchkiss RH, Ames DP (2015) Introducing a low-head
dam fatality database and internet information portal. JAWRA
51(5):1453–1459
Kinziger AP, Nakamoto RJ, Anderson EC, Harvey BC (2011) Small
founding number and low genetic diversity in an introduced
species exhibiting limited invasion success (speckled dace,
Rhinichthys osculus). Ecol Evol 1:73–84
Leberg PL (1990) Influence of genetic variability on population
growth: implications for conservation. J Fish Biol 37(suppl
A):193–195
Lessard JL, Hayes DB (2003) Effects of elevated water temperature
on fish and macroinvertebrate communities below small dams.
River Res Appl 19:721–732
Ligon FK, Dietrich WE, Trush WJ (1995) Downstream ecological
effects of dams. Bioscience 45:183–192
Linam GW, Mayes KB, Saunders KS (1993) Habitat utilization and
population size estimate of fountain darters Etheostoma fonticola
in the Comal River, Texas. Texas J Sci 45:341–348
Lowe WH, Allendorf FW (2010) What can genetics tell us about
population connectivity? Mol Ecol 19:3038–3051
McCraney WT, Goldsmith G, Jacobs DK, Kinziger AP (2010)
Rampant drift in artificially fragmented populations of the
endangered tidewater goby (Eucyclogobius newberryi). Mol
Ecol 19:3315–3327
Peery MZ, Kirby R, Reid BN, Stoelting R, Doucet-Be
¨er E, Robinson
S, Va
´squez-Carrillo C, Pauli JN, Palsbøll PJ (2012) Reliability of
genetic bottleneck tests for detecting recent population declines.
Mol Ecol 21:3403–3418
Pritchard JK, Stephans M, Donnelly P (2000) Inference of population
structure using multilocus genotype data. Genetics 155:945–959
Rice WR (1989) Analyzing tables of statistical tests. Evol 43:223–225
Robinson J, Simmons J, Williams A, Moyer G (2013) Population
structure and genetic diversity in the endangered bluemask darter
(Etheostoma akatulo). Conserv Genet 14:79–92
Rosenberg NA (2004) Distruct: a program for the graphical display of
population structure. Mol Ecol Notes 4:137–138
Rousset F (2008) Genepop ‘007: a complete reimplementation of the
Genepop software for Windows and Linux. Mol Ecol Res
8:103–106
Santucci VJ, Gephard SR, Pescitelli SM (2005) Effects of multiple
low-head dams on fish, macroinvertebrates, habitat, and water
quality in the Fox River, Illinois. N Am J Fish Manag 25:975–992
Schenck JR, Whiteside BG (1976) Distribution, habitat preference,
and population size estimate of Etheostoma fonticola. Copeia
4:697–703
Schwartz MK, Luikart G, Waples RS (2007) Genetic monitoring as a
promising tool for conservation and management. Trends Ecol
Evol 22:25–33
Sterling K, Reed D, Noonan B, Warren M Jr (2012) Genetic effects of
habitat fragmentation and population isolation on Etheostoma
raneyi (Percidae). Conserv Genet 13:859–872
Stockwell CA, Vinyard GL (2000) Life history variation in recently
established populations of western mosquitofish (Gambusia
affinis). West N Am Nat 60:273–280
Tiemann JS, Gillette DP, Wildhaber ML, Edds DR (2004) Effects of
lowhead dams on riffle-dwelling fishes and macroinvertebrates
in a Midwestern river. Trans Am Fish Soc 133:705–717
US Fish and Wildlife Service (1996) San Marcos and Comal springs
and associated aquatic ecosystems (Revised) recovery plan. U.S.
Fish and Wildlife Service, Albuquerque, NM. http://ecos.fws.
gov/docs/recovery_plan/960214.pdf. Accessed 17 Dec 2015
Va
¨ha
¨J, Erkinaro J, Niemela E, Primmer CR (2007) Life-history and
habitat features influence the within-river genetic structure of
Atlantic salmon. Mol Ecol 16:2638–2654
Waples RS (2005) Genetic estimates of contemporary effective
population size: to what time periods do the estimates apply?
Mol Ecol 14:3335–3352
Weir BS, Cockerham CC (1984) Estimating F-statistics for the
analysis of population structure. Evol 38:1358–1370
Wofford JEB, Gresswell RE, Banks MA (2005) Influence of barriers
to movement on within-watershed genetic variation of coastal
cutthroat trout. Ecol Appl 15:628–637
Wright S (1943) Isolation by distance. Genetics 28:114–138
Conserv Genet
123
Author's personal copy